University of Southern California Dissertations and Theses

Genetic basis of diet-dependent responses across the lifespan in Caenorhabditis elegans

(USC Thesis Other)

Genetic basis of diet-dependent responses across the lifespan in Caenorhabditis elegans

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Copyright 2023 Nicole L. Stuhr Genetic basis of diet-dependent responses across the lifespan in Caenorhabditis elegans by Nicole L. Stuhr A Dissertation Presented to the FACULTY OF THE USC GRADUATE SCHOOL UNIVERSITY OF SOUTHERN CALIFORNIA In Partial Fulfillment of the Requirements for the Degree DOCTOR OF PHILOSOPHY (MOLECULAR BIOLOGY) August 2023 ii This work is dedicated to all the people who have supported me every step of the way. Thank you. iii Acknowledgements I would like to thank all the people that have been crucial during my scientific journey: Dr. Sean Curran, an amazing mentor and boss, who has constantly been supportive and encouraging during my PhD studies. Thank you for making the lab environment somewhere I wanted to be every day of the week. From all of the lab field trips catching Pokémon pre- pandemic, to playing exploding kittens via Zoom during lockdown, to the educational lab trips post-pandemic, it has always been an enjoyable experience working in your lab. I can’t thank you enough for the opportunity to work in your laboratory for the last five year. Your mentorship has made all of my accomplishments possible, and I will use your teaching and mentoring skills as an example for how I want to be when I am a PI. To my committee members Dr. Carolyn Phillips, Dr. John Tower, and Dr. Bérénice Benayoun, thank you for your guidance and invaluable suggestions during committee update meetings. To the former and current Curran lab members, thank you for creating an amazing workplace that I loved coming into on a daily basis. It was always a blast being in and around everyone. Thank you for the constant support and incredible feedback during lab meetings. You all contributed to an engaging and supportive work environment. Hans, who helped me get my bearings when I first joined the lab. Amy, An, and Sierra for all the cat shenanigans and cuddle sessions with the kitties. Beaut & Sarah, who have been the people I can always go to for support in and out of the lab. Chris, who has been with me through every step of the way. To my family, who have always supported my interest in science and continue to encourage me to follow my dreams. To my Molecular Biology cohort for continuously staying friends and always making time for dinners and hangouts to de-stress. Lastly, to my cats Mogwai and Mr. Moose for their stress-relieving furry companionship. iv Table of Contents Dedication………………………………………………………………………………………………….ii Acknowledgements ....................................................................................................................... iii List of Tables ............................................................................................................................... viii List of Figures ............................................................................................................................. viii Abstract ......................................................................................................................................... xi Chapter 1: Introduction ................................................................................................................. 1 Chapter 2: Bacterial diets differentially alter lifespan and healthspan trajectories in C. elegans ................................................................................................................................... 18 Introduction ............................................................................................................................. 20 Results .................................................................................................................................... 22 Discussion ............................................................................................................................... 34 Materials & Methods ............................................................................................................... 39 Tables ...................................................................................................................................... 48 Figues ..................................................................................................................................... 49 Supplemental Figures ............................................................................................................. 66 Supplemental Tables ............................................................................................................... 74 Acknowledgements ................................................................................................................. 75 Chapter 3: Avoidance behavior of C. elegans in the presence of Methylobacterium and its link to fatty acid biosynthesis pathways ...................................................................................... 76 Abstract ................................................................................................................................... 77 Results/Discussion .................................................................................................................. 78 Materials & Methods ............................................................................................................... 85 Tables ...................................................................................................................................... 87 Figures .................................................................................................................................... 88 Supplemental Figures ............................................................................................................. 95 Chapter 4: A dicer-related helicase opposes the age-related pathology from SKN-1 activation in ASI neurons ............................................................................................................ 99 Abstract ................................................................................................................................. 100 Introduction ........................................................................................................................... 101 Results .................................................................................................................................. 103 Discussion ............................................................................................................................. 112 Materials & Methods ............................................................................................................. 115 Figures .................................................................................................................................. 121 Supplemental Figures ........................................................................................................... 126 Acknowledgements ............................................................................................................... 131 v Chapter 5: Genetic variation in ALDH4A1 is associated with muscle health over the lifespan and across species ................................................................................................................... 132 Abstract ................................................................................................................................. 133 Introduction ........................................................................................................................... 134 Results/Discussion ................................................................................................................ 136 Materials & Methods ............................................................................................................. 144 Figures .................................................................................................................................. 151 Tables .................................................................................................................................... 156 Supplemental Figures ........................................................................................................... 159 Acknowledgements ............................................................................................................... 163 Chapter 6: Lipid Quantification in Caenorhabditis elegans by Nile Red and Oil Red O Staining ..................................................................................................................................... 164 Abstract ................................................................................................................................. 165 Background ........................................................................................................................... 166 Materials & Methods ............................................................................................................. 167 Procedure ............................................................................................................................. 170 Figures .................................................................................................................................. 178 Notes ..................................................................................................................................... 182 Recipes ................................................................................................................................. 183 Acknowledgements ............................................................................................................... 185 Chapter 7: Different methods of killing bacteria diets differentially influence Caenorhabditis elegans physiology ................................................................................................................... 186 Abstract ................................................................................................................................. 187 Description ............................................................................................................................ 188 Methods ................................................................................................................................ 190 Figures .................................................................................................................................. 193 Acknowledgements ............................................................................................................... 194 Chapter 8: Ether Lipid Biosynthesis Promotes Lifespan Extension and Enables Diverse Prolongevity Paradigms in Caenorhabditis elegans ................................................................. 195 Abstract ................................................................................................................................. 197 Introduction ........................................................................................................................... 198 Results .................................................................................................................................. 201 Discussion ............................................................................................................................. 212 Materials & Methods ............................................................................................................. 216 Figures .................................................................................................................................. 228 Supplemental Figures ........................................................................................................... 240 Acknowledgements ............................................................................................................... 253 vi Chapter 9: Riboflavin Depletion Promotes Longevity and Metabolic Hormesis in Caenorhabditis elegans ............................................................................................................ 254 Abstract ................................................................................................................................. 256 Introduction ........................................................................................................................... 257 Results .................................................................................................................................. 260 Discussion ............................................................................................................................. 270 Materials & Methods ............................................................................................................. 275 Figures .................................................................................................................................. 285 Supplemental Figures ........................................................................................................... 297 Acknowledgments ................................................................................................................. 306 References ............................................................................................................................... 307 vii List of Tables Chapter 1 Table 1. Physiological attributes in C. elegans that are influenced by bacterial diet.......................7 Table 2. Wildtype worm phenotypes in the presence of killed E. coli OP50.................................10 Table 3. Gene-diet pairs in C. elegans induce diet-dependent physiological changes................15 Chapter 2 Table 1. Bacterial diet exposure leads to changes in C. elegans physiology...............................48 Table S1. Gene Ontology Terms for RNAseq analysis in L4 C. elegans on bacterial diets..........74 Chapter 3 Table 1. List of C. elegans strains used in this study....................................................................87 Chapter 5 Table 1. Top SNPs associated with specific phenotypes............................................................156 Table 2. SNPs remaining after filtering for minor allele frequency and pruning based on Linkage disequilibrium...............................................................................................................157 Table 3. Replication across ethnic subsamples in the HRS.......................................................158 viii List of Figures Chapter 1 Figure 1. Dietary requirements of C. elegans.................................................................................2 Figure 2. E. coli strains used as diets for C. elegans in the laboratory...........................................4 Chapter 2 Figure 1. Characterization of bacterial diets.................................................................................49 Figure 2. Gene expression analysis of L4 C. elegans on each bacterial food source after Thirty generations........................................................................................................................50 Figure 3. Developmental timing of C. elegans is dependent upon bacterial diet..........................52 Figure 4. Fat content and distribution vary depending on the food C. elegans are exposed to....54 Figure 5. C. elegans raised on Red and Yellow have decreased reproductive output.................56 Figure 6. Lifespan of C. elegans on each bacterial diet................................................................58 Figure 7. C. elegans have a food-dependent decline in thrashing with age.................................60 Figure 8. C. elegans food choice..................................................................................................62 Figure 9. The introduction of the Red bacteria at the post-reproductive stage in C. elegans Extends lifespan.........................................................................................................................64 Figure S1. Bacterial growth conditions and metabolite concentrations.......................................66 Figure S2. Bacterial growth curve and images............................................................................67 Figure S3. Time in each developmental stage and worm area is altered based on bacterial diet raised on................................................................................................................................68 Figure S4. L4 pumping and Oil Red O staining...........................................................................70 Figure S5. Reproductive timing is altered in C. elegans raised on the different bacterial diets....71 Figure S6. Thrashing declines with age in a food-dependent manner.........................................72 Figure S7. Bacterial diet combinations compared to C. elegans raised on OP50........................73 Chapter 3 Figure 1. Behavior of C. elegans on the Red/Methylobacterium longevity-promoting diet...........88 Figure 2. AWA, AWB, and AWC neurons mediate avoidance responses to Methylobacterium...90 Figure 3. ODR-1 signaling is required for the avoidance response towards Methylobacterium...91 Figure 4. Avoidance response is due to a difference in metabolites found in Methylobacterium and not a volatile chemical produced by the bacteria...................................................................92 Figure 5. Disruption of the fatty acid biosynthesis pathway changes the response of worms to Methylobacterium.........................................................................................................94 Figure S1. Food choice in wildtype worms is independent from the diet that the worms are raised on......................................................................................................................................95 Figure S2. Functionality of chemosensory neurons is important for the Methylobacterium avoidance response.......................................................................................97 Figure S3. Chemosensory mutants show a variety of responses towards Methylobacterium, some move away while some stay...............................................................................................98 Chapter 4 Figure 1. Cell autonomous activity of SKN-1gf in ASI neurons...................................................121 Figure 2. Neddylation regulates nuclear SKN-1 stabilization in the intestine.............................122 Figure 3. SKN-1 activity in ASI neurons mediates peripheral stress responses.........................123 Figure 4. DRH-1 activation delays SKN-1gf-dependent healthspan decline..............................124 Figure 5. Intestinal DRH-1 activation reduced transcriptional load of activated SKN-1..............125 Figure S1. Gain-of-function mutations drive SKN-1 activation...................................................126 Figure S2. Ubiquitin-related pathways mediate SKN-1-dependent activities.............................127 ix Figure S3. Cell non-autonomous impact of SKN-1 activation in neurons...................................128 Figure S4. Identification and characterization of drh-1gf mutation..............................................129 Figure S5. drh-1gf effects are tied to transcriptional regulation...................................................130 Chapter 5 Figure 1. Mutation of alh-6 uniquely activates age-dependent and activation of gst-4p::gfp oxidative stress reporter in muscle...........................................................................151 Figure 2. ALDH4A1 variants associate with human age-related phenotypes for change in muscle function..........................................................................................................................153 Figure 3. Effects of ALDH4A1 variation on phenotypes representing association with change in aging-related function in a normative, population-based sample of older adults.........154 Figure 4. alh-6 mutations accelerate loss of muscle function.....................................................155 Figure S1.1. Novel alleles of alh-6 induce muscle specific activation of the gst-4p::gfp oxidative stress reporter.............................................................................................................159 Figure S1.2. Location of amino acid substitutions in alh-6 mutants...........................................160 Figure S3. Association between rs77608580 and ALDH4A1 gene expression levels in whole blood................................................................................................................................161 Figure S4. alh-6 mutations accelerate loss of muscle function..................................................162 Chapter 6 Figure 1. Representative images of Oil Red O...........................................................................178 Figure 2. Representative images of Nile Red Staining...............................................................179 Figure 3. Representative images of embryo Nile Red Staining.................................................180 Figure 4. Outlining and calculation of total lipid density using ImageJ.......................................181 Chapter 7 Figure 1. Differential effects of PFA-killed and UV/Antibiotic-killed E. coli on C. elegans physiology.....................................................................................................................193 Chapter 8 Figure 1. Genes responsible for ether lipid biosynthesis are necessary for biguanide-induced lifespan extension........................................................................................228 Figure 2. Phenformin treatment of C. elegans leads to increased abundance of multiple alkyl and alkenyl ether lipids.......................................................................................................230 Figure 3. Peroxisomal protein import, fatty acid elongases, and fatty acid desaturases are required for pro-longevity effects of biguanides..........................................................................232 Figure 4. Genes involved in ether lipid biosynthesis are required for lifespan extension in multiple longevity paradigms......................................................................................................234 Figure 5. FARD-1 overexpression is sufficient to extend lifespan by modulating ether lipid synthesis............................................................................................................................235 Figure 6. Phenformin modulates systemic lipid metabolism through an ether lipid-skn-1 signaling relay............................................................................................................................237 Figure 7. Schematic representation for the role of lipid ether biosynthetic machinery in multiple pro-longevity paradigms...............................................................................................239 Figure S1.1. Reduced function of genes responsible for ether lipid biosynthesis partially suppresses biguanide effects of growth and lifespan without affecting biguanide levels.........................................................................................................................240 Figure S1.2. The use of FuDR in lifespan analyses does not confound the observed epistasis between the ether lipid machinery and biguanide-mediated lifespan extension......................................................................................................................242 Figure S2.1. Biguanide treatment modulates abundance of fatty acids in C. elegans...............243 x Figure S2.2. FARD-1::RFP localized to intestinal lipid droplets and peroxisomes and is not positively regulated at the RNA or protein level by phenformin.........................................244 Figure S2.3. Disruption of bacterial growth and metabolism does not prevent biguanide-mediated induction of ether lipid synthesis................................................................246 Figure S2.4. Inactivation of ether lipid machinery disrupts biguanide-mediated lifespan extension independent of effects on bacterial growth or metabolism.........................................248 Figure S4.1. Ether lipid biosynthetic genes are not necessary for daf-2-dependent lifespan extension and fard-1 overexpression extends lifespan in a manner dependent upon ether lipid biosynthesis......................................................................................................249 Figure S5.1. Genetic induction of ferroptosis does not impact fard-1 overexpression nor biguanide-mediated lifespan extension................................................................................250 Figure S6.1. Biguanides do not activate gst-4 expression irrespective of bacterial diet.............251 Chapter 9 Figure 1. rft-1 RNAi depletes flavins and extends lifespan........................................................285 Figure 2. Riboflavin depletion promotes longevity by activating FOXO/daf-16..........................287 Figure 3. Riboflavin depletion alters cellular energetics.............................................................289 Figure 4. Activation of the mitochondrial unfolded protein response is required for riboflavin depletion to promote longevity....................................................................................291 Figure 5. Riboflavin depletion alters lipid metabolism................................................................293 Figure 6. Riboflavin depletion mimics features of dietary restriction..........................................295 Figure S1. rft-1 knockdown phenotypes.....................................................................................297 Figure S2. Activation of DAF-16 by riboflavin deficiency............................................................299 Figure S3. Metabolite two-photon imaging and LC/MS quantification in riboflavin depletion.....301 Figure S4. hsp-6::GFP expression categories............................................................................303 Figure S5. Lipid analyses with riboflavin depletion indicate increased fat mass........................304 xi Abstract Food is imperative to fuel growth and essential cellular functions for all organisms. All animals receive crucial nutrients from diet to support development, metabolism, aging, and survival. Diet is an extremely variable aspect in life due to the expansive number of options within an organism’s environment. Although this idea isn’t a novel concept because the phrase “You are what you eat” was coined in the 1800’s and is still commonly used in popular culture, understanding how diet can influence phenotypic outcomes has yet to be elucidated. Caenorhabditis elegans have been a common model organism used in diet studies within the laboratory environment due to the shared core metabolic pathway with mammals. In order to diversify the menu available to culture C. elegans in the lab, we have isolated and cultured three such microbes: Methylobacterium, Xanthomonas, and Sphingomonas. The nutritional composition of these bacterial foods is unique, and when fed to C. elegans, can differentially alter multiple life history traits including development, reproduction, and metabolism. In light of the influence each food source has on specific physiological attributes, we comprehensively assessed the impact of these bacteria on animal health and devised a blueprint for utilizing different food combinations over the lifespan, in order to promote longevity. The expansion of the bacterial food options to use in the laboratory will provide a critical tool to better understand the complexities of bacterial diets and subsequent changes in physiology and gene expression. 1 Chapter 1: Introduction 1. Caenorhabditis elegans is an exceptional model to study dietary effects Food is imperative to fuel growth and essential cellular functions for all organisms. All animals receive crucial nutrients from diet to support development, metabolism, aging, and survival. Diet is an extremely variable aspect in life due to the expansive number of options within an organism’s environment. Throughout life, nutrition changes over time due to the evolution of food preferences and the types of diets available, leading to diets never being universally equal across populations. Organisms must learn to distinguish between diets of lower and higher quality to have the best chance for survival. Discovering that diet in the short term can impact metabolism and energy while long term can influence aging and age-related diseases has led to the increase in studies aiming to elucidate why diet holds such profound influence over health and longevity. Although this idea isn’t a novel concept because the phrase “You are what you eat” was coined in the 1800’s and is still commonly used in popular culture, understanding how diet can influence phenotypic outcomes has yet to be elucidated. Caenorhabditis elegans have been a common model organism used in diet studies 1-4 within the laboratory environment due to the shared core metabolic pathway with mammals 2 . With the use of these bacterivore nematodes, many studies have demonstrated how diet plays a role in determining a multitude of physiological outcomes including: developmental timing 5,6 , reproduction 7,8 , healthspan 9,10 , and lifespan 11,12 . Additionally, other studies have shown that diet can induce changes in metabolic profiles 11,13,14 , fat content 2,5,15,16 , feeding behavior 5,12,17 , and the transcriptome 4,18,19 . It has been shown that these effects can be exerted immediately 1,4,5,12,20-24 as well as being capable of influencing future generations 25-31 . C. elegans is an attractive model for dietary studies aimed at understanding how specific diets can impact both health and longevity for numerous reasons. Firstly, C. elegans are cultured in the laboratory environment on a monoculture of bacteria 32-34 , which makes studying the effects of different diets as easy as changing the strain or genera of bacteria for the food 2 source. Second, C. elegans have invariant, short developmental and reproductive periods followed by an averaged 3-week lifespan, which facilitates rapid identification of dietary effects on these traits 35,36 . Third, worms are transparent multicellular organisms that share 60-80% of all human genes and several core metabolic pathways found in humans 37-42 . Fourth, the entire C. elegans genome is sequenced 43 and an expansive collection of knowledge is documented from this organism. For these reasons and more, C. elegans is an exceptional model and a critical resource to realize the impact that diet can have on physiology and gene expression. 2. Worm nutrition C. elegans are free-living nematodes found in soil rich in bacterial content from decaying fruits and rotting vegetation. Less often, worms have also been isolated from slug digestive tracts found in France and Northern Germany 44,45 . Isolation of wild populations allows for characterization of genetic variation underlying diverse phenotypic traits, but also sheds insight into what food sources C. elegans have in their natural environment 44-49 . This is one of the reasons why more and more studies go out and look for natural isolates of C. elegans and why collection protocols have been established 50,51 . One may ask why C. elegans is a good model to study diet and nutrition and the answer is that worms and humans require similar sets of essential nutrients and share basic metabolic pathways 52 . A review recently published in 2019 does a spectacular job or outlining the requirements of C. elegans 53 . In brief, many of the dietary requirements of C. elegans has been discovered during the attempt to create a chemically defined media for axenic culturing 54-56 . C. elegans do not require an essential protein or peptide in its diet except for heme 54,56,57 , unless this protein is the source for required amino acids. Amino acids that must be present and are necessary for C. elegans growth and maintenance are arginine, histidine, lysine, tryptophan, phenylalanine, methionine, threonine, leucine, isoleucine, Figure 1. Dietary requirements of C. elegans. C. elegans require 10 amino acids, 10 vitamins and 2 minerals for normal growth and survival in the lab and the wild. 3 and valine. Interestingly, all but arginine represents essential amino acids for humans 58 . When it comes to macronutrients, carbohydrates are an important dietary constituent because without them C. elegans growth is heavily impaired 55,59,60 . Early studies in the creation of a chemically defined media for C. elegans also identified sterols as important for growth and reproduction 54,56,57 . The requirement for dietary sterols in the laboratory environment is fulfilled with the addition of cholesterol in the plates 61 while decaying plants or fungal material most likely are the source in the natural environment 33 . C. elegans require folate 62 and all other vitamins of the B complex (riboflavin, thiamine, pyridoxine, niacinamide, pantothenic acid, biotin, and cobalamin) 63 for normal growth and development. Potassium and magnesium are absolutely essential minerals required for C. elegans development and growth 53 . Although C. elegans display some important dietary differences from humans, they are still a highly suitable model for genetics of animal nutrition and metabolism. 3. The natural environment of C. elegans differs drastically from that of the laboratory environment C. elegans natural environment C. elegans have been found to inhabit a multitude of environments throughout the world 46-48 . Although they are often found in human-associated habitats like botanical gardens, orchards, and compost heaps, more recently they have also been isolated from forests and scrubland 44,45,49 . The one common characteristic of these habitats is the presence of rotting vegetation rich in microbe species. In nature, C. elegans feed on many species of bacteria found in the soil. Isolation of C. elegans intestines have also revealed that they may feed on some eukaryotes, most commonly yeast 45 . Regardless of where the bacterial food sources come from, the most commonly found bacterial phyla are Proteobacteria, Firmicutes, Bacteroidetes, and Actinobacteria 16 . 4 Recent studies have leveraged these newly discovered bacterial diets to investigate physiological traits that have been commonly studied in the laboratory environment using E. coli 16,64,65 . There has even been a recent push in establishing a collection of bacteria that has been isolated from C. elegans gut in order to start using C. elegans as a model for microbiome research 66 . The reason for this expansion of bacterial diets used in the laboratory environment is because various microbiota genera are able to colonize in the gut and lead to beneficial effects on growth and stress resistance. Protection from pathogenic infection, development, fertility, immunity, and energy metabolism have all been shown to be influenced by bacterial diets in the natural environment 67-70 . Along with these physiological responses, C. elegans behavior is also influenced by bacterial diet exposure. They have been shown to distinguish between beneficial and detrimental food sources and demonstrated food choice behavior preferring more beneficial diets that promote accelerated developmental timing, health, and longevity 67,71,72 .Overall, these studies have led to a greater understanding of the natural history and environment of C. elegans, proving beneficial in informing us of its physiology, behavior, and immunity among other common physiological characteristics. C. elegans laboratory environment C. elegans maintenance in the laboratory environment is on a standardized monoculture of Escherichia coli and requires the routine removal of spontaneously occurring bacterial contaminants. E. coli is used in the laboratory environment not because of its association with C. elegans in the wild, but rather due to its availability in the lab when C. elegans were established as a model organism 73,74 . The uracil auxotroph was found to be a good laboratory diet due to the limited growth on plates, making microscopic analysis easier. A monoculture of E. coli OP50 B strain is used in order to limit experimental variation across research groups. Most of the experimental data thus far that has been collected by the C. elegans community using E. coli Figure 2. E. coli strains used as diets for C. elegans in the laboratory. 5 OP50 as the worm’s diet, including data related to gene expression, development, lifespan and aging. Yet, one of the most powerful tools in C. elegans is RNA-interference (RNAi), which uses the E. coli HT115 K-12 strain. Two other E. coli diets used to grow C. elegans in the lab include HB101 B/K-12 strain hybrid, and a strain derived from OP50 B strain DA837 (outlined in Figure 2). Due to the use of these two distinct bacterial diets in the field, experimental outcomes have led to the understanding that dietary composition and amounts of metabolites and micronutrients available are influencing phenotypic outcomes 5,11,12,17,21,22,24 . To make things even more complex, C. elegans are less likely to encounter E. coli and more commonly found with other bacterial genera in their natural habitat. Therefore, due to C. elegans exposure to naturally occurring bacteria being limited to the natural environment, many aspects of C. elegans biology is masked and/or completely undetectable in the laboratory environment, stressing the importance of expanding bacterial diets used in laboratory settings. Different E. coli strains lead to changes in physiology Even with the lack of bacterial diversity in C. elegans laboratory environment, it has been shown that altering the E. coli species used for experiments can impact physiology in the worms. Previous studies have shown that these different E. coli species do contain different amounts and compositions of nutrients, eliciting varying dietary responses 15,75 . Bacterial diets supply macronutrients to the worms such as carbohydrates, fats and proteins, all of which are needed to generate energy for cellular processes. Gross analysis of these macronutrients have shown that E. coli HB101 and HT115 both contain three to five fold higher levels of total carbohydrates 15 . More detailed analysis of metabolites has shown that the HT115 diet has increases in the amino acids aspartate, glutamate, and lysine while other metabolites increased in concentration include betaine, glucose, lactate, and o-phosphocholie. Intriguingly, HT115 also has formate, oxaloacetate, and propionate, which were not detected in OP50 75 . Furthermore, other studies have begun to detect other metabolites specific to certain E. coli strains important for organismal 6 physiology, including vitamin B12 76,77 , where OP50 is mildly deficient, and tryptophan 10 , where higher levels are detected in HT115 and HB101. Particular diets can influence a plethora of phenotypic outcomes across organisms. One such attribute that can be changed by changing the bacterial diet that worms eat is lifespan. Brooks et al. 2009 observed how the lifespan of worms is significantly increased in the presences of either HT115 or HB101 when compared to worms living on OP50 15 . They also found that this extension of lifespan was lost in the presence of another E. coli strain derived from OP50, DA837, which has been previously used in food preference and satiety studies 20 . This lifespan extension has been further corroborated in a multitude of other studies 4,12,75,78-80 . Healthspan is another attribute that has recently jumped to the forefront of C. elegans research due to the development of many tools allowing for the separation of lifespan effects from alterations in healthspan 81-83 . Similar to humans, C. elegans lose muscle integrity as they age 81 , so movement can be analyzed as a proxy for healthspan. Many studies have shown that this too can be influenced not only by genetic backgrounds, but also by diet the worms are exposed to throughout life. Worms raised on HT115 have shown opposing results between swimming and crawling speed, where crawling has been shown to be decreased 84 while swimming speed has been shown to be increased 4 . HB101-fed animals show an increase in swimming rate 4 . Another muscle that can be measured for healthspan is the pharyngeal muscle due to the number of pumps per minute, which also shows a decline as C. elegans age 4 , which doesn’t seem to be diet-dependent 4 . Although there are other assays that can aid in determining healthspan, autofluorescence levels and the SMURF assay being two, little has been done in literature to examine how these can be influenced by bacterial diet alone. Future experiments should be done to test these healthspan metrics in the context of diet to better understand the full effects diet can have not only on muscle function but other age-related pathologies. 7 Many other physical traits have been examined in the context of diet and are outlined in Table 1 below. The only physiological attribute that has not been reported as changed in the presence of different E. coli strains is reproduction 4,5,12,14,15,75 . This is most likely due to the fact that C. elegans hermaphrodites display extremely efficient internal fertilization with the amount of progeny controlled by the number of sperm produced 85 and this process is tightly regulated 86 . Regardless, it appears that many attributes in C. elegans can be regulated by changes in diet and available nutrients. It is important to note that many of these findings were made in wildtype worms as a control for other experiments involving genetic mutants. Before 2020, there hadn’t been a head- to-head comparison for the three commonly used E. coli strains in the lab (OP50, HT115, and HB101) examining their effects on organismal physiology 4 and there are still many questions that need to be answered. Table 1. Physiological attributes in C. elegans that are influenced by bacterial diet. Physiological Attribute Bacterial Diet Trend Developmental Timing 4,5,16,78,84 HT115 Accelerated developmental timing HB101 Accelerated developmental timing Reproduction 4,5,12,14,15,75 HT115 Similar brood size HB101 Lifespan 4,12,15,75,78-80 HT115 Extended lifespan HB101 Extended lifespan DA837 Normal lifespan Healthspan 4,17,84 HT115 Slower crawling speed, faster swimming speed early in life that declines with age, normal pharyngeal pumping HB101 Faster swimming speed early in life, normal pharyngeal pumping Fat Content 4,15 HT115 Altered fat storage HB101 Decreased fat storage Stress Survival 87 HT115 Enhanced survival in heat stress, increased survival on juglone treatment, increased survival on hydrogen peroxide treatment, prolonged survival on pathogen HB101 - All comparisons made to wildtype C. elegans raised on E. coli OP50. “-” indicates no data was found for this physiological attribute in the context of a particular diet. 8 Live and dead bacteria impact C. elegans in different ways C. elegans feed on bacteria, which serves as a source of food and nutrition, providing carbohydrates, fats, and proteins to support growth and daily activities 88 . Worms eat anywhere from 10 3 -10 5 bacteria per day 89 , so bacterial quality and metabolism can play a large role in regulating behavior and life traits. In their natural environment, C. elegans search for higher quality food and constantly distinguish between pathogenic and nonpathogenic bacteria in order to find the best diet for survival 90-92 . However, it has recently been discovered that bacterial activity has the capability of influencing multiple behaviors and physiological traits 21,75,93,94 , which stresses the importance of either a defined diet or a diet consisting of non-proliferating bacteria. Some studies have even gone to show that proliferating bacteria are detrimental due to proliferation and bacterial colonization of the gut, leading to the eventual demise of the worm 89 . This has led to an increase in studies using killed bacteria, a process that allows bacteria to grow in an overnight culture before being treated, by either heat 95 , ultraviolet irradiation 95,96 , antibiotics 97 , or more recently, paraformaldehyde 19 . Using data gathered by the variety of bacterial-killing techniques, we can begin to decipher how only the nutrients within the bacterial food, and not the host-microbe interactions, can influence physiological changes. C. elegans fed heat-killed bacteria undergo developmental arrest, inducing a protective response similar to a starvation response. When arrested worms are then placed in the presence of live OP50, they recovered to adults and lay viable progeny 98 . When given the option, worms are found less often in the presence of heat-killed bacteria and more often on live bacteria, unless the live food source has been completely depleted. This food choice and developmental delay is partially but significantly rescued with the addition of vitamin B2, suggesting that vitamin B2 levels are very low in heat-killed bacteria. Vitamin B2 is important for regulating the ASP-13 and ASP-14 proteases in addition to the ELT-2 GATA factor, important for C. elegans usage of dead bacteria. These processes were found to be mediated by TORC1, where vitamin B2 promotes TORC1 and 9 protease activity through FAD and ATP production. This FAD-ATP-TORC1-ELT-2 pathway dictates food uptake and foraging behaviors by discriminating against low-quality food, similar to that of heat-killed bacteria 98 . This study demonstrates the importance of bacteria in providing micronutrients to the worms to promote health and fitness. Many caveats exist with using heat-killed bacteria, leading to researchers finding alternative methods for killing food sources for C. elegans research. Other methods have included Ultraviolet (UV) irradiation, which generally involves exposing plates seeded with live bacteria to UV using a crosslinker. These studies have shown that UV-killed bacteria do not effect fat content 99 but do significantly extend lifespan 96,97,100 . This process is thought to occur because E. coli is slightly pathogenic and killing it leaves it nonpathogenic and promotes lifespan. This is similar to studies using antibiotics to kill bacteria, which also extends C. elegans lifespan 96,97 . While both methods have confirmed the reproductive death of bacteria, neither have confirmed if intrinsic metabolic activity persists in these microbes, which could lead to difficulties distinguishing what phenotypes in C. elegans are due to bacterial metabolism and what are due to the food source itself. Heat, ultraviolet irradiation, and antibiotics are all effective methods in preventing bacterial replication, but the bacteria can still remain metabolically active 101,102 . Within the last few years, a new killing technique was established using paraformaldehyde (PFA) which successfully killed bacteria rapidly and reproducibly, removing the metabolic activity altogether 19 . PFA acts by permeabilizing the cell, crosslinking proteins, and creating a mesh-like structure within the cell, which makes cells no longer viable without lysing or destroying their inner structure 103,104 . This maintenance of cell structure is crucial in keeping the bacteria edible for worms. Beydoun and colleagues showed that PFA treatment prevents both replication and metabolic activity in pathogenic and nonpathogenic bacteria. They continue in their research and demonstrate that worms still prefer live bacteria over the PFA-killed bacteria, which is similar to studies that have 10 killed bacteria in other ways 98,105,106 . Intriguingly, this method is better overall because it does not have a significant effect on brood size or lifespan, contrary to heat-killed bacteria, suggesting that the worms are overall healthy. The main difference is a delay in developmental timing, but this delay is better than the worms not developing into adulthood at all, which occurs on heat-killed bacteria 19 . Overall, this new method is promising, especially for use in studies aiming to answer questions about metabolism and aging in C. elegans. Table 2. Wildtype worm phenotypes in the presence of killed E. coli OP50. Phenotype Heat-Killed UV-Killed Antibiotic- Killed PFA-Killed Metabolomics - - - Abundance of 27 metabolites changed – Enrichment of phenylalanine, tyrosine, tryptophan, arginine, and proline 19 Development Arrested development 98 - - Delayed development 19 Reproduction - - - Normal reproduction 19 Lifespan Normal lifespan 106 Longer lifespan 96,97,100 Longer lifespan 96,97 Normal lifespan 19 Food Choice Less preferred than live bacteria and decreased dwelling 98 - - Less preferred than live bacteria 19 Worm Size Normal length 105 - - - Fat Content - Normal fat content 99 - - “-” indicates no data was found for this phenotype. Although the PFA-killed bacteria technique is promising when it comes to removing the effect of proliferating bacteria on C. elegans physiology, it still leads to delayed development and changes in metabolomics. One way around this would be by defining a diet, but this has proven difficult. This would remove the factor of a living organism and allow for a better understanding on how diet, and not interactions with a microbe, are influencing changes in traits. Many groups have attempted to generate a chemically defined diet that is bacteria free, but these diets exert effects on several life history traits and phenotypes 107,108 . It’s also important to note that metabolically 11 active bacteria have been found to produce metabolites that are essential for C. elegans survival 109,110 . Further studies are required to identify bacterial-provided products required for C. elegans growth, development, and overall survival. Luckily more and more studies are identifying what these factors are with the use of mutant E. coli strains 96,100,110-112 . 4. Gene-diet interactions discovered in C. elegans Changes in C. elegans bacterial food source reveal gene functions specific to diet An emerging area of research in the C. elegans field is focused on elucidating how diet can influence life history traits like aging and longevity. These studies have been facilitated in part by the discovery of “gene-diet” pairs where the function of a gene becomes discernable on a particular diet 2 . By expanding the bacterial diets available to C. elegans in the laboratory environment, studies have started to identify specific genes that control physiology in the presence of particular bacterial diet 5,12,15,79,113,114 . Feeding C. elegans different diets has revealed novel signaling pathways unique gene functions in the presence of a particular diet. These gene- diet pair interactions accentuate not only the power of diet on physiology and genetics, but also highlight the complexity of this system. One of the first gene-diet pairs to be established in the field was rict-1, a component of the Target of Rapamycin complex 2 (TORC2) that influences developmental timing, metabolism, and lifespan in a diet-dependent manner 5 . rict-1 mutants have slower developmental timing in the presence of E. coli HB101 diet, display a lower fat content as shown with both BODIPY and Oil Red O fat staining, and an extended lifespan when compared to mutants raised on the E. coli OP50 diet 5 . The alteration of multiple physiological attributes in the presence of a mutant that has been shown to influence feeding behavior led to Soukas et al. to further investigate the feeding behavior of worms on these different food sources, revealing that despite E. coli HB101 being rich in carbohydrates compared to E. coli OP50 15 , worms were leaner due to an alteration in feeding behavior and the worms spending less time on the bacterial lawn. Overall, it was concluded that rict-1 is an integral part in a pathway regulating feeding 12 behavior in the presence of different bacterial diets. Demonstrating that diet-sensing can influence and organism’s ability to adapt to food sources. An additional gene important for neuroendocrine signaling and food-seeking behavior pathways is an intestinal peptide transporter pept-1. In 2009 pept-1 was implicated in regulating fat storage in C. elegans fed different diets containing varying levels of carbohydrates and fats 15 . pept-1 mutants showed no differences in fat storage with fed either E. coli OP50 or HB101, but they did show reduced reproductive output on the HB101 diet. These results suggest that pept-1 doesn’t necessarily change overall fat content, but through feeding rates may prevent adequate assimilation of nutrients leading to reduced progeny output 15 . Soon after these data were published, Maier et al. identified two more genes important in dietary-sensing displaying diet- dependent effects on lifespan. nmur-1, a mammalian homolog of the neuromedin U receptor and osm-3, a gene encoding kinesin motor protein important for sensory cilia formation 12 . Both mutants exhibit normal lifespan in the presence of E. coli HT115 but a shortened lifespan when raised on OP50. Co-expression patterns were observed between the two mutants in sensory neurons, suggesting that both genes play an important role in the processing of sensory information the worm receives when introduced to different bacterial diets 12 . Altering dietary components is a clear way to profoundly affect metabolism and lipid content in organisms. This is clearly seen in the SKN-1 gain-of-function (skn-1gf) mutants in a diet- dependent manner through the distribution of lipids. When skn-1gf mutants are fed the E. coli OP50 diet, the worms undergo this age-dependent somatic depletion of fat phenotype that demonstrates the loss of lipids in the soma but maintenance of lipids in the germline post- reproductively. However, this is not seen in the skn-1gf mutants fed the HT115 diet 113,115 . This demonstrates a trade-off in C. elegans in response to dietary carbohydrates which mobilizes lipids from the intestine to the germline to support fecundity at the cost of survival and longevity. 13 Reminiscent of another gene-diet interaction, alh-6 mutants also undergo changes in lipid homeostasis dependent upon the diet they are fed. alh-6 is a mitochondrial proline metabolism gene that encodes for 1-pyrroline-5-carboxylate dehydrogenase (P5CDH), a mitochondrial enzyme needed to catalyze the P5C to glutamate reaction to avoid accumulation of reactive oxygen species (ROS) 22,23,113 . Interestingly, the alh-6 mutants showed a premature aging phenotype when raised on the E. coli OP50 diet but not when raised on HT115. Intriguingly, the authors also tested if this change in lifespan was dependent upon the life stage of exposure to the diet. When alh-6 mutants are exposed to OP50 between larval stage 3 and larval stage 4 into adulthood, this shortened lifespan is diminished, indicating that shortened lifespan in these mutants on OP50 is timing-dependent. This suggests that there may be more critical life stages throughout development that can be influenced by diet and cause premature aging through increased ROS and abnormal mitochondrial morphology 22,23,113 . Furthermore, these data signify the capacity of diet to not only induce changes in physiology, but to also exert changes in organelle morphology as we age. mrpl-2 is a mitochondrial ribosomal protein orthologous to human MRPL2 which, when mutated, can activate the mitochondrial unfolded protein response (UPR mt ) and change mitochondrial functionality in a diet-dependent manner 114 . The effects of the mrpl-2 mutant on E. coli OP50 were also shown to go beyond mitochondrial UPR activation. Amin and collogues also showed that oxygen consumption levels and ATP levels were elevated in mutants on OP50, accompanied by reduced mitochondrial membrane potential. Since these traits are associated with increased mitochondrial dysfunction, they also looked to see how the amount of reactive oxygen species (ROS) within the cells was changing, showing that there was reduced oxidative damage 114 . Furthermore, mrpl-2 also increases lifespan and healthspan on the OP50 diet in addition to conferring extended survival on the pathogen Pseudomonas aeruginosa. Further exploration 14 revealed the mechanism involved in activating the UPR mt in mrpl-2 mutants on OP50 was associated with vitamin B12 availability, because when OP50 was supplemented with B12, phenotypes were lost 114 . This study highlights the importance of not only examining overall dietary intake when considering causes for changes in physiology, but to also look more specifically at small changes in diet, like micronutrient content. Vitamin B12 availability seems to play a key role in gene-diet interactions because flr-4, a serine- threonine gene originally identified as a regulator of fluoride resistance, induces physiological changes in the presence of vitamin B12 80,116 . flr-4 mutants display an increased lifespan when fed E. coli HT115 compared to the mutants on OP50, but also have an increased healthspan as measured by lipofuscin pigment accumulation, motility, and muscle nuclei integrity 79 . This seems to be mediated by an increase expression of the cytoprotective xenobiotic detoxification pathway (XPD) since the mutant on HT115 activates genes in this pathway. The increase in lifespan and healthspan is shown to be dependent upon multiple components of the p38 MAPK pathway as well 79 . These phenotypes are lost on diets lacking in B12 or when diets are genetically modified, highlighting the importance of varying micronutrient content in maintaining C. elegans adaptive capacity to different diets. Tryptophan is another bacterially secreted metabolite that has been shown to influence physiology in specific C. elegans mutant backgrounds. In 2020, Brinkmann et al. discovered that bacterially produced tryptophan was responsible for lifespan extension of an ahr-1 mutant. ahr-1 is an aryl hydrocarbon receptor (AhR) that forms a heterodimer with C. elegans Arnt homolog AHA-1 and binds xenobiotic responsive elements (XREs). By exposing ahr-1 mutants to different E. coli diets, it was revealed that AHR-1 affects aging in a diet-dependent manner. ahr-1 extends lifespan of worms raised on the E. coli HT115 diet but not on the OP50 diet. Moreover, loss of ahr-1 extended both the lifespan and healthspan of two age-associated disease models, namely animals with 15 muscle expression of polyglutamine (polyQ40) and alpha-synuclein 2 . For all phenotypic alterations, the HT115 bacteria must be metabolically active and produce tryptophan; supplementation of tryptophan abolished the differences observed in lifespan and healthspan between wildtype worms and ahr-1 mutants. nhr-114, nuclear hormone receptor 114, has also been shown to have physiological attributes change in response to dietary levels of tryptophan 24 . Gracida and Eckmann found that the absence of nhr-114 on certain bacterial diets led to sterility; animals fed the E. coli OP50 diet were sterile while nhr-114 mutants on both HT115 and HB101 laid viable progeny. Further investigation revealed that this phenotype is due to prevalent cellular defects in the germline. The proliferative germ cell pool fails to expand in nhr-114 at larval stage 3 (L3), leading to accumulation of cell-division defects, reducing competence for both meiotic differentiation and gamete production. Additionally, this was found to be tryptophan-dependent because supplementation of less-optimal diets compensates for the loss of nhr-114 function 24 . This relationship between dietary tryptophan, ahr-1 and nhr-114 accentuates the importance of diet and available nutrients in eliciting physiological responses. Table 3. Gene-diet pairs in C. elegans induce diet-dependent physiological changes. Gene Description Bacterial Diet Impact rict-1 5 Component of the Target of Rapamycin complex 2 (TORC2) OP50 Increased fat, shortened lifespan HT115 Decreased fat, lengthened lifespan HB101 Decreased fat, lengthened lifespan pept-1 15 Intestinal peptide transporter OP50 Normal brood size HB101 Reduced brood size nmur-1 12 Mammalian homolog of the neuromedin U receptor OP50 Lengthened lifespan HT115 Normal lifespan osm-3 12 Kinesin motor protein OP50 Lengthened lifespan HT115 Normal lifespan alh- 6 22,23,113 Proline metabolism: 1- pyrroline-5-carboxylate dehydrogenase (P5CDH) OP50 Shortened lifespan, increased lipid depletion under acute starvation, abnormal mitochondrial morphology, increased ROS production, decreased ATP production 16 HT115 Normal lifespan, normal lipid depletion under acute starvation, normal mitochondrial morphology, normal ROS production, normal ATP production skn- 1 113,115 Transcription factor ortholog to mammalian Nuclear factor- erythroid-related factor (NRF) OP50 Age-dependent somatic depletion of fat (Asdf) HT115 Non-Asdf flr-4 79,80 Serine-threonine gene OP50 Normal lifespan, normal health HT115 Lengthened lifespan, increased health mrpl-2 114 Mitochondrial ribosomal protein gene OP50 Increased oxygen consumption, increased levels of ATP, reduced levels of oxidative damage, lengthened lifespan, improved muscle function, increased survival after pathogen infection HT115 Normal oxygen consumption, normal levels of ATP, unchanged levels of oxidative damage, normal lifespan, normal muscle function, normal survival after pathogen infection HB101 Normal oxygen consumption, normal levels of ATP, unchanged levels of oxidative damage, normal lifespan, normal muscle function, normal survival after pathogen infection ahr-1 10 Aryl hydrocarbon receptor OP50 Normal lifespan, normal healthspan HT115 Lengthened lifespan, extended healthspan nhr-114 24 Nuclear hormone receptor, mammalian ortholog of HNF4 OP50 Sterile HT115 Fertile HB101 Fertile These studies further emphasize the importance diet has on physiological outcomes in organisms. Not only can changing the overall dietary intake alter physiology, but making small adjustments in micronutrient content can be just as powerful 52,117 . A core reason for why these gene-diet interactions were discovered was due to a common experimental design that exists across C. elegans laboratories: raising C. elegans on one diet (E. coli OP50) and then doing forward genetic screens with another diet (E. coli HT115). This is due to the fact that HT115 was used in the engineering of the first feeding RNA-interference (RNAi) libraries mainly due to HT115 being deficient in the dsRNA specific endonuclease RNAse III 118-120 . It wasn’t until 2015 that a new 17 RNAi OP50 strain was developed, which has revealed new results in screens that were thought to be fully saturated 84 . With these new tools, the future of the field indeed looks bright. By incorporating multiple diets, the genetic pathways underlying diet-dependent regulation of physiological attributes will start to be revealed. 5. Conclusions and perspectives Even though much work has been done in identifying gene-diet interactions in C. elegans and characterizing the effects of different bacterial diets on life history traits, the molecular mechanisms behind these responses to diet have yet to be fully understood. Dissecting the interplay of diet, metabolism, and other physiological attributes remains an important goal due to the increasing number of foods that become available within society. Further research is required in understanding what important metabolites are necessary for worms and currently are provided by the live diet. Studies that expand the bacterial diets in the laboratory environment by isolating bacterial contaminants or brining bacteria from the natural environment, will aid in a greater understanding of how diet can be leverage for the greater good of organisms. Work with natural isolates of C. elegans will contribute to the identification of novel gene-diet pairs that are normally masked in the laboratory environment by the wildtype “N2” strain. The development of an effective bacterial-killing method or a defined diet that doesn’t change C. elegans physiological attributes will be helpful in dissecting these mechanisms. Altogether, these studies will contribute to the ever-growing objective of identifying diets or metabolites that can be used to delay aging, the onset of age-related diseases, and extend lifespan. Furthermore, by identification of important dietary components and the genetic pathways crucial for healthspan and lifespan, we may be able to start using diet as a nutraceutical to target further diseases that affect populations of individuals throughout the world. 18 Chapter 2: Bacterial diets differentially alter lifespan and healthspan trajectories in C. elegans *This is a version of a manuscript published in Nature Communications Biology Authors: Nicole L. Stuhr 1,2 and Sean P. Curran 1,2,3,* Affiliations: 1 Leonard Davis School of Gerontology, University of Southern California, 3715 McClintock Ave, Los Angeles, CA, 90089 USA 2 Dornsife College of Letters, Arts, and Science, Department of Molecular and Computational Biology, University of Southern California, 1050 Childs Way, Los Angeles, CA, 90089 USA Norris Comprehensive Cancer Center, University of Southern California, 1441 Eastlake Ave, Los Angeles, CA, 90033 * Corresponding author: Sean P. Curran E-mail: spcurran@usc.edu Running Title: Diet strategies to manipulate health Keywords: diet, lifespan, healthspan, aging, development, reproduction, metabolism, nutraceutical 19 ABSTRACT Diet is one of the more variable aspects in life due to the variety of options that organisms are exposed to in their natural habitats. In the laboratory, C. elegans are raised on bacterial monocultures, traditionally the E. coli B strain OP50, and spontaneously occurring microbial contaminants are removed to limit experimental variability because diet - including the presence of contaminants, can exert a potent influence over animal physiology. In order to diversify the menu available to culture C. elegans in the lab, we have isolated and cultured three such microbes: Methylobacterium, Xanthomonas, and Sphingomonas. The nutritional composition of these bacterial foods is unique, and when fed to C. elegans, can differentially alter multiple life history traits including development, reproduction, and metabolism. In light of the influence each food source has on specific physiological attributes, we comprehensively assessed the impact of these bacteria on animal health and devised a blueprint for utilizing different food combinations over the lifespan, in order to promote longevity. The expansion of the bacterial food options to use in the laboratory will provide a critical tool to better understand the complexities of bacterial diets and subsequent changes in physiology and gene expression. 20 INTRODUCTION Since the discovery that aging can be improved by dietary, genetic, and pharmacological interventions, there has been an increase in studies aiming to elucidate how each of these interventions can promote long life and healthy aging. Many dietary interventions including calorie restriction, intermittent fasting, and dietary restriction 1 have been shown to not only increase maximal lifespan, but average lifespan of the cohort and healthspan as well. New diets and fads have become popularized within society 2,3 , however, these dietary fads focus on removing one aspect of diet (carbohydrates, sugars, fats, proteins, etc.) instead of focusing on the overarching complexity of food. Food is imperative for all organisms in order to provide nourishment to fuel growth and essential cellular functions. Diet is one of the more variable aspects in life due to the vast options that organisms are regularly exposed to in their natural habitats. Survival within these environments require organisms to select high quality food sources from a range of nutritiously diverse alternatives. For centuries, society has accepted that diet is important for health and longevity; “You are what you eat.” However, our understanding of why diet holds such profound influence over our health and how we can use this knowledge to improve overall health and longevity requires further study. Many diet studies have employed the use of the model organism Caenorhabditis elegans 4-6 due in part to the many shared core metabolic pathways with mammals 6 . Understanding how the introduction of specific dietary foods impact both health and lifespan in the worm may aid in explaining the variability in the rates of aging and severity of age-related disease. These bacterivore nematodes can be used for the identification of dietary effects due to their invariant and short developmental and reproductive periods, which are followed by an averaged 3-week lifespan 7 . 21 The impact of bacterial diets on physiology is vast and certainly integrates into multiple life history traits including development 8 , reproduction 9 , healthspan 10 , and longevity 4,11-13 . Additional studies have identified diet-induced phenotypic effects that are accompanied by measurable differences in metabolic profiles 11,14,15 , fat content 5,6,8,16 , and feeding behavior 8,13,17,18 . C. elegans interact with a variety of bacterial species in their natural environment, including Proteobacteria, Bacteroidetes, Firmicutes, and Actinobacteria 19 . Under normal laboratory conditions, C. elegans are cultured using the standardized bacterial species Escherichia coli OP50. This bacterium was not chosen because of its association with nematodes in the wild, but because of availability in the laboratory setting 16 . Maintenance of C. elegans in the laboratory includes the removal of spontaneously occurring bacterial contaminants in order to limit experimental variations. The lack of exposure of C. elegans to naturally occurring bacterial species have led to many aspects of C. elegans biology becoming undetectable in the artificial laboratory environment 16,19-22 . We routinely noticed that C. elegans are found feeding on contaminants, when present, rather than the E. coli food source supplied. With this in mind, we looked at these contaminants as an opportunity to examine the impact that different food sources have on physiology. We identified the genera of three such “contaminants” as bacterium that can be found in C. elegans natural habitats. As such, we decided to use these three microbes alongside the three most commonly used E. coli fed to worms, to expand the menu of bacterial diets and provide insight into the contribution that different foods have on C. elegans physiology 4,23 . Our studies provide a comprehensive assessment of changes in physiology and transcriptomic signatures as a result of C. elegans exposure to bacteria found in the laboratory environment versus their natural environment. To our knowledge, this is the first side-by-side comparison documenting how these six bacteria differentially effect multiple key aspects of C. elegans physiology over the lifespan. These data represent a critical resource to realize the impact different bacterial foods can have on multiple aspects of C. elegans physiology and gene expression. 22 RESULTS Identification and characterization of three distinct bacterial diets In their natural environment, C. elegans need to adapt to a variety of microbes they might encounter as a source of food. The necessity to cope with varied food sources have likely shaped multiple aspects of their biology, which are masked in the artificial laboratory environment where they are fed the unnatural diet of E. coli 7 . In order to elucidate the underlying connection between bacterial diet and physiology, it is important to evaluate how different foods influence behavior and physiological attributes. We isolated and cultured three bacterial species as noteworthy C. elegans foods that we call “Red”, “Orange”, and “Yellow” due to their pigments when grown on plates (Fig. 1a). Using 16S ribosomal sequence alignment, we identified the genus of each new bacterial diet: Red is Methylobacterium; Orange is Xanthomonas, and Yellow is Sphingomonas. Intriguingly, these three bacterial diet genera can be found in C. elegans natural environments 16 . We next identified optimal growth parameters to ensure monoculture growth for each bacteria in LB broth and on plates (Supplementary Fig. 1a). Although most species were able to grow at 37°C on LB solid media and in LB liquid media, Red/Methylobacterium required a growth temperature of 30°C on LB solid media, while Yellow/Sphingomonas required a growth temperature of 26°C in LB liquid media. The growth rate of each bacterial species was similar, except Red/Methylobacterium, which was markedly reduced (Fig. 1b and Supplementary Fig. 2a). Preceding research has identified many distinct differences in macronutrient composition between the three E. coli strains used to raise C. elegans in the laboratory. Not only do these strains differ in carbohydrate levels 24,25 , but also contain varying levels of vitamin B12 25 , dietary folate and tryptophan 26-28 . The dissimilarity between these bacterial compositions has been shown to alter many phenotypic attributes in C. elegans. To assess the nutritional value of the bacterial diets 23 used in our studies, we first performed bomb calorimetry to define the total caloric composition of each bacteria. Surprisingly, the total caloric composition was not significantly different in any of the bacteria in comparison to the E. coli OP50/B bacterial diet. In order to better assess the differential nutrient composition of each bacteria, we next measured the concentrations of glucose, glycerol, glycogen, triglyceride, and water to define a nutritional profile for each bacterial food source (Fig. 1c and Supplementary Fig. 1b-g). Orange/Xanthomonas was the most significantly different food among the three distinct bacterial diets because glucose, glycerol, glycogen, and water content were all higher than the levels found in OP50/B. Red/Methylobacterium displayed a similar trend with an increase in glycerol, glucose, and water content, but not triglycerides. Finally, Yellow/Sphingomonas only carried more glycerol and water content, compared to E. coli OP50/B. Interestingly, E. coli HB101/B&K-12 was the only bacterial diet with an increase in triglyceride content relative to E. coli OP50/B. We also noted that glucose and glycogen content were highest in Orange/Xanthomonas, glycerol and triglyceride content were highest in E. coli HB101/B&K-12, and water content was highest in Yellow/Sphingomonas. Taken together, Red/Methylobacterium, Orange/Xanthomonas, and Yellow/Sphingomonas represent three new potential C. elegans food sources, with similar calorie content, but differential nutritional composition to E. coli OP50/B that could be used to investigate dietary effects on animal physiology. Bacterial diets direct unique transcriptional signatures We next confirmed that worms could be cultured on each bacteria as the sole source of nutrition to enable the systematic characterization of the impact of these foods on C. elegans health. After maintaining animals on each bacterial food for more than 30 generations, we then compared animals reared on these new laboratory bacterial foods alongside populations grown on the three most common E. coli foods (OP50, HT115 and HB101). 24 Because diet can impact multiple cellular processes, we first assessed the steady state transcriptional response to each bacterial diet by RNA-seq (Fig. 2a and Supplementary Data 1). Each food source evoked a unique transcriptional signature (Figs. 2b-f and Supplementary Data 1) when fed to the animal over multiple generations, which revealed the impacts of food on multiple physiological responses. We compared the relative expression for each gene to age- matched animals fed OP50 and discovered that there were 743 genes altered in animals fed HT115, 3512 genes altered in animals fed HB101, 455 genes altered in animals fed Red, 1149 genes altered in animals fed Orange, and 6033 genes altered in animals fed Yellow. Of these, 102 genes were uniquely altered in animals fed HT115, 320 on HB101, 40 on Red, 24 on Orange, and strikingly 2667 on Yellow (Fig. 2g and Supplementary Data 1). In addition to finding genes that were shared on two, three, and four food types, we also identified 140 genes that were altered on all five bacterial foods, as compared to OP50-fed animals. With the analysis of these 140 genes, we were able to see that in terms of transcriptional response, Red and HT115 were more similar to each other and more closely related to OP50 while Orange, Yellow, and HB101 shared many similarities (Fig. 2h). An analysis of the Gene Ontology (GO)-terms of the unique genes that were differentially expressed in animals fed each bacteria type revealed specific molecular signatures for each diet (Supplementary Table 1). Notably, we observed expression changes for genes that could alter multiple physiological processes including development, metabolism, reproduction, and aging. In consideration of this enrichment, we explored the impact of the different bacterial food sources on each of these critical physiological attributes. Food-induced acceleration of developmental timing Our ability to culture animals on each bacterial food revealed that they were sufficient to support life, but as food can influence multiple life history traits, we further examined the physiological 25 responses to the foods relative to the standard OP50 food. Bacterial diet can impact developmental timing 24 and as such, we asked whether any bacterial food altered the time to reach each larval stage using the mlt-10::GFP-PEST reporter strain, which marks the transition of each animal as it progresses through its four developmental molts 29 ; noting that on OP50 larval development occurs with invariant timing (Fig. 3a). We determined that all bacterial foods facilitated successful development to reproductive adulthood but remarkably, each food source resulted in faster development relative to OP50-reared animals (Figs. 3b-f and Supplementary Fig. 3). We noted three important variables in this precocious development: first, bacteria could accelerate development of specific larval stages – HT115 (Fig. 3b) and Red (Fig. 3c) resulted in early transition from larval stage 2 (L2) to larval stage 3 (L3), Orange (Fig. 3d) resulted in early transition from L3 to larval stage 4 (L4), and HB101 (Fig. 3e) and Yellow (Fig. 3f) resulted in early transition from L4 to adulthood. Second, with the exception of the precocious transition between two larval stages listed above, the time spent in each larval stage (peak to peak time) was one to three hours shorter compared to OP50 raised animals (Supplementary Fig. 3). Third, the time to molt (trough to trough time) was mostly unchanged (Fig. 3a). In addition, we measured animal size at each developmental stage, which showed some small but significant differences in body size on different bacterial foods, similar to observations from other studies 30-33 . Worms raised on HT115 contained a bigger area from L4-day 1 adult stage, Orange worms had a smaller area until day 2 of adulthood, and the other foods resulted in a fluctuation of larger and smaller areas throughout developmental stages, relative to OP50 (Supplementary Fig. 3). These data reveal that animals fed the new Red, Orange, and Yellow bacterial diets or the commonly used E. coli foods HT115 or HB101, reach reproductive maturity faster, as compared to the standard OP50 food source, which could influence animal fitness. Due to both the difference in total time to reproduction and time to progress through each developmental stage, we examined the expression of C. elegans genes important for 26 development (Figs. 3g-h). Based on GO-term classifications, worms fed the Yellow bacteria had the greatest number of differentially expressed developmental genes, while worms fed Red bacteria had the fewest. Intriguingly, of the 140 genes with shared expression changes between the five diets, 12 of these genes are involved in the molting cycle (sqt-2,sqt-3,ptr-4,ptr-18,mlt- 7,mlt-10,qua-1,bli-1,dpy-3,rol-6,acn-1 and noah-1) 29,34 . Bacterial diets differentially affect lipid homeostasis Diet directly impacts overall metabolism, which is under tight genetic control 23,35 . The unique metabolic profiles of each bacteria drove us to explore how these different microbial diets could affect organismal metabolism. As changes in intracellular lipid stores are easily measured by microscopy, we stained age-matched populations of fixed animals with two lipophilic dyes: Nile Red (NR) to measure total lipid content (Figs. 4a-g) and Oil Red O (ORO) to identify lipid distribution (Figs. 4h-n and Supplementary Fig. 4). Relative to OP50-fed animals (Fig. 4a), overall lipid levels were increased in animals fed HT115 (Figs. 4b,g) and Orange (Figs. 4e,g), but were markedly decreased in HB101 (Figs. 4c,g), Red (Figs. 4d,g), and Yellow (Figs. 4f-g). Under most bacterial foods, lipid distribution across tissues were similar (Figs. 4h-l,n and Supplementary Fig. 4), except for animals fed the Yellow bacteria, which resulted in the depletion of intestinal lipids in late reproductive adults, a phenotype known as age-dependent somatic depletion of fat (Asdf) at day 3 of adulthood (Figs. 4m-n) 36,37 . The fact that animals fed each of the bacteria developed faster than OP50-reared animals indicates animals were not nutritionally deprived, which can result in developmental delay 19,38-42 . Nevertheless, different microbes have been shown to alter rates of pharyngeal pumping, which regulates ingestion of bacteria 43 . Pharyngeal pumping rates were examined in both the L4 stage and day 1 adult animals, but none of the bacterial diets significantly altered pharyngeal pumping (Fig. 4o and Supplementary Fig. 4). As a complement to the pharyngeal pumping analyses, we 27 wanted to measure overall food intake from the L4 stage of development to day 4 of adulthood (Fig. 4p). Similar to the pharyngeal pumping data, the food intake assay also demonstrated that the amount of food being eaten by a single worm at a certain point in development was not significantly different. This indicates that worms are eating similar amounts of food throughout their lives, which suggests that fat phenotypes observed above are not due to the amount of food ingested, but rather the composition and nutritional aspect of the bacteria. Because each bacterial diet represented a specific metabolic profile and C. elegans reared on these bacteria differed in the amounts of stored intracellular lipids, we examined the expression of genes that regulate lipid metabolism and homeostasis (Figs. 4q-r). Of the total number of genes that were differentially expressed on each bacterial food source, about 1/3 of the genes in worms raised on each food were related to metabolism. Feeding animals the Yellow bacteria evoked the largest number of metabolism-related gene changes, which correlated with our phenotypic observation that animals fed this food store the lowest levels of fat at the L4 stage (Fig. 4g) and it was the only bacteria to drive the Asdf phenotype later in life (Fig. 4m). Among the genes differentially expressed on these new food sources were fat-5 and fat-7, which may suggest changes in lipid biosynthesis pathways, as well as multiple lipases (Lipl-class), suggesting changes in lipid utilization (Fig. 4g). Red and Yellow reduce reproductive output We next investigated how each bacterial diet influenced fertility by measuring the total number of viable progeny laid by individual animals over a ten-day reproductive span. The total number of progeny was reduced by ~25% when animals were fed the Red bacteria, modestly reduced (~10%) in animals fed the Yellow bacteria, and unchanged on the other microbial foods (Fig. 5a). Although animals reached peak reproductive output at approximately the same time, animals fed Red and HB101 ceased reproduction sooner than all other groups (Fig. 5b and Supplementary 28 Fig. 5). Reproductive output of C. elegans hermaphrodites is determined by the number of spermatids generated at the L4 larval stage 44,45 . To determine if the decreased progeny production and early loss of reproductive output were due to diminished sperm availability in Red-reared worms, we mated hermaphrodites with males, which increases total reproductive output (Fig. 5c). Males raised on either OP50 or Red could increase total reproductive output similarly when mated to OP50-reared or Red-reared hermaphrodites, which indicated that sperm are functional when in excess. Surprisingly, when mated, hermaphrodites raised on Red have significantly more progeny as compared to hermaphrodites raised on OP50. We also counted the number of unfertilized oocytes laid by unmated hermaphrodites, which revealed HT115, HB101, Red, and Orange bacteria resulted in markedly reduced expulsion of unfertilized gametes (Fig. 5d) while hermaphrodites on Yellow were similar to animals fed OP50. An analysis of the reproduction- related genes that were differentially changed in animals fed each of the bacterial foods revealed several bacteria-specific changes (Figs. 5e-f). Bacterial foods differentially alter organismal life expectancy and health The quantity of diet ingested has been shown to influence lifespan across organisms 46,47 . Similarly, diet quality and composition can also influence healthspan, but the underlying mechanisms of healthspan improvement, and its relationship to diet, remain underdeveloped. As such, additional models to explore how diet can impact life- and healthspan are of great interest. To this end, we examined impact of each new food on organismal lifespan. Similar to previous studies 13 animals raised on HB101 had a modest increase in mean lifespan, with the most significant impact on the last quartile of life (Fig. 6a and Supplementary Data 2). In contrast, worms raised on HT115, Red, and Yellow displayed significantly increased lifespan (Figs. 6b-d and Supplementary Data 2) while animals raised on Orange were short-lived (Fig. 6e and Supplementary Data 2). Taken together, our results show that Red, Yellow, and Orange represent 29 three bacterial foods of sufficient nutritional quality to accelerate development, but differentially alter life expectancy in a food-dependent manner. We examined the transcriptional profiles of genes previously annotated to be involved in lifespan and discovered several interesting trends that may explain the lifespan-altering effects of each bacterial diet (Figs. 6f-g). Feeding the Yellow bacteria results in an increase in lifespan and, correspondingly, genes in the insulin-like signaling pathway (daf-2/Insulin receptor and age- 1/PI3K) were downregulated, while DAF-16/FoxO target genes (Dao and Dod) were upregulated. Similarly, although less significantly, animals fed Red bacteria have increased expression of several genes regulated by DAF-16. Moreover, HT115, Red and Yellow all extend lifespan and share 11 genes with altered expression, all of which are upregulated. In previous studies, these 11 genes (abu-1, abu-11, abu-14, gst-10, kat-1, lys-1, pqn-2, pqn-54, lpr-5, glf-1, bus-8) lead to a shortened lifespan phenotype when the expression is reduced by RNAi 34 . Finally, feeding worms the Orange bacteria significantly decreased lifespan and although we did not identify any known longevity genes to display transcriptional changes, we found that kgb-1, pha-4, and vhp-1 were all up regulated on this diet, while previous studies have observed extended lifespan when these genes are targeted by RNAi 34 . Because studies done in the past have shown that dietary restriction can influence lifespan 1 , we decided to also look at a subset of genes shown to differ between ad libitum and dietary restricted conditions 48 . Interestingly, there was a similar proportion of genes in worms on each bacterial diet that were differentially expressed (Figs. 6h-i). Furthermore, Red and HT115 were more closely related to the OP50-raised worm genetic profile while HB101, Orange, and Yellow contained a higher number of genes that were both up and downregulated. Collectively, while longevity is a complex phenotype, feeding any of these bacterial diets can evoke transcriptional changes in genes associated with life expectancy. 30 Since lifespan can be uncoupled from healthspan 10,49 we examined the impact of each bacterial diet on animal muscle function via thrashing as a surrogate for health (Figs. 7a-e and Supplementary Fig. 6). We measured thrashing rate at five specific timepoints that were selected for their significance in relation to other phenotypes: L4 larval development stage, based on the RNAseq data and fat staining analysis; day 1 of adulthood, based on the pharyngeal pumping assay; day 3 of adulthood, based on the presentation of the Asdf phenotype; day 8 of adulthood, due to 50% of the Orange population being dead at this time point; and day 11 of adulthood when 50% of the OP50 population had perished. Surprisingly, animals raised on HT115, HB101, Red, and Yellow had faster muscle movement while Orange were indistinguishable from OP50-fed animals at L4 stage (Fig. 7a and Supplementary Data 3) and at day 1 of adulthood (Fig. 7b and Supplementary Data 3). Although most day 3 animals had similar muscle function, animals fed HT115 and Orange had modestly slower movements (Fig. 7c and Supplementary Data 3). By day 8 of adulthood, Red-fed animals displayed significantly slower movements (Fig. 7d and Supplementary Data 3). Day 11 adults raised on HT115, HB101, Yellow, and Orange moved half as fast as they did at the end of development, similar to OP50-raised control animals, but Red- fed animals we significantly slower (~75% reduction) (Fig. 7e and Supplementary Data 3). Taken together, despite some bacteria enhancing early life muscle function, age-related decline remained similar, and was even enhanced in animals fed Red (Fig. 7f). It is notable that when examining the GO-term analysis of the 30 shared genes that are deregulated on all five of the bacterial foods, as compared to OP50, all 30 genes have been annotated as causing movement defects in C. elegans when the expression of that gene is altered 50-52 . C. elegans are least often found dwelling on the Red bacteria when given the choice of other bacteria Previously it has been shown that C. elegans display behaviors indicative of preference for bacterial diets that aid in better development, reproduction, and lifespan. Previous studies have 31 suggested that dietary choices are made based on both the quality of food and the impact on the survival of the bacterial-feeding nematodes. Importantly, many animals select their food according to their environmental and dietary requirements 53,54 . The overall trend of these past studies suggests C. elegans are choosing food sources aiding in higher fitness. Knowing that the six bacterial foods used in this study are capable of exerting diverse phenotypic changes (Table 1), we wanted to determine if foods we observed as more beneficial for longevity and healthspan would be what the worms were found feeding on when given multiple options. We set up a food choice assay to test this hypothesis with the use of OP50-reared worms (Fig. 8a). The food choice assay had the option of all six bacterial diets (Fig. 8b). When OP50-reared worms had the option to choose from any of the six bacteria, a trend was observed where the Red bacteria would be the only food source lacking dwelling worms. Taken together, this data suggests that C. elegans will choose any of the five bacterial foods over that of the Red bacteria, which is interesting since the Red bacteria promotes longevity. The standard monoculture of OP50 in the laboratory environment is generally nonpathogenic and wild type worms typically remain on the bacterial lawn until all bacteria is eaten. In the case of pathogenic bacteria, C. elegans tend to initially eat the bacteria when encountered and then move off of the lawn in a span of a couple hours, a term commonly called lawn avoidance 55,56 . In order to determine possible reasons for why C. elegans are less commonly found dwelling on the Red bacteria, we employed a lawn avoidance assay, which counts the number of worms residing on and off of a bacterial lawn at the L4 stage (Fig. 8c) and day 1 adult stage (Fig. 8d) of development. These assays revealed that the Red bacteria does not induce an avoidance response at either L4 or the day 1 adult stage. In fact, the proportion of worms found on the Red bacterial lawn is equivalent to the proportion of worms found on most of the other bacterial lawns. The one outlier to this is the Orange bacteria during the L4 stage, which has about forty percent of the animals residing off of the bacterial lawn. Knowing that pathogenic bacteria evoke an immune response, 32 we asked if this was reflected in the RNA sequencing data. We examined a list of genes that are commonly found to be altered in worms when in the presence of pathogenic bacteria and found that the proportion of genes differentially expressed in worms on each bacterial diet ranged from 3-10% (Fig. 8e). When comparing the expression profiles of these immune response genes, we found that HT115 and Red bacteria resulted in similar gene expression responses while Orange and Yellow were closer to HB101 (Fig. 8f). Bacterial diet exposure at different timepoints in development alter lifespan trajectories Previous studies have demonstrated that effects of calorie restriction (CR) can be realized even when initiated later in life 57,58 . Moreover, when fed ad libitum after a period of CR, mortality is shifted as if CR never occurred. With this model in mind, we asked if switching bacteria type, rather than abundance, could be wielded to alter lifespan outcomes. To address these questions, we switched growth conditions at major life stages (development, reproduction, post- reproduction) and measured lifespan (Fig. 9a). We used both Orange bacteria, which decreases lifespan, and Red bacteria, which increases lifespan, as the basis of our model. Remarkably, after assessing eight combinations of bacterial diet switching, we discovered that the bacteria fed during development (food 1) and reproductive period (food 2) contributed little to overall lifespan and the last bacteria exposed (food 3) exerted the most impact. In brief, when animals were exposed to the Red bacteria after experiencing the Orange bacteria, the normally shortened lifespan resulting from ingestion of Orange is suppressed (Figs. 9b-d and Supplementary Data 4). Conversely, exposure to Orange suppresses the normally extended lifespan linked to whole- life feeding of the Red bacteria (Figs. 9e-g and Supplementary Data 4). Intriguingly, when compared to animals raised on OP50, ingestion of Red bacteria post-reproductively, regardless of ingestion of Orange bacteria at any other life stage, results in a relatively normal lifespan; not shortened (Supplementary Fig. 7). Moreover, animals that eat the Red bacteria post- reproductively have an extended lifespan, which is further enhanced if the food is introduced after 33 development. Taken together, our studies reveal that expanding the available foods for C. elegans is a powerful tool to study the impact of food on lifespan and healthspan. Future studies to integrate genetic analyses to define new gene-diet pairs 6,59 and gene-environment interactions in general, will be of significant interest. 34 DISCUSSION C. elegans is a well-established model to study diet and aging. Here we augment that model by introducing a comprehensive phenotypic analysis of C. elegans fed E. coli and three laboratory bacteria originally isolated from contaminated plates: Methylobacterium, Xanthomonas, and Sphingomonas. Interestingly, these microbes have been found in the normal C. elegans environments 16 and when compared to the three most common E. coli strains OP50, HT115, HB101, our study is a critical tool to aid in our understanding of how these foods influence physiological and transcriptomic responses over the lifespan 11,16 . Previous works have identified how C. elegans react to the three standard E. coli foods provided in the lab 5,8,59-63 and yet, to our knowledge, a “head-to-head" comparison of the age-related and healthspan-relevant outcomes that result from feeding these bacteria, is lacking. Because of this, we decided to investigate the effects of bacterial diet on both physiological attributes and transcriptional signatures of C. elegans raised on bacterial species for thirty generations, to avoid acute stress responses to the food. Recent studies have shown that bacterial diet can alter transcriptional responses in C. elegans 64,65 . Our work supports this observation with three new menu options for C. elegans culture. Although each of these foods evokes a unique transcriptional signature (Fig. 2), several specific classes of genes are shared among multiple, or even all of the bacterial diets. Our RNAseq analysis was limited to gene expression changes observed as animals enter adulthood, prior to reproduction. Given the extent of changes in reproductive capacity, health (movement and fat), and aging, future work to examine how gene expression is altered on each food over the lifespan will be of great interest. Moreover, a fine-tuned analysis of transcription at each development stage will be informative based on our observation that different foods can accelerate the transition of animals across specific developmental stages. Nevertheless, it is clear that the standard laboratory E. coli strains and our three new bacterial foods induce a 35 transcriptional response of phenotypically relevant genes. Knowing that even different strains of E. coli can differentially impact physiology and gene expression, we acknowledge that our findings may not be generalizable for all Methylobacterium, Xanthomonas, and Sphingomonas and are potentially strain dependent. Bacteria serve as a live food source for C. elegans both in their natural environment and laboratory setting. Prior studies have shown animals pumping at similar rates in the presence of bacteria they can and cannot eat 12,66 . The rates of pharyngeal pumping were not significantly different on any of the bacteria in L4 stage and day 1 adult animals (Fig. 4o and Supplementary Fig. 4a). Because pharyngeal pumping may not be the best way to measure overall consumption of food, we employed a food intake assay to measure the quantity of food ingested by worms fed each bacterial diet. These data demonstrated that worms were eating at similar rates on the bacterial diets between the L4 larva and day 4 adult stages (Fig. 4p). We also found that when examining the bacterial load in day 1 adults, it appears that all bacterial diets are able to be broken down and digested by the worms due to the lack of colony growth on LB plates after the lysis of individual worms (Supplementary Fig. 4h). In support of this data, we also observed faster developmental timing (Fig. 3) and large-sized broods (Fig. 5a), indicating that each bacterium is providing sufficient nutrition to the worms feeding on them. Nevertheless, based on the similarities of animals eating Red and animals undergoing dietary restriction, it remains possible that this bacterium allows ad libitum ingestion with the physiological benefits of reduced eating, which is further supported by the Red bacteria causing worms to have the highest proportion of dietary restriction-related genes deregulated in comparison to the other bacterial diets (Figs. 6h-i). Clearly, changing food sources can potently impact multiple phenotypic attributes and several of these aging-relevant phenotypes are interrelated. For example, reproduction, fat, and stress resistance are intrinsically tied to overall life expectancy 9 . In support of this model, the Red and 36 Yellow bacteria, which reduce overall reproduction (Fig. 5a), indeed increase lifespan (Fig. 6). However, our study reveals that this relationship is more complex as these two foods have different effects on lipid storage and Yellow evokes an age-dependent somatic depletion of fat (Asdf) response (Fig. 4), which is a phenotype observed in animals exposed to pathogens 36,37 . It may be that C. elegans perceive the ingested Yellow bacteria as a pathogen, however the increased lifespan that follows on this diet, suggests this food is health-promoting. The Yellow bacteria also does not produce a lawn avoidance phenotype at L4 or day 1 of adulthood (Figs. 8c-d), which is a common phenotype that occurs within hours of introduction of worms to pathogenic bacteria. Additionally, these data are supported by the RNAseq data analysis of immune system-related genes, which shows that the Yellow bacteria induces a proportionally smaller number of gene expression changes compared to the other bacterial diets used in this study (Figs. 8e-f). Lastly, genetic and environmental mechanisms that delay developmental timing have been tied to increased longevity in adulthood 40,67,68 . Each of the bacterial diets we tested results in faster development into a reproductive adult (Fig. 3), but only HB101, HT115, Red, and Yellow increase lifespan, while Orange decreases lifespan. Taken together our study reveals that each bacterial food can exert a specific life history changing response, but also questions previously established models of aging. Many aspects of the bacterial communities, besides that of their nutritional composition, have been shown to influence multiple attributes in C. elegans, including food preference, feeding rates, brood size, and lifespan 69,70 . We decided to explore the food preference aspect by carrying out a food choice assay that contained all six of our bacterial diets. We hypothesized that bacterial diets shown to extend lifespan, like Red and Yellow (Figs. 6c-d), would have higher levels of dwelling worms while the Orange bacteria would have fewer dwelling worms due to the shortened lifespan (Fig. 6e). Interestingly, in the food choice assay, we saw that worms are less often found on the Red bacteria and more often found on the other bacteria. Finding worms more often on both 37 Yellow and Orange was somewhat surprising due to the Asdf phenotype at day 3 of adulthood in Yellow-raised worms (Figs. 4m-n) and the shortened lifespan on Orange (Fig. 6e). Both of these phenotypes are consistent with phenotypes when worms are exposed to pathogens. However, the pathogenic effect of specific bacteria is often associated with intestinal colonization 71-73 , which we did not observe (Supplementary Fig. 4h). Moreover, the increased lifespan of animals fed the Yellow bacteria disagrees with the idea that the Yellow bacteria is pathogenic. Similarly, the lack of significant bacterial lawn avoidance at day 1 of adulthood when animals are reared on Orange bacteria suggests the Orange diet isn’t perceived as detrimental. Nevertheless, these results motivate future experimentation to understand why worms are found less frequently on a longevity-promoting bacterium and more frequently on a bacterium that decreases lifespan. Inspired by previously described dietary interventions that are capable of extending lifespan in multiple species 47,74,75 , we wanted to investigate whether, instead of keeping animals on one bacterial diet their entire life, altering food exposure at critical life stages could affect overall lifespan. We asked whether these foods could alter lifespan without any previous generational exposure. Strikingly, acute exposure to these bacterial foods was capable of altering lifespan (Fig. 9), but intriguingly the magnitude of the response differed from animals that have been on these foods for 30+ generations (Fig. 6). Furthermore, our study revealed that the bacteria fed during the post-reproductive period exerted the strongest influence over the lifespan of the cohort. This result, however, is potentially confounded by the amount of time (longitudinally) that is spent on each bacteria. Regardless, our study reveals that any potential early-life (development and reproductive span) exposure to foods that normally shorten lifespan can be mitigated by eating a lifespan-promoting food option. This result is reminiscent of previous studies showing the mortality rate of dietary restricted (DR)-treated animals is accelerated when switched to an ad libitum diet while mortality rate is reduced when ad libitum-fed animals are switched to a DR diet 76 . In our case, calories are perhaps not different on Red or Orange, but rather the overall nutritional 38 composition of the bacteria is different. Given that in our model animals are chronically exposed to as much food as they can eat throughout their life, we posit that if a similar human diet were discovered that this treatment protocol would be more accessible as food restriction is difficult, socially and psychologically 77-79 . Ultimately, this study reveals the impact that diet can have on both physiology and the transcriptomics of animals that thrive on them. These alterations caused by differential bacterial diet exposure can not only be seen at the surface level in the diverse phenotypes that present themselves, but also at the genetic level, causing fluctuations in gene expression important for multiple physiological processes including development, fat content, reproduction, healthspan, and lifespan (Table 1). These discoveries in the worm can aid in understanding how dietary exposure influences different phenotypes and continuing work in the effects of bacterial diet on overall health and aging will unequivocally contribute to more personalized diets to promote healthier and longer lives in individuals. Taken together, our study expands the menu of bacterial diets available to researchers in the laboratory, identifies bacteria with the ability to drive unique physiological outcomes, and provides a food quality approach to better understand the complexities of gene-diet interactions for health over the lifespan. 39 MATERIALS & METHODS C. elegans strains and maintenance C. elegans were raised on 6 cm nematode growth media (NGM) plates supplemented with streptomycin and seeded with each bacterial diet. For experiments, nematode growth media plates without streptomycin were seeded with each bacterium at the optical density of 0.8 A600. All worm strains were grown at 20°C and unstarved for at least three generations before being used. Strains used in this study were N2 Bristol (wild type) and GR1395 (mgIs49[mlt-10p::gfp- pest, ttx-3::gfp]IV]). Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). E. coli strains used: OP50, HT115(DE3), HB101. Red, Yellow, and Orange bacteria were isolated from stock plates in the laboratory and selected for with antibiotics before inoculating. Red, Orange, and Yellow were sequenced using the 16S primer pair 337F (GACTCCTACGGGAGGCWGCAG) and 805R (GACTACCAGGGTATCTAATC) and identified using the blastn suite on the NCBI website. Bacterial Growth Curves Three test tubes with 5mL of LB without antibiotics were inoculated with each bacterium before being place in 37°C (except for Yellow, which was placed at 26°C). Optical density measurements were taken in duplicate every hour for 12 hours. The final measurement was taken 24 hours later. Optical density curves were made by averaging the readings together after performing the experiment in biological triplicate. The antibiotic versus no antibiotic growth curves were conducted in a similar manner. Three test tubes with 5mL of LB with and without appropriate antibiotics (OP50/HB101 with streptomycin and HT115/Red/Orange/Yellow with ampicillin) were inoculated with each bacterium before 40 being place in 37°C (except for Yellow, which was placed at 26°C). Optical density measurements were taken in duplicate at 12 hours and 24 hours after inoculation. Optical density curves were made by averaging the readings together after performing the experiment in biological triplicate. Metabolite Kits Bacteria was grown overnight in LB liquid culture with corresponding antibiotics. The next day, cultures were collected at the log phase of growth, spun down, and bacteria was washed in water three times before being spun down and frozen at -80°C until use. Red bacteria samples were seeded onto LB plates and then allowed to grow overnight before scraping off and collecting. Bacteria were homogenized based on metabolite kit instructions and corresponding metabolites were measured via Bio Vison kit instructions. Both glycogen and glucose measurements were obtained from the Glycogen Colorimetric/Fluorometric Assay Kit (K646) and the triglyceride and glycerol measurements were obtained from the Triglyceride Quantification Colorimetric/Fluorometric Kit (K622). Bomb calorimetry Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, seeded onto LB plates at 0.8 optical density, and allowed to grow overnight. Bacteria was then collected with autoclaved water from the seeded plates and spun down to remove excess water. Samples were frozen at -80°C until being sent in triplicate to the Department of Nutrition Sciences at the University of Alabama at Birmingham for sample drying and bomb calorimetry. Water content measurements were also obtained after measuring wet weight and dry weight of the samples. 41 RNA-sequencing Wild type worms grown on each food for at least 30 generations were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped on NGM plates seeded with 0.8 OD of each bacterial diet. 48 hours post drop, L4 animals were washed 3 times with M9 buffer and frozen in TRI reagent at -80°C until use. Animals were homogenized and RNA extraction was performed via the Zymo Direct-zol RNA Miniprep kit (Cat. #R2052). Qubit™ RNA BR Assay Kit was used to determine RNA concentration. The RNA samples were sequenced and read counts were reported by Novogene. Read counts were then used for differential expression (DE) analysis using the R package DESeq2 created using R version 3.5.2. Statistically significant genes were chosen based on the adjust p-values that were calculated with the DESeq2 package. Genes were selected if their p-value<0.01. Microscopy All images in this study were acquired using ZEN software and Zeiss Axio Imager. Worm area comparisons were imaged at 10x magnification (L2-L3 and L4 stage worms) and 5x magnification (day 1 adult and day 2 adult) with the DIC filter. Worm areas were measured in ImageJ using the polygon tool. For GFP reporter strains, worms were mounted in M9 with 10mM levamisole and imaged with DIC and GFP filters. For staining of bacteria, Biotium BactoView TM Live Fluorescent Stain was used to visualize the DNA inside the cells. Developmental timing by mlt-10::gfp GR1395 worms grown on each bacterial food for at least 20 generations were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. One 24-well plate per food source, each containing a single L1 worm, were visualized by fluorescence microscopy every hour for 55 hours. Each hour worms were scored as green (molting) or non-green (not 42 molting). Wells without worms, wells with two worms and worms that crawled to the side of the plate were censored. Lifespan analysis Worm strains on each bacterial diet were egg prepped to generate a synchronous L1 population and dropped on the corresponding food. Worms were kept at 20°C and moved every other day during progeny output (day 2, day 4, day 6, day 8). Every diet was moved at the same time. Worms were scored daily for survival by gentle prodding with a platinum wire. Animals that burst or crawled to the side of the plate were censored and discarded from this assay. Reproduction assay Wild type worms grown on each bacteria for at least 30 generations were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped on NGM plates seeded with each bacterial diet (OD 0.8). L4s hermaphrodites were singled 48 hours later onto individual plates and moved every 12 hours until egg laying ceased. Progeny were counted 48 hours after the singled hermaphrodite was moved to another plate. Progeny were removed from the plate during counting in order to ensure accuracy. Unfertilized oocytes were counted 24 hours after the singled hermaphrodite was moved to another plate. Mated reproduction Male wild type populations were raised for at least 10 generations before mating to wild type hermaphrodite populations grown for 30+ generations on each bacteria. Wild type males raised on Red and OP50 were egg prepped at the same time as wild type hermaphrodite populations raised on Red and OP50 and allowed to hatch overnight in M9 to create synchronous L1 populations. The next day, L1s were dropped onto each corresponding diet and allowed to grow 43 for 48 hours into L4s. L4 hermaphrodites were singled onto a plate with 30mL of OP50 or Red bacteria seeded in the center of the NGM plate. A single male was placed with each hermaphrodite and allowed to mate for 24 hours before the hermaphrodite was moved to a new plate and allowed to egg lay. Hermaphrodites were moved every day to a new plate until egg laying ceases. Progeny were counted 48 hours after moving the hermaphrodite and plates were checked for ~50% males to ensure the hermaphrodite was mated. Progeny were removed from the plate during counting in order to ensure accuracy. Nile Red Staining Worms were grown on each bacterial diet for 48 or 72 hours, to the L4 stage or day 1 adult stage respectively and washed with 1x phosphate-buffered saline with Tween detergent (PBST) wash buffer. Worms were then rocked for 3 minutes in 40% isopropyl alcohol before being pelleted and treated with Nile Red in 40% isopropyl alcohol for 2 hours. Worms were pelleted after 2 hours and washed in PBST for 30 minutes before being imaged at 10x magnification (L4) or 5x magnification (day 1 adult) with DIC and GFP filters on the Zeiss Axio Imager. Fluorescence is measured via corrected total cell fluorescence (CTCF) via ImageJ and Microsoft Excel. CTCF = Worm Integrated Density-(Area of selected cell X Mean fluorescence of background readings) and normalized to the control of OP50-reared worms. Oil Red O Staining Worms were grown on each bacterial diet for 48 or 120 hours, to the L4 stage or day 3 adult stage respectively and washed with PBST. Worms were then rocked for 3 minutes in 40% isopropyl alcohol before being pelleted and treated with Oil Red O in diH2O for 2 hours. Worms were pelleted after 2 hours and washed in PBST for 30 minutes before being imaged at 10x magnification (L4) or 5x magnification (day 3 adult) with the DIC filter on the Zeiss Axio Imager Erc color camera. 44 Asdf Quantification ORO-stained worms were placed on glass slides and a coverslip was placed over the sample. Worms were scored, and images were taken using a Zeiss microscope at 10× magnification. Fat levels of worms were placed into 3 categories: non-Asdf, intermediate, and Asdf. Non-Asdf worms display no loss of fat and are stained a dark red throughout most of the body (somatic and germ cells). Intermediate worms display significant fat loss from the somatic tissues, with portions of the intestine being clear, but ORO-stained fat deposits are still visible (somatic < germ cells). Asdf worms have had most, if not all, observable somatic fat deposits depleted (germ cells only). Pharyngeal pumping assays Wild type worms grown on each bacterial diet for at least 30 generations were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped on NGM plates seeded with a bacterial diet (OP50, HT115, HB101, Red, Orange or Yellow). Before imaging, worms were singled onto plates seeded with a bacterial diet 1-2 hours before videoing pumping. For L4 pumping analysis, 10-12 worms were imaged 48 hours later via Movie Recorder at 8ms exposure using the ZEN 2 software at 10X magnification (Zeiss Axio Imager). For day 1 adult pumping analysis, 10-12 worms were imaged 72 hours after dropping L1s. Food intake Food intake experiments were adapted from Gomez-Amaro et al. 2015 80 . Food intake was assessed in NGM liquid media without streptomycin in flat-bottom, optically clear 96-well plates with 150 mL total volume. Plates contained 10-40 worms per well. Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the 45 log phase, seeded onto NGM plates at 0.8 optical density, and allowed to grow overnight. Bacteria was then collected with autoclaved water from the seeded plates and spun down to wash and resuspended in water. Age-synchronized nematodes were seeded as L1 larvae and grown at 20°. Plates were sealed with parafilm to prevent evaporation and rocked continuously to prevent drowning of the nematodes. 5-fluoro-2’-deoxyuridine (FUDR) was added 48 hours after seeding at a final concentration of 0.12 mM. OD600 of each well was measured using a plate reader every 24 hours starting at L4 stage and ending at Day 4 of adulthood (144 hours after dropping L1s). The fraction of animals alive was scored microscopically at Day 4 of adulthood. Food intake per worms was calculated as bacterial clearance divided by worm number in well. Measurements were then normalized to the L4 to Day 1 Adult clearance rate for each bacterial diet. M9 thrashing assays Wild type worms grown on each bacterial diet for at least 30 generations were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped on NGM plates seeded with each bacterial diet (OD 0.8). Worms were grown on each bacterial diet until the L4, day 1 adult, day 3 adult, day 8 adult or day 11 adult stage. Worms were then moved to an unseeded NGM plate to remove bacteria from the cuticle of the worms. After an hour, worms were washed with M9 and dropped in 5mL M9 drops onto a fresh NGM plate. After one minute, 10-12 worms were imaged via Movie Recorder at 50ms exposure using the ZEN 2 software (Zeiss Axio Imager). Food choice assays Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, 30 mL of each bacterium was seeded onto NGM plates with no antibiotics at 0.8 optical density, and allowed to grow overnight. In the 6 food 46 choice assays, all bacteria were seeded 2 cm from the center point on a 6 cm plate. For the 6 food choice assay, seeding order was OP50, Red, HT115, Orange, HB101, and then Yellow. Once food choice assay plates were seeded and allowed to grow overnight, OP50-reared worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped into the center of the NGM plate and counted. The plate was then checked at 1, 24 and 48 hours to observe the location of worms. If worms were found on the bacterial lawn, then those worms were counted as on that food. Worms found outside bacterial lawns were counted as not on food. The proportion of worms found on each food or off of food was then calculated and graphed. Each assay was done in biological triplicate with technical triplicates for a total of nine plates. Bacterial load assay Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, seeded onto NGM plates at 0.8 optical density, and allowed to grow overnight. Worms reared on each bacterial food were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped onto NGM plates with each bacterial diet. Worms were allowed to grow until the day 1 adult stage (72 hours post-drop). Worms were then washed with water and then dropped on an unseeded NGM plate. Worms were allowed to crawl for an hour before washing again and dropping onto a new unseeded NGM plate. Worms were allowed to crawl for an hour before lysing 24 individual worms in 10 mL of worm lysis buffer. 5 mL of the supernatant was then seeded onto LB plates and allowed to grow 48 hours at 37°. Plates were checked at 24 and 48 hours for growth. This was done in duplicate for a total of 48 individual worms per bacterial diet. 47 Lawn avoidance assay Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, seeded onto NGM plates at 0.8 optical density, and allowed to grow overnight. Worms reared on each bacterial food were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped onto NGM plates with each bacterial diet and counted. Plates were checked 48 hours later at the L4 stage and the number of worms on and off food were counted. This was repeated at the day 1 adult stage 24 hours later as well. The proportion of worms was calculated and graphed using Graphpad Prism 8. Statistics and Reproducibility Data are presented as mean ± SEM. Comparisons and significance were analyzed in Graphpad Prism 8. Comparisons between more than two groups were done using ANOVA. For multiple comparisons, Tukey’s multiple comparison test was used and p-values are *p<0.05 **p<0.01 *** p<0.001 ****<0.0001. Lifespan comparisons were done with Log-rank test. Sample size and replicate number for each experiment can be found in figures and corresponding figure legends. This information is also in the experimental methods. Exact values for graphs found in the main figures can be found in Supplementary Data 5. DATA AVAILABILITY RNA-sequencing data are deposited into the Gene Expression Omnibus (accession no. GSE152794). All other relevant data is available upon request from the corresponding author. 48 TABLES Table 1. Bacterial diet exposure leads to changes in C. elegans physiology. OP50 HT115 HB101 Red Orange Yellow Development timing to adult a 55 hours ↓ ↓ ↓ ↓ ↓ Lifespan (mean) b,* 23 days ↑↑↑ ↑ ↑↑↑ ↓↓↓ ↑↑↑ Fat Fat Content (L4) c,* control ↑↑↑↑ ↓↓↓↓ ↓↓↓↓ ↑↑ ↓↓↓↓ Fat Distribution (Day 3 adult) d control n.s. n.s. n.s. n.s. Asdf Reproduction Output e,* 295±8 n.s. n.s. ↓↓↓ n.s. ↓ Period Length f 5 days n.s. 4 days 3 days n.s. n.s. Unfertilized oocytes g,* 172±18 ↓↓↓↓ ↓↓↓↓ ↓↓↓↓ ↓↓↓↓ n.s. Pharyngeal Pumping h,* L4 stage 228 ppm n.s. n.s. n.s. n.s. n.s. Day 1 adult 271 ppm n.s. n.s. n.s. n.s. n.s. Food intake i,* development control n.s. n.s. n.s. n.s. n.s. adulthood control n.s. n.s. n.s. n.s. n.s. Movement (Thrashing) j,* L4 stage 95 ↑↑↑↑ ↑↑ ↑↑↑↑ n.s. ↑↑↑↑ Day 1 adult 94 ↑↑↑↑ ↑↑↑↑ ↑↑↑↑ n.s. ↑↑↑↑ Day 3 adult 103 ↓ n.s. n.s. ↓ n.s. Day 8 adult 78 n.s. n.s. ↓↓↓↓ n.s. n.s. Day 11 adult 47 n.s. n.s. ↓↓↓↓ ↓ n.s. Food preference Choice k control n.s. n.s. ↓ n.s. n.s. Avoidance L4 stage l,* 2% n.s. n.s. n.s. ↑↑↑↑ n.s. Avoidance Day 1 adult m,* 14% n.s. n.s. n.s. n.s. n.s. a. C. elegans developmental timing to adulthood as measured using the mlt-10 reporter strain: mgIs49[mlt-10p::gfp-pest, ttx-3::gfp]IV]. b. Wild type C. elegans mean lifespan on each bacterial diet. HT115, Red and Yellow extend lifespan while Orange decreases lifespan. c. Fat content of wild type C. elegans measured by Nile Red fluorescent stain at the L4 stage of development. d. Fat distribution of C. elegans raised on each bacterial food at day 3 of adulthood. The Yellow-reared worms exhibit an age-somatic depletion of fat (Asdf) phenotype. e. Wild type C. elegans reproductive output of viable progeny on each bacterial diet. Output of progeny is decreased in worms on the Red and Yellow foods. f. Number of days wild type C. elegans lay viable progeny. g. Number of unfertilized oocytes laid by worms on each bacterial diet. This number is significantly reduced in worms raised on all diets compared to OP50. h. Pumping rate of wild type C. elegans on each bacterial food at the L4 and day 1 adult stage showed no significant difference from OP50-reared worms. i. Food intake of wild type worms from L4 to day 4 of adulthood was not significantly different when raised on any of the bacterial diets. j. Thrashing rate measurements at different stages of development in wild type C. elegans. Each stage was chosen due to its relation to other phenotypes observed. k. When presented with all bacterial diets, wild type worms are less often found dwelling on the Red bacteria from L1-L4. l. Wild type worms at the L4 stage avoid the Orange bacterial lawn at a higher rate than worms on other bacterial diets. m. Wild type C. elegans at day 1 of adulthood have similar avoidance of the bacterial lawn despite the bacterial diet that is present. *Arrows represent significance. Up arrows are equivalent to increased/longer while down arrows correspond to decreased/shorter. The number of arrows relates to the p-value: 1 arrow, p<0.05; two arrows, p<0.01; three arrows, p<0.001; four arrows, p<0.0001; n.s., not significant. 49 FIGURES Figure 1. Characterization of bacterial diets. (a) Bacterial streak plate with the six bacterial diets fed to C. elegans. (b) Growth curves of the bacteria with antibiotics, with optical density measurements every hour for 12 hours and another measurement taken at 24 hours. (c) Principal component analysis (PCA) of metabolite concentrations in the different bacterial diets. Bacteria were collected during the log phase of growth for bomb calorimetry (water content) and metabolite kits. 50 Figure 2. Gene expression analysis of L4 C. elegans on each bacterial food source after thirty generations. (a) Volcano plot of all differentially expressed genes in each food relative to OP50-reared worms. All genes considered to be significant have a p-value <0.01. (b-f) Volcano plots of each individual food with all significant genes that are differentially expressed. (g) Venn diagram of all significant genes. Shows the number of genes shared between two, three, four and five of the bacterial diets, along with the number of genes unique to each bacterium. (h) Heat map contains genes that are shared between all five bacterial diets. Gene Ontology (GO) terms for uniquely-expressed genes are noted below and split between genes that were up-regulated and 51 down-regulated. More information for the GO-term enrichment analysis of RNAseq data in larval stage 4 animals can be found in Supplementary Tables 1 and 2. 52 Figure 3. Developmental timing of C. elegans is dependent upon bacterial diet. (a) The molting reporter strain mgls49[mlt-10p::gfp-pest ttx-3p::gfp] IV allows for the visualization of 53 molting between larval stages of development into adulthood. The time between peaks represents the period at each larval stage, while time between troughs represent time spent molting. Analysis was carried out after this strain was raised on each bacterial food for twenty generations. (b-f) Developmental timing of worms raised on each food relative to OP50. All diets showed accelerated development into adulthood when compared to worms raised on OP50. (g) Volcano plot showing how many of the differentially expressed genes are related to development. The number of genes on each bacterial food are labeled in the legend. All genes have a p-value <0.01. (h) Heat map of development genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. 54 Figure 4. Fat content and distribution vary depending on the food C. elegans are exposed to. (a-f) Nile Red staining of L4 worms, scale bar 50 μm. HT115 (b) and Orange (e) have significantly higher fat content while HB101 (c), Red (d) and Yellow (f) have lower fat contents. (g) Quantification of Nile Red staining with comparisons made to the OP50 control. GFP fluorescence is measured and then normalized to area before being normalized to OP50. Statistical 55 comparisons by Tukey’s multiple comparison test. ***, p<0.001; ****, p<0.0001. (h-m) Oil Red O lipid staining in day 3 adult C. elegans. OP50, HT115, HB101, Red, and Orange show similar lipid distribution of fat throughout the animal (representative image, scale bar 50 μm). (m) Yellow- reared worms display age-dependent somatic depletion of fat (Asdf) with loss of fat in the intestine while fat is retained in the germline (representative image). (n) Quantification of lipid distribution across tissues. (o) Pumping in day 1 adult worms is not significantly different between any of the bacterial foods and OP50. (p) Food intake of C. elegans raised on each bacterial food in liquid nematode growth media. (q) Volcano plot of differentially expressed metabolism genes in all bacterial diets compared to OP50. All genes have a p-value <0.01. (r) Heat map of metabolism genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. All studies were performed in biological triplicate. 56 Figure 5. C. elegans raised on Red and Yellow have decreased reproductive output. (a) Hermaphrodites have reduced brood size when fed Red or Yellow bacteria. (b) Reproductive timing is also altered, revealing differences between all bacterial diets and OP50 in total output each day of their reproductive span. HB101 and Red halt output of viable progeny before OP50, 57 HT115, Orange and Yellow. (c) OP50 and Red hermaphrodites mated with OP50 or Red males yield similar number of progeny. (d) Yellow has similar number of unfertilized oocytes compared to OP50. HT115, HB101, Red and Orange have significantly fewer unfertilized oocytes. Statistical comparisons by Tukey’s multiple comparison test. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. (e) Volcano plot of differentially expressed reproduction genes on all bacterial foods compared to OP50. The number of significant genes have a p-value of <0.01. (f) Heat map of reproduction genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. 58 Figure 6. Lifespan of C. elegans on each bacterial diet. (a-e) Lifespan comparisons of OP50 versus each bacterial diet. HB101 (a) had a very small significant difference in lifespan compared to OP50. HT115 (b), Red (c) and Yellow (d) worms had increased lifespans and Orange (e) worms had a shorter lifespan. Lifespan comparisons between OP50-reared worms and the other diets 59 made with Log-rank test. For lifespan quartile comparisons, refer to Supplementary Data 2. OP50 n=155, HT115 n=51, HB101 n=53, Red n=113, Orange n=122, Yellow n=105. (f) Volcano plot of differentially expressed genes related to lifespan and stress on all bacterial foods compared to OP50. The number of significant genes have a p-value of <0.01. (g) Heat map of lifespan and stress-related genes that are significant with a p-value <0.01 in at least one of the five bacterial diets. (h) Volcano plot of differentially expressed genes related to dietary restriction on all bacterial foods compared to OP50. The number of significant genes have a p-value of <0.01. (i) Heat map of dietary restriction-related genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. 60 Figure 7. C. elegans have a food-dependent decline in thrashing with age. (a-e) The rate of thrashing in worms raised on each bacterial diet at different developmental stages. All comparisons were made to OP50-reared worms. (a) All foods except for Orange showed a significantly higher thrashing rate at the L4 stage. (b) All foods except for Orange showed a significantly higher thrashing rate at the day 1 adult stage. (c) HT115 and Orange thrashed 61 significantly less than OP50 while the other bacterial diets were similar in thrashing rate at the day 3 adult stage. (d) Red and Orange had significantly lower thrashing rates compared to OP50 and all other bacterial foods at the day 8 adult stage. (e) Red, Orange, and Yellow had lower thrashing rates at day 11 adult when compared to OP50. Statistical comparisons by Tukey’s multiple comparison test. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. (f) Average rate of thrashing over time on each bacterial diet. All studies performed in biological triplicate; refer to Supplementary Data 3 for n for each comparison. (g) Volcano plot of differentially expressed locomotion genes on all bacterial foods compared to OP50. The number of significant genes have a p-value of <0.01. (h) Heat map of locomotion genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. 62 Figure 8. C. elegans food choice. (a) Schematic of the food choice assay. (b) Food choice of C. elegans raised on OP50 when presented with all six bacterial foods as a representative proportion of worms found on each food; refer to Supplementary Data 5 for individual data points. (c) Proportion of worms on and off of the bacterial lawn at the L4 stage of development. n=200/replicate for a total of 5 replicates. (d) Proportion of worms on and off of the bacterial lawn 63 at day 1 of adulthood. n=200/replicate for a total of 5 replicates. (e) Volcano plot of differentially expressed immune-related genes on all bacterial foods compared to OP50. The number of significant genes have a p-value of <0.01. (f) Heat map of immune-related genes that are significant with a p-value <0.01 in at least one of the five bacterial diets, relative to the OP50 bacterial diet. 64 Figure 9. The introduction of the Red bacteria at the post-reproductive stage in C. elegans extends lifespan. (a) Schematic of the bacterial diet as a nutraceutical experiment. Food 0 was OP50 for all worms. After synchronization and allowing L1s to hatch overnight, L1s were dropped on bacterial food 1 and moved at L4 to bacterial food 2 before being moved to bacterial food 3 at 65 day 3 of adulthood. Lifespans were then measured. (b-g) Lifespan curves of C. elegans on each of the different bacterial diet combinations. Lifespan comparisons between the bacterial diet combinations and Red-only, and Orange-only were made with Log-rank test; refer to Supplementary Data 4. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 66 SUPPLEMENTAL FIGURES Supplemental Figure 1. Bacterial growth conditions and metabolite concentrations. (a) Table with optimal growth conditions for each bacterial diet. (b-g) Metabolite concentrations in each bacterium relative to OP50. The different metabolites measured were glucose (b), glycerol (c), glycogen (d), triglycerides (e) and water content (f). Statistical comparisons by Tukey’s multiple comparison test. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. All studies were performed in biological triplicate. (g) Calorie content of all bacteria were similar. 67 Supplemental Figure 2. Bacterial growth curve and images. (a) Bacterial growth curve comparison between growth in LB with and without antibiotics with no significant difference in growth rate. (b-g) Microscopic images of bacteria after being scraped off of a plate seeded at the optical density of 0.8 A600. All bacteria are observably similar or smaller in size compared to OP50 bacteria. The bacteria are (b) OP50, (c) HT115, (d) HB101, (e) Red. (f) Orange, and (g) Yellow. (b’-g’) Images of stained bacteria with BactoView TM Live Fluorescent dye that stains the DNA. Scale bar 5 μm. 68 Supplemental Figure 3. Time in each developmental stage and worm area is altered based on bacterial diet raised on. (a) Table showing differences in time to each molt along with hours spend in each stage of development. (b-e) Area measurements of worms on each bacteria relative to OP50. (b) Orange-reared worms showed a slightly smaller area than OP50 and other bacteria. (c) C. elegans on HT115 and HB101 had significantly larger areas at the L4 stage. (d) Worms 69 raised on HT115 have a larger area than worms raised on OP50 at day 1 of adulthood. HB101, Orange and Yellow worms have smaller area at day 1 of adulthood. (e) C. elegans raised on Red and Orange have larger areas at day 2 of adulthood compared to OP50 and the other diets. Statistical comparisons by Tukey’s multiple comparison test. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. All studies were performed in biological triplicate. 70 Supplemental Figure 4. L4 Pumping and Oil Red O staining. (a) L4 pumping of worms raised on each bacterial diet. There was no significant difference between worms raised on the different diets. Statistical comparisons by Tukey’s multiple comparison test. Study was performed in biological triplicate. (b-g) Oil Red O staining for lipid distribution in L4 C. elegans. Scale bar 50 μm. (h) Bacterial load experiment that counted the number of colonies that grew on an LB plate after 24 and 48 hours. n=48 individual worms per bacterial food. 71 Supplemental Figure 5. Reproductive timing is altered in C. elegans raised on the different bacterial diets. Reproductive output of individual worms (each line represents a different worm) at each day of its reproductive span. The thicker green line (a) and black line (b-f) on the graph represents the average output per day of reproduction. Worms were moved every 24 hours and progeny were counted. Each bacterial food had a reproductive peak from day 1 to day 2 of adulthood and then started producing fewer progeny per day until day 5 of adulthood. HB101 (c) and Red (d) worms have a significant decline in reproductive output by day 3 of adulthood. 72 Supplemental Figure 6. Thrashing declines with age in a food-dependent manner. C. elegans at each stage of development on each bacterial diet. Comparisons were made to day 1 of adulthood. Statistical comparisons by Tukey’s multiple comparison test. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. All studies were performed in biological triplicate. (a) OP50 thrashing starts to show a decline at day 11 of adulthood. (b) HT115 thrashing declines starting at day 3 of adulthood and continues to do so up until day 11 of adulthood. (c, e-f) HB101, Orange, and Yellow decline in thrashing rate at day 8 of adulthood and continues into day 11 of adulthood. (d) Red worms decline in thrashing from day 1 to day 11 of adulthood. Red-reared worms have the most significant rate of decline in thrashing compared to the other diets. 73 Supplemental Figure 7. Bacterial diet combinations compared to C. elegans raised on OP50. Lifespan comparisons of OP50 versus each nutraceutical diet combination. Lifespan comparisons made with Log-rank test (supplementary data 4). 74 SUPPLEMENTAL TABLES Table S1. Gene Ontology (GO) Terms for RNAseq analysis in L4 C. elegans on bacterial diets. 75 ACKNOWLEDGEMENTS H. Dalton for early work on this project including isolation of new bacterial contaminants, C-A Yen for experimental assistance, J. Gonzalez for technical assistance, A. Hammerquist for critical reading of the manuscript, A. Frand for the strain GR1395 (mgIs49[mlt-10p::gfp-pest, ttx- 3::gfp]IV]), and members of the Curran lab for thoughtful suggestions. AUTHOR CONTRIBUTIONS S.P.C. designed the study; N.L.S. performed the experiments; N.L.S. and S.P.C. analyzed data. S.P.C. wrote the draft manuscript, and N.L.S. and S.P.C. revised the final manuscript. This work was funded by the NIH R01GM109028 and R01AG058610 to S.P.C. and T32AG052374 and T32GM118289 to N.L.S. COMPETING INTERESTS The authors declare no competing interests. 76 Chapter 3: Avoidance behavior of C. elegans in the presence of Methylobacterium and its link to fatty acid biosynthesis pathways *This chapter is a part of a work in progress. Authors: Nicole L. Stuhr 1,2 , Chris D. Turner 1,2 , and Sean P. Curran 1,2,3,* Affiliations: 1 Leonard Davis School of Gerontology, University of Southern California, 3715 McClintock Ave, Los Angeles, CA, 90089 USA 2 Dornsife College of Letters, Arts, and Science, Department of Molecular and Computational Biology, University of Southern California, 1050 Childs Way, Los Angeles, CA, 90089 USA Norris Comprehensive Cancer Center, University of Southern California, 1441 Eastlake Ave, Los Angeles, CA, 90033 * Corresponding author: Sean P. Curran E-mail: spcurran@usc.edu 77 ABSTRACT Organisms utilize sophisticated neurocircuitry to aid in the detection of optimal food sources to maintain physiological homeostasis across the lifespan. Due to the variety of nutrients available in the natural environment, Caenorhabditis elegans have evolved optimal foraging techniques to distinguish between detrimental and beneficial microbes in order to thrive and survive in their habitats. Methylobacterium is a previously established nutritious and lifespan-promoting bacterial genera that drives faster development and longevity, but despite the many healthspan promoting effects, wild-type C. elegans will consistently pick any other available food to eat suggesting other factors beyond health can supersede healthy choices. Here we examine the neuronal and molecular underpinnings associated with development and longevity-promoting food choice behaviors. We employed computational approaches to integrate metabolomic and transcriptomic profiling and identified unique signatures that influence C. elegans food choice dynamics. A screen of established regulators of olfaction identified AWB and AWC neurons as critical regulators that mediate this novel diet-gene pair for developmental timing. Taken together, our work defines the neurocircuitry of complex behaviors, like healthy food selection, that can impact multiple life history traits. 78 RESULTS/DISCUSSION Behavior of wildtype C. elegans in the presence of lifespan-promoting bacteria C. elegans must adapt to a variety of microbes in their natural environment and distinguish between what food sources are beneficial and detrimental 1,2 . The necessity to interact with a multitude of microorganisms has shaped many aspects of their biology that are masked in the laboratory environment. In our previous publication, we conducted a head-to-head comparison between bacterial diets in C. elegans laboratory environment (E. coli strains OP50, HT115, and HB101) compared to bacterial diets found in the natural environment (Methylobacterium/Red, Xanthomonas/Orange, and Sphingomonas/Yellow). We found that changing the food source alone and maintaining a single host genetic background led changes in numerous physiological attributes and alterations of gene expression profiles in the worms raised on the diets 3 . We also identified possible behavioral changes associated with the different bacterial diets as measured by a food choice assay with all six bacterial diets present. We found that worms were found dwelling on all bacterial diets more often than the Red bacteria. In order to elucidate the underlying connection between diet and behavior, we decided to investigate this further with a more comprehensive approach. We conducted pairwise comparisons between the Red bacteria and other five bacterial diets. Previous studies have shown that worms are capable of choosing bacterial diets that are more beneficial for growth and sustaining life 1,2 , so we wanted to see if that was the case with these bacterial diets as well. We first synchronized wildtype worms that were grown up on the OP50 diet only (Figure 1A) and then dropped them on the five food choice pairwise comparisons – Red vs. OP50, Red vs. HT115, Red vs. HB101, Red vs. Orange, and Red vs. Yellow (Figure 1B). We found that when given the option, worms were found more often on any bacteria other than Red, as indicated by positive food choice indices. We checked this response from 6 hours post-drop 79 (L1s) and again at 24, 48, and 72 hours. By Day 1 of adulthood, all food choice indices were positive, indicating more than half of the worms dropped on the plate dwelling on the other bacterial food source over Red. We also tested to see if the food source worms were raised on was influencing this choice and found that this movement away from the Red bacteria is independent from food source worms are raised on (Figure S1 A-F). One caveat to pairwise comparison plates is that food choice is sometimes left up to chance, the plates are large and the worms may move in one direction over another. To attempt to limit variation that may occur due to this, we developed two additional assays that force the worms to interact with one food source before choosing to leave and explore the other food source (Figure 1C-E). In addition to the OP50 vs. Red pairwise comparison, we also have the OP50 Line and Red Line Plates. OP50 Line plates force the worms to first encounter the OP50 food source in the form of a line before having to cross over and move to the Red bacteria. Red Line plates are the same except worms encounter the Red bacteria first. When we measure the food choice index of worms dropped on either of these three plates, we start to see a trend. When wildtype worms face the pairwise comparison (Figure 1F), they stay on OP50 more often, which is the same of the worms dropped on the OP50 Line Plate (Figure 1G). Intriguingly, when worms are dropped on the Red Line plate, worms first go to the Red bacteria and by 24 hours later we see worms are moving off the food source in search for something else. By 72 hours, more than half of the worms are found off of the Red line and on the OP50 spot on the other side (Figure 1H). Again we tested to see if this was dependent of diet worms are raised on and saw that no matter what food source the worms are used to, they all show this avoidance response towards Red, as indicated by movement away from Red and towards OP50 in the Red Line plates (Figure S1 G-J). 80 Mutants in the chemosensory pathways display enhanced movement away from the Red bacteria Chemosensation in C. elegans has been studied immensely, demonstrating that there are volatile cues worms receive that either lead to an attractive or aversion response 4 . Previous publications have shown that worms not only are able to pick up on these cues, but learn to avoid what may be toxic in order to promote better health and survival 5 . C. elegans rely on five main neuron pairs in order to sense chemicals in their environment – AWA and AWC neurons are responsible for attraction while AWB, ASH, and ASI neurons are more involved in repulsive responses 6,7 . In order to determine if there is something being sensed by worms that is present in the Red bacteria, we decided to first test to chemosensory mutants that are involved in multiple neuronal pathways. egl-30 and egl-3 are expressed in all five chemosensory neurons, so we decided to test mutants and examine their food choice response. In both cases, when we have a loss-of-function (lf) mutation in egl-3 or a gain-of-function mutation in egl-30 we see that the movement away from the Red bacteria is enhanced when compared to wildtype worms (Figure 2A-C). This holds true in all three of the food choice assays, and this difference can be seen as early as the 24 hour timepoint (Figure S2 A-C). Due to an enhanced response in two mutants involved in multiple chemosensory pathways, we decided to take advantage of the five strains available to the C. elegans community that have ablated neuronal function in one of the five neuron pairs responsible for sensing volatile odors. When we raised these strains in OP50 and then synchronized and dropped them on the three food choice assays, we observed varying responses. We found that when AWA neurons lost functionality, meaning the attractive response provided by AWA neurons is lost, that worms are found moving towards OP50 faster than that of wildtype worms on both the OP50 vs. Red Plates (Figure 2D) and Red Line Plates (Figure 2F). Although none of the other ablation strains showed significant differences from wildtype on the OP50 vs. Red Plates or OP50 Line Plates (Figure 2D- 81 E), we still see a preference towards OP50. The Red Line Plates actually showed a much more significant responses, demonstrating that not only AWA, but also the AWB and AWC neurons are important for the behavior since all three neuronal ablations had enhanced movement away from the Red bacteria (Figure 2F). We see that instead of having to wait until Day 1 of adulthood for wildtype worms to move past the 50-50 line of food choice index, AWB ablated neurons leads to a difference at 24 hours and ASH ablated neurons shows a difference starting at 48 hours post- drop (Figure S2 F-H). In order to rule out that these mutants just have a faster movement speed, contributing to their faster response, we measured crawling speed at each timepoint (Figure 2G- H and Figure S2 D-E). We found that despite some mutants moving faster away from Red at earlier timepoints that these mutants had a slower speed than wildtype, which indicated to us that this food choice is independent from movement speed. Before testing a bunch of other mutants within the olfactory pathway, we first wanted to see if we could rescue the phenotypes observed with the egl-3lf and egl-30gf mutants. For egl-3lf mutants we decided to pan-neuronally rescue egl-3 in the egl-3lf mutant using the rgef-1p promoter. We found that expressing egl-3 pan-neuronally actually returned the food choice index back to wildtype levels (Figure 2I, Figure S2 J,L). We then tested to see if we could recapitulate the phenotype of egl-30gf by pan-neuronally expressing egl-30 in the wildtype background of C. elegans. We found that in most cases we were able to (Figure S2 I,K). These results indicate that altering functionality of chemosensory pathways is enough to change food choice preference of worms when given a choice between OP50 and Methylobacterium. ODR-1 is important for the movement away from the Red bacteria Due to the promising results indicating that the response away from the Red bacteria has to do with chemosensation, we decided to test and see if we could find any mutants involved in sensing of odorants with one of the five neuron pairs that actually made the worms stay on the Red 82 bacteria. We decided to test 16 different mutants that were anywhere from single to quintuple mutants and examined their response on the three food choice assays between OP50 and Red (Figure S3 A-C). We found that some mutants enhanced the response away from Red (which was expected since we observed this with a few neuronal ablation mutants along with egl-30gf and egl-3lf), but intriguingly, we also identified a few mutants that stayed on Red. One such mutant was odr-1lf, which was exciting because it showed the mutants staying on Red in the Red Line Plates (Figure 3C). Instead of worms moving away like wildtype, it appears that the odr-1lf mutants are fine being in the presence of the Red bacteria. To demonstrate that the loss-of- function mutation in the odr-1lf mutant was causal for this phenotype, we rescued odr-1 in the odr- 1lf mutant pan-neuronally and saw that this lack of movement away from Red was restored to wildtype and these transgenic lines led to worms moving away from the Red bacteria (Figure 3D- F). Methylobacterium contains different metabolites that are responsible for the avoidance phenotype observed in C. elegans Earlier we saw that AWA, AWB, and AWC neuronal function is important for the behavioral phenotype of worms on the Red bacteria. One interesting thing to note is the ODR-1 is actually expressed only in the AWB and AWC neurons, which have opposing functions since one neuron pair is involved in repulsion and the other attraction. The thing that they have in common is allowing for the sensing of volatile odors. The question that remains is are the odors from something that the Red bacteria is producing or is the response because of something that is in the bacteria itself. To test this, we decided to grow up OP50 and Red bacteria in an overnight culture and then spin the bacteria down and resuspend OP50 in the Red supernatant. This experiment would allow to test to see if something that is being produced by the Red bacteria (found in the supernatant of the overnight culture) is causing this avoidance response. We have similar plate set ups where it was OP50 vs. OP50 resuspended in the Red supernatant (Figure 83 4A). In all three assays we actually see that worms are no longer avoiding either of the bacteria on the plate when given a choice (Figure 4B). This led us to believe that it wasn’t something that was being produced by the Red bacteria that was causing the avoidance response but rather it was a metabolite in the bacteria itself. Previous studies have shown that even the different strains of E. coli have different levels of metabolites like carbohydrates, fats, and sugars 8,9 . In order to examine the differences between Red and OP50 (along with the other four bacterial diets since worms are only found avoiding the Red bacteria), we performed untargeted mass spectrometry to identify changes in metabolites within the diets. We found that there were quite a few changes in diets when compared to the standard OP50 diet (Figure 4C). When we compared the metabolites that were either changed to be lower or higher in the bacteria compared to OP50, we saw that there were only 8 metabolites unique to the Red bacteria (Figure 4D). Of these 8 metabolites, only 3 were highly expressed – valeric acid, palmitic acid, and sailcyclic acid. All three of these metabolites were expressed 5 or more fold higher in the Red bacteria, so we decided to see what would happen if we added these metabolites to OP50 and examined the food choice response. When we gave the worms a choice between OP50 and OP50 with one of the three metabolites, we saw that the movement away from a metabolite high in Red was only recapitulated when palmitic acid was added to the OP50 food. We know that palmitic acid is a crucial portion of the fatty acid biosynthesis pathway and it is converted into stearic acid and then oleic acid (Figure 4F). So we tested to see if this phenotype could be recapitulated with the addition of stearic acid, which indeed it was (Figure 4G). To our knowledge, this is a unique instance where worms are avoiding a food source because of the lipid content found in the food. 84 Disruption of fatty acid biosynthesis can cause enhanced avoidance of Methylobacterium Palmitic acid is a precursor to quite a few other fatty acids, and the pathway is quite extensive (Figure 5A). We decided to investigate how disruption of this pathway would influence C. elegans food choice between OP50 and the Red bacteria. Luckily, the enzymes involved in the conversion of one fatty acid to another are well studied so we know what mutants would be useful to further examine. We tested food choice and found that none of the mutants led to the response of staying on the Red bacteria like odr-1lf, however, many of the mutants had increased movement away from the Red bacteria similar to egl-3lf and egl-30gf. We saw that of the mutants tested, elo-1lf and fat-4lf had very strong responses against Red (Figure 5B-G). What is interesting is that loss- of-function of elo-1 actually would lead to an increase in the amount of palmitic acid because these mutants are not able to convert palmitic acid to stearic acid. We know that palmitic acid can cause an avoidance response, so this data supports the data where we added palmitic acid to OP50. Altogether, although a few more experiments are needed to complete this story, we believe that we have discovered a novel mechanism involving C. elegans avoidance of lipids. A recent study came out and showed that the presence of certain fatty acids could lead to neuronal rewiring, which is something that we are looking in to. We hope to identity lipids as signaling molecules involved in chemosensation and the control of avoidance or attractive responses towards food sources found in C. elegans natural environments. 85 MATERIALS & METHODS C. elegans strains and maintenance C. elegans were raised on 6 cm nematode growth media (NGM) plates supplemented with streptomycin and seeded with each bacterial diet. For experiments, nematode growth media plates without streptomycin were seeded with each bacterium at the optical density of 0.8 A600. All worm strains were grown at 20°C and unstarved for at least three generations before being used. Strains used in this study are outlined in Table 1 below. E. coli strains used: OP50, HT115(DE3), HB101. Red, Yellow, and Orange bacteria were isolated from stock plates in the laboratory and selected for with antibiotics before inoculating. Red, Orange, and Yellow were sequenced using the 16S primer pair 337F (GACTCCTACGGGAGGCWGCAG) and 805R (GACTACCAGGGTATCTAATC) and identified using the blastn suite on the NCBI website. Movement Measurements – Crawling Worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with OP50. Worms were then allowed to grow until each time point (48 h post-drop for L4s, 72 h post-drop for Day 1 Adults, 120 h post-drop for Day 3 adults and 168 h post-drop for Day 5 adults). Once worms were the required stage of development, 30-50 worms were washed off of a plate in 50 uL of M9 with a M9+triton coated P1000 tip and dropped onto an unseeded NGM plate. The M9 was allowed to dissipate, and worms roamed on the unseeded plate for 1 hour before imaging crawling. Crawling was imaged with the MBF Bioscience WormLab microscope and analysis was performed with WormLab version 2022. Worm crawling on the plate was imaged for 1 minute for each condition at 7.5 ms. Worm crawling was analyzed with the software and only worms that moved for at least 90% of the time were included in the analysis. 86 Food choice assays Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, 30uL of each bacterium was seeded onto NGM plates with no antibiotics at 0.8 optical density and allowed to grow overnight. All bacteria were seeded 2 cm from the center point on a 6 cm plate. For the line plates, a glass Pasteur pipette was bent with a flame at a 90 degree angle and then used to transfer bacteria as a line in the center of the plate. The line was 2 cm from the drop point and 2 cm from the spot of food. Once food choice assay plates were seeded and allowed to grow overnight, worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped into the center of the NGM plate and counted. The plate was then checked at 6, 24, 48, and 72 hours to observe the location of worms. If worms were found on the bacterial lawn, then those worms were counted as on that food. Worms found outside bacterial lawns were counted as not on food. The proportion of worms found on each food or off of food was then calculated and graphed. Each assay was done in biological triplicate with technical triplicates for a total of nine plates. Statistics and Reproducibility Data are presented as mean ± SEM. Comparisons and significance were analyzed in Graphpad Prism 8. Comparisons between more than two groups were done using ANOVA. For multiple comparisons, Tukey’s multiple comparison test was used and p-values are *p<0.05 **p<0.01 *** p<0.001 ****<0.0001. Lifespan comparisons were done with Log-rank test. Sample size and replicate number for each experiment can be found in figures and corresponding figure legends. This information is also in the experimental methods. Exact values for graphs found in the main figures can be found in Supplementary Data 5. 87 TABLES TABLE 1. List of C. elegans strains used in this study. Bristol N2 (wildtype) C. elegans wild isolate CE1047 egl-30(ep271) VC671 egl-3(ok979) CX4 odr-7(ky4) – AWA Ablated JN1715 pels1715[str-1p::mCasp-1 + unc-122p::GFP] -- AWB Ablated PY7502 oyls85[ceh-36p::TU#813 + ceh-36p::TU#814 + srtx-1p::GFP + unc- 122p::DsRed] -- AWC Ablated JN1713 pels7113[sra-6p::mCasp-1 + unc-122p::mCherry] -- ASH Ablated PY7505 oyls84[gpa-4p::TU#813 + gcy-27p::TU#814 + gcy-27p::GFP + unc- 122p::DsRed] -- ASI Ablated CX2065 odr-1(n1936) BX14 elo-1(wa7) BX24 fat-1(wa9) BX106 fat-6(tm331) BX110 fat-6(tm331);fat-5(tm420) BX160 fat-7(wa36) fat-5(tm420) RB1795 fat-1(ok2323) VC138 elo-1(gk48) BX107 fat-5(tm420) RB1031 fat-4(ok958) BX17 fat-4(wa14) RB969 fat-2(ok873) BX52 fat-4(wa14) fat-1(wa9) NL2105 gpa-3(pk35) odr-3(n1605) MT4810 odr-3(n2046) NL335 gpa-3(pk35) NL348 gpa-2(pk16) gpa-3(pk35) CX2205 odr-3(n2150) CX3222 odr-3(n1605) GJ7 gpa-2(pk16) gpa-3(pk35) gpa-13(pk1270); gpa-5(pk376) gpa-6(pk480) NL334 gpa-2(pk16) CB1033 che-2(e1033) CX3937 lim-4(ky403) CX5922 kyls140[str-2p::GFP + lin-15(+)];ceh-36(ky640) MT1071 egl-21(n476) SP1205 dyf-1(mn335) SP1234 dyf-2(m160) DR47 daf-11(m47) 88 FIGURES Figure 1. Behavior of C. elegans on the Red/Methylobacterium longevity-promoting diet. (A) Schematic for food choice assays and formula for food choice index. Worms were grown on OP50 for multiple generations before egg prepping and dropping synchronized L1s onto five food choice comparison plates. When more worms are on the Red bacteria they have a more negative food choice index and when more worms are found on the other diet the food choice index is positive. (B) Food choice of worms raised on OP50 when given the option of Red or another food source. Worms are found more often residing on any bacteria other than Red when given the option. (C-E) Development of OP50 vs. Red food choice assays. (C) Pairwise choice assay between OP50 and Red, (D) OP50 Line Plates, and (E) Red Line Plates. (F-H) Food choice of wildtype worms on the three food choice assays. Worms reside on OP50 more often on the OP50 vs. Red Plates (F) along with the OP50 Line Plates (G). Wildtype worms interact with the Red A. B. C. D. E. F. G. H. 89 Line and by 72 hours (Day 1 Adult) more than half the worms on the plate move away from Red to the OP50 spot. 90 Figure 2. AWA, AWB and AWC neurons mediate avoidance responses to Methylobacterium. (A-C) egl-30gf and egl-3lf mutants show an increased preference towards OP50 when compared to wildtype worms on OP50 vs. Red plates (A), OP50 Line Plates (B) and the Red Line Plates (C). Statistical significance was measured at the 72 hours post-drop timepoint. (D-F) Testing of neuronal ablation strains on the food choice assays. AWA ablated neurons lead to an increased preference towards OP50 on OP50 vs. Red Plates (D) and Red Line Plates (F) while AWB ablated and AWC ablated neurons show increased preference towards OP50 on the Red Line Plates (F). Crawling speed at the 6 hours timepoint (G) and the 72 hour timepoint (H) of the neuronal ablation strains was slightly slower in AWA and AWB strains earlier on, but speeds up by 72 hours so this food choice response is independent from movement speed. (I) Pan-neuronal rescue of egl-3lf with rgef-1p::egl-3wt resulting in the food choice on Red Line Plates returning back to wildtype levels in two separate transgenic lines. (J) Pan-neuronal expression of the egl-30gf construct in WT worms had varying results. A. B. C. D. E. F. G. H. I. J. 91 Figure 3. ODR-1 signaling is required for the avoidance response towards Methylobacterium. (A-C) Food choice of odr-1lf mutants when given the option between OP50 and the Red bacteria. On OP50 vs. Red Plates (A) and OP50 Line Plates (B) the food choice response is similar to wildtype worms. When worms are forced to encounter the Red bacteria first with the Red Line Plates (C) odr-1lf mutants stay on the Red bacteria instead of moving away like wildtype worms. (D-F) Pan-neuronal rescue of odr-1lf with rgef-1p::odr-1wt returns the food choice back to wildtype on all food choice assays including the Red Line Plate (F). F. D. E. A. B. C. 92 Figure 4. Avoidance response is due to a difference in metabolites found in Methylobacterium and not a volatile chemical produced by the bacteria. (A) Food choice assays made with OP50 and OP50 resuspended in media where Methylobacterium was grown. (B) OP50 vs. OP50 with Red Supernatant plates revealed that the avoidance response of wildtype worms towards the Red bacteria is not due to something being produced by the bacteria because the avoidance response is lost. (C) Metabolic profiling of the six bacterial diets revealed many metabolites that were either up or down relative to OP50. The Red bacteria contained 264 metabolites at higher concentrations relative to OP50, while it contained only 104 metabolites that were down. (D) Comparison of metabolites up or down in each bacteria relative to OP50 revealed 8 unique metabolites found in the Red bacteria. (E) Testing of the three metabolites that were highly expressed in the Red bacteria – valeric acid, palmitic acid, and sailcyclic acid. Food choice between OP50 and OP50 supplemented with 40mM of one of the three metabolites revealed that worms are found to highly avoid palmitic acid but not valeric or sailcyclic acid. (F) Fat biosynthesis HT115 HB101 Red Orange Yellow A. B. C. D. F. E. G 93 pathway shows that palmitic acid is converted into stearic acid normally, so we decided to test stearic acid as well and saw a similar response (G). 94 Figure 5. Disruption of the fatty acid biosynthesis pathway changes the response of worms to Methylobacterium. (A) Fatty acid biosynthesis pathway detailed with enzymes important for the conversion of one fat species to another. (B) Testing mutants in this pathway on OP50 vs. Red Plates and OP50 Line Plates (C) did not reveal a large amount of significant changes compared to wildtype worms. (D) Red Line Plates showed that quite a few of the mutants have a faster response away from the Red bacteria, especially elo-1lf mutants and fat-4lf mutants. (E-G) Significance of food choice index compared to wildtype worms on each food choice assay at each timepoint. Gray cells represent more of a preference towards OP50 than compared to wildtype worms while red cells represent more of a preference towards the Red bacteria. A. B. C. D. E. F. G. 95 SUPPLEMENTAL FIGURES Figure 1 – Supplemental. Food choice in wildtype worms is independent from the diet that the worms are raised on. (A) Schematic for raising wildtype worms on each bacterial diet for multiple generations before egg prep synchronization at the L1 stage when worms were dropped on food choice plates. Regardless of the diet worms were raised on, worms were found to prefer OP50 (B), HT115 (C), HB101 (D), Orange/Xanthomonas (E), or Yellow/Sphingomonas (F) over A. B. C. D. E. F. G. H. I. J. 96 the Red/Methylobacterium diet. (G) Schematic for raising wildtype worms on each bacterial diet for multiple generations before egg prep synchronization at the L1 stage when worms were dropped on food choice plates. Regardless of the diet worms were raised on, worms were found to prefer OP50 on OP50 vs. Red Plates (H) and OP50 Line Plates (I). Worms also made an effort to move away from the Red bacteria on the Red Line Plates (J). 97 Figure 2 – Supplemental. Functionality of chemosensory neurons is important for the Methylobacterium avoidance response. (A-C) Timepoint comparisons of wildtype, egl-30gf, and egl-3lf strains on the OP50 vs. Red Plates (A), OP50 Line Plates (B), and Red Line Plates (C). Crawling speed of neuronal ablation mutants at the 24 hours timepoint (D) and 48 hour timepoint (E). (F-H) Timepoint comparisons of wildtype and neuronal ablation mutants on OP50 vs. Red Plates (F), OP50 Line Plates (G), and Red Line Plates (H). Pan-neuronal expression of the egl-30gf construct in wildtype worms on OP50 vs. Red Plates (I) and OP50 Line Plates (K). Pan-neuronal rescue of egl-3lf with rgef-1p::egl-3wt on OP50 vs. Red Plates (J) and OP50 Line Plates (L). K. L. H. I. J. D. E. F. G. A. B. C. 98 Figure 3 – Supplemental. Chemosensory mutants show a variety of responses towards Methylobacterium, some move away while some stay. (A-C) A plethora of chemosensory mutants were tested for food choice on the three food choice assays. (A) OP50 vs. Red Plates showed quite a few mutants with an enhanced response away from Red. (B) OP50 Line Plates showed similar results, highlighting some mutants that stayed on OP50 for all the time during the assay. (C) Red Line Plates were the most informative, demonstrating that a few dye-filling defective mutants and odr-1lf stayed on Red while all other olfaction mutants had an increased movement towards OP50. (D-F) Timepoint comparisons between wildtype and odr-1lf for all three of the food choice assays. A. B. C. E. D. F. 99 Chapter 4: A dicer-related helicase opposes the age-related pathology from SKN-1 activation in ASI neurons *This chapter is a version of a manuscript Authors: Chris D. Turner 1,2* , Nicole L. Stuhr 1,2 *, Carmen M. Ramos 1,2 , Bennett T. Van Camp 1 , Sean P. Curran 1,2,# *Authors contributed equally to this work Affiliations: 1 Leonard Davis School of Gerontology, University of Southern California, Los Angeles, CA. 90089 2 Molecular and Computational Biology, University of Southern California, Los Angeles, CA. 90089 # correspondence to spcurran@usc.edu Short Title: Intestinal DRH-1 opposes neuronal SKN-1 activation Short Description: The dicer-related helicase participates in a cell non-autonomous program to mitigate age-related pathology stemming from transcriptional overload in two ASI neurons. 100 ABSTRACT Coordination of cellular responses to stress are essential for health across the lifespan. The transcription factor SKN-1 is an essential homeostat that mediates the response to environmental stress conditions and cellular dysfunction, but constitutive activation of SKN-1 drives premature aging thus revealing the importance of turning off this cytoprotective response. Here we identify how SKN-1 activation in two ciliated ASI neurons in C. elegans results in an overload in organismal transcriptional capacity that drives pleiotropic outcomes in peripheral tissues. An increase in the expression of established SKN-1 stress response and lipid metabolism gene classes of RNA in the ASI neurons, in addition to the increased expression of several classes of non-coding RNA, define a molecular signature of animals with constitutive SKN-1 activation and diminished healthspan. We reveal neddylation as a mediator of SKN-1 within intestinal cells. Moreover, RNAi-independent activity of the dicer-related DExD/H-box helicase, drh-1, in the intestine, can oppose the effects of aberrant SKN-1 transcriptional activation and delays age- dependent decline in health. Taken together, our results uncover a cell non-autonomous circuit to maintain organism-level homeostasis in response to excessive SKN-1 transcriptional activity in the sensory nervous system. 101 INTRODUCTION A central node in the response to xenobiotics and oxidative stress is the cytoprotective transcription factor SKN-1/NRF2, which binds to antioxidant response elements in the promoters of genes, including phase 2 detoxification and antioxidant synthesis enzymes, that are important for survival under a wide range of stress conditions 1-6 . SKN-1 also plays critical roles in embryonic development by specifying the early blastomere identity and formation of the pharyngeal and intestinal tissues 7-11 . Recently, SKN-1 activity has been demonstrated to coordinate cellular metabolism, particularly lipid metabolism and mobilization 12-18 . Taken together, these studies highlight the complex roles that SKN-1 plays in organismal homeostasis across the lifespan. Despite the essentiality for development and in the response to stress with age, constitutive activation of SKN-1 has been demonstrated to be pleiotropic for health 13,15,17-20 . Although the molecular basis of this health detriment remains elusive, it is clear that turning off SKN-1 activity is equally important to turning it on. SKN-1 abundance and turnover are regulated by the ubiquitin proteasome system, mediated by the CUL-4/WDR-23 ubiquitin ligase complex. Whereas the loss of skn-1 is maternal-effect embryonic lethal 7,21 , and whole-life loss of wdr-23 leads to constitutive SKN-1 activation that results in sickness, the adult-specific inactivation of wdr-23 can lead to increased lifespan 22 . Relatedly, gain-of-function (gf) mutations in skn-1 lead to enhanced resistance to acute exposure to oxidative stress early in life, but this stress resistance phenotype is lost when animals reach the post-reproductive period of life 15 . Tied to these changes in stress resistance and reproduction, these skn-1gf alleles also result in the loss of somatic lipid reserves and a significant diminishment of adult lifespan 23 that is tied to the activation of pathogen resistance responses 13 . In light of each of these discoveries, SKN-1 plays critical roles from embryogenesis through death, which requires sophisticated regulatory mechanisms. 102 Fifteen years ago, SKN-1 was shown to function in ASI neurons to mediate longevity responses to caloric restriction 24 , but multicopy transgenic reporters suggest that SKN-1 activity in non- neuronal tissues is activated in response to stress 11 . Moreover, use of transgenic reporters designed to measure SKN-1 transcriptional responses are activated in cells beyond ASI neurons that suggest that either multicopy transgenics do not accurately reflect physiological SKN-1 responses or that cell non-autonomous pathways responding to SKN-1 activation exists but have yet to be elucidated. To address these questions, we employed genome editing to tag the endogenous skn-1 locus with GFP in wild type and skn-1gf mutant animals and additionally, replaced the commonly used multicopy extrachromosomal transgenic arrays with single copy tissue-specific constructs that more accurately measure physiological responses. We discovered that SKN-1 expression, even in animals with genetically-encoded constitutive SKN-1 activation, remains detectable only in ASI neurons, which is sufficient to coordinate changes in stress and metabolic homeostatic responses. Comparing the transcriptomic landscapes in whole animals, and specifically transcriptional responses to SKN-1 activation specifically in ASI neurons, we further discovered that the pleiotropic outcomes from constitutive SKN-1 activation can be delayed by the dicer- related helicase that can temporarily mitigate the transcriptional overload in response to SKN-1 activation. 103 RESULTS SKN-1 activation phenotypes from two ASI neurons Previously, activation of SKN-1 has been demonstrated to drive the age-dependent reallocation of somatic lipids such that somatic lipids are depleted while germline lipids are maintained during reproduction that subsequently alters organismal stress resistance and reproduction 15 . In wild type animals, this phenotype is one of the earliest detectable response upon exposure to pathogenic bacteria like Pseudomonas aeruginosa 13 . Although the environment is a potent driver of this response, genetic mutations that activate SKN-1, including gain-of-function (gf) mutations in skn-1 (Figure S1A)are potent drivers of this change in lipid partitioning 18,25,26 ; the causality of this single nucleotide mutation on lipid partition was confirmed by genome editing (Figure S1B- C). To understand how constitutive SKN-1 activation drives premature aging and diminished health, we used CRISPR-Cas9 genomic editing to tag the SKN-1 locus in wild type animals and animals harboring the skn-1gf(lax188) allele; hereafter referred to as skn-1gf. We used this endogenously-tagged version of SKN-1 and SKN-1gf to identify direct transcriptional targets by ChIPseq. RNAseq analyses reveal upregulation of several gene classes including phase II detoxification, host defense factors and lipid metabolism in skn-1gf animals. Additionally, significant enrichment of promoters regulating proteostasis factors (e.g., ribosomal subunits, ubiquitin ligases, and chaperones) were recovered by ChIPseq (Figure 1A-B). These direct targets included 70 ribosomal subunits (rps-1 through rps-27 and rpl-1 through rpl-43), ubl-1, ubq- 1, lgg-2, hsp-1, hsp-3, hsp-6, and ndk-1, and several genes enriched in neurons (Figure 1B). We investigated whether the low number of recovered targets bound by SKN-1 could be explained by the expression pattern of the SKN-1gf protein. Surprisingly, and unlike animals with activated SKN-1 stemming from loss of the negative regulator wdr-23 20,22,27-33 or animals exposed to 104 exogenous toxicants 32,34-37 that have SKN-1 stabilized in the intestine, the expression of SKN- 1wt-GFP and SKN-1gf-GFP were only detectable in ASI neurons (Figure 1C,D). Previous transcriptional profiling of animals with activated SKN-1 was performed using RNA samples derived from whole animals. Since SKN-1wt and SKN-1gf protein were only detectable in ASI neurons, we enriched ASI neuronal populations and performed RNAseq. Among the gene targets specifically upregulated in ASI neurons of skn-1gf mutants (Figure 1E, Table S1) were canonical glutathione s-transferases (gst-4 and gst-30), pathogen response genes (irg-5, C55B7.3), and several neuron enriched genes (gcy-7, C50C3.19, nspd-10, and zig-2). Taken together, these data suggest that skn-1gf activity may change steady state stress homeostats in ASI neurons. Neddylation regulates intestinal SKN-1 stability SKN-1 coordinates stress adaptation in response to a variety of cellular insults 2,11-13,15,17-20,38-46 and although skn-1gf mutant animals are stress resistant, several details surrounding the molecular basis of this response remain unclear. Traditionally, when a stressor is encountered, SKN-1 protein is stabilized, translocated to the nucleus, and then mediates the transcription of genes that will mitigate the current stress condition 11,29,33 . To reconcile the difference in the robust pan-tissue transcriptional response measured in the skn-1gf mutant with the inability to detect stabilized SKN-1gf expression outside of the ASI neurons (Figure 1C-D), we measured several characteristics of SKN-1 dynamics. First, we tested the stabilization of SKN-1 in response to oxidative stress by acute exposure to hydrogen peroxide (H2O2), which resulted in the predicted accumulation of SKN-1wt-GFP and SKN-1gf-GFP in the intestine of treated animals (Figure 2A- B). Despite robust resistance to hydrogen peroxide exposure, we noted a delayed accumulation of SKN-1gf-GFP in the intestine as compared to SKN-1wt-GFP, suggesting an improved capacity to turn over the SKN-1gf protein outside of the nervous system. 105 The stabilization of SKN-1 in response to oxidative stress is linked to its turnover by the CUL4- DDB-1-WDR23 E3 ubiquitin ligase that targets SKN-1 to the ubiquitin proteosome system (UPS) for degradation 20,33,41,47 . As such, we next exposed animals expressing SKN-1wt-GFP or SKN- 1gf-GFP to wdr-23 RNAi and observed accumulation of SKN-1wt-GFP and SKN-1gf-GFP in the intestine (Figure S2A-D). This result confirms that the UPS-mediated control of SKN-1 protein is not perturbed in skn-1gf mutants, which supports previous findings that the increase in transcriptional output stemming in skn-1gf mutants is additive with loss of wdr-23 18 . Although regulation of SKN-1 by the CUL4/DDB1/WDR23 E3 ligase and the ubiquitin proteasome system is well established, the proteostasis network is tightly regulated by the coordinated actions of several pathways 48 , including the post-translation modification of proteins by the ubiquitin-like molecules NED-8/NEDD8 49 and SMO-1/SUMO 50 . We used RNAi to inhibit ubiquitinylation, neddylation, and sumoylation components of proteostasis machinery to examine their effects on SKN-1wt and SKN-1gf mutants (Figure 2C-F, Figure S2E-N). Intriguingly, RNAi targeting uba-1, the E1 ubiquitin activating enzyme (Figure 2C-D), and ned-8 (neural precursor cell expressed, developmentally down-regulated 8), the ubiquitin like modifier (Figure 2E-F), stabilized SKN-1gf- GFP but not SKN-1wt-GFP within the nucleus of intestinal cells. The preferential stabilization of SKN-1gf-GFP was not observed with RNAi targeting the sumoylation pathway (Figure S2I-N), which suggests the observed differential stabilization of SKN-1gf relative to SKN-1wt is not a generalized sensitivity to proteostatic stress, but instead a specific response to the ubiquitin and neddylation pathways (Figure 2G). SKN-1gf activity in ASI neurons drives cell non-autonomous decline with age Our finding that SKN-1gf protein is only detectable outside of the ASI neurons only when proteostasis is perturbed raised questions as to where SKN-1gf activity is actually needed to 106 establish early life stress resistance and eventual health decline if left unchecked. To generate physiologically relevant models for tissue specific manipulation of SKN-1wt and SKN-1gf activities, we developed several new strains (Figure 3A) for tissue-specific expression under established promoter elements as well as also tissue-specific degradation by tagging the endogenous skn-1 locus with an auxin-inducible degron (AID) tag for auxin-mediated degradation of SKN-1wt or SKN-1gf protein only in tissues expressing of TIR1 51 . In light of the well-established role that SKN-1 plays in cytoprotection against oxidative stress 11,52 , we tested whether the expression of skn-1gf in a single tissue was sufficient to establish resistance to hydrogen peroxide as previously documented for the skn-1gf mutant 15 . Both skn- 1B and skn-1C regulate oxidative stress resistance and longevity 11,24,53,54 ; however, because the lax188 gain-of-function mutation does not alter the SKN-1b polypeptide (Figure S1A) we focused on the skn-1C isoform. Although previously thought that SKN-1b activity in ASI was sufficient to drive oxidative stress resistance 11,24 , restricted expression of skn-1gf in ASI neurons was sufficient to recapitulate the resistance to acute exposure to hydrogen peroxide (Figure 3B), suggesting that the gain-of-function allele is active in ASI and can stimulate organism-level protection from oxidative insult. The ability of ASI restricted expression of the skn-1gf c-isoform to mediate an organism-level oxidative stress response suggests a cell non-autonomous action stemming from SKN-1 activation. Although we are only able to detect SKN-1wt-GFP and SKN-1gf-GFP in the ASI neurons, our initial isolation of the skn-1gf mutant was due to the robust activation of the gst- 4p::gfp reporter that was induced across multiple tissue types in the animal 18 . As such, we next tested whether tissue-specific expression of skn-1gf was sufficient to activate the gst-4p::gfp reporter cell non-autonomously (Figure 3C, Figure S3A-D). Although muscle-specific expression (myo-3p) of skn-1gf resulted in muscle restricted activation of gst-4p::gfp, ASI specific expression, 107 under the control of the gpa-4 promoter, induced gst-4p::gfp activation beyond the two ASI neurons, including expression in the intestine and body wall muscle. Animals with pan-neuronal expression (rgef-1p) displayed a similar pattern of gst-4p::gfp activation except that multiple neurons displayed activation of the reporter. In contrast, intestine specific expression (vha-6p) resulted in gst-4p::gfp reporter activation only in the intestine and body wall muscle. We next examined whether cell type specific expression of skn-1gf could stimulate the age- dependent somatic depletion of lipids (Asdf). Although SKN-1gf activity in ASI neurons was sufficient to recapitulate oxidative stress resistance, it was not sufficient to induce somatic lipid depletion at day 3 of adulthood (Figure 3D, Figure S3E-G); however, we did document a trend toward increased somatic lipid depletion in the population at day 5 of adulthood (Figure 3SH-K). Although we could not identify sufficiency for any single tissue for somatic lipid depletion, we next examined which tissues, if any, were necessary for somatic lipid depletion by auxin-mediated degradation. Based on previous reports with this system 51,55 we expected and observed some auxin-independent degradation effects on lipid depletion (Figure S3L). However, the addition of auxin greatly enhanced the suppression of lipid depletion in the intestine-specific TIR1 strain by 21%, while pan-neuronal and ASI-specific TIR1 strains resulted in a 40% and 16% suppression of lipid depletion, respectively (Figure 3E, Figure S3L-S, Table S2). Collectively, these results reveal that skn-1 activity in neurons are required for the phenotypes associated with constitutive SKN-1 activation and can initiate a cell non-autonomous response in peripheral tissues like the intestine, which normally removes activated SKN-1 through ubiquitin and neddylation proteostasis pathways. 108 DRH-1 activation delays healthspan decline from SKN-1 activity Our data suggest that activation of skn-1 in the ASI neurons is sufficient to drive systemic changes in oxidative stress resistance throughout the organism and is needed, at least in part, for the somatic lipid depletion that accompanies SKN-1 activation with age. To identify mediators of the peripheral response to skn-1gf activity in ASI neurons, we performed an unbiased genetic screen with ethyl methanesulfonate to recover suppressors of the activation of the gst-4p::gfp reporter outside of the nervous system in skn-1gf mutants (Figure 4A). To our surprise, we recovered a suppressor mutant in the F1 generation of the screen and confirmed the dominant nature of this allele by subsequent backcrossing into the unmutagenized parental strain. We identified insertion- deletion polymorphisms 56 linked to the dominant suppressor mutation that mapped to the center of chromosome IV (LGIV) (Figure 4B). Genome-wide sequencing (GWS) of the suppressor mutant genomic DNA revealed twenty-six missense mutations within this region (Figure S4A). Only RNAi targeting the dicer-related helicase gene, drh-1, restored the peripheral gst-4p::gfp expression observed in the parental skn-1gf mutant (Figure 4C-D, Figure S4B-G), which also suggest the drh-1gf(lax257) mutation is gain-of-function; hereafter referred to as drh-1gf. We confirmed causality and dominance of the drh-1gf allele by transgenesis (Figure 4E-F). The drh- 1 locus encodes for two predicted isoforms, DRH-1A and DRH-1B that are 1037 and 779 amino acids in length, respectively. The drh-1gf(lax257) mutation changes glycine 474 in DRH-1A and glycine 216 in DRH-1B to arginine, which are on the surface of the predicted DRH-1 protein structure in a region in the predicted C. elegans DRH-1 protein (Figure 4G) that resembles the bridging domain found in RIG-I that participates in RNA recognition 57,58 . We next examined the impact of the drh-1gf on the age-related healthspan phenotypes influenced by skn-1gf; age-dependent somatic lipid depletion (Figure 4H-I, Figure S4H-Q, Table S2), oxidative stress resistance (Figure 4J-K), and movement (Figure 4L-N, Table S3). skn-1gf drh- 1gf double mutant animals display a significant reduction of somatic lipid depletion at day 3 of 109 adulthood (Figure 4H), whereas the phenotype nears complete penetrance in skn-1gf mutants. Remarkably, the suppression of somatic lipid depletion was not maintained at day 5 of adulthood where animals harboring the drh-1gf allele were indistinguishable from age-matched skn-1gf single mutant animals (Figure 4I); thus, the effect of the drh-1gf mutation is to delay the impact of the skn-1gf allele. In addition to changing age-dependent distribution of lipids, the skn-1gf mutation has a paradoxical effect on oxidative stress resistance where skn-1gf mutant animals are more resistant to acute exposure to oxidative stress at day 1 of adulthood, as compared to WT (Figure 4J), but at day 3 of adulthood, when somatic lipid depletion is complete, skn-1gf mutant animals are much more sensitive to the same exposure of oxidant than age-matched wild type animals (Figure 4K). The drh-1gf mutation partially reversed the effects of the skn-1gf allele back to wild type (Figure 4J-K). Finally, skn-1gf mutant animals display movement defects compared to wild type animals that is characterized by reduced speed and reduced mean amplitude of body bends when crawling starting early in life (Figure 4L, Table S3) that persists throughout adulthood (Figure 4M-N). Early on, skn-1gf drh-1gf mutant animals show an intermediary movement defect that is indicative of partial suppression of the skn-1gf mutation which becomes progressively worse and more fully resembles the skn-1gf single mutants later in adulthood. Taken together, these results reveal the ability to delay the diminished health stemming from SKN-1 activation with age by an activating mutation in the dicer-related helicase, drh-1. Intestinal DRH-1 reduces transcriptional overload from SKN-1 activation Universally represented across all of the genetic mutants with enhanced SKN-1 activity is diminished health with age. Although previous work has demonstrated that the reduced healthspan is associated with the transcriptional activity of SKN-1 13 , our understanding of why constitutive transcription is debilitating requires additional detail. We first examined whether the suppression of the gst-4p::gfp transcriptional reporter outside of the nervous system by drh-1gf 110 was not an artefact of this simple transgenic reporter. We measured the expression of endogenous gst-4 transcripts and multiple phase II detoxification genes 59 (Figure 5A-G, Table S4), that are strongly induced in skn-1gf 13,18 and regulated by SKN-1 under normal conditions 60,61 and found that they were significantly reduced in the skn-1gf drh-1gf double mutant and their expression pattern opposes the age related change in the skn-1gf single mutant animals. Next, we examined how drh-1gf influences other transcriptional stress responses in the context of constitutive SKN-1gf activity. Previous work identified several RNAi conditions that induce SKN-1 activation 22,46,60,61 , including RNAi targeting RNA polymerases themselves and regulators of transcription 60 . RNAi targeting several transcriptional regulators (e.g., rpc-1, F30A10.9, and tif- 1A) results in hyperactivation of the gst-4p::gfp reporter which does not occur in the context of the drh-1gf mutation (Figure 5H, Figure S5A ). As such, drh-1gf attenuates the response to increased transcriptional load resulting from the activation of cytoprotective transcription factors like SKN-1. DRH-1 is an ortholog of the mammalian retinoic acid-inducible gene I (RIG-I)-like receptors which can detect viral double-stranded RNA (dsRNA) and promote antiviral defense. In C. elegans, drh- 1 is required for antiviral RNA interference 62-64 , but also plays a separable signaling role in the regulation of transcription immune responses to viral infection 65 . We tested RNAi targeting the CEr1 regulator of viral RNA, cerv-1, which strongly induces gst-4p::gfp in the skn-1gf mutant but not in skn-1gf drh-1gf mutants (Figure 5H, Figure S5A). Additionally, we confirmed that the reduced activation of the gst-4p::gfp reporter was not dependent on the activation of the RNAi- related machinery and as such, unlikely a result of RNAi-induced transgene silencing (Figure S5B). Taken together, these results support a role for drh-1 in the regulation of SKN-1 dependent transcriptional homeostasis. 111 In addition to stress response and cellular metabolism genes that are activated in skn-1gf mutants but repressed by drh-1gf (Figure S5C), we noted an increase in the expression of multiple transcripts enriched for double-stranded secondary structure (e.g., ncRNAs, 21U-RNA, tRNA, snRNA, snoRNA, and pseudogenes) in skn-1gf mutants but are repressed in the skn-1gf drh-1gf double mutant (Figure 5I-Q, Figure S5C). Taken together, these data reveal a new model where activated DRH-1 works to oppose the increased transcriptional load stemming from constitutive SKN-1 activation (Figure 5W, Figure S5D). Although the drh-1gf allele can temporarily counteract SKN-1 activity, if SKN-1 is not attenuated, the ability of DRH-1 to counterbalance SKN-1 activation is eventually overcome. We noted that the classes of non-coding RNAs were not significantly enriched in the genes upregulated from the ASI neuron population suggesting that the regulation of the ncRNAs occurs outside of the nervous system. As such, we were curious where skn-1gf activity was needed to suppress the transcriptional load in the skn-1gf mutant background. DRH- 1 activity in the intestine has been linked to its RNAi-independent roles in gene expression {33264285; 30640956; 31619561}. We expressed drh-1gf exclusively in the intestine, which mitigated much of the skn-1gf-induced activation of the gst-4p::gfp reporter (Figure 5R-S). We noted a significant change in the expression of multiple genes that mediate signaling across tissues mediated by the drh-1gf allele in the context of the skn-1gf background (Figure 5T-V) that might mediate the cell non-autonomous responses to activated SKN-1. Our results demonstrate that the genetic activation of the dicer-related helicase opposes the activity of SKN-1 in ASI neurons to reestablish transcriptional homeostasis at the organismal level that improves age- dependent health status (Figure 5W, Figure S5D). 112 DISCUSSION The paradoxical discovery that the activation of cytoprotective transcription factors can be the cause of diminished health over the lifespan points to the complexity of the homeostats that have evolved to ensure survival. Our work supports the notion that turning off cytoprotection when it is not needed is perhaps equally as important as our ability to turn it on when it is; a corollary to turning a faucet on and off to fill a sink with water. SKN-1 is the C. elegans ortholog of the NRF2/NFE2L2 family of transcription factors in mammals. In cancer patients, activation of NRF2/NFE2L2 in transformed cells results in chemo- and radiation therapy resistance 66-68 because the players in phase II detoxification that are normally induced by NRF2 in normal cells to protect against toxic conditions protect the cancer cells from treatment. Beyond this action of the protein products of SKN-1/NRF2, our study revealed that several non-coding RNA transcripts that are induced when SKN-1 is activated may also drive phenotypic decline with age. Although first reported a decade ago, the molecular basis of how a single amino acid substitution in SKN-1 renders this cytoprotective transcription factor constitutively active remains elusive. Our inability to detect activated SKN-1 stabilized at steady state in cells beyond the ASI neurons was suggestive that either SKN-1 is not functioning in those cell and tissue types or that the skn-1gf mutation alters the homeostatic turnover of the gain-of-function protein. Our finding that mildly impairing the ubiquitin proteasome system allows for the detection of intestinal nuclear-localized SKN-1gf but not SKN-1wt supports the latter model. Moreover, the differential hypersensitivity to changes in the ubiquitin-like modifiers of the proteostatic network, reveals a previously undescribed role for post-translational modification of neddylation on SKN-1 activity. The stabilization of SKN-1gf in the nucleus when neddylation or ubiquitinylation is impaired supports recent evidence demonstrating that alternating chains of NEDD8-Ubiquitin can mediate nuclear proteotoxic stress 69 . In addition, NEDD8 conjugation to proteins can decrease stability and is an established mechanism to regulate transcription factor function 70 and neddylation can activate 113 cullin ring ligases, like Cul4, through conformational change which promotes cellular proteostasis 71 . Beyond the intestine, our work also reveals that while activation of SKN-1 in a single tissue is insufficient to recapitulate all the phenotypes associated with skn-1gf, we discovered that neurons are essential for this response. We describe how the activation in the two ASI sensory neurons, is required to initiate a cell non-autonomous program in the periphery that drives age-related pathology. Specifically, the activation of SKN-1 in the ASI neurons results in a change in the proteostatic balance within intestinal cells that drives the rapid depletion of intracellular lipid stores in these cells to fuel reproduction 15 , which supports the recent model of Morimoto and colleagues that the proteostasis network is critical for reproductive success 72 . We also demonstrate how intestinal activation of the dicer related helicase, drh-1, a regulator of both RNA interference and transcriptional responses to pathogen attack that was previously not connected to SKN-1 activity, can oppose the phenotypes of SKN-1 activation. Animals with constitutive activation of SKN-1 display increased expression of several RNAs with complex secondary structures and are also sensitive to changes in the expression of regulators of multiple RNA polymerases and the cellular proteostasis machinery 46 ; a response that is abolished by the drh-1gf allele. Thus, DRH-1 acts as a regulator of this RNA homeostat; acting as the drain valve in our analogy and can prevent the sink from overflowing. Intriguingly, activation of DRH-1 only provides a temporary relief to oppose the pathology stemming from SKN-1 activation, since if SKN-1 activity is not turned off, the negative consequences of transcriptional overload still emerge, albeit delayed. Remarkably, the activation of skn-1gf in two neurons is sufficient to drive this systemic response that drives multisystem functional decline with age. 114 Several key questions remain but our results promote the practice of matching exogenous manipulations of cellular cytoprotective responses to physiological signals stemming from cell non-autonomous pleiotropic consequences in distal cells. It remains possible that SKN-1 activation in two or more tissues or that multiple isoforms, which have different subcellular roles, are necessary to recapitulate the age-related decline in health observed in the skn-1gf genetic mutants. The partial mitigation of the skn-1gf response when drh-1gf is expressed only in the intestine further exemplifies the complexity of SKN-1 signaling and reveals a new layer of cell non-autonomous control in maintaining organismal homeostasis. Moreover, our work provides a framework to refine our models and include alternative approaches that can alleviate the consequences of dysregulated transcriptional activities. 115 MATERIALS & METHODS C. elegans Strains and Maintenance. C. elegans were raised on 6 cm nematode growth media (NGM) plates supplemented with streptomycin and seeded with OP50. All worm strains were grown at 20°C and unstarved for at least three generations before being used. Strains used in this study include: WT, N2 Bristol strain; SPC168, dvIs19[gst-4p::gfp]; skn-1gf(lax188); SPC572, dvIs19[gst-4p::gfp]; skn-1gf(lax188) drh-1gf(lax257); SPC2005, skn-1(lax188syb2619); SPC2004, skn-1gf(syb2597); SPC2058, ttTi4348-Pvha-6-skn-1a(lax188); SPC2065, ttTi5605-Prgef-1-skn-1c(lax188); SPC2067, ttTi5605-Pgpa-4-skn-1c(lax188); SPC2047, skn-1gf-AID; DV3803, ges-1p::TIR1; DV3805, rgef-1p::TIR1; SPC2048, gpa-4p::TIR1; SPC597, Ex[vha-6p::drh-1(lax257)]; skn-1gf; gst-4p::gfp. Single, double, and triple mutants were obtained by standard genetic crosses. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). Genetic mapping of the drh-1(lax257) mutation was performed as previously described 18,19,38,39,41,73 by linkage of the gst-4p::gfp suppression of the skn-1gf(lax188) phenotype by using the InDel mapping primer set 56 . Transgenic and genome editing. CRISPR/Cas9 genome editing was used to revert the skn- 1(lax188) mutation to WT in skn-1gf mutant animals and separately to create an independent gain-of-function allele in wild type animals. To generate tissue-specific expression of drh-1gf, the coding and 3’UTR of drh-1(lax257) was cloned between the vha-6p 5’ regulatory sequence and an SL2::WrmScarlet marker and injected along with a myo-2p::mCherry marker into skn-1gf; gst- 4p::gfp animals. Pharyngeal expression of mCherry was used to identify transgenic and non- transgenic siblings for imaging of the gst-4p::gfp reporter. RNAi treatment. RNAi treatment was performed as previously described 74 . Briefly, HT115 bacteria containing specific double stranded RNA-expression plasmids were seeded on NGM 116 plates containing 5mM isopropyl-b-D-thiogalactoside and 50µgml -1 carbenicillin. RNAi was induced at room temperature for 24 h. Synchronized L1 animals were added to those plates to knockdown indicated genes. DiI staining. DiI staining was performed as previously described 75 . In brief, worms were washed with M9 and then put on a rotator overnight in 10ug/mL DiI in M9 at 20°C. Worms were then washed with M9 twice and imaged on an agar pad. All centrifugation was done at 106 rcf for 30 seconds. Oil Red O Staining. Oil Red O fat staining was conducted as outlined in Stuhr et al. 2022 56 . In brief, worms were egg prepped and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria and raised to 120 h (Day 3 Adult stage) or 168 h (Day 5 Adult stage). Worms were washed off plates with PBS+triton, then rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with ORO in diH2O for 2 h. Worms were pelleted after 2 h and washed in PBS+triton for 30 min before being imaged at 20x magnification with LAS X software and Leica Thunder Imager flexacam C3 color camera. For tissue-specific degradation experiments, worms were egg prepped and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria and raised to 48 h (L4 stage) and then moved to experiment plates with vehicle 4mM ethanol or 4mM auxin. Worms were moved to new plates every day until 144 h post- drop (Day 4 Adults). Worms were washed off plates with PBS+triton, then rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with ORO in diH2O for 2 h. Worms were pelleted after 2 h and washed in PBS+triton for 30 min before being imaged at 20x magnification with LAS X software and Leica Thunder Imager flexacam C3 color camera. 117 Asdf Quantification. ORO-stained worms were placed on glass slides and a coverslip was placed over the sample. Worms were scored, as previously described in Stuhr et al. 2022 76 . Worms were scored and images were taken with LAS X software and Leica Thunder Imager flexacam C3 color camera. Fat levels of worms were placed into two categories: non-Asdf and Asdf. Non-Asdf worms display no loss of fat and are stained dark red throughout most of the body (somatic and germ cells). Asdf worms had most, if not all, observable somatic fat deposits depleted (germ cells only) or significant fat loss from the somatic tissues with portions of the intestine being clear (somatic < germ). Nile Red Staining. Nile Red fat staining was conducted as outlined in Stuhr et al. 2022 76 . In brief, worms were egg prepper and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria and raised to 48 h (L4 stage). Worms were washed off plates with PBS+triton, rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with Nile Red in 40% isopropyl alcohol for 2 h. Worms were pelleted after 2 h and washed in PBS+triton for 30 min before being images at 10X magnification with ZEN Software and Zen Axio Imager with the DIC and GFP filter. Fluorescence is measured via corrected total cell fluorescence (CTCF) via ImageJ and Microsoft Excel. CTCF = Worm Integrated Density-(Area of selected cell X Mean fluorescence of background readings) and normalized to the control. Hydrogen peroxide treatment. Conducted as previously described 13 . Briefly, synchronous worm populations at either Day 1 adulthood or 80HPF were washed 3x with M9+.01% Triton. 500uL of 20mM H2O2 was then added to 600uL of worms+M9+.01% Triton. Worms were then incubated on a rotator at 20°C for 25 minutes before being rinsed 3x with M9+.01% Triton. 118 RNAseq Analysis. RNAseq analysis was conducted as outlined in Stuhr & Curran 2020 77 . Worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, L1s were dropped on to seeded NGM plates and allowed to grow 48 h, 72 h, 120 h or 168 h (L4, Day 1 Adult, Day 3 Adult , Day 5 Adult) before collection. Animals were washed 3 times with M9 buffer and frozen in TRI reagent at -80°C until use. Animals were homogenized and RNA extraction was performed via the Zymo Direct-zol RNA Miniprep kit (Cat. #R2052). Qubit™ RNA BR Assay Kit was used to determine RNA concentration. The RNA samples were sequenced and read counts were reported by Novogene. Read counts were then used for differential expression (DE) analysis using the R package DESeq2 created using R version 3.5.2. Statistically significant genes were chosen based on the adjust p-values that were calculated with the DESeq2 package. Gene Ontology was analyzed using the most recent version of WormCat 2.0 78 . Chromatin Immunoprecipitation (ChIP). Chromatin was prepared as in Nhan et al. 2019 and Wormbook (Modern techniques for the analysis of chromatin and nuclear organization in C. elegans). Approximately 1,000,000 L4 synchronized animals were washed in M9 and collected into lysis buffer and flash frozen in liquid nitrogen. Worm pellets were ground via mortar and pestle and resuspended with 1.1% formaldehyde to crosslink proteins. Chromatin was fragmented via sonication and SKN-1::GFP was pulled down by overnight incubation with GFP affinity beads. Associated immunocomplexes were eluted by heat denaturing and DNA fragments were purified. Purified DNA fragments were then used as input for sequencing library preparation using the Diagenode MicroPlex Library prep kit V2. Bioinformatic analyses were done using the following software for ChIP-seq. Quality trimming and Adapter sequences were trimmed from raw paired end reads using Trim Galore package v 0.6.4. Reads were mapped to the C. elegans reference genome using bwa v 0.7.17. BAM files were sorted with Samtools v 1.10. BAM files were sorted to contain only uniquely mapped reads using Sambamba v 0.6.8. Peak calls were made using 119 MACS2 v2.2.7.1 Peak files were feature annotated using Chipseeker Bioconductor package in R using the annotate Peak function ASI-Enriched RNAseq. Performed as previously described 79,80 . In brief, Approximately 250,000 L4 synchronized WT and skn-1gf animals with GFP tagged ASI neurons (daf-7p::gfp) were washed with M9 6 times to remove residual bacteria. Animals were then pelleted and Cell Isolation Buffer(20mM HEPES, 0.25% SDS 200mM DTT 3% Sucrose pH8) was added to worms. Worms were incubated in Cell Isolation Buffer for 2min. Initial lysis was quenched by washing with M9 6 times. Worm pellet was resuspended in 20mg/ml Pronase and digested for 20 mins with vigorous pipetting every 5mins through a P1000 tip. Pronase digestion was quenched by resuspending in FBS. Cells were pelleted by centrifuging at 550rcf and resuspended in fresh FBS, Cells were passed through a 40 micron cell strainer. DAPI was added to cells to assess viability. Cells were then sorted on a Bio-Rad S3e FACS system. Neurons were homogenized and RNA extraction was performed via the Zymo Direct-zol RNA Miniprep kit (Cat. #R2052). Qubit™ RNA BR Assay Kit was used to determine RNA concentration. Low input RNA libraries were prepped using the Ovation SoLo RNA library kit from Tecan Genomics. The RNA libraries were sequenced by Novogene. Raw paired-end reads were quality trimmed and adapter trimmed using timmomatic v0.39. Quality trimmed reads were aligned to the C. elegans reference genome using STAR 2.7.6a. Mapped reads were counted via HTseq v2.0.2 union mode. Read counts were then used for differential expression (DE) analysis using the R package DESeq2 created using R version 3.5.2. Statistically significant genes were chosen based on the adjust p-values that were calculated with the DESeq2 package. Gene Ontology was analyzed using the most recent version of WormCat 2.0 78 . Movement Measurements - Crawling. Worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates 120 seeded with OP50. Worms were then allowed to grow until each time point (48 h post-drop for L4s, 72 h post-drop for Day 1 Adults, 120 h post-drop for Day 3 adults and 168 h post-drop for Day 5 adults). Once worms were the required stage of development, 30-50 worms were washed off of a plate in 50 uL of M9 with a M9+triton coated P1000 tip and dropped onto an unseeded NGM plate. The M9 was allowed to dissipate, and worms roamed on the unseeded plate for 1 hour before imaging crawling. Crawling was imaged with the MBF Bioscience WormLab microscope and analysis was performed with WormLab version 2022. Worm crawling on the plate was imaged for 1 minute for each condition at 7.5 ms. Worm crawling was analyzed with the software and only worms that moved for at least 90% of the time were included in the analysis. Protein prediction. Phyre2 81 was used to predict the structure of DRH-1wt and DRH-1gf; WT structure prediction file 25950237 and Missence3D (image with cyan and magenta in same structure, predicted changes in protein structure 30995449. 121 FIGURES Figure 1. Cell autonomous activity of SKN-1gf in ASI neurons. (A) Population of WT or skn- 1gf mutants were used for whole-animal RNAseq and ChIPseq. Genes with altered expression in skn-1gf mutants and identified with occupancy of SKN-1gf on the promoter region were analyzed in WormCat 2.0 (list of genetic loci can be found in Table S1) to reveal enriched classes (B). Expression of SKN-1wt-GFP (C) and SKN-1gf-GFP (D) are indistinguishable and restricted to ASI neurons; co-stained with DiI (red) that marks ciliated neurons; arrows designate GFP in ASI cell bodies (green). (E) WormCat 2.0 analysis of genes upregulated in FACS-enriched ASI neuron populations from skn-1gf animals. 122 Figure 2. Neddylation regulates nuclear SKN-1 stabilization in the intestine. Acute exposure to hydrogen peroxide drives nuclear accumulation outside of the ASI neurons (white arrows) for SKN-1wt-GFP (A) and SKN-1gf-GFP (B) within intestinal nuclei (yellow arrows). RNAi of uba-1 (C,D) and ned-8 (E,F) stabilizes SKN-1gf-GFP but not SKN-1wt-GFP in the intestine. (G) Cartoon of the role of ubiquitinylation and neddylation on nuclear SKN-1 stability. 123 Figure 3. SKN-1 activity in ASI neurons mediates peripheral stress responses. (A) Cartoon representation of strains for tissue-specific regulation of skn-1gf expression. (B) Expression of skn-1gf from gpa-4p (ASI neurons), but not vha-6p (intestine), or rgef-1p (pan-neuronal) can establish resistance to acute exposure to H2O2. (C) Tissue-specific expression of skn-1gf results in the cell non-autonomous activation of the gst-4p::gfp reporter but is not sufficient to drive somatic lipid depletion (D). Pan-neuronal, intestinal, and ASI neuron-specific degradation of SKN- 1gf can partially suppress somatic lipid depletion; n=300; N=3 per condition. A E C B D Auxin-mediated degradation SKN-1gf-AID Tissue-specific skn-1gf expression Intestine gst-4p::GFP induced gpa-4p vha-6p skn-1gf(lax188) WT Neurons Muscle Hypodermis Germline rgef-1p promoter::skn-1gf(c) promoter::skn-1gf(c) skn-1gf-AID promoter::TIR1 Intestine - vha-6p pan-neuronal - rgef-1p ASI neurons- gpa-4p 124 Figure 4. DRH-1 activation delays SKN-1gf-dependent healthspan decline. (A) Cartoon schematic of EMS genetic screen for suppressors of skn-1gf activation of gst-4p::gfp. (B) Genetic linkage maps the lax257 suppressor to LGIV. As compared to control RNAi (C) drh-1 RNAi abolishes the suppression of drh-1gf (D). Ectopic expression of drh-1gf (E) suppresses skn-1gf activation of gst-4p::gfp as compared to non-transgenic siblings (F). (G) Predicted structure and amino acid substitution (wt-cyan; gf-magenta) in DRH-1gf. drh-1gf suppresses the somatic lipid depletion phenotype of skn-1gf mutant at day 3 (H) but not day 5 (I) of adulthood. The resistance to acute H2O2 by skn-1gf exposure at day 1 of adulthood (J) and the sensitivity at day 3 of adulthood (K) is reversed by drh-1gf. (L-N) The suppression of the impaired movement phenotype of skn-1gf by drh-1gf at the L4 larval stage (L) is progressively abrogated at day 1 (M) and day 3 (N) of adulthood; n=50; N=3 per condition. A B J H +16.5 LGIV -22.9 -7.3 -5.1 +2 +4 +6.9 +10.9 +12.4 +14.2 26 mutations identified in GWS RNAi candidates G L M N EMS skn-1gf; gst-4p::gfp I K C E F skn-1gf; gst-4p::gfp skn-1gf drh-1gf;gst-4p::gfp control RNAi Ex drh-1gf NTS E F drh-1 RNAi D 125 Figure 5. Intestinal DRH-1 activation reduces transcriptional load of activated SKN-1. (A- G) drh-1gf (magenta) suppresses the activation of established SKN-1 targets in skn-1gf mutant animals (blue). (H) drh-1gf mutations abolishes the increased sensitivity of skn-1gf mutant animals to RNAi inhibition of transcriptional regulators. (I) WormCat 2.0 analysis of genes activated by skn-1gf and suppressed by drh-1gf. (J-Q) drh-1gf suppresses the increased expression of ncRNA in skn-gf mutants. The activation of the gst-4p::gfp in skn-1gf animals (R) is suppressed by intestinal expression of drh-1gf (S). (T-V) drh-1gf suppresses the increased expression of signaling molecules. (W) Cartoon schematic of cell non-autonomous signaling by skn-1gf in ASI neurons and drh-1gf in the intestine. 126 SUPPLEMENTAL FIGURES Figure S1. Gain-of-function mutations drive SKN-1 activation. (A) Cartoon of genomic location of skn-1 mutations. A strain with a de novo synthesized skn-1gf allele but not a strain with the skn-1gf allele reverted to WT display somatic lipid depletion at day 3 (B-D) and day 5 (E- G). SKN-1wt-GFP (H-J) and SKN-1gf-GFP (K-M) are expressed in ASI neurons (I,L) that overlaps with ciliated neurons (J,M). 127 Figure S2. Ubiquitin-related pathways mediate SKN-1-dependent activities. (A-D) RNAi of wdr-23 results in nuclear accumulation of SKN-1wt-GFP and SKN-1gf-GFP in the intestine as compared to control RNAi. (E-N) RNAi of ubc-12, smo-1, uba-2, ubc-9 does not result in intestinal accumulation of SKN-1wt-GFP or SKN-1gf-GFP, similar to control RNAi treated animals. 128 Figure S3. Cell non-autonomous impact of SKN-1 activation in neurons. Tissue-specific expression of skn-1gf isoform c in the intestine (A), muscle (B), pan-neuronal (C), or ASI neurons (D) results in the expression of the gst-4p::gfp reporter in multiple tissues. (E-K) Tissue-specific expression of skn-1gf isoform c does not result in somatic lipid depletion as compared to wt and skn-1gf mutant animals (K). (L-S) Tissue-specific and auxin-dependent degradation of SKN-1gf suppresses somatic lipid depletion (L) that does not significantly alter total intracellular lipid levels (M). 129 Figure S4. Identification and characterization of a drh-1gf mutation. (A-G) RNAi screen identifies reducing drh-1 expression in lax257 as causal for suppression of skn-1gf effects on gst- 4p::gfp activation. (H-M) Representative images of ORO staining revealing drh-1gf suppression of day 3 somatic lipid depletion but not day 5. (N-P) Representative images of Nile red staining of lipids and quantification at L4 stage; n=100; N=3 per condition (Q). 130 Figure S5. drh-1gf effects are tied to transcriptional regulation. (A-B) drh-1gf mutation abolishes the increased sensitivity of skn-1gf mutant animals to RNAi inhibition of transcriptional regulators, but the suppression by drh-1gf does not depend upon classical regulators of RNAi. (C) drh-1gf (magenta) suppresses the increased expression of genes induced in skn-gf mutants (blue). (D) Cartoon model of the reduction in transcriptional load by drh-1gf in the context of unchecked skn-1gf activity. 131 ACKNOWLEDGEMENTS We thank J Gonzalez, M Lynn, S Ledgerwood for technical assistance. This work was funded by the NIH R01AG058610 to SPC, F31GM137587 to CDT, F31AG077873 to NLS, and T32AG052374 to NLS and BTVC. Author contributions: Conceptualization: SPC; Methodology: CDT, NLS, CMR, BTVC, and SPC; Investigation: CDT, NLS, CMR, BTVC, SPC; Visualization: CDT, NLS, CMR, BTVC, SPC; Supervision: SPC; Writing (original draft): CDT and SPC; Writing (reviewing & editing): CDT, NLS, CMR, BTVC, SPC Competing interests: All authors declare that they have no competing interests 132 Chapter 5: Genetic variation in ALDH4A1 is associated with muscle health over the lifespan and across species *This chapter is a version of a manuscript published in eLife Authors: Osvaldo Villa 1,* , Nicole L. Stuhr 1,2,* , Chia-An Yen 1,2,* , Eileen M. Crimmins 1 , Thalida Em Arpawong 1 , and Sean P. Curran 1,2,3,# Affiliations: 1 Leonard Davis School of Gerontology, University of Southern California, 3715 McClintock Ave, Los Angeles, CA, 90089 USA 2 Dornsife College of Letters, Arts, and Science, Department of Molecular and Computational Biology, University of Southern California, 1050 Childs Way, Los Angeles, CA, 90089 USA 3 Norris Comprehensive Cancer Center, University of Southern California, 1441 Eastlake Ave, Los Angeles, CA, 90033 USA *, equal contribution # Corresponding author: Sean P. Curran E-mail: spcurran@usc.edu Keywords: sarcopenia; muscle; mitochondria, proline catabolism, aging; ALDH4A1; alh-6; C. elegans; human, HRS, GeneWAS Running title: Genetic variation in mitochondrial P5C dehydrogenase is associated with age- related muscle decline 133 ABSTRACT The influence of genetic variation on the aging process, including the incidence and severity of age-related diseases, is complex. Here we define the evolutionarily conserved mitochondrial enzyme ALH-6/ALDH4A1 as a predictive biomarker for age-related changes in muscle health by combining C. elegans genetics and a gene-wide association scanning (GeneWAS) from older human participants of the US Health and Retirement Study (HRS). In a screen for mutations that activate oxidative stress responses, specifically in the muscle of C. elegans, we identified 96 independent genetic mutants harboring loss-of-function alleles of alh-6, exclusively. Each of these genetic mutations mapped to the ALH-6 polypeptide and led to the age-dependent loss of muscle health. Intriguingly, genetic variants in ALDH4A1 show associations with age-related muscle-related function in humans. Taken together, our work uncovers mitochondrial alh- 6/ALDH4A1 as a critical component to impact normal muscle aging across species and a predictive biomarker for muscle health over the lifespan. 134 INTRODUCTION Sarcopenia is defined as the age-related degeneration of skeletal muscle mass and is characterized by a progressive decline in strength and performance 82 . This syndrome is prevalent in older adults and has been estimated by large scale studies to afflict 5-13% of people aged 60- 70 years and expands to 50% of those aged 80 and above 83 . Loss of muscle function is associated with a decline in quality of life and higher mortality and morbidity rates due to increased chance of falls and fractures 84,85 . Sarcopenia is linked to risk factors, such as a sedentary lifestyle, lack of exercise, and a diet deficient in protein and micronutrients 84 . However, several aspects of the molecular basis of the age-dependent decline in muscle health remains unknown. Although age-related muscle function is clearly linked to frailty 86 , previously, different etiologies of clinical weakness led to discrepancies in the definitions of sarcopenia 87,88 . Furthermore, the identification of human genetic loci that influence age-related functions has traditionally been difficult to characterize due to the methodological difficulties in longitudinal assessments; the prevalence of sarcopenia for example begins in the 4 th decade of life 89 . The US Health and Retirement Study (HRS) is a nationally representative survey of adults aged 50 years and older and has proven to be an invaluable dataset for investigating the normal aging processes 90-93 . New skeletal muscle cutpoints for identifying elevated risk for physical disability in older adults 94 has enabled cross-sectional analyses to identify cohorts of HRS participants with age-related decline in muscle function (e.g. grip strength (GS) basic activities of daily living (ADL) and instrumental ADL (IADL)) 95 . In Caenorhabditis elegans, mutation of the conserved proline catabolic gene alh-6 (88% identity to ALDH4A1 in humans) leads to premature aging and impaired muscle mitochondrial function 19 . Proline catabolism functions in a two-step reaction, beginning with the conversion of proline to 1- pyrroline-5-carboxylate (P5C) which is catalyzed by proline dehydrogenase, PRDH-1; 135 subsequently, P5C dehydrogenase, ALH-6, catalyzes the conversion of P5C to glutamate. alh-6 expression is observed in body wall muscle, pharyngeal muscle, and neurons 19 , and when alh-6 is mutated, the activation of gst-4p::gfp oxidative stress reporter is predominantly observed in the body wall muscle tissue and only in post-reproductive adults 96 . alh-6 mutants have increased levels of P5C; the accumulation of this toxic metabolic intermediate leads to an increase in reactive oxygen species (ROS), including H2O 2 19 , which then activates cytoprotective responses, impairs mitochondrial activity, and drives cellular dysfunction 17,19,38,39 . Several studies have linked disease states that drive morbidity and mortality with genomic variation through genome-wide association studies (GWAS) 97-99 and non-human models have been utilized to test how single genes can drive phenotypes that mimic the disease state in humans 100-102 . However, biological testing of genetic association studies for the normal human aging process remains underrepresented. The recent expansion of the HRS data to include genotyping of participants has enabled scans to test associations between normal aging phenotypes and variation across genes 103 . In this study, we exploit the facile genetic tractability of C. elegans with the rich genetic and phenotypic data available in the HRS to reveal genetic variation in alh-6/ALDH4A1 as a predictive indicator of muscle-related functionality in later life. 136 RESULTS/DISCUSSION While the strong induction of oxidative stress reporter activity in the musculature was linked to mutation of the mitochondrial P5C dehydrogenase gene, and not observed in other genetic mutants 13,15,17,20,38,39,41 , the breadth of genetic mutations that could induce stress responses in muscle was unknown. In order to identify additional genetic components of this age-related muscle phenotype, we performed an ethyl methanesulfonate (EMS) mutagenesis screen selecting for the same age-dependent activation of the gst-4p::gfp reporter in the musculature. We screened the progeny of ~4000 mutagenized F1 animals and isolated 96 mutant animals with age-dependent activation of the gst-4p::gfp reporter restricted to the body wall musculature, which phenocopies the alh-6(lax105) mutant (Figure 1a, Figure 1 - figure supplement 1). To rule out additional loss-of-function alleles of alh-6 we performed genetic complementation (cis-trans) testing with our established alh-6(lax105) allele; surprisingly, all 96 new mutations failed to complement and as such were all loss-of-function alleles of alh-6. To catalog these mutations, we began sequencing the alh-6 genomic locus in each of the mutants isolated. After sequencing approximately half of the mutants, we noted the repeated independent isolation of several distinct molecular lesions in alh-6: E78Stop (lax903, lax918, lax930), E447K (lax916, lax920, lax929, lax934), G527R (lax914, lax932, lax933, lax947), etc. (Figure 1b). The lack of diversity in genes uncovered and the independent isolation of identical alleles multiple times from this unbiased screen strongly suggest genetic saturation and specificity of this response to animals with defective mitochondrial proline catabolism. In addition, several mutations mapped to discreet regions of the linear ALH-6 polypeptide, including G152/K153, G199/E201/G202, and E418/G419, which may define critical domains in the folded protein. Imaging at day 3 of adulthood revealed that each mutant was phenotypically identical to lax105 in the activation of the gst-4p::gfp stress reporter in the bodywall muscle (Figure 1 - figure supplement 1), but with varying intensity (Figure 1c). We next mapped the location of each amino 137 acid mutated in our panel of alh-6 mutants on the structure of the ALH-6 protein (Figure 1 – figure supplement 2) 81,104 , which enabled a prediction model of the impact of each missense mutation (Figure 1 – figure supplement 3). Most missense mutations were predicted to maintain the overall structure (no structural damage), which suggests the associated phenotypes derive from a range of reduction of function mutations. Since the degree of mitochondrial dysfunction can influence both beneficial and detrimental physiological outcomes 105,106 , this collection of mutants provides a model to understand the complex role mitochondria play in organismal health over the lifespan. Taken together, these data reveal that the age-dependent and muscle activation of the gst-4p::gfp is driven specifically by mutations in mitochondrial alh-6. Based on the striking specificity of the muscle-restricted and age-dependent activation of the gst- 4p::gfp stress reporter in C. elegans harboring mutations in alh-6 19,39 , combined with the high degree of conservation in mitochondrial metabolism pathways across metazoans 17 , we reasoned that ALDH4A1 genetic variants would associate with phenotypes indexing normative, longitudinal changes in human aging-related functionality, specifically those that involve usage of different muscle groups. To test this hypothesis, we performed gene-wide association scanning (GeneWAS) adjusting for relevant covariates and indicators of population stratification in the US Health and Retirement Study (HRS); a nationally representative longitudinal study of >36,000 adults over age 50 in the US 91,107 . HRS collects biological and genetic samples on sub-sets of participants and assesses physical and psychosocial measures of all study participants in older adulthood; including multiple measures of muscle-related functionality (Figure 2 – source data 1). The human phenotypes, represented in the HRS index normative changes in aging-related physiological ability. There were 53 single nucleotide polymorphism (SNP) markers within the ALDH4A1 region that are on the Illumina Omni array representing 273 human SNPs in the gene . While measures like grip strength are more commonly used to assess muscle health, our inclusion of another phenotype represents changes in complex physiological process that are 138 influenced by the musculature and also other systems (e.g. metrics of walking speed can also be influenced by neurological factors). As such, future work to assess the role of neuronal alh- 6/ALDH4a1 will be important. Nevertheless, the observed decline in muscle-related measures with age is relevant. Overall, two associations between variants within ALDH4A1 and two phenotypes were detected and surpassed the respective empirical p-value thresholds, determined by permutation testing (Table 1 and Table 2). These demonstrate a pattern of association between ALDH4A1 variation and two independent phenotypes (Table 1). Because each of the SNPs within the ALDH4A1 region represents a tag, or marker SNP for human variation within the locus, this GeneWAS was unable to directly identify a causal variant; however, the indexing of variations within the same gene suggests conserved associations within this aging human cohort. With this study design, we did not intend to find a single genetic variant that would explain functionality of a specific muscle group. Specifically, we chose to include common measures of physical functioning that index aging-related decline. We calculated phenotypes for decline in functionality over time because they are more robust for testing genetic associations, represent normative aging processes in human samples compared to single-time point assessments, and index broader human functionality. (Figure 2 – source data 1). These results indicate that variants within the Aldh4a1 locus affect an individual’s performance on basic ambulatory movements such as speed of walking short distances or ability to exert hand grip strength. rs77608580 was significantly associated with change in gait speed over time (Figure 2a). Specifically, with each additional A allele, there was an average increase in gait speed of 0.052 meters per second per year compared to other same-aged individuals without the allele (p-value = 0.0025, surpassing the empirical p-value threshold of 0.006). This was assessed among 139 N=3,319 older individuals with a mean age of 73.0 years (SD=5.9) and mean gait speed of 0.80 meters per second (SD=0.25), or 2.6 ft/sec (Figure 3a). Measures of muscle health, such as grip strength, are effective biomarkers of overall health in older populations 108,109 . rs28665699 was significantly associated with an increase in grip strength over time (Figure 2b); with each additional A allele, there was an average increase in grip strength by 0.045 kg weight per year while holding all other characteristics constant (age, sex, and use of the dominant hand for gripping; p-value = 0.0009, surpassing the empirical p-value threshold of 0.0019). This was assessed among N=5,228 older individuals with a mean age of 68.9 years (SD=10.4), mean grip strength of 30.21 kg (SD=11.1), and average level of decline in grip strength at 2.31 kg per year (SD=5.37). If calculated as change over a ten-year period, those with 1 or 2 effect alleles would have stronger grip by 0.5 and 1.0 kg compared to those without an effect allele, respectively. The allele therefore is associated with a slower rate of decline in grip strength over a decade of age (Figure 3b). These effects are examples where variation in the gene contributing to phenotypes that represent different age-related change in functionality; overall we find there are small effects associated with each phenotype, but there are possible pleiotropic effects, and environmental or behavioral factors contributing. It is not known if any one of the identified ALDH4A1 SNPs is a causal variant or if they mark a different variant within the ALDH4A1 gene that was not represented on the HRS array. Regardless, these results collectively support a true association between ALDH4A1 and age-related physical function. We tested replication of the top two SNPs from the GeneWAS across ethnic subsamples in the HRS by calculating a common effect size across the samples. We did this by completing a fixed effects and random effects meta-analysis using PLINK software 110,111 (Table 3). For one SNP, the 140 minor allele frequency in the African ancestry sample was below 1% and thus the subsample was excluded from the meta-analysis. The Cochrane’s Q statistic (Q), as an indicator of variance across sample effect sizes, and the heterogeneity index (I), which quantifies dispersion across samples indicate random effects analysis fit the data better for gait speed decline, thus we focus on results from random effects to account for differences in effect sizes by sample (e.g., the I index indicates 64.95% of the observed variance between samples is due to differences in effect sizes between samples). Given these results, the common effect size calculated for grip strength decline still suggest significance of these associations with SNPs in the gene, whereas the effect for gait speed decline remained for the European ancestry cohort only and not across subsamples. Genetic data obtained from similarly large international cohorts studies (e.g., English Longitudinal Study of Ageing (ELSA; https://www.elsa-project.ac.uk); Irish Longitudinal Study on Ageing (TILDA; https://tilda.tcd.ie/); cohorts in the Survey of Health, Ageing and Retirement in Europe (SHARE; https://g2aging.org/overviews?study=share-aut), or Northern Ireland Cohort for the Longitudinal Study of Ageing (NICOLA; https://www.qub.ac.uk/sites/NICOLA/AboutNICOLA/); and others who are aged 50 and older will enable additional replication and additional cross comparisons. To test how genetic variation in P5C dehydrogenase can influence age-related muscle function, we returned to our collection of C. elegans strains harboring mutations in alh-6. We measured individual animal movement speed as a function of muscle health with age 112 . Only animals harboring mutations of ALH-6 at position G152R(lax940), K153T(lax924), S523F(lax928), and G527R(lax933) resulted in a significant loss of movement speed at larval stage 4; just prior to adulthood (Figure 4a). However, with the exception of S230F(lax907) and Y427N(lax918), all mutants tested displayed a significant reduction in movement speed at day 3 of adulthood (Figure 4b). As a secondary measure of muscle function in our panel of alh-6 mutants, we measured changes in swimming performance (Figure 4 – figure supplement 1), which has documented 141 effects on animal health and longevity 113 . It is established that swimming is a more energetically demanding activity than crawling on a plate 114 . Intriguingly, the effect of the canonical alh- 6(lax105) mutation on swimming was less pronounced than that observed for crawling speed and our panel of alh-6 mutants displayed differences in developmental and adult swimming performance. Taken together these data support the age-specific acceleration of muscle decline in mitochondrial proline catabolism mutants, which is conserved from nematodes to humans. The traits analyzed in the HRS came from a population-based study and were not assessed to allow us to identify physiological degeneration in specific muscles, rather to index and track overall age-related decline in functionality over time. It is widely accepted that genetic variation underlying these aging-related traits are highly polygenic. Thus, it is not expected that a single variant within a gene would be identified to drive these phenotypic results in humans. It is likely that small effects of multiple SNPs across multiple genes, including within the same gene, and with non-additive effects (e.g., gene-by-environment effects 115 ), contribute to the resulting phenotypes. With this use of genewide-association scanning approach, it is only possible to identify variants associated with overall effects. Without identifying a causal SNP, we can only aggregate available data to suggest what contributes to a biological pathway. For example, through exploitation of the publicly available Genotype-Tissue Expression (GTEx) database 116 , we found that one of the tag SNPs in ALDH4A1, rs77608580, was significantly associated with differential ALDH4A1 expression levels through the association with an expression quantitative trait locus (eQTL) in whole blood (Figure 3, figure supplement 1). Further experimental studies to reveal the downstream effect(s) of altered gene expression and/or specific muscle functionality phenotyping, is required to address mechanistic questions pertaining to unique muscle groups and muscle-specific activities. With the goal of better understanding the relationships between the phenotypes and potential disease, or system functionality, investigating more than one phenotype is an important strength 142 117,118 . Several studies have now demonstrated the biological utility of invoking multiple phenotypes for genetic association scans 119,120 . Invoking more than one phenotype and multiple SNPs for genetic testing, however, brings forth new challenges when needing to consider multiple hypothesis-testing burden, or type 1 error, while not missing underlying associations from overly stringent significance criteria that typically assume independent genetic variants and phenotypes. Prior studies that have used a multiple phenotype approach to investigate upwards of a thousand phenotypes do so with a completely agnostic design, used for exploratory hypothesis-generation 119,120 . In contrast, with the current study, we sought to use the selected phenotypes in human data to cross-validate findings, with hypotheses generated from a model organism. Thus the current study uses a more targeted, hypothesis-driven approach, by limiting a study to two selected phenotypes and variants within one identified genomic region. With this design, we retain the potential for detecting possible areas of genetic pleiotropy (i.e., genetic variation on more than one phenotype) and possiblities for identifying mechanistic pathways leading to normative aging- related muscle functioning and decline. Existing biomarkers of muscle health that can accurately predict muscle health later in life are extremely scarce due to limited data in human aging and an incomplete understanding of the molecular basis of sarcopenia. Moreover, facile approaches to experimentally validate the hypothesis generated from deep human genetic variation datasets are scarce. Understanding the diversity of genetic variation underlying sarcopenia, as well as their corresponding phenotypic outcomes, will be critical for providing accurate risk assessments for family planning and genetic counseling of older adults. We have established a powerful experimental platform that synergistically utilizes the data rich resources of the US Health and Retirement study with the genetically tractible and methodologically rich C. elegans model. We anticipate this new research paradigm will be a formidable tool for collaboration between computational and bench sciensts. Although the etiology of human disease is complex and multifactorial, we have used a 143 combination of classical C. elegans genetics and human genetic association studies to define genetic variation in alh-6/ALDH4A1 as a new biomarker of age-related muscle health in human. 144 MATERIALS & METHODS C. elegans Strains and Maintenance C. elegans were cultured using standard techniques at 20°C 121 . The following strains were used: wild type (WT) N2 Bristol, SPC321 [alh-6(lax105)], CL2166[gst-4p::gfp], SPC223 [alh- 6(lax105);gst-4p::gfp], SPC542 [alh-6(lax917);gst-4p::gfp], SPC531 [alh-6(lax906);gst-4p::gfp], SPC528 [alh-6(lax903);gst-4p::gfp], SPC552 [alh-6(lax927);gst-4p::gfp], SPC561 [alh- 6(lax937);gst-4p::gfp], SPC564 [alh-6(lax940);gst-4p::gfp], SPC549 [alh-6(lax924);gst-4p::gfp], SPC566 [alh-6(lax945);gst-4p::gfp], SPC540 [alh-6(lax915);gst-4p::gfp], SPC562 [alh- 6(lax938);gst-4p::gfp], SPC563 [alh-6(lax939);gst-4p::gfp], SPC546 [alh-6(lax921);gst-4p::gfp], SPC527 [alh-6(lax902);gst-4p::gfp], SPC529 [alh-6(lax904);gst-4p::gfp], SPC536 [alh- 6(lax911);gst-4p::gfp], SPC534 [alh-6(lax909);gst-4p::gfp], SPC559 [alh-6(lax935);gst-4p::gfp], SPC532 [alh-6(lax907);gst-4p::gfp], SPC569 [alh-6(lax993);gst-4p::gfp], SPC544 [alh- 6(lax919);gst-4p::gfp], SPC562 [alh-6(lax938);gst-4p::gfp], SPC551 [alh-6(lax926);gst-4p::gfp], SPC530 [alh-6(lax905);gst-4p::gfp], SPC533 [alh-6(lax908);gst-4p::gfp], SPC548 [alh- 6(lax923);gst-4p::gfp], SPC550 [alh-6(lax925);gst-4p::gfp], SPC565 [alh-6(lax941);gst-4p::gfp], SPC538 [alh-6(lax913);gst-4p::gfp], SPC543 [alh-6(lax918);gst-4p::gfp], SPC541 [alh- 6(lax916);gst-4p::gfp], SPC545 [alh-6(lax920);gst-4p::gfp], SPC554 [alh-6(lax929);gst-4p::gfp], SPC558 [alh-6(lax934);gst-4p::gfp], SPC526 [alh-6(lax901);gst-4p::gfp], SPC568 [alh- 6(lax992);gst-4p::gfp], SPC535 [alh-6(lax910);gst-4p::gfp], SPC553 [alh-6(lax928);gst-4p::gfp], SPC539 [alh-6(lax914);gst-4p::gfp], SPC556 [alh-6(lax932);gst-4p::gfp], SPC557 [alh- 6(lax933);gst-4p::gfp], SPC567 [alh-6(lax947);gst-4p::gfp]. Double mutants were generated by standard genetic techniques. E. coli strains used were as follows: OP50/E.coli B for standard growth. All genetic mutants were backcrossed at least 4X prior to phenotypic analyses. 145 Genetic Complementation (cis-trans) Testing Hermaphrodites from each isolated mutant that phenocopied the alh-6(lax105)-like, age-related activation of the gst-4p::gfp reporter in the musculature were mated to SPC223 [alh-6(lax105);gst- 4p::gfp] males. F1 progeny were screened at day 3 of adulthood for the alh-6(lax105)-like phenotype, which indicates a failure of the alh-6(lax105) allele to complement the mutation in the new mutant strain; thus the new mutant harbors a loss-of-function allele in alh-6. DNA Sequencing of alh-6 Genetic Mutants Approximately 200 adult worms were collected and washed with M9. Animals were homogenized and genomic DNA was extracted using the Zymo Research Quick-DNA Miniprep kit (Cat. #D3025). The entire alh-6 genomic sequence (ATG to stop) was amplified by PCR and cloned in a linearized pMiniT 2.0 vector (NEB PCR Cloning Kit, Cat. #E1202S). Plasmid DNA was purified using the Zymo Research Zyppy Plasmid Miniprep kit (Cat. D4019) and sequenced. Microscopy Zeiss Axio Imager and ZEN software were used to acquire all images used in this study. For GFP reporter strains, worms were mounted in M9 with 10mM levamisole and imaged with DIC and GFP filters. Worm areas were measured in ImageJ software (National Institutes of Health, Bethesda, MD) using the polygon tool. HRS Human Samples The US Health and Retirement Study (HRS 91,92 ) is a nationally representative, longitudinal sample of adults aged 50 years and older, who have been interviewed every two years, beginning in 1992. Because the HRS is nationally representative, including households across the country and the surveyed sample now includes over 36,000 participants, it is often used to calculate national 146 prevalence rates for specific conditions for older adults, including physical and mental health outcomes, cognitive outcomes, as well as financial and social indicators. The sample for the current study is comprised of a subset of the HRS for which genetic data were collected, as described below. To reduce potential issues with population stratification, the GeneWAS in this study was limited to individuals of primarily European ancestry. The final subsample varied depending on the phenotype. The subsample for each phenotype ranged from N=3,319 (for change in gait speed) to N=9,907 (for walking across a room), with the proportion of women at 58.5%. Genotyping Data For HRS, genotype data were accessed from the National Center for Biotechnology Information Genotypes and Phenotypes Database (dbGaP 93 ). DNA samples from HRS participants were collected in two waves. In 2006, the first wave was collected from buccal swabs using the Qiagen Autopure method (Qiagen, Valencia, CA). In 2008, the second wave was collected using Oragene saliva kits and extraction method. Both waves were genotyped by the NIH Center for Inherited Disease Research (CIDR; Johns Hopkins University) using the HumanOmni2.5 arrays from Illumina (San Diego, CA). Raw data from both phases were clustered and called together. HRS followed standard quality control recommendations to exclude samples and markers that obtained questionable data, including CIDR technical filters 122 , removing SNPs that were duplicates, had missing call rates ≥ 2%, > 4 discordant calls, > 1 Mendelian error, deviations from Hardy-Weinberg equilibrium (at p-value < 10 -4 in European samples), and sex differences in allelic frequency ≥ 0.2). Further detail is provided in HRS documentation 107 . Applying these criteria to the gene region, on chromosome 1, (NC_000001.10): 19,194,787 - 19,232,430 resulted in available data on 70 SNPs within the ALDH4A1 region that are on the Illumina array to represent 273 human SNPs in the gene . With the goal of evaluating whether representative marker SNPs across the 147 gene are associated with the phenotypes of interest, we implemented a pruning procedure, which sequentially scans SNPs in linkage disequilibrium (LD), and performs thinning to subset to more independent SNPs fbased on a given threshold of correlation between SNPs and between linear combinations of SNPs. To achieve this, SNPs were first filtered to retain 53 SNPs that had a minor allele frequency at 0.01 or greater. We then pruned by recursively removing SNPs within a sliding window of 25 (i.e., 25 consecutive SNPs), shifted the window with 5 SNPs forward, and set the variance inflation factor threshold at 2. This yielded 21 SNPs for consideration (Table 2). Statistical Analysis of HRS Data Set Following SNP extraction, we followed analytical steps of Phenotype construction Gene-wide association scans (GeneWASs), and SNP evauation. HRS Phenotype Construction HRS phenotype construction was completed to calculate common measures of normal age- related muscle decline in functionality over time. Figure 2 – source data 1 shows the HRS data years from which phenotypes were calculated and details on how the variable is defined, and score or variable range. Datasets from multiple survey years were merged to get repeated assessments of variables on the same individuals. Phenotypes were calculated based on consensus following a review of the literature on assessments for age-related outcomes for variables implemented in the HRS and similar population-based surveys of aging. Further background for coding of specific phenotypes are described in detail previously for gait 123-125 . Phenotypes for grip strength decline and gait speed decline were assessed as change in performance on those tasks over time. Change was calculated by taking the score from the most recent assessment and subtracting the score from the first assessment for each person, within the respective years listed. Additional descriptive statistics on phenotypes can be provided. Phenotypes were calculated using SAS 9.4. 148 GeneWAS GeneWAS occurred through separate linear regression scans, under an additive model, adjusting for relevant covariates and indicators of population stratification as described below. Population Stratification - As with any statistical analysis of association, if the correlation between dependent and independent variables differs for subpopulations, this may result in spurious genetic associations 126 . To reduce such type 1 error, we conducted the GeneWAS adjusting for population structure as indicated by latent factors from principal components analysis (PCA) 127,128 . Detailed descriptions of the processes employed for running PCA, including SNP selection, are provided by HRS, and follow methods outlined by Patterson and colleagues) 128 . Two PCAs were run. The first PCA included 1,230 HapMap anchors from various ancestries and were used to test against self-reported race and ethnic classifications. Several corrections to the dataset were made based on this analysis. The second PCA was run on the corrected dataset, on unrelated individuals and excluding HapMap anchors, to create eigenvectors to serve as covariates and adjust for population stratification in association tests. From the second PCA, the first two eigenvalues with the highest values accounted for less than 4.5% of the overall genetic variance, with additional components (3-8) increasing this minimally, by a total of ~1.0% 107 . Based on these analyses, we opted for a strategy that does not ignore population substructure, but also does not over-correct, and adjusted for the first four PCs in all analyses. When coupling this approach of adjusting for principal components with all quality control procedures performed, excluding any related individuals and limiting the dataset for ancestral homogeneity, we reduce the likelihood of false associations resulting from population stratification) 127-133 . Regression models and other covariates - When conducting regressions on phenotypes indicating change over time, additional adjustments were made using covariates for baseline levels, number 149 of years during which change was calculated, and variables shown to affect outcomes. For example, with change in gait speed, a linear regression scan was run adjusting for sex, age at the first assessment point, number of years of follow-up, baseline walking speed, and floor type in addition to principal components. All GeneWAS were completed using PLINK 2.0 111 . The strength of the associations, as indicated by effect sizes and p-values, are not directly comparable for each phenotype because the sample sizes differed by phenotype. Thus, the strength of an association does not reflect how strong a SNP effects one phenotype compared to another. Because we did find more than one variant associated with the phenotypes, we are more confident that these results were not due to type 1 error. SNP Evaluation We evaluated SNP associations in the GeneWAS by p-value. With the number of SNPs and primary phenotypes in this study, strict Bonferroni correction would yield an adjusted multiple test- correction p-value threshold of 0.0012 (for 21x2 tests). However, Bonferroni correction such as these are too conservative because of the correlations among SNPs 134,135 and the cross-validation approach. To address the correlation among SNPs, we implement a pruning schema and calculate empirical p-value thresholds, through permutation 134-137 . Permutation is a process whereby necessary correlations between SNPs and phenotypes are intentionally shuffled so that p-values for the shuffled (null) data are compared to the non-shuffled data. This permutation is repeated multiple times in order to determine an empirical p-value 135,137,138 , a calculated threshold at which a test result is less likely to achieve significance by chance alone. Thus, when performing 1,000 permutations using PLINK and max(T) option 137 , the empirical p-value thresholds of 0.0019 for grip strength decline and 0.006 for gait speed decline were observed for determining gene- wide significance. For SNP comparisons, we used R (CRAN; https://www.r-project.org). 150 Gene Expression The Genotype-Tissue Expression (GTEx) database 116 , the most comprehensive, publicly- available resource for tissue-specific gene expression data, was used to evaluate whether there was evidence for regulatory functions of SNPs within the gene. We entered the top SNPs into GTEx to assess relationships with differential gene expression. WormLab Measurements As previously described 112 , but in brief; 15-20 animals were moved to a NGM stock plate without E. coli OP50 and recorded in WormLab software (MBF Bioscience) for 2 minutes. Swimming Measurements As previously described 77 , but in brief; 15-20 worms were moved to an unseeded NGM stock plate for one hour. Then worms were washed with M9 into 5µL drops onto a fresh NGM plate. After one minute, 15-20 worms were imaged via Movie Recorder at 50 ms exposure using ZEN 2 software (Zeiss Axio Imager). Statistical Analysis of alh-6 Genetic Mutants Data are presented as mean ± SD. Comparisons and significance were analyzed in GraphPad Prism 8. Comparisons between more than two groups were done using ANOVA. 151 FIGURES Figure 1. Mutation of alh-6 uniquely activates age-dependent and activation of the gst- 4p::gfp oxidative stress reporter in muscle. (a) Schematic representation of genetic screen for mutants that phenocopy alh-6(lax105). (b) Schematic representation of the ALH-6 protein with the molecular identity of mutants isolated and sequenced annotated. Alleles that were 152 selected for additional functional tests of muscle function (Fig. 4) are highlighted in red and the location of the canonical alh-6(lax105) allele is highlighted in green. These alleles represent all the sequenced mutations in alh-6 that were isolated from the EMS screen. (c) Quantification of stress reporter activation in the muscle in the new alh-6 mutant alleles, as measured by the intensity of GFP fluorescence from the oxidative stress reporter gst-4p::gfp (see Supplemental for Figure 1 for representative images). t-test relative to gst-4p::gfp reporter animals (control); *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 153 Figure 2. ALDH4A1 variants associate with human age-related phenotypes for change in muscle function. Plot of association between variants in the ALDH4A1 gene and normative aging-related muscle decline in (a) gait speed and (b) grip strength in the U.S. Health and Retirement Study (HRS). The x-axis shows the beta estimate for the effect of each SNP, represented by a dot, on the phenotype. The y-axis shows the log of the p-value for the association between the SNP and the phenotype. SNPs that surpassed the empirical p-value threshold, shown as a red line, for decline in gait speed (empirical p-value = 0.006) and grip strength (empirical p-value = 0.0019) are depicted as red dots. SNPs that surpassed a suggestive threshold (p-value = 0.009 for gait speed) are depicted as purple dots. 154 Figure 3. Effects of ALDH4A1 variation on phenotypes representing association with change in aging-related function in a normative, population-based sample of older adults. (a) Change in gait speed over 10 years. Effect of SNP rs77608580 on aging-related changes in gait speed (b=0.052, p=0.0025). Over the span of 1 decade, on average, those with 1 or 2 effect alleles will have faster gait speeds with a difference of 0.52 and 1.04 m/sec, respectively, compared to those without an effect allele. (b) Decline in grip strength over 10 years. Variation in ALDH4A1 (SNP rs28665699) is inversely associated with decline in aging- related grip strength (b=-0.045, p=0.0009). Individuals with 1 or 2 effect alleles have slower progression of weakened grip strength over 10 years by 0.5 and 1.0 kg respectively, compared to the same aged individuals without the effect allele. 155 Figure 4. alh-6 mutations accelerate loss of muscle function. WormLab software analysis of adjusted center point speed of individual animals of the given genotypes at the L4 stage (a) or day 3 of adulthood (b). Brown-Forsythe and Welch ANOVA test with Dunnett’s T3 multiple comparisons test, with individual variances computed for each comparison. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 156 TABLES Table 1. Top SNPs associated with specific phenotypes Phenotype SNP name Location Ref. Allele Minor allele Freq. 1 Scan N 2 Effect Size 3 P-value Grip strength decline rs28665699 1920018 5 A A 0.014 5228 -0.045 9.1E-04 Gait speed decline rs77608580 1919696 8 A G 0.017 3319 0.052 2.5E-03 1. Freq = frequency of minor allele as reported by 1000 genomes. 2. N = Sample size of scan for the phenotype and SNP. 3. Effect sizes provided are standardized regression coefficients. 157 Table 2. SNPs remaining after filtering for minor allele frequency and pruning based on linkage disequilibrium. SNP location reference allele minor allele frequency rs28652778 19194995 A 0.20 rs28405179 19195143 A 0.03 rs111289603 19195492 G 0.03 kgp2515954 19195951 A 0.02 rs77608580 19196968 A 0.04 rs9699485 19197237 G 0.02 rs3935824 19197849 G 0.18 rs28665699 19200185 A 0.03 rs28493067 19203333 A 0.35 rs6426814 19204173 A 0.19 rs35285457 19205258 A 0.14 rs7365978 19206020 A 0.21 rs28508407 19210018 A 0.27 rs113232075 19211163 G 0.02 rs9426718 19213022 A 0.02 rs4911985 19215440 G 0.22 rs28582076 19217295 G 0.02 rs11484743 19219987 C 0.02 rs17492518 19221621 A 0.04 rs4912044 19230263 A 0.18 rs79251057 19231130 A 0.04 158 Table 3. Replication across ethnic subsamples in the HRS. Phenotype SNP name location minor allele European Ancestry (N) 1 African Ancestr y (N) 1 Hispanic Ancestry (N) 1 Fixed Effect p- value 2 Random Effect P-value 3 Fixed Effect 2 : OR or beta Random Effect 3 : OR or beta Q 4 I 5 Grip strength decline rs2866 5699 19200185 A 5228 -- 409 0.001 50 0.00150 -0.0418 -0.0418 0.3 341 0.0 0 Gait speed decline rs7760 8580 19196968 A 3319 381 237 0.007 75 0.72900 0.0424 0.0146 0.0 577 64. 95 1. N: sample size by group included in the meta-analysis 2. Fixed effect: p-value and effect size 3. Random effect: p-value and effect size 4. Cochrane’s Q statistic: indicator of variance across sample effect sizes 5. I: heterogeneity index to quantify dispersion across samples 159 SUPPLEMENTAL FIGURES Figure 1 – figure supplement 1. Novel alleles of alh-6 induce muscle specific activation of the gst-4p::gfp stress reporter reporter. GFP fluorescence images of gst-4p::gfp animals harboring alh-6 mutations, as indicated. scale bar = 100m. Quantification of fluorescence is shown in Fig 1c. 160 Figure 1 – figure supplement 2. Location of amino acid substitutions in alh-6 mutants. Location of individual missense mutations of ALH-6 mutants on the predicted structure of the wild type ALH-6 protein by Phyre2 104 . Mutated residues are colored (purple) and circled. 161 Figure 3 – figure supplement 1. Association between rs77608580 and ALDH4A1 gene expression levels in whole blood. Normalized gene expression levels for ALDH4A1 (y-axis) by SNP genotype (x-axis) is shown by violin plot. Also on the x-axis is the sample size by genotype in parenthesis. Violin plots show the density plot of the data (green cloud) with the median of the data shown by the white line of the black box plot within, the lower and upper border of the box plot corresponding to the first and third quartiles, respectively. The black dots represent sample points for the genotype in which there were too few samples to depict by box plot. A linear regression model was used to estimate the mean difference in expression levels, calculated as a Normalized Effect Size (NES) to compare the alternative allele, G, to the minor allele, A. Figure and data source: GTEx Analysis Release V8 (dbGaP Accession phs000424.v8.p2) 116 162 Figure 4 – figure supplement 1. alh-6 mutations accelerate loss of muscle function. Rate of thrashing for individual animals of the given genotypes at the L4 stage (a) or day 3 of adulthood (b). Brown-Forsythe and Welch test with Dunnett’s T3 multiple comparisons test, with individual variances computed for each comparison. *, p<0.05; **, p<0.01; ***, p<0.001; ****, p<0.0001. 163 ACKNOWLEDGEMENTS We thank J. Gonzalez for technical assistance, Dr. W. Mack for statistical consultation, and Drs. R. Irwin and C. Duangjan for critical reading of the manuscript. Some strains were provided by the CGC, which is funded by the NIH Office of Research Infrastructure Programs (P40 OD010440). This work was funded by the NIH R01 AG058610 and RF1 AG063947 to S.P.C., T32 AG052374 to O.V. and N.L.S. and T32 GM118289 to N.L.S. This study was supported in part by funding from The National Institute on Aging, through the USC-Buck Nathan Shock Center (P30 AG068345). The National Institute on Aging has supported the collection of both survey and genotype data for the Health and Retirement Study through co-operative agreement U01 AG009740. The datasets are produced by the University of Michigan, Ann Arbor. The HRS phenotypic data files are public use datasets, available through: https://hrs.isr.umich.edu/data- products/access-to-public-data.The HRS genotype data are available to authorized researchers: https://www.ncbi.nlm.nih.gov/projects/gap/cgi-bin/study.cgi?study_id=phs000428.v2.p2 Declarations of Interest: The authors declare no competing interests. 164 Chapter 6: Lipid Quantification in Caenorhabditis elegans by Nile Red and Oil Red O Staining *This chapter is a version of a manuscript published in bio-protocol Authors: Nicole L. Stuhr 1,2,# , James D. Nhan 1,2,# , Amy M. Hammerquist 1,2 , Bennett Van Camp 1 , and Sean P. Curran 1,2,3 Affiliations: 1 Leonard Davis School of Gerontology, University of Southern California, Los Angeles, United States 2 Department of Molecular and Computational Biology, Dornsife College of Letters, Arts and Science, University of Southern California, Los Angeles, United States 3 Norris Comprehensive Cancer Center, Keck School of Medicine, University of Southern California, Los Angeles, United States For Correspondence: For correspondence: spcurran@usc.edu Keywords: Lipids, C. elegans, Fat, Distribution, Abundance, Age-dependent somatic depletion of fat (Asdf), Nile Red, Oil Red O # Contributed equally to this work 165 ABSTRACT The ability to stain lipid stores in vivo allows for the facile assessment of metabolic status in individuals of a population following genetic and environment manipulation or pharmacological treatment. In the animal model, Caenorhabditis elegans, lipids are stored in and mobilized from intracellular lipid droplets in the intestinal and hypodermal tissues. The abundance, size, and distribution of these lipids can be readily assessed by two staining methods for neutral lipids: Oil Red O, which can define tissue distribution and lipid droplet size or Nile Red, which can also quantitatively measure lipid droplet abundance. C. elegans are a useful animal model in studying pathways relating to aging, fat storage, and metabolism whose transparent nature allows for easy microscopic assessment of lipid droplets. This is done by fixation and permeabilization, staining with NR or ORO, image capture on a microscope, and computational identification and quantification of lipid droplets in individuals within a cohort. To ensure reproducibility of measures of lipid, we provide a detailed protocol to measure intracellular lipid dynamics in C. elegans. Graphical Abstract: Flow chart depicting the preparation of C. elegans for fat staining protocols. 166 BACKGROUND In C. elegans, lipid homeostasis is one of the many cellular processes that declines with age 15,139,140 . C. elegans are a popular model organism to study lipid metabolism because of the multiple techniques that have been developed to determine lipid content of worms 141 . Analysis of specific lipid species can be conducted with the use of high-performance liquid chromatography- mass spectrometry and gas-chromatography-mass spectrometry 142,143 . Although useful in determining lipid extract complexity, these methods require time for analysis and expensive machinery that isn’t readily available. Here we describe two methods that take advantage of the transparency of the worm for visualizing the amount and tissue distribution of intracellular lipids. C. elegans can be stained with many lipophilic dyes; two of the most readily available, easy-to- use, and biochemically validated compounds are Nile Red and Oil Red O 15,19,77,143,144 . Nile Red, 9-diethylamino-5H-benzo[α]phenoxazine-5-one, is a lipophilic dye that stains intracellular neutral lipid droplets 145 . Nile Red allows for the quantification of lipid abundance due to the emission of green light following excitation. Oil Red O is a fat-soluble dye that stains neutral lipids a bright red color, allowing for qualitative assessment of lipid droplet size and lipid distribution 146 . Staining C. elegans with these two different dyes allows for the determination of alterations in quantity and distribution of lipids. 167 MATERIALS & REAGENTS 1. Microscope slides (VWR, catalog number: 48312-004) 2. Sterile 6 cm Petri dishes (VWR, catalog number: 25373-085) 3. Platinum wire (Tritech Research, catalog number: PT-9010) 4. Eyelash brush (consisting of a human eyelash attached to a Pasteur pipette, generic, with tape, generic) 5. Microscope slide cover slips (VWR, catalog number: 48366-227) 6. Parafilm (Sigma-Aldrich, catalog number: P7543) 7. 200 μl micropipette tips (Genesee Scientific, catalog number: 24-151RL) 8. 1000 μl micropipette tips (Genesee Scientific, catalog number: 23-165R) 9. E. coli OP50-1 (available from Caenorhabditis Genetics Center) 10. Worm strains (available from Caenorhabditis Genetics Center) 11. Sodium Chloride (Fisher Scientific, catalog number: 02-004-047) 12. Peptone (BD, catalog number: 211820) 13. Bacto Agar (BD, catalog number: 214040) 14. Cholesterol (Sigma-Aldrich, catalog number: C8667) 15. Ethanol (VWR, catalog number: 89125-172) 16. Calcium Chloride (Sigma-Aldrich, catalog number: C3881) 17. Magnesium Sulfate (Sigma-Aldrich, catalog number: M2773) 18. Potassium Phosphate dibasic (Sigma-Aldrich, catalog number: P5504) 19. Potassium Phosphate monobasic (Sigma-Aldrich, catalog number: P0662) 20. Streptomycin sulfate (Sigma-Aldrich, catalog number: S6501) 21. LB powder (Teknova, catalog number: L9315) 22. Potassium Chloride (Sigma-Aldrich, catalog number: P3911) 23. Triton X-100 (Sigma-Aldrich, catalog number: X100-100 ml) 24. Isopropanol (BHD, catalog number: BDH1133) 168 25. DAPI (Sigma-Aldrich, catalog number: D9542) 26. Dimethyl sulfoxide (DMSO) (Sigma-Aldrich, catalog number: D2650-5X10ML) 27. Nail Polish (clear, generic) 28. Oil Red O (Alfa Aesar, catalog number: A12989) 29. Nile Red for microscopy (Sigma-Aldrich, catalog number: 72485) 30. Acetone 99.5% (BDH, catalog number: 1101-1LP) 31. 15 ml Centrifuge Tubes with EZ Flip Cap (ThermoFisher Scientific, catalog number: 362694) 32. 1.5 ml Microtube (Axygen Scientific, catalog number: MCT-150-C) 33. 10 ml Disposable Syringe (VWR, catalog number: 76290-382) 34. 0.2 μm Syringe Filters (VWR, catalog number: 28145-477) 35. Nematode Growth Medium (NGM) plates (see Recipes) 36. Phosphate buffered saline (PBS) (see Recipes) 37. PBS + 0.01% Triton X-100 (PBST) (see Recipes) 38. M9 (see Recipes) 39. Oil Red O stock solution (see Recipes) 40. Nile Red stock solution (see Recipes) 41. DAPI stock solution (see Recipes) Equipment 1. 2-20 μl micropipette (Gilson FA10003M) 2. 20-200 μl micropipette (Gilson FA10005M) 3. 100-1000 μl micropipette (Gilson FA10006M) 4. Microcentrifuge for 1.5 ml tubes (Eppendorf, model: 5430) 5. Worm incubator (Generic, maintain at 20 °C) 6. Tube rotator (Thermo Scientific, catalog number: 88881001) 7. Compound microscope with DIC, DAPI, and GFP filters, and 5× and 10× objectives (Zeiss, 169 model: AxioScope5) 8. Color camera (Zeiss AxioCam MRm) 9. Digital camera (Zeiss AxioCam ERc5s) 10. Stir plates, generic 11. Magnetic stir bars, generic Software 1. Imaging software Dependent on the microscope used (in this case, we used a Zeiss Axioscope and the associated ZEN imaging software) 2. ImageJ (available from NIH) 3. Data analysis software We used GraphPad prism, although any software capable of t-test will suffice. ANOVA may be helpful if comparing multiple conditions, but not strictly necessary. 170 PROCEDURE A. Prepare hermaphrodite worms to use for staining. 1. Maintain the growth of strains needed for Oil Red O fat staining and/or Nile Red fat staining. Allow populations to be unstarved for 3+ generations to avoid changes due to epigenetics. 2. Ramp up worm strains that will be used for fat staining (Nile Red and/or Oil Red O). 3. Synchronize populations by alkaline hypochlorite treatment overnight, as described before 13 . 4. Drop larval stage 1 worms onto freshly seeded 6 cm NGM plates. Aim for 100 worms per plate. 5. 48 hours after dropping the worms (unless the strains are developmentally delayed or staining Day 3 adults), L4 fat staining can occur. 6. If imaging Day 3 adults, wash adult worms with M9 to new plates 72 hours-post drop and 96 hours post-drop to avoid having the progeny overtake the experiment worms. Adult animals can be enriched from larvae and eggs by allowing the larger adults to settle by gravity in a 15 ml centrifuge tube. 7. 120 hours after dropping the worms (unless strains are developmentally delayed), Day 3 adult fat staining can occur. B. Oil Red O fat staining 1. Prepare Oil Red O staining solution. a. Make staining solution: Dilute Oil Red O stock solution in autoclaved water (600 μl Oil Red O stock solution per 400 μl water). Depending on the volume, prepare in either a 1.5 ml Eppendorf tube or 15 ml centrifuge tube. Note: 600 μl is used per sample. Make more than necessary to account for volume lost during filtration (Protocol B.1.d.). b. Wrap the top of the tube with parafilm. 171 c. Place Oil Red O staining solution on a tube rotator overnight. Note: Although preparing the stain overnight is preferred, stain can be prepared up to two hours before being used in this protocol. d. The next day, filter the Oil Red O staining solution in a 10 ml plastic syringe attached to a 0.2 μm syringe filter into a new centrifuge tube. e. Optional. Co-staining samples with DAP will aid in the identification of tissues (germline versus intestine). Dilute 1 mg/ml DAPI stock solution to 10 μg/ml in the Oil Red O staining solution (10 μl DAPI per 1 ml Oil Red O staining solution). 2. Stain C. elegans with Oil Red O staining solution. a. Wash off synchronized worm populations from 6 cm petri dishes with NGM seeded with E. coli OP50-1 with PBS + 0.01% Triton X-100 (PBST) into a 1.5 ml tube. 1. For each strain, you will need ~100 worms. b. Centrifuge for 30 seconds at 500 rpm. Remove supernatant to 0.1 ml (use marks on the tube as a guide). c. Wash 1-2 additional times with PBST. Note: These washes remove E. coli from the worms to be stained. Wash until the supernatant is clear and not cloudy with bacteria. d. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. e. Add 600 μl of 60% isopropanol and rotate for 3 minutes at room temperature to fix worms. f. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. g. Add 600 μl of Oil Red O staining solution and incubate for two hours while rotating at room temperature. h. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. i. Add 600 μl PBST and incubate for 30 minutes while rotating at room temperature. j. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. 172 k. Using a cut pipette tip (to broaden opening), pipette 14 μl of worms in PBST onto microscope slide. 1. If you want to line worms up, do so with the eyelash brush before covering with a cover slip. l. Cover with a cover slip and seal the edges with clear nail polish. Allow to dry for 5-10 minutes. m. Image at 10× with DIC and DAPI filters using the Zeiss AxioCam ERc5s. Representative images are shown in Figure 1. 1. Image at least 100 worms per strain per replicate for qualitative analysis of non- Asdf versus Asdf (age-dependent somatic depletion of fat). C. Nile Red fat staining of post-embryonic animals 1. Wash off synchronized worm populations from 6 cm petri dishes with NGM seeded with E. coli OP50-1 with PBS + 0.01% Triton X-100 (PBST) into a 1.5 ml tube. 1. For each strain, you will need ~100 worms. 2. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml (use marks on the tube as a guide). 3. Wash 1-2 additional times with PBST. Note: These washes remove E. coli from the worms to be stained. Wash until the supernatant is clear and not cloudy with bacteria. 4. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. 5. Add 600 μl of 40% isopropanol and rotate for 3 minutes at room temperature to fix worms. 6. While worms are fixing, make Nile Red staining solution: dilute Nile Red stock solution in 40% isopropanol (6 μl Nile Red stock solution per 1 ml 40% isopropanol). a. Optional. Co-staining samples with DAP will aid in the identification of tissues (germline versus intestine). Add 1 mg/ml DAPI stock solution (10 μl in 1 ml Nile Red staining solution). 173 7. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. 8. Add 600 μl of Nile Red staining solution and incubate for two hours in the dark. Note: Before incubation, flick the bottom of the tube to resuspend the pellet in the staining solution. 9. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. 10. Add 600 μl PBST and incubate for 30 minutes in the dark. 11. Centrifuge for 1 minute at 500 rpm. Remove supernatant to 0.1 ml. 12. Using a cut pipette tip (to broaden opening), pipette 14 μl of worms in PBST onto microscope slide. a. If you want to line the worms up, do so with the eyelash brush before covering with a cover slip. 13. Cover with a cover slip and seal the edges with clear nail polish. Allow to dry in the dark for 5-10 minutes. 14. Image at 10× with DIC, DAPI (Blue) and AlexaFluor488 (Green) filters using the Zeiss AxioCam MRm. Representative images are shown in Figure 2. a. Image at least 50 worms per strain per replicate for quantitative analysis. Note: The Nile Red fat staining is imaged in the GFP/AlexaFluor488 channel and the DAPI is imaged in the DAPI channel. D. Nile Red fat staining of embryonic animals Note: This protocol is similar to the Nile Red fat staining of post-embryonic animals with a few minor differences. 1. Synchronize populations by alkaline hypochlorite treatment. Instead of allowing animals to hatch overnight, aspirate down to 1 ml and transfer M9 buffer with embryos to a 1.5 ml centrifuge tube. 2. Spin down for 1 minute at 5000 rpm. Remove as much as the supernatant as possible without disrupting the pellet. 174 3. Wash sample with PBST. Spin down for 1 minute at 5000 rpm and remove supernatant to 0.1 ml. 4. Add 600 μl of 40% isopropanol and rotate overnight at room temperature to fix embryos. 5. The next day, make Nile Red staining solution: dilute Nile Red stock solution in 40% isopropanol (6 μl Nile Red stock solution per 1 ml 40% isopropanol). a. Add 1 mg/ml DAPI stock solution (10 μl in 1 ml Nile Red staining solution). 6. Centrifuge for 1 minute at 5000 rpm. Remove supernatant to 0.1 ml. 7. Add 600 μl of Nile Red staining solution and incubate for two hours in the dark. Note: Before incubation, flick the bottom of the tube to resuspend the pellet in the staining solution. 8. Centrifuge for 1 minute at 5000 rpm. Remove supernatant to 0.1 ml. 9. Add 600 μl PBST and incubate for 30 minutes in the dark. 10. Centrifuge for 1 minute at 5000 rpm. Remove supernatant to 0.1 ml. 11. Pipette 14 μl of embryos in PBST onto microscope slide. 12. Cover with a cover slip and seal the edges with clear nail polish. Allow to dry in the dark for 5-10 minutes. 13. Image at 40-63× with DIC, DAPI and GFP (AlexaFluor488 – Green) filters using the Zeiss AxioCam MRm. Representative images are shown in Figure 3. a. Image at least 50 eggs per strain per replicate for quantitative analysis. Note: The Nile Red fat staining is imaged in the GFP/AlexaFluor488 channel and the DAPI is imaged in the DAPI channel. Data analysis 1. Download Fiji by ImageJ to open and analyze the .czi files from the Zen imaging software: https://imagej.net/Fiji/Downloads 2. Open Fiji. Open the .czi image file to analyze (File – Open – Select File – Open). 3. The “Bio-Formats Import Options” window will pop up. Format the following way: 175 a. For stacking view, “View stack with: Hyperstack”. b. For color options, “Color mode: Custom”. c. No other options need to be selected. d. Once the stacking view and color options have been updated, press “OK”. 4. The “Console” window will pop up. You can ignore this window. 5. The “Bio-Formats Color Customization” window will pop up. Select a color for each channel used to image. a. Example: i. DIC 1. “Series 0 Channel 0 Red”: 255 2. “Series 0 Channel 0 Green”: 255 3. “Series 0 Channel 0 Blue”: 255 Note: These settings will make the DIC look like it does when you are imaging (no colors added) ii. DAPI 1. “Series 0 Channel 1 Red”: 0 2. “Series 0 Channel 1 Green”: 0 3. “Series 0 Channel 1 Blue”: 255 Note: These settings will make everything imaged in the DAPI channel blue in color. iii. GFP 1. “Series 0 Channel 2 Red”: 0 2. “Series 0 Channel 2 Green”: 255 3. Series 0 Channel 2 Blue”: 0 Note: These settings will make everything imaged in the GFP/AlexaFluor488 channel green in color. 176 6. Press “OK”. Now the image file will open with a slider to move from one channel to another. 7. Open the “ROI Manager” window: Edit – Selection – Add to Manager 8. Open the “Results” window: Analyze – Measure. a. Make sure the “Results” window has the following measurements: “Area”, “Mean”, and “IntDen”. b. If these are not shown in the “Results” window, right click the gray bar in the “Results” window and select “Set Measurements…” c. Select “Area”, “Integrated density”, and “Mean gray value”. Press “OK”. 9. Move to the Nile Red fat staining channel. Select the oval tool and draw a small shape in an area with just the background in it (no worms). Measure the background: Analyze – Measure. Background measurement will show up in the “Results” window. Note: You need a new background measurement for each image analyzed. If you have multiple worms in one image, you can use the same background measurement for all of them. You can use this oval measurement again if you save it in the “ROI Manager” window. a. While the oval is still drawn on the image, select the “Add [t]” option in the “ROI Manager” window. b. Rename to “Background” to remember to use for future images: Highlight number, select “Rename…”, rename the image and press “OK”. 10. Move to the DIC channel. Select the polygon tool. Outline the worm completely. Note: In order to include a worm for analysis, the whole worm must be present in the image. 11. With the worm still outlined, move to the Nile Red fat staining channel. Add the outline to the “ROI Manager” (“Add [t]”). Measure the fluorescence in the outlined worm (Analyze – Measure). 177 12. Close image: File – Close. 13. Repeat for all other images in the analysis. a. Open image: File – Open. b. Move to the Nile Red fat staining channel. c. Background measurement: Select ROI in “ROI Manager” labelled background. Measure background: Analyze – Measure. d. Worm measurement: Move to DIC channel, select polygon tool and outline the worm. Move to the Nile Red fat staining channel, add to the “ROI Manager” (“Add [t]”). Measure the worm fluorescence: Analyze – Measure. e. Close image: File – Close. 14. Once finished with a strain/condition, save the ROIs in the “ROI Manager”: Highlight all of the ROIs and select: “More…” then “Save” and save. 15. Save all the measurements by selecting all measurements and copying to a spreadsheet where the area-corrected worm fluorescent intensity (CTCF) will be calculated. 16. To calculate the CTCF for each image: Multiply the background area by the background mean. Subtract this value from the worm integrated densitiy. Divide by the worm area. The final value is the CTCF. ( )−( ∗ ) 17. To normalize all samples to the control, calculate the average CTCF value for the control samples. Divide all CTCF values (including the controls) by the average control CTCF. Now everything is normalized to the control. 18. Plot CTCF values on a graph. If there are multiple samples, either a t-test or ANOVA can be performed to determine significance between the samples’ fat content. Representative quantification of data is shown in Figure 4. 178 FIGURES Figure 1. Representative images of skn-1gf mutant animals fixed and stained at larval stage 4 (A-B) or day 3 adulthood (C-D) and visualized by DAPI-stained nuclei (A,C) and Oil Red O stained lipids (B,D). scale bar = 50mm 179 Figure 2. Representative images of wild-type (N2 Bristol) (A-C and G-I) and SKN-1gf mutant animals (D-F and J-L) fixed and stained at larval stage 4 (A-F) or day 3 of adulthood (G-L) and visualized by DIC (A,D,G,J), DAPI-stained nuclei (B,E,H,K), and Nile Red stained lipids (C,F,I,L). scale bar = 50mm. 180 Figure 3. Representative images of fixed and stained Wild-type (N2 Bristol) embryos visualized by DIC (A), DAPI-stained nuclei (B), and Nile Red stained lipids (C). scale bar = 10mm 181 Figure 4. One worm has been outlined (in white) in ImageJ for size analysis and lipid density (total fluorescence). scale bar = 50mm. Quanitification of lipid abundance in populations of animals of the indicated genotypes (n=100 for (B) and n=50 for (D); representative imgages found in Figure 2). 182 NOTES Note 1. NGM stock plates should be freshly prepared and freshly seeded for all experiments. Note 2. Staining solutions should be prepared the day of staining (NR) or the day before staining (ORO). Note 3. Always include a control group to make comparisons within each experimental group. Note 4. Although the washing steps use PBS + 0.01% Triton X-100, worms and embryos tend to stick to the side of the tubes. It is recommended to have over double the population required for imaging to compensate for the loss of some of the sample. 183 RECIPES 1. Nematode Growth Media (NGM) plates a. Prepare stock solutions of 1 M MgSO 4, 5 mg/ml cholesterol, 1 M KH2PO4, 1 M CaCl2, and 2.5% (w/v) streptomycin. i. Dissolve 120.366 g of MgSO4 per 1 L water (1 M). Filter sterilize. ii. Dissolve 5 g of cholesterol per 1 L 100% ethanol (5 mg/ml). Store at 4 °C. iii. Dissolve 136.086 g of KH2PO4 per 1 L water (1 M). Filter sterilize. iv. Dissolve 110.98 g of CaCl2 per 1 L water (1 M). Filter sterilize. v. Dissolve 25 g of streptomycin sulfate per 1 L water (2.5%). Filter sterilize. Store at 4 °C. b. Add 3 g of NaCl, 17 g of agar, 2.5 g of peptone, and 950 ml of water to a glass flask. Include a stir bar. c. Cover with foil and autoclave to sterilize and dissolve agar. d. Cool, with stirring, to 55-60 °C. e. Add 1 ml of 1 M MgSO4, 1 ml of 5 mg/ml cholesterol, 25 ml of 1 M KH2PO4,, 1 ml of 1 M CaCl2, and 7.5 ml of 2.5% streptomycin. f. Stir 15 minutes. g. Dispense 11 ml into 6 cm Petri dishes using sterile technique. h. Allow to solidify overnight. i. Seed stock NGM plates with 250 μl of OP50-1 overnight culture (grown in LB + streptomycin, without shaking). j. Allow plates to dry (covered) 2-3 days before use. 2. Phosphate buffered saline (PBS) a. Add 26.5 g of Na2HPO4·H2O, 80 g of NaCl, 2 g of KCl, and 2 g of KH2PO4 to a glass bottle. b. Bring volume to 1 L with water. 184 c. Sterilize by autoclaving. 3. PBS + 0.01% Triton X-100 (PBST) a. Using a cut pipette tip (to broaden opening), add 100 μl of Triton X-100 to 1 L sterile PBS. b. Mix well (do not shake). 4. M9 a. Add 30 g of KH2PO4, 60 g of Na2HPO4, 50 g of NaCl, and 120 mg of MgSO4 to a glass bottle. b. Bring volume to 1 L with water. c. Sterilize by autoclaving. Precipitation of calcium salts may occur after autoclaving. Re-dissolving may take several days. Agitation helps. 5. DAPI stock solution a. Dissolve 1 mg/ml in DMSO. 6. Oil Red O stock solution a. Add 1.5 g Oil Red O to 300 ml 100% isopropanol in a glass bottle for a final concentration of 5 mg/ml. Add a stir bar. b. Let stir overnight. Store at room temperature in the dark (can be stored for months). 7. Nile Red stock solution a. Cover a glass bottle with foil. b. Add 100 mg Nile Red for microscopy to 200 ml 100% acetone in a glass bottle. Add a stir bar. c. Let stir overnight in the dark. Store at room temperature in a dark place (can be stored for months). 185 ACKNOWLEDGEMENTS Protocol is derived from the original research paper, Nhan et al. “Redirection of SKN-1 abates the negative metabolic outcomes of a perceived pathogen infection” Proc Natl Acad Sci U S A. 2019 Oct 29;116(44):22322-22330. doi: 10.1073/pnas.1909666116 13 . This work was funded by the NIH R01AG058610 to S.P.C., T32AG052374 to N.L.S and B.V.C., T32GM118289 to N.L.S., and T32AG000037 to J.D.N. and A.M.H. Competing interests: The authors declare no competing interests. Ethics: No human or vertebrate animal subjects are used in this study. 186 Chapter 7: Different methods of killing bacteria diets differentially influence Caenorhabditis elegans physiology *This chapter is a version of a manuscript submitted to micropublication Biology Authors: Nicole L. Stuhr 1,2 and Sean P. Curran 2,* Affiliations: 1. Department of Molecular and Computational Biology, University of Southern California, Los Angeles, CA. 90089 2. Leonard Davis School of Gerontology, University of Southern California, Los Angeles, CA. 90089 Correspondence: *, correspondence to spcurran@usc.edu 187 ABSTRACT Across species, diet plays a critical role in most, if not all life history traits. Caenorhabditis elegans is an important and facile organism for research across modalities, but the use of live bacteria as sources of nutrition can exert pleiotropic outcomes that stem from the action of host-pathogen defenses. Recently, a powerful new approach to readily generate dead and metabolically inactive Escherichia coli was developed that enabled reproducible measures of health across the lifespan. Here we further characterize additional comparisons of developmental and physiological parameters of animals fed either bacteria killed by treatment with ultraviolet (UV) light and bactericidal antibiotics or low-dose paraformaldehyde (PFA). Unlike bacteria killed by UV/Antibiotic treatment, PFA-killed diets resulted in a 25% reduction in body size just prior to adulthood and an overall reduction in stored intracellular lipids. Moreover, a small but reproducible number of animals fed PFA-killed bacteria display age-dependent depletion of somatic lipids, which does not normally occur on live bacteria or bacteria killed by UV/antibiotics. Lastly, animals fed PFA-treated, but not UV-antibiotic treated bacteria display a 20% increase in crawling speed. Taken together, these new more thoroughly define the physiological impact two methodologies to prepare C. elegans diets that should be considered during experimental design. 188 DESCRIPTION The metabolic activity of live bacteria as a food source can significantly impact physiological outcomes in animals fed these diets, particularly in pharmacological studies and when the goal is to understand the complexities of metabolism and nutrition 147,148 .. Although multiple methods have previously been reported several techniques have limitations that confound the interpretation of the data; for example, heat-killed bacteria can be difficult for C. elegans to ingest and results in calorie restricted state that changes developmental timing 148 , UV irradiated bacteria is low throughput and results in incomplete and inconsistent killing 149 , yet still can influence the lifespan of C. elegans 147,150,151 , and antibiotic treatment can differ based on the bactericidal or bacteriolytic nature of the drug and also results in altered lifespan 147,150 . The effects on lifespan are intriguing but add an additional variable of host-pathogen responses that can make uncoupling the effects of genetics, pharmacological treatment, or nutritional changes more difficult. A recent study optimized an effective and reproducible method for killing bacteria using paraformaldehyde (PFA) at low doses 152 . This study highlights not only the consistency of killing bacteria, but also demonstrates that PFA-killed bacteria do not replicate but are also metabolically inactive. Importantly, despite PFA-killed E. coli slightly delaying larval development to adulthood, other life history traits like reproduction and lifespan, are not changed 152 , suggesting worms are healthy when fed this preparation of food. One parameter missing from this otherwise thorough investigation of the impact that PFA-killed E. coli has on C. elegans health is an assessment of lipid homeostasis and additional healthspan measures such as movement speed; here we provide additional data to fill this knowledge gap. We investigated these parameters in C. elegans fed an E. coli diet killed by UV/antibiotic exposure as compared to a low dose (0.25%) PFA treatment. We confirmed the slight developmental delay reported in animals fed PFA-killed OP50, but subsequently compared animals at the L4 development stage for overall body size. We found that when compared to animals fed live OP50 and UV/antibiotic killed OP50, animals raised on PFA-killed E. coli OP50 were significantly smaller at the L4 larval stage (Figure 1A). We next performed fixed Nile red staining to quantify the abundance of intracellular lipid pools, which revealed animals fed PFA-killed OP50 stored less lipids relative to animals fed live bacteria (Figure 1B-C). Notably, 189 the magnitude of this difference is more significant when compared to animals fed UV/antibiotic-killed bacteria, which display a modest increase in intracellular lipid pools. We confirmed that the growth and lipid storage phenotypes were not due to the presence of residual PFA in the preparation of the bacterial, because the decrease in body size and lipid content persists when the PFA-killed bacteria are prepared with additional washes with M9 buffer. Bacterial diets can also influence where lipids are distributed between somatic and germ tissues 13,15 , and as such, we also performed Oil-red-O staining to qualitatively assess the location of lipid stores in animals with age. Worms fed PFA-killed E. coli display a small but significant increase in the incidence of somatic depletion of fat, while maintaining germline lipid pools at day 4 of adulthood. This result is intriguing because depletion of somatic lipid stores can also occur when animals are exposed to pathogenic bacteria 13 . Taken together, these data reveal that suggests that despite PFA-killed bacteria being metabolically inactive, when provided as the sole source of nutrition to C. elegans it maintains the capacity to induce metabolic change. Finally, we tested how the different dead bacterial diets can influence overall health by measuring movement speed as a surrogate for muscle function 112 . We measured the average speed of wildtype worms at the L4 stage on plates without E. coli after being raised on either live OP50 or killed OP50 from the L1 to L4 larval stages (Figure 1E-F). We found that raising worms on PFA-treated OP50 led to a significant increase in average crawling speed (Figure 1E). This increase in crawling speed was not observed in animals raised on the UV/Antibiotic-killed OP50 (Figure 1F). Taken together, our results show that metabolic activity of the bacterial diet can influence C. elegans metabolism and healthspan. Importantly, these are two phenotypes that have not yet been addressed in the field live and dead bacterial diet studies of C. elegans 147,150,152,153 . These results fill knowledge gaps in the field and emphasize the importance of bacterial respiration on C. elegans physiology which should be taken into consideration for study design. 190 METHODS C. elegans strains and maintenance C. elegans were raised on 6 cm nematode growth media (NGM) plates supplemented with streptomycin and seeded with live OP50. For experiments, nematode growth media plates without streptomycin were seeded with live OP50, UV/Antibiotic-killed OP50 or PFA-killed bacteria. The N2 Bristol (wildtype) worm strain was grown at 20°C and unstarved for at least three generations before being used. Some strains were provided by the CGC, which is funded by NIH Office of Research Infrastructure Programs (P40 OD010440). Killed-OP50 Treatments After culturing, 400 mL of live bacteria was aliquoted into 1000mL Erlenmeyer flasks. The first flask contained live OP50. 10% PFA was added to the second flask to get the desired final concentration of PFA (0.25% PFA). PFA-treated bacteria was shaken at 37 °C in a shaking incubator at 200 rpm for 2 hours to allow for mixing of PFA and sufficient exposure. 4% kanamycin and 4% ampicillin was added to the culture and incubated in a shaking incubator at 200 rpm for 12 hours. All conditions were then spun down and resuspended in M9 at 1x concentration for live OP50 and 5x concentration for PFA-treated and UV/Antibiotic-treated OP50. PFA-treated bacteria was washed one additional time in M9 before resuspension. The “PFA Wash” condition had two subsequent wash steps in M9 before resuspension at 5x concentration. UV/Antibiotic-killed OP50 was then exposed to 25000 mJ of UV before all bacteria were seeded on NGM plates and left to dry for 48 hours before the experiments. Nile Red Staining Nile Red fat staining was conducted as outlined in Stuhr et al. 2022 77 . In brief, worms were egg prepper and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria and raised to 48 h (L4 stage). Worms were washed off plates with PBS+triton, rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with Nile Red in 40% isopropyl alcohol for 2 h. Worms were pelleted after 2 h and washed in PBS+triton for 30 min before being images at 10X magnification with ZEN Software and Zen Axio Imager with the DIC and GFP filter. Fluorescence is 191 measured via corrected total cell fluorescence (CTCF) via ImageJ and Microsoft Excel. CTCF = Worm Integrated Density-(Area of selected cell X Mean fluorescence of background readings) and normalized to the control. Oil Red O Staining Oil Red O fat staining was conducted as outlined in Stuhr et al. 2022 154 . In brief, worms were egg prepped and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria and raised to 120 h (Day 3 Adult stage). Worms were washed off plates with PBS+triton, then rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with ORO in diH2O for 2 h. Worms were pelleted after 2 h and washed in PBS+triton for 30 min before being imaged at 20x magnification with LAS X software and Leica Thunder Imager flexacam C3 color camera. Lipid distribution ORO-stained worms were placed on glass slides and a coverslip was placed over the sample. Worms were scored, as previously described in Stuhr et al. 2022 154 ; the age-dependent somatic depletion of fat (Asdf) phenotype is characterized by the loss of detection of ORO-stained lipids in the soma but presence of lipids in the germline. Worms were scored and images were taken with LAS X software and Leica Thunder Imager flexacam C3 color camera. Fat levels of worms were placed into two categories: non-Asdf and Asdf. Non- Asdf worms display no loss of fat and are stained dark red throughout most of the body (somatic and germ cells). Asdf worms had most, if not all, observable somatic fat deposits depleted (germ cells only) or significant fat loss from the somatic tissues with portions of the intestine being clear (somatic < germ). Movement Measurements – Crawling Worms were egg prepped and eggs were allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria. Worms were then allowed to grow until each time point (48 h post-drop for L4s). Once worms were the required stage of development, 30-50 worms were washed off of a plate in 50 uL of M9 with a M9+triton coated P1000 tip and dropped onto an unseeded NGM plate. The M9 was allowed to dissipate, and worms roamed on the unseeded plate for 1 hour before 192 imaging crawling. Crawling was imaged with the MBF Bioscience WormLab microscope and analysis was performed with WormLab version 2022. Worm crawling on the plate was imaged for 1 minute for each condition at 7.5 ms. Worm crawling was analyzed with the software and only worms that moved for at least 90% of the time were included in the analysis. 193 FIGURES Figure 1. Differential effects of PFA-killed and UV/Antibiotic-killed E. coli on C. elegans physiology. The method of killing the C. elegans OP50 diet prior to feeding can differentially effect body size (A), stored intracellular lipids as visualized by fixed Nile red staining (B) normalized to body area (C), distribution of stored lipid pools (D), and animal movement speed in early life (E-F). Live PFA PFA Wash UVAb A. B. E. F. C. D. 194 ACKNOWLEDGEMENTS We thank S Ledgerwood for technical assistance. This work was funded by the NIH R01AG058610 to SPC, F31AG077873 to NLS, and T32AG052374 to NLS. Author Contributions: Conceptualization: SPC; Methodology: NLS and SPC; Investigation: NLS and SPC; Visualization: NLS and SPC; Supervision: SPC; Writing (original draft): NLS and SPC; Writing (reviewing & editing): NLS and SPC Competing interests: All authors declare that they have no competing interests 195 Chapter 8: Ether Lipid Biosynthesis Promotes Lifespan Extension and Enables Diverse Prolongevity Paradigms in Caenorhabditis elegans *This chapter is a version of a manuscript under review at eLife. This project was done in collaboration with Alexander Soukas’ group, investigating how ether lipid biosynthesis effects lifespan and metabolism in C. elegans. I was responsible for data collected in Figure 6, including experimental design and writing the results, discussion, and methods section. Authors: Lucydalila Cedillo 1,2,3 , Fasih M. Ahsan 1,2,3 , Sainan Li 1,2 , Nicole L. Stuhr 7 , Yifei Zhou 1,2 , Yuyao Zhang 1,2 , Adebanjo Adedoja 1,2,3 , Luke M. Murphy 1,2,3 , Armen Yerevanian 1,2 , Sinclair Emans 1,2 , Khoi Dao 4 , Zhaozhi Li 5 , Nicholas D. Peterson 6 , Jeramie Watrous 4 , Mohit Jain 4 , Sudeshna Das 5 , Read Pukkila-Worley 6 , Sean P. Curran 7 and Alexander A. Soukas 1,2, * Affiliations: 1 Center for Genomic Medicine and Diabetes Unit, Endocrine Division, Department of Medicine, Massachusetts General Hospital and Harvard Medical School, Boston MA 02114 2 Broad Institute of Harvard and MIT, Cambridge, MA 02142 3 Program in Biological and Biomedical Sciences, Division of Medical Sciences, Harvard Medical School, Boston, MA 02114 4 Department of Medicine and Pharmacology, University of California San Diego, San Diego, CA 92093 196 5 Biomedical Informatics Core, Massachusetts General Hospital and Harvard Medical School, Cambridge, MA 02139 6 Program in Innate Immunity, Division of Infectious Diseases and Immunology, University of Massachusetts Medical School, Worcester, MA, 01655 7 Leonard Davis School of Gerontology, University of Southern California, Los Angeles, CA, 90089 For Correspondence: Alexander A. Soukas Email: asoukas@mgh.harvard.edu Keywords: metformin; aging; ether lipids; C. elegans; metabolism 197 ABSTRACT Biguanides, including the world’s most prescribed drug for type 2 diabetes, metformin, not only lower blood sugar, but also promote longevity in preclinical models. Epidemiologic studies in humans parallel these findings, indicating favorable effects of metformin on longevity and on reducing the incidence and morbidity associated with aging-related diseases. Despite this promise, the full spectrum of molecular effectors responsible for these health benefits remains elusive. Through unbiased screening in C. elegans, we uncovered a role for genes necessary for ether lipid biosynthesis in the favorable effects of biguanides. We demonstrate that biguanides prompt lifespan extension by stimulating ether lipid biogenesis. Loss of the ether lipid biosynthetic machinery also mitigates lifespan extension attributable to dietary restriction, target of rapamycin (TOR) inhibition, and mitochondrial electron transport chain inhibition. A possible mechanistic explanation for this finding is that ether lipids are required for activation of longevity-promoting, metabolic stress defenses downstream of the conserved transcription factor Nrf/skn-1. In alignment with these findings, overexpression of a single, key, ether lipid biosynthetic enzyme, fard-1/FAR1, is sufficient to promote lifespan extension. These findings illuminate the ether lipid biosynthetic machinery as a novel therapeutic target to promote healthy aging. 198 INTRODUCTION Metformin is the first line therapy for type 2 diabetes and the most frequently prescribed oral hypoglycemic medication worldwide 155 . Human epidemiologic studies note an association between metformin use and decreased incidence of cancer 156,157 . In addition, metformin extends lifespan in invertebrate and vertebrate models 158-160 , and therefore may reduce aging-related diseases in humans 161 . Nonetheless, our understanding of the molecular pathways governing the health-promoting effects of metformin is only just beginning to emerge. Our previous work identified a conserved signaling axis connecting mitochondria, the nuclear pore complex, and mTORC1 inhibition that is required for metformin-mediated extension of lifespan in C. elegans and inhibition of growth in worms and human cancer cells 162 . The energy sensor AMP-activated protein kinase (AMPK) is not necessary for metformin-induced growth inhibition in C. elegans but is required for the drug’s pro-longevity effects 158,160,163 . Consistently, mechanistic studies indicate that the longevity-promoting transcription factor SKN-1/Nrf is required for biguanide-mediated lifespan extension 158,160 . The relationship of these metformin longevity response elements to each other and their hierarchy in the biguanide response relay remains unknown. Thus, the mechanisms by which metformin exacts its beneficial effects on health are likely to be branching and complex. The importance of ether lipids, a major structural component of cell membranes, to aging and longevity is not fully established. Ether lipids are involved in the maintenance of general membrane fluidity and in the formation of lipid rafts within microdomains, which are important for promotion of membrane fusion and cellular signaling 164-166 . Ether lipids have broad roles in the regulation of cell differentiation 167-170 , cellular signaling 171,172 , and reduction of oxidative stress through their action as antioxidants 173-176 . Humans deficient in ether lipid biogenesis suffer from rhizomelic chondrodysplasia punctata (RCDP), a rare genetic disorder, which results in skeletal and facial abnormalities, psychomotor retardation, and is uniformly fatal typically before patients 199 reach their teenage years 177 . Thus, current evidence linking alterations in ether lipid levels to aging and longevity in humans is strictly correlative 178,179 . Ether lipids, which are structurally distinct from canonical phospholipids, have a unique biosynthetic pathway through which a fatty alcohol is conjugated to the glycerol backbone at the sn-1 position via an ether linkage. Ether lipid precursors are first synthesized by enzymes associated with the membranes of peroxisomes 180-182 . The main enzymes involved in ether lipid biosynthesis within the peroxisomal matrix are glyceronephosphate O-acyltransferase (GNPAT) and alkylglycerone phosphate synthase (AGPS). Fatty acyl-CoA reductase 1 (FAR1) supplies most of the fatty alcohols used to generate the ether linkage in the precursor, 1-O-alkyl-glycerol- 3-phosphate. This precursor is then trafficked to the endoplasmic reticulum (ER) for acyl chain remodeling to produce various ether lipid products 183 . In C. elegans, loss-of-function mutations of any of the three main enzymes involved in human ether lipid biosynthesis, acl-7/GNPAT, ads- 1/AGPS, and fard-1/FAR1, results in an inability to produce ether linked lipids, as in humans, and has been reported to shorten lifespan 184,185 . Worms and human cells deficient in ether lipids exhibit compensatory changes in phospholipid species, including increases in phosphatidylethanolamines and phosphatidylcholines containing saturated fatty acids 169,186 . However, in contrast to humans, ether lipid deficient nematodes develop to adulthood at a normal rate, providing an opportunity to determine the biological roles of ether lipids in aging and longevity without pleiotropies associated with developmental rate. Here, we show that the ether lipid biosynthetic machinery is necessary for lifespan extension stimulated by metformin or the related biguanide phenformin in C. elegans. Metabolomic analysis indicates that phenformin treatment drives increases in multiple phosphatidylethanolamine-containing ether lipids in a manner independent of bacterial-host interactions. Interestingly, requirement for the ether lipid biosynthetic genes extends to multiple genetic longevity paradigms, including defective mitochondrial electron transport function (isp-1), defective pharyngeal pumping/ caloric restriction (eat-2), and compromises in mTOR complex 1 200 activation (raga-1). We show that overexpressing FARD-1, the enzyme that supplies all the fatty alcohols for ether lipid biogenesis in C. elegans, extends lifespan, supportive of the idea that alterations in the ether lipid landscape alone is sufficient to promote healthy aging. Mechanistically, ether lipids promote longevity downstream of biguanide action through activation of metabolic stress defenses and somatic lipid redistribution through the transcription factor SKN- 1/Nrf. These data suggest that a heretofore unappreciated role for ether lipids is to promote organismal-level, longevity-promoting stress defenses. 201 RESULTS Genes responsible for ether lipid biosynthesis are necessary for biguanide-induced lifespan extension A prior screen of ~1000 metabolic genes for RNA interference (RNAi) knockdowns that interfere with the growth-inhibitory properties of a supraphysiologic 160 mM dose of metformin in C. elegans (utilized to maximize the sensitivity and specificity of our assay to identify true epistatic candidates) 162 , yielded fard-1 and acl-7, which are required for ether lipid biosynthesis. Ether lipids are distinguished from canonical phospholipids, as the latter contain exclusively fatty acids conjugated to glycerol, whereas ether lipids contain a fatty alcohol conjugated to the glycerol backbone at the sn-1 position via an ether linkage (Figure 1A). Confirming our screen results, granular quantitative analysis following RNAi knockdown of fard-1 and acl-7 reveal significant resistance to biguanide-induced growth inhibition. (Figure 1—figure supplement 1A). Our lab has previously demonstrated that biguanide effects on growth in C. elegans share significant overlap mechanistically with the machinery by which metformin extends lifespan in the worm, thus suggesting that modulation of ether lipid biosynthesis may also be responsible for the lifespan- extending properties of the drug 162 . Indeed, loss-of-function mutations in any of three genes encoding enzymes required for ether lipid biosynthesis, fard-1, acl-7 or ads-1, significantly abrogates lifespan extension induced by lifespan extending doses of metformin (50 mM) and the related biguanide phenformin (4.5 mM) (Figure 1B—G). Loss-of-function of ads-1 and acl-7 may display a modest increase in lifespan with metformin administration but display a percentage median lifespan increase significantly reduced in comparison to wildtype controls. (Figure 1B— G, and throughout manuscript see Supplementary file 1 for all tabular survival statistics and biological replicates). Confirming that these mutations confer resistance to metformin by compromising ether lipid synthetic capacity, RNAi knockdowns of fard-1 and acl-7 in wildtype worms also partially impair lifespan extension promoted by phenformin (Figure 1—figure supplement 1B—C). This dependency is not confounded by chemical inhibition of reproduction, 202 as lifespan analyses performed without the use of the thymidylate synthase inhibitor 5-fluoro-2′- deoxyuridine (FUdR) reveal similar abrogation of biguanide-mediated lifespan extension with inactivation of the ether lipid synthesis machinery (Figure 1—figure supplement 2A—F) 187 . Studies from this point forward are presented predominantly with phenformin because phenformin is more readily absorbed without need for a specific transporter, unlike metformin 162,188,189 , and our experience indicates more consistent lifespan extension with phenformin in C. elegans. Because ether lipids are a major structural component of cell membranes, one possibility is that deficiencies in ether lipid synthesis compromises drug action by reducing biguanide bioavailability in the worm. To test this, we compared the relative levels of biguanides present in vehicle- and biguanide- treated wild type animals to the three ether lipid synthesis mutants by liquid chromatography—tandem mass spectrometry (LC-MS/MS). A comparison of normalized concentrations of phenformin across all four strains shows that phenformin abundance is quantitatively similar across wild type and the three ether lipid mutant strains (Figure 1H and Figure 1—figure supplement 1D). Similar results were obtained when comparing levels of metformin wild type and ether lipid mutant animals (Figure 1I and Figure 1—figure supplement 1E). Thus, a deficiency in ether lipid synthesis does not significantly impact levels of biguanides in C. elegans. Phenformin induces changes in ether lipid levels We reasoned that if biguanides require ether lipid biosynthesis to promote lifespan extension, that phenformin may promote synthesis of one or more ether lipids. To investigate the impact of biguanides on ether lipids at a high level, we first utilized gas chromatography-mass spectrometry (GC-MS) analysis. We first recapitulated the observation that fard-1 mutants show absence of 18- carbon containing fatty acid derivatives (dimethylacetals, or DMAs, which indicate alkenyl ether lipid or plasmalogen levels) and an accumulation of stearate (18:0) relative to wild type controls 203 by GC-MS (Figure 2A—B) 185 . We then asked if phenformin impacts the levels of 18-carbon alkenyl ether lipids in wild type animals and if those corresponding changes are absent in fard-1 mutants. Strikingly, phenformin-treated wild type worms display a significant increase in 18:0 DMA versus vehicle, whereas no such increase is evident in drug treated fard-1 worms (Figure 2C). In addition, relative proportions of stearic acid (18:0) levels within the total fatty acid pool are significantly increased in fard-1 mutants treated with phenformin versus vehicle treated fard-1 controls (Figure 2D). In comparison, the relative proportion of stearic acid does not rise in phenformin treated wild type animals, suggesting that stearate is being utilized for ether lipid production. Analysis of the total fatty acid pool by GC-MS (Figure 2—figure supplement 1) indicates that aside from several fatty acids (e.g. 18:2), the most pronounced differences were in the plasmalogen pool. In alignment, an assessment of levels of additional alkenyl fatty alcohols in phenformin-treated, wild type animals indicate a parallel, significant increase in the less abundant 16:0 DMA and 18:1 DMA species (Figure 2E). We conclude that phenformin treatment leads to an overall increase of alkenyl ether lipid levels in C. elegans. To investigate relative changes in individual ether lipid abundance in response to phenformin at high resolution, we utilized LC-MS/MS analysis. Using this method, we detected 20 alkyl and alkenyl phosphatidylethanolamine-based ether lipids previously noted to be the most abundant ether lipids in C. elegans 184,185 (Figure 2F—G and Figure 2-source data 1). This analysis indicates that phenformin treatment results in a significant increase in normalized abundance of four ether lipids, PE(O-16:0/18:1), PE(O-18:0/18:3), PE(O-18:0/20:2), and PE(P-18:1/18:1), even when corrected for multiple hypothesis testing. Most ether lipids measured display mean levels that increase with phenformin treatment, though most are either nominally significant or exhibit a nonsignificant trend because of the strict threshold required to reach significance when correcting for multiple hypotheses. Finally, phosphatidylethanolamine ether lipid abundances were extremely low in fard-1, acl-7 and ads-1 mutants and unchanged by phenformin treatment, unlike 204 in wild type animals (Figure 2F and Figure 2-source data 1). In aggregate, these data indicate that phenformin treatment leads to increased abundance of multiple ether lipid species in C. elegans. Peroxisomal ether lipid synthesis is essential to the biological action of phenformin In order to begin to understand the governance of ether lipid biosynthesis by biguanides, we examined the expression of a C. elegans FARD-1::RFP translational reporter, under the control of its own promoter (Figure 2—figure supplement 2A). Exogenously expressed FARD-1 (fard-1 oe1) is expressed in the intestine and localizes near structures resembling lipid droplets by Nomarski microscopy (Figure 2—figure supplement 2B). Given that ether lipid biogenesis occurs between peroxisomes and the endoplasmic reticulum (ER) 180-183 , we crossed this FARD-1::RFP reporter to an animal bearing a GFP reporter that illuminates peroxisomes in the intestine (GFP fused to a C-terminal peroxisomal targeting sequence 1 (PTS1)) to determine if localization of FARD-1 is regulated by biguanides. FARD-1 does not possess a predicted C-terminal PTS1, similar to ACL-7 and ADS-1. At baseline, FARD-1::RFP fluorescence partially overlaps with peroxisomally-targeted GFP (Figure 2—figure supplement 2C). Colocalization analysis indicates that treatment with phenformin does not change the amount of overlap between FARD-1::RFP and GFP::PTS1 relative to vehicle treated controls (Figure 2—figure supplement 2D). To confirm our earlier observation that suggests FARD-1 colocalization with lipid droplets, we used confocal imaging to assess the spatial distribution of an integrated FARD-1::RFP reporter (fard-1 oe3) in C. elegans fed C1-BODIPY-C12 to label lipid droplets (and treated with glo-4 RNAi to remove BODIPY-positive lysosome-related organelles) 190-192 We found that FARD-1::RFP fluorescence directly surrounds some, but not all, BODIPY-positive lipid droplets in the worm intestine (Figure 2—figure supplement 2E). However, as with peroxisomes, phenformin does not alter the number of lipid droplets that are surrounded by FARD-1 or its distribution around lipid droplets (data not shown). Finally, FARD-1::RFP localizes into web-like structures in the fard-1(oe3) reporter that 205 may represent smooth endoplasmic reticulum versus another cellular tubular vesicular network (Figure 2—figure supplement 2F), and this localization is also not altered by biguanide treatment. Thus, the regulation of ether lipid biosynthesis does not appear to be via differential localization of FARD-1. Expression of mRNAs encoding FARD-1, ACL-7, and ADS-1 are all decreased or unchanged in abundance upon treatment with biguanide via quantitative RT-PCR (Figure 2—figure supplement 2G—L), suggesting that ether lipids are not increased in phenformin treatment through a transcriptional mechanism. A parallel decrease in overall levels of FARD-1::RFP protein of fard- 1(oe1) transgenics was seen with phenformin treatment (Figure 2—figure supplement 2M). These seemingly paradoxical data are likely consistent with post-translational negative feedback of ether lipids on the ether lipid biosynthetic pathway, as has been previously reported 193 . To affirm that the peroxisome is an essential site of ether lipid production in biguanide action, we disrupted peroxisomal protein targeting and examined phenformin-stimulated lifespan extension. Indeed, either prx-5 or prx-19 RNAi impair lifespan extension prompted by phenformin fully or partially, respectively (Figure 3A—B). PRX-5 is involved in protein import into the peroxisomal matrix and PRX-19 is involved in proper sorting of proteins for peroxisomal biogenesis. Thus, either disruption of ether lipid biosynthetic machinery or of a principal site of ether lipid biosynthesis impairs phenformin’s pro-longevity benefit. Biguanide-mediated ether lipid synthesis is necessary for a pro-longevity benefit irrespective of bacterial growth or metabolism Previous studies into the biological action of metformin have suggested that biguanides mediate their lifespan extending properties in the nematode through alterations in growth and metabolism of their bacterial food source 194,195 . To evaluate whether ether lipid synthesis is induced by 206 biguanides through a direct effect in the worm or via alterations in bacterial-host dynamics, we leveraged a robust, established methodology to chemically kill and metabolically inactivate the C. elegans OP50-1 food source prior to seeding on NGM plates. 1% paraformaldehyde was used to completely kill OP50-1 cultures prior to seeding, confirmed through bacterial titer analysis (Figure 2 – figure supplement 3A). Wildtype adult day 1 nematodes treated with phenformin and grown in the metabolically inactive OP50-1 food source reveal that biguanides can completely reduce somatic free fatty acid levels irrespective of OP50-1 growth status (Figure 2 – figure supplement 3B-C). Notably, despite this significant reduction of fatty acids, biguanides preferentially protect levels of FA 16:0 DMA and FA 18:1 DMA, again irrespective of bacterial growth status (Figure 2 – figure supplement 3-D-E). Thus, we conclude that biguanides increase/protect ether lipid levels through direct action in the nematode, rather than through indirect effects in the bacterial food source. We then hypothesized that disruption of ether lipid biosynthesis may also abrogate biguanide-mediated lifespan extension irrespective of bacterial growth and metabolism. Indeed, ads-1 deficient mutant nematodes completely blunt both metformin and phenformin-mediated lifespan extension irrespective of bacterial growth conditions (Figure 2 – figure supplement 4A- F). Combined, these data suggest that biguanides increase ether lipid levels and require activated ether lipid machinery to exert pro-longevity benefits through direct action in the nematode. Fatty acid elongases and desaturases are positive effectors of biguanide-mediated lifespan extension Most mature ether lipid species contain a fatty acid in the sn-2 position linked by an ester bond 196 . The majority of fatty acids conjugated in ether lipids are largely synthesized endogenously in C. elegans by fatty acid desaturases and fatty acid elongases 197,198 (Figure 3C). Thus, we hypothesized that some of these desaturases and elongases may also contribute mechanistically to biguanide-mediated lifespan extension. Indeed, RNAi knockdown of three fatty acid desaturases and two fatty acid elongases in phenformin-treated C. elegans blunted phenformin- 207 stimulated lifespan extension relative to empty vector controls (Figure 3D—H). Notably, these five genes all contribute to the production of fatty acids 18-20 carbons in length with three or more double bonds. Although knockdown of fatty acid desaturases and elongases in C. elegans results in inherent lifespan extension on vehicle relative to wild type controls on empty vector RNAi as has been previously reported 199,200 , RNAi knockdown of fat-3, fat-4, elo-1, and elo-2 mitigate phenformin-driven lifespan extension (Figure 3E—H). RNAi knockdown of fat-1 reduces the percentage lifespan extension with biguanide treatment to ~17.5% (as compared to ~55.5% for EV RNAi controls), although the maximal lifespan extension of fat-1 RNAi treated animals with phenformin is non-significant compared to EV control (Figure 3D). These results suggest the tantalizing possibility that specific fatty acid desaturases and elongases promote biguanide- mediated lifespan extension through contribution of long and polyunsaturated fatty acids to the synthesis of ether lipids, though a mechanistically distinct role is also possible. Genes involved in ether lipid biosynthesis are required in multiple longevity paradigms Given the critical role of ether lipids in the response to biguanides, we hypothesized that these molecules may also play a broader role in diverse longevity paradigms involving metabolic or nutrient-sensing pathways. C. elegans mutant strains that exhibit 1) reduced mitochondrial function (isp-1), 2) disrupted mTORC1 signaling (raga-1), 3) abnormal pharyngeal pumping resulting in a dietary restricted-like state (eat-2), or 4) inhibition of insulin/insulin-like growth factor-1 signaling (daf-2), all result in extension of lifespan 201-204 . To determine whether requirement for the ether lipid biosynthetic machinery in aging generalizes to these other lifespan extension paradigms, we knocked down all three ether lipid biosynthetic enzymes by RNAi in wild type C. elegans and four long-lived genetic mutants: raga-1, isp-1, eat-2, and daf-2. Knockdown of fard-1, acl-7, and ads-1 by RNAi results in suppression of lifespan extension in isp-1, raga-1, and eat-2 mutants (Figure 4A—C). However, knockdown of ether lipid synthesis genes by RNAi did not impact lifespan extension in daf-2 mutants (Figure 4—figure supplement 1A). Thus, the 208 ether lipid biosynthetic machinery plays a broad role in lifespan extension, and, importantly, does not non-selectively shorten lifespan by making animals generally unfit. Overexpression of fard-1 is sufficient to promote lifespan extension To determine whether stimulation of ether lipid biosynthesis is sufficient to prompt lifespan extension, we tested the effect of overexpression (oe) of the sole C. elegans fatty acid reductase that synthesizes fatty alcohols for ether lipid biogenesis, fard-1, on lifespan. Strikingly, fard-1(oe1) alone significantly extends lifespan (Figure 5A). This result is similar in a second, independent fard-1(oe2) transgenic line (Figure 5B). To confirm that fard-1(oe) lifespan extension is dependent upon ether lipid biosynthesis, we knocked down fard-1, acl-7, and ads-1 by RNAi in the fard- 1(oe1) transgenic strain. As predicted, knockdown of three ether lipid biosynthetic enzymes leads to significant suppression of fard-1(oe1) lifespan extension (Figure 5C and Figure 4—figure supplement 1B—C). To determine whether lifespan extension attributable to fard-1(oe) shares genetic dependencies with biguanide-mediated longevity, we independently knocked down skn-1/Nrf, aak-2/AMPK and daf-16/FoxO by RNAi in a fard-1(oe) background. While skn-1/Nrf and aak-2/AMPK have previously been demonstrated to be necessary for metformin-stimulated lifespan extension, daf- 16/FoxO has not 160,205 . Lifespan extension attributable to fard-1(oe1) is suppressed by these three gene knockdowns (Figure 5D—F), indicating that it is mechanistically similar, but not identical, to biguanide-mediated lifespan extension 158,160 . In line with this finding, we observe that both phenformin treatment and FARD-1 overexpression in fard-1(oe3) animals non-additively reduce endogenous mRNA expression of fard-1 by ~40% (Figure 5G), while fard-1(oe3) animals display an exaggerated increase in exogenous fard-1 expression that is decreased ~60% by phenformin treatment (Figure 5H). This suggests that while biguanides may share mechanistic overlaps with FARD-1 overexpression in post-translational negative feedback regulation of 209 endogenous fard-1 mRNA expression, exogenous overexpression of FARD-1 through both multicopy genome integration and unc-54 3’UTR activity may account for the distinct differences in the required longevity dependencies. In aggregate, these results support the notion that ether lipids are an important requirement in multiple, diverse longevity paradigms, and further that fard- 1(oe) promotes mechanistically distinct lifespan extension in C. elegans. To characterize shifts in ether lipids related to pro-longevity effects, we performed comparative GC-MS-based fatty acid profiling of our integrated fard-1(oe) animals. Levels of 16:0 and 18:1 alkenyl ether lipids (indicated by DMAs on GC-MS analysis), are significantly increased in fard- 1(oe3) transgenic animals versus wild type worms (Figure 5I). By comparison, 18:0 DMA ether lipids were not increased, indicating that the ether lipid pool has both similarities and differences between fard-1 overexpression and phenformin treatment. Echoing the analysis seen with phenformin treatment, few differences were found in a comparison of the relative abundance of fatty acids within the total lipid pool for fard-1(oe3) and wild type worms (Figure 5J). Those exhibiting increases in fard-1(oe) include the polyunsaturated fatty acids (PUFA) 20:4 arachidonate and 20:5 eicosapentaenoate (EPA). This suggests either that PUFAs play a mechanistic role in lifespan extension in fard-1(oe) or that they are increased because of longevity-promoting activity of ether-lipids. Ether lipids do not promote lifespan extension by modulating ferroptosis Ether lipids have been reported to be protective against ferroptosis, an iron-dependent form of programmed cell death characterized by the accumulation of lipid peroxides 206,207 . In order to determine whether ether lipids promote longevity downstream of biguanide action by modulating ferroptosis, we knocked down members of the glutathione peroxidase (GPX) family in animals overexpressing integrated fard-1 (fard-1 oe3 and fard-1 oe4), as has been previously reported to genetically facilitate lipid peroxidation and ferroptosis 207,208 (Figure 5—figure supplement 1A—C). 210 This analysis indicates that gpx-1 (ortholog of human GPX4) RNAi leads to variable lifespan extension relative to wild type controls and exhibits non-additive lifespan extension with fard-1(oe) (Figure 5—figure supplement 1A). Neither gpx-6 nor gpx-7 knockdown impacts lifespan extension in fard-1(oe) animals (Figure 5—figure supplement 1B—C). Further, GPX family RNAi do not negatively impact lifespan extension reproducibly downstream of phenformin (Figure 5—figure supplement 1D—F). We conclude that genetic triggers that induce ferroptosis do not impact phenformin-prompted or fard-1(oe) lifespan extension, and thus it is unlikely that either extend lifespan by suppressing ferroptosis. The ether lipid biosynthetic machinery operates upstream of the stress responsive factor, skn-1/Nrf, to enable lifespan extension in response to biguanides We noted when analyzing FARD-1 protein localization that somatic lipid droplets are generally less numerous in BODIPY-stained phenformin-treated animals vs. vehicle. Indeed, quantitative analysis indicates that intestinal lipid droplets are significantly less numerous following phenformin treatment (in glo-4 RNAi-treated FARD-1::RFP transgenics (fard-1 oe3) fed C1-BODIPY-C12 to label lipid droplets, Figure 6A). We previously reported that gain-of-function mutations in the nutrient- and stress-responsive transcription factor skn-1/Nrf prompt age-dependent, somatic depletion of fat (Asdf) 13,15 . This, together with early adult decreases in lipid droplet numbers, suggested to us that phenformin may prompt longevity by activating metabolic stress defenses in a skn-1-dependent manner. Strikingly, we found that phenformin treatment produces Asdf at day three of adulthood, a phenotype that is quantitatively abrogated in skn-1 loss-of-function mutant animals (Figure 6B-C). The phenformin mediated Asdf-phenotype is quantitatively analogous to and non-additive with skn-1 gain-of-function mutants (Figure 6D—E). Interestingly, FARD-1 overexpressing animals additionally display an intermediate Asdf phenotype, with moderate enhancement by phenformin treatment (Figure 6D-E). Compellingly, loss-of-function mutations in any of the three ether lipid biosynthetic genes completely prevent the phenformin-mediated Asdf 211 phenotype (Figure 6D—E). Our previous work determined that SKN-1 activates a metabolic stress defense response to drive somatic lipid depletion through enhancing lipid utilization and innate immunity gene expression that opposes canonical oxidative stress responses 13 . Indeed, we show that phenformin treatment reduces expression of the canonical oxidative stress response gene gst-4 irrespective of bacterial diet source (Figure 6 – figure supplement 1A-B), while reciprocally inducing expression of the innate immune response gene dod-24 in a SKN-1, ether lipid machinery dependent manner (Figure 6F and Figure 6 – figure supplement 1C). Together with the observation that promotion of lifespan extension by both phenformin and fard-1(oe) require skn-1, these data suggest that biguanides activate an ether lipid-skn-1 signaling relay to drive longevity-associated metabolic shifts. 212 DISCUSSION In an unbiased RNAi screen of ~1000 metabolic genes, we identified ether lipid biosynthesis as critical to the growth-inhibitory and longevity-promoting effects of metformin in C. elegans. Our results show that the biguanide phenformin promotes lifespan extension by stimulating biogenesis of ether lipids through direct action in the nematode, prompting longevity-promoting metabolic stress defenses mediated by skn-1. The broad importance of ether lipids is demonstrated by their requirement in multiple diverse paradigms of lifespan extension. Our findings also indicate that ether lipid modulation through overexpression of fard-1 is also sufficient to promote longevity. Thus, ether lipids form a heretofore unappreciated lynchpin of lifespan modulation and are sufficient to support healthy aging through multiple central longevity effectors, including skn-1. Differences in ether lipid abundance and composition are correlated with diseases of aging. The uniform lethality associated with human genetic ether lipid deficiency, as in the case of patients diagnosed with RCDP and Zellweger syndrome, has made it difficult to study the role of ether lipids in aging and aging-associated diseases 209-212 . Nonetheless, observational studies demonstrate decreases in certain plasmalogen species in Alzheimer’s Disease, suggesting a probable link between ether lipids and aging-related pathologies 213-215 . Ether lipids have conflicting roles in cancer; while loss of the ether lipid biosynthetic machinery profits cancer cell survival by enhancing resistance to ferroptosis 206 , in other contexts, ether lipid deficiency results in impaired pathogenicity in various human cancer cells 186,207 . Cancer cells generally have higher levels of ether lipids compared to normal cells, leading others to suggest that ether lipids confer pro-survival benefit 186,216,217 . However, certain ether lipid species have also been reported to have anti-tumor properties 218,219 . Thus, in line with the results we present here, it is critical to understand ether lipids in context. Future work will need to focus on the impact of specific ether lipid species rather than the whole class en masse to understand which may play a beneficial versus detrimental role in health. 213 Studies in long-lived animal models suggest there is an association between ether lipid content and animal longevity, such as in the naked mole-rat (Heterocephalus glaber) 220 and the mud clam Arctica islandica 221 . Higher plasmalogen levels in naked mole-rat tissues versus mice are speculated to contribute to protection of cellular membranes via a reduction of oxidative stress 220 . Similarly, exceptionally long-lived humans harbor higher levels of phosphatidylcholine- derived, short chained alkyl ether lipids and a lower levels of phosphatidylethanolamine-derived longer chained plasmalogens 179 , but these associations are of unclear functional significance. Although it is clear that ether lipid deficiency in C. elegans prevents longevity downstream of mitochondrial electron transport chain dysfunction, mTOR deficiency, caloric restriction, and biguanides alike, the precise lipid(s) conferring this activity remains unknown. Each of these longevity paradigms have features of nutrient deficiency, energy stress, or nutrient sensing, so it is possible that ether lipids are at least part of the common effector arm conferring benefit in aging to various forms of metabolic stress. Our results suggest that unsaturated fatty acids and phosphatidylethanolamine ether lipids are essential to the health promoting effects of biguanides. Although we see major shifts in abundance of alkenyl ether lipids, genetic evidence of the necessity of ether lipids, and requirement for the synthesis of mono- and poly-unsaturated fatty acids in biguanide-induced longevity, determination of the specific lipids necessary for promoting healthy aging awaits the ability to modulate the level of specific ether lipids. Additionally, disruption of ether lipid biosynthesis has been shown to increase the proportion of stearate (18:0) and other saturated fatty acids 185 . Thus, at this time, we cannot rule out the possibility that biguanide-stimulated alterations in ether lipid biosynthesis serves to divert accumulation of lipid species that are detrimental to lifespan, for instance, saturated fatty acids. Nonetheless, considering our finding that ether lipids prompt metabolic stress defenses, this alternative mechanism is less likely. Definitive proof will require a deeper 214 understanding of the regulation of specific steps dictating the synthesis and modification of ether lipids of different fatty alcohol and fatty acid composition. Based upon our findings, ether lipid synthesis is likely to be regulated post-translationally by biguanide treatment. The demonstrated increase in plasmalogens and specific ether lipids are both consistent with increases in activity of the ether lipid biosynthetic machinery. While we do not understand the mechanism for the increased activity of ether lipid synthesizing enzymes, the decreases in mRNAs for acl-7, ads-1, and fard-1 and protein for FARD-1 treated with phenformin invokes negative feedback consistent with previous work showing that higher levels of ether lipids promotes proteasomal degradation of peroxisomal Far1 protein 193 . Co-localization of the fatty alcohol reductase, FARD-1, with both peroxisomes and lipid droplets is similarly not impacted by biguanides. We cannot rule out the possibility, however, that the exogenous, overexpressed nature of FARD-1::RFP in these experiments may result in a hyperactivated ether lipid biosynthesis state, thereby locking FARD-1::RFP localization in an activated configuration that cannot be further induced with biguanide treatment. Future studies leveraging endogenously tagged FARD-1 animals will be required to resolve this caveat. Finally, further investigation into the precise molecular interactions between FARD-1 protein and other organelles will be required to further understand how FARD-1 and the other ether lipid biosynthetic enzymes are regulated by biguanides and in aging. Strikingly, our data demonstrate for the first time that ether lipids are required for phenformin to activate metabolic defenses downstream of the stress- and metabolism-responsive transcription factor skn-1/NRF. Phenformin drives age-dependent somatic depletion of fat (Asdf), a phenotype we previously reported upon genetic activation of skn-1 13,15 . Based upon our own work, biguanides do not stimulate canonical skn-1 antioxidant defenses such as gst-4 expression, in contrast to the subtle effects seen in the existing literature 158,160 . Indeed, we observe a significant 215 decrease in gst-4 expression with phenformin treatment, reciprocally balanced by increased innate immune dod-24 expression in a skn-1, ether lipid machinery dependent manner. We suggest that skn-1 is uniquely required for metabolic stress defenses downstream of metformin such as Asdf, rather than canonical oxidative or proteostatic defenses. The requirement for ether lipids in Asdf activation by phenformin confirm that this class of lipids plays a heretofore unappreciated role in a distinct form of skn-1 activation mimicked by genetic forms of skn-1 activation that we have previously reported 13,15 . In aggregate, data presented here indicate that ether lipid biosynthesis plays a broader role in aging than previously described. The necessity of the ether lipid machinery in metformin- and phenformin-stimulated lifespan extension and in multiple longevity paradigms indicates that ether lipids serve as a lynchpin through which lifespan is modulated (Figure 7A—B). Our demonstration that overexpression of FARD-1 alone results in lifespan extension provides an exciting opportunity to identify ether lipids that promote health and the effector mechanisms through which they act. Finally, these results support the exciting possibility that modulation of ether lipids pharmacologically or even dietarily may provide a new potential therapeutic target in aging and aging-related diseases. 216 MATERIALS & METHODS C. elegans genetics. Strains were maintained at 20°C grown on E. coli OP50-1 (RRID: WB- STRAIN:WBStrain00041971) for all experiments unless otherwise indicated. The following strains were used in this study: N2 (wild type strain, RRID: WB-STRAIN:WBStrain00000001), BX275 fard-1(wa28) [G261D] (RRID: WB-STRAIN:WBStrain00004025), BX259 acl-7(wa20) [R234C] (RRID: WS-STRAIN: WBStrain00004024), BX10 ads-1(wa3) [G454D] (RRID: WB- STRAIN:WBStrain00004007), CB1370 daf-2(e1370) (RRID: WB-STRAIN: WBStrain00004309), MQ989 isp-1(qm150) (RRID: WB-STRAIN: WBStrain00026672), VC533 raga-1(ok701) (RRID: WB-STRAIN: WBStrain00035849), DA465 eat-2(da465) (RRID:WB-STRAIN: WBStrain00005463), MGH48 mgIs43[ges-1p::GFP::PTS1], SPC168 skn-1(lax188) (skn-1gf, RRID: WB-STRAIN: WBStrain00034420), CF3556 agIs6[dod-24p::GFP] (RRID:WB- STRAIN:WBStrain00004921), CL2166 dvIs19 [(pAF15)gst-4p::GFP::NLS] (RRID:WB- STRAIN:WBStrain00005102), and EU31 skn-1(zu135) (skn-1lf, RRID: WB-STRAIN: WB- STRAIN:WBStrain00007251). BX275, BX259, and BX10 strains contain missense mutations that result in loss-of-function of the ether lipid biosynthesis, as previously described 222 . For fard- 1 overexpression, the following strains were generated: MGH471 alxEx122[fard-1p::FARD- 1::mRFP::HA unc-54 3'UTR myo-2p::GFP] (fard-1 oe1), MGH472 alxEx135[fard-1p::FARD- 1::mRFP::HA unc-54 3'UTR myo-2p::GFP] (fard-1 oe2), MGH605 alxIs45[fard-1p::FARD- 1::mRFP::HA::unc-54 3'UTR myo-2p::GFP] (fard-1 oe3), and MGH606 alxIs46[fard- 1p::FARD-1::mRFP::HA::unc-54 3’UTR myo-2p::GFP] (fard-1 oe4). All strains for fard-1 overexpression were backcrossed 8x to N2 Bristol. For colocalization analysis with peroxisomally targeted GFP, we crossed MGH48 and MGH471 to generate the strain: MGH607 mgIs43[ges-1p::GFP::PTS1]; alxEx122[fard-1p::FARD-1::mRFP::HA::unc-54 3'UTR myo- 2p::GFP] (noted in text as GFP::PTS1; FARD-1::RFP). 217 Generation of fard-1 C. elegans transgenic lines. For FARD-1 expression, the entire genomic sequence of the fard-1 locus (3659 bp), including introns and exons, plus 4910 bp of promoter were amplified and cloned into a modified Fire vector driving fard-1 fused to mRFP and a HA epitope tag at the C-terminus. The following cloning primers were used: F: 5’-TGCATGCCTGCAGGTCGACTTTGACAAAAGTTCTGTTGCCG-3’ and R: 5’-TTTGGGTCCTTTGGCCAATCGCTTTTTTGAAGATACCGAGAATAATCC-3’. The FARD-1 overexpression construct was injected at 10 ng/μL (alxEx122) and 18 ng/μL (alxEx135) into the gonad of wild type adult animals with salmon sperm DNA as a carrier and 1.5 ng/μL myo-2p::GFP as a co-injection marker. alxEx122 was subsequently integrated by UV irradiation and 8x backcrossed to N2 Bristol to obtain MGH605 and MGH606. RNA interference (RNAi) assays. RNAi clones were isolated from a genome-wide E. coli RNAi library (generated in strain HT115(DE3), RRID: WB-STRAIN:WBStrain00041079), sequence verified, and fed to animals as described 223 . RNAi feeding plates (6 cm) were prepared using a standard NGM recipe with 5 mM isopropyl-B–D-thiogalactopyranoside and 200 μg/ml carbenicillin. RNAi clones were grown for 15 hours in Luria Broth (LB) containing 100 μg/ml carbenicillin with shaking at 37°C. The stationary phase culture was then collected, concentrated through centrifugation, the supernatant was discarded, and the pellet was resuspended in LB to 20% of the original culture volume; 250 μl of each RNAi clone concentrate was added to RNAi plates and allowed to dry at least 24 hours prior to adding biguanide. Drug treatment was added to seeded RNAi plates and allowed to dry at least 3 hours before adding worms. Longevity assays. Lifespan analysis was conducted at 20°C, as previously described 224 . Briefly, 218 synchronized L1 animals were seeded onto NGM (for mutant treatment) or RNAi plates (for RNAi) and allowed to grow until the L4 to YA transition stage. On day 0 of adulthood as indexed in the figure legend, ~50-60 L4/YA worms per plate (unless otherwise noted) were transferred onto fresh NGM or RNAi plates. These NGM and RNAi plates were supplemented with 30 μM and 100 μM 5-fluorodeoxyuridine (FUdR) to suppress progeny production, respectively. For biguanide treatment, about ~55-60 synchronized L1 animals (unless otherwise noted) were seeded onto plates containing 50 mM metformin or 4.5 mM phenformin. Based upon power calculations for log-rank analysis, minimum N of 50 (per group) was chosen to satisfy a = 0.05, b = 0.2, and effect size = 20% difference in lifespan 225 . At the L4/YA stage, these worms were transferred to plates containing biguanide treatment and FUdR for the remainder of their life. For experiments performed without the use of FUdR, animals were transferred to freshly seeded RNAi and drug supplemented plates every two days between day 0 and day 10 of adulthood, ensuring no crossover contamination of progeny or laid eggs on the lifespan plates until the animals cease the reproductive stage. Dead worms were counted every other day, and scoring investigators were blinded as to the experimental group/treatment until the conclusion of each experiment. All lifespans performed include same-day wildtype (N2) controls examined simultaneously with experimental test animals in each study. Statistical analysis was performed with online OASIS2 resources 226 . Body Size Determination of C. elegans. We measured worm body size in response to biguanide treatment by imaging as previously described 162 . Egg prep synchronized wild type worms were treated with empty vector (L4440) or ether lipid biosynthesis machinery RNAi and treated with vehicle (ddH2O) or 160mM metformin. After ~65-70 hours, worms were transferred into a 96 well plate, washed 3x with M9, and paralyzed in M9 buffer with 1 mg/ml levamisole (L9756-10G, Sigma-Aldrich). Once immobilized, brightfield imaging was performed at 5X magnification on a Leica DM6000 microscope within 5 minutes of transferring to a 96 well Teflon imaging slide. We 219 determined the maximal, longitudinal cross-sectional area of the imaged worms by using MetaMorph software for a minimum of ~80 animals per condition in each experiment. Results of a single experiment is shown. Each experiment was performed at least twice, and results were consistent between experiments. GC/MS lipidomics. Lipid extraction and GC/MS of extracted, acid-methanol-derivatized lipids was performed as described previously 227,228 . Briefly, 5000 synchronous mid-L4 animals were sonicated with a probe sonicator on high intensity in a microfuge tube in 100-250 microliters total volume. Following sonication, lipids were extracted in 3:1 methanol: methylene chloride following the addition of acetyl chloride in sealed borosilicate glass tubes, which were then incubated in a 75°C water bath for 1 hour. Derivatized fatty acids and fatty alcohols were neutralized with 7% potassium carbonate, extracted with hexane, and washed with acetonitrile prior to evaporation under nitrogen. Lipids were resuspended in 200 microliters of hexane and analyzed on an Agilent GC/MS equipped with a Supelcowax-10 column as previously described 227 . Fatty acids and alcohols are indicated as the normalized peak area of the total of derivatized fatty acids and alcohols detected in the sample. Based upon power calculation for pairwise comparison, a minimum N of 3 biological replicates (per group) was chosen to satisfy a = 0.05, b = 0.2, and effect size = 50% with s = 20%. Analyses were blinded to the investigator conducting the experiment and mass spectrometry calculations until the conclusion of each experiment when aggregate statistics were computed. LC/MS-MS lipidomics. Wild type, fard-1, acl-7, and ads-1 worm mutants were collected using conditions that enabled our reported longevity phenotypes. Briefly, collection for LC/MS-MS processing comprised of 3 replicates of these 4 strains that were independently treated with vehicle (ddH2O) and 4.5mM phenformin on 10cm NGM plates. Based upon power calculations, 220 as for GC/MS, a minimum N of 3 biological replicates (per group) was chosen to satisfy a = 0.05, b = 0.2, and effect size = 50% with s = 20%, though the power is only expected to hold for the first significant difference detected. Analyses were blinded to the investigator conducting the experiment and mass spectrometry calculations until the conclusion of each experiment when aggregate statistics were computed. A total of ~6,000 animals (2 x 10cmM plates, 3,000 worms per plate) were utilized per sample. These worms were washed with M9 (4x), concentrated into 200 μL of M9, and then flash frozen with liquid nitrogen in 1.5mL Eppendorf microcentrifuge tubes. Worm pellets were transferred to 2 mL impact resistant homogenization tubes containing 300 mg of 1 mm zirconium beads and 1 mL of 90:10 ethanol:water. Using a Precellys 24 tissue homogenizer, samples were homogenized in three 10 second cycles at 6400 Hz followed by 2 minutes of sonication. Samples were then placed at -20 °C for one hour to facilitate protein precipitation. Samples were transferred to 1.5 mL microfuge tubes and centrifuged at 14,000 g for 10 minutes at 4 °C. After centrifugation, 120 µL of supernatant was dried in vacuo and resuspended in 120 µL of 80:20 methanol:water containing internal standards 1 ng/µL CUDA and 1 ng/µL MAPCHO-12-d38. Lipidomic data was acquired by injecting 20 µL of sample onto a Phenomenex Kinetex F5 2.6 µm (2.1 x 100 mm) column at 40 °C and flowing at 0.35 mL/min. Metabolites were eluted using (A) water containing 0.1% formic acid and (B) acetonitrile:isopropanol (50:50) containing 0.1% formic acid using the following gradient: 0% B from 0-1 min, 0-50% B from 1-6 mins, 50-100% B from 6 to 17 minutes and 100% B hold from 17- 20 mins. Compounds were detected using a Thermo Scientific QExactive Orbitrap mass spectrometer equipped with a heated electrospray ionization (HESI) source operating in positive and negative ion mode with the following source parameters: sheath gas flow of 40 units, aux gas flow of 15 units, sweep gas flow of 2 units, spray voltage of +/-3.5 kV, capillary temperature of 265°C, aux gas temp of 350°C, S-lens RF at 45. Data was collected using an MS1 scan event followed by 4 DDA scan events using an isolation window of 1.0 m/z and a normalized collision 221 energy of 30 arbitrary units. For MS1 scan events, scan range of m/z 100-1500, mass resolution of 17.5k, AGC of 1e 6 and inject time of 50 ms was used. For tandem MS acquisition, mass resolution of 17.5 k, AGC 5e 5 and inject time of 80 ms was used. Data was collected using Thermo Xcalibur software (version 4.1.31.9) and analyzed using Thermo QualBrowser (version 4.1.31.9) as well as MZmine 2.36. Statistical analysis of metabolomics data. All visualization and significance testing of metabolomics was conducted using the MetaboAnalyst 5.0 package 229 . Mass integration values for 9,192 compounds were extracted from full-scan LC-MS/MS measurements of L4 to young adult (YA) transition wild type (N2 Bristol), ads-1(wa3), acl-7(wa20), and fard-1(wa28) animals treated from L1 hatch with vehicle, 4.5 mM phenformin, or 50mM metformin. Missing and zero values in the data matrix were imputed via replacement with 1/5th of the minimum positive value for each variable. Abundance values were subsequently filtered based on interquartile range (reducing the compound list to the 2500 most variable compounds), and log10 transformed. Quantile normalization was then performed, followed with division by the standard deviation of each variable (auto-scaling). Normalized abundance values for were then extracted based upon MS/MS signatures for phosphatidylethanolamine ether lipids and assessed for statistical significance via one-way ANOVA followed by false discovery rate (FDR) control using the Benjamini-Hochberg (BH) method 230 . Post-hoc testing was then performed using Fisher’s LSD to evaluate pairwise comparison significance. Metabolites were considered differentially abundant in any one condition with an FDR controlled P value < 0.05. The top 25 metabolites across treatment (ranked by ANOVA f statistic and FDR value) were visualized using a heatmap of Euclidean distance measurements, with Ward clustering of samples and normalized compound abundances included. All mass integration values for identified phosphatidylethanolamine containing ether lipids, normalized abundance values, and log- transformed, normalized abundance values, are included in this manuscript as Figure 2-source 222 data 1. These same data have been made publicly available and can be found at Dryad at: https://datadryad.org/stash/share/tZw0MURwnUaWP6Y6maavIpvz0tQIvJhRSjhapMSmcmY Quantitative RT-PCR. To assess changes in mRNA levels of fard-1 (both native 3' UTR and Exon 5-6 spanning junctions), acl-7, and ads-1 in response to biguanide treatment, we used quantitative RT-PCR as previously described 162 . Briefly, synchronized wild type (N2) or fard-1(oe3) L1 animals were seeded onto OP50-1 NGM plates containing vehicle (ddH2O), 50 mM metformin, or 4.5 mM phenformin. ~1600 worms were collected from 4 6cm plates per replicate, per condition (with no more than 400 worms seeded per plate to prevent overcrowding). n = 3 biological replicates. Worms were collected at the L4 to YA transition (for wildtype analysis) or at Adult Day 1 (for fard- 1(oe3) analysis) using M9 buffer and washed an additional 3X, allowing worms to settle by gravity between washes. Total RNA was extracted using TRIzol and phenol-chloroform extraction. Reverse transcription was performed with the Quantitect Reverse Transcription kit (Qiagen). qRT- PCR was conducted in triplicate using Quantitect SYBR Green PCR reagent (Qiagen) following manufacturer instructions on a Bio-Rad CFX96 Real-Time PCR system (Bio-Rad). If not processed immediately, worms were flash frozen in liquid nitrogen and kept in −80 °C until RNA preparation. The sequences for primer sets used in C. elegans are: act-1: F: 5’-TGCTGATCGTATGCAGAAGG-3’ and R: 5’-TAGATCCTCCGATCCAGACG-3’ pmp-3: F: 5’-GTTCCCGTGTTCATCACTCAT-3’ and R: 5’-ACACCGTCGAGAAGCTGTAGA-3’ 223 fard-1 (spanning Exons 5-6) : F: 5’-ACAAGTCACCAATGGCTCCAC-3’ and R: 5’-GCTTTGGTCAGAGTGTAGGTG-3’ fard-1 (native 3’ UTR): F: 5’-cgatagtgtgtctgttgattgtga-3’ and R: 5’-agttattgttgatgagagagtgcg-3’ acl-7: F: 5’-GTTTATGGCTGGCGTGTTG-3’ and R: 5’-CGGAGAAGACAGCCCAGTAG-3’ ads-1: F: 5’-GCGATTAACAAGGACGGACA-3’ and R: 5’-CGATGCCCAAGTAGTTCTCG-3’. Expression levels of tested genes were presented as normalized fold changes to the mRNA abundance of act-1 or pmp-3 for C. elegans by the ΔΔCt method. FARD-1 overexpression reporter fluorescence intensity analysis. To assess changes in levels of fluorescent FARD-1 protein in response to biguanide treatment, we used the strain MGH471 alxEx122[fard-1p::FARD-1::mRFP::HA unc-54 3'UTR myo-2p::GFP] (fard-1 oe1). In brief, egg prep synchronized L1 FARD-1::RFP transgenic worms were treated with vehicle (ddH2O) or 4.5 mM phenformin, paralyzed with 1 mg/ml of levamisole, and then imaged in 96-well format with a Leica DM6000 microscope outfitted with a mCherry filter set and MMAF software. These imaging experiments were carried out in biological triplicate with ~10 animals imaged per replicate. Images 224 were qualitatively assessed to obtain conclusions and results were consistent between independent replicates. Colocalization analysis of FARD-1::RFP and peroxisomally targeted GFP Colocalization of GFP and RFP expression in vehicle or phenformin treated MGH607 was performed by Coloc2 (Fiji) on images taken on a Leica Thunder microscopy system. Since FARD- 1::RFP in MGH607 is exogenously expressed, we performed 3 hour egg lays with ~30 gravid hermaphrodites expressing both GFP::PTS1 and FARD-1::RFP to synchronize L1s. The eggs were treated with vehicle (ddH2O) or 4.5mM phenformin immediately after gravid hermaphrodites were removed, dried in a laminar flow hood, and allowed to incubate at 20°C until the worms were young adult/early day 1 adults. To prepare for imaging, only worms expressing both GFP::PTS1 and FARD-1::RFP were picked onto slides containing dried 2% agar pads, immobilized in ~5 μL of 2.5mM levamisole solution and covered with a cover slip. Images of the upper, mid, and lower intestine were taken for 30 individual worms per condition (15 worms per replicate for 2 biological replicates). We generated Pearson’s r values to assess the extent to which intestinal RFP and GFP overlap in each region of all samples. All Pearson’s r values were combined to generate 4 individual averages (1 per condition) to perform an unpaired t-test. Lipid droplet analysis. The strain MGH605 alxIs45[fard-1p::FARD-1::mRFP ::HA::unc-54 3'UTR myo-2p::GFP] (fard-1 oe3) was used for this analysis. Preparation of worms for imaging was similar to our longevity assays but modified to incorporate staining of lipid droplets. Briefly, 6cm RNAi plates were seeded with 250μL bacteria expressing glo-4 RNAi [5X] and allowed to incubate for 24 hours at 20°C. 1μM of green C1-BODIPY-C12 (D-3823, Invitrogen) diluted in 100μL 1X phosphate buffer saline (PBS, pH 7.2) was then added to the RNAi bacteria lawn as in 224 .The plates were immediately dried in a dark laminar flow hood, wrapped in aluminum foil to prevent photobleaching, and allowed to incubate at 20°C for 24-48 hours. These plates were 225 treated with vehicle (ddH2O) or 4.5mM phenformin as mentioned previously (while kept away from light.) Egg prep synchronized worms were dropped onto plates and grown to day 1 adult stage. To prepare for confocal imaging, animals were rapidly picked onto slides containing dried 2% agar pads, immobilized in ~5 μL of 2.5mM levamisole solution and covered with a cover slip. Lipid droplets were imaged by Zeiss LSM 800 Airyscan within 5 minutes of slide placement. Z-stacked images were obtained for the intestine near the tail end of 14 glo-4, vehicle treated and 19 glo-4, phenformin treated worms (2 biological replicates per condition). 5 planes were extracted (planes 1,2,4,5, and 9) using ImageJ for all samples. For lipid droplet counting, quantification was performed using CellProfiler 4.2.1 231 where lipid droplets were identified as primary objects. The min/max range for typical object diameters was 3-67 pixels, and those objects outside of the diameter range were discarded. Planes were excluded entirely if the pipeline did not accurately capture individual lipid droplets for the vast majority of objects. Oil-Red-O Staining. Oil-red-O (ORO) fat staining was conducted as outlined in 154 , In brief, worms were synchronized by bleach prep and allowed to hatch overnight for a synchronous L1 population. The next day, worms were dropped onto plates seeded with bacteria with or without phenformin and raised to 120 hours (day 3 adult stage). Worms were washed off plates with PBST, then rocked for 3 min in 40% isopropyl alcohol before being pelleted and treated with ORO in diH2O for 2 hours. Worms were pelleted after 2 hours and washed in PBST for 30 min before being imaged at 5x magnification with the DIC filter on the Zeiss Axio Imager Erc color camera. A minimum of 200 worms in total (across 3 independent biological replicates) were assessed per condition for final quantification and evaluation. Generation of metabolically inactive E. coli for lipidomic and lifespan studies. PFA killing of OP50-1 E. coli was performed as previously described with slight modifications 152 . 50 mL aliquots of OP50-1 liquid cultures grown overnight in LB media supplemented with 25 μg/mL streptomycin 226 were dispensed into 250 mL Erlenmeyer flasks. Either 1x PBS (Life Technologies) for mock treatment or 4% paraformaldehyde (Sigma Aldrich) diluted in 1x PBS was added to each flask for a final concentration of 1% (v/v). Bacteria were then shaken in 37°C at 210 rpm for 2 hours to enable PFA inactivation. Cultures were then aseptically transferred into 50 mL conical centrifuge tubes, and then washed 6x with sterile PBS to remove residual PBS or PFA solution. After the final wash, bacterial pellets were then 10x concentrated in LB media supplemented with 25 μg/mL streptomycin, and 300 μL seeded onto freshly prepared NGM plates. Plates were allowed to dry for 2 days prior to use for GC/MS or lifespan analyses. A standard culture of OP50-1 grown overnight was similarly 10x concentrated and seeded as a ‘Live OP50-1’ control to compare to mock-treated and PFA-treated bacterial conditions. Bacterial titer calculations were performed as previously described 152 , removing an 10 μL aliquot of culture prior to plate seeding, diluting ten times in ten-fold serial dilutions, and subsequently dispensing the 100 μL dilutions onto LB agar plates and aseptically spread across the surface evenly. Plates were incubated at 37°C overnight before counting colonies for colony forming units (CFU) and titer calculations. Asdf Quantification. ORO-stained worms were placed on glass slides and a coverslip was placed over the sample. Worms were scored, as previously described 154 . Worms were scored and images were taken with the Zeiss Axio Imager Erc color camera at 5x magnification. Fat levels of worms were placed into three categories: non-Asdf, intermediate, and Asdf. Non-Asdf worms display no loss of fat and are stained dark red throughout most of the body (somatic and germ cells). Intermediate worms display significant fat loss from the somatic tissues, with portions of the intestine being clear, but ORO-stained fat deposits are still visible (somatic < germ cells). Asdf worms had most, if not all, observable somatic fat deposits depleted (germ cells only). Fluorescence reporter imaging and quantification. GFP imaging of CF3556 agIs6[dod- 24p::GFP] and CL2166 dvIs19 [(pAF15)gst-4p::GFP::NLS] animals was performed using a fully 227 automated, high speed fluorescence Leica THUNDER 3D imaging station at 5x magnification. Egg prep synchronized dod-24p::GFP or gst-4p::NLS::GFP animals were dropped onto NGM plates seeded with OP50-1, or RNAi plates seeded with L4440 (EV), skn-1, fard-1, acl-7, or ads- 1 HT115 RNAi clones, treated with vehicle (water) or 4.5 mM phenformin, and grown to Adult Day 1 stage. Animals were rapidly picked onto slides containing dried 2% agar pads, immobilized in ~5 μL of 2.5mM levamisole solution and covered with a coverslip. Quantification was performed using ImageJ/FIJI 232 , in which at least 10 animals per condition per replicate were randomly polygon traced, collected into an ROI manager, and measured for mean fluorescence intensity (MFI). MFI values per condition per replicate (n = 3) were aggregated using Prism 9 (Graphpad) for visualization and subsequent statistical analysis. Quantification and statistical analysis. Unless otherwise indicated, the statistical differences between control and experimental groups were determined by two-tailed students t-test (two groups), one-way ANOVA (more than two groups), or two-way ANOVA (two independent experimental variables), with corrected P values < 0.05 considered significant. Analyses conducting more than two comparisons were always corrected for multiple hypothesis testing. The log rank test was used to determine significance in lifespan analyses using online OASIS2 (https://sbi.postech.ac.kr/oasis2/). 228 FIGURES 229 Figure 1. Genes responsible for ether lipid biosynthesis are necessary for biguanide- induced lifespan extension. (A) C. elegans ether lipid synthesis is catalyzed by three enzymes: fatty acyl reductase FARD-1, acyltransferase ACL-7 and alkylglycerone phosphate synthase ADS-1 (adapted from 185,196 ). The latter two are localized to the peroxisomal lumen. (B-D) Missense, loss of function mutations in fard-1 (B), acl-7 (C), and ads-1 (D) in C. elegans suppress phenformin-induced lifespan extension. (E-G) A deficiency of ether lipid synthesis in fard-1 (E), acl-7 (F), and ads-1 (G) worm mutants blunts metformin-induced lifespan extension. Results are representative of 3 biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. Note that B-D and E-G contain the same wildtype controls as they are visualized from the same replicate of the study. See also Figure 1-figure supplement 1 and refer to Supplementary file 1 for tabular survival data and biological replicates. (H-I) Normalized concentrations of phenformin (H) and metformin (I) in vehicle, 4.5mM phenformin, or 50 mM metformin treated wild type (wt) C. elegans versus fard-1, acl-7, and ads-1 mutants. n = 3 biological replicates; ***, P < 0.004 by two-tailed students t-test with Bonferroni correction for multiple hypothesis testing. Box represents 75 th /25 th percentiles, while whisker represents higher/lower hinge +/- [1.5 * interquartile range (IQR)]. 230 Figure 2. Phenformin treatment of C. elegans leads to increased abundance of multiple alkyl and alkenyl ether lipids. (A-B) Loss-of-function fard-1 mutants have significant reduction 231 in 18:0 fatty alcohols derivatized from 18-carbon containing alkenyl ether lipids (dimethylacetal, DMA) by GC/MS (A) and accumulation of the saturated fatty acid stearate (18:0, B). (C) Wild type worms treated with 4.5 mM phenformin display a significant increase in 18:0 DMA relative to vehicle control, indicative of higher levels of alkenyl ether lipids, with levels remaining essentially undetectable in fard-1 mutants on vehicle or drug. (D) 4.5mM phenformin treatment does not impact stearate levels in wild type worms, however it does result in a greater accumulation of stearate in fard-1 mutants. For A-D, **, P < 0.01; ****, P < 0.0001, by t-test (A-B) or two-way ANOVA (C-D), n = 3 biological replicates. (E) 4.5mM phenformin treatment results in a significant increase in 16:0 DMA and 18:1 DMA in wild type worms, relative to vehicle-treated controls (*, P < 0.05; **, P < 0.01, by multiple t-tests, with two-stage linear step-up procedure of Benjamini, Krieger and Yekutieli. n = 3 biological replicates. (F) Heat map of normalized ether lipid abundance following phenformin treatment in wild type C. elegans indicates an overall increase in ether lipids relative to vehicle treated controls, and this shift is absent in ether lipid deficient mutants. All metabolites shown have an FDR adjusted P < 0.05 by one-way ANOVA followed by Fisher’s LSD post-hoc testing for WT versus fard-1, ads-1, and acl-7 mutants. (G) LC-MS analysis shows that phosphatidylethanolamine-containing ether lipids detected exhibited a general trend towards increased abundance in wild type worms treated with 4.5mM phenformin. Four of these ether lipids reached statistical significance: PE(O-16:0/18:1), PE(O- 18:0/18:3), PE(O-18:0/20:2), and PE(P-18:1/18:1). Eleven of the ether lipids detected are of the alkyl-type (indicated by “O” in their name prior to fatty alcohol designation) whereas 9 are of the alkenyl-type (plasmalogen, indicated by “P” in their name prior to the fatty alcohol designation) ether lipids. For G, *, P < 0.05; **, P < 0.01; ****, P < 0.0001, by multiple t-tests, with multiple hypothesis testing correction by two-stage step-up method of Benjamini, Krieger, and Yekutieli, n = 3 biological replicates. See Figure 2-source data 1 for raw and normalized mass spectrometry data. 232 Figure 3. Peroxisomal protein import, fatty acid elongases, and fatty acid desaturases are required for the pro-longevity effects of biguanides. (A-B) Knockdown of prx-5 (A) and prx- 19 (B) by RNAi eliminates or significantly suppresses phenformin-mediated lifespan extension. (C) Schematic representation of the mono- (MUFA) and poly-unsaturated fatty acid (PUFA) synthesis pathway in C. elegans (adapted from 233 ). (D-H) RNAi of three fatty acid desaturases (D-F) and two fatty acid elongases (G and H) involved in the synthesis of 18- and 20-carbon polyunsaturated fatty acids blunt phenformin-mediated lifespan extension in wild type worms. Colored symbols for elo and fat genes (vs. those in black and white) in Figure 3C indicates those that inhibit phenformin lifespan extension when knocked down by RNAi. For A, B and D-H, results are representative of 2-3 biological replicates. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by log- rank analysis. Note that D-G contain the same wildtype controls as they are visualized from the C16:0 C18:0 C18:1n9 C18:2n6 C18:3n6 C20:3n6 C20:4n6 C16:1n7 C18:1n7 C18:3n3 C20:3n3 C18:4n3 C20:4n3 C20:5n3 fat-5 elo-2 elo-1 fat-6 fat-2 fat-7 fat-1 elo-1 fat-3 elo-1 elo-2 fat-3 elo-1 elo-2 elo-2 fat-1 fat-1 fat-4 fat-4 Figure 3 C B E F G n A H eh phen 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (fat-4 RNAi) +v wt (fat-4 RNAi) + vs. vs. **** **** ns vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (elo-2 RNAi) +veh wt (elo-2 RNAi) +phen vs. vs. **** ns **** vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +phen wt (prx-5 RNAi) +veh wt (prx-5 RNAi) +phe wt (EV RNAi) +veh vs. vs. *** ns **** vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (prx-19 RNAi) +phen wt (EV RNAi) +veh wt (EV RNAi) +phen wt (prx-19 RNAi) +veh vs. vs. **** ** **** vs. D 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (fat-1 RNAi) +veh wt (fat-1 RNAi) +phen vs. vs. **** ns ** vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (fat-3 RNAi) +veh wt (fat-3 RNAi) +phen vs. vs. **** *** ns vs. n eh phe 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (elo-1 RNAi) +v wt (elo-1 RNAi) + vs. vs. **** ns **** vs. 233 same replicate of the study. See also Supplementary file 1 for tabular survival data and biological replicates. 234 Figure 4. Genes involved in ether lipid biosynthesis are required for lifespan extension in multiple longevity paradigms. (A-C) isp-1, raga-1, and eat-2 mutants display extended lifespan relative to wild type animals that is suppressed by RNAi knockdown of any of the three members of the ether lipid biosynthetic pathway. Results are representative of 3 biological replicates. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. See also Figure 4—figure supplement 1 and Supplementary file 1 for tabular survival data and biological replicates. Figure 4 C B A 0 10 20 30 40 50 0 20 Days 40 60 80 100 Percent Survival wt (EV RNAi) isp-1 (EV RNAi) isp-1 (fard-1 RNAi) isp-1 (acl-7 RNAi) isp-1 (ads-1 RNAi) vs. vs. ** *** *** vs. **** vs. 0 10 20 30 40 50 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) raga-1 (EV RNAi) raga-1 (fard-1 RNAi) raga-1 (acl-7 RNAi) raga-1 (ads-1 RNAi) vs. vs. **** **** **** vs. **** vs. eat-2 (fard-1 RNAi) 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) eat-2 (EV RNAi) eat-2 (acl-7 RNAi) eat-2 (ads-1 RNAi) vs. vs. ** **** **** vs. **** vs. 235 Figure 5. FARD-1 overexpression is sufficient to extend lifespan by modulating ether lipid synthesis. (A-B) Two independently generated fard-1 overexpression (fard-1 oe1 and fard-1 oe2) transgenic strains exhibit lifespan extension that is not further extended by concomitant Figure 5 B E D F A C 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) wt (ads-1 RNAi) fard-1 (oe1) (EV RNAi) fard-1 (oe1) (ads-1 RNAi) vs. ns vs. **** ns vs. vs. **** 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt +veh fard-1 (oe1) +veh wt +phen fard-1 (oe1) +phen vs. *** *** vs. vs. ns ns vs. 0 10 20 30 Days 0 20 40 60 80 100 Percent Survival fard-1 (oe1) (skn-1 RNAi) wt (EV RNAi) fard-1 (oe1) (EV RNAi) wt (skn-1 RNAi) vs. ns vs. **** ns vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival fard-1 (oe1) (daf-16 RNAi) wt (EV RNAi) fard-1 (oe1) (EV RNAi) wt (daf-16 RNAi) vs. ns vs. **** ns vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt +veh fard-1 (oe2) +veh wt +phen fard-1 (oe2) +phen vs. **** *** vs. vs. ns * vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival fard-1 (oe1) (aak-2 RNAi) wt (EV RNAi) fard-1 (oe1) (EVRNAi) wt (aak -2 RNAi) vs. ns vs. **** ns vs. I 16:0 18:0 18:1 0.016 0.031 0.063 0.125 0.250 0.500 1.000 2.000 4.000 Relative Abundance (normalized) wt fard-1 (oe3) Alkenyl Fatty Alcohols (DMA) * * 14:0 15:iso 15:iso 16:0 17:cyclo 16:1 17:iso 17:iso 17:cyclo 17:cyclo 18:0 18:1 oleate 18:1 vaccenate 19:iso 18:2 18:2 18:2 19:iso 19:cyclo 19:cyclo 19:cyclo 19:cyclo 19:cyclo 20:0 19:cyclo 17:cyclo 20:1 20:2 20:3 DGLA 20:4 ARA 20:3 11,14,17 ETA 20:4 20:5 EPA 22:0 0 5 10 15 20 25 Percent of Total Fatty Acid Pool * *** ** **** **** *** ** J WT fard-1 (oe3) G H N2 fard-1 (oe3) 0.0 0.5 1.0 1.5 fard-1 Native 3' UTR Relative Fold Change (normalized to pmp-3) ! !! ns N2 fard-1 (oe3) 0 50 100 150 200 Relative Fold Change (normalized to pmp-3) fard-1 Exons 5-6 !!!! ns !!!! 236 phenformin treatment. (C) RNAi knockdown of ads-1 fully suppresses fard-1(oe1) lifespan extension, indicating that the fard-1(oe)-mediated lifespan extension is dependent upon ether lipid synthesis. (D-F) RNAi of skn-1 (D), aak-2 (E), and daf-16 (F) suppress fard-1(oe1)-mediated lifespan extension. For A-F, results are representative of 2-3 biological replicates. *, P < 0.05; ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. Note that E-F contain the same wildtype controls as they are visualized from the same replicate of the study. See also Figure 4—figure supplement 1 and Supplementary file 1 for tabular survival data and biological replicates. (G-H) qRT-PCR analysis of wildtype and fard-1(oe3) animals treated with vehicle or phenformin until Adult Day 1 reveals that both biguanide treatment and fard-1 exogenous overexpression results in an equivalent reduction of native fard-1 gene expression, as indicated by primers targeting the native 3’ UTR (H) and the junction spanning exons 5-6 (I) of fard-1. For G-H, n = 3 biological replicates. *, P < 0.05; **, P < 0.01; ****, P < 0.0001 by two-way ANOVA followed by Tukey’s multiple comparisons test. (I) Worms overexpressing a backcrossed, integrated FARD-1 (fard-1 oe3) display a significant increase in 16:0 and 18:1 but not 18:0 alkenyl ether lipids by GC/MS. (J) Comparison of the total fatty acid pool indicates that the polyunsaturated fatty acids 20:4 arachidonic acid (ARA) and 20:5 eicosapentaenoic acid (EPA) are significantly increased in fard- 1 overexpressing (fard-1 oe3) worms vs. wild type animal, while several isomethyl (iso) and cyclopropyl (cyclo) fatty acids change in opposing directions. For I-J, n = 3 biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by multiple t-tests (with multiple hypothesis correction by two-stage step-up method of Benjamini, Krieger, and Yekutieli). 237 Figure 6. Phenformin modulates systemic lipid metabolism through an ether lipid-skn-1 signaling relay. (A) The number of intestinal, C1-BODIPY-C12 labeled lipid droplets are significantly lower in phenformin treated animals versus vehicle (FARD-1::RFP reporter transgenic (fard-1 oe3) worms are also treated with glo-4 RNAi to remove BODIPY-positive lysosome-related organelles). n=2 biological replicates. *, P < 0.05 by unpaired t-test. (B-C) Oil- red-O staining of day 3 adult phenformin treated wildtype and skn-1lf(zu135) animals reveals that the total loss of function of SKN-1 completely abrogates the phenformin-induced age-dependent 238 somatic depletion of fat (Asdf) phenotype. Quantification (C) reveals that skn-1lf(zu135) decreases the proportion of Asdf animals relative to wildtype controls treated with phenformin. For B-C, data represented n=3 biological replicates. (D-E) Oil-red-O staining of day 3 adult phenformin treated wild type animals indicates that drug treatment leads to Asdf, as previously reported for skn-1 gain of function mutants (skn-1 gf), suggesting that phenformin activates Asdf downstream of skn-1. Quantification (E) indicates that the proportion of Asdf animals is non additively increased by phenformin treatment in a skn-1gf mutant, and that phenformin is no longer able to activate Asdf in 3 independent ether lipid deficient mutants (ads-1, acl-7, and fard- 1). Additionally, FARD-1 overexpression results in an Asdf phenotype, moderately strengthened by phenformin treatment. For D-E, n=3 biological replicates. (F) Phenformin treatment induces intestinal expression of dod-24, an established SKN-1 response target and innate immune effector, as indicated by increased dod-24p::GFP expression, in both OP50-1 and HT115 bacterial diets. RNAi knockdown of skn-1, fard-1, acl-7, and ads-1 all prevent phenformin-mediated induction of dod-24p::GFP. Quantification performed with at least 30 animals in each condition (10 animals assayed per replicate for 3 biological independent experiments). ns, P > 0.05; ****, P < 0.0001 by two-way ANOVA followed by Tukey’s multiple comparisons test. 239 Figure 7. Schematic representation for the role of the ether lipid biosynthetic machinery in multiple pro-longevity paradigms. (A) Model of ether lipid action in biguanide-prompted lifespan extension. Activation of ether lipid biosynthesis leads to longevity-promoting activity of metabolic stress defenses downstream of the transcription factor skn-1. (B) Model portraying a broader than previously appreciated role of ether lipids in longevity downstream of biguanides, mitochondrial electron transport inhibition, mTORC1 inhibition, and eat-2 mutation-mediated dietary restriction (EAT-2 DR). Dashed lines for DAF-16 indicate its requirement for FARD-1 overexpression-, but not biguanide-mediated lifespan extension, suggesting a context-dependent role for DAF- 16/FOXO in mediating pro-longevity outcomes through modulation of ether lipid levels. Figure 7 Biguanides Ether Lipid Biosythesis Longevity SKN-1 SKN-1 A B Mitochondria mTORC1 EAT-2 DR Biguanides Ether Lipid Biosythesis Longevity AAK-2 SKN-1 DAF-16 240 SUPPLEMENTAL FIGURES Figure 1—figure supplement 1. Reduced function of genes responsible for ether lipid biosynthesis partially suppresses biguanide effects of growth and lifespan without affecting biguanide levels. (A) RNAi to fard-1 and acl-7 induce C. elegans resistance to growth inhibition by 160 mM metformin treatment. *, P < 0.05, by two-way ANOVA, n = 2 biological replicates. (B-C) RNAi knockdown of fard-1 (B) and acl-7 (C) in C. elegans partially suppresses phenformin’s effect on lifespan extension. For B and C, results are representative of 3 biological replicates. Note that B-C contain the same wildtype controls as they are visualized from the same replicate of the study. ****, P < 0.0001 by log-rank analysis; for tabular survival data and biological replicates see also Supplementary file 1. (D) Log fold change (LogFC) of phenformin abundance in samples treated with 4.5 mM phenformin versus vehicle reveals that the increase in phenformin Figure 1—figure supplement 1 B EV (RNAi) fard-1 (RNAi) acl-7 (RNAi) 0.0 0.5 1.0 1.5 Relative body area Vehicle Metformin * * A C 0.6 0.8 1.0 ads-1 phenformin - ads-1 vehicle contrast logFC genotype ads-1 acl-7 fard-1 wt Phenformin acl-7 phenformin - acl-7 vehicle fard-1 phenformin - fard-1 vehicle wt phenformin - wt vehicle D 0.2 0.4 0.6 0.8 contrast logFC genotype ads-1 acl-7 fard-1 wt Metformin ads-1 metformin - ads-1 vehicle acl-7 metformin - acl-7 vehicle fard-1 metformin - fard-1 vehicle wt metformin - wt vehicle E 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (fard-1 RNAi) +veh wt (fard-1 RNAi) +phen vs. vs. **** **** **** vs. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) +veh wt (EV RNAi) +phen wt (acl-7 RNAi) +veh wt (acl-7 RNAi) +phen vs. vs. **** **** **** vs. 241 levels in wild type and three ether lipid deficient mutants is similar. (E) LogFC of metformin abundance in samples treated with 50 mM metformin versus vehicle show that metformin increases are similar across all 4 strains. Bars represent mean and 95% confidence intervals. 242 Figure 1 – figure supplement 2. The use of FUdR in lifespan analyses does not confound the observed epistases between the ether lipid machinery and biguanide-mediated lifespan extension. Lifespans performed without the use of FUdR to inhibit progeny formation corroborate that a deficiency of ether lipid synthesis in fard-1 (A-B), acl-7 (C-D), and ads-1 (E-F) worm mutants negates both metformin (top row) and phenformin (bottom row)-induced lifespan extension. Results are representative of 3 biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. Note that A/D, B/E, and C/F contain the same vehicle controls as they are visualized from the same replicate of the study. Please refer to Supplementary file 1 for tabular survival data and biological replicate summary statistics. 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +phen ads-1 +veh ads-1 +phen Figure 1—figure supplement 2 B C D E A F 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +met fard-1 +veh fard-1 +met 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +phen fard-1 +veh fard-1 +phen 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +met acl-7 +veh acl-7 +met 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +phen acl-7 +veh acl-7 +phen 0 10 20 30 0 20 40 60 80 100 Days Percent Survival N2 +veh N2 +met ads-1 +veh ads-1 +met vs. vs. **** **** ns vs. vs. vs. **** **** ns vs. vs. vs. **** **** **** vs. vs. vs. **** ns **** vs. vs. vs. **** **** **** vs. vs. vs. **** * **** vs. No FUdR No FUdR No FUdR No FUdR No FUdR No FUdR 243 Figure 2—figure supplement 1. Biguanide treatment modulates abundance of fatty acids in C. elegans. A comparison of the percent of the total fatty acid pool for 33 fatty acids shows that 7 fatty acids are significantly altered in phenformin treated wild type worms. n = 3 biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by multiple t-tests (corrected for multiple hypothesis testing with two-stage step-up method of Benjamini, Krieger, and Yekutieli). Figure 2—figure supplement 1 A fard-1 reporter 10 (actually #5) +0mM phen (slide 1) fard-1 RFP #10 +4.5mM phen 14:0 15:iso 16:0 17:cyclo 16:1 17:iso 17:cyclo 17:cyclo 18:0 19:cyclo 18:1 oleate 18:1 vaccenate 19:iso 18:2 18:2 18:2 18:3 18:3 19:cyclo 19:cyclo 20:0 19:cyclo 17:cyclo 20:1 20:2 20:3 DGLA 18:3 20:3 19:cyclo 20:3 20:4 ARA 20:5 EPA 22:0 0 5 10 15 20 25 30 35 Percent of Total Fatty Acid Pool wt (+) veh wt (+) phen *** **** **** **** * **** ** 244 Figure 2—figure supplement 2. FARD-1::RFP localizes to intestinal lipid droplets and peroxisomes and is not positively regulated at the RNA or protein level by phenformin. (A) Diagram of the C. elegans FARD-1::RFP overexpression reporter. (B) FARD-1::RFP (fard-1 oe1) exhibits intestinal expression in C. elegans. FARD-1 displays a cytoplasmic distribution and an 245 association with structures resembling lipid droplets (B, arrows). (C) Co-expression of FARD- 1::RFP and peroxisomally targeted GFP::PTS1 in transgenic animals indicates partial colocalization of FARD-1 with peroxisomes in intestine. (D) Superplot displays colocalization of RFP and GFP in vehicle or phenformin treated GFP::PTS1; FARD-1::RFP transgenics (N= 20 total worms assessed; 5 worms per condition; 3 images per worm (upper/mid/lower intestine)) for a total of 15 images (dots) per replicate; blue = replicate 1, orange = replicate 2). Correlation coefficients were separately calculated for each biological replicate and the mean is represented for each pool (blue or orange triangle). These two means were then used to calculate the average (horizontal bar), standard error of the mean (error bars), and P value. Analysis of the average Pearson’s r values demonstrates no significant difference between colocalization of FARD-1::RFP and GFP::PTS1 in vehicle or phenformin-treated worms. n = 2 biological replicates. (E) Confocal imaging of an integrated FARD-1::RFP reporter (fard-1 oe3) in C. elegans stained with C1- BODIPY-C12 (treated with glo-4 RNAi to remove BODIPY positive lysosome related organelles) demonstrates localization of FARD-1 protein to the surface of lipid droplets in the worm intestine. (F) In fard-1(oe3) transgenics, confocal imaging indicates FARD-1::RFP organization into web- like structures and bright punctae that represent the intersection of these “webs”. These structures may represent smooth endoplasmic reticulum. Images were taken using a Zeiss Plan- Apochromat 63x/1.4 Oil DIC M27 objective with a 2.0 scan zoom for each field. (G-I) Levels of fard-1, acl-7, and ads-1 mRNA decrease in wild type C. elegans treated with 4.5 mM phenformin versus vehicle. n = 3 biological replicates; ns, not significant; *, P < 0.05 by unpaired t-test. (J-L) Levels of fard-1, acl-7, and ads-1 mRNA decrease in wild type C. elegans treated with 50 mM metformin versus vehicle. n = 3 biological replicates; *, P < 0.05; **, P < 0.01; ***, P < 0.001 by unpaired t-test. (M) Phenformin (4.5 mM) results in decreased expression of the FARD-1::RFP translational reporter (fard-1 oe1). n = 3 biological replicates; total assessed: N = 30 worms per condition (10 worms per replicate). 246 Figure 2 – figure supplement 3. Disruption of bacterial growth and metabolism does not prevent biguanide-mediated induction of ether lipid synthesis. (A) Bacterial titer assay measuring viability of OP50-1 treated with standard seeding conditions (Live OP50-1), treated with 1% PBS for 2 hours (Mock Treated OP50-1 [2 hr]), or treated with 1% PFA for 2 hours (1% PFA Treated OP50-1 [2 hr]). Data represent mean +/- SEM, n = 3 biological replicates. ns, P > -2 -1 0 1 2 3 4 0 1 2 3 log 2 (Phenformin/Vehicle) -log 10 (FDR) 1% PFA Treated OP50-1 (2 hr) FA 16:0 DMA FA 18:1 DMA FA 18:0 DMA B Figure 2 - figure supplement 3 A Live OP50-1 Mock Treated OP50-1 (2 hr) 1% PFA Treated OP50-1 (2 hr) 0 1 !10 10 2!10 10 3!10 10 4!10 10 5!10 10 Colony Forming Units (CFU) per milliliter (mL) ns ! -6 -4 -2 0 2 4 -4 -2 0 2 4 Component 1 (51.9%) Component 2 (10.9%) Sparse PLS-DA Classification C D E F G Live Mock PFA Vehicle Phenformin -4 -2 0 2 4 0 1 2 3 4 Mock Treated OP50-1 (2 hr) log 2 (Phenformin/Vehicle) -log 10 (FDR) FA 16:0 DMA FA 18:1 DMA FA 18:0 DMA -4 -2 0 2 4 0 1 2 3 4 5 Live OP50-1 log 2 (Phenformin/Vehicle) -log 10 (FDR) FA 18:0 DMA FA 16:0 DMA FA 18:1 DMA 247 0.05; *, P < 0.05 by one-way ANOVA followed by Dunnett’s multiple comparisons test. (B) Sparse partial-least squares linear discriminant analysis (PLS-DA) of total sum normalized AUC for lipids measured using FAME GC/MS in wildtype animals treated with vehicle/4.5 mM phenformin until Adult Day 1, and grown either on Live OP50-1, Mock Treated OP50-1 (2 hr) or 1% PFA Treated OP50-1 (2 hr). Samples separate predominantly on Component 1 by drug treatment. n = 3 biological replicates. (C) Combined total area values for all free fatty acids identified in samples collected in (B) reveal that biguanides reduce total free fatty acid abundance irrespective of bacterial growth conditions. Data represent mean +/- SEM, n = 3 biological replicates for each condition. **, P < 0.01; ****, P < 0.0001 by two-way ANOVA followed by Tukey’s multiple comparisons testing. (D-F) Volcano plots for all differentially expressed lipids reveal that ether lipids are preferentially sustained despite a global loss of somatic lipids observed, irrespective of bacterial growth conditions. Fold change and false discovery rate (FDR) calculations were performed with t-tests followed by Benjamini-Hochberg FDR adjustment using MetaboAnalyst 3.0. (G) Total sum normalized AUC measurement of FA 16:0 DMA levels across bacterial growth conditions and drug treatments reveal that biguanides increase 16:0 DMA levels irrespective of the bacterial growth and metabolic conditions. Data represent mean +/- SEM. Ns, P > 0.05, *; P < 0.05 by two-way ANOVA followed by Tukey’s multiple comparisons testing. 248 Figure 2 – figure supplement 4. Inactivation of ether lipid machinery disrupts biguanide- mediated lifespan extension independent of effects on bacterial growth or metabolism. Lifespan analyses of N2 (wildtype) or ads-1 mutant animals grown on Live OP50-1 (A-B), mock treated OP50-1 for two hours (C-D), or 1% PFA treated OP50-1 for two hours (E-F) reveal that ads-1-mediated ether lipid deficiency disrupts metformin (top row) or phenformin (bottom row) mediated lifespan extension independent of bacterial treatment. Results are representative of 2 biological replicates. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. Note that the results from panels A-B, C-D, and E-F share the same same-day wildtype controls as they originate from the same replicate. Please refer to Supplementary file 1 for tabular survival data and biological replicate summary statistics. 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +met ads-1 +veh ads-1 +met Figure 2 - figure supplement 4 C E B D A F vs. vs. ** ns **** vs. 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +phen ads-1 +veh ads-1 +phen 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +met ads-1 +veh ads-1 +met 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +phen ads-1 +veh ads-1 +phen 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +met ads-1 +veh ads-1 +met 0 10 20 30 40 0 50 100 Days Percent Survival N2 +veh N2 +phen ads-1 +veh ads-1 +phen vs. vs. * * **** vs. vs. vs. **** ns **** vs. vs. vs. **** * **** vs. vs. vs. ** ns **** vs. vs. vs. **** ns **** vs. Live OP50-1 Mock Treated OP50-1 (2 hr) 1% PFA Treated OP50-1 (2 hr) 249 Figure 4—figure supplement 1. Ether lipid biosynthetic genes are not necessary for daf-2- dependent lifespan extension, and fard-1 overexpression extends lifespan in a manner dependent upon ether lipid biosynthesis. (A) daf-2 mutants display extended lifespan relative to wild type animals. RNAi knockdown of fard-1, acl-7, and ads-1 does not impact lifespan extension in these mutants. (B-C) RNAi knockdown of fard-1 (B) and acl-7 (C) suppresses fard-1 overexpression(oe1)-associated lifespan extension. For A-C, results are representative of 2-3 biological replicates. ns, P > 0.05; ***, P < 0.001; ****; P < 0.0001 by log-rank analysis. For tabular survival data and biological replicates see also Supplementary file 1. Figure 4—figure supplement 1 A C B 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) wt (acl-7 RNAi) fard-1 (oe1) (EV RNAi) fard-1 (oe1) (acl-7 RNAi) 0 10 20 30 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) wt (fard-1 RNAi) fard-1 (oe1) (EV RNAi) fard-1 (oe1) (fard-1 RNAi) vs. ns vs. **** ns vs. vs. ns vs. **** ns vs. 0 10 20 30 40 50 60 0 20 40 60 80 100 Days Percent Survival wt (EV RNAi) daf-2 (EV RNAi) daf-2 (fard-1 RNAi) daf-2 (acl-7 RNAi) daf-2 (ads-1 RNAi) vs. *** vs. ns ns vs. ns vs. vs. **** vs. **** 250 Figure 5—figure supplement 1. Genetic induction of ferroptosis does not impact fard-1 overexpression nor biguanide-mediated lifespan extension. (A-C) Independent knockdown of glutathione peroxidases, gpx-1 (A), gpx-6 (B), or gpx-7 (C) by RNAi does not mitigate lifespan extension by integrated fard-1 overexpression (fard-1 oe3 and fard-1 oe4), as would be expected if fard-1 overexpression extended lifespan by lowering ferroptosis. gpx-1 unexpectedly extends lifespan in a non-additive manner with fard-1(oe). (D-F) Similarly, knockdown of gpx-1 (D), gpx-6 (E), or gpx-7 (F) by RNAi do not suppress phenformin-mediated lifespan extension. For A-F, results are representative of 2-3 biological replicates. Note that A-F contain the same wildtype (EV RNAi) controls as they are visualized from the same replicate of the study. ***, P < 0.001; ****, P < 0.0001 by log-rank analysis. For tabular survival data and biological replicates see also Supplementary file 1. Figure 5—figure supplement 1 B E D F A J C 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival fard-1 (oe4) (gpx-1 RNAi) wt (EV RNAi) fard-1 (oe4) (EV RNAi) wt (gpx-1 RNAi) 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival fard-1 (oe4) (gpx-6 RNAi) wt (EV RNAi) fard-1 (oe4) (EV RNAi) wt (gpx-6 RNAi) 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival fard-1 (oe3) (gpx-7 RNAi) wt (EV RNAi) fard-1 (oe3) (EV RNAi) wt (gpx-7 RNAi) vs. ns vs. vs. **** **** vs. *** vs. ns vs. vs. **** vs. ns vs. vs. **** ns ns 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival wt (gpx-6 RNAi) +phen wt (EV RNAi) +veh wt (EV RNAi) +phen wt (gpx-6 RNAi) +veh 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival wt (gpx-7 RNAi) +phen wt (EV RNAi) +veh wt (EV RNAi) +phen wt (gpx-7 RNAi) +veh 0 10 20 30 40 0 20 40 60 80 100 Days Percent Survival wt (gpx-1 RNAi) +phen wt (EV RNAi) +veh wt (EV RNAi) +phen wt (gpx-1 RNAi) +veh vs. ns vs. vs. **** **** vs. **** vs. ns vs. vs. **** ns vs. **** vs. vs. vs. **** ns vs. **** **** vs. *** vs. *** 251 Figure 6 – figure supplement 1. Biguanides do not activate gst-4 expression irrespective of bacterial diet. (A-B) GFP quantification of gst-4p::NLS::GFP animals treated from hatching with vehicle, 50 mM metformin, or 4.5 mM phenformin on either OP50-1 seeded NGM plates or EV HT115 seeded RNAi plates, and imaged at Adult Day 1. For (B), data represent the mean +/- SEM of at least 30 animals per condition (at least 10 animals per replicate combined from 3 biologically independent experiments). ns, P > 0.05; *, P < 0.05; **** P < 0.0001 by two-way ANOVA followed by Tukey’s multiple comparisons testing. (C) Representative images of dod- 24p::GFP animals treated from hatching with vehicle or 4.5 mM phenformin and grown on OP50- Figure 6 — figure supplement 1 A. 0 100 200 300 Mean Fluorescence Intensity (a.u.) gst-4p::GFP::NLS OP50-1 (NGM) HT115 (EV RNAi) Vehicle (AD1) + 50 mM Metformin (AD1) + 4.5 mM Phenformin (AD1) ns ! ns !!!! B. OP50-1 (NGM) HT115 (EV RNAi) Vehicle Metformin Phenformin gst-4p::GFP::NLS (AD1) C. Vehicle Phenformin OP50-1 (NGM) HT115 (EV RNAi) skn-1 RNAi fard-1 RNAi acl-7 RNAi ads-1 RNAi dod-24p::GFP (AD1) 252 1 seeded NGM plates, or RNAi plates seeded with EV, skn-1, fard-1, acl-7, or ads-1 RNAi and imaged at Adult Day 1, as quantified in Figure 6F. 253 ACKNOWLEDGEMENTS We thank Talia Hart, Dr. Gary Ruvkun, Dr. Eric Greer, and Dr. Keith Blackwell for discussions and constructive criticisms. This work was funded by NIH/NIA Grants R01AG058259 and R01AG69677 (to A.A.S.) and R01AG058610 (to S.P.C.), by the Weissman Family MGH Research Scholar Award (to A.A.S.), by a NSF GRFP Award 1000253984 (to L.C.), and by NIH/NIAID R01AI130289 (to R.P.W.), and by IRACDA NIH Grant K12GM106996 (to L.C.). Thanks to the University of Southern California and Buck Institute Nathan Shock Center (P30AG068345) for providing core services and support. Thanks to the NIH/NIDDK-funded NORC of Harvard (P30DK040561) and the NIH/NIDDK-funded Boston-Area DERC (P30DK057521) for core services. Some strains were provided by the CGC, funded by the NIH Office of Research Infrastructure Programs (P40OD010440), and the C. elegans Knockout Consortium. Figures 1A, 3C, and 7 were created with BioRender.com. Competing Interest Statement: The authors declare that no competing interests exist. Author Contributions: Conceptualization, LC, AAS; Methodology, LC, SL, FMA, YZho, YZha, AA, JW, MJ, NS, SPC, AAS; Validation: LC, SL, LM, FMA, AAS; Formal Analysis: LC, FMA, JW, MJ, ZL, SD, NDP, RPW, AAS; Investigation: LC, SL, FMA, NS, YZho, YZha, AA, LM, AY, SE, KD, JW, MJ, NDP, RPW, SPC, AAS; Writing – Original Draft: LC, AAS; Writing – Review & Editing: LC, SL, FMA, NS, YZho, YZha, AA, LM, AY, SE, KD, JW, MJ, ZL, SD, NDP, RPW, SPC, AAS; Visualization: LC, FMA, ZL, NS, SD, SPC, AAS; Supervision: AAS; Funding Acquisition: AAS. 254 Chapter 9: Riboflavin Depletion Promotes Longevity and Metabolic Hormesis in Caenorhabditis elegans *This chapter is a version of a manuscript published in Aging Cell. This project was done in collaboration with Alexander Soukas’ group, investigating how riboflavin depletion can promote longevity. I was responsible for data collected in Figure 1, including designing the experiments and writing the results, discussion, and necessary supplemental section. Authors: Armen Yerevanian 1,2 , Luke M. Murphy 1,2 , Sinclair Emans 1,2 , Yifei Zhou 1,2 , Fasih M. Ahsan 1,2 , Daniel Baker 1,2 , Sainan Li 1,2 , Adebanjo Adedoja 1,2 , Lucydalila Cedillo 1,2 , Nicole L. Stuhr 3 , Einstein Gnanatheepam 4 , Khoi Dao 5 , Mohit Jain 5 , Sean P. Curran 3 , Irene Georgakoudi 4 , Alexander A. Soukas 1,2,6 Affiliations: 1 Department of Medicine, Diabetes Unit and Center for Genomic Medicine, Massachusetts General Hospital, Boston, MA 02114 2 Department of Medicine, Harvard Medical School, Boston, MA 02115 3 Leonard Davis School of Gerontology, University of Southern California, Los Angeles, CA 90089 4 Department of Biomedical Engineering, Tufts University School of Engineering, Medford, MA 02155 5 Department of Medicine and Pharmacology, University of California San Diego, San Diego, CA 92023 255 6 Broad Institute of Harvard and MIT, Cambridge, MA 02142 Correspondence: asoukas@mgh.harvard.edu Keywords: Riboflavin, UPR mt , dietary restriction, longevity, rft-1, AMPK, FOXO, C. elegans 256 ABSTRACT Riboflavin is an essential cofactor in many enzymatic processes and in the production of flavin adenine dinucleotide (FAD). Here we report that the partial depletion of riboflavin through knockdown of the C. elegans riboflavin transporter 1 (rft-1) promotes metabolic health by reducing intracellular flavin concentrations. Knockdown of rft-1 significantly increases lifespan in a manner dependent, AMP-activated protein kinase (AMPK)/aak-2, the mitochondrial unfolded protein response, and on FOXO/daf-16. Riboflavin depletion promotes altered energetic and redox states and increases adiposity, independent of lifespan genetic dependencies. Riboflavin depleted animals also exhibit activation of caloric restriction reporters without any reduction in caloric intake. Our findings indicate that riboflavin depletion activates an integrated hormetic response that promotes lifespan and healthspan in C. elegans. 257 INTRODUCTION Healthy mitochondrial function requires the coordination of multiple cellular inputs including sufficient energetic substrates, amino acids and micronutrients. Vitamin cofactors are key to metabolic processes such as the citric acid cycle, electron transport chain, and energy shuttling into the cytosol. One family of vitamins that participate in mitochondrial physiology are the flavins 234 . The flavin co-factors include flavin mononucleotide (FMN) and flavin adenine dinucleotide (FAD) and are essential for redox chemistry and electron shuttling 235 . FAD is classically known to serve as an electron acceptor in the conversion of succinate to fumarate by succinate dehydrogenase in the citric acid cycle, as well as an electron donor to complex II of the electron transport chain. The flavins are also cofactors for multiple enzyme classes including the oxidoreductases and the fatty acid dehydrogenases 236 . They are derived from riboflavin, a water soluble ribitol derivative also known as vitamin B2. Riboflavin is an essential nutrient for all animals, and must be acquired either from food sources or from commensal gut flora 237 . The animal kingdom has evolved specific transporters to import riboflavin from the gut lumen and to transport them intracellularly 238 . In humans and mice, three isoforms of these transporters are expressed by three distinct genes: Slc52A1, Slc52A2 and Slc52A3 239-241 . Disruption in the function of these transporters is known to induce pathology through cellular riboflavin depletion 242 . Congenital deficiency in these transporters is associated with clinical syndromes in humans including Brown-Vialetto-Van Laere syndrome, where patients experience progressive neurologic deficits, and is treated successfully with extremely large doses of riboflavin 243,244 . 258 The C. elegans genome harbors two riboflavin transporter orthologs, rft-1 and rft-2 245 . Previous work has shown that the riboflavin transporters have differential expression and phenotypes based on knockdown. 245 . rft-1 expression is localized to the intestine and knockdown leads to a complete elimination of brood size. rft-2 expression occurs in the pharynx and intestine but is associated with reduced brood size. While loss of function of C. elegans orthologs of riboflavin transporters are known to lead to embryonic lethality, it is not known whether depletion of riboflavin after early embryonic and larval development has deleterious consequences. Based upon published literature, riboflavin deficiency is predicted to have broad metabolic impacts on the organism, including reducing cellular respiration by impacting the citric acid cycle and mitochondrial electron transport activity, as well as reducing the enzymatic function of a wide variety of oxidoreductase enzymes important for anabolic activity. It remains unknown whether the impact of such broad metabolic perturbations following development are deleterious, or whether there are advantageous aspects based upon activation of energetic stress responses under conditions of low riboflavin and FAD levels. Hormetic responses and extended lifespan due to energetic disruption are a well described phenomenon in C. elegans 246,247 . The traditional focus on co-factor biology has been that increasing micronutrient intake provides beneficial effects to the organism. We took the contrarian view that like macronutrient restriction, micronutrient restriction, under the right circumstances, could produce beneficial effects through creating energetic stress or other means, with riboflavin being an obvious candidate given its prominent role in metabolism. In the present study, we ask whether flavin co-factor depletion via disrupting the normal uptake of riboflavin can promote advantageous, metabolic stress defenses such as those activated by dietary restriction and mitochondrial energetic stress. We determine that physiologic riboflavin depletion alters cellular energetics and activates key longevity 259 factors AMPK, FOXO, and the mitochondrial unfolded protein response (UPR mt ). Riboflavin depletion via riboflavin transporter knockdown extends lifespan and promotes healthspan in C. elegans, dependent upon a AMPK-UPR mt -FOXO signaling relay. These data suggest that riboflavin depletion provides metabolic benefits and by leveraging factors mobilized by caloric restriction and energetic stress without actual reductions in caloric intake. Further, despite the deleterious impacts of monogenic diseases in riboflavin transporters, selective knockdown of transporters and alteration of riboflavin physiology may provide translational opportunities to manage energetic states and thus metabolic diseases of aging. 260 RESULTS Knockdown of rft-1 Promotes Longevity via Riboflavin Depletion The lack of brood and prominent intestinal expression suggested that rft-1 had the most prominent metabolic phenotype, so we examined consequences of its knockdown 248 . rft- 1 depletion via RNA interference (RNAi) prompts a significant, 25% increase in lifespan (Figure 1a). RNAi knockdown of rft-1 reduces the transporter’s mRNA by approximately 70%, suggesting that partial riboflavin transporter deficiency rather than complete knockdown induces this phenotype. (Figure S1a). mRNA encoding the paralogous riboflavin transporter rft-2 trends non-significantly upwards with rft-1 RNAi, which indicates that there is not significant compensation for rft-1 knockdown by rft-2, and confirms the specificity of the RNAi-based knockdown of rft-1 (Figure S1a) To confirm that the longevity phenotype of the rft-1 knockdown animals is due to reduced riboflavin uptake rather than a non-canonical effect of the transporter, we administered high dose riboflavin supplementation to attempt to overcome the deficit in transport. As expected, high doses of riboflavin abrogate the lifespan increase attributable to rft-1 RNAi (Figure 1a). This strongly suggests that riboflavin is the etiologic factor in the rft-1 RNAi phenotype, and further that depletion of riboflavin due to transporter deficiency is the source of lifespan extension. Worm lysates of young adult worms treated with rft-1 RNAi exhibit marked reductions in riboflavin, FMN and FAD levels as assessed by quantitative liquid chromatography/mass spectrometry (LC/MS) (Figure 1b). In parallel, we evaluated the expression of enzymes key to flavin co-enzyme synthesis including riboflavin kinase (rfk-1) and FAD synthetase (flad-1). rft-1 RNAi was not associated with reductions in expression of flad-1 and rfk-1 which produce FAD and FMN respectively, suggesting 261 stoichiometric depletion of FMN/FAD rather than reductions in the enzymes that govern production (Figure S1a). Previous descriptions of brood size deficits in rft-1 knockdown animals suggests that the germline may play a role in the phenotype 39,148,245 . We utilized rrf-1 (somatic RNAi blunted, germline RNAi competent) and ppw-1 (somatic RNAi competent, germline RNAi incompetent) mutants and examined brood size and lifespan on rft-1 RNAi. The loss of brood size is dependent on somatic action of rft-1 (i.e. normalized in rrf-1 mutants which do not efficiently conduct somatic RNAi). This suggests that a somatic process is altering metabolic and reproductive capacity (Figure 1c). Lifespan was also dependent on somatic action of rft-1, as rrf-1 mutants did not exhibit lifespan extension on rft-1 RNAi (Figure 1d). Germline loss of RNAi extended lifespan further on rft-1, suggesting that preserved germline uptake of riboflavin accentuates somatic depletion of riboflavin and potentiates the lifespan extension. (Figure 1e). rft-1 RNA no longer extends lifespan in a long-lived glp-1 mutant lacking germline stem cells and the effect is modestly blunted, suggesting that the germline is required to act as a riboflavin sink for beneficial effects of rft-1 RNAi to manifest (Figure 1f). The reduction in brood size raised the question of whether germ line stem cells and oocyte production is altered by riboflavin depletion. We examined the presence of the germline stem cells and oocytes via DAPI staining, which revealed an intact germline and oocyte production (Figure S1b). Animals treated with rft- 1 RNAi exhibit normal developmental timing (Figure 1g), indicating that delays in development do not play a role in lifespan extension with rft-1 knockdown. 262 In order to quantify temporal dynamics of somatic riboflavin-related gene expression as a proxy for flavin biology across the lifespan, we examined expression of riboflavin transporter and FAD synthetic enzymes in germlineless glp-4 mutants. Somatic rft-1 and rft-2 mRNAs increase in glp-4 animals at Day 4 but return back to normal levels through aging at Day 10. Rfk and flad-1 expression increase through the life of animals (Figure S1c). In spite of this, rft-1 RNAi-prompted lifespan extension is dependent upon whole life RNAi, as post-developmental RNAi fails to extend lifespan (Figure S1d). The most probable explanation for this observation is that rft-1 knockdown very likely needs to be manifest during L4 to young adult development when rapid germline expansion is operative to effectively “steal” and deplete somatic riboflavin levels. Riboflavin deficiency is associated with neurologic sequelae in mammals, and we wanted to verify that feeding behavior and motility is unchanged compared to control animals 148 . Knockdown of rft-1 does not affect food intake as measured up to adult day 3 by a food disappearance assay (Figure 1h). Pharyngeal pumping rates are also unchanged. (Figure S1e). Max crawling speed increases with rft-1 RNAi (vs. vector control) in L4 larvae, but is similar between adult day 1 and adult day 3 animals (Figure 1i). Average crawling speed shows similar patterns (Figure S1f). Animals spend equivalent amounts of time on food at L4 and Adult Day 1 when treated with vector vs. rft-1 RNAi (Figure S1g). Animal length and width was also similar across the lifespan. (Figure S1h) These findings suggest that animals treated with rft-1 knockdown exhibit normal developmental rate to adulthood, robust eating and size, and healthy activity levels in adulthood 263 suggesting that the nutritional depletion of riboflavin does not reduce healthspan to achieve lifespan extension. Riboflavin Depletion Promotes Longevity Through FOXO FOXO is known to act genetically downstream of nutrient-sending manipulations that extend lifespan 249,250 and we hypothesized that it may be activated in the context of riboflavin depletion. Indeed, loss of function in the sole C. elegans FOXO ortholog, daf- 16, is epistatic to lifespan extension with rft-1 RNAi, with or without the presence of supplemental riboflavin (Figure 2a). A DAF-16::GFP translational reporter demonstrates greater nuclear localization at adult day 1 and day 3 with riboflavin depletion, suggesting that rft-1 RNAi activates DAF-16 (Figure 2b and S2a). Transgenic DAF-16::GFP worms treated with empty vector and rft-1 RNAi as synchronous L1 were subsequently transferred at adult day 1 to plates with and without riboflavin. By Adult Day 3, additional riboflavin completely abrogates the DAF-16 nuclear localization evident in in rft-1 RNAi treated animals (Figure S2b). This indicates that the consequences of riboflavin depletion on daf-16 nuclear localization are reversible in adulthood. Confirming increased transcriptional activity of DAF-16 in the setting of riboflavin depletion, RNAi of rft-1 leads to the upregulation of a sod-3p::GFP reporter (Figure 2c). This activation is present through adult day 7, indicating consistent FOXO activation post-developmentally (Figure S2c). The activation of FOXO suggests either suppression of insulin-like/PI-3 kinase signaling (canonical FOXO activation) or activation via another mechanism. We examined whether 264 insulin signaling was playing a role by examining the effect of rft-1 RNAi on akt-1 and pdk- 1 gain of function mutants 251 . These animals are short lived due to constitutive inhibition of FOXO. The akt-1 gain-of-function mutation abrogates lifespan extension attributable to rft-1 RNAi (Figure 2d). The pdk-1 gain-of-function mutant on rft-1 RNAi, conversely, still exhibits lifespan extension (Figure S2d). This indicates that suppression of DAF-16 activity via augmented Akt signaling abrogates lifespan extension prompted by riboflavin deficiency. Riboflavin Depletion Alters Cellular Redox Ratio and Energetics We suspected that AMPK may be similarly activated by energetic stress with riboflavin deficiency, and further that AMPK activation may be mechanistically linked to lifespan extension with rft-1 RNAi. Knockdown of rft-1 fails to promote longevity with loss of function in the AMPKa catalytic subunit aak-2 (Figure 3a), and this pattern is not altered by addition of riboflavin (Figure 3b). Lifespan is still extended in long-lived aak-2oe animals with rft-1 RNAi (versus the same animals on vector RNAi control), suggesting that aak-2 is necessary but not sufficient for lifespan extension (Figure 3c). Consistent with AMPK activation by riboflavin depletion, phospho-AMPK T172 levels are increased in rft-1 RNAi treated animals, and this effect is abrogated by the addition of riboflavin (Figure 3d). In order to determine whether AMPK is required for activation of DAF-16 under riboflavin depletion, we examined DAF-16::GFP nuclear localization with and without functional aak-2. Under rft-1 RNAi conditions, DAF-16 nuclear localization still increases in the absence of aak-2 (Figure S3a). 265 Activation of AMPK suggests that modulation of cellular energetics might play a role in the longevity phenotype seen with the rft-1 knockdown. We hypothesized that reductions in flavin cofactor (FAD, FMN) concentrations induce mitochondrial stress responses due to changes in organellar energetics by altering redox state. We examined the impact of riboflavin depletion on the redox ratio utilizing label-free multiphoton microscopy and fluorescence lifetime imaging (FLIM) of intestinal cells in control and rft-1 RNAi treated animals (Figure S3b-c). Animals treated with rft-1 RNAi versus empty vector control- treated animals exhibit a significant decrease in the optical redox ratio, defined as the ratio of FAD/(NAD(P)H + FAD) calculated based on the autofluorescence signatures of the corresponding co-enzymes. There is also an increase in mitochondrial clustering, suggesting altered mitochondrial energetics in an oxidized state and morphologic changes to the mitochondrial network (Figure 3e) 252 . A significant decrease in the NAD(P)H protein bound fraction suggests decreased levels of glutaminolysis and enhanced utilization of the glutathione pathway 253 . An increase in the FAD bound fraction suggested that overall FAD depletion was causing aggressive capture of flavin co-factors by enzymatic machinery (Figure 3e). LC/MS metabolomics of rft-1 RNAi treated animals indicates significant changes in multiple metabolic pathways. Increases in purine catabolism metabolites are present, including xanthine, hypoxanthine, and guanosine (Figure 3f). Pathway enrichment analysis reveals that other than riboflavin metabolism, riboflavin deficiency leads to significant impact on metabolites in the glutathione and purine metabolic pathways (Figure S3d). Components of the citric acid cycle including citrate, isocitrate, and α-ketoglutarate, as well as ATP, are reduced by riboflavin depletion. Glutamate and glutamine levels are also reduced suggestive of disruptions in glutamine 266 synthesis (Figure 3f). Riboflavin Depletion Activates the Mitochondrial Unfolded Protein Response The frank changes in energetic status, altered redox ratio, and presence of mitochondrial clustering all suggested that mitochondrial stress responses may also be contributing to the longevity response to riboflavin deficiency. Indeed, the UPR mt is activated by rft-1 knockdown, as evidenced by induction of hsp-6p::GFP 254 on days 1 and 3 of adulthood, and this effect is abrogated by the addition of riboflavin (Figure 4a). Full activation of the UPR mt is known to require the transcription factor ATFS-1, which translocates to the nucleus to activate stress response pathways 255 . Established target genes of ATFS-1, including cdr-2, hrg-9, and C07G1.7 256 are upregulated with rft-1 knockdown, with the previously undescribed ATFS-1 target gene C07G1.7 exhibiting a 2000-fold increase (Figure 4b). The UPR mt activation is necessary for lifespan extension, as atfs-1 loss of function animals, which have lower lifespans compared to wild-type at baseline, do not exhibit lifespan extension with rft-1 knockdown (Figure 4c). In the setting of gain-of- function mutations in atfs-1, which lead to shortened lifespan 257 , rft-1 knockdown still promotes significant extension of lifespan (Figure 4d). This indicates that the UPR mt response is necessary but not sufficient for the riboflavin depletion longevity phenotype. We wished to evaluate the relationship between the UPR mt and DAF-16 and examined nuclear localization of DAF-16 in atfs-1 mutants. rft-1 RNAi in atfs-1;DAF-16::GFP animals revealed abrogated nuclear localization, suggesting that DAF-16 activation is partially dependent on atfs-1 and the UPR mt (Figure 4e). We evaluated sod-3 and hsp-6 expression in aak-2 and atfs-1 animals to determine if the transcriptional responses to 267 daf-16 and atfs-1 were present in these mutant strains with riboflavin depletion. QPCR revealed that the upregulation of hsp-6 and sod-3 with rft-1 RNAi are both dependent upon both atfs-1 and aak-2. (Figure 4f) Riboflavin Depletion Alters Somatic Lipid Stores The long-lived phenotype of riboflavin depletion and the role of flavin cofactors in beta- oxidation suggests that changes in lipid composition may be manifest following rft-1 knockdown. We hypothesized that changes in lipid metabolism occur upstream of or in parallel to FOXO activation, due to changes in enzymatic function (such as reduced activity of lipid dehydrogenases). RNAi of rft-1 induces significant increases in fat mass in both the intestine and the germline in adult day 1 worms, as exhibited by fixative-based oil-red-O and Nile red staining (Figure 5a, S5a) We examined the epistatic relationships of the UPRmt, FOXO and AMPK with regards to this high lipid phenotype. We also evaluated whether key lipid regulating pathways such as MTOR and nhr-49 play an important role in the lipid biology of riboflavin depletion. Epistasis analysis indicates that rft-1 RNAi increases fat mass in daf-16, aak-2, nhr-49 and raga-1 animals but not in atfs-1 or rict-1 animals by fixative-based Nile red staining (Figure 5b). Post-developmental rft-1 RNAi beginning at young adult stage in enhanced RNAi eri-1 animals also increases fat mass, indicating that riboflavin deficiency does not impact lipid metabolism exclusively through a developmental pleiotropy (Figure 5b). Confirming these observations and further delineating the nature of the lipids increased in abundance following rft-1 RNAi, stimulated Raman scattering (SRS) analysis of live 268 adult day 1 animals indicates increased total signal of unsaturated fatty acids and the unsaturated to total lipid ratio in riboflavin deficiency (Figure S5b) 258,259 . By gas chromatography/mass spectrometry (GC/MS) of triglyceride and phospholipid fractions separated by solid phase extraction, global triglyceride stores increase by 40% in both young adult and adult day 1 rft-1 RNAi-treated animals, consistent with the spectroscopic imaging and fixative-based lipid staining (Figure 5c). While only small changes are evident by young adulthood (Figure S5c), by adult day 1 animals exhibit significant differences in their lipid composition, with increases in unsaturated and branched chain fatty acids, and reductions in cyclopropyl fatty acids in both triglyceride and phospholipid fractions (Figure 5d, S5c). Due to its role in lipid oxidation we examined whether rft-1 lifespan extension was dependent upon nhr-49. rft-1 RNAi significantly extends the lifespan of nhr-49 mutants (Figure 5e). Due to increases in branched chain fatty acid synthesis, we examined the expression of acdh-1, which is a known branched chain dehydrogenase in elegans and that has been previously reported as a dietary sensor 260 . An acdh-1 promoter GFP reporter is significantly increased ~70% with rft-1 RNAi at adult day 1 (Figure S5d). In order to begin to determine whether unsaturated fatty acids are elevated in riboflavin deficiency owing to increased production vs utilization, we examined expression of the fatty acid desaturase fat-7 261 . Riboflavin depletion does not promote changes in fat-7 expression early in life but preserves it with aging (Figure S5e). Riboflavin Depletion Activates Dietary Restriction Pathways The long-lived phenotype of riboflavin depletion, concomitant with decreases in energetics, AMPK activation, and impairment in lipid beta-oxidation, suggested to us that 269 riboflavin depletion mimics some features of a dietary restriction-like phenotype. The acs- 2 and bigr-1 genes are well established to be transcriptionally upregulated during periods of caloric restriction in C. elegans 262,263 . rft-1 RNAi induced bigr-1::RFP and acs-2p::GFP expression with age (Figure 6a). We sought to assess whether other canonical caloric restriction factors and processes were involved in riboflavin depletion. We examined eat- 2 animals, which have extended lifespan owing to defective pharyngeal pumping, and noted that the animals experience further lifespan extension with rft-1 RNAi (Figure 6b) We also examined the C. elegans FOXA homolog pha-4, known to be epistatic to caloric restriction-mediated longevity 264 . Lifespan extension with rft-1 RNAi is dependent on pha- 4/FOXA (Figure 6c). Inhibition of target of rapamycin (TOR) signaling is also important in the response to dietary restriction. Thus, we examined whether mutants in the TOR complex 1 (TORC1) and TOR complex 2 (TORC2) pathways exhibit longevity with riboflavin depletion. RAGA/raga-1 mutants, which have defects in TORC1 activation, experience lifespan extension on rft-1 RNAi (Figure 6d). To further determine whether altered TORC1 activity is required for the hormetic effects of riboflavin depletion, we used a strain of elegans that contains a knock-in, humanized S6K, which permits immunoblotting for phospho-S6K to determine the activity of TORC1. No difference is evident in p-S6K between rft-1 RNAi and empty vector control, suggesting that altered TORC1 signaling is not essential to the biological response to riboflavin depletion (Figure 6e). In contrast, loss of function mutations in the essential TORC2 subunit rict-1 experience no lifespan extension with rft-1 RNAi, indicating that riboflavin depletion requires TORC2 activity to exact its favorable effects (Figure 6f). 270 DISCUSSION Vitamins, as essential co-factors for life, have traditionally been viewed as highly beneficial entities independent of their concentrations. This is particularly true of the B- vitamins, which are water soluble and do not exhibit significant toxicities at moderate supraphysiologic doses. Our work in C. elegans counters the notion that more is always better, as depletion of key enzymatic co-factor riboflavin can trigger metabolic and physiologic stress responses that are hormetic in nature and extend lifespan. We were initially concerned that the reduction of the flavin cofactors such as FAD and FMN would be frankly toxic to the organism. This was particularly true when LC/MS revealed that FAD levels in the rft-1 treated animals were 80-90% below normal. We anticipated that the loss of FAD to this level would prevent the function of succinate dehydrogenase and the ability of the electron transport chain (ETC) to absorb electrons produced by the citric acid cycle. Our data does indeed suggest disruptions in mitochondrial respiration, with decreased redox ratios, mitochondrial clustering, reduced ATP production and activation of the UPR mt . Unexpectedly, however, riboflavin depletion has lifespan dependencies that differ from of classic ETC disruption, such as in cco-1 and frh-1 (frataxin) knockdowns 265,266 . Previous examinations of cco-1 and frh-1 RNAi have shown that lifespan extension with ETC disruption is AMPK and FOXO independent 265,266 and atfs-1 independent 257 . These epistatic relationships to lifespan extension suggest a different “flavor” of UPR mt , and that riboflavin depletion does not represent a solitary poisoning of the ETC as seen through complex I-IV knockout or knockdown. Riboflavin depletion is likely having pleiotropic effects leading to alternative ways of activating UPR mt 271 (ie enzymatic disruption), and that the function of the UPR mt under these circumstances may be managing stress responses that are not purely energetic in origin. The central role of AMPK in the lifespan extension suggests that energetic perturbation is still relevant to riboflavin depletion-mediated longevity. Riboflavin depleted animals exhibit normal food consumption, normal TORC1 activation, and elevated triglyceride stores. Despite evidence for ample macronutrient availability, the animal experiences apparent energetic deficits with activation of AMPK during rft-1 RNAi. This begged the question as to whether the animal is activating dietary restriction (DR) pathways independent of true DR. The lifespan dependencies on FOXO/daf-16 and FOXA/pha-4, as well as activation of canonical “starvation” reporters acs-2 and bigr-1, provide evidence that the animal is utilizing these pathways in the context of micronutrient depletion only. Amino acid sensing and dietary restriction via essential nutrients such as methionine have been previously described to extend lifespan 150 . Depletion of canonical vitamin co-factors however, has proven deleterious in previous investigations. Depletion of biotin, B12 derivatives, and folate have previously shown to shorten lifespan 267,268 . To our knowledge, our work is the first to show that depletion of a vitamin co-factor can mimic features of dietary restriction and extend lifespan using shared molecular mechanisms. We anticipated that the relationships between the UPR mt , AMPK activity and daf-16 activation would be epistatic based on the hypothesized most proximal impact of having mitochondrial dysfunction. We noted however that FOXO nuclear localization occurs independent of AMPK but partially dependent on atfs-1, and that upregulation of sod-3 is 272 also dependent on atfs-1. Transcriptional activation of hsp-6 is unexpectedly abrogated in aak-2 mutants during riboflavin depletion, suggesting that AMPK activity is necessary for potentiation of the UPR mt . This complex interplay of these dependencies hints that there is a concerted cellular response to reductions of the flavin co-factors. This creates the exciting opportunity to explore for flavin sensing molecular systems that converge on these key stress factors either upstream or downstream and may provide new avenues to activate pro-longevity paradigms. (Figure 6g). The lipid phenotypes provides some clues to the nature of flavin sensing. Elevated triglyceride stores following riboflavin depletion is independent of canonical lifespan regulating pathways such as FOXO, AMPK and TORC1. This decoupling of fat mass and lifespan suggest that the lipid phenotype may be regulated by enzymatic processing of lipids upstream of the energetic stress axes. The exceptions to this were the atfs-1 and rict-1 mutants. Phenotypes associated with UPR mt activation are known to induce lipid accumulation 269,270 . Recent work has identified NHR-80 as a key regulator of citrate sensing and lipid accumulation in the UPR mt phenotype 269 . The relevance of atfs-1 activity to dve-1 and ubl-5 function suggests that the UPR mt may be instrumental in the communication of flavin depletion and related citric acid cycle disruptions on organismal energetics. The lack of fat mass increase in rict-1 mutants, which at baseline exhibit higher lipid content, suggests either a dependency or inability for riboflavin depletion to overcome the excess lipid accumulation associated with TORC2 knockout. TORC2 has been described as a nutritional sensor that regulates lipid biogenesis 271 , and it is entirely plausible that there are distinct inputs in mitochondrial energetics, mito-stress and TORC2 273 activation that are governed by flavin biology. Further investigation would be beneficial to identify whether TORC2 can directly sense changes in flavin levels, as this would have significant implications for the nutritional regulation of anabolic signals in senescence and cancer. rft-1 RNAi does not impact developmental rate and metabolic phenotypes manifest most impressively at the young adult to adult day 1 transition. This is in parallel to the growth and development of the germline and the oocytes. This pattern suggests either that the larval stages are relatively resistant to riboflavin depletion, or, more likely, contain and accumulate sufficient flavin cofactors at time of egg lay and during early larval development (prior to rft-1 knockdown) to proceed through development normally. We surmise that somatic growth dilutes endogenous flavin cofactors, subsequently inducing the favorable effects of riboflavin deficiency uniquely in adulthood. Further, development of the germline and riboflavin shunting into oocytes in late larval and early adult stages may prompt further, rapid riboflavin depletion, inducing the metabolic stress required to induce the phenotype identified in this work. This was particularly reflected in the germline RNAi deficient animals which revealed more pronounced lifespan extension compared to total body RNAi. It is worthy of mention that riboflavin deficiency leads to sterility, but this sterility is not accompanied by a decrease in germline stem cell numbers. Thus, effects on the germline are unlikely to be responsible for the shifts in lifespan and fat mass evident in riboflavin deficiency. 274 The presence of a post-developmental fat increase with depletion of riboflavin suggests that acute depletion in adulthood has important impacts that are likely different from depletion starting at larval stages. The most likely etiology for these post-developmental changes are alterations in enzymatic activity due to loss of flavin co-factors. The flavin co-factors are important for a wide variety of enzymes particularly those associated with oxido-reductase functions including the fatty acid dehydrogenases. The ‘flavoproteome’ is an established set of enzymes requiring FAD, FMN or riboflavin to function 236 . The impact of riboflavin depletion globally on the proteome is likely to produce stoichiometric shifts in key metabolites that will alter global physiology. Differential utilization of different dehydrogenases (branched chain vs. long chain) may also explain the unique lipid phenotype that is produced with riboflavin depletion. Using metabolomics, we identified other examples of likely enzymatic effects, with evidence of reductions in purine catabolism likely due to loss-of-function in xanthine dehydrogenase. Alterations in xanthine metabolism have been previously described as beneficial and lifespan extending 272 . The impact of riboflavin depletion on enzymatic processes is complex and there are likely to be both hormetic and harmful impacts of this process. Identifying the enzymatic pathways where riboflavin depletion provides beneficial versus detrimental impacts will provide new insights into mechanistic targets promoting longevity. We suggest that further investigations into the functions of the flavoproteome and flavin biology will serve to identify new therapeutic and investigational targets for the metabolism of aging and aging associated diseases. 275 MATERIALS & METHODS C. elegans Genetics Strains were maintained at 20 o C on nematode growth medium (NGM) plates seeded with E. coli OP50. All experiments were conducted at 20 o C unless otherwise specified. The following strains were utilized: wild type (N2 Bristol ancestral), NL3511 ppw-1(pk1425), NL2098 rrf-1(pk1417), daf-16(mgDF47), TJ356 zls356[daf-16p::daf-16a/bGFP+rol- 6(su1006), CF1553 muls84[(pAD76)sod-3p::GFP+rol-6(su1006)], GR1318 pdk- 1(mg142gf), GR1310 akt-1(mg144gf), RB754 aak-2(ok524), VC3201 atfs-1(gk3094), QC118 atfs-1(et18), SJ4100 zcls13[hsp-6p::GFP+lin-15(+)], DMS303 nls590[fat-7p::fat- 7::GFP +lin15(+)], VL749 wwls24[acdh-1p::GFP +unc-119(+)] MGH266 rict-1(mg451), VC222 raga-1 (ok386), MGH559 aak-2(ok754);zls356[daf-16p::daf-16a/b::GFP+rol- 6(su1006)], MGH249 alxls19 [bigr-1::bigr-1::mRFP3-HA;myo-2p::GFP], WBM392 Is[Pacs-2::GFP+rol-6(su1006)], MGH430 rsks-1(alx48 humanized S6K hydrophobic motif). MGH113 nhr-49(nr2041), CB4037 glp-1(e2141), DA465 eat-2(ad465), WBM60 uthls248[aak-2p::aak-2(genomic aa1-321)::GFP::unc-54 3’UTR+myo-2p::tdTOMATO], MGH600 atfs-1(gk3094);zls356[daf-16p::daf-16a/b::GFP+ rol-6(su1006)]. E. coli Strains Non-RNAi experiments were all conducted on NGM plates containing E. coli OP50-1 (CGC) as the food source and used 3-7 days after seeding. Cultures of E. coli OP50 were grown in Luria Broth (LB) for 15 hrs. at 37 o C without shaking and seeded directly onto NGM plates. RNA interference experiments were conducted using E. coli HT115(DE3) bacteria (Ahringer library) as the food source. Clones were isolated from the primary RNAi library and plated on ampicillin/tetracycline plates. Individual clones were grown in LB 276 broth for 15 hours with shaking. Cultures were concentrated 1:5 and seeded directly onto NGM plates containing 5 mM isopropyl-B-D-thiogalactopyranoside and 200 mg/ml carbenicillin. Plates were used 1-5 days after seeding. All RNAi clones were sequence verified prior to utilization. Riboflavin solution or vehicle was applied to the plate and allowed to dry for at least 30 minutes prior to seeding with animals. Riboflavin Treatment Culture grade riboflavin (Sigma-Aldrich) was dissolved in 50mM NaOH to a concentration of 13.3 mM (5 mg/ml). Fully seeded plates were treated with 500ul riboflavin solution (final concentration 665 µM) and allowed to dry on the plate prior to worm placement. Vehicle plates were treated with 500ul 50mM NaOH solution. Longevity assays Lifespan analysis was conducted at 20 o C except where indicated. Gravid adults were grown on NGM plates and isolated eggs were incubated overnight in M9 solution for hatching. Synchronized L1 animals were seeded unto RNAi plates and allowed to grow until the adult stage. Adult animals were subsequently transferred to fresh RNAi plates every other day until post-reproductive stage where they were maintained on a single plate. Dead worms were counted daily or every other day. Statistical analysis for survival curves was performed using OASIS2 software 226 . 277 Development Assays Synchronized L1s were prepared via bleach prep and plated on RNAi plates containing empty vector or rft-1 RNAi grown at 20 o C. Larvae were examined every 2 hours starting 41 hours after drop and scored for their transition to adulthood by the appearance of the vulvar slit. Brood Size Fifty synchronized L1 animals of each strain were dropped on EV and rft-1 RNAi treated plates. Two days later, 2 young adult animals from each condition were dropped onto new EV and rft-1 RNAi plates respectively. These animals were transferred every 2 days until the two adults from each condition became post-reproductive. All animals on residual plates were counted once they reached L4/YA to calculate brood size. Pharyngeal Pumping Pumping rate was determined using a Sony camera attached to a Nikon SMZ1500 microscope that recorded 10 well fed Day 1 adult animals per genotype. Pharyngeal contractions in 15 second time periods for 4 technical replicates were counted (by slowing video playback speed by 4x) for each animal using OpenShot and pumping rates per minute were calculated. Food Intake and Activity Assays Food intake experiments were adapted from 77 . Food intake was assessed in RNAi liquid media without antibiotics in flat-bottom, optically clear 96-well plates with 150 μL total 278 volume. Age-synchronized nematodes were seeded as L1 larvae and grown at 20°. 5- Fluoro-2′-deoxyuridine (FUDR) was added 48 h after seeding at a final concentration of 0.12 mM. OD600 of each well was measured using a plate reader every 24 h starting at L1 stage and ending at day 5 of adulthood (168 h after dropping L1s). The fraction of animals alive was scored microscopically every day until the last day of the assay. Food intake per worms was calculated as bacterial clearance divided by worm number in well. Measurements were then normalized to the L4 to day 1 adult clearance rate for each condition. Lawn avoidance assays were conducted as described in 77 . Bacteria was grown overnight in liquid culture of LB with corresponding antibiotics. The next day, bacteria were collected at the log phase, seeded onto RNAi plates at 5X concentration, dried, and allowed to grow overnight at 20°. L1s were dropped onto RNAi plates with each. Plates were checked 48 h later at the L4 stage, and the number of worms on and off food were counted. For size and movement assays, 30-50 worms were washed off of a plate in 50 uL of M9 with a M9+triton coated P1000 tip and dropped onto an unseeded RNAi plate. The M9 was allowed to absorb and worms roamed on the unseeded plate for 1.5 hours before imaging crawling. Crawling was imaged with the MBF Bioscience WormLab microscope and analysis was performed with WormLab version 2022. Worm crawling on the plate was imaged for 1 minute for each condition at 7.5 ms. Worm crawling was analyzed with the software and only worms that moved for at least 90% of the time were included in the analysis. Oil-Red-O and Nile Red Staining Lipid staining protocol was adapted from Escorcia et al 26 . Adult day 1 animals were collected via washing and washed twice with M9 solution. Animals were then fixed with 279 40% isopropanol for 3 minutes with shaking. For oil-red-O staining, working solution of oil-red-O was generated from stock solution and fixed animals were stained in working solution for two hours. Animals were subsequently placed in M9 solution to remove excess stain and were imaged on a Leica Thunder multichannel microscope to generate composite images. For Nile red staining, Nile red working solution was generated by mixing 6 ul/Nile red stock solution in 1 ml isopropanol. Animals were stained in working solution for two hours followed by 30 minutes of wash in M9 solution. Nile red imaging was performed on the Leica Thunder GFP setting with 10ms exposure at 5X magnification. Western Blotting Worms were isolated by washing with M9 buffer and centrifuged into a pellet. Worm lysates were prepared by adding RIPA buffer and proteinase inhibitor cocktail (Roche) followed by water bath sonication in a Diagenode Bioruptor XL 4 at maximum strength for 15 minutes. Lysates were cleared of debris via centrifugation at 21,000g for 15 minutes at 4 o C and supernatant was collected. Protein concentration as measured using the Pierce BCA Assay (Thermo Fisher). Lysate was subsequently mixed with 4X Laemmli buffer (Bio-Rad) and boiled for 10 minutes. Samples were run on SDS-PAGE protocol (Bio-Rad) and transferred to nitrocellulose membrane via wet transfer at 100V for 1 hour. Immunoblotting was performed according to primary antibody manufacturer’s protocols. Secondary antibody treatment utilized goat -anti-rabbit HRP conjugate or goat-anti- mouse-HRP conjugate (GE Healthcare) at 1:10,000 and 1:5000 dilutions, respectively. HRP chemiluminescence was detected via West-Pico substrate (Thermo Pierce). The 280 western blot results shown are representative of at least two experiments. Primary antibodies used were the following: Rabbit monoclonal anti-Phospho-AMPKα (Thr172), Cell Signaling Technology Rabbit monoclonal anti-p70 Phospho-S6 Kinase (Thr389), Cell Signaling Technology Mouse monoclonal anti-Actin, Abcam Quantitative RT-PCR Worms samples were flash frozen in liquid nitrogen and kept in -80°C until RNA preparation. Samples were lysed through the use of metal beads and the Tissuelyser (Qiagen) Total RNA was extracted using RNAzol RT (Molecular Research Center) according to manufacturer instructions. RNA was treated with RNAse free DNAse prior to reverse transcription with the Quantitect reverse transcription kit (Qiagen). Quantitative RT-PCR was conducted in triplicate using a Quantitect SYBR Green PCR reagent (Qiagen) following manufacturer instructions on a Bio-Rad CFX96 Real-Time PCR system (Bio-Rad) Expression levels of tested genes were presented as normalized fold changes to the mRNA abundance of control genes indicated in the figures by the δδCt method. The primers used for the qPCR are as follows: rft-1 forward: GCTATTGTTCAGATCGCGTGC rft-1 reverse: CAGAGACCCAATTGACAAATACATGC rft-2 forward: CGGGAGTTGTTCAGATCGCT rft-2 reverse: GAGTCCCAGTTGACAACAGCA rfk forward: TGTTGGAAAAAGAAACGAAAGAA 281 rfk reverse: TCGATTAAAATTCGGTAACAACG flad-1 forward: TGCCTGGAGTTCCAAAATTC flad-1 reverse: GAAGGGCTGGGTGTTTTACA C07G1.7 forward: GCTGAAGAAGCTTCAACCGTAG C07G1.7 reverse: TCTCGTGTCAATTCCGGTCT hrg-9 forward: TGGAATATTGAGTGGCGTTG hrg-9 reverse: CCTCCTCTACTTGGTGCATGT cdr-2 forward: CGAGCCTCATTTGGAAAGAA cdr-2 reverse: GCATCTGCCGCTGTAACTTT sod-3 forward: GCAATCTACTGCTCGCACTG sod-3 reverse: TGCATGATTTCATGGCTGAT hsp-6 forward: CGAAGACCCAGAGGTTCAAA hsp-6 reverse: AATGCTCCAACCTGAGATGG GC/MS Lipid Analysis Lipid extraction and GC/MS of extracted, acid-methanol derivatized lipids was performed as described previously 227 . Briefly, 10,000 synchronous mid-L4 animals were sonicated with a probe sonicator on high intensity in a microfuge tube in 100-250 microliters total volume. Following sonication, lipids were extracted in 3:1 methanol: methylene chloride following the addition of acetyl chloride in sealed borosilicate glass tubes, which were then incubated in a 75°C water bath for 1 hour. Derivatized fatty acids and fatty alcohols were neutralized with 7% potassium carbonate, extracted with hexane, and washed with 282 acetonitrile prior to evaporation under nitrogen. Lipids were resuspended in 200 microliters of hexane and analyzed on an Agilent GC/MS equipped with a Supelcowax- 10 column as previously described 228 . Fatty acids were indicated as the normalized peak area of the total of derivatized fatty acids detected in the sample, normalized by recovery of spiked-in, standard phospholipid and triglyceride. LC/MS Metabolite Analysis Four biologic replicates of adult day 1 wild-type worms treated either with empty vector or rft-1 RNAi were collected with M9 wash and frozen by liquid nitrogen into a worm pellet. Polar metabolites of homogenized worms were analyzed using a Thermo QExactive orbitrap mass spectrometer coupled to a Thermo Vanquish UPLC system, as previously described 273 . Bioactive lipids metabolites were profiled on the same system, as previous described 274 . Collected data were imported into the mzMine 2 software suite for analysis (version 2.53). Metabolites were annotated by using an in-house library of commercially available standards. Please see supplemental methods for detailed methods. All mass integration values, normalized abundance values, significance testing scores, and pathway enrichment scores are included in this manuscript as Supplementary Table 2. Quantification and Statistical Analysis All western blotting quantifications were conducted on Bio-Rad Image Lab. Intensity analysis for fluorescent images was performed on ImageJ. Statistical analyses were performed using Prism (GraphPad Software). The statistical differences between control and experimental groups were determined by two-tailed students t-test (two groups), one- 283 way ANOVA (more than two groups), two-way ANOVA (two independent experimental variables), or log-rank (survival analyses) as indicated in each figure legend, with numbers of samples indicated and corrected P values < 0.05 considered significant. Fluorescence Microscopy DIC, brightfield and fluorescence Imaging of animals was performed utilizing the Leica Thunder microscopy system. Animals were picked onto a slide containing agar and 2.5mM levimasole solution. Imaging was performed within 5 minutes of slide placement. GFP and RFP Images were taken at 10ms exposure at 30% FIM and at 5X magnification, unless otherwise specified. Fluorescence intensity for quantification was calculated utilizing ImageJ software. For signal intensity experiments, quantification was performed on 20 worms (10 worms of two biological replicates). Two-Photon and Fluorescence Lifetime Imaging Wild type and mutant C. elegans were immobilized for fluorescence imaging using a previously proposed protocol 275 . Endogenous two-photon excited fluorescence (TPEF) images of C. elegans were acquired using a laser scanning microscope (Leica TCS SP8, Wetzlar, Germany) equipped with a femtosecond laser (Insight Deep See, Spectra Physics, Mountain View, CA). Fluorescence lifetime images (512 × 512 pixels) of C. elegans were acquired using the same excitation and emission settings as for intensity NAD(P)H and FAD images and a PicoHarp 300 time-correlated single photon counting unit (PicoQuant, Berlin, Germany) integrated in the Leica SP8 system. Please see Supplemental methods for further details on methods and analysis. 284 Stimulated Raman Scattering Imaging Stimulated Raman scattering (SRS) images of C. elegans were acquired using a laser scanning confocal microscope (Leica TCS SP8, Wetzlar, Germany) equipped with a picosecond NIR laser (picoEmerald, APE, Berlin, Germany). SRS images were acquired in the wavenumber range of 2798 cm -1 to 3103 cm -1 with an interval of 6 cm -1 . The Nd:VAN 1064.2 nm output was used as the SRS Stokes beam and the OPO beam tuned from 800 nm to 820 nm with step size of 0.4 nm was used as the pump laser. SRS images (620 × 620 microns x 51 wavenumbers, 512 × 512 pixels x 51 wavenumbers, 0.75 zoom) were acquired using a water immersion objective HCX IRAPO L 25x/0.95 NA with pixel dwell time of 4.9 μs. The pixels corresponding to regions occupied by C elegans were identified by implementing a global threshold of 300 (intensities ranged from 0 to 800). The SRS spectrum of each remaining pixel was normalized by the maximum SRS value of the entire field spectral image. To estimate the relative unsaturation level of the lipids in C. elegans, a ratio metric approach was adapted 258,259,276,277 . Specifically, the ratio of the area under the SRS spectrum for wavenumbers spanning 2991 and 3022 cm- 1 and wavenumbers spanning 2830 and 2870 cm-1 ,was estimated to represent the relative unsaturation levels 258,259,276,277 . Both fluorescence and SRS images were calibrated for laser power before analysis. 285 FIGURES FIGURE 1. rft-1 RNAi depletes flavins and extends lifespan. (a) Knockdown of the riboflavin transporter rft-1 via RNA interference significantly extends lifespan in wild type (wt) animals versus empty vector RNAi (EV). Addition of 665 µM riboflavin abrogates this lifespan extension (see table S1 for tabular data and replicates). (b) LC-MS/MS analysis of worm lysates collected at adult day 1 treated with rft-1 RNAi reveals significant reductions in organismal riboflavin, FMN, and FAD concentrations. (c) Brood size is diminished in rft-1 RNAi treated animals and remains suppressed in ppw-1 animals (somatic RNAi competent, germline RNAi incompetent) but is rescued in rrf-1 animals (b) (a) (c) (d) (e) (f) 35 37 39 41 43 45 47 49 51 53 59 61 63 0.0 0.2 0.4 0.6 0.8 1.0 Hour Since Hatching Fraction Reaching Adulthood wt (EV RNAi) + vehicle wt (EV RNAi) + riboflavin wt (rft-1 RNAi) + vehicle wt (rft-1 RNAi) + riboflavin vs ns vs vs vs ns ns ns Riboflavin FMN FAD 0 100000 200000 300000 Normalized Concentration (A.U) EV RNAi rft-1 RNAi ✱ ✱ ✱ 0 10 20 30 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) wt (EV RNAi) + Riboflavin wt (rft-1 RNAi) + Riboflavin vs ns vs vs ** ns vs ** vs ns 0 10 20 30 40 0.00 0.25 0.50 0.75 1.00 Days Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) rrf-1 (EV RNAi) rrf-1 (rft-1 RNAi) vs vs vs vs * * ns * 0 10 20 30 40 0.00 0.25 0.50 0.75 1.00 Days Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) ppw-1 (EV RNAi) ppw-1 (rft-1 RNAi) vs vs vs vs * * ns * 0 10 20 30 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) glp-1 (EV RNAi) glp-1 (rft-1RNAi) vs vs vs vs * ns * * wt rrf-1 ppw-1 0 100 200 300 400 Brood Size (# progeny) EV RNAi rft-1 RNAi ✱ ✱ (g) (h) (i) L4 Adult Day 1 Adult Day 3 0.0 0.5 1.0 1.5 Max Speed Max Speed (body length/s) EV RNAi rft-1 RNAi ✱ ns ns L1L2 L3 L4 Adult Day 1 Adult Day 2 Adult Day 3 Adult Day 4 -0.002 0.000 0.002 0.004 0.006 Food Intake Arbitrary Units (A.U) EV RNAi rft-1 RNAi 286 (somatic RNAi blunted, germline RNAi competent), suggesting a somatic site of rft-1 action. (d) rrf-1 animals exhibit lifespans equivalent to wt on rft-1 RNAi.(e) ppw-1 mutants experience lifespan extension with rft-1 RNAi. (f) Germline deficient glp-1 animals (kept at the non-permissive temperature, 25°C) do not experience additional lifespan extension with rft-1 RNAi. (g) Developmental rate from the first larval stage to adulthood is unchanged in animals treated with rft-1 RNAi, with or without riboflavin. (h) Food intake for animals treated with EV and rft-1 RNAi is not different across development and adulthood. (i) Maximum speed levels are higher in L4 and intact for adult day 1 and day 3 animals treated with rft-1 RNAi compared to controls. For lifespans results represent at least two biological replicates. See table S1 for tabular data and replicates of survival analyses. * indicates P < 0.05, **, P < 0.001 by log-rank analysis (a through g), standard two-way ANOVA (b,c,h,i). Bars represent means ± SEM. NS, non-significant. 287 FIGURE 2. Riboflavin depletion promotes longevity by activating FOXO/daf-16. (a) Lifespan extension with knockdown of the rft-1 transporter is abrogated in daf-16 mutants, with and without supplemental riboflavin. (b) Nuclear localization of DAF-16 was increased following rft-1 RNAi in a riboflavin-dependent manner as assessed by a DAF- 16::GFP translational reporter. Animals were binned into no nuclear localization, low levels of localization and high levels of localization (see Fig. S2 for representative 0 10 20 30 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) daf-16 (EV RNAi) daf-16 (rft-1 RNAi) vs vs vs vs No Riboflavin * * ns * 0 10 20 30 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) daf-16 (EV RNAi) daf-16 (rft-1 RNAi) vs vs vs vs Suppl. Riboflavin ns * ns ns 0 10 20 30 40 0.0 0.5 1.0 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) akt-1GOF (EV RNAi) akt-1GOF (rft-1 RNAi) vs vs vs * vs * * * Young Adult Adult Day 1 Adult Day 3 0 20 40 60 80 100 120 EV RNAi + vehicle Individuals None Low High Young Adult Adult Day 1 Adult Day 3 0 20 40 60 80 100 120 rft-1 RNAi + vehicle Individuals None Low High Young Adult Adult Day 1 Adult Day 3 0 20 40 60 80 100 120 rft-1 RNAi + riboflavin Individuals None Low High Young Adult Adult Day 1 Adult Day 3 0 20 40 60 80 100 120 EV RNAi + riboflavin Individuals None Low High Adult Day 1 Adult Day 3 Adult Day 5 Adult Day 7 0 10 20 30 40 50 60 Signal Intensity (Arbitrary Units) EV RNAi rft-1 RNAi ✱ ns ✱ ✱ Fold Change 1.16 2.02 2.06 2.05 sod-3p::GFP (a) (b) (c) (d) DAF-16::GFP Nuclear Localization 288 images). N = 120 animals per condition, representative of two biological replicates. (c) A sod-3p::GFP transcriptional reporter indicates increased activity of DAF-16 significantly over the lifespan of rft-1 RNAi treated animals. (d) A chromosomally-located akt-1 gain- of-function mutation blunts the longevity response to riboflavin depletion vs. wt. For a-d, results are representative of at least two biological replicates. * indicates P < 0.05 by log- rank analysis (a and d), and by two-way ANOVA followed by Sidak’s multiple comparisons post-hoc test (c). See table S1 for tabular data and replicates of survival analyses. Bars represent means ± SEM. 289 FIGURE 3. Riboflavin depletion alters cellular energetics. (a) Lifespan extension with rft- 1 RNAi is abrogated in AMPK/aak-2 mutants. (b) Addition of riboflavin has a deleterious effect on lifespan in aak-2 mutants. (c) Lifespan extension with rft-1 RNAi is preserved in animals with overexpression of aak-2 subunit aa 1-321 (aak-2oe) (d) Western blotting of phospho-AMPK Thr172 in lysates collected from young adult animals indicates activation following RNAi to rft-1, an effect abrogated by the addition of riboflavin. (e) Box plots of results from two-photon and fluorescence lifetime imaging, including organismal redox ratio, mitochondrial clustering, NAD(P)H bound fraction and FAD bound fraction for EV 290 and rft-1 RNAi treated animals. Riboflavin depletion decreases the redox ratio and increases intestinal mitochondrial clustering and the FAD bound fraction. (f) Volcano plot and heatmap of differentially abundant metabolites quantified by LC-MS reveals reductions in citric acid metabolites including citric acid, isocitric acid and α-ketoglutarate following riboflavin depletion. Purine metabolites including xanthine, hypoxanthine and guanosine are enriched with rft-1 RNAi. Representation of citric acid metabolites impacted by riboflavin depletion. See table S1 for tabular data and replicates of survival analyses. For (a-d) results are representative of two biological replicates. For (e) results are from 8 worms of two biological replicates. * indicates P < 0.01 by log-rank analysis (a-c) and by two tailed t-test (e). Box and whisker plots (e) indicate median and 5/95 th percentile respectively. See table S1 for tabular data and replicates of survival analyses. 291 FIGURE 4. Activation of the mitochondrial unfolded protein response (UPR mt ) is required for riboflavin depletion to promote longevity. (a) RNAi to rft-1 promotes activation of hsp- 6p::GFP progressively on adult days 1 and 3, an effect reversed by riboflavin supplementation (for binning, N = 120 worms per condition, representative of two biological replicates) Images above at 40X, binning performed at 10X magnification. See Figure S4 for binning examples. (b) Quantitative RT-PCR of established atfs-1 target EV RNAi with Vehicle rft-1 RNAi with Vehicle rft-1 RNAi with Riboflavin hsp-6p::GFP AD1 EV RNAi rft-1 RNAi EV RNAi + Riboflavin rft-1 RNAi + Riboflavin 0 5 10 15 hrg-9 Fold Change ✱ EV RNAi rft-1 RNAi EV RNAi + Riboflavin rft-1 RNAi + RIboflavin 0 5 10 15 20 25 cdr-2 Fold Change ✱ EV RNAi rft-1 RNAi EV RNAi + Riboflavin rft-1 RNAi + RIboflavin 0 500 1000 1500 2000 2500 Fold Change ✱ C07G1.7 EV RNAi rft-1RNAi 0 20 40 60 80 100 120 Young Adult Individuals None Low High EV RNAi rft-1RNAi 0 20 40 60 80 100 120 Adult Day 1 Individuals None Low High EV RNAi rft-1RNAi 0 20 40 60 80 100 120 Adult Day 3 Individuals None Low High 0 5 10 15 20 25 30 35 40 45 0.0 0.2 0.4 0.6 0.8 1.0 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) atfs-1 (EV RNAi) atfs-1 (rft-1 RNAi) vs vs vs * vs * ns * 0 5 10 15 20 25 30 35 40 45 0.0 0.2 0.4 0.6 0.8 1.0 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) atfs-1GOF (EV RNAi) atfs-1GOF (rft-1 RNAi) vs vs vs * vs * * * (a) (b) (c) (d) EV RNAi rft-1 RNAi 0 20 40 60 80 100 120 Adult Day 1 Individuals None Low High EV RNAi rft-1 RNAi 0 20 40 60 80 100 120 Adult Day 3 Individuals None Low High atfs-1, DAF-16::GFP Nuclear Localization (e) (f) hsp-6 sod-3 0 1 2 3 4 5 Fold Change rft-1 vs EV RNAi N2 rict-1 atfs-1 aak-2 ✱ ✱ ✱ ✱ hsp-6p::GFP Intensity 292 genes indicates marked upregulation with riboflavin depletion. (c) atfs-1 loss of function mutants do not experience lifespan extension with riboflavin depletion. (d) Gain of function mutants in atfs-1 are short lived but preserve responsiveness to rft-1 RNAi. (e) Atfs- 1;DAF::16GFP animals exhibit decreased but present nuclear localization on rft-1 RNAi at both adult day 1 and 3. N=120 animals. (f) QPCR of hsp-6 and sod-3 exhibit upregulation of transcripts in wild type animals treated with rft-1 RNAi and loss of upregulation in atfs-1 and aak-2 mutants. For (b,f), results are representative of at least three biological replicates. See table S1 for tabular data and replicates of survival analyses. * and ** indicate P < 0.01 by two-way ANOVA of ΔCt values (b,f), and by log- rank analysis (c and d). (e) representative of population of 60 worms from two biological replicates. Bars represent means ± SEM. 293 FIGURE 5. Riboflavin depletion alters lipid metabolism. (a) Fixative-based Nile red staining reveals marked upregulation of lipids in both soma and germline of aak-2, daf-16 and raga-1 animals (b) Quantification of images and fold change differences between EV and rft-1 RNAi reveal significant increases in fat mass with multiple mutants including a post developmental rft-1 RNAi in eri-1 enhanced RNAi mutants and with larval exposure to rft-1 RNAi in multiple mutants including daf-16, aak-2, nhr-49, atfs-1(GOF), and raga- 1. atfs-1 and rict-1 loss-of-function mutants do not exhibit increased fat on rft-1 RNAi. (c) Lipid analysis via GC/MS reveals an increase in overall fat stores (triglyceride/phospholipid ratio). (d) Shifts towards unsaturated fatty acid side chains and branched chain lipids in phospholipid and triglyceride fractions in aggregate (e) nhr-49 mutants exhibit intact lifespan extension with rft-1 RNAi. * indicates P < 0.05 by students 2-tailed t-test (c) and by two-way ANOVA (d) and by log rank analysis (e). Bars (c,d) Saturated Fatty Acids Unsaturated Fatty Acids Branched Chain Cyclopropyl 0 10 20 30 40 50 60 Triglycerides % Abundance EV RNAi rft-1 RNAi ✱ ✱ ✱ ✱ FC: 0.79 FC: 1.10 FC: 1.17 FC: 0.89 Saturated Fatty Acids Unsaturated Fatty Acids Branched Chain Cyclopropyl 0 20 40 60 80 Phospholipids % Abdundance EV RNAi rft-1 RNAi ✱ ✱ ✱ ✱ FC: 0.84 FC: 1.03 FC: 1.42 FC: 0.83 aak-2 (EV RNAi) aak-2 (rft-1 RNAi) daf-16 (EV RNAi) daf-16 (rft-1 RNAi) raga-1 (EV RNAi) raga-1 (rft-1 RNAi) EV RNAi rft-1 RNAi 0.0 0.5 1.0 1.5 2.0 2.5 Adult Day 1 Lipid Content Triglyceride/Phospholipid Arbitrary Units (A.U) ✱ FC: 1.42 (a) (b) (c) (d) 0 10 20 30 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) nhr-49 (EV RNAi) nhr-49 (rft-1 RNAi) vs vs vs vs * * * * (e) 0.6 0.8 1.0 1.2 1.4 1.6 1.8 Fold Change Intensity (rft-1 vs EV RNAi) wt eri-1 Post Dev daf-16 aak-2 atfs-1 atfs-1GOF raga-1 rict-1 nhr-49 294 represent means ± SEM. Dots in (b) represent individual biological replicates as fold change values over EV treated animals. 295 FIGURE 6. Riboflavin depletion mimics features of dietary restriction. (a) Imaging of reporters known to be activated with dietary restriction (bigr-1::RFP, right, and acs- 2p::GFP, left) indicates activation under rft-1 RNAi that progresses with age. (b) Lifespan extension with rft-1 RNAi occurs in eat-2 mutants. (c) Lifespan extension is abrogated in pha-4 loss-of function mutants versus temperature sensitive smg-1 mutant controls. (d) TORC1 mutant raga-1 exhibits lifespan extension with rft-1 RNAi. (e) Western blot of adult day 1 animals containing humanized S6K (permitting western blotting for TORC1- mediated phosphorylation of S6K T389 ) reveals no change in phospho-S6K concentrations between control and riboflavin depleted worms. (f) TORC2 mutant rict-1 does not exhibit lifespan extension with riboflavin depletion. (g) Model of riboflavin depletion impact on EV RNAi rft-1 RNAi 0 1 2 3 4 5 Arbitrary Units (A.U) Protein Concentration ns + + + + + + EV RNAi rft-1 RNAi p-S6K Actin (a) (b) 0 10 20 30 40 50 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive smg-1 (EV RNAi) smg-1 (rft-1 RNAi) smg-1; pha-4 (EV RNAi) smg-1; pha-4 (rft-1 RNAi) vs vs vs vs * * ns * EV RNAi rft-1 RNAi Young Adult Adult Day 1 Adult Day 3 0 50 100 150 200 250 300 bigr-1::RFP Signal Intensity (Arbitrary Units) EV RNAi rft-1 RNAi ✱ ✱ ✱ 0 5 10 15 20 25 30 35 40 45 50 55 0.0 0.2 0.4 0.6 0.8 1.0 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) raga-1(EV RNAi) raga-1(rft-1 RNAi) vs vs vs * vs * * ns 0 5 10 15 20 25 30 35 40 45 0.0 0.2 0.4 0.6 0.8 1.0 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) rict-1 (EV RNAi) rict-1 (rft-1 RNAi) vs vs vs * vs ns * * (c) (d) (e) (f) Young Adult Adult Day 1 Adult Day 3 0 10 20 30 40 50 60 70 80 acs-2p::GFP Signal Intensity (Arbitrary Units) EV RNAi rft-1 RNAi ✱ - - - - - - (g) 0 10 20 30 40 0.00 0.25 0.50 0.75 1.00 Day Fraction Alive wt (EV RNAi) wt (rft-1 RNAi) eat-2 (EV RNAi) eat-2 (rft-1 RNAi) vs vs vs vs * * * * 296 metabolism and longevity. Knockdown of the riboflavin transporter leads to depletion of flavin co-factors, influencing the energetic status of the animal and affecting global enzymatic activities dependent on FAD and FMN. This creates an integrated metabolic signaling response that promotes increased triglyceride stores and extends lifespan. See table S1 for tabular data and replicates of survival analyses. * indicates P < 0.05 by two- way ANOVA (a), students T-test (e) and log-rank test (b,c and d). Bars represent means ± SEM. 297 SUPPLEMENTAL FIGURES FIGURE S1. (a). Knockdown of rft-1 produces 70% reduction in transcript levels by quantitative RT-PCR and does not induce corresponding reductions in orthologous transporter rft-2 mRNA. N change in riboflavin kinase (rfk) and FAD synthetase (flad-1) levels are noted with rft-1 knockdown. (b) Differential interference contrast (DIC) and fluorescence imaging of DAPI stained EV and rft-1 RNAi treated adult day 1 animals reveals similar germ line stem cell morphology and the presence of oocytes. (c) RT-PCR of riboflavin pathway genes across the lifespan in germline deficient glp-4 animals reveals 298 an initial increase in riboflavin transporter expression at day 4, with subsequent decrease with aging. Rfk and flad-1 expression increases throughout the lifespan. (d) Post- developmental or larval only RNAi of rft-1 does not extend lifespan. (e) Pharyngeal pumping assay on adult day 1 animals reveals no difference between EV and rft-1 RNAi. (f) Crawling speed at L4, adult day 1 and adult day 3 revealed intact speed for rft-1 RNAi treated animals. (g) L4 and adult day 1 animals exhibit similar proportions of time off the bacterial lawn. (h) Body length and width is similar for EV and rft-1 RNAi animals across the lifespan. Bars represent means ± SEM. * represents p<0.05 for two-way ANOVA of ΔCt values (a,c), for log rank analysis (d), and for two way ANOVA with Sidak’s multiple comparison test (f-h). 299 FIGURE S2. Activation of DAF-16 by riboflavin deficiency. (a) Representative images of DAF-16::GFP high, low, and no (none) nuclear localization at 10X magnification (b) Transfer of rft-1 RNAi treated animals at adult day 1 to empty vector (EV) RNAi plates or riboflavin supplemented plates shows that nuclear localization of DAF-16 is reversed by removal from rft-1 RNAi and the addition of riboflavin. (c) Representative images of sod- 3p::GFP animals over the lifespan, with significant increases evident from adult day 3 onward with rft-1 RNAi. Persistent activation is seen in intestine, pharynx, and vulva (d) 300 Riboflavin depletion promotes lifespan extension in pdk-1 gain of function animals. * indicates P < 0.05 by log rank analysis (d). 301 FIGURE S3. Metabolite two-photon imaging and LC/MS quantification in riboflavin depletion. (a) Binning of images of aak-2;DAF-16::GFP animals treated with EV and rft-1 RNAi reveals persistent nuclear localization with riboflavin depletion both at adult day 1 and 3. (b) Representative images of total autofluorescence of NAD(P)H, FAD, and redox ratio map of C. elegans (1-4 respectively). Images of NAD(P)H, FAD, and redox map are shown only in masked region. (c) Representative phasors (top panels) and corresponding LLIF coded images (bottom panels) for NAD(P)H and FAD Lifetime Images. NAD(P)H images were acquired using 755 nm excitation/ 460 nm detection and FAD images were acquired using 860nm excitation/525 nm detection (d) Summary plot of quantitative enrichment analysis (QEA) of metabolites between EV treated and rft-1 RNAi treated 302 animals reveals enrichment of pathways involved in riboflavin, glutathione and purine metabolism (n=4, P<0.01). 303 FIGURE S4. (a) Representative images of hsp-6p::GFP expression categories in Figure 4a. 304 FIGURE S5. Lipid analyses with riboflavin depletion indicate increased fat mass. (a) Oil- red-O staining of adult day 1 animals treated with EV and rft-1. (b) Pseudo color map of (1, 4) total unsaturated fatty acid (area under curve between wavenumbers 2991 and 3022 cm -1 ), (2, 5) total fatty acid (area under curve between wavenumbers 2830 and 305 2870 cm -1 ) and (3, 6) ratio of TUFA and TFA for EV treated (4-6) and rft-1 RNAi treated (1-3) C. elegans acquired using SRS imaging. (c) GC/MS analysis of triglyceride and phospholipid fractions in adult day 1 and young adult worms. Ratios represent fold change of worms treated with rft-1 RNAi compared to EV treated animals (d) Imaging of acdh-1p::GFP animals reveals activation of acdh-1 expression with rft-1 RNAi. (e) Imaging of fat-7p::GFP animals reveals unchanged expression of fat-7 until adult day 5, where rft-1 RNAi preserves an aging-related decrease in desaturase expression. * indicates P <0.05 by two-tailed Student’s t-test (d) and by two-way ANOVA (c,e). Bars represent means ± SE. 306 ACKNOWLEDGEMENTS We thank Dr. Yuyao Zhang and Talia Hart for their creative input. This work was funded by the NIH/NIA Grants R01AG058259 and R01AG69677 and the Weissman Family MGH Research Scholar Award (to Alexander Soukas) and NIH/NIA grant R01AG058610 (to Sean Curran) and USC/Buck Nathan Shock Center grant P30AG068345. This work was also funded by the Foundation for Women’s Wellness, the Human Growth Foundation and NIH/NIDDK F32DK124948 (to Armen Yerevanian). Thanks to the Nutrition Obesity Research Center (NORC) at Harvard (P30DK040561) for core services. Some strains were provided by the CGC, which is funded by the NIH Office of Research Infrastructure Programs (P40OD010440). Figure 3f, 6g and graphical abstract were generated by BioRender.com. SRS and two photon images were acquired using microscopes acquired through NIH S10 OD021624 and NSF Major Research Instrumentation 1531683 grants. Conflict of Interest Statement: The authors report no competing interests. Author Contributions: Conceptualization: AY, LM, DB, AAS; Methodology: AY, LM, DB, SE; Validation: AY, LM, DB, SE; Investigation: AY, LM, SE, DB, SL, YZ, AA, LC, NS, AAS; Formal Analyses: AY, EG, KD, MJ, IG, AAS; Writing: AY, AAS; Review and Editing: AY, FA, SE, LM, YZ, SL, LC, AA, DB, NS, EG, MJ, IG, AAS; Funding: AY, IG, SPC, AAS. 307 References 1 Motohashi, H. & Yamamoto, M. Nrf2-Keap1 defines a physiologically important stress response mechanism. Trends in molecular medicine 10, 549-557, doi:10.1016/j.molmed.2004.09.003 (2004). 2 Sykiotis, G. P. & Bohmann, D. Stress-activated cap'n'collar transcription factors in aging and human disease. Sci Signal 3, re3, doi:10.1126/scisignal.3112re3 (2010). 3 Pomatto, L. C. D. et al. Limitations to adaptive homeostasis in an hyperoxia-induced model of accelerated ageing. Redox Biol 24, 101194, doi:10.1016/j.redox.2019.101194 (2019). 4 Tonelli, C., Chio, I. I. C. & Tuveson, D. A. Transcriptional Regulation by Nrf2. Antioxid Redox Signal 29, 1727-1745, doi:10.1089/ars.2017.7342 (2018). 5 Zhang, M. et al. Emerging roles of Nrf2 and phase II antioxidant enzymes in neuroprotection. Progress in neurobiology 100, 30-47, doi:10.1016/j.pneurobio.2012.09.003 (2013). 6 Joshi, G. & Johnson, J. A. The Nrf2-ARE pathway: a valuable therapeutic target for the treatment of neurodegenerative diseases. Recent Pat CNS Drug Discov 7, 218-229 (2012). 7 Bowerman, B., Draper, B. W., Mello, C. C. & Priess, J. R. The maternal gene skn-1 encodes a protein that is distributed unequally in early C. elegans embryos. Cell 74, 443- 452 (1993). 8 Bowerman, B., Eaton, B. A. & Priess, J. R. skn-1, a maternally expressed gene required to specify the fate of ventral blastomeres in the early C. elegans embryo. Cell 68, 1061- 1075 (1992). 9 Priess, J. R. & Hirsh, D. I. Caenorhabditis elegans morphogenesis: the role of the cytoskeleton in elongation of the embryo. Dev Biol 117, 156-173, doi:10.1016/0012- 1606(86)90358-1 (1986). 10 Maduro, M. F., Meneghini, M. D., Bowerman, B., Broitman-Maduro, G. & Rothman, J. H. Restriction of mesendoderm to a single blastomere by the combined action of SKN-1 and a GSK-3beta homolog is mediated by MED-1 and -2 in C. elegans. Molecular cell 7, 475-485 (2001). 11 Blackwell, T. K., Steinbaugh, M. J., Hourihan, J. M., Ewald, C. Y. & Isik, M. SKN-1/Nrf, stress responses, and aging in Caenorhabditis elegans. Free Radic Biol Med 88, 290- 301, doi:10.1016/j.freeradbiomed.2015.06.008 (2015). 12 Lehrbach, N. J. & Ruvkun, G. Endoplasmic reticulum-associated SKN-1A/Nrf1 mediates a cytoplasmic unfolded protein response and promotes longevity. Elife 8, doi:10.7554/eLife.44425 (2019). 308 13 Nhan, J. D. et al. Redirection of SKN-1 abates the negative metabolic outcomes of a perceived pathogen infection. Proc Natl Acad Sci U S A 116, 22322-22330, doi:10.1073/pnas.1909666116 (2019). 14 Li, L. et al. Glucose negatively affects Nrf2/SKN-1-mediated innate immunity in C. elegans. Aging 10, 3089-3103, doi:10.18632/aging.101610 (2018). 15 Lynn, D. A. et al. Omega-3 and -6 fatty acids allocate somatic and germline lipids to ensure fitness during nutrient and oxidative stress in Caenorhabditis elegans. Proc Natl Acad Sci U S A 112, 15378-15383, doi:10.1073/pnas.1514012112 (2015). 16 Steinbaugh, M. J. et al. Lipid-mediated regulation of SKN-1/Nrf in response to germ cell absence. Elife 4, doi:10.7554/eLife.07836 (2015). 17 Pang, S., Lynn, D. A., Lo, J. Y., Paek, J. & Curran, S. P. SKN-1 and Nrf2 couples proline catabolism with lipid metabolism during nutrient deprivation. Nat Commun 5, 5048, doi:10.1038/ncomms6048 (2014). 18 Paek, J. et al. Mitochondrial SKN-1/Nrf mediates a conserved starvation response. Cell Metab 16, 526-537, doi:10.1016/j.cmet.2012.09.007 (2012). 19 Pang, S. & Curran, S. P. Adaptive Capacity to Bacterial Diet Modulates Aging in C. elegans. Cell Metab 19, 221-231, doi:10.1016/j.cmet.2013.12.005 (2014). 20 Spatola, B. N., Lo, J. Y., Wang, B. & Curran, S. P. Nuclear and cytoplasmic WDR-23 isoforms mediate differential effects on GEN-1 and SKN-1 substrates. Sci Rep 9, 11783, doi:10.1038/s41598-019-48286-y (2019). 21 Maduro, M. F., Broitman-Maduro, G., Mengarelli, I. & Rothman, J. H. Maternal deployment of the embryonic SKN-1-->MED-1,2 cell specification pathway in C. elegans. Developmental biology 301, 590-601, doi:10.1016/j.ydbio.2006.08.029 (2007). 22 Curran, S. P. & Ruvkun, G. Lifespan regulation by evolutionarily conserved genes essential for viability. PLoS Genet 3, e56, doi:10.1371/journal.pgen.0030056 (2007). 23 Lynn, D. A. & Curran, S. P . The SKN-1 hunger games: May the odds be ever in your favor. Worm 4, e1078959, doi:10.1080/21624054.2015.1078959 (2015). 24 Bishop, N. A. & Guarente, L. Two neurons mediate diet-restriction-induced longevity in C. elegans. Nature 447, 545-549, doi:10.1038/nature05904 (2007). 25 Nhan, J. D. & Curran, S. P. Metabolic Assessment of Lipid Abundance and Distribution. Methods Mol Biol 2144, 103-110, doi:10.1007/978-1-0716-0592-9_9 (2020). 26 Escorcia, W., Ruter, D. L., Nhan, J. & Curran, S. P. Quantification of Lipid Abundance and Evaluation of Lipid Distribution in Caenorhabditis elegans by Nile Red and Oil Red O Staining. J Vis Exp, doi:10.3791/57352 (2018). 27 Tacutu, R. et al. Prediction of C. elegans longevity genes by human and worm longevity networks. PloS one 7, e48282, doi:10.1371/journal.pone.0048282 (2012). 309 28 Xu, Z. et al. WDR-23 and SKN-1/Nrf2 Coordinate with the BLI-3 Dual Oxidase in Response to Iodide-Triggered Oxidative Stress. G3 (Bethesda) 8, 3515-3527, doi:10.1534/g3.118.200586 (2018). 29 Tang, L. & Choe, K. P . Characterization of skn-1/wdr-23 phenotypes in Caenorhabditis elegans; pleiotrophy, aging, glutathione, and interactions with other longevity pathways. Mech Ageing Dev 149, 88-98, doi:10.1016/j.mad.2015.06.001 (2015). 30 Staab, T. A., Evgrafov, O., Knowles, J. A. & Sieburth, D. Regulation of synaptic nlg- 1/neuroligin abundance by the skn-1/Nrf stress response pathway protects against oxidative stress. PLoS Genet 10, e1004100, doi:10.1371/journal.pgen.1004100 (2014). 31 Staab, T. A. et al. The conserved SKN-1/Nrf2 stress response pathway regulates synaptic function in Caenorhabditis elegans. PLoS Genet 9, e1003354, doi:10.1371/journal.pgen.1003354 (2013). 32 Hasegawa, K. & Miwa, J. Genetic and cellular characterization of Caenorhabditis elegans mutants abnormal in the regulation of many phase II enzymes. PloS one 5, e11194, doi:10.1371/journal.pone.0011194 [doi] (2010). 33 Choe, K. P., Przybysz, A. J. & Strange, K. The WD40 repeat protein WDR-23 functions with the CUL4/DDB1 ubiquitin ligase to regulate nuclear abundance and activity of SKN- 1 in Caenorhabditis elegans. Molecular and cellular biology 29, 2704-2715, doi:10.1128/MCB.01811-08 (2009). 35 Satoh, T., McKercher, S. R. & Lipton, S. A. Nrf2/ARE-mediated antioxidant actions of pro-electrophilic drugs. Free Radic Biol Med 65, 645-657, doi:10.1016/j.freeradbiomed.2013.07.022 (2013). 36 Abiko, Y., Miura, T., Phuc, B. H., Shinkai, Y. & Kumagai, Y. Participation of covalent modification of Keap1 in the activation of Nrf2 by tert-butylbenzoquinone, an electrophilic metabolite of butylated hydroxyanisole. Toxicology and applied pharmacology 255, 32- 39, doi:10.1016/j.taap.2011.05.013 (2011). 37 Hong, F., Sekhar, K. R., Freeman, M. L. & Liebler, D. C. Specific patterns of electrophile adduction trigger Keap1 ubiquitination and Nrf2 activation. The Journal of biological chemistry 280, 31768-31775, doi:10.1074/jbc.M503346200 (2005). 38 Yen, C. A. & Curran, S. P . Incomplete proline catabolism drives premature sperm aging. Aging Cell 20, e13308, doi:10.1111/acel.13308 (2021). 39 Yen, C. A., Ruter, D. L., Turner, C. D., Pang, S. & Curran, S. P . Loss of flavin adenine dinucleotide (FAD) impairs sperm function and male reproductive advantage in C. elegans. Elife 9, doi:10.7554/eLife.52899 (2020). 40 Haghani, A. et al. Air Pollution Alters Caenorhabditis elegans Development and Lifespan: Responses to Traffic-Related Nanoparticulate Matter. J Gerontol A Biol Sci Med Sci 74, 1189-1197, doi:10.1093/gerona/glz063 (2019). 41 Lo, J. Y., Spatola, B. N. & Curran, S. P. WDR23 regulates NRF2 independently of KEAP1. PLoS Genet 13, e1006762, doi:10.1371/journal.pgen.1006762 (2017). 310 42 Naji, A. et al. The activation of the oxidative stress response transcription factor SKN-1 in Caenorhabditis elegans by mitis group streptococci. PLOS ONE 13, e0202233, doi:10.1371/journal.pone.0202233 (2018). 43 Dehghan, E. et al. Hydralazine induces stress resistance and extends C. elegans lifespan by activating the NRF2/SKN-1 signalling pathway. Nature Communications 8, 2223, doi:10.1038/s41467-017-02394-3 (2017). 44 Li, H. et al. O-GlcNAcylation of SKN-1 modulates the lifespan and oxidative stress resistance in Caenorhabditis elegans. Sci Rep 7, 43601, doi:10.1038/srep43601 (2017). 45 Glover-Cutter, K. M., Lin, S. & Blackwell, T. K. Integration of the unfolded protein and oxidative stress responses through SKN-1/Nrf. PLoS Genet 9, e1003701, doi:10.1371/journal.pgen.1003701 (2013). 46 Li, X. et al. Specific SKN-1/Nrf stress responses to perturbations in translation elongation and proteasome activity. PLoS genetics 7, e1002119, doi:10.1371/journal.pgen.1002119 (2011). 47 Wu, C.-W., Deonarine, A., Przybysz, A., Strange, K. & Choe, K. P . The Skp1 Homologs SKR-1/2 Are Required for the Caenorhabditis elegans SKN-1 Antioxidant/Detoxification Response Independently of p38 MAPK. PLOS Genetics 12, e1006361, doi:10.1371/journal.pgen.1006361 (2016). 48 Hipp, M. S., Kasturi, P. & Hartl, F. U. The proteostasis network and its decline in ageing. Nat Rev Mol Cell Biol 20, 421-435, doi:10.1038/s41580-019-0101-y (2019). 49 Enchev, R. I., Schulman, B. A. & Peter, M. Protein neddylation: beyond cullin-RING ligases. Nat Rev Mol Cell Biol 16, 30-44, doi:10.1038/nrm3919 (2015). 50 Vertegaal, A. C. O. Signalling mechanisms and cellular functions of SUMO. Nat Rev Mol Cell Biol 23, 715-731, doi:10.1038/s41580-022-00500-y (2022). 51 Zhang, L., Ward, J. D., Cheng, Z. & Dernburg, A. F. The auxin-inducible degradation (AID) system enables versatile conditional protein depletion in C. elegans. Development 142, 4374-4384, doi:10.1242/dev.129635 (2015). 52 Ewe, C. K., Alok, G. & Rothman, J. H. Stressful development: integrating endoderm development, stress, and longevity. Dev Biol 471, 34-48, doi:10.1016/j.ydbio.2020.12.002 (2021). 53 An, J. H. & Blackwell, T. K. SKN-1 links C. elegans mesendodermal specification to a conserved oxidative stress response. Genes & development 17, 1882-1893, doi:10.1101/gad.1107803 (2003). 54 Tullet, J. M. et al. Direct inhibition of the longevity-promoting factor SKN-1 by insulin-like signaling in C. elegans. Cell 132, 1025-1038, doi:10.1016/j.cell.2008.01.030 (2008). 55 Ashley, G. E. et al. An expanded auxin-inducible degron toolkit for Caenorhabditis elegans. Genetics 217, doi:10.1093/genetics/iyab006 (2021). 56 Hwang, H. Y. & Wang, J. Fast genetic mapping using insertion-deletion polymorphisms in Caenorhabditis elegans. Sci Rep 11, 11017, doi:10.1038/s41598-021-90190-x (2021). 311 57 Wang, W. & Pyle, A. M. The RIG-I receptor adopts two different conformations for distinguishing host from viral RNA ligands. Mol Cell 82, 4131-4144.e4136, doi:10.1016/j.molcel.2022.09.029 (2022). 58 Kowalinski, E. et al. Structural basis for the activation of innate immune pattern- recognition receptor RIG-I by viral RNA. Cell 147, 423-435, doi:10.1016/j.cell.2011.09.039 (2011). 59 Lindblom, T. H. & Dodd, A. K. Xenobiotic detoxification in the nematode Caenorhabditis elegans. J Exp Zool A Comp Exp Biol 305, 720-730, doi:10.1002/jez.a.324 (2006). 60 Wang, J. et al. RNAi screening implicates a SKN-1-dependent transcriptional response in stress resistance and longevity deriving from translation inhibition. PLoS genetics 6, doi:10.1371/journal.pgen.1001048 (2010). 61 Oliveira, R. P. et al. Condition-adapted stress and longevity gene regulation by Caenorhabditis elegans SKN-1/Nrf. Aging cell 8, 524-541, doi:10.1111/j.1474- 9726.2009.00501.x (2009). 62 Wang, J. & Barr, M. M. RNA interference in Caenorhabditis elegans. Methods Enzymol 392, 36-55, doi:10.1016/S0076-6879(04)92003-4 (2005). 63 Guo, X., Zhang, R., Wang, J., Ding, S. W. & Lu, R. Homologous RIG-I-like helicase proteins direct RNAi-mediated antiviral immunity in C. elegans by distinct mechanisms. Proc Natl Acad Sci U S A 110, 16085-16090, doi:10.1073/pnas.1307453110 (2013). 64 Gammon, D. B. et al. The Antiviral RNA Interference Response Provides Resistance to Lethal Arbovirus Infection and Vertical Transmission in Caenorhabditis elegans. Curr Biol 27, 795-806, doi:10.1016/j.cub.2017.02.004 (2017). 65 Sowa, J. N. et al. The Caenorhabditis elegans RIG-I Homolog DRH-1 Mediates the Intracellular Pathogen Response upon Viral Infection. J Virol 94, doi:10.1128/JVI.01173- 19 (2020). 66 Wang, R. et al. Hypermethylation of the Keap1 gene in human lung cancer cell lines and lung cancer tissues. Biochemical and biophysical research communications 373, 151- 154, doi:10.1016/j.bbrc.2008.06.004 (2008). 67 Wang, X. J. et al. Activation of Nrf2 by arsenite and monomethylarsonous acid is independent of Keap1-C151: enhanced Keap1-Cul3 interaction. Toxicology and applied pharmacology 230, 383-389, doi:10.1016/j.taap.2008.03.003 (2008). 68 Wang, X. J. et al. Nrf2 enhances resistance of cancer cells to chemotherapeutic drugs, the dark side of Nrf2. Carcinogenesis 29, 1235-1243, doi:10.1093/carcin/bgn095 (2008). 69 Maghames, C. M. et al. NEDDylation promotes nuclear protein aggregation and protects the Ubiquitin Proteasome System upon proteotoxic stress. Nat Commun 9, 4376, doi:10.1038/s41467-018-06365-0 (2018). 70 Loftus, S. J., Liu, G., Carr, S. M., Munro, S. & La Thangue, N. B. NEDDylation regulates E2F-1-dependent transcription. EMBO Rep 13, 811-818, doi:10.1038/embor.2012.113 (2012). 312 71 Nguyen, H. C., Wang, W. & Xiong, Y. Cullin-RING E3 Ubiquitin Ligases: Bridges to Destruction. Subcell Biochem 83, 323-347, doi:10.1007/978-3-319-46503-6_12 (2017). 72 Sala, A. J. & Morimoto, R. I. Protecting the future: balancing proteostasis for reproduction. Trends Cell Biol 32, 202-215, doi:10.1016/j.tcb.2021.09.009 (2022). 73 Villa, O. et al. Genetic variation in. Elife 11, doi:10.7554/eLife.74308 (2022). 74 Dalton, H. M. & Curran, S. P. Hypodermal responses to protein synthesis inhibition induce systemic developmental arrest and AMPK-dependent survival in Caenorhabditis elegans. PLoS Genet 14, e1007520, doi:10.1371/journal.pgen.1007520 (2018). 75 Altun, Z. F., Herndon, L.A., Wolkow, C.A., Crocker, C., Lints, R. and Hall, D.H. (ed.s). WormAtlas. http://www.wormatlas.org (2002-2021). 76 Stuhr, N. L. et al. Rapid Lipid Quantification in. Bio Protoc 12, e4340, doi:10.21769/BioProtoc.4340 (2022). 77 Stuhr, N. L. & Curran, S. P. Bacterial diets differentially alter lifespan and healthspan trajectories in C. elegans. Commun Biol 3, 653, doi:10.1038/s42003-020-01379-1 (2020). 78 Holdorf, A. D. et al. WormCat: An Online Tool for Annotation and Visualization of. Genetics 214, 279-294, doi:10.1534/genetics.119.302919 (2020). 79 Taylor, S. R. et al. Molecular topography of an entire nervous system. Cell 184, 4329- 4347.e4323, doi:10.1016/j.cell.2021.06.023 (2021). 80 Kaletsky, R. et al. The C. elegans adult neuronal IIS/FOXO transcriptome reveals adult phenotype regulators. Nature 529, 92-96, doi:10.1038/nature16483 (2016). 81 Ittisoponpisan, S. et al. Can Predicted Protein 3D Structures Provide Reliable Insights into whether Missense Variants Are Disease Associated? J Mol Biol 431, 2197-2212, doi:10.1016/j.jmb.2019.04.009 (2019). 82 Santilli, V., Bernetti, A., Mangone, M. & Paoloni, M. Clinical definition of sarcopenia. Clin Cases Miner Bone Metab 11, 177-180 (2014). 83 von Haehling, S., Morley, J. E. & Anker, S. D. An overview of sarcopenia: facts and numbers on prevalence and clinical impact. J Cachexia Sarcopenia Muscle 1, 129-133, doi:10.1007/s13539-010-0014-2 (2010). 84 Tsekoura, M., Kastrinis, A., Katsoulaki, M., Billis, E. & Gliatis, J. Sarcopenia and Its Impact on Quality of Life. Adv Exp Med Biol 987, 213-218, doi:10.1007/978-3-319- 57379-3_19 (2017). 85 Ahmadpanah, M. et al. No evidence for metabolic syndrome and lipid profile differences in patients suffering from bipolar I disorder with and without suicide attempts. Int J Psychiatry Clin Pract 19, 168-173, doi:10.3109/13651501.2015.1049277 (2015). 313 86 Dent, E. et al. Physical Frailty: ICFSR International Clinical Practice Guidelines for Identification and Management. J Nutr Health Aging 23, 771-787, doi:10.1007/s12603- 019-1273-z (2019). 87 Batsis, J. A. et al. Variation in the prevalence of sarcopenia and sarcopenic obesity in older adults associated with different research definitions: dual-energy X-ray absorptiometry data from the National Health and Nutrition Examination Survey 1999- 2004. J Am Geriatr Soc 61, 974-980, doi:10.1111/jgs.12260 (2013). 88 Cruz-Jentoft, A. J. et al. Sarcopenia: European consensus on definition and diagnosis: Report of the European Working Group on Sarcopenia in Older People. Age Ageing 39, 412-423, doi:10.1093/ageing/afq034 (2010). 89 Sayer, A. A. et al. The developmental origins of sarcopenia. J Nutr Health Aging 12, 427- 432, doi:10.1007/BF02982703 (2008). 90 Fisher, G. G. & Ryan, L. H. Overview of the Health and Retirement Study and Introduction to the Special Issue. Work Aging Retire 4, 1-9, doi:10.1093/workar/wax032 (2018). 91 Sonnega, A. et al. Cohort Profile: the Health and Retirement Study (HRS). Int J Epidemiol 43, 576-585, doi:10.1093/ije/dyu067 (2014). 92 Juster, F. T. & Suzman, R. M. An Overview of the Health and Retirement Study. Journal of Human Resources 30, S7-56 (1995). 93 Health and Retirement Study (HRS), <https://www.ncbi.nlm.nih.gov/projects/gap/cgi- bin/study.cgi?study_id=phs000428.v2.p2> ( 94 Janssen, I., Baumgartner, R. N., Ross, R., Rosenberg, I. H. & Roubenoff, R. Skeletal muscle cutpoints associated with elevated physical disability risk in older men and women. Am J Epidemiol 159, 413-421, doi:10.1093/aje/kwh058 (2004). 95 MacEwan, J. P. et al. Measuring Sarcopenia Severity in Older Adults and the Value of Effective Interventions. J Nutr Health Aging 22, 1253-1258, doi:10.1007/s12603-018- 1104-7 (2018). 96 Tang, H. & Pang, S. Proline Catabolism Modulates Innate Immunity in Caenorhabditis elegans. Cell Rep 17, 2837-2844, doi:10.1016/j.celrep.2016.11.038 (2016). 97 Timmers, P . R. et al. Genomics of 1 million parent lifespans implicates novel pathways and common diseases and distinguishes survival chances. Elife 8, doi:10.7554/eLife.39856 (2019). 98 Tam, V . et al. Benefits and limitations of genome-wide association studies. Nat Rev Genet 20, 467-484, doi:10.1038/s41576-019-0127-1 (2019). 99 Manolio, T. A. et al. Finding the missing heritability of complex diseases. Nature 461, 747-753, doi:10.1038/nature08494 (2009). 314 100 Ke, W. et al. Genes in human obesity loci are causal obesity genes in C. elegans. PLoS Genet 17, e1009736, doi:10.1371/journal.pgen.1009736 (2021). 101 Song, L. et al. Lymphangiogenic therapy prevents cardiac dysfunction by ameliorating inflammation and hypertension. Elife 9, doi:10.7554/eLife.58376 (2020). 102 Teumer, A. et al. Genome-wide association meta-analyses and fine-mapping elucidate pathways influencing albuminuria. Nat Commun 10, 4130, doi:10.1038/s41467-019- 11576-0 (2019). 103 Liu, Z. et al. Associations of genetics, behaviors, and life course circumstances with a novel aging and healthspan measure: Evidence from the Health and Retirement Study. PLoS Med 16, e1002827, doi:10.1371/journal.pmed.1002827 (2019). 104 Kelley, L. A., Mezulis, S., Yates, C. M., Wass, M. N. & Sternberg, M. J. The Phyre2 web portal for protein modeling, prediction and analysis. Nat Protoc 10, 845-858, doi:10.1038/nprot.2015.053 (2015). 105 Wang, Y . & Hekimi, S. Mitochondrial dysfunction and longevity in animals: Untangling the knot. Science 350, 1204-1207, doi:10.1126/science.aac4357 (2015). 106 Shields, H. J., Traa, A. & Van Raamsdonk, J. M. Beneficial and Detrimental Effects of Reactive Oxygen Species on Lifespan: A Comprehensive Review of Comparative and Experimental Studies. Front Cell Dev Biol 9, 628157, doi:10.3389/fcell.2021.628157 (2021). 107 Weir, D. R., University of Michigan. Quality Control Report for Genotypic Data. (2013). 108 Carson, R. G. Get a grip: individual variations in grip strength are a marker of brain health. Neurobiol Aging 71, 189-222, doi:10.1016/j.neurobiolaging.2018.07.023 (2018). 109 McGrath, R. et al. What are the association patterns between handgrip strength and adverse health conditions? A topical review. SAGE Open Med 8, 2050312120910358, doi:10.1177/2050312120910358 (2020). 110 Rentería, M. E., Cortes, A. & Medland, S. E. Using PLINK for Genome-Wide Association Studies (GWAS) and data analysis. Methods Mol Biol 1019, 193-213, doi:10.1007/978-1- 62703-447-0_8 (2013). 111 Purcell, S. et al. PLINK: a tool set for whole-genome association and population-based linkage analyses. Am J Hum Genet 81, 559-575, doi:10.1086/519795 (2007). 112 Roussel, N., Sprenger, J., Tappan, S. J. & Glaser, J. R. Robust tracking and quantification of C. elegans body shape and locomotion through coiling, entanglement, and omega bends. Worm 3, e982437, doi:10.4161/21624054.2014.982437 (2014). 113 Laranjeiro, R. et al. Swim exercise in Caenorhabditis elegans extends neuromuscular and gut healthspan, enhances learning ability, and protects against neurodegeneration. Proc Natl Acad Sci U S A 116, 23829-23839, doi:10.1073/pnas.1909210116 (2019). 315 114 Laranjeiro, R., Harinath, G., Burke, D., Braeckman, B. P. & Driscoll, M. Single swim sessions in C. elegans induce key features of mammalian exercise. BMC Biol 15, 30, doi:10.1186/s12915-017-0368-4 (2017). 115 Yen, C. A. & Curran, S. P . Gene-diet interactions and aging in C. elegans. Exp Gerontol 86, 106-112, doi:10.1016/j.exger.2016.02.012 (2016). 116 Battle, A. et al. Genetic effects on gene expression across human tissues. Nature 550, 204-213, doi:10.1038/nature24277 (2017). 117 Dey, R., Schmidt, E. M., Abecasis, G. R. & Lee, S. A Fast and Accurate Algorithm to Test for Binary Phenotypes and Its Application to PheWAS. Am J Hum Genet 101, 37-49, doi:10.1016/j.ajhg.2017.05.014 (2017). 118 Pendergrass, S. A. et al. Phenome-Wide Association Studies: Embracing Complexity for Discovery. Hum Hered 79, 111-123, doi:10.1159/000381851 (2015). 119 Hall, M. A. et al. Detection of pleiotropy through a Phenome-wide association study (PheWAS) of epidemiologic data as part of the Environmental Architecture for Genes Linked to Environment (EAGLE) study. PLoS Genet 10, e1004678, doi:10.1371/journal.pgen.1004678 (2014). 120 Pendergrass, S. A. et al. Phenome-wide association study (PheWAS) for detection of pleiotropy within the Population Architecture using Genomics and Epidemiology (PAGE) Network. PLoS Genet 9, e1003087, doi:10.1371/journal.pgen.1003087 (2013). 121 Brenner, S. The genetics of Caenorhabditis elegans. Genetics 77, 71-94, doi:10.1093/genetics/77.1.71 (1974). 122 Laurie, C. C. et al. Quality control and quality assurance in genotypic data for genome- wide association studies. Genet Epidemiol 34, 591-602, doi:10.1002/gepi.20516 (2010). 123 Wu, C., Smit, E., Peralta, C. A., Sarathy, H. & Odden, M. C. Functional Status Modifies the Association of Blood Pressure with Death in Elders: Health and Retirement Study. J Am Geriatr Soc 65, 1482-1489, doi:10.1111/jgs.14816 (2017). 124 Batsis, J. A., Germain, C. M., Vasquez, E. & Bartels, S. J. Prevalence of weakness and its relationship with limitations based on the Foundations for the National Institutes for Health project: data from the Health and Retirement Study. Eur J Clin Nutr 70, 1168- 1173, doi:10.1038/ejcn.2016.90 (2016). 125 Kim, D. H. et al. Validation of a Claims-Based Frailty Index Against Physical Performance and Adverse Health Outcomes in the Health and Retirement Study. J Gerontol A Biol Sci Med Sci 74, 1271-1276, doi:10.1093/gerona/gly197 (2019). 126 Novembre, J. et al. Genes mirror geography within Europe. Nature 456, 98-101, doi:10.1038/nature07331 (2008). 127 Tian, C., Gregersen, P . K. & Seldin, M. F. Accounting for ancestry: population substructure and genome-wide association studies. Hum Mol Genet 17, R143-150, doi:10.1093/hmg/ddn268 (2008). 316 128 Price, A. L. et al. Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet 38, 904-909, doi:10.1038/ng1847 (2006). 129 Li, Q. & Yu, K. Improved correction for population stratification in genome-wide association studies by identifying hidden population structures. Genet Epidemiol 32, 215-226, doi:10.1002/gepi.20296 (2008). 130 Serre, D. et al. Correction of population stratification in large multi-ethnic association studies. PLoS One 3, e1382, doi:10.1371/journal.pone.0001382 (2008). 131 Zhang, S., Zhu, X. & Zhao, H. On a semiparametric test to detect associations between quantitative traits and candidate genes using unrelated individuals. Genet Epidemiol 24, 44-56, doi:10.1002/gepi.10196 (2003). 132 Zhu, X., Zhang, S., Zhao, H. & Cooper, R. S. Association mapping, using a mixture model for complex traits. Genet Epidemiol 23, 181-196, doi:10.1002/gepi.210 (2002). 133 Price, A. L., Zaitlen, N. A., Reich, D. & Patterson, N. New approaches to population stratification in genome-wide association studies. Nat Rev Genet 11, 459-463, doi:10.1038/nrg2813 (2010). 134 Han, B., Kang, H. M. & Eskin, E. Rapid and accurate multiple testing correction and power estimation for millions of correlated markers. PLoS Genet 5, e1000456, doi:10.1371/journal.pgen.1000456 (2009). 135 Sham, P. C. & Purcell, S. M. Statistical power and significance testing in large-scale genetic studies. Nat Rev Genet 15, 335-346, doi:10.1038/nrg3706 (2014). 136 Dudbridge, F. & Gusnanto, A. Estimation of significance thresholds for genomewide association scans. Genet Epidemiol 32, 227-234, doi:10.1002/gepi.20297 (2008). 137 Pahl, R. & Schäfer, H. PERMORY: an LD-exploiting permutation test algorithm for powerful genome-wide association testing. Bioinformatics 26, 2093-2100, doi:10.1093/bioinformatics/btq399 (2010). 138 North, B. V., Curtis, D. & Sham, P. C. A note on the calculation of empirical P values from Monte Carlo procedures. Am J Hum Genet 71, 439-441, doi:10.1086/341527 (2002). 139 Johnson, A. A. & Stolzing, A. The role of lipid metabolism in aging, lifespan regulation, and age-related disease. Aging Cell, e13048, doi:10.1111/acel.13048 (2019). 140 Hammerquist, A. M., Escorcia, W. & Curran, S. P. Maf1 regulates intracellular lipid homeostasis in response to DNA damage response activation. Mol Biol Cell 32, 1086- 1093, doi:10.1091/mbc.E20-06-0378 (2021). 141 Zhang, Y. et al. Comparative genomics and functional study of lipid metabolic genes in Caenorhabditis elegans. BMC Genomics 14, 164, doi:10.1186/1471-2164-14-164 (2013). 142 Castro, C. et al. A metabolomic strategy defines the regulation of lipid content and global metabolism by Delta9 desaturases in Caenorhabditis elegans. BMC Genomics 13, 36, doi:10.1186/1471-2164-13-36 (2012). 317 143 Pino, E. C., Webster, C. M., Carr, C. E. & Soukas, A. A. Biochemical and high throughput microscopic assessment of fat mass in Caenorhabditis elegans. J Vis Exp, doi:10.3791/50180 (2013). 144 Webster, C. M. et al. Genome-wide RNAi Screen for Fat Regulatory Genes in C. elegans Identifies a Proteostasis-AMPK Axis Critical for Starvation Survival. Cell Rep 20, 627- 640, doi:10.1016/j.celrep.2017.06.068 (2017). 145 Greenspan, P., Mayer, E. P. & Fowler, S. D. Nile red: a selective fluorescent stain for intracellular lipid droplets. J Cell Biol 100, 965-973, doi:10.1083/jcb.100.3.965 (1985). 146 O'Rourke, E. J., Soukas, A. A., Carr, C. E. & Ruvkun, G. C. elegans major fats are stored in vesicles distinct from lysosome-related organelles. Cell Metab 10, 430-435, doi:10.1016/j.cmet.2009.10.002 (2009). 147 Garigan, D. et al. Genetic analysis of tissue aging in Caenorhabditis elegans: a role for heat-shock factor and bacterial proliferation. Genetics 161, 1101-1112, doi:10.1093/genetics/161.3.1101 (2002). 148 Qi, B., Kniazeva, M. & Han, M. A vitamin-B2-sensing mechanism that regulates gut protease activity to impact animal's food behavior and growth. Elife 6, doi:10.7554/eLife.26243 (2017). 149 Gems, D. & Riddle, D. L. Genetic, behavioral and environmental determinants of male longevity in Caenorhabditis elegans. Genetics 154, 1597-1610, doi:10.1093/genetics/154.4.1597 (2000). 150 Cabreiro, F. et al. Metformin retards aging in C. elegans by altering microbial folate and methionine metabolism. Cell 153, 228-239, doi:10.1016/j.cell.2013.02.035 (2013). 151 Sanchez-Blanco, A. et al. Dietary and microbiome factors determine longevity in Caenorhabditis elegans. Aging (Albany NY) 8, 1513-1539, doi:10.18632/aging.101008 (2016). 152 Beydoun, S. et al. An alternative food source for metabolism and longevity studies in Caenorhabditis elegans. Commun Biol 4, 258, doi:10.1038/s42003-021-01764-4 (2021). 153 Scott, T. A. et al. Host-Microbe Co-metabolism Dictates Cancer Drug Efficacy in C. elegans. Cell 169, 442-456 e418, doi:10.1016/j.cell.2017.03.040 (2017). 154 Stuhr, N. L. et al. Rapid Lipid Quantification in Caenorhabditis elegans by Oil Red O and Nile Red Staining. Bio Protoc 12, e4340, doi:10.21769/BioProtoc.4340 (2022). 155 Inzucchi, S. E. et al. Management of hyperglycaemia in type 2 diabetes: a patient- centered approach. Position statement of the American Diabetes Association (ADA) and the European Association for the Study of Diabetes (EASD). Diabetologia 55, 1577-1596 (2012). 156 Evans, J. M., Donnelly, L. A., Emslie-Smith, A. M., Alessi, D. R. & Morris, A. D. Metformin and reduced risk of cancer in diabetic patients. Bmj 330, 1304-1305 (2005). 318 157 Yuan, P . et al. Phenformin enhances the therapeutic benefit of BRAFV600E inhibition in melanoma. Proceedings of the National Academy of Sciences 110, 18226-18231 (2013). 158 Cabreiro, F. et al. Metformin retards aging in C. elegans by altering microbial folate and methionine metabolism. Cell 153, 228-239 (2013). 159 Martin-Montalvo, A. et al. Metformin improves healthspan and lifespan in mice. Nature communications 4, 1-9 (2013). 160 Onken, B. & Driscoll, M. Metformin induces a dietary restriction–like state and the oxidative stress response to extend C. elegans healthspan via AMPK, LKB1, and SKN-1. PloS one 5, e8758 (2010). 161 Barzilai, N., Crandall, J. P., Kritchevsky, S. B. & Espeland, M. A. Metformin as a tool to target aging. Cell metabolism 23, 1060-1065 (2016). 162 Wu, L. et al. An ancient, unified mechanism for metformin growth inhibition in C. elegans and cancer. Cell 167, 1705-1718. e1713 (2016). 163 Chen, J. et al. Metformin extends C. elegans lifespan through lysosomal pathway. Elife 6, e31268 (2017). 164 Glaser, P. E. & Gross, R. W. Plasmenylethanolamine facilitates rapid membrane fusion: a stopped-flow kinetic investigation correlating the propensity of a major plasma membrane constituent to adopt an HII phase with its ability to promote membrane fusion. Biochemistry 33, 5805-5812 (1994). 165 Komljenovic, D. et al. Disruption of blood-testis barrier dynamics in ether-lipid-deficient mice. Cell and tissue research 337, 281-299 (2009). 166 Marrink, S.-J. & Mark, A. E. Molecular view of hexagonal phase formation in phospholipid membranes. Biophysical journal 87, 3894-3900 (2004). 167 Davies, S. S. et al. Oxidized alkyl phospholipids are specific, high affinity peroxisome proliferator-activated receptor γ ligands and agonists. Journal of Biological Chemistry 276, 16015-16023 (2001). 168 Facciotti, F. et al. Peroxisome-derived lipids are self antigens that stimulate invariant natural killer T cells in the thymus. Nature immunology 13, 474-480 (2012). 169 Rodemer, C. et al. Inactivation of ether lipid biosynthesis causes male infertility, defects in eye development and optic nerve hypoplasia in mice. Hum Mol Genet 12, 1881-1895, doi:10.1093/hmg/ddg191 (2003). 170 Teigler, A., Komljenovic, D., Draguhn, A., Gorgas, K. & Just, W. W. Defects in myelination, paranode organization and Purkinje cell innervation in the ether lipid- deficient mouse cerebellum. Hum Mol Genet 18, 1897-1908, doi:10.1093/hmg/ddp110 (2009). 319 171 Thukkani, A. K. et al. Reactive chlorinating species produced during neutrophil activation target tissue plasmalogens: production of the chemoattractant, 2-chlorohexadecanal. J Biol Chem 277, 3842-3849, doi:10.1074/jbc.M109489200 (2002). 172 Albert, C. J. et al. Eosinophil peroxidase-derived reactive brominating species target the vinyl ether bond of plasmalogens generating a novel chemoattractant, alpha-bromo fatty aldehyde. J Biol Chem 278, 8942-8950, doi:10.1074/jbc.m211634200 (2003). 173 Morand, O. H., Zoeller, R. A. & Raetz, C. R. Disappearance of plasmalogens from membranes of animal cells subjected to photosensitized oxidation. J Biol Chem 263, 11597-11606 (1988). 174 Zoeller, R. A., Morand, O. H. & Raetz, C. R. A possible role for plasmalogens in protecting animal cells against photosensitized killing. J Biol Chem 263, 11590-11596 (1988). 175 Reiss, D., Beyer, K. & Engelmann, B. Delayed oxidative degradation of polyunsaturated diacyl phospholipids in the presence of plasmalogen phospholipids in vitro. Biochem J 323 ( Pt 3), 807-814, doi:10.1042/bj3230807 (1997). 176 Maeba, R., Sawada, Y., Shimasaki, H., Takahashi, I. & Ueta, N. Ethanolamine plasmalogens protect cholesterol-rich liposomal membranes from oxidation caused by free radicals. Chem Phys Lipids 120, 145-151, doi:10.1016/s0009-3084(02)00101-9 (2002). 177 White, A. L., Modaff, P., Holland-Morris, F. & Pauli, R. M. Natural history of rhizomelic chondrodysplasia punctata. Am J Med Genet A 118a, 332-342, doi:10.1002/ajmg.a.20009 (2003). 178 Gonzalez-Covarrubias, V. et al. Lipidomics of familial longevity. Aging Cell 12, 426-434, doi:10.1111/acel.12064 (2013). 179 Pradas, I. et al. Exceptional human longevity is associated with a specific plasma phenotype of ether lipids. Redox biology 21, 101127 (2019). 180 Ghosh, M. K. & Hajra, A. K. Subcellular distribution and properties of acyl/alkyl dihydroxyacetone phosphate reductase in rodent livers. Archives of biochemistry and biophysics 245, 523-530 (1986). 181 Hardeman, D. & Van den Bosch, H. Topography of ether phospholipid biosynthesis. Biochimica et Biophysica Acta (BBA)-Lipids and Lipid Metabolism 1006, 1-8 (1989). 182 Singh, H., Beckman, K. & Poulos, A. Exclusive localization in peroxisomes of dihydroxyacetone phosphate acyltransferase and alkyl-dihydroxyacetone phosphate synthase in rat liver. Journal of lipid research 34, 467-477 (1993). 183 Hua, R. et al. VAPs and ACBD5 tether peroxisomes to the ER for peroxisome maintenance and lipid homeostasis. Journal of Cell Biology 216, 367-377 (2017). 184 Drechsler, R., Chen, S. W., Dancy, B. C., Mehrabkhani, L. & Olsen, C. P. HPLC-Based Mass Spectrometry Characterizes the Phospholipid Alterations in Ether-Linked Lipid 320 Deficiency Models Following Oxidative Stress. PLoS One 11, e0167229, doi:10.1371/journal.pone.0167229 (2016). 185 Shi, X. et al. A Caenorhabditis elegans model for ether lipid biosynthesis and function. Journal of lipid research 57, 265-275 (2016). 186 Benjamin, D. I. et al. Ether lipid generating enzyme AGPS alters the balance of structural and signaling lipids to fuel cancer pathogenicity. Proc Natl Acad Sci U S A 110, 14912- 14917, doi:10.1073/pnas.1310894110 (2013). 187 Van Raamsdonk, J. M. & Hekimi, S. FUdR causes a twofold increase in the lifespan of the mitochondrial mutant gas-1. Mech Ageing Dev 132, 519-521, doi:10.1016/j.mad.2011.08.006 (2011). 188 Sogame, Y., Kitamura, A., Yabuki, M. & Komuro, S. A comparison of uptake of metformin and phenformin mediated by hOCT1 in human hepatocytes. Biopharm Drug Dispos 30, 476-484, doi:10.1002/bdd.684 (2009). 189 Segal, E. D. et al. Relevance of the OCT1 transporter to the antineoplastic effect of biguanides. Biochem Biophys Res Commun 414, 694-699, doi:10.1016/j.bbrc.2011.09.134 (2011). 190 Hermann, G. J. et al. Genetic analysis of lysosomal trafficking in Caenorhabditis elegans. Mol Biol Cell 16, 3273-3288, doi:10.1091/mbc.e05-01-0060 (2005). 191 Zhang, S. O., Trimble, R., Guo, F. & Mak, H. Y. Lipid droplets as ubiquitous fat storage organelles in C. elegans. BMC Cell Biol 11, 96, doi:10.1186/1471-2121-11-96 (2010). 192 Zhang, S. O. et al. Genetic and dietary regulation of lipid droplet expansion in Caenorhabditis elegans. Proc Natl Acad Sci U S A 107, 4640-4645, doi:10.1073/pnas.0912308107 (2010). 193 Honsho, M., Asaoku, S. & Fujiki, Y. Posttranslational regulation of fatty acyl-CoA reductase 1, Far1, controls ether glycerophospholipid synthesis. J Biol Chem 285, 8537- 8542, doi:10.1074/jbc.M109.083311 (2010). 194 Cabreiro, F. et al. Metformin retards aging in C. elegans by altering microbial folate and methionine metabolism. Cell 153, 228-239, doi:10.1016/j.cell.2013.02.035 (2013). 195 Pryor, R. et al. Host-Microbe-Drug-Nutrient Screen Identifies Bacterial Effectors of Metformin Therapy. Cell 178, 1299-1312 e1229, doi:10.1016/j.cell.2019.08.003 (2019). 196 Dean, J. M. & Lodhi, I. J. Structural and functional roles of ether lipids. Protein Cell 9, 196-206, doi:10.1007/s13238-017-0423-5 (2018). 197 Perez, C. L. & Van Gilst, M. R. A 13C isotope labeling strategy reveals the influence of insulin signaling on lipogenesis in C. elegans. Cell Metab 8, 266-274, doi:10.1016/j.cmet.2008.08.007 (2008). 198 Perez, M. A. & Watts, J. L. Worms, Fat, and Death: Caenorhabditis elegans Lipid Metabolites Regulate Cell Death. Metabolites 11, doi:10.3390/metabo11020125 (2021). 321 199 Shmookler Reis, R. J. et al. Modulation of lipid biosynthesis contributes to stress resistance and longevity of C. elegans mutants. Aging (Albany NY) 3, 125-147, doi:10.18632/aging.100275 (2011). 200 Horikawa, M., Nomura, T., Hashimoto, T. & Sakamoto, K. Elongation and desaturation of fatty acids are critical in growth, lipid metabolism and ontogeny of Caenorhabditis elegans. J Biochem 144, 149-158, doi:10.1093/jb/mvn055 (2008). 201 Apfeld, J., O'Connor, G., McDonagh, T., DiStefano, P . S. & Curtis, R. The AMP-activated protein kinase AAK-2 links energy levels and insulin-like signals to lifespan in C. elegans. Genes Dev 18, 3004-3009, doi:10.1101/gad.1255404 (2004). 202 Curtis, R., O'Connor, G. & DiStefano, P. S. Aging networks in Caenorhabditis elegans: AMP-activated protein kinase (aak-2) links multiple aging and metabolism pathways. Aging Cell 5, 119-126, doi:https://doi.org/10.1111/j.1474-9726.2006.00205.x (2006). 203 Senchuk, M. M. et al. Activation of DAF-16/FOXO by reactive oxygen species contributes to longevity in long-lived mitochondrial mutants in Caenorhabditis elegans. PLoS Genet 14, e1007268, doi:10.1371/journal.pgen.1007268 (2018). 204 Schreiber, M. A., Pierce-Shimomura, J. T., Chan, S., Parry, D. & McIntire, S. L. Manipulation of behavioral decline in Caenorhabditis elegans with the Rag GTPase raga-1. PLoS Genet 6, e1000972, doi:10.1371/journal.pgen.1000972 (2010). 205 Kenyon, C., Chang, J., Gensch, E., Rudner, A. & Tabtiang, R. A C. elegans mutant that lives twice as long as wild type. Nature 366, 461-464, doi:10.1038/366461a0 (1993). 206 Zou, Y. et al. Plasticity of ether lipids promotes ferroptosis susceptibility and evasion. Nature 585, 603-608, doi:10.1038/s41586-020-2732-8 (2020). 207 Perez, M. A., Magtanong, L., Dixon, S. J. & Watts, J. L. Dietary Lipids Induce Ferroptosis in Caenorhabditiselegans and Human Cancer Cells. Dev Cell 54, 447-454.e444, doi:10.1016/j.devcel.2020.06.019 (2020). 208 Sakamoto, T., Maebayashi, K., Nakagawa, Y. & Imai, H. Deletion of the four phospholipid hydroperoxide glutathione peroxidase genes accelerates aging in Caenorhabditis elegans. Genes Cells 19, 778-792, doi:10.1111/gtc.12175 (2014). 209 Braverman, N. et al. Human PEX7 encodes the peroxisomal PTS2 receptor and is responsible for rhizomelic chondrodysplasia punctata. Nature genetics 15, 369-376 (1997). 210 Motley, A. M. et al. Rhizomelic chondrodysplasia punctata is a peroxisomal protein targeting disease caused by a non-functional PTS2 receptor. Nature genetics 15, 377- 380 (1997). 211 Purdue, P. E., Zhang, J. W., Skoneczny, M. & Lazarow, P. B. Rhizomelic chondrodysplasia punctata is caused by deficiency of human PEX7, a homologue of the yeast PTS2 receptor. Nature genetics 15, 381-384 (1997). 322 212 Itzkovitz, B. et al. Functional characterization of novel mutations in GNPAT and AGPS, causing rhizomelic chondrodysplasia punctata (RCDP) types 2 and 3. Hum Mutat 33, 189-197, doi:10.1002/humu.21623 (2012). 213 Grimm, M. O. et al. From brain to food: analysis of phosphatidylcholins, lyso- phosphatidylcholins and phosphatidylcholin-plasmalogens derivates in Alzheimer's disease human post mortem brains and mice model via mass spectrometry. J Chromatogr A 1218, 7713-7722, doi:10.1016/j.chroma.2011.07.073 (2011). 214 Goodenowe, D. B. et al. Peripheral ethanolamine plasmalogen deficiency: a logical causative factor in Alzheimer's disease and dementia. J Lipid Res 48, 2485-2498, doi:10.1194/jlr.P700023-JLR200 (2007). 215 Han, X., Holtzman, D. M. & McKeel, D. W., Jr. Plasmalogen deficiency in early Alzheimer's disease subjects and in animal models: molecular characterization using electrospray ionization mass spectrometry. J Neurochem 77, 1168-1180, doi:10.1046/j.1471-4159.2001.00332.x (2001). 216 Albert, D. H. & Anderson, C. E. Ether-linked glycerolipids in human brain tumors. Lipids 12, 188-192, doi:10.1007/bf02533292 (1977). 217 Snyder, F. & Wood, R. Alkyl and alk-1-enyl ethers of glycerol in lipids from normal and neoplastic human tissues. Cancer Res 29, 251-257 (1969). 218 Jaffrès, P. A. et al. Alkyl ether lipids, ion channels and lipid raft reorganization in cancer therapy. Pharmacol Ther 165, 114-131, doi:10.1016/j.pharmthera.2016.06.003 (2016). 219 Arthur, G. & Bittman, R. Glycosylated antitumor ether lipids: activity and mechanism of action. Anticancer Agents Med Chem 14, 592-606, doi:10.2174/1871520614666140309231144 (2014). 220 Mitchell, T. W., Buffenstein, R. & Hulbert, A. J. Membrane phospholipid composition may contribute to exceptional longevity of the naked mole-rat (Heterocephalus glaber): a comparative study using shotgun lipidomics. Exp Gerontol 42, 1053-1062, doi:10.1016/j.exger.2007.09.004 (2007). 221 Munro, D. & Blier, P. U. The extreme longevity of Arctica islandica is associated with increased peroxidation resistance in mitochondrial membranes. Aging Cell 11, 845-855, doi:10.1111/j.1474-9726.2012.00847.x (2012). 222 Shi, X. et al. A Caenorhabditis elegans model for ether lipid biosynthesis and function. J Lipid Res 57, 265-275, doi:10.1194/jlr.M064808 (2016). 223 Kamath, R. S. & Ahringer, J. Genome-wide RNAi screening in Caenorhabditis elegans. Methods 30, 313-321 (2003). 224 Soukas, A. A., Kane, E. A., Carr, C. E., Melo, J. A. & Ruvkun, G. Rictor/TORC2 regulates fat metabolism, feeding, growth, and life span in Caenorhabditis elegans. Genes & development 23, 496-511 (2009). 323 225 Petrascheck, M. & Miller, D. L. Computational Analysis of Lifespan Experiment Reproducibility. Front Genet 8, 92, doi:10.3389/fgene.2017.00092 (2017). 226 Han, S. K. et al. OASIS 2: online application for survival analysis 2 with features for the analysis of maximal lifespan and healthspan in aging research. Oncotarget 7, 56147- 56152, doi:10.18632/oncotarget.11269 (2016). 227 Pino, E. C. & Soukas, A. A. Quantitative Profiling of Lipid Species in Caenorhabditis elegans with Gas Chromatography-Mass Spectrometry. Methods Mol Biol 2144, 111- 123, doi:10.1007/978-1-0716-0592-9_10 (2020). 228 Pino, E. C., Webster, C. M., Carr, C. E. & Soukas, A. A. Biochemical and high throughput microscopic assessment of fat mass in Caenorhabditis elegans. J Vis Exp, doi:10.3791/50180 (2013). 229 Pang, Z. et al. MetaboAnalyst 5.0: narrowing the gap between raw spectra and functional insights. Nucleic Acids Research 49, W388-W396, doi:10.1093/nar/gkab382 (2021). 230 Benjamini, Y. & Hochberg, Y. Controlling the False Discovery Rate: A Practical and Powerful Approach to Multiple Testing. Journal of the Royal Statistical Society. Series B (Methodological) 57, 289-300 (1995). 231 Stirling, D. R. et al. CellProfiler 4: improvements in speed, utility and usability. BMC Bioinformatics 22, 433, doi:10.1186/s12859-021-04344-9 (2021). 232 Schindelin, J. et al. Fiji: an open-source platform for biological-image analysis. Nat Methods 9, 676-682, doi:10.1038/nmeth.2019 (2012). 233 Watts, J. L. Using Caenorhabditis elegans to Uncover Conserved Functions of Omega-3 and Omega-6 Fatty Acids. Journal of Clinical Medicine 5, 19 (2016). 234 Mansoorabadi, S. O., Thibodeaux, C. J. & Liu, H. W. The diverse roles of flavin coenzymes--nature's most versatile thespians. J Org Chem 72, 6329-6342, doi:10.1021/jo0703092 (2007). 235 Massey, V. Introduction: flavoprotein structure and mechanism. FASEB J 9, 473-475, doi:10.1096/fasebj.9.7.7737454 (1995). 236 Lienhart, W. D., Gudipati, V. & Macheroux, P. The human flavoproteome. Arch Biochem Biophys 535, 150-162, doi:10.1016/j.abb.2013.02.015 (2013). 237 Powers, H. J. Riboflavin (vitamin B-2) and health. Am J Clin Nutr 77, 1352-1360, doi:10.1093/ajcn/77.6.1352 (2003). 238 Moriyama, Y. Riboflavin transporter is finally identified. J Biochem 150, 341-343, doi:10.1093/jb/mvr095 (2011). 239 Yonezawa, A., Masuda, S., Katsura, T . & Inui, K. Identification and functional characterization of a novel human and rat riboflavin transporter, RFT1. Am J Physiol Cell Physiol 295, C632-641, doi:10.1152/ajpcell.00019.2008 (2008). 324 240 Yamamoto, S. et al. Identification and functional characterization of rat riboflavin transporter 2. J Biochem 145, 437-443, doi:10.1093/jb/mvn181 (2009). 241 Subramanian, V. S. et al. Differential expression of human riboflavin transporters -1, -2, and -3 in polarized epithelia: a key role for hRFT-2 in intestinal riboflavin uptake. Biochim Biophys Acta 1808, 3016-3021, doi:10.1016/j.bbamem.2011.08.004 (2011). 242 Nabokina, S. M., Subramanian, V. S. & Said, H. M. Effect of clinical mutations on functionality of the human riboflavin transporter-2 (hRFT-2). Mol Genet Metab 105, 652- 657, doi:10.1016/j.ymgme.2011.12.021 (2012). 243 Dipti, S. et al. Brown-Vialetto-Van Laere syndrome; variability in age at onset and disease progression highlighting the phenotypic overlap with Fazio-Londe disease. Brain Dev 27, 443-446, doi:10.1016/j.braindev.2004.10.003 (2005). 244 Spagnoli, C. & De Sousa, C. Brown-Vialetto-Van Laere syndrome and Fazio-Londe disease - treatable motor neuron diseases of childhood. Dev Med Child Neurol 54, 292- 293, doi:10.1111/j.1469-8749.2011.04179.x (2012). 245 Biswas, A., Elmatari, D., Rothman, J., LaMunyon, C. W. & Said, H. M. Identification and functional characterization of the Caenorhabditis elegans riboflavin transporters rft-1 and rft-2. PLoS One 8, e58190, doi:10.1371/journal.pone.0058190 (2013). 246 Dillin, A. et al. Rates of behavior and aging specified by mitochondrial function during development. Science 298, 2398-2401, doi:10.1126/science.1077780 (2002). 247 Feng, J., Bussiere, F. & Hekimi, S. Mitochondrial electron transport is a key determinant of life span in Caenorhabditis elegans. Dev Cell 1, 633-644, doi:10.1016/s1534- 5807(01)00071-5 (2001). 248 Gandhimathi, K., Karthi, S., Manimaran, P., Varalakshmi, P. & Ashokkumar, B. Riboflavin transporter-2 (rft-2) of Caenorhabditis elegans: Adaptive and developmental regulation. J Biosci 40, 257-268, doi:10.1007/s12038-015-9512-x (2015). 249 Greer, E. L. et al. An AMPK-FOXO pathway mediates longevity induced by a novel method of dietary restriction in C. elegans. Curr Biol 17, 1646-1656, doi:10.1016/j.cub.2007.08.047 (2007). 250 Lee, S. S., Kennedy, S., Tolonen, A. C. & Ruvkun, G. DAF-16 target genes that control C. elegans life-span and metabolism. Science 300, 644-647, doi:10.1126/science.1083614 (2003). 251 Paradis, S. & Ruvkun, G. Caenorhabditis elegans Akt/PKB transduces insulin receptor- like signals from AGE-1 PI3 kinase to the DAF-16 transcription factor. Genes Dev 12, 2488-2498, doi:10.1101/gad.12.16.2488 (1998). 252 Liu, Z. et al. Mapping metabolic changes by noninvasive, multiparametric, high- resolution imaging using endogenous contrast. Sci Adv 4, eaap9302, doi:10.1126/sciadv.aap9302 (2018). 325 253 Alonzo, C. A. et al. Two-photon excited fluorescence of intrinsic fluorophores enables label-free assessment of adipose tissue function. Sci Rep 6, 31012, doi:10.1038/srep31012 (2016). 254 Kimura, K., Tanaka, N., Nakamura, N., Takano, S. & Ohkuma, S. Knockdown of mitochondrial heat shock protein 70 promotes progeria-like phenotypes in caenorhabditis elegans. J Biol Chem 282, 5910-5918, doi:10.1074/jbc.M609025200 (2007). 255 Wu, Z. et al. Mitochondrial unfolded protein response transcription factor ATFS-1 promotes longevity in a long-lived mitochondrial mutant through activation of stress response pathways. BMC Biol 16, 147, doi:10.1186/s12915-018-0615-3 (2018). 256 Soo, S. K. & Van Raamsdonk, J. M. High confidence ATFS-1 target genes for quantifying activation of the mitochondrial unfolded protein response. MicroPubl Biol 2021, doi:10.17912/micropub.biology.000484 (2021). 257 Bennett, C. F. et al. Activation of the mitochondrial unfolded protein response does not predict longevity in Caenorhabditis elegans. Nat Commun 5, 3483, doi:10.1038/ncomms4483 (2014). 258 Nieva, C. et al. The lipid phenotype of breast cancer cells characterized by Raman microspectroscopy: towards a stratification of malignancy. PLoS One 7, e46456, doi:10.1371/journal.pone.0046456 (2012). 259 Potcoava, M. C., Futia, G. L., Aughenbaugh, J., Schlaepfer, I. R. & Gibson, E. A. Raman and coherent anti-Stokes Raman scattering microscopy studies of changes in lipid content and composition in hormone-treated breast and prostate cancer cells. J Biomed Opt 19, 111605, doi:10.1117/1.Jbo.19.11.111605 (2014). 260 Watson, E., MacNeil, L. T., Arda, H. E., Zhu, L. J. & Walhout, A. J. M. Integration of metabolic and gene regulatory networks modulates the C. elegans dietary response. Cell 153, 253-266, doi:10.1016/j.cell.2013.02.050 (2013). 261 Han, S. et al. Mono-unsaturated fatty acids link H3K4me3 modifiers to C. elegans lifespan. Nature 544, 185-190, doi:10.1038/nature21686 (2017). 262 Van Gilst, M. R., Hadjivassiliou, H. & Yamamoto, K. R. A Caenorhabditis elegans nutrient response system partially dependent on nuclear receptor NHR-49. Proc Natl Acad Sci U S A 102, 13496-13501, doi:10.1073/pnas.0506234102 (2005). 263 Wu, L. et al. An Ancient, Unified Mechanism for Metformin Growth Inhibition in C. elegans and Cancer. Cell 167, 1705-1718 e1713, doi:10.1016/j.cell.2016.11.055 (2016). 264 Panowski, S. H., Wolff, S., Aguilaniu, H., Durieux, J. & Dillin, A. PHA-4/Foxa mediates diet-restriction-induced longevity of C. elegans. Nature 447, 550-555, doi:10.1038/nature05837 (2007). 265 Durieux, J., Wolff, S. & Dillin, A. The cell-non-autonomous nature of electron transport chain-mediated longevity. Cell 144, 79-91, doi:10.1016/j.cell.2010.12.016 (2011). 326 266 Schiavi, A. et al. Autophagy induction extends lifespan and reduces lipid content in response to frataxin silencing in C. elegans. Exp Gerontol 48, 191-201, doi:10.1016/j.exger.2012.12.002 (2013). 267 Austin, M. U. et al. Knockout of the folate transporter folt-1 causes germline and somatic defects in C. elegans. BMC Dev Biol 10, 46, doi:10.1186/1471-213X-10-46 (2010). 268 Bito, T., Matsunaga, Y., Yabuta, Y., Kawano, T. & Watanabe, F. Vitamin B12 deficiency in Caenorhabditis elegans results in loss of fertility, extended life cycle, and reduced lifespan. FEBS Open Bio 3, 112-117, doi:10.1016/j.fob.2013.01.008 (2013). 269 Yang, R. et al. NHR-80 senses the mitochondrial UPR to rewire citrate metabolism for lipid accumulation in Caenorhabditis elegans. Cell Rep 38, 110206, doi:10.1016/j.celrep.2021.110206 (2022). 270 Kim, H. E. et al. Lipid Biosynthesis Coordinates a Mitochondrial-to-Cytosolic Stress Response. Cell 166, 1539-1552 e1516, doi:10.1016/j.cell.2016.08.027 (2016). 271 Soukas, A. A., Kane, E. A., Carr, C. E., Melo, J. A. & Ruvkun, G. Rictor/TORC2 regulates fat metabolism, feeding, growth, and life span in Caenorhabditis elegans. Genes Dev 23, 496-511, doi:10.1101/gad.1775409 (2009). 272 Gioran, A. et al. Multi-omics identify xanthine as a pro-survival metabolite for nematodes with mitochondrial dysfunction. EMBO J 38, doi:10.15252/embj.201899558 (2019). 273 Garratt, M. et al. Male lifespan extension with 17-alpha estradiol is linked to a sex- specific metabolomic response modulated by gonadal hormones in mice. Aging Cell 17, e12786, doi:10.1111/acel.12786 (2018). 274 Lagerborg, K. A., Watrous, J. D., Cheng, S. & Jain, M. High-Throughput Measure of Bioactive Lipids Using Non-targeted Mass Spectrometry. Methods Mol Biol 1862, 17-35, doi:10.1007/978-1-4939-8769-6_2 (2019). 275 Kim, E., Sun, L., Gabel, C. V. & Fang-Yen, C. Long-term imaging of Caenorhabditis elegans using nanoparticle-mediated immobilization. PLoS One 8, e53419, doi:10.1371/journal.pone.0053419 (2013). 276 Verma, S. P . & Wallach, D. F . Raman spectra of some saturated, unsaturated and deuterated C18 fatty acids in the HCH-deformation and CH-stretching regions. Biochim Biophys Acta 486, 217-227, doi:10.1016/0005-2760(77)90018-2 (1977). 277 Freudiger, C. W. et al. Highly specific label-free molecular imaging with spectrally tailored excitation stimulated Raman scattering (STE-SRS) microscopy. Nat Photonics 5, 103-109, doi:10.1038/nphoton.2010.294 (2011).

Abstract (if available)

Abstract Food is imperative to fuel growth and essential cellular functions for all organisms. All animals receive crucial nutrients from diet to support development, metabolism, aging, and survival. Diet is an extremely variable aspect in life due to the expansive number of options within an organism’s environment. Although this idea isn’t a novel concept because the phrase “You are what you eat” was coined in the 1800’s and is still commonly used in popular culture, understanding how diet can influence phenotypic outcomes has yet to be elucidated. Caenorhabditis elegans have been a common model organism used in diet studies within the laboratory environment due to the shared core metabolic pathway with mammals. In order to diversify the menu available to culture C. elegans in the lab, we have isolated and cultured three such microbes: Methylobacterium, Xanthomonas, and Sphingomonas. The nutritional composition of these bacterial foods is unique, and when fed to C. elegans, can differentially alter multiple life history traits including development, reproduction, and metabolism. In light of the influence each food source has on specific physiological attributes, we comprehensively assessed the impact of these bacteria on animal health and devised a blueprint for utilizing different food combinations over the lifespan, in order to promote longevity. The expansion of the bacterial food options to use in the laboratory will provide a critical tool to better understand the complexities of bacterial diets and subsequent changes in physiology and gene expression.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

SKN-1 coordination of stress adaptation, metabolism, and resource allocation in Caenorhabditis elegans

PDF

Regulation of Caenorhabditis elegans small RNA pathways: an examination of Argonaute protein RNA binding and post-translational modifications in C. elegans germline

PDF

Characterization of genetic and physiological responses to environmental stress in Caenorhabditis elegans across the lifespan

PDF

SIMR-1 facilitates robust silencing of piRNA target loci in the C. elegans germline

PDF

Defining novel molecular mechanisms to restore balance of a stress response homeostat across spatial and temporal boundaries

PDF

The role of mitochondria in the male reproductive capacity of Caenorhabditis elegans

PDF

Elucidating the mechanism behind the negative outcomes associated with hyperactive SKN-1

PDF

Dissecting novel roles for MAFR-1 in reproduction and metabolic homeostasis

PDF

Phenotypic and multi-omic characterization of novel C. elegans models of Alzheimer's disease

PDF

The organization of small RNA pathways within C. elegans germ granules: mutator foci formation, regulation, and interaction

PDF

Alpha-ketoglutarate, an endogenous metabolite, extends lifespan and compresses morbidity in aging mice

PDF

TLR4-mediated innate immune response and neuroinflammation: focus on APOE and obesity

PDF

Swimming exercise reduces native ⍺-synuclein protein species in a transgenic C. elegans model of Parkinson’s disease and confers downstream protective effects independent of food restriction…

PDF

Computational approaches to identify genetic regulators of aging and late-life mortality

PDF

WDR23-mediated regulation of the DNA damage response, oxidative stress resistance, and glucose metabolism

PDF

Genomic and physiological characterization of Escherichia coli evolving in long-term batch culture

PDF

Defining the functional roles of neurotransmitters and neuropeptides in neural circuits

PDF

Packaging and delivery of plasmids by bacterial extracellular vesicles

PDF

Novel roles of the WDR23-DDB1-CUL4 ubiquitin ligase complex in cytoprotection

PDF

Characterizing the physiological roles and regulatory mechanisms of Maf1

Asset Metadata

Creator Stuhr, Nicole Lynn (author)

Core Title Genetic basis of diet-dependent responses across the lifespan in Caenorhabditis elegans

School College of Letters, Arts and Sciences

Degree Doctor of Philosophy

Degree Program Molecular Biology

Degree Conferral Date 2023-08

Publication Date 06/08/2023

Defense Date 06/08/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aging,C elegans,diet,gene-diet pairs,healthspan,lifespan,OAI-PMH Harvest,Physiology,worms

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Curran, Sean P. (committee chair), Benayoun, Berenice (committee member), Phillips, Carolyn (committee member), Tower, John (committee member)

Creator Email nstuhr@usc.edu,nstuhr15@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113169864

Unique identifier UC113169864

Identifier etd-StuhrNicol-11941.pdf (filename)

Legacy Identifier etd-StuhrNicol-11941

Document Type Dissertation

Format theses (aat)

Rights Stuhr, Nicole Lynn

Internet Media Type application/pdf

Type texts

Source 20230612-usctheses-batch-1054 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu