University of Southern California Dissertations and Theses

The regulation, roles, and mechanism of action of mitochondrial-derived-peptides (MDPs) in aging

(USC Thesis Other)

The regulation, roles, and mechanism of action of mitochondrial-derived-peptides (MDPs) in aging

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content THE REGULATION, ROLES, AND MECHANISM OF ACTION OF
MITOCHONDRIAL-DERIVED-PEPTIDES (MDPS) IN AGING
by
Jialin Xiao
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
in Partial Fulfilment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(BIOLOGY OF AGING)
December 2018

i
ACKNOWLEDGEMENTS
As graduation approaches, I have reflected upon my time at USC, not only the time at my
bench doing experiments, but also the time in conference brainstorming with my colleagues,
mentor, and realize how wonderful and rewarding these five years have been to me. I would like
to thank my advisor Dr. Pinchas Cohen for presenting the tremendous opportunity to study as his
PhD student. There is a lot to learn from him: his research approach is tactical and innovative;
his great personality and effective time management; above all he is fulfilled with a true passion
for the pursuit of knowledge. Thank you for always supporting my ideas, for giving me the
freedom to explore and grow throughout my five years in the lab.
Much gratitude must be given to my graduate committee members: for their time,
suggestion and support.
Furthermore, I would like to thank all the Cohen group members I have overlapped with:
Su-jeong Kim, Kelvin Yen, Junxiang Wan, Hemal Mehta, Brendan Miller, Jennifer Zeng,
Richard Wong and Noel Guerrero. I appreciate your company: I enjoy publishing papers and
making jokes about Hassy together.
To all the other excellent collaborators I have worked with over the years – Dr. Andrew
Hoffman, Dr. Stephen Freedland and Dr. Christina Wang – I would like to extend my
appreciation. I would not have overcome the challenges completing this thesis not having help
from them.
I would like to thank my parents, Weiwen Xiao and Xianming Huang, you have always
shown tremendous patience and support every time I felt depressed, discouraged or perplexed,

ii
and motivated me to reach for the stars and the moon. Thank you for always believing in me and
support every decision I made. The achievement of this degree is as much mine as it is yours.
To my supporting friends that are still striving in their PhD studies – Xiaoyu Cai, Yuan
Zhang – I wish you a fun and fruitful journey.
To my undergraduate volunteer Hao Zheng, who offered not only help with experiments,
but friendship. Thank you for listening to my inner feelings, sharing with me your happiness and
sorrow.
No one deserves any more credit than Zhiyao Lu, who has shown extreme tolerance of my
undoubtedly immature conduct and provided unconditional support towards my finishing this
challenging thesis. I am forever grateful, and I look forward to our years ahead.

iii
TABLE OF CONTENTS
TABLE OF CONTENTS .......................................................................................................... iii
LIST OF FIGURES ..................................................................................................................vii
LIST OF TABLES ...................................................................................................................... x
ABBREVIATIONS ................................................................................................................... xi
ABSTRACT ........................................................................................................................... xvii
CHAPTER 1 INTRODUCTION ................................................................................................. 1
MITOCHONDRIA AND AGING ..................................................................................................... 1
Overview of Mitochondria .................................................................................................... 1
The Role of Mitochondria in Aging and in Age-related Diseases ........................................... 2
Retrograde Signaling ............................................................................................................ 8
MITOCHONDRIA-DERIVED-PEPTIDES (MDPS) ......................................................................... 11
Overview of Mitochondria-Derived-Peptides (MDPs)......................................................... 11
Humanin ............................................................................................................................. 12
MOTS-c .............................................................................................................................. 16
Small-Humanin-Like-Peptides (SHLPs) .............................................................................. 17
The Mitochondrial Origin of Humanin, MOTS-c and SHLPs .............................................. 19
MITOCHONDRIA-DERIVED-PEPTIDES (MDPS), AGING AND AGE-RELATED DISEASES ............... 21
Overview of MDPs and Age-Related Diseases .................................................................... 21
Cancer ................................................................................................................................ 21
Neurodegenerative Disorders ............................................................................................. 23
Cardiovascular Diseases .................................................................................................... 24

iv
Metabolic Syndromes ......................................................................................................... 25
MITOCHONDRIA-DERIVED-PEPTIDES (MDPS) AND METABOLISM ............................................ 28
Cellular Bioenergetics ........................................................................................................ 28
Amino Acid, Lipid, and Nucleotide Metabolism .................................................................. 30
Systemic Glucose Homeostasis and Adiposity ..................................................................... 30
Humanin and IGF-I ............................................................................................................ 34
MITOCHONDRIA BIOLOGY AND PROSTATE CANCER ETHNIC DISPARITY ................................... 38
Introduction ........................................................................................................................ 38
Mitochondrial-Related Genetic Alterations in Prostate Cancer .......................................... 41
Mitochondrial Content and Proteomic Changes in Prostate Cancer ................................... 48
Epigenetics and Mitochondria in Prostate Cancer .............................................................. 52
Environmental Factors in Prostate Cancer ......................................................................... 54
CHAPTER 2 HUMANIN AS A DIETARY MIMETIC ............................................................ 57
ABSTRACT .............................................................................................................................. 57
BACKGROUND ........................................................................................................................ 59
RESULTS ................................................................................................................................ 65
HNG Suppresses Plasma IGF-I and Increases IGFBP-1 Levels, while Enhancing CP-
Induced Tumor Suppression in Tumor-Bearing Mice .......................................................... 65
Long-Term HNG Treatment Promotes Visceral Fat and Weight Loss Without an Overall
Reduction in Food Intake .................................................................................................... 67
HNG Treatment Improved Metabolic Markers and Reduced Inflammation ......................... 69
Effects of HNG on Memory, Recognition and Motor Function ............................................ 70
HNG Activated AMPK Pathway and Mitochondrial Biogenesis in Muscle .......................... 73

v
The Molecular Mechanism of HNG Induced Mitochondrial Biogenesis and ATP Generation
........................................................................................................................................... 77
DISCUSSION............................................................................................................................ 83
Differential Stress Resistance (DSR) by Humanin ............................................................... 83
Humanin Acts As a CR-Mimetic and Extends Healthspan ................................................... 87
Humanin Activates AMPK Pathway and Improves Muscle Bioenergetics............................ 90
MATERIALS AND METHODS .................................................................................................... 94
CHAPTER 3 CHARACTERIZATION OF SMALL-HUMANIN-LIKE PEPTIDE 2 (SHLP2)100
ABSTRACT ............................................................................................................................ 100
BACKGROUND ...................................................................................................................... 102
RESULTS .............................................................................................................................. 108
Identification and Validation of Small Open Reading Frames (sORFs) Within Mitochondrial
16S rRNA ......................................................................................................................... 108
SHLP2 is an Insulin Sensitizer Acting both Peripherally and Centrally............................. 110
SHLP2 is a Bioactive Peptide that Modulates Mitochondrial Function ............................. 110
SHLP2 Inhibits IAPP Misfolding ...................................................................................... 113
Low Circulating Levels of SHLP2 as Novel Biomarker for Prostate Cancer Risk .............. 115
Multivariate Analysis of Metabolomic Profile Alterations in Response to SHLP2 Treatment
......................................................................................................................................... 122
Changes in Metabolite Profile with HNG or SHLP2 Treatment ........................................ 127
DISCUSSION.......................................................................................................................... 135
Implications of SHLP2 in Cellular Function in Age-Related Diseases............................... 135
SHLP2 is a Central and Peripheral Insulin Sensitizer ....................................................... 136

vi
SHLP2 in Aging ................................................................................................................ 137
SHLP2 and IAPP Interaction ............................................................................................ 137
SHLP2 as A Novel Prostate Cancer Biomarker................................................................. 139
SHLP2 Treatment Reduced γ-Glutamyl Cycle and ROS .................................................... 143
MATERIALS AND METHODS .................................................................................................. 147
FUTURE DIRECTIONS ............................................................................................................ 159
SUPPLEMENTAL FIGURES AND TABLES .................................................................................. 160
REFERENCES ........................................................................................................................ 162

vii
LIST OF FIGURES
Figure 1-1 Illustrative Diagram of How Mitochondrial DNA Mutation and Damage Can Lead to
Dysregulation of Cellular Parameters and Fitness ................................................................. 7
Figure 1-2 The Overview of Mitochondrial Retrograde Signaling Pathways.............................. 10
Figure 1-3 Locations of Mitochondrial-Derived-Peptides (MDPs) in The Mitochondrial Genome
........................................................................................................................................... 12
Figure 1-4 Critical Function of Each Amino Acid Residue in Humanin ..................................... 15
Figure 1-5 Schematic Illustration of Humanin-, MOTS-C- and SHLPs-Mediated Signaling
Pathways and Biological Functions..................................................................................... 16
Figure 1-6 Schematic Diagram of GH, IGF-I, and Humanin Regulation .................................... 36
Figure 1-7 Mitochondria-Related Factors Interact and Contribute to Prostate Cancer ................ 56
Figure 2-1 The Overlapping Effects of Humanin and CR .......................................................... 62
Figure 2-2 Humanin Exerts CR-Mimetic Effects by Suppressing IGF-I and Increasing IGFBP-1
in tumor-bearing mice......................................................................................................... 66
Figure 2-3 Humanin Failed to Alter Circulating IGF-I and IGFBP-1 Levels in Normal Mice .... 67
Figure 2-4 Humanin Promotes Lean Body Mass, Reduces Body Weight and Visceral Fat
Without Changing Food Intake ........................................................................................... 69
Figure 2-5 HNG Treatment Suppresses IGF-I and Circulating Glucose, Reduces Pro-
Inflammatory Cytokines ..................................................................................................... 71

viii
Figure 2-6 HNG Improves Motor Function, Hippocampal-Dependent Learning, and Working
Memory, Possibly by Reducing Activated Microglia .......................................................... 72
Figure 2-7 HNG is Potent Inducer of AMPK in vivo and in vitro .............................................. 74
Figure 2-8 HNG Treatment Enhances Mitochondrial Biogenesis, Demonstrated by Upregulation
of MtDNA Copy Number and Elevation of Mito-Gene Expression ..................................... 76
Figure 2-9 HNG Increases NRF-1 Binding Affinity to TFAM Promoter ................................... 79
Figure 2-10 Inhibition of AMPK Diminished HNG-Induced ATP Production ........................... 81
Figure 2-11 Humanin Acts as A Caloric Restriction (CR) Mimetic and Improves Healthspan ... 93
Figure 3-1 Identification and Validation of Small-Open-Reading-Frames (sORFs) Within the
Mitochondrial 16S Ribosomal RNA (rRNA) Gene ........................................................... 109
Figure 3-2 SHLP2 Modulates Mitochondrial Membrane Potential and H2O2 release ............... 112
Figure 3-3 Mitochondrial-Derived Peptides HNG and SHLP2 Inhibit IAPP Fibrilization ........ 114
Figure 3-4 The Distribution of SHLP2 Levels Stratified by Race or Outcome ......................... 119
Figure 3-5 The Distribution of SHLP2 Levels and a Cut-Off at 350-pg/ml .............................. 121
Figure 3-6 ROC Curve and AUC Statistics Before and After Adding SHLP2 in the Model ..... 122
Figure 3-7 Principle Component Analysis (PCA) of Selected Metabolic Pathways and
Hierarchical Clustering of All Measured Metabolites ........................................................ 124
Figure 3-8 Random Forest Classification Using Named Metabolites in Plasma of Control
Compared to Plasma of HNG and SHLP2 Treated Mice ................................................... 126

ix
Figure 3-9 The Metabolite Profile of the Transmethylation, Transulfuration and Gamma-
Glutamyl Cycle ................................................................................................................ 130
Figure 3-10 Major Intermediates in the Gamma-Glutamyl Cycle Decreased, Suggesting
Reduction in Oxidative Stress Burden ............................................................................... 131
Figure 3-11 The Metabolite Profile of the Sphingolipid Metabolism ....................................... 134

x
LIST OF TABLES
Table 1-1 Locations and Sequences of Published MDPs............................................................ 19
Table 2-1Changes in Acylcarnitines, Malonate and Glucose Levels When Treated with HNG .. 82
Table 3-1 Baseline Characteristics in Cases and Controls ........................................................ 115
Table 3-2 Association between SHLP2 and other variables (SHLP2 cut-off at 350-pg/ml) ...... 117
Table 3-3 Association Between SHLP2 and Overall Risk of Cancer and Risk of Cancer Grade,
Stratified by Race ............................................................................................................. 120

xi
ABBREVIATIONS
ORF: Open-Reading-Frame
MDP: Mitochondrial-Derived-Peptide
SHLP: Small-Humanin-like-Peptide
MOTS-c: Mitochondrial Open-Reading-Frame of the 12 S rRNA-c
AMPK: AMP-Activated-Protein-Kinase
CR: Caloric Restriction
IGF-I: Insulin-like Growth Factor-I
mtDNA: Mitochondrial DNA
ddPCR: Digital Droplet PCR
COBRA: Combined Bisulfite Restriction Analysis
IAPP: Islet Amyloid Polypeptide
OXPHOS: Oxidative Phosphorylation
mRNA: Messenger RNA
tRNA: Transfer RNA
rRNA: Ribosomal RNA
MELAS: Mitochondrial Encephalomyopathy, Lactic Acidosis and Stroke-
like Episodes
TFAM: Mitochondrial Transcription Factor A
HMG: High-Mobility Group
TFB2M: Mitochondrial Transcription Factor B2
MERRF: Myoclonic Epilepsy Associated with Ragged-Red Fibers
POLG: Mitochondrial Polymerase Gamma

xii
ROS: Reactive Oxygen Species
MFRTA: Mitochondrial Free-Radical Theory of Aging
RC: Respiratory Chain
SASP: Senescence-Associated Secretory Phenotype
ETC: Electron Transport Chain
MiDAS: Mitochondrial Dysfunction-Associated Senescence
IL: Interleukin
Rho-zero cells: Mitochondrial DNA-Depleted Cells
AD: Alzheimer’s Disease
PD: Parkinson’s Disease
ALS: Amyotrophic Lateral Sclerosis
LRRK2: Leucine-Rich Repeat Kinase-2
COX: Cytochrome Oxidase
TCA Cycle: Tricarboxylic Acid Cycle
HIF-1: Hypoxia Inducible Factor-1
UPR
mt
: Mitochondrial Unfolded Protein Response
FAD: Familial Alzheimer’s Disease
APP: Amyloid Precursor Protein
CNTFR: Ciliary Neurotrophic Factor Receptor
gp130: Glycoprotein 130
JAK: Janus Kinase
STAT: Signal Transducer and Activator of Transcription
AKT: Protein Kinase B

xiii
ERK: Extracellular Signal-Regulated Kinase
FPRL: N-Formyl Peptide Receptor-like
AICAR: 5-Aminoimidazole-4-Carboxamide Ribonucleotide
IS-GDR: Insulin-Stimulated Glucose Disposal Rate
GTT: Glucose Tolerance Test
HFD: High Fat Diet
RANKL: Receptor Activator of Nuclear Factor-κB Ligand
MRSA: Methicillin-Resistant S. aureus
MCP: Monocyte Chemoattractant Protein
NUMT: Nuclear Mitochondrial DNA Segment
MPP8: M-Phase Phosphoprotein 8
PARP: Poly (ADP-Ribose) Polymerase
p75NTR: 75-kDa Neurotrophin Receptor
I/R: Ischemia and Reperfusion
CVD: Cardiovascular Disease
T2DM: Type-2-Diabetes Mellitus
T1DM: Type-1-Diabetes Mellitus
GIR: Glucose Infusion Rate
GSIS: Glucose-Stimulated Insulin Secretion
NOD: Non-Obese Diabetic
OCR: Oxygen Consumption Rate
hRPE Cells: Retinal Pigment Epithelial Cells
LID Mice: Liver-Specific IGF-I Deficient Mice

xiv
GHRH: Growth Hormone Releasing Hormone
CP: Cyclophosphamide
SNP: Single Nucleotide Polymorphism
INDEL: Insertion/Deletion
PIN: Prostatic Intraepithelial Neoplasia
LCM: Laser Capture Microdissection
ND3: NADH Dehydrogenase 3
COI: Cytochrome Oxidase Subunit I
TMA: Tissue Microarray
ZIP: Zrt- and Irt-like Protein
SOD: Superoxide Dismutase
HSP: Heat-Shock Protein
TET: Ten-Eleven Translocation methylcytosine Dioxygenase
MSC: Mesenchymal Stem Cell
GSTP: Glutathione S-Transferase Pi
IGFBP-1: Insulin-like Growth Factor Binding Protein-1
IIS: Insulin and Insulin-like Growth Factor-I Signaling
TOR: Target of Rapamycin
FOXO: Forkhead Box O
IRS: Insulin Receptor Substrate
SNF1: Sucrose Non-Fermenting 1 Kinase
LKB1: Liver Kinase B1
WBC: White Blood Cells

xv
SSC: Spermatogonial Stem Cells
PGC-1α: Peroxisome Proliferator-Activated Receptor-γ Coactivator-1
LonP1: Lon Protease-1
SLC25A5: Solute Carrier
UQCRC2: Ubiquinol-Cytochrome C Reductase Core Protein II
NRF-1: Nuclear Respiratory Factor-1
ChIP: Chromatin Immunoprecipitation
TNF-α: Tumor Necrosis Factor-α
LPS: Liposaccharide
SAB: Spontaneous Alternative Behavior
EB: Escape Box
DIO: Diet-Induced Obesity
ER: Endoplasmic Reticulum
ThT: Thioflavin T
EPR: Electron Paramagnetic Resonance Spectroscopy
CD: Circular Dichroism
PSA: Prostate-Specific Antigen
STS: Staurosporine
CSF: Cerebro-Spinal Fluid
SDSL: Site-Directed Spin Labelling
BMI: Body Mass Index
DRE: Digital Rectal Examination
AUC: Area under the Curve

xvi
PCA: Principle Component Analysis
RF: Random Forest
GGT: Gamma-Glutamyl Transferase
SAM: S-Adenosylmethionine
S1P: Sphingosine-1-Phosphate

xvii
ABSTRACT
The discovery of humanin nearly two decades ago has ushered in a new interest in
mitochondrial biology and has combined several nascent fields of research. Over the past decade,
with advancement of the current proteomic technologies and progress in the field of small open-
reading-frames (ORFs), the concept of mitochondrial small ORFs has been established. This
innovative concept opened up the possibility that the mitochondrial genome encodes for more
than 13 proteins and contains molecular instructions for biological functions besides energy
production. Humanin is the first of several mitochondrial-derived-peptides (MDPs) that are
originated within small, alternative ORFs of the mitochondrial genome. Humanin has a number
of different cytoprotective and metaboloprotective effects while the recently discovered small
humanin-like peptides (SHLP) 1-6 have similar and distinct properties compared to humanin.
The discovery of another MDP called MOTS-c as an exercise mimetic and activator of AMP-
activated protein kinase (AMPK) suggests that these peptides will have an important role in
metabolism and could be used as future therapeutics.
Humanin and other MDPs are small circulating peptides that have important signaling
functions and participate in the mitochondrial retrograde signaling events. While nearly 200
papers were published on humanin and other MDPs since their discovery, most of them have
focused on the protective effects of this molecule in vitro and in vivo, showing its ability to
ameliorate damage induced by multiple stressors and disease conditions. Yet little is known
about their fundamental mechanisms: for example, the enigma of humanin expression in the
mitochondria is still largely unsolved. Similarly, the actual mechanism of action of humanin and
the target organs involved remain unclear. To fill this gap, I utilized rodent models and tissue
culture to examine the systemic mechanism involved, particularly for the caloric-restriction

xviii
(CR)-mimetic effects. Chapter 2 summarized the how humanin modulated three major pathways
closely associated with CR: the insulin-like growth factor-I (IGF-I) pathway, the AMPK
pathway and the oxidative stress to prolong healthspan and confer differential protection, which
are established benefits of dietary interventions.
My next main project involves the characterization of a family of humanin-like peptides
called SHLPs. They exhibit potent signaling effects, such as regulation of mitochondrial
metabolism and glucose homeostasis. My research interest mainly revolved around SHLP2,
since it resembles humanin in terms of signaling and biological functions. I demonstrated the
presence of polyadenylated mRNA transcripts of SHLP2 from the 16S rRNA region of
mitochondrial genome. Furthermore, I found it regulated mitochondrial oxidative stress and
improved multiple metabolic biomarkers such as circulating sphingolipids and key intermediates
in the gamma-glutamyl cycle, which provides mechanistic explanation of my finding that low
circulating levels of SHLP2 can serve as a novel biomarker for prostate cancer risk. Similarities
between humanin and SHLP2 were characterized: they both interact with islet amyloid
polypeptide (IAPP) structurally and interfere IAPP aggregation; SHLP2 and humanin also have
overlapping metabolic targets.
Our observation of differential regulation of SHLP2 levels in black and white men coupled
with the known differential risk for prostate cancer incidence and mortality rates between races is
very intriguing. Since SHLP2 is derived from mitochondria, the lower levels of SHLP2 in black
men are most likely resulted from mitochondrial genetic factors such as black-specific single
nucleotide polymorphisms (SNPs), heteroplasmy or mitochondrial DNA (mtDNA) copy number
changes. Therefore, the identification of prostate cancer-specific mitochondrial genomic markers
will be extremely helpful in understanding racial disparity of this disease. With the help of next-

xix
generation sequencing and bioinformatic tools, I did a comprehensive analysis of 192 DNA
samples from a pre-existing patient cohort at the Durham VA. Unfortunately, I failed to establish
any correlation between prostate cancer and mutation load or heteroplasmy. Yet, there were few
mitochondrial SNPs overrepresented in prostate cancer cases.
While mtDNA damage and mutations have long been implicated in the aging process and a
number of mitochondrial signaling pathways can induce cellular senescence, the role of
mitochondrial epigenetics has largely been unexplored. Especially, how mitochondrial epigenetic
regulations affect MDP expression. Lastly, I combined several complementary methods
including bisulfite sequencing and combined bisulfite restriction analysis (COBRA) to study
alterations of mtDNA methylation patterns during cellular senescence and in response to air
pollution. In addition, I optimized a digital droplet PCR (ddPCR)-based method to quantify CpG
methylation percentages. This novel method avoids the step of bisulfite conversion, thus
enabling rapid methylation quantification with higher sensitivity and lower cost.

1
CHAPTER 1 INTRODUCTION
Mitochondria and Aging
Overview of Mitochondria
Mitochondria are the primary sites for generating energy in most eukaryotic cells. As a cellular
organelle with endosymbiotic origin, the mitochondrion possesses several unique features,
including enclosed double membrane and harboring its own genome (1). The double-stranded
circular mitochondrial DNA (mtDNA) spans about 16.5kbp, and encodes 13 full size proteins, all
of which are essential components of the mitochondrial oxidative phosphorylation (OXPHOS)
system (2). The mitochondrial genome also encodes 13 messenger RNAs (mRNAs), 22 transfer
RNAs (tRNAs) and 2 ribosomal RNAs (rRNAs) that are required for mitochondrial translational
machinery. The D-loop is the only regulatory region which contains sequences for DNA
replication and three promoters required for transcription initiation (3). Polycistronic primary
transcripts are generated from each strand (heavy strand and light strand) and subsequently
processed to give rise to mature rRNAs, tRNAs, and mRNAs (4). Mutations and deletions of
mtDNA can lead to pathologies. For example, mitochondrial encephalomyopathy, lactic acidosis
and stroke-like episodes (MELAS) is caused by mutations in mitochondrial encoded NADH
dehydrogenases and mitochondrial tRNAs (5). Therefore, the maintenance of mtDNA integrity
and regulation mitochondrial gene expression are critical for mitochondrial functionality, hence
metabolic fitness of the cell. As for mtDNA replication, the DNA polymerase (pol) γ is the sole
DNA polymerase in mammalian mitochondria (6). Different from nuclear DNA, which only
replicates once during cell division, mtDNA needs to be continuously replicated due to the fact
that mitochondria undergo constant fission and fusion (7,8). Pol γ has DNA polymerase (9), 3’-
5’ exonuclease (10) and 5′dRP lyase activities (11), thereby conferring high base-substitution

2
fidelity. The mitochondrial transcription factor A (TFAM) is the first mitochondrial transcription
factor identified and is vital for expression, maintenance and organization of the mitochondrial
genome (12). TFAM protein contains two tandem high-mobility group (HMG)-box domains
connected with a basic linker region and a C-terminal tail essential for physical interaction with
the mitochondrial transcription factor B2 (TFB2M) to initiate transcription (13). Similar to other
HMG proteins, TFAM binds, unwinds and bends DNA efficiently without sequence selectivity.
With its structural characteristics, TFAM also serves as a “histone-like” protein to package
mtDNA into compact nucleoids (14).
Despite mitochondria’s simple genetic organization, many aspects of mitochondrial biology are
poorly understood. Studies on classic mitochondrial diseases mainly focused on the mtDNA
mutations or mutations in nuclear mitochondrial genes that disturb mtDNA maintenance, such as
MELAS, MERRF syndrome (myoclonic epilepsy associated with ragged-red fibers) and POLG-
related disorders (15,16). It was previously believed that all mtDNA diseases lead to OXPHOS
dysfunctions and expected to have similar manifestations, however, the clinical phenotypes of
these diseases are very distinct (17). These facts indicate that the disruption of mtDNA can give
rise to dysregulation of mitochondrial functions in addition to energy production. In other words,
mitochondrial genome does not only contain molecular instructions for assembly of essential
respiratory complexes, but also other factors determining critical mitochondrial functions.
The Role of Mitochondria in Aging and in Age-related Diseases
Increasing evidence has suggested that mitochondrial dysfunction is correlated with a number of
age-related diseases, such as diabetes, cancer and neurodegeneration. However, how decline in
mitochondrial function leads to these diseases is not yet understood. Biochemical reactions
taking place in mitochondria have the potential to damage cells because electron leakage from

3
the respiratory chain can result in reactive oxygen species (ROS) production. According to the
mitochondrial free-radical theory of aging (MFRTA), decline in mitochondrial integrity can
amplify the oxidative stress that drives the aging process (18). This theory is supported by the
observations that mitochondrial ROS production increases with age and that mutations of
mtDNA accumulate during aging (19,20). These somatic mtDNA mutations create a “vicious
cycle”, which in turn results in a further increase in ROS production and accumulated oxidative
damage to proteins, lipids, and DNA (21). In fact, mitochondrial DNA (mtDNA) mutations were
shown to directly promote aging. The very first experimental proof came from the mtDNA
mutator mice. These mice express a proofreading-deficient mtDNA polymerase (Polg
mut
) and
have extensive mtDNA mutations. Most importantly, they display myriad pre-mature aging
phenotype (22,23). However, the mtDNA mutator mice have little or no increase in ROS levels
and oxidative damage, despite respiratory chain (RC) function is severely influenced (23,24).
This finding argues against MFRTA theory and the “vicious cycle” concept. Clearly, much about
the role of mitochondria in aging needs to be explored, and it is necessary to consider
mitochondrial functions other than ROS production.
Speaking of aging, it is an organismal deterioration process resulted from a chain of molecular
events. There is no single protein or pathway dictating aging, and aging is accompanied with vast
changes in biochemicals and signaling pathways inter-related with each other. Due to the
complexity of aging research at the organismal level, researchers sought to obtain more
knowledge from cells, the “building blocks of life”, and found that senescent cells accumulate
and the risk of developing neurodegenerative disorders, chronic metabolic diseases and other
types of organ deteriorations also increases during aging (25,26). Compelling evidence also
suggests that the accumulation of senescent cells in tissue stem cell pool or progenitor cell pool

4
over time leads to tissue degeneration (27). The causal relationship between cellular senescence
and aging is further supported by the study demonstrating that clearance of senescence cells
delayed manifestation of aging phenotypes in mouse model (28). Moreover, the depletion of
senescent marker p16
Ink4a
alleviates stem cell aging and some age-related phenotypes (29).
Mitochondrion, the “powerhouse of the cell”, becomes the key player in the massive remodeling
of cellular energetics during cellular senescence. Its involvement allows cells to shift from
proliferation to synthesis of senescence-associated secretory phenotype (SASP) factors (30).
However, it is still debatable whether mitochondria directly cause senescence, or simply serve as
an effector. A number of mitochondrial related events occur during senescence, including
elevated ROS levels and enlarged mitochondria as a result of increased fusion (31). While some
cellular senescence models have hyper-activated mitochondria and increased OXPHOS activity
(32), other studies have identified mitochondrial stress as a potent inducer of senescence (33). On
one hand, a study conducted by the Wiley et al., 2016 demonstrated that different forms of
electron transport chain (ETC) inhibition altered NAD
+
/NADH and AMP/ATP ratios, the
subsequent AMP-activated protein kinase (AMPK) activation led to p53-induced senescence.
They further characterized a distinct secretory phenotype accompanied with the mitochondrial
dysfunction associated senescence (MiDAS), which lacks interleukin-1 (IL-1) dependent
inflammatory arm and has potent paracrine effects (33). On the other hand, another group found
out the absence of mitochondria alleviated a spectrum of senescence phenotypes without
affecting ATP production (34). Although there is a discrepancy revolving around the exact role
of mitochondria in senescence, both studies confirmed that the secretion of certain pro-
inflammatory SASP factors is closely associated with mitochondria. The studies on senescence

5
and mitochondria clearly demonstrated mitochondrial dysfunctions can influence the cell via
secondary messengers such as NAD
+
and ATP.
Tumorigenesis and senescence are two antagonizing mechanisms that occur during aging. While
the number of senescent cells increases with age, the incidence of neoplasia also increases. This
divergence can be partially explained by carcinogenesis-favoring microenvironment created by
secretion of pro-inflammatory SASP factors from senescent cells (35). Mitochondria are
involved in the secretion of certain pro-inflammatory SASP factors and closely associated with
cancer. For this reason, it is reasonable to hypothesize that mitochondria are at the nexus of the
signaling pathways leading to the two biological phenomena. The first line of evidence
supporting a link between cancer and mitochondria is the “Warburg effect”, where cancer cells
rely heavily on aerobic glycolysis instead of oxidative phosphorylation to generate energy (36).
This phenomenon could be an adaptation of cancer cells to hypoxic environment and higher
demand on proliferation (37,38). It could also be a result of mitochondrial damage, such as
OXPHOS dysfunction caused by mtDNA mutations. Although the causal relationship between
cancer and mitochondria has not yet been fully established, mitochondrial impairment is clearly
involved in the pathophysiology of multiple types of cancer (39). The mutation rate of mtDNA is
higher than that of nuclear DNA and this high mutation rate results in part from the mtDNA’s
lack of protective histones and an inefficient DNA repair system (40). mtDNA mutations lead to
altered protein transcription/translation, which causes electron transport chain dysfunctions and
increases the production of ROS. As a result, ROS exacerbates protein and mtDNA damage – a
“vicious cycle” (41). Both large-deletion and point mutation in mtDNA have been found in
cancer. Point mutations in mtDNA might influence the activity of certain proteins encoded in
mtDNA, depending on the site of these mutations. If the mutation is in a tRNA or an rRNA, it

6
may affect the activity of multiple mitochondrial encoded proteins. If the mutation is in the
control region, it may affect replication and transcription. Cells with a high proportion of large-
deletion mutant mtDNA might have characteristics similar to those with depleted mtDNA.
Experimentally-derived mtDNA-depleted cells (Rho-zero cells) are highly resistant to apoptosis
(42), and some of them display invasive phenotypes (43). These observations suggest that
dysregulated mitochondrial function as well as mtDNA are critical for cancer cells in the process
of acquiring apoptosis resistance and invasiveness.
As the critical regulator of energy production and cell death, mitochondria have a central role to
play in age-related neurodegeneration. Although neurodegenerative diseases such as Alzheimer’s
disease (AD), Parkinson’s disease (PD), amyotrophic lateral sclerosis (ALS) are heterogeneous
in terms of genetics and symptoms, a common feature is neuron death through intrinsic
mitochondrial apoptotic pathway (44). Moreover, the brain represents only 2% of the body
weight but accounts for 20% of total body oxygen consumption. This enormous energy demand
is largely driven by neurons as they require energy to maintain ion gradients for the generation of
action potentials (45). For this reason, neuronal health and function are highly dependent on
mitochondrial OXPHOS as the energy requirement is continuous; even brief periods of oxygen
or glucose deprivation result in neuronal death. As mentioned previously, the accumulation of
mtDNA mutations and deletions during aging is correlated with decline in mitochondrial
function, which might be the underlying explanation for the increasing risk of neurodegeneration
with age (46). However, correlative studies do not indicate causation. The conditioned TFAM
knock-out (KO) mice in dopamine neurons have reduced mtDNA copy number and OXPHOS
deficiency also showed parkinsonism phenotype (47). While mtDNA mutator mice with marked

7
disruption in respiratory function have progeric phenotypes without neuropathology (48),
humans with POLG mutations exhibit parkinsonism (49). The difference between rodent model
and human is not yet understood, nevertheless, majority of these mouse models display different
extents of mitochondrial dysfunction (50). For example, PD mouse models incorporating
transgenes harboring dominant mutations in leucine-rich repeat kinase-2 (LRRK2) and α-
synuclein independently cause PD, albeit all induce mitochondrial abnormalities (51,52). Studies
on AD also support that neuronal mitochondrial bioenergetic defects such as cytochrome c
oxidase (COX) activity, oxygen consumption, and ROS production precede or drive Aβ
production and plaque formation (53–55).
As aging is a degenerative process contributed by a range of biological and environmental
factors. These factors are often interrelated, thus forming a dynamic network of interacting
pathways.

Figure 1-1 Illustrative Diagram of How Mitochondrial DNA Mutation and Damage Can Lead to
Dysregulation of Cellular Parameters and Fitness
Mitochondria, the power plant and the signaling hub of a cell, are therefore at the nexus of this
interaction network of aging. The study of aging and aging-related pathologies will undoubtedly
Mitochondrion
Nucleus
Mutations & Damages ↑
NADH/NAD
+
↑
ATP/ADP↓
OXPHOS function ↓
Redox Homeostasis ↓
ROS effects
Senescence ↑
Tumorigenesis ↑
Repair↓
Aging

8
involve research on mitochondrial biology. Mitochondria regulate ATP production,
NADH/NAD
+
ratio, redox homeostasis and calcium signaling, which is pivotal to cellular and
metabolic fitness (Figure 1-1). Although it is not clear whether mitochondrial dysfunction
precedes aging or the other way around, improvement on mitochondrial health provides a
promising therapeutic avenue in the fight against aging.
Retrograde Signaling
The citric acid cycle, also commonly referred to as the TCA cycle, generates metabolites and
reducing equivalents such as NADH, FADH2. Electrons from reducing equivalents are then fed
into the mitochondrial ETC and the electron transfer is coupled with the transfer of protons (H
+

ions) across the inner membrane followed by ATP synthesis. The two essential functions of
mitochondria to generate energy and to support biosynthesis are balanced to meet the cellular
needs, which suggests the existence of an effective communication mechanism between
mitochondria and other organelles. Clearly, mitochondria receive signals in response to stress
and metabolic changes (anterograde signaling), most notably through nuclear-encoded proteins
responsible for nearly every mitochondrial process. Moreover, cytosolic calcium can enter into
mitochondria to regulate bioenergetics (56). Emerging data also suggest that mitochondria are
not just receiving the signals but also actively sending signals back to the cytosol and nucleus
(retrograde signaling) (57). One critical function of retrograde signaling is to ensure that cell and
mitochondria are coordinated in a way that enables cells to commit biological processes with
input from the mitochondria, harmonizing the metabolic demands of the cell and the ability of
the mitochondria to meet them. The very first observation of mitochondrial retrograde signaling
is mitochondrial release of cytochrome c to induce apoptosis (58). Mitochondria are major sites
of superoxide and hydrogen peroxide generation. The two small molecules have robust signaling

9
roles. Hydrogen peroxide can diffuse into the cytoplasm to modify reactive thiol side chains
(e.g., cysteine) of proteins (59). These redox modifications are able to alter protein structure and
affect its activity. Superoxide, on the other hand, is negatively charged and hence cannot easily
diffuse across cell membranes. However, it may still participate in signaling events in the
cytoplasm via undefined pathways. Apart from directly modulate protein structure, ROS also
trigger hypoxic gene expression through enhancing binding of hypoxia inducible factor 1 (HIF-
1) to promoters during hypoxia (60). Furthermore, Cross talk between mitochondria and the
nucleus plays an important role in the aging process, and mitochondrial stress signaling
constitutes a major component of stress response. Mitochondrial genetic and metabolic stress
cause induction of mitochondria-specific heat shock proteins and promoting cytosolic calcium-
dependent signaling (61,62). The mitochondrial unfolded protein response is one of the
retrograde signaling pathways which increase mitochondrial-localized chaperons and proteases
to recover the mitochondrial protein homeostasis (63,64). Interestingly, the mitochondrial
unfolded protein response (UPR
mt
) also modulates cellular metabolism including increase in
glycolysis and decreases the gene expression of TCA cycle and OXPHOS potentially to reduce
mitochondrial stress and alter cellular metabolism to promote survival (64,65). The
aforementioned pathways contribute to mitochondrial retrograde signaling. Their signals are very
diverse, including small molecules, metabolites, ions and proteins, which are sensed and
transmitted through proteins such as kinases/phosphatases to alter cellular functions (Figure 1-2).
In the past decade, it has been reported that the export of mitochondrial peptides can also
contribute to retrograde signaling: non-assembled mitochondrial proteins are degraded to small
peptides that are released into the cytoplasm in an ATP- and temperature-dependent manner (66).
The accumulated peptides outside mitochondria can then trigger proteotoxic stress and

10
downstream UPR
mt
effects. The concept of MDPs is different from the former phenomenon.
Distinct from the peptides generated by degradation of full proteins, MDPs are encoded by small
open reading frames (sORFs) and alternative open reading frames (altORFs) that are located in
mtDNA. Moreover, MDPs are “mitokines” in the form of peptides that are secreted and act as
extracellular signals to affect different organs of the body. To date, multiple MDPs have been
discovered and characterized, including humanin, small-humanin-like peptides (SHLPs) and the
mitochondrial ORF within the 12S rRNA c (MOTS-c). They have diverse cellular targets and
functions, which will be discussed in detail in the following section.

Figure 1-2 The Overview of Mitochondrial Retrograde Signaling Pathways
2
nd
MESSENGERS TARGETS
Calcineurin, Kinases
Kinases, ATPases,
Adenylyl cyclases
GSH/GSSG,
Cys/CySS Redox-
sensitive enzymes
PARPs, Sirtuins
CELLULAREFFECTS
• Activation of signaling
pathways
• (De)Activation of
nuclear genes
• Chromatin remodeling
• Cytoprotection
• Activation of AMPK,
Akt pathways
• Enhancing
mitochondrial functions

11
Mitochondria-Derived-Peptides (MDPs)
Overview of Mitochondria-Derived-Peptides (MDPs)
sORFs and altORFs are ubiquitously present in all genomes, but their biological functions have
not been thoroughly studied (67,68). The translation of some of them have been confirmed (69),
however, the coding capacity of the mitochondrial sORFs has long been overlooked. Previously,
it was commonly accepted that the human mtDNA was an extremely compact genome that only
encoded 13 essential components of the electron transport chain and 24 structural RNAs required
for their translation (70). The mitochondrial genome displays exceptional economy of
organization, with no introns and very few noncoding nucleotides between coding sequences
(71). Nearly two decades ago, Hashimoto et al., 2001 and Ikonen et al., 2003 independently
discovered humanin, the first of several MDPs that are found within small, alternative reading
frame of the mitochondrial 16S rRNA (Figure 1-3). Following the discovery of humanin, more
studies established its protective effects against various disease-relevant insults. The discovery of
humanin paved the way for a more detailed exploration of other MDPs; MOTS-c was found to
be an exercise mimetic and a potent activator of AMPK encoded from an altORF in the
mitochondrial 12S rRNA (72), while the recently discovered SHLP 1-6 have similar and distinct
properties compared to humanin (73). These peptides can be secreted into extracellular space and
delivered to other organs through circulation, which confers them more diverse roles than
communication within the cell. With mitochondria being the hub of cellular signaling, research
on MDPs may provide targets for future therapeutics.

12

Figure 1-3 Locations of Mitochondrial-Derived-Peptides (MDPs) in The Mitochondrial Genome
Humanin
Humanin is a “mitokine” that has a number of different cytoprotective and metaboloprotective
effects (74). The Nishimoto group identified the humanin cDNA sequence from a functional
expression screen. Clones that protected cells from cell death induced by a mutant form of
amyloid precursor protein (APP), a possible cause of Alzheimer’s disease, were compared and
they all shared a 75 bp sORF. This sORF is identical to a region of the mitochondrial 16S rRNA
and encodes a 24-amino acid peptide, which was called “humanin” in hopes that it would restore
the humanity back to patients with Alzheimer’s disease (75,76). Further experiments proved that
the overexpression of humanin can suppress neuronal death caused by Familial Alzheimer’s
disease (FAD) proteins including amyloid precursor protein (APP), presenilin1 and 2 (75). In a
separate, independent screen by the Reed lab, they discovered the mechanism by which humanin
SHLP3 SHLP2 SHLP4/1
SHLP5
SHLP6
Humanin
Human Mitochondrial
Genome
16S ribosomal RNA

13
acts as an anti-apoptotic molecule (77). Humanin was cloned as a Bax binding partner from a
yeast two-hybrid screen and the interaction was confirmed by co-immunoprecipitation. Although
other Bcl-2 family proteins such as Bcl-2 and Bcl-B share structural similarity with Bax, they did
not interact with humanin. Furthermore, humanin could suppress apoptosis in a Bax-dependent
way by preventing Bax translocation from cytosol to mitochondria, but apoptosis induced by
Bax-independent stimulus such as necrosis induction was not suppressed by humanin (77). As
the first MDP to be discovered, humanin has been found to play a diverse role in a number of
different processes. At almost the same time, a study by Ikonen et al., 2003 shed light on how
humanin, insulin-like growth factor I (IGF-I) and apoptosis could be related. We cloned humanin
from another yeast two-hybrid screen when searching for IGF binding factor 3 (IGFBP-3)
interacting proteins (78). The affinity and specificity of binding between the two molecules were
determined both in vitro and in vivo. IGF-binding proteins are a group of proteins that regulate
IGF-I bioavailability by acting as carriers. In particular, IGFBP-3, the most abundant one,
accounts for 80% of all IGF binding and can participate in regulation of cell survival both
dependently or independently of IGF-I (79). The levels of IGFBP-3 are up-regulated by pro-
apoptotic signals and IGFBP-3 has been shown to induce apoptosis (80). On the other hand, IGF-
I is known to have anti-apoptotic effect and can protect against Aβ-induced neuronal death in an
IGFBP-3 sensitive fashion, as 10nM IGFBP-3 is sufficient to completely abrogate IGF-I
mediated protection (81). As will be discussed in more detail below, humanin and its analogues
also have effects on metabolism and have been shown to increase glucose stimulated insulin
release and can decrease body weight gain and visceral fat (82,83).
In the past year, several new papers of humanin have further established the positive effects of
humanin in a number of domains. Thummasorn et al., 2016 have confirmed that humanin

14
treatment can protect against ischemia/reperfusion injury and they further showed that this may
be due to a decrease in reactive oxygen species generation (84–86). Two other papers showed the
importance of humanin in neurocognition by showing that it can prevent diazepam induced
memory dysfunction as well as act as an anxiolytic agent (87,88). Gidlund et al. 2016 discovered
that muscle humanin levels are increased during resistance training compared to control and that
aerobic exercise (Nordic walking), but did not have an effect on circulating levels (89). In a
population of pre-eclampsia patients, it was found that humanin levels were elevated and this
group and the authors of the paper suggested that this could be in response to the cardiovascular
stress occurring (90).
Human is a 24- or 21-amino acid peptide depending on if its mRNA is translated in the
cytoplasm or mitochondria, respectively. It is not yet clear where the exact site of translation is,
but both synthesized forms of humanin showed anti-apoptotic effects (77). A recent study from
Parharkova et al. 2015 suggests that at least in rats, translation occurs within the mitochondria
(91). Because humanin is relatively short, systematic single amino acid substitution has been
performed on each residue to study the importance of each amino acid (Figure 1-4). Residues 3-
19 were found to be the “core domain” for humanin’s neuro-protective ability (92). Furthermore,
this region may also directly bind to Aβ, thus preventing Aβ from self-aggregating (93).
Modification of the “core domain” yielded potent humanin analogues: replacement of Ser14 with
glycine gave rise to an enhanced form of humanin (S14G/HNG) with increased potency over
1,000-fold (94). Replacement of Ser14 by D-form serine also improved humanin’s function (95).
With regards to the IGFBP-3 binding ability, Phe6 and Lys21 were identified as essential sites.
Substitution of Phe6 by alanine completely negated the interaction between humanin and
IGFBP-3, whereas Lys21 to Ala conversion only prevented binding at lower concentrations of

15
IGFBP-3 (78). HNG-F6A, is another humanin analogue with Ser14 to glycine and Phe6 to
alanine modifications. As a non-IGFBP-3 binding peptide with more potent cyto-protective
function, HNG-F6A has been shown to regulate insulin action in β cells and stimulate glucose-
stimulated insulin secretion (82). Moreover, HNG-F6A treatment prevented endothelial
dysfunction and atherosclerosis progression by down-regulating oxidative stress and apoptosis in
the developing plaque (96). The amino acid residues Leu9, Leu10, Leu11, Pro19 and Val20 were
found to be critical for humanin secretion. Other humanin analogues have also been
characterized (92), including HN-C8P (abolished binding to Bax), HN-S7A (no self-
dimerization) and HN-L12A (humanin antagonist) and these humanin analogues are able to
inhibit cyclophosphamide-induced germ cell apoptosis with the exception HN-L12A (97).

Figure 1-4 Critical Function of Each Amino Acid Residue in Humanin
While humanin exerts some of its actions via direct interaction with signaling molecules such as
Bax and IGFBP-3, it also participates in receptor-mediated signaling. Upon secretion, humanin is
believed to activate one of two membrane receptors. The first receptor described is the ciliary
M A P R G F S C L L L L T S E I D L P V K R R A
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Non-essential
Apoptosis protection
IGFBP3 binding
Secretion
Confers increased potency

16
neurotrophic factor receptor (CNTFR)/glycoprotein 130 (gp130)/WSX1 tripartite receptor that
then activates Janus kinase (JAK), signal transducer and activator of transcription (STAT),
protein kinase B (AKT), and extracellular signal-regulated kinase (ERK) (98) (Figure 1-5). The
N-formyl peptide receptor-like 1 and 2 (FPRL1, 2) receptors have also been shown to be
activated by humanin and they also signal to ERK (99,100). As the first MDP discovered,
humanin has been the most comprehensively investigated and both structural and functional
aspects have been described. Its function as a cytoprotective, anti-apoptotic peptide has been
thoroughly assessed, while its function in cognition is still being examined.

Figure 1-5 Schematic Illustration of Humanin-, MOTS-C- and SHLPs-Mediated Signaling
Pathways and Biological Functions
MOTS-c
In addition to humanin, an in silico search of the mitochondrial genome revealed several
additional potential MDPs. The mitochondrial-derived peptide called MOTS-c is a 16-amino
AICAR
gp130
CNTFR
WSX-1
JAK
STAT-3
PI3K AKT
MEK ERK
AMPK
MetabolicHomeostasis
InsulinSensitivity
Cytoprotection
MitochondrialRespiration
Cytoprotection
OxidativeStressReduction

17
acid peptide located in the 12S rRNA gene and was recently discovered by Lee et. al., 2015.
Although MOTS-c is encoded in the mitochondrial genome, it must be translated in the
cytoplasm for translation according to the mitochondrial genetic code would result in premature
stop codon. MOTS-c regulates insulin sensitivity and metabolic homeostasis via AMPK,
increases 5-aminoimidazole-4-carboxamide ribonucleotide (AICAR) levels (Figure 1-5) (72).
Reduction of this AMPK activation by chemical compounds or siRNA abolished the enhanced
glucose-stimulated glycolytic response (72). AMPK is a central regulator of cellular energy
homeostasis, it senses AMP: ATP ratio and activates essential pathway to promote energy
production (101). In vivo MOTS-c infusion significantly increased glucose clearance and insulin-
stimulated glucose disposal rate (IS-GDR) under glucose tolerance test (GTT) and clamp studies
(72). MOTS-c further prevents high-fat-diet induced obesity and insulin resistance in CD-1 mice,
as well as prevents high fat diet (HFD)-induced obesity by hypeinsulinemia independent of
caloric intake in C57BL/6 mice. Ming et al. 2016 showed that MOTS-c inhibits receptor
activator of nuclear factor-κB ligand (RANKL) -induced osteoclast formation, represses
osteoclast differentiation via AMPK activation in vitro, and suppress the ovariectomy-induced
bone loss in mice (102). Experiments on mice infected Methicillin-resistant S. aureus (MRSA)
also proved MOTS-c can inhibit ERK signaling while promote STAT3 phosphorylation (103).
More recently, the m.1382A>C polymorphism in MOTS-c coding region, which is unique for
the Northeast Asian population and would cause a Lys14Gln replacement, was found to be
related to longevity in Japanese people through its putative endocrine action (104).
Small-Humanin-Like-Peptides (SHLPs)
After the discovery of humanin, more focus has been given to the exploration of mitochondrial
peptides. With the advancement of bioinformatics. recent mitochondrial transcriptome analyses

18
revealed the existence of multiple small mRNAs transcribed from mtDNA. Cobb et al. 2016
recently reported the existence of another 6 small humanin-like peptides, named SHLPs
(SHLP1-6), within the same 16S rRNA gene in which humanin is located (73). SHLP2 and
SHLP3 share similar protective effects with humanin. They both improved mitochondrial
metabolism by increasing oxygen consumption rate and reduced apoptosis and the generation of
ROS in vitro. Just as humanin increases mitochondrial biogenesis, SHLP2 and SHLP3 may also
increase both mitochondrial biogenesis and oxygen consumption rate. Alternatively, this increase
in oxygen consumption rate could be due to increased uncoupling. Because mitochondrial
oxygen consumption is coupled to ATP production, the increase in energy production and its
TCA cycle metabolites may enhance mitochondrial metabolism. SHLP2 and SHLP3 also
enhanced 3T3-L1 pre-adipocyte differentiation. Intracerebral infusion of SHLP2 increased
glucose uptake and suppresses hepatic glucose production, suggesting that it functions as an
insulin sensitizer both peripherally and centrally (73). Further supporting their role as insulin
sensitizers. SHLP2 injection increased leptin levels but had no effect on the inflammatory
biomarkers interleukin 6 (IL-6) and monocyte chemoattractant protein-1 (MCP-1). On the other
hand, SHLP3 elevated both metabolic and inflammatory biomarkers. ERK and STAT3 are
phosphorylated upon humanin stimulation (105), similarly, SHLP2 activated both ERK and
STAT3 pathways in NIT-1 β cells in a time-dependent manner but with different kinetics than
those induced by humanin. SHLP3 activated ERK at a later time point, but did not activate
STAT3 phosphorylation, suggesting different mechanisms of protection mediated by SHLP2 and
SHLP3 (73). Similar to humanin, the circulating levels of MOTS-c and SHLP2 decline with age
(72,73), indicating that they are potential regulators of aging.

19
Table 1-1 Locations and Sequences of Published MDPs
Peptide
Amino
acids
Predicted
MW
Location Sequence
Humanin 24 2687.26 2634-2707
MAPRGFSCLLLLTSEIDLPVKRRA
MOTS-c 16 2174.60 1343-1393
MRWQEMGYIFYPRKLR
SHLP1 24 2393.59 2561-2490
MCHWAGGASNTGDARGDVFGKQAG
SHLP2 26 3017.54 2170-2092
MGVKFFTLSTRFFPSVQRAVPLWTNS
SHLP3 38 4380.15 1821-1707
MLGYNFSSFPCGTISIAPGFNFYRLYFIW
VNGLAKVVW
SHLP4 26 3131.85 2524-2446
MLEVMFLVNRRGKICRVPFTFFNLSL
SHLP5 24 2565.95 2856-2785
MYCSEVGFCSEVAPTEIFNAGLVV
SHLP6 20 2385.87 2992-3051
MLDQDIPMVQPLLKVRLFND
The Mitochondrial Origin of Humanin, MOTS-c and SHLPs
As endosymbiotic organelles, mitochondria are thought to have transferred most of its genetic
information to the host nucleus during evolution. These mitochondrial-derived sequences of
different sizes are integrated in the nuclear genome, which are described as “nuclear
mitochondrial DNA segment (NUMT)” (106). The existence of NUMTs presents a huge
challenge for researchers to confirm the mitochondrial origin of MDPS. There are NUMTs
highly similar to humanin sequence in chromosomes 1-13, but none of them are fully identical to
the original cDNA clone by Nishimoto et al., 2001, or by Ikonen et al., 2003, both of which were
100% identical to the mitochondrial sequence (107). However, the debate over whether humanin

20
can be translated from NUMTs never ceases. As for MOTS-c and SHLPs, immunoblots using
corresponding antibodies showed that their expression levels were diminished in Hela cells
devoid of mtDNA (Hela-ρ0) (72,73).

21
Mitochondria-Derived-Peptides (MDPs), Aging and Age-Related Diseases
Overview of MDPs and Age-Related Diseases
In recent years, research in the field of sORFs have broadened our understanding about
proteomics (108). mtDNA is a very compact genome, but the mitochondrial sORFs concept adds
additional complexity to the information it contains. In fact, it is now recognized that the
mitochondrial genome is able to encode more than 13 proteins, and MDPs are as additional
peptides can be derived from mitochondrial sORFs existing as “nested genes” (109). Many of
these small peptides were identified as important signaling molecules involved in aging and
aging associated diseases. As mentioned previously, the circulating levels of humanin, MOTS-c
and SHLPs are negatively correlated with age, which further corroborates their role in aging.
Mitochondrial dysfunction is one of the nine hallmarks of aging and appears to be a common
factor connecting the other hallmarks. As discussed earlier, mitochondria are closely associated
with senescence, neurodegeneration, cancer and the etiology of other age-related diseases.
MDPs, as important mitochondrial retrograde signals, directly reflect the status of mitochondrial
health. Therefore, it is unsurprising that all MDPs discovered to date either demonstrated altered
expression under age-related pathological conditions, or ameliorated symptoms when
administered exogenously. Here we will summarize relevant literature on the relationship
between MDPs and various cancers, cardiovascular disease, metabolic syndromes and
neurodegeneration. Furthermore, critical assessment of these information will also be included.
Cancer
Advancing age is the greatest risk factor for cancer overall (110). On the cellular level, aging is
associated with increasing number of senescent cells, genomic instability, dysregulation of
signaling and cellular energetics. Most of the damages are also contributing factors to cancer

22
incidence/development. Cancer is characterized by non-controllable cell growth with the
potential to spread to other tissues, whereas humanin’s primary effect is to prevent cells from
apoptosis. This raised concern that humanin might contribute to cancer development by helping
cancer cells circumvent cell death. Maximov et al., 2002 reported the up-regulation of expression
of the mitochondrial 16S rRNA gene in non-Hodgkin’s lymphoma, which is indicative of higher
expression of humanin in cancer cells (111). Later, the same group identified M-phase
phosphoprotein 8 (MPP8) as a binding protein for humanin through yeast two-hybrid system
(112). MPP8 has been shown to direct DNA methylation to suppress tumor suppressor genes and
promote metastasis. Based on these findings, researchers speculated that humanin may be up-
regulated and interact with oncoprotein to play a role in oncogenesis. Furthermore, humanin was
detected via immunostaining in four of the seven patients with cutaneous T-cell lymphoma but
not in healthy subjects. Intriguingly, humanin was found to be localized in macrophages and
granulocytes, but not in tumor cells (113). Consistent with this notion, a number of studies also
reported increased expression of the mitochondrial16S rRNA; Lu et al., 1992 used differential
hybridization between HT-29 humanin colon adenocarcinoma cells and normal cells, and found
increased levels of 16S rRNA (114), similar pattern was observed in hepatoma cells (115). The
limitation of most of the aforementioned studies, as noted by the authors, was the lack of protein
level measurements of humanin. Elevation of 16S rRNA expression does not necessarily indicate
higher levels of humanin, because humanin peptide abundance is also regulated by mRNA
processing, translation efficiency and protein turnover. Additionally, the functional consequence
of the interaction between humanin and MPP8 has not been elucidated (112), thus making
humanin’s role as an oncopeptide ambiguous.

23
Contrary to the potential role of humanin as “accomplice” to tumorigenesis and invasion of the
cancer cells, Eriksson et al., 2014 and colleagues demonstrated that, when HNG (a potent
humanin analogue) was co-injected with bortezomib into mice with human tumor xenograft, the
side effects of bortezomib was inhibited without affecting its chemotherapeutic effects (116).
They also showed that HNG prevented bortezomib-induced apoptosis by preventing Bax and
poly (ADP-ribose) polymerase (PARP) activation. Bortezomib is a proteasome inhibitor
currently used as anti-cancer drug to treat multiple myeloma and other cancers. It is known for
inducing apoptosis in growth plate chondrocytes and causing bone growth retardation in treated
mice (117,118). More interestingly, HNG alone delayed tumor growth and cancer cell doubling
time in medulloblastoma and neuroblastoma in vivo (116).
So far, the data on other MDPs and cancer are insufficient. AMPK, the primary target of MOTS-
c, has been implicated in cancer as an oncogene and tumor suppressor based on the context
(119). Further investigation of the relationship between MOTS-c and cancer will yield valuable
insight. More studies on humanin are also needed to decipher the exact role of this peptide in
cancer; whether the upregulation of humanin is a compensatory response or a cause of cancer.
Neurodegenerative Disorders
The Humanin ORF was discovered from the protected region of an AD brain, which inspired
researchers to focus on the effects of humanin on neurodegenerative diseases. Both in vivo and in
vitro experiments supported the protective role of humanin in AD-related pathology. It was
demonstrated that HN could antagonize neurotoxicity induced by a wide range of FAD genes,
including mutant APP, PS1, PS2, and other Aβ peptides (Aβ 1-42 and Aβ 25-35) in vitro (75).
Humanin does not inhibit secretion of Aβ peptides but mediates its effects via the activation of
JAK2/STAT3 pro-survival pathway (120). Independent study by Tsukamoto et al., 2006 also

24
demonstrated that the protective effect of humanin were negated by blocking the 75-kDa
neurotrophin receptor (p75NTR). In addition, it was found that the secretion of humanin is
required for the neuro-protective effects, as the non-secreted mutant humanin sequestered inside
the cell did not exhibit any protection (75). It stabilizes mitochondrial potential and prevents
release of cytochrome c (121). Moreover, HNG (a potent humanin analogue) was shown to
disaggregate Aβ fibrils in vitro (122), which provides an alternative mechanistic explanation for
humanin effects on antagonizing Aβ toxicity. As for in vivo evidence, intra-cerebro-ventricular
injection prevented impairment of spatial working memory caused by Aβ25-35 (123). In
addition, intraperitoneal injection of HNG also ameliorated behavioral deficits induced by Aβ25-
35 by reducing neuro-inflammation and apoptosis (124). Humanin treatment also had beneficial
effects in different AD mouse models, including APPswe/PS1dE9 mice, APPswe/tauP310L/PS-
1M146V triple transgenic mice and Tg2576 mice (125–127). In the triple transgenic mouse
model, however, no changes in tau phosphorylation levels were observed, which suggests that
the cytoprotective effect of humanin is independent of tau phosphorylation and aggregation. The
role of humanin was also studied in other neurological diseases, such as stroke. Humanin
treatment has been shown to reduce infarct volume, prevent neuronal cell death and improve
neurological function in ischemia and reperfusion (I/R) mouse model (128). The neuro-
protective effect of humanin on I/R injury was confirmed by several independent studies
(86,129).
Cardiovascular Diseases
Besides neurological disorders, cardiovascular disease (CVD) is another age-related disease that
has become a major focus of humanin research. Among all the tissues, heart expressed the
highest level of humanin examined by qPCR and immunodetection (130). Endogenous humanin

25
increased after myocardial I/R in mice, and atherosclerotic plaques of the coronary arteries
expressed humanin as shown by staining. Despite the fact that humanin increased in myocardial
I/R, exogenous humanin treatment attenuated myocardial I/R injury symptoms in mice, as shown
by dose-dependent decrease in infarct size and improved cardiac performance including ejection
fraction, end-systolic volume, end-diastolic volume and cardiac output (85,131). It was also
observed that HNG protected cardiac myoblasts from oxidative stress-induced cell death by
acutely increasing activity of antioxidants through involvement of non-receptor tyrosine kinases.
Furthermore, lower humanin levels are associated with coronary endothelial dysfunction (132).
Consistent with this observation, HN treatment prevented endothelial cell apoptosis caused by
Ox-LDL-induced oxidative stress (133), and preserved endothelial function in ApoE deficient
mice on high cholesterol diet by enhancing eNOS activity (96).
Metabolic Syndromes
Type-2-diabetes mellitus (T2DM) is one of the most common metabolic diseases and its
incidence increases with age. Aging is associated with decreased beta-cell proliferative capacity
and enhanced sensitivity to apoptosis (134). When functional beta-cells are deficient, insulin
secretion is impaired. Moreover, the most accepted hypothesis underlying T2DM is insulin
resistance. Both peripheral insulin resistance and impaired insulin secretion contribute to the
pathogenesis of T2DM in aging (135).
Decline in insulin action is also associated with AD. Humanin is identified as a neuroprotective
peptide and also an IGFBP-3 binding protein, which attracted interest in studying the role of
humanin in glucose homeostasis. Hyperinsulinemic-euglycemic clamp is the gold standard
method to assess insulin sensitivity specifically at the level of liver and muscle. Rats
continuously administered with humanin via intra-cerebroventricular infusion during

26
hyperinsulinemic-euglycemic clamps required higher glucose infusion rate (GIR) to maintain
normoglycemia, which indicates decreased hepatic glucose output and increased glucose uptake
in the skeletal muscle, demonstrating increased insulin sensitivity (136). Another important
finding from the same study is the central action of humanin was mediated through activation of
hypothalamic STAT-3 signaling pathway. Continuous intravenous infusion of HNGF6A, a
humanin analogue with greater stability, higher potency, and non-IGFBP-3 binding ability,
during hyperinsulinemic-euglycemic clamp increased GIR, peripheral glucose uptake, and
suppressed hepatic glucose production. Furthermore, a single intravenous injection of HNGF6A
significantly lowered the blood glucose in Zucker diabetic fatty rats (82). Additionally, humanin
increased glucose uptake into the β cells and enhanced metabolism by promoting glucose
oxidation. The in vivo consequence is an increased glucose-stimulated insulin secretion (GSIS)
in cultured beta cells and in islets isolated from wild type and diabetic mice (82). These studies
indicate a role for humanin in glucose homeostasis through both improved insulin action and
increased insulin secretion.
As for type-1-diabetes mellitus (T1DM), decreased beta-cell proliferation and beta-cell apoptosis
are important in the pathogenesis. Humanin’s anti-apoptosis effect is not restricted to neurons,
therefore, researchers sought to expand its function in other diseases or tissues. The non-obese
diabetic (NOD) mice have been extensively utilized as a model of humanin T1DM. When treated
with humanin for 6 weeks normalized glucose tolerance and treatment for 20 weeks, NOD mice
were prevented against or showed delayed onset of diabetes. Moreover, their pancreas displayed
decreased lymphocyte infiltration in the islets and decreased apoptosis.
With respect to humanin expression during metabolic syndromes, plasma humanin levels have
been associated with hyperglycemia. A significant decrease in humanin was observed in the

27
impaired fasting glucose group compared to control (137). It is also found that humanin
expression was increased in skeletal muscles from patients with MELAS, and also increased in
small arteries (138).
MOTS-c is identified as an “exercise-mimetic” peptide that targets the skeletal muscle and
influences glucose metabolism. As such, MOTS-c is implicated in a range of metabolic
regulations, such as obesity, diabetes, and aging. Acute treatment of this peptide via
intraperitoneal injection significantly enhanced performance in GTT and reduced non-fasting
glucose levels in normal mice. In the HFD-induced insulin resistance model, MOTS-c treatment
prevented the mice from getting obese and insulin resistant. Moreover, the treatment also
reduced visceral fat content and hepatic steatosis (139). Speaking of endogenous MOTS-c
regulation, mice fasted for 48 hours had reduced plasma and skeletal muscle levels of MOTS-c.
Circulating MOTS-c were lower in patients with endothelial dysfunction. Furthermore, plasma
MOTS-c levels were positively correlated with microvascular and epicardial coronary
endothelial function.

28
Mitochondria-Derived-Peptides (MDPs) and Metabolism
Cellular Bioenergetics
Mitochondria are the primary energy source for all cellular functions. Mitochondria couple the
oxidation of nutritional substrates with ATP synthesis. Carbohydrates, fats, and proteins are
broken down to glucose, free fatty acids, and amino acids that can be utilized by mitochondria to
produce ATP. Glycolysis generates 2 ATP whereas mitochondrial oxidation of pyruvate derived
from glucose and palmitate derived from fatty acids generates 31.5 and 113 ATP respectively
(140). These metabolic intermediates are translocated into mitochondrial matrix and undergo
TCA cycle and oxidative phosphorylation. The TCA cycle generates NADH and FADH2 that are
fed into an electron transport chain to provide electrons. The electron transport chain complexes
transfer electrons to oxygen and concomitantly pumps protons across the inner mitochondrial
membrane to generate proton gradient. ATP synthase, then, utilizes the proton-motive force to
produce ATP. Therefore, this proton gradient is crucial for the ATP synthesis and directly
correlates with the oxygen consumption rate (OCR) by the electron transport chain. As
mitochondria are the primary source of cellular ATP, mitochondrial quality control mechanisms
are required for cellular fitness. For example, mitophagy removes damaged mitochondria that
lose their membrane potential (141). In addition to mitochondrial turnover, mitochondrial
peptides are produced to preserve essential functions related to energy production. Humanin
directly enhances mitochondrial bioenergetics by increasing basal OCR, maximum respiration,
respiration capacity, and ATP production in retinal pigment epithelial (hRPE) cells (142). Thus,
humanin protected hRPE cells from oxidative damage (tBh treatment) via inhibition of loss of
mitochondrial bioenergetics. Increased mitochondrial biogenesis is one possible mechanism of
how cells increase mitochondrial bioenergetics. Humanin increases the copy number of mtDNA,

29
the number of mitochondria, and the expression level of mitochondrial transcription factors,
suggesting that humanin increases mitochondrial biogenesis (142). On the other hand, humanin
also suppresses the increase in mtDNA copy number in serum-deprived lymphocytes (143). The
mtDNA copy number possibly increases in this situation to support metabolic demands as well
as to compensate for the serum-deprivation-induced impairment of mitochondria. Humanin
treated cells showed higher metabolic activity beyond control and mtDNA copy number could be
decreased to maintain their cellular homeostasis. The humanin analogue HNG also increases
ATP production and mitochondrial bioenergetics. In addition, HNG increases the mitochondrial
membrane potential in H9C2 myoblast cells, and it rescues the loss of membrane potential in
response to H2O2 treatment (144). Since increased mitochondrial membrane potential is linked to
elevated cellular ATP production, in turn, the cellular ATP level is elevated in the presence of
HNG in H9C2 cells. Substitution of Phe6 by Alanine completely negated the interaction between
humanin or HNG and IGFBP-3 and generated an enhanced form of humanin (HNF6A or
HNGF6A) (78). HNGF6A regulates glucose metabolism and energy production. HNGF6A
promoted the glucose-induced GLUT2 transporter translocation to the plasma membrane, and
increased glucose oxidation and ATP production in bTC3 cells (82). HNGF6A also elevated the
mitochondrial membrane potential in the cells. Similar to humanin, both SHLP2 and SHLP3
increase mitochondrial respiration and ATP production (73).
MOTS-c, which is encoded from the 12S rRNA region of mitochondria, has also been reported
to influence mitochondrial metabolism. MOTS-c administration increased glucose uptake and
glycolysis, whereas suppressed mitochondrial respiration in cultured cells and skeletal muscle.
This resembles a Crabtree effect-like phenomenon, which is decreased mitochondrial oxygen
consumption rate in response to high glucose uptake (72). Furthermore, SIRT1 siRNA reduces

30
the glucose-stimulated glycolytic response, which suggests that AMPK and SIRT1 play roles in
MOTS-c actions on cellular bioenergetics.
Amino Acid, Lipid, and Nucleotide Metabolism
In addition to producing ATP via the TCA cycle and oxidative phosphorylation, mitochondria
play important roles in amino acid, lipid, and nucleotide metabolism. Thus, the TCA cycle
metabolites are utilized for building of macromolecules. For example, a-ketoglutarate and
oxaloacetate can be transported into the cytosol and are utilized for de novo protein and
nucleotide synthesis (145). In addition, citrate can be transported into the cytosol and is utilized
for protein acetylation as well as de novo fatty-acid synthesis (146). One of the mitochondrial
peptides MOTS-c is closely associated with amino acid and lipid metabolism. MOTS-c
suppresses the folate-methionine cycle and levels of 5-methyl-tetrahydrofolate, the most
abundant form of active folate, is notably decreased. MOTS-c also activates AMPK by blocking
de novo purine biosynthesis, resulting in an accumulation of endogenous AICAR. Moreover,
MOTS-c affects fatty acid metabolism via the AICAR-AMPK pathway. As AMPK is the cellular
signaling hub for balancing fuel usage and energy demand, MOTS-c stimulates carnitine
shuttling, reduces levels of essential fatty acids, and increases the b-oxidation intermediates (72).
In addition, MOTS-c increases metabolite levels of NAD
+
, glycolysis and the pentose phosphate
pathway.
Systemic Glucose Homeostasis and Adiposity
ATP and metabolites generated from mitochondrial respiration modulate insulin secretion in
pancreatic b-cells. In b-cells, glucose-stimulated ATP production increases the ATP/ADP ratio,
resulting in closing of the ATP-dependent K
+
channel in β-cells. This closure of ATP-dependent

31
K+ channel leads to membrane depolarization and activation of voltage-dependent calcium
channel (147). The resulting increase in the cytoplasmic calcium concentration leads to
exocytosis of insulin in pancreatic b-cells. Patients with mitochondrial DNA mutations exhibit
impaired b-cell function and inhibition of mitochondrial metabolism can inhibit glucose-
stimulated insulin secretion (148,149). In addition to ATP, metabolites from mitochondrial
metabolism including malonyl CoA, long-chain acyl CoA, and NADPH modulate insulin
secretion by inhibiting ATP-dependent K
+
channel (150). Additionally, glutamate is generated in
the mitochondria from a-ketoglutarate and directly stimulates insulin exocytosis (151).
Mitochondria are closely associated with insulin function as well as insulin secretion. For
example, abnormal morphology, decreased mitochondrial number, decreased mitochondrial
oxidative enzymes and lower ATP production were commonly found in insulin-resistant
metabolic tissues including skeletal muscle, liver and fat (152,153). Elevated circulating free
fatty acids accumulated in these tissues will also decrease insulin-stimulated glucose disposal
(154–156). This impaired insulin signaling is a major cause of insulin resistance because it not
only affects insulin-stimulated glucose metabolism in skeletal muscle but also impairs other
actions of insulin in diverse tissues including liver, adipose tissue, and heart. Therefore, glucose
and lipid metabolism in the mitochondria is important in insulin signaling and glucose
homeostasis. Mitochondrial bioenergetics and metabolism are closely associated with insulin
signaling and glucose homeostasis (157,158). As MDPs modulate cellular bioenergetics and
metabolism in vitro, they also show systemic regulation of metabolism in vivo. Multiple animal
model studies administering humanin and its analogs support the crucial role of humanin in
glucose homeostasis. Intra-cerebroventricular administration of humanin showed increased
insulin sensitivity in the liver and muscle, causing the reduction of hepatic glucose production

32
and increased insulin-mediated AKT signaling and fatty acid metabolism signaling (136). These
effects were modulated by humanin-mediated STAT3 activation in hypothalamus. Peripheral
administration of humanin also enhanced peripheral glucose uptake and suppressed hepatic
glucose production. HNGF6A as well as HNG and humanin, but not HNF6A, shows insulin
sensitizing effects. HNGF6A increased glucose-stimulated-insulin secretion in isolated islets
from both normal and db/db mice and in mouse pancreatic cells (bTC3) (82). Elevated ATP
production and the shuttling of aspartate aminotransferases of the malate-aspartate NADH are
key mechanisms of HNGF6A regulation of insulin secretion. A hyperglycemic clamp study
found that HNGF6A enhanced glucose-stimulated-insulin secretion in young Sprague-Dawley
rats (136). Additionally, HNGF6A significantly lowers blood glucose in Zucker diabetic fatty
rats. The direct effect of HNGF6A on isolated islets and bTC3 suggests that HNGF6A mitigates
some of the metabolic abnormalities present in islets in type 2 diabetes (105). Recently, a new
role of humanin in lipid metabolism was revealed. Intraperitoneal administration of HNG
decreases body weight gain, visceral fat, and hepatic triglyceride accumulation in high-fat diet-
fed mice (83). The decrease in hepatic triglyceride accumulation is caused by increased activity
of hepatic microsomal triglyceride transfer protein and increased hepatic triglyceride secretion.
Vagotomy method removes part of the vagus nerve, and the vagotomized mice injected with
HNG blocked the humanin’s effect of both intravenous and intra-cerebroventricular infusion on
hepatic triglyceride secretion. These results suggest that the effect of humanin is mediated
through hypothalamus as the vagus nerve serving as an efferent from hypothalamus to the liver,
but not by a neuroendocrine signal.
SHLP2 and SHLP3 have insulin sensitizing effects in vitro and in vivo. Both SHLP2 and SHLP3
accelerated 3T3-L1 cell (a murine pre-adipocyte cell line) differentiation in the presence of

33
insulin (73). This suggests that SHLP 2 and 3 promote cellular differentiation and enhance
insulin sensitivity in adipose tissue. SHLP2, but not SHLP3, enhances the insulin sensitizing
effect of hepatic glucose production suppression and increased glucose disposal in peripheral
tissues. Both ATP and mitochondrial respiration metabolites are equally important for insulin
secretion. Although both SHLP2 and SHLP3 enhance ATP production, the different metabolites
modulated by SHLP2 and SHLP3 could be the possible mechanism differentiating the distinct
effects of SHLP2 and SHLP3 in vivo. Further investigation to address the mechanism is required.
MOTS-c enhances whole body insulin sensitivity, acting primarily through the muscle. MOTS-c
increases the insulin-stimulated glucose disposal rate, an indicator of enhanced skeletal muscle
insulin sensitivity, but does not alter the rate of hepatic glucose production (72). Insulin-
mediated AKT signaling is elevated in the muscle isolated from MOTS-c injected C57BL/6 mice
and differentiated L6 rat myotubes overexpressing MOTS-c have accelerated glucose uptake,
and enhanced glucose-stimulated and maximum glycolytic rate. The role of MOTS-c on
enhancing insulin sensitivity and glucose homeostasis has also been examined in HFD fed CD-1
mice. MOTS-c treated mice showed reduced weight gain in HFD-fed mice but did not show any
difference in food intake. This result suggests that MOTS-c may increase the metabolic rate of
these mice and experiments using metabolic cages found that HFD-fed mice treated with MOTS-
c showed increased respiratory exchange ratio, reflecting increased glucose utilization. MOTS-c
treated mice also generated significantly more heat that may also partially account for the
increased energy expenditure. Hepatic lipid accumulation was dramatically reduced in HFD-fed
mice treated with MOTS-c and MOTS-c prevented HFD-induced hyperinsulinemia, indicating
improved glucose homeostasis. Furthermore, MOTS-c promoted AMPK activation and GLUT4
expression in the skeletal muscles of HFD-fed mice.

34
Humanin's physiological effects are well established and changes in glucose utilization and
insulin sensitization have been found. Hypothalamic signaling is central to these effects as is
STAT3 signaling. In contrast, the physiological effects of MOTS-c and the SHLPs have yet to be
thoroughly established. While there are hints of a mechanism, much more research will be
required to discover the signaling pathways activated by these MDPs.
Humanin and IGF-I
GH/IGF-I signaling has been well characterized in the aging process (159). Emerging evidence
suggests that mitochondrial factors including mtDNA and mitochondrial derived peptides could
also be a key factor in regulation of aging and aging-related diseases. Although the regulation of
endogenous humanin is still unclear it has been reported that circulating humanin level declines
with age (136). A recent paper explored possible interactions of the GH/IGF-I signaling pathway
and humanin in genetically modified mouse models in the GH/IGF-I axis and human (160).
These mice models have a unique GH and IGF-I profile, and their corresponding humanin levels
suggest that IGF-I negatively regulates humanin level in the plasma, but not GH per se. GH-
transgenic mice (GH-Tg), which showed a 70% reduction of plasma humanin levels, also had
increased levels of GH and IGF-I and exhibit increased body size, various symptoms of
premature aging, and a shortened lifespan by more than 50% (161). Ames dwarf (Prop-1
df
) mice,
which showed a 40% increase in plasma level of humanin, have non-detectable level of GH and
IGF-I and exhibit decreased body size, delayed symptoms of aging, and increased lifespan (162).
Liver-specific IGF-I deficient (LID) mice, which had 45% increase in plasma humanin level,
show elevated levels of GH and 75% reduction of IGF-I and exhibit normal growth and lifespan
(163). IGFBP3 knockout (BP3KO) mice, which showed a 70% decrease in plasma humanin
levels, exhibit no change in GH and no difference in body weight but show reduced lifespan and

35
enhanced IGF-I activity due to an increase in the free IGF-I (164). The direct treatment of rhGH
and rhIGF-I in male C57BL/6 mice showed reduction of humanin level in plasma, further
suggesting that GH inhibits humanin level via IGF-I (160). Plasma sample from GH-deficient
children who were being evaluated for their short stature showed that humanin levels were
negatively correlated to IGF-I levels (160). Plasma sample from an Ecuadorian corhort with GH
receptor deficiency, very low levels of IGF-I, and extreme short stature showed that humanin
levels were 80% elevated compared to normal matched Ecuadorian relatives (165). These results
suggest that humanin levels are directly down-regulated by IGF-I. Notably, humanin and IGF-I
levels simultaneously decrease with age. This could be due to the age-dependent accumulation of
mitochondrial DNA damages as well as the impairment of mitochondria quality control in mice
and humans (166). The regulation of humanin in response to IGF-I in vitro is different from in
vivo experiment. According to a study, rat humanin level measured in the testis, the endogenous
humanin level increased in response to GH and IGF-I in cultured Leydig cells. Humanin
promoted the survival of Leydig cells in culture and enhanced the rate of steroidogenesis by
Leydig cells cooperatively with IGF-I (167). Humanin's effect on steroidogenesis may be due to
the increased survival of IGF-I responding cells in the culture rather than an enhanced rate of
steroidogenesis. Differential receptor expression patterns of these aged cells may also explain the
differential action of humanin on these cells, but further studies are required to clarify this point
of view.

36

Figure 1-6 Schematic Diagram of GH, IGF-I, and Humanin Regulation
GHRH stimulates pituitary GH release which stimulates IGF-I production mainly in the liver.
IGF-I forms negative feedback loop to GH release by stimulating SRIF, which inhibits pituitary
GH release, and by inhibiting pituitary GH release directly. IGF-I suppresses humanin level in
the plasma, and vice versa humanin suppresses circulating IGF-I level. Both IGF-I and humanin
binds to IGFBP-3. Somatotropin release-inhibiting factor (SRIF); GH-releasing hormone
(GHRH).
The effects of humanin analogues on plasma IGFBP3 and IGF-I levels were examined in mice
(168). According to the study, HNG decreased the plasma IGFBP3 levels over time, decreasing
to 80% of baseline after 3hrs. In contrast, an IGFBP3-nonbinding analogue, HNGF6A, did not
change the plasma IGFBP3 levels. Similarly, plasma IGF-I levels were differentially regulated
by HNG and HNGF6A. IGF-I levels decreased approximately 30% when HNG was
administered whereas they remained unchanged by HNGF6A in wild type mice and also
remained unchanged when HNG was injected in IGFBP3 KO mice. These results support that
humanin’s binding affinity to IGFBP3 leads to the enhanced clearance and reduced circulating
levels of IGF-I and IGFBP3. As humanin does not inhibit the binding affinity between IGF-I and

37
IGFBP-3, Humanin’s binding to IGFBP3 may alter the stability of the ternary complexes
(168,169). Furthermore, a study demonstrating that humanin enhances cyclophosphamide (CP)-
induced suppression of lung melanoma metastases showed that administration of HNG with CP
suppressed plasma IGF-I levels compared to CP alone (170). These results suggest that the
reduction of IGF-I mediated by humanin may cooperatively enhance the tumor suppressive
effect of CP.
As the new field of mitochondrial derived peptides emerges, the role of these peptides as
signaling molecules is clearer. Humanin’s overlapping function with IGF-I as well as its apparent
control by and control of IGF-I/GH suggests that humanin is a new player in IGF-I/GH
signaling. Although the exact mechanism by which this occurs is still under investigation, future
studies will need to determine these details.

38
Mitochondria Biology and Prostate Cancer Ethnic Disparity
Introduction
Prostate cancer is the second most diagnosed cancer and the fifth-leading cause of cancer death
in men worldwide (171). The treatment and subsequent monitoring of prostate cancer patients
place a substantial burden on the health care system (172). Racial disparities of prostate cancer
have long been recognized, but their underlying causes need further research. Although
socioeconomic factors could account for such differences to a certain extent, it is now
increasingly accepted that such disparity may have genetic and molecular basis (173). In the
United States, black men have the highest incidence and mortality rates from prostate cancer,
followed by white and Hispanic men, with Asians and Pacific islanders having the lowest rates.
Furthermore, black men are often diagnosed with more advanced and aggressive prostate cancer
compared to any other racial/ethnic group (174). As for prostate cancer mortality, the highest
death rate was observed in the Caribbean, whereas the lowest rates were reported in most regions
of Asia (175). Although the geographical variation in prostate cancer epidemiology can be
partially explained by accessibility of screening tools and medical care, it is also influenced by
genetic, lifestyle and other environmental factors (176). The molecular pathology of prostate
cancer is complex. Besides somatic mutations and chromosomal abnormalities which cause
dysregulation of oncogenes and tumor repressor genes, changes in expression of growth factors
and their receptors are involved in prostate cancer pathogenesis (177). In addition, metabolic
syndrome including obesity and diabetes can also act modify for prostate cancer risk (178).
Understanding how each of these different factors influence disease incidence and mortality is
key to improving the current prevention, detection and treatment strategies.

39
With substantial research conducted on nuclear encoded pathways, an emerging body of work on
prostate cancer suggests a role for mitochondrion in this disease. Mitochondria are known as the
“power plant” of the cell, and play a central role in many cellular functions (179). Mitochondrial
functions are closely associated with aging and cancer progression. Being one of the nine
hallmarks of aging, mitochondrial function has long been recognized to decline with age (180).
Moreover, animal models with disrupted mtDNA integrity suggest that mtDNA mutations
promote aging and lead to respiratory dysfunction (181). The first line of evidence supporting a
link between cancer and mitochondria is the “Warburg effect”, where cancer cells rely heavily
on aerobic glycolysis instead of oxidative phosphorylation to generate energy (36). The mutation
rate of mtDNA is higher than that of nuclear DNA and this high mutation rate results in part
from the mtDNA’s lack of protective histones and an inefficient DNA repair system (40).
mtDNA mutations lead to altered protein transcription/translation, which causes electron
transport chain dysfunctions and increases the production of ROS. As a result, ROS exacerbates
protein and mtDNA damage – a “vicious cycle” (41). Both large-deletion and point mutation in
mtDNA have been found in prostate cancer. Point mutations in mtDNA might influence the
activity of certain proteins encoded in mtDNA, depending on the site of these mutations. If the
mutation is in a tRNA or an rRNA, it may affect the activity of multiple mitochondrial encoded
proteins. If the mutation is in the control region, it may affect replication and transcription.
mtDNA is a very compact genome, but the mito-sORFs concept adds additional complexity to
the information it contains. In fact, it is now recognized that the mitochondrial genome is able to
encode more than 13 proteins, as additional peptides can be derived from mitochondrial sORFs
existing as “nested genes” (109). The first MDP discovered was humanin, which was identified
as an important signaling molecule (182,183) involved in protection from neurodegeneration

40
(184), atherosclerosis (133), and diabetes (105) as well as chemotherapy side effects in cancer
models (185). More members of the MDP family, including MOTS-c and SHLPs were later
characterized as exercise mimetics and insulin sensitizers (72,186). Therefore, mutations in
mtDNA may not only affect the OXPHOS system but also impair normal cellular homeostasis by
changing MDP structure or expression. Cells with a high proportion of large-deletion mutant
mtDNA might have characteristics similar to those with depleted mtDNA. Experimentally-
derived Rho-zero cells are highly resistant to apoptosis (42), and some of them display invasive
phenotypes (43). These observations suggest that dysregulated mitochondrial function as well as
mtDNA are critical for cancer cells in the process of acquiring apoptosis resistance and
invasiveness. Taken together, increases in mtDNA mutations together with the depletion of
normal mtDNA have been shown to reduce the levels of some or all of the 13 enzymes required
for OXPHOS, MDPs important for retrograde signaling, and might affect mitochondrial
bioenergetics, redox regulation of cells, and Ca
2+
homeostasis.
A number of molecular factors specific to tumor cells, including genetic and epigenetic
modifications were found to be associated with prostate cancer racial disparities. Such genetic
alterations include single nucleotide polymorphisms (SNPs) as well as insertion/deletions
(INDELs). Epigenetic alterations include histone modification and DNA methylation, and
overexpression and/or suppression of miRNAs (173) leading to altered tumor microenvironment
(187). Chronic inflammation and oxidative stress will trigger biological responses that favor
tumor invasion and create a supporting tissue environment for prostate cancer (188). As
described earlier, mtDNA mutations and mitochondria integrity are directly related to prostate
cancer pathology. It is thus reasonable to speculate that prostate cancer health disparities may
have a mitochondrial component. In contrast to the nuclear genome, the mitochondrial genome

41
demonstrates considerable geographic diversity and ethnic specificity, and therefore serves as an
important check point for examining race-specific risk (189). Moreover, the compactness of
mitochondrial genome enables relatively cheaper and more rapid identification of prostate
cancer-associated mutations and epigenetic modifications. Mitochondrial genetics, energy
production/fuel utilization, apoptosis, and retrograde signaling all potentially contribute to
prostate cancer etiology and perhaps disparities and will be further discussed in detail below.
Mitochondrial-Related Genetic Alterations in Prostate Cancer
Structurally, the mtDNA is composed of a coding region packed with protein encoding genes,
ribosomal RNAs and transfer RNAs, and a non-coding displacement loop (D-loop) (2). The
naked mtDNA is not protected by packaging proteins or the surveillance of efficient repair
mechanisms (190). In addition, it is in close proximity to high levels of oxidative stress. Thus,
mtDNA somatically mutates at a much higher rate than nuclear DNA, with the D-loop being the
most polymorphic region (191–193). Both somatic mutations and active mutagenesis of mtDNA
in neoplastic lesions have been observed in prostate cancer. Since each mitochondrion contains 3
to 10 copies of mtDNA and each cell has hundreds of mitochondria, mutations in mtDNA may
result in the presence of more than one version of mitochondrial genome within a cell or
individual, a state called heteroplasmy (194). These mutations will either lead to inhibition of
electron transport chain, increase in mitochondrial ROS release, or create a more tumorigenic
microenvironment. The mitochondrial D-loop is a “hotspot” for mutagenesis, and has two
hypervariable regions (195). The origin of replication and two origins of transcription of mtDNA
reside in the D-loop, hence mutations in this control region could have adverse impact on the
integrity of mtDNA and expression of mitochondrial proteins and peptides. Genetic alterations in
the D-loop have been reported in several prostate cancer studies. Analysis of mitochondrial D-

42
loop sequences were done using mtDNA isolated from pure populations of prostate cancer cells,
benign epithelial gland cells, and prostatic intraepithelial neoplasia (PIN) cells (where available)
(196,197). The specific cell populations were obtained by using laser capture microdissection
(LCM) from 16 prostatectomy specimens. The mitochondrial sequence from benign epithelial
gland cells from the same individual was used as reference. In 90% of the tumors and/or PIN
samples, heteroplasmic and homoplasmic somatic mutations were detected in 34 nucleotide
positions, 30 of which were substitutions and four of which were INDELs. Two D-loop
regions—a mononucleotide sequence repeat, and a dinucleotide microsatellite segment—were
more susceptible to INDELs. However, the mutations were not correlated with tumor grade or
age of the patients. Apart from prostate cancer, other cancer types such as cervical carcinoma
and colorectal cancer also reported high frequency of D-loop alterations and mitochondrial
microsatellite instability (192,198).
Multiple SNPs and INDELs were also found in the coding region of cancer patients’ mtDNA and
are dramatically overrepresented in prostate cancer cases. The coding region of the
mitochondrial genome is also susceptible to mutagenesis, and such mutations might be important
for prostate cancer risk assessment. For example, mtDNA G10398A polymorphism has been
associated with invasive breast cancer in black women and with a higher risk of prostate cancer
in black men (199–201). Cybrid cells harboring this specific mutation showed elevated ROS
production, increased apoptosis resistance and tumorigenicity. G10398A resides in the gene
encoding NADH dehydrogenase 3 (ND3), a core component of mitochondrial complex I (199).
The G to A polymorphism in ND3 causes the substitution of alanine by threonine in amino acid
114 in the ND3 polypeptide. Oddly, this substitution was accompanied by increased level of
20kD complex I subunit and enhanced complex I activity. With an increase in complex I activity

43
and level of complex III activity remained unchanged, the electron transport chain may be in a
more reduced state, thereby increasing the retention time of electrons and providing more
opportunity for univalent reduction of oxygen, as complexes I and III are the major sources of
ROS in cells. Further experiments proved G10398A cybrids were able to evade etoposide-
induced apoptosis and demonstrated greater ability to metastasize. The potential mechanism for
this is that mtDNA mutation-induced OXPHOS-activity-decline will stimulate ROS release,
consequently, the elevated oxidative stress will change tumor microenvironment and confer
invasiveness. Another independent study also corroborated the mtDNA mutation and ROS
observation. The pathogenic mtDNA mutation (np 8993G) was introduced into PC3 cells by
cybrid formation. Wild-type and mutant cells were injected into nude mice and allowed to grow.
It was found that after 110 days the tumor volume was more than seven times greater in those
animals receiving mutant cells compared with wild-type cells (202). This finding demonstrates
the potential in vivo relevance of ROS in prostate tumor growth. Several studies have shown that
the mutations in the mitochondrial cytochrome c oxidase subunit 1 (COI) were correlated with
prostate cancer. A comprehensive population-based study of mtDNA mutations in prostate
cancer was reported by Petros et al., 2005. In this study, the entire mitochondrial genome of one
individual tumor was first sequenced. In total, 38 substitution mutations were seen, 31 of which
were reported polymorphisms and the rest were new mutations. The analysis of 260 prostate
cancer patients revealed that 31 (12%) had an inherited mutation in COI compared to <2% of
non-cancer controls. They also analyzed a population sample of 1019 European and African
mtDNA sequences, and the frequency of COI mutations was found to be 7.8%, which was still
lower than the 12% observed in the prostate cancer cohort. The over-representation of COI
polymorphisms in prostate cancer cases suggests that mutations in the COI gene might be a risk

44
factor for developing prostate cancer (202). Given that COI polymorphisms are more frequent in
African mtDNA than in the rest of the world, these SNPs might be genetic determinants for
prostate cancer racial disparities. The importance of this mutation analysis is the large sample
size, which permitted an epidemiological study of the mtDNA mutations. However, in vitro or in
vivo experiments have not yet been conducted and thus mechanisms could not be easily
identified. The presence of somatic mutations in mitochondrial-tRNA were also found to be
associated with elevated PSA levels in prostate cancer patients (203). Mutations in tRNAs can
disrupt mitochondrial gene processing, aminoacylation, and translation. The accumulation of
nonfunctional mitochondrial proteins could affect mitochondrial integrity and metabolic fitness,
leading to increased leaking of PSA into the blood circulation. All the genetic variations
mentioned above are summarized in Table 1-2.
Table 1-2 Known mtDNA Mutations Associated with Prostate Cancer
Mitochondrial-
Encoded Proteins
Functions Modifications in PCa Ref
D-loop Replication/transcription origins
Substitutions and
INDELs
(193,19
4)
MT-ND3 NADH Dehydrogenase 3 Substitutions
(195,19
6,204)
MT-ATP6
ATP Synthase, translocation of
protons and generation of energy
Substitutions (199)
MT-COI Cytochrome c oxidase I
Substitutions and
upregulated expression
(199,20
5)
MT-tRNAs
Ribosomal functions,
Protein translation
Substitutions (200)

45
Although the mitochondrial coding region is intronless, sORF can exist as “genes within genes”.
Previously, silent mutations in the mitochondrial genome were often considered as harmless, as
they do not change amino acid composition of the 13 proteins. Nowadays, with the concept of
small ORF-derived peptides and MDPs fully recognized, these silent mutations may no longer be
“silent”. They could change MDP structure or even cause truncation of peptides. Humanin is the
first MDP identified, and it is encoded within a sORF in the 16S rRNA gene. This 24-amino acid
peptide exerts cytoprotective and metabolon-enhancing activity by interacting with IGFBP-3,
BAX/BID directly or through receptor-mediated signaling (77,78). The SNP G2706A is one of
the two ancestral mitochondrial mutations that define mitochondrial Haplogroup H. This
polymorphism sits in the 16S rRNA and the stop codon of humanin, therefore it might change
the shape of the RNA of the mitochondrial ribosome, and the protective peptide humanin. The
minor allele (G:G) is predominantly found in Europe, while A:A is found in Asia and Africa
(213). The racially disparate distribution of this SNP might provide a molecular mechanism for
prostate cancer ethnic disparity. The A1382C polymorphism in the MOTS-c coding region is
Nuclear-encoded
Mitochondrial
Proteins

GSTP1 Antioxidant mechanism Hypermethylation (206)
ZIP2 Zinc transporter
Downregulated
expression
(207,20
8)
SOD2 Antioxidant mechanism Substitutions (209)
HSP60
Mitochondrial chaperon, facilitates
proper protein-folding
Downregulated
expression
(210)
POLG Mitochondrial DNA polymerase Hypermethylation (211)
PPARGC1A Mitochondrial biogenesis Hypermethylation (212)

46
specific to Northeast Asian population, and might be a putative explanation for exceptional
longevity of Japanese people (104). Although it might not be directly linked with prostate cancer,
it is associated with aging, which is the greatest risk factor for cancer. In a newly reported study,
we showed that low circulating levels of the MDP SHLP2 represent a novel biomarker for
prostate cancer risk (214). This study also found that the SHLP2 levels in white men were higher
than those of black men, which suggests a possible molecular link between prostate cancer risk
and ethnicity. SHLP2 is mitochondrially encoded and also regulates mitochondrial function, and
therefore low SHLP2 levels might represent mitochondrial dysfunction. Alternatively, SHLP2
may reduce prostate cancer risk by acting as an insulin sensitizer which it has been demonstrated
to be in mouse models.
Numerous studies have focused on the overall mutational load of mitochondria and
mitochondrial haplogroups. Next-generation sequencing of mtDNA from prostate tissue biopsies
and matched blood of 115 men revealed a positive correlation between the total burden of
acquired mtDNA variation and elevated Gleason Score at diagnosis and biochemical relapse
(215). Two additional studies also proved that the accumulation of mutations over the entire
mitochondrial genome contributes to prostate cancer progression (216). However, whether there
is difference in mitochondrial mutational load between ethnic groups is not known.
Mitochondrial haplogroups are used to trace the matrilineal inheritance and represent the major
phylogenetic branches. Previous studies have suggested that inherited mitochondrial genome
variation, which defines population-specific haplogroups, may be associated with prostate cancer
risk. Specifically, a study of 221 White North American men showed increased risk in
haplogroup U for developing prostate cancer [odds ratio 1.95] (217). Whereas in two other
independent studies, common European-derived mtDNA haplogroups are not correlated with

47
increased prostate cancer risk or specimen Gleason Score (205). In addition, no association
between mitochondrial haplogroups and prostate cancer was found in a Korean population or a
Colombian population (218,219). Apart from population studies, a number of groups have used
cybrid models to support the concept that mtDNA haplogroups can influence nuclear gene
expression, rates of cell growth and cell behavior. The cellular energetics and gene expression
profile were compared between European H cybrids and the African L cybrids in one study
(213). It was found that mtDNA haplogroup variants can greatly influence the efficiency of
respiration, irrespective the nuclei. The African L cybrids exhibited lower mtDNA copy number
but higher mitochondrial gene expression. The L cybrids also have lower ATP turnover rates and
lower spare respiratory capacity, which suggests that they may not be able to respond to stress as
readily as the H cybrids. The potential association between the African-derived mtDNA
haplogroups and aggressive prostate cancer remains to be confirmed, but the cybrid study may
provide insight into the mitochondrial determinants of prostate cancer health disparity.
One limitation of current research on mtDNA mutations and prostate cancer is the lack of
comprehensive study of the association between mtDNA mutations and tumor grade. Among the
previously described studies, most of them fail to mention the association between mtDNA
mutations and Gleason grade. As black men are often diagnosed with more aggressive prostate
cancer, it would be ideal to identify a mtDNA mutation that is linked with Gleason grade. Most
of the reported mtDNA alterations in prostate cancer to date only involve individuals with
Gleason grades 5–7, the most frequent clinical grade at presentation. Therefore, future studies
need to include more subjects with lower grades to ascertain whether there is an association
between specific mitochondrial mutations or mutation load and Gleason grade or tumor stage.

48
Intratumor heterogeneity and tumor evolution study by the help of single-cell sequencing will
also be an invaluable source of information.
Mitochondrial Content and Proteomic Changes in Prostate Cancer
Mitochondrial genetics is not the only potential molecular determinant of mitochondrial related
prostate cancer health disparity. As mentioned earlier, tumor growth is accompanied by cellular
adaptations to hypoxic environment, cell cycle control and excessive proliferation. Mitochondria
are key organelles for energy production with a critical role in cell survival and apoptosis. Thus,
transformation of normal cells to cancer cells usually involves modulation of mitochondrial
protein expression. Mitochondrial proteins consist of mitochondrially translated proteins and
nuclear encoded mitochondrial enzymes. As for the 13 polypeptides and MDPs, their expression
is not only affected by mtDNA mutations, but also by the changes in mitochondrial content and
transcription/translation machinery. It has been reported that mitochondrial biogenesis and
quantity of mitochondria varied between control and prostate cancer patients. Grupp et al., 2013
found that the mitochondrial content was tightly linked to pathological features of prostate
cancer (220). A tissue microarray (TMA) of 11,152 prostate cancer specimens were analyzed in
order to evaluate the clinical significance of mitochondria abundance. The mitochondrial content
was determined by staining with anti-MTCOII antibody. The results demonstrated marked
increase of mitochondrial content in cancer cells when compared with normal prostate epithelial
cells. Moreover, higher numbers of mitochondria were observed with increasing tumor grade and
stage, which suggests that increase in mitochondrial content is necessary or supportive for cancer
development and progression. The literature also suggests that increase in mtDNA copy number
and mitochondrial abundance is required for cancer cell proliferation, and prostate cancer
progression is accompanied by the activation of core mitochondrial biogenesis regulators (221).

49
These findings may seem contradictory to the mitochondrial damage theory of prostate cancer,
which claims that mitochondrial function decline is a cause of prostate cancer. Nevertheless,
there is a possible hypothesis for this discrepancy. Normal prostate epithelial cells accumulate
mtDNA damage over time, which may lead to loss of mitochondrial integrity and transformation
into cancerous phenotypes. As a result, mitochondrial biogenesis pathways are activated as a
compensatory mechanism, thereby increasing the quantity of dysfunctional mitochondria and
facilitating cancer cell proliferation. In fact, the upregulation of mtDNA copy number and
mitochondrial gene expression in response to DNA damage has been reported in various cancer
cases (222). On the contrary, there is evidence supporting a negative association between
mitochondrial content and cancer risk. It has been reported that the reduced mtDNA content has
been associated with poor prognosis of prostate cancer in black men (223). Biswas et al., 1999
reported that decreased mtDNA content could activate the NF-kB/Rel factors (207). Activation
of NF-kB signaling confers apoptosis resistance and plays pivotal role in cancer malignant
transformation. In addition, activation of the AKT pathway by mtDNA deficiency could also
inhibit cell apoptosis (224). These results indicate that mitochondrial content and quality have to
be maintained at an optimal level. Higher and lower levels of mitochondrial mass might be a risk
factor for prostate cancer. As mentioned earlier, different mitochondrial haplogroups
demonstrated different levels of mtDNA copy number, which suggests that the difference in
mtDNA level between ethnic groups may be a determinant of prostate cancer risk or grade.
The prostate glandular epithelial cells experience drastic metabolic transformation during
tumorigenesis, and the transformation must involve biochemical reactions occurred in the
mitochondria and regulation at the proteomic level. In other words, proteomic changes could
facilitate prostate cancer progression. About 80% of prostate malignancies occur in the

50
peripheral zone of the prostate gland, where the cells are highly specialized secretory cells.
Different from cells in the central zone and the rest of the body, peripheral zone cells accumulate
substantial amounts of zinc, due to high activity of the zinc uptake transporters (208). The
presence of high zinc concentrations in the mitochondria of the peripheral zone cells inhibits the
activity of the enzyme mitochondrial aconitase, which prevents the oxidation of citrate through
the Krebs cycle. In contrast, malignant cells exhibit marked decrease in zinc and citrate levels.
The reduction in mitochondrial zinc levels renders prostate cells energy efficient, as the
mitochondrial aconitase is no longer inhibited and production of ATP resumes. Moreover, the
low levels of zinc lead to elimination of the pro-apoptotic effect of zinc. Together with the
increased energy production, the suppression of apoptosis promotes proliferation of malignant
cells. The zinc homeostasis is tightly regulated by zinc transporters. Out of all the zinc
transporters, Zrt- and Irt-like protein-1 (ZIP1) was shown to be the major zinc uptake transporter.
Furthermore, it was shown that zinc levels and ZIP1 expression were strongly correlated in the
prostate. During prostate cancer development, both the decrease of ZIP1 expression and zinc
depletion occur at early stages of the disease (225). Recent studies discovered that ZIP2 was
abundantly present in normal prostate epithelial cells and in benign prostate hyperplasia, while
its expression was downregulated in prostate cancer tumor cells (226). Furthermore, it was found
that ZIP2 was significantly lower in black patients, who generally have higher risk to develop
prostate cancer compared to age-matched white individuals. Consistent with this notion, ZIP2
levels were found to be more reduced in tissue samples from black men compared to Gleason
grade-matched samples from white males (227).
An alternative hypothesis for zinc depletion induced malignant transformation is that normal
prostate glandular epithelial cells have truncated Krebs cycle and low levels of ROS. These cells

51
may have low oxidative stress tolerance when they encounter this metabolic switch and start to
produce more ROS. The mitochondrial antioxidant pathways and enzymes are therefore
important for preventing “the vicious cycle” from progressing. Kang et al., 2007 studied the
association between genetic polymorphisms in three isoforms of superoxide dismutase (SOD)
and prostate cancer (209). Supposedly, dysfunctional antioxidant enzymes are most likely to
exacerbate mitochondrial damage and promote cancer development. However, it was found that
the Ala variant at SOD2 Ex2+24T>C(V16A), despite having higher superoxide scavenging
activity, was associated with elevated prostate cancer risk in Caucasians. This finding is
incompatible with the original hypothesis. One possible explanation is that higher SOD catalytic
activity results in more superoxide to H2O2 conversions. H2O2 accumulation in the cell may also
induce DNA damage and hence stimulate carcinogenesis. Stratification by quartiles of dietary
and supplemental vitamin E intake demonstrated that smoking and vitamin E intake were
modifiers for prostate cancer risk among SOD2 Ala variant carriers; Ala variant carriers who are
smokers and have low vitamin E intake are associated with moderately increased risk of prostate
cancer. No significant association with prostate cancer was observed for polymorphic variants
in SOD3 or SOD1.
Mitochondrial heat-shock proteins (HSPs) and proteases are known to be upregulated to maintain
protein folding homeostasis and mitochondrial function. This is mediated by UPR
mt
, a stress
response induced by perturbations in the protein-folding environment. It has been reported that
the UPR
mt
is different between cells derived from white and black men with prostate cancer. In
fact, gene expression of UPR
mt
chaperone proteins HSP60 and DNAj was elevated in cells from
Caucasian compared with cells from African (210). Transcriptional levels of HSP70 and HSP90
were unchanged. Since HSP60 is an essential chaperon protein for the translocation and proper

52
refolding of proteins from the cytosol to mitochondria, deficiency in HSP60 in cells from
African may disrupt the UPR
mt
and mitochondria restoration. Far cancer cells, UPR
mt
can send
pro-apoptotic signal via MAPK activation (228). As a result, the patients of African descent may
have defective apoptosis pathway and more aggressive tumorigenesis. Thus, the differential
features of UPR
mt
might be able to explain racial difference in prostate cancer.
Epigenetics and Mitochondria in Prostate Cancer
DNA methylation is the addition of a methyl group to the cytosine within cytosine guanine
dinucleotides (CpGs). DNA methylation at multiple adjacent CpG sites on promoter region
ultimately causes gene silencing by blocking the access of transcriptional factors and/or
activators to the target sites. MtDNA methylation has been a controversial topic. Bisulfite
sequencing analysis and use of methylation sensitive and insensitive restriction enzymes have
been implemented to study mitochondrial epigenetics. One aspect of the argument against the
existence of mtDNA methylation has been the absence of the indispensable factors for carrying
out this process. However, over the last few years, the identification of methyl donors,
methyltransferases, and ten-eleven translocation methylcytosine dioxygenase 1 (TET1) was
reported in mitochondria (229–231). Recently there have been stronger evidence indicating the
existence of epigenetic modifications of mitochondrial genome. For instance, one study used
methylated DNA immunoprecipitation datasets from the NIH Roadmap Epigenomics project to
examine the mitochondrial methylation in a wide range of tissue and cell types, and found
distinct but conserved methylation patterns (232). Another study compared the mtDNA
methylation of normal mesenchymal stem cells (MSCs) with that of senescent MSCs using the
combined bisulfite restriction analysis (COBRA). They reported several CpG sites exhibiting
differential methylation percentages when cells undergo senescence (233). Currently,

53
comprehensive study on mitochondrial DNA methylation in prostate cancer etiology is missing.
Given the importance of mitochondria in prostate cancer, this is an aspect worth in-depth
investigation.
Importantly, most mitochondria proteins are nuclear encoded proteins. Thus, epigenetic
modifications on promoter regions of these nuclear-encoded mitochondrial genes are likely to
alter mitochondrial behavior. One study screened 899 mitochondrial-acting nuclear genes and
identified 636 genes exhibiting tissue-specific and differentially methylated patterns (206). This
study further proved that the tissue-dependent mitochondrial functions were regulated by the
methylation status of nuclear mito-genes. Moreover, another study has concluded that DNA
methylation levels of the CpG islands in nuclear encoded mitochondrial genes were lower than
that of non-mitochondrial genes (234). The latter finding was then cross-validated by using the
data from the HumanMethylation450 BeadChip containing probes that lay within 1000bp of
transcription start sites. Most probes (>5000) at the transcription start sites of mitochondrial
proteins showed hypomethylation, while 400 probes showed hypermethylation (235). Abundant
evidence has accumulated to suggest that epigenetic alterations are significantly different in the
prostate tumors of black and white men. Of all the genes examined in prostate tumors so far,
glutathione S-transferase Pi 1(GSTP1) is a mitochondrial antioxidant enzyme (236). Comparison
of GSTP1 methylation status with their clinical and pathological outcomes showed that black
men carrying GSTP1 hypermethylation were 13.3 times more likely to have prostate cancer,
whereas in white men, this ratio was only 3.8 (237). This study suggested that the methylation
status of GSTP1 might serve as a putative biomarker for prostate cancer diagnostics in African-
descent populations. Apart from epigenetic regulation of mitochondrial enzymes, methylation of
mitochondrial proteins responsible for mtDNA replication has also been observed. POLG is

54
involved in the replication of mtDNA and has proofreading functions. It was reported that POLG
expression was regulated by the methylation of the POLG gene transcription start site. As an
essential gene dictating mtDNA replication, the POLG methylation levels were found to be
negatively associated with mtDNA copy number (211). Methylation of the PPARGC1A gene,
which is the master regulator of mitochondrial biogenesis, has also been correlated negatively
with mtDNA copy number (212). The rationale is relatively straightforward; inhibition of
mitochondrial biogenesis via methylation of PPARGC1A will result in suppression of genes
promoting mitochondria fission and mtDNA replication, thereby reducing mtDNA copy number.
Moreover, evidence gathered from cybrid experiments have revealed that the mitochondria can
affect nuclear methylation patterns (238). Given the importance of epigenetics to prostate cancer
regulation, and the uniqueness of mitochondria in this disease, more attention and endeavor on
this topic are needed to decipher molecular basis of prostate cancer and its ethnic disparity.
Environmental Factors in Prostate Cancer
The most important environmental factors to consider are diet and obesity, as diet is highly
variable across ethnicities, and obesity has been consistently associated with an increased risk of
prostate cancer aggressiveness and mortality. Among the most obvious characteristics of the
Western diet are intake of high caloric and fatty foods. Several epidemiological, interventional,
and animal model studies have not only indicated strong correlations between fat consumption
and the rate of prostate cancer mortality (239,240), but also proved that low-fat diet can slow the
rate of LAPC-4 prostate cancer xenograft growth (241). Saturated fat, which is largely fat from
animal sources, has been correlated strongly with prostate cancer mortality and increased
aggressiveness, although no statistically significant racial differences were observed (242). High
intake of dietary fat can contribute to prostate cancer in several different ways, such as

55
influencing androgen signaling, the IGF pathway and cell proliferation (243). Although
molecular explanation of how fat intake can cause alterations in mitochondria is not fully
understood, it is speculated that the fatty acids metabolism plays a critical role (244). The
mitochondria are where fatty acids are broken down by beta-oxidation and enter Krebs cycle.
The difference in richness of saturated fat in diet may place differential levels of burden on
mitochondria. The high levels of fat content may cause mitochondrial damage and decline,
which later contributes to prostate cancer initiation. Each ethnic group has different diet
preference, thereby possessing different levels of prostate cancer risk.
Exercise is also a key determinant of cancer risk and can affect mitochondrial function
(245,246). Physical activity has been shown to improve patient survival after diagnosis and is
linked with reduced risks of cancer. As for prostate cancer, limited evidence has been provided.
One study evaluated physical activity of patients suffering from non-metastatic prostate cancer
and found out that physical activity improved cancer survival as well as overall survival (247).
Another study of 4623 men with localized prostate cancer confirmed the former finding (248).
The underlying molecular pathway involves IGF axis, p53 and mitochondrial apoptosis pathway.
The serum from men with regular exercise was used to culture LNCaP cells and exhibited higher
levels of apoptosis than cells cultured in serum from sedentary controls. The exercise serum
stimulated cells also showed higher expression of p53 and Bcl-2 (249).
Previous studies demonstrated that serum from men consuming a low-fat diet and undergoing
exercise intervention reduced LNCaP cell growth and induced apoptosis in vitro (250). Exercise
also decreased serum IGF-I and increased serum IGF binding protein-1 (IGFBP-1), which
suggests that IGF-I and IGFBP-1 levels are mediated by diet and exercise. In fact, reduction of
IGF-I levels in the circulation has long been proposed as a mechanism of caloric restriction (CR),

56
which is a pro-longevity regimen. The mitochondrial peptide humanin and MOTS-c are CR- and
exercise-mimetic, respectively. Humanin is a new player in the regulation of the GH-IGF-I axis
and interacts with IGFBP-3 to decrease circulating IGF-I levels (251). In other words, decline in
mitochondrial integrity and reduction in humanin levels may disrupt the diet-IGF-I pathway. As
a result, the suppression of IGF-I by humanin will be altered and the resultant elevated IGF-I
levels will promote malignant cell growth (Figure 1-7). The same rationale applies for exercise-
mediated anti-tumorigenesis effects by MOTS-c (72). Information concerning human genetic
polymorphisms in the genetic pathways involved in hormone synthesis, metabolism, and action
is now accumulating very rapidly. Together with the SNPs found in the MDP coding regions,
more comprehensive mitochondrial sequencing studies are required to unravel the roles of MDPs
in prostate cancer.

Figure 1-7 Mitochondria-Related Factors Interact and Contribute to Prostate Cancer

57
CHAPTER 2 HUMANIN AS A DIETARY MIMETIC
Abstract
Dietary regimens including CR and fasting are associated with various beneficial effects on
physiological processes that are relevant to aging. As the most effective and reproducible dietary
intervention, CR delays the onsets of age-related diseases and preserves key biological functions
such as insulin sensitization, inhibition of tumor progression, hence extends healthspan.
Mitochondria play a uniquely important role in aging as its genome is damaged progressively
with age accumulating mtDNA deletions and point mutations leading to a decline in
mitochondrial integrity, which compromises its functions, leading to diabetes and age-related
neurodegenerative disease. Furthermore, humans and mice with mitochondrial genetic diseases
exhibit phenotypes that resemble premature aging.
Mitochondria have been proposed to participate in CR-induced benefits, but the exact
mechanisms involved remain elusive. We have been studying a mitochondrial-derived peptide,
humanin, which has effects on delaying diabetes and Alzheimer’s disease as well as promoting
longevity. Since humanin and CR have many overlapping benefits, we set to determine the
molecular mechanism connecting mitochondria, humanin and dietary intervention. By utilizing
in vitro and in vivo models, we found that humanin treatment suppressed circulating IGF-I while
increasing IGFBP-1, which resembled the effects of CR. Body weights of these mice were
significantly reduced and fasting insulin levels were slightly reduced in humanin-treated mice,
compared with controls. Furthermore, humanin injection was shown to ameliorate side-effects
induced by CP, likely through differential protection via suppression of the insulin and IGF-I
signaling (IIS). Previously, we have shown that growth hormone and IGF-I inhibit humanin
levels in mice and humans, identifying a novel connection between the GH/IGF system and the

58
mitochondria. Humanin also activated AMPK, and enhanced mitochondrial biogenesis in
skeletal muscle, delaying age-associated motor function decline. Subsequent behavioral study
and metabolomic study further suggested the similarities between humanin and CR.
Taken together, these data suggest that humanin exhibits CR-mimetic effects by suppression of
the IIS pathway and activation of AMPK, leading to protection against chemotherapy and
prolonged healthspan. The results indicate that the mitochondrial humanin system and the GH-
IGF pathway have counter-regulatory activities, each inhibiting the levels of the other hormonal
pathway: GH and IGF-I suppress humanin levels while, humanin inhibits insulin & IGF-I and
promotes IGFBP-1 production. All these interrelated and distinct responses define a regulatory
circuitry that links mitochondrial retrograde signaling, GH/IGF-I and aging.

59
Background
Caloric restriction (CR) is a well-established intervention that extends lifespan in various model
organisms ranging from yeast to rodents (252). Despite numerous studies were conducted to
decipher the detailed mechanism, there is no definitive answer. Down regulation of the IIS has
been proposed as an important mechanism underlying CR, in rodents, the GH/IGF-I axis
reduction has been widely accepted as one of the effectors of CR and lifespan extension (253).
Target of rapamycin (TOR) pathway (254), AMPK signaling (255), and Sirtuins (256) also
received much attention from studies trying to link these pathways to the benefits of CR. It is
widely accepted that CR delays the onset of age-related decline in many species and reduces the
incidence of age-related diseases such as cancer, diabetes, cardiovascular disease,
atherosclerosis, and neurological disorders. CR modulates physiology and metabolic
homeostasis, thus affecting animal behavior, metabolic activities, stress response and insulin
sensitivity (257).
Studies on invertebrates such as C. elegans and Drosophila have yielded valuable insights into
the molecular mechanism of aging. The first line of evidence linking longevity and components
of the insulin or insulin-like signaling pathways is from genetic studies on C. elegans. Klass et
al., 1983 identified five mutants with altered lifespan, which were later mapped to a single
genetic locus age-1 (258). The age-1 gene is homologous to the phosphatidylinositol 3-kinase
(PI3K), which can be activated by IGF-I and promotes cell proliferation. Several years after the
identification of age-1, another mutation daf-2 was found to extend lifespan by 100%. Whereas
the phenotypes of age-1 and daf-2 mutants can be suppressed by mutation in daf-16 gene (259).
Subsequent molecular cloning revealed that daf-2 encoded for an IIS receptor, while daf-16
encoded for a forkhead box O (FOXO) transcription factor downstream of the IIS/PI3K pathway

60
(260). Clancy et al., 2001 also found that reduced IIS signaling increased lifespan in Drosophila,
suggesting it might be an evolutionarily conserved mechanism. In this experiment, mutations in
the chico gene (a homolog of insulin receptor substrate (IRS)), Inr (fruit fly equivalent of daf-2)
and Dp110 (fruit fly equivalent of age-1) resulted in long-lived mutants (261). In worms and
flies there is only a single insulin/IGF receptor, while in mammals there are several homologous
receptors. GH and IGF-I are important growth promoting hormones with pleiotropic effects; the
production of IGF-I is stimulated by pituitary GH. Circulating IGF-I is produced mainly by the
liver, but autocrine and paracrine secretion of IGF-I also occurs in many tissues (262). Several
genetic models with suppressed or disrupted GH/IGF-I axis exhibit extended longevity. The
Ames mice and Snell mice both harbor mutations in transcription factors regulating pituitary
development, which lead to combined pituitary hormone deficiency (thyroid stimulating
hormone, prolactin and GH) and increase in lifespan (male, 49%; female, 64%) (263). The Laron
dwarf mice lacking GH receptor/binding protein genes have 90% reduction in serum IGF-I and
display 38% and 55% increase in mean lifespan in females and males, respectively (264). As
observed in genetic models with altered IIS signaling, caloric restricted animals also exhibit
reduced body weight, decline in IGF-I levels and improved insulin sensitivity. Moreover,
microarray analysis of gene expression pattern changes in Ames dwarfs and caloric restricted
mice liver tissue showed considerable overlap between regulated genes (265). Therefore, CR
promotes longevity through sophisticated but poorly understood mechanisms that appear to
overlap partially with those responsible for the effects of IIS signaling on aging. However, CR
further increased lifespan in Ames dwarf mice (266), confirming that the overlap between the
pathways influenced by dietary and IIS-related genetic interventions is only partial, and

61
application of CR and IIS reduction act in addictive fashion via the same downstream
mechanism to fully exploit life extension capacity.
The AMPK is a central nutrient sensor, which is activated upon an increase in the AMP/ATP
ratio (low energy status). When activated by phosphorylation, the phospho-AMPK can enhance
mitochondrial metabolism, inhibit the mTOR signaling and regulate other pathways involved in
longevity. In other words, AMPK activation can be triggered by CR as CR induces energy stress,
which leads to the speculation that AMPK may be another mediator of the beneficial effects of
CR (255). The yeast homolog of AMPK, the sucrose non-fermenting 1 kinase (SNF1), can
translocate into the nucleus to initiate transcriptional activation of non-fermentable carbon
metabolism (267). CR in yeast is often achieved by glucose restriction, and the beneficial effects
of yeast glucose restriction are related to enhanced respiratory function. Therefore, Snf1 is likely
involved in CR and aging. Similar conclusion was also drawn from C. elegans; worms
overexpressing AMPK (aak-2) lived longer than controls, and glucose restriction increased
enzyme activity and mitochondrial respiration (268). In addition, it was reported that the pro-
longevity effect of CR was dependent on aak-2 (269). Studies in Drosophila also demonstrated
that overexpression of liver kinase B1 (LKB1), the upstream kinase for AMPK, elongated
lifespan (270). Furthermore, the tissue specific overexpression of AMPK in muscle and
abdominal fat extended the fly lifespan, and that supplementation of adenosine, which suppresses
AMPK activity, could reverse the beneficial effects of CR (271).
Humanin, and its functional analogs have been extensively studied under different age-related
disease models. The identification of this micropeptide and characterization of its role in neuro-
protection in AD have inspired researchers to study the potential role of humanin and apply as
therapeutics in various neurological and metabolic diseases (83). Humanin and its analogs

62
ameliorated the symptoms of age-related diseases including T2DM (82,272), cardiovascular
diseases (85,273,274), memory loss and neurodegeneration (275,276), stroke (277), and
inflammation (278). The similarities between CR and humanin effects inspired the speculation
that humanin might be the link connecting mitochondrial metabolism, CR and aging (Figure 2-
1). Therefore, we set to investigate the common signaling pathways shared between humanin and
CR and explore the potential CR-mimetic functions of humanin.

Figure 2-1 The Overlapping Effects of Humanin and CR
A number of effects exhibited by humanin treatment are also observed in caloric restricted
animals, including delayed onsets of several age-related diseases, suppression of inflammation,
and so on. However, some CR effects such as lifespan and healthspan extension have not been
studied in humanin-related experiments.
Firstly, we decided to assess the relationship between humanin and GH/IGF-I axis. Although it
was previously demonstrated that the circulating levels of humanin is negatively associated with
the activity of GH/IGF-I, the effects of humanin on IGF-I are not thoroughly studied. CR has
been shown to suppress tumor growth and reduce cancer incidence (279,280). IGF-I promotes
v Protection against
Alzheimer’s disease
v Insulin sensitization
v Cytoprotection
v Anti-inflammation
Effects of Humanin Effects of CR
v Increased healthspan
v Longevity
v Protection against
chemotherapy toxicity
v Enhanced
mitochondrial OCR

63
cell proliferation and tumorigenesis, the mechanism behind the protective effects of CR may
depend on the reduction of this growth factor. Moreover, the supplementation of IGF-I to caloric
restricted tumor-bearing mice abrogated the protective effect of CR on neoplastic progression
(281). Therefore, we wanted to examine whether administration of HNG (a potent humanin
analog) can mimic the effects of CR. We studied the effects of HNG with or without a
chemotherapeutic agent cyclophosphamide (CP) on cancer in mice. In this study, the mouse
metastatic lung melanoma model was used as a prototype of a metastatic cancer. Inoculation of
mouse melanoma cells induced lung metastases that could be easily quantified by counting
melanoma nodules in the lung. We selected CP as a chemotherapeutic agent in this study because
1) CP is common example of an alkylating chemotherapeutic agent used in mouse studies, 2) the
dose and time course of CP on suppression of spermatogenesis is very well defined (282,283), 3)
CP decreases peripheral leukocytes (284), and 4) CP treatment suppresses metastatic lung
melanomas (285). Since humanin has anti-apoptotic effect, which has been hypothesized to be an
“oncopeptide”. We further explored whether this feature of humanin would interfere with CR-
mimetic cancer suppression effects. Our study showed that 1) HNG protected against
chemotherapy-induced toxicity on germ cells and leukocytes in normal or tumor-bearing
immunocompetent mice, and addressed the critical question of whether HNG may prevent the
killing of cancer cells by chemotherapy; 2) HNG by itself modestly suppressed the number of
lung metastases and enhanced the CP-induced cancer suppression; 3) HNG may act as a caloric-
restriction mimetic by suppressing IGF-I and increasing IGFBP-1 levels thus inhibiting the
cancer growth. Coupling the prevention of chemotherapy-induced germ cell loss and reducing
leucopenia, whereas enhancing chemotherapy efficacy renders humanin and its analogues
promising adjuvants to cancer treatment.

64
Later, we conducted a long-term animal experiment in which middle-aged mice (18-month-old)
were subject to twice a week HNG intraperitoneal injection. Since humanin levels decrease with
age, our hypothesis is that the supplementation of HNG in aged mice may elongate their
healthspan, hence lifespan. The mice were allowed to age until their natural death occurred or
terminated at 32-month-old. Body weight and food intake were constantly monitored, and key
metabolic biomarkers including IGF-I, insulin and glucose were measured. From this study, we
confirmed previous conclusion that humanin and IGF-I counteract, and overlapping metabolic
effects of humanin and CR. From the animal behavioral assays, we found that the HNG-treated
mice exhibited delayed the age-related decline in motor function, which indicated that the
skeletal muscle functions were preserved by supplementation of humanin. The molecular
mechanism underlying this effect was later carefully studied and found to involve AMPK
activation.

65
Results
HNG Suppresses Plasma IGF-I and Increases IGFBP-1 Levels, while Enhancing CP-
Induced Tumor Suppression in Tumor-Bearing Mice
In collaboration with the Wang lab, we explored whether HNG (humanin analog) can have
differential chemotherapy protection as fasting in cancer models. The C57BL/6J mouse allograft
models with melanoma were generated in the Wang lab. HNG (P = 0.036) or CP (P = 0.027)
treatment alone significantly suppressed plasma IGF-I levels. Addition of HNG to CP further
suppressed IGF-I (P = 0.02) levels compared with CP alone (Figure 2-2 A). Cotreatment of HNG
with CP increases plasma IGFBP-1 levels when compared with CP alone (P ≤ 0.05) (Figure 2-2
B). CP treatment significantly suppressed the number of white blood cells (WBCs),
granulocytes, monocytes, and lymphocytes compared with tumor-bearing control mice.
Importantly, addition of HNG to CP regimen alleviated CP-induced cell death and restored
WBC, granulocyte and monocyte counts to the levels observed in control mice. Moreover,
treatment with HNG or CP significantly decreased the number of metastatic tumors in the lungs
as compared with control tumor-bearing mice, addition of HNG to CP treatment further
decreased number of tumors compared with CP treatment alone. Refer to (170) for other relevant
data and information.

66

Figure 2-2 Humanin Exerts CR-Mimetic Effects by Suppressing IGF-I and Increasing IGFBP-1
in tumor-bearing mice
Murine IGF-I (A), and murine IGFBP-1 (B) levels in the control (NT), HNG-treated alone
(HNG), CP alone (CP), and combined CP with HNG-treated (CP+HNG) mice (n = 5 mice per
group). Values are mean ± SEM.

Mice were maintained in UCLA Harbor by the Wang lab, plasma were extracted and sent to the
Cohen lab for circulating factor measurements by the Aging Biomarker Core.
To confirm whether the effects can be recapitulated in normal mice, we did a similar study using
2-month-old healthy male C57BL/6J mice and a CR (30% reduction in food intake) group was
added as a positive control. However, the suppression of IGF-I and increase in IGFBP-1 were
not observed in HNG-treated normal mice (Figure 2-3 A, B). When the IGF-I levels in normal
mice and tumor-bearing mice were later compared, we found that the tumor increased circulating
IGF-I levels (Figure 2-3 C).

67

Figure 2-3 Humanin Failed to Alter Circulating IGF-I and IGFBP-1 Levels in Normal Mice
Murine IGF-I (A), and murine IGFBP-1 (B) levels in the control, HNG-treated (HNG), caloric-
restricted (CR) mice (n = 5 mice per group). (C) murine IGF-I levels in normal mice, tumor-
bearing mice treated with or without CP/CP+HNG. Values are mean ± SEM, *p< 0.05, **p <
0.01, ***p<0.001.

Animal experiment was conducted by me, circulating factors were measurements (ELISA) by
the Aging Biomarker Core.
Long-Term HNG Treatment Promotes Visceral Fat and Weight Loss Without an Overall
Reduction in Food Intake
Control
HNG
CR (30%)
n.s.
**
mIGFBP-I (ng/mL)
**
n.s.
Control
HNG
CR (30%)
mIGF-I (ng/mL)
A B
Con: Normal mice
NT: Non-treated tumor-bearing mice
CP: Cyclophosphamide treated
CH0.05: CP + HNG (0.05mg/kg)
CH0.5: CP + HNG (0.5mg/kg)
CH5: CP + HNG (5mg/kg)
mIGF-I (ng/mL)
*** ** ***
*
C

68
We conducted a 14-month study of female C57BL/6 mice starting at 18 months of age using
twice-weekly humanin injections (4mg/kg) to assess the effects on lifespan and healthspan (n =
50). Mice were fed on ad lib diet starting at 18 months of age until their natural death or
sacrifice. HNG was injected intraperitoneally and the control mice were subject to vehicle
(water) injection. Since the start of the experiment, both HNG and control group gradually
increased body weight until 22 months of age (670 days), then experienced continuous decrease.
The body weights of HNG group were lower than that of the controls between 19 and 25 months
(2-way ANOVA, P < 0.0001), then gradually converged (Figure 2-4 B). Interestingly, there is no
difference in food intake throughout the study (Figure 2-4 D). To investigate diet-induced body
composition changes, we evaluated lean body mass and body fat localization by microCT. At 28
months, HNG group mice had reduced visceral adipose tissue compared to control mice (P <
0.01), although subcutaneous adipose tissue percentage was not significantly different
(Supplemental Figure S1). Moreover, the lean body mass of HNG treated mice were higher than
controls (P < 0.01), suggesting this body weight change was primarily due to a loss of visceral
fat. These observations were consistent with the notion that humanin functions as a CR mimetic
in terms of body weight and composition modulation.

69

Figure 2-4 Humanin Promotes Lean Body Mass, Reduces Body Weight and Visceral Fat
Without Changing Food Intake
(A) Visceral fat percentage at 28 months of age (n = 6 per group). (B) Lean body mass at 28
months of age (n = 6 per group). (C) Mouse body weight profile. (D) Food intake as consumed
kcal/g of bodyweight. Values are mean ± SEM, n is variable during the course of experiment as
natural death of mice occurred. Student’s t-test is used to test difference in (A) and (C), 2-way
ANOVA is used to test difference between humanin and control cohorts in (B) and (D). *p<0.05,
**p<0.01.
HNG Treatment Improved Metabolic Markers and Reduced Inflammation
Furthermore, similar to dietary restriction, HNG-treated mice (n = 8) had a significant decrease
in blood glucose and IGF-I levels with a concomitant increase in IGFBP-1 than control mice (n =
6) (Figure 2-5 A, B). The results also agree with the previous cancer study in which we found
reduction of IGF-I by 30% and elevated IGFBP-1, confirming humanin’s role in suppressing

70
GH/IGF-I axis. Inflammation play a key role in the development of many age-associated
disorders including AD, PD and T2DM (286). HNG mice also had several improved
inflammatory markers similar to CR with decreases in both IL-6 and IL-1 beta levels (Figure 2-5
C, D). The levels of IL-10 (P < 0.05) and INF-gamma (P = 0.0521) in HNG treated group were
higher than that of control.
Effects of HNG on Memory, Recognition and Motor Function
Several physical and cognitive tests were performed on this cohort with a number of
improvements found in the HNG-treated group. To measure spatial learning and memory, Barnes
maze, a hippocampus-dependent cognitive task requiring spatial reference memory to locate a
unique escape box by learning and memorizing visual clues (287), was conducted. The HNG-
treated mice were more successful at this task than control mice (n = 10) (Figure 2-6 B),
reflecting a change in search strategy with HNG-treated mice switching to a spatial learning
strategy. Working memory, assessed with a Y-maze test (288), showed that HNG-treated mice
were once again significantly better than control as mice in the HNG cohort displayed enhanced
spontaneous alternating behavior (SAB %) compared to control mice with no difference in the
total number of arm entries (a measure of activity) (Figure 2-6 A). A rotarod test showed that
HN-treated mice stayed on for a longer period of time on the accelerating rotarod (289),
suggesting an improvement in balance and motor coordination (Figure 2-6 C). We also evaluated
whether their performance improved during subsequent trials. The mice from the HNG group
performed consistently better (staying on the rotarod longer) than the control mice, although the
rate of learning was similar in the two cohorts (Figure 2-6 C, trial 2 - 6). Finally, reduced
expression of the pro-inflammatory microglial marker Iba-1 was seen (Figure 2-6 D, HNG: n =
8, control: n = 6). As seen in circulating inflammatory markers, humanin effectively reduced the

71
inflammatory burden of the whole body including the brain. Overall, multiple healthspan
parameters were improved in the humanin treated group with a remarkable overlap between the
improvements found in this treatment group with those reported for CR. These results further
support our hypothesis that humanin is a CR-mimetic and indicate that it has a healthspan-
promoting effect in mammals.

Figure 2-5 HNG Treatment Suppresses IGF-I and Circulating Glucose, Reduces Pro-
Inflammatory Cytokines
P = 0.0521

72
(A) Blood glucose measured by glucose meter at 23-month old. (B) Plasma murine IGF-I levels
measured by ELISA (control: n = 6; HNG: n = 8). (C-F) Plasma cytokine levels measured by
Meso-scale discovery (MSD) assays (control: n = 6; HNG: n = 8). Plasma samples are collected
at 32-month of age (time at sacrifice). Values are mean ± SEM. Levels are compared by using
Student’s t-test, *p<0.05.

Figure 2-6 HNG Improves Motor Function, Hippocampal-Dependent Learning, and Working
Memory, Possibly by Reducing Activated Microglia
(A) Spontaneous alternation behavior (SAB) at 23 months (n = 11 per group). (B) Success rate
in the Barnes maze at 23 months (n = 7–12 per group). (C) Best rotarod performance score at 23
months (n = 18 per group). (D) Quantification of Iba-1 staining in control and HNG mouse
brains at 32-month-old (time of sacrifice, n = 6 per group). Values are mean ± SEM. Statistics are
computed in (A) and (D) by Student’s t-test, in (B) and (C) by 2-way ANOVA, *p<0.05.

The animal experiment was led by Hemal Mehta, Cohen lab members including me participated
in mice maintenance, body weight and food weight measurements, body composition

73
measurement and tissue collection. Circulating factors were measured by the Aging Biomarker
Core, and the animal behavioral assays were performed by the Mouse Phenotyping Core.

HNG Activated AMPK Pathway and Mitochondrial Biogenesis in Muscle
The dramatic increase in rotarod performance of humanin-treated old mice attracted our
attention, we then focused on deciphering the molecular mechanism of humanin-induced
improvement on muscle function. In fact, we have observed similar effects in animals treated
with another MDP – MOTS-c. This particular peptide exhibits exercise-mimetic effects via
activating AMPK pathway. While each MDP possesses unique biological functions and
characteristics, they can certainly share common target pathways as they all belong to the
mitochondrial peptide family. In addition, AMPK exists as the central regulator of skeletal
muscle energy metabolism. The activation of AMPK requires two steps: reversible
phosphorylation of Thr172 in the α subunit and stimulatory allosteric binding of AMP within the
C-terminal region of the γ subunit (290). CR can induce energy deprivation and the increase in
AMP and ADP relative to ATP, leading to phosphorylation of AMPK (291). Therefore, we
decided to examine whether AMPK is involved in HNG-induced enhancement of muscle
function. We found that in the longevity study, 14-month of twice-weekly injection of HNG
(5mg/kg body weight) was sufficient to activate AMPK in the skeletal and heart muscle (Figure
2-7 A, B). Later, we set to find out whether this can be replicated in cell culture system. As
skeletal muscle is the most responsive tissue in terms of AMPK activation, differentiated C2C12
myotubes were chosen as they produce characteristic muscle proteins (292). Experiments using
30µM HNG to treat differentiated C2C12 myotubes further confirmed humanin-induced AMPK
activation in vitro (Figure 2-7 C). Intriguingly, we found that HNG was able to trigger AMPK
phosphorylation as early as after 30 minutes of treatment, although the effect was less prominent

74
than the effect after 1-hour of treatment. Moreover, this AMPK activation effect was sustained at
24-hour.

Figure 2-7 HNG is Potent Inducer of AMPK in vivo and in vitro
Western blot analysis was performed to assess the total protein levels and the phosphorylation
states of AMPKα (Thr172) in (A) Skeletal muscle isolated from control and HNG-treated mice
(n = 4 per group). (B) Heart muscle isolated from control and HNG-treated mice (n = 4 per

75
group). (C) C2C12 myotubes at different time points (n = 3 per group). Quantifications were
performed using Image J, values are mean ± SEM. Statistics are calculated by Student’s t-test or
one-way ANOVA for multiple comparison, *p<0.05, **p<0.01, ***p<0.001.
AMPK activation usually results in activation of catabolic pathways to generate more ATP for
the energy homeostasis. One of the major effector downstream AMPK pathway is the
peroxisome proliferator-activated receptor-γ coactivator-1α (PGC-1α). The positive correlation
between PGC-1α expression and mitochondrial biogenesis has already been established, and
there is a causative link between PGC-1α and mtDNA content and key mitochondrial protein
expression. The RT-qPCR was utilized to examine the expression levels of PGC-1α mRNA in
skeletal muscle tissues isolated from HNG-treated mice and control mice at 32 months of age (n
= 4 per group). Consistent with the previous finding of AMPK activation, HNG treatment
induced PGC-1α mRNA (P < 0.05) and led to increased mtDNA copy number (P < 0.05) (Figure
2-8 A, B). The increase in PGC-1α protein level in HNG-treated skeletal muscle was also
confirmed using immunoblot (Figure 2-8 E). The finding of mitochondrial copy number and
PGC-1α elevation in the skeletal muscle from HNG group suggests that the long-term
administration of humanin promoted mitochondrial biogenesis. Through analyzing the
expression of four essential nuclear-encoded mitochondrial genes: Lon protease-1 (LonP1),
solute carrier family 25 member 5 (SLC25A5), ubiquinol-cytochrome c reductase core protein II
(UQCRC2) and mitochondrial transcription factor-A (TFAM) by RT-qPCR and immunoblot, we
corroborated humanin-induced mitochondrial biogenesis as the mRNAs and protein levels of the
aforementioned proteins were also elevated (Figure 2-8 C, E) in response to long-term HNG
administration. Moreover, consistent with the rise in TFAM levels, the overall mitochondrial
transcription levels also increased as shown by higher levels of mitochondrially-transcribed
genes including cytochrome b, cytochrome c oxidase II (COII) and NADH dehydrogenase 1
(NDI) (Figure 2-8 D).

76

Figure 2-8 HNG Treatment Enhances Mitochondrial Biogenesis, Demonstrated by Upregulation
of MtDNA Copy Number and Elevation of Mito-Gene Expression

77
(A) Relative mtDNA copy number in skeletal muscle tissue (control as baseline, 100%, (n = 4
per group). (B) Relative expression of PGC-1α mRNA measured by RT-qPCR (n = 4 per group).
(C) Relative expression of LonP1, Slc25A5, TFAM and Uqcrc2 mRNA measured by RT-qPCR
(n = 4 per group). (D) Relative expression of mitochondrial encoded genes including cytochrome
B, COII and ND1Western blot analysis was performed to assess the total protein levels of LonP1
and TFAM in skeletal muscle isolated from control and HNG-treated mice, GAPDH is used as
loading control (n = 4 per group). Quantifications were performed using Image J. Values are
mean ± SEM, levels are compared by using Student’s t-test, *p<0.05, **p<0.01.
The Molecular Mechanism of HNG Induced Mitochondrial Biogenesis and ATP
Generation
The promoters of aforementioned nuclear-encoded mitochondrial proteins (LonP1, TFAM,
SLC25A5, and UQCRC2) all have the nuclear respiratory factor-1 (NRF-1) binding sites.
Furthermore, NRF-1 is also one of the most important transcription factor targets of PGC-1α
coactivator. Therefore, we investigated how NRF-1 levels and functions were altered in HNG-
treated mice. We first hypothesized that humanin action involves increased NRF-1 translocation
into the nucleus to initiate transcription of these genes. To test this hypothesis, subcellular
fractionation was performed using the skeletal muscle tissue isolated from the 32-month old
control and HNG mice (n = 3 per group). Purity of nuclear fractions and cytosolic fractions were
confirmed by immuno-detection of nuclear and cytosolic markers (Figure 2-9 A). However, by
comparing the abundance of NRF-1 protein in the nuclei or cytosol from HNG-treated mouse
skeletal muscle with that of the controls, we failed to observe any changes in the translocation of
NRF-1 (Figure 2-9 B). Thus, the evidence failed to support the hypothesis that humanin triggers
NRF-1 translocation into the nucleus.
In the previous experiment, it was confirmed that HNG induced activation of AMPK and
increased expression of PGC-1α. Several independent studies established the relationship
between PGC-1α and the NRFs: PGC-1α modulates the NRFs via physical interaction and PGC-

78
1α-induced mitochondrial biogenesis requires NRFs. It is already known that PGC-1α enters
nucleus to bind NRF-1 and acts as a co-activator to enhance NRF-1 binding to the promoters.
For example, exercise caused increase in PGC-1α protein expression and NRF-1 binding to the
δ-aminolevulinate synthase (δ-ALAS) promoter (293). Therefore, our next hypothesis is that
PGC-1α can increase NRF-1 binding to the promoters of genes encoding key mitochondrially
localized proteins such as TFAM (Figure 2-9 C). To test this hypothesis, chromatin-
immunoprecipitation (ChIP) assay was performed and HNG group had higher NRF-1 binding
affinity to the TFAM promoter (Figure 2-9 C, D). Since PGC-1α is also a potent transcriptional
inducer of NRF-2, we also examined levels of NRF-2 in skeletal muscle tissues and the amount
of NRF-2 was non-detectable (data not shown).

79

Figure 2-9 HNG Increases NRF-1 Binding Affinity to TFAM Promoter
Subcellular fractionation was performed on skeletal muscle tissue isolated from control and
HNG-treated mice (n = 3 per group). (A) Purities of cytosolic and nuclear fractions tested by
presence/absence of cytosolic marker (GAPDH) and nuclear marker (Lamin B1). (B) Western
analysis of NRF-1 abundance in nucleus (upper panel) and cytosol (lower panel). (C) ChIP
results demonstrating increased NRF-1 binding to TFAM promoter, statistics computed by t-test.
(D) Hypothetical pathway of humanin-induced PGC-1α coactivation of NRF-1.
The next step is to see whether there is a causative link between AMPK, PGC-1α and
mitochondrial biogenesis. We chose to use mammalian cells as our experimental system.
However, we found that short term treatment (24-hour) of HNG was able to increase PGC-1α
levels but not to elicit mitochondrial biogenesis as the key mitochondrial genes were not changed

80
(Supplemental Figure S2). We suspected that short-term HNG treatment is not enough to lead to
increase in levels of LonP1, TFAM and other mitochondrial proteins. This result is not
unexpected, as mitochondrial biogenesis induced by AMPK and PGC-1α is a molecular
adaptation process which takes time. Through subsequent experiments, we found that short-term
treatment of HNG (24-hour) in the C2C12 myotubes triggered AMPK activation and
downstream PGC-1α elevation, which was enough to induce higher ATP generation (Figure 2-
10). we further inhibited AMPK activation by using the chemical inhibitor compound C, as a
result, the induction of PGC-1α expression and ATP production by humanin was diminished.
Therefore, we established that the central energy sensor AMPK and its downstream PGC-1α are
the key effectors of humanin-mediated effects, especially muscle bioenergetics.
So far, we have demonstrated that the short-term HNG administration increased ATP generation
without increasing mitochondrial biogenesis, it is still interesting to find out the mechanism by
which humanin enhances energy production acutely through AMPK pathway. Therefore, we
took diet-induced-obese (DIO) mice and treated with HNG for 3 days (daily injection, 5mg/kg
body weight). The plasma samples were collected and the plasma metabolite profile in response
to humanin administration was carefully analyzed. We found that several long-chain
acylcarnitines were lower in HNG-treated mice (Table 2-1). Carnitine is conjugated to fatty
acids, by the enzyme carnitine palmitoyl transferase (CPT), to facilitate their transport across the
mitochondrial membrane (294). The activity of CPT is considered a rate-limiting step in fatty
acid b-oxidation. Elevated levels of acylcarnitines in cells can indicate increased levels of fatty
acid b-oxidation. Increases of acylcarnitines in plasma can be an indication of inefficient fatty
acid b-oxidation, with subsequent export of excess acylcarnitines from cells ending up in the
blood stream. Under these circumstances, the humanin-dependent decrease in acylcarnitines

81
might be consistent with a higher efficiency of mitochondrial oxidative phosphorylation hence
ATP production. Moreover, the increase in plasma glucose and decline in malonate may indicate
that acute humanin administration causes skeletal muscle fuel selection by increasing fatty acid
oxidation while temporarily shutting down glucose oxidation.

Figure 2-10 Inhibition of AMPK Diminished HNG-Induced ATP Production

82
(A) Western blot demonstrating inhibition of HNG-induced pAMPK increase by chemical
inhibitor compound C (10µM) for 24 hours. (B) Western blot demonstrating inhibition of HNG-
induced PGC-1α increase by chemical inhibitor compound C (10µM) for 24 hours. (C) ATP
production increased by HNG treatment (30µM) for 24 hours, and the effect is negated by
AMPK inhibition via compound C. Quantifications were performed using Image J. Values are
mean ± SEM, levels are compared by using One-way ANOVA, *p<0.05, **p<0.01, ***p<0.001.
Table 2-1 Changes in Acylcarnitines, Malonate and Glucose Levels When Treated with HNG
Metabolite Pathway Fold Change p-value
3-hydroxybutyrylcarnitine Fatty Acid Metabolism 0.389398497 0.0097612
acetylcarnitine Fatty Acid Metabolism 0.786074069 0.24119123
adipoylcarnitine Fatty Acid Metabolism 0.540135203 0.03218876
cis-4-decenoyl carnitine Fatty Acid Metabolism 0.681189913 0.11951888
decanoylcarnitine Fatty Acid Metabolism 0.770855872 0.20550292
hexanoylcarnitine Fatty Acid Metabolism 0.621088294 0.21047661
laurylcarnitine Fatty Acid Metabolism 0.768928417 0.28484918
linoleoylcarnitine Fatty Acid Metabolism 0.718161314 0.02515596
myristoleoylcarnitine Fatty Acid Metabolism 0.707817384 0.08797376
myristoylcarnitine Fatty Acid Metabolism 0.731927383 0.15555153
octanoylcarnitine Fatty Acid Metabolism 0.86180001 0.55624984
oleoylcarnitine Fatty Acid Metabolism 0.70764819 0.00949586
palmitoleoylcarnitine Fatty Acid Metabolism 0.657882714 0.04376562
palmitoylcarnitine Fatty Acid Metabolism 0.818761593 0.11279927
stearoylcarnitine Fatty Acid Metabolism 0.925099802 0.20267983
malonate Fatty Acid Synthesis 0.740656665 0.04868544
glucose Glucose Metabolism 1.289298323 0.04696272

Green: indicates significant difference (p≤0.05), metabolite ratio of < 1.00
Red: indicates significant difference (p≤0.05), metabolite ratio of ≥ 1.00

83
Discussion
Differential Stress Resistance (DSR) by Humanin
Recent advances in the understanding of cancer cell biology and multiple modality treatment
have led to dramatically improved cancer survival (295). To improve the quality of life in cancer
survivors, there is an increasing need for cytoprotective agents that reduce the treatment-related
toxicity for normal tissues without affecting treatment efficacy on cancers. Most
chemotherapeutic agents cause acute adverse effects on bone marrow suppression and chronic
adverse effects, including subfertility/infertility, growth defects in children, and cognitive
dysfunction in long term cancer survivors (296). Recent reports describe the new role of the
mitochondrial peptide humanin in protecting against chemotherapy-induced adverse effects on
bone growth and germ cell apoptosis (97,116,297). A recent report also suggests that HNG by
itself might have anticancer effects (116). CR, which is the most effective dietary regimen that
increases lifespan, has also been shown to reduce cancer incidence (298). However, in clinical
cancer treatment, CR is not recommended as an adjuvant therapy due to the problems associated
with long-term caloric shortage. In the collaborative project with the Wang lab we investigated
whether a potent humanin analog HNG will protect against chemotherapy-induced toxicity while
exerting additive or synergistic effects on chemotherapy-induced metastatic tumor suppression,
in a way that mimics the effects of CR but does not require reduction in caloric intake.
It has been already demonstrated that a single intraperitoneal injection of either humanin or its
potent analog HNG attenuated CP-induced male germ cell apoptosis within 12 hours in healthy
rats (297), and within 24 hours in healthy mice (97). HNG was able to protect against CP-
induced spermatogonia, spermatocytes and round spermatids apoptosis in rodents (297). In this
study, we showed that daily intraperitoneal injection of HNG for 2 weeks significantly reduced

84
CP-induced male germ cell apoptosis in metastatic tumor-bearing mice. Moreover, this effect is
sustained with repeated dosing of CP, as we demonstrated that HNG daily subcutaneous
injection for 4 weeks significantly increased the number of spermatogonial stem cells (SSCs) and
cauda epididymal sperm count in mice with repeated doses of CP treatment. We provided the
first evidence that co-administration of HNG with CP significantly attenuated the CP-induced
suppression of the number of total WBCs, granulocytes and monocytes in healthy and metastatic
tumor-bearing mice while enhancing the therapeutic effect of CP on tumor, respectively. The
protective actions of humanin against stress-induced apoptosis of many cells initially led to the
concerns that it might promote tumor growth. We found in this study that HNG by itself
suppressed the number of metastatic lung melanoma. Moreover, we demonstrated that HNG
enhanced the suppressive effects of CP on the number of metastatic lung melanomas as
compared with CP treatment alone. Eriksson et al., 2014 showed that HNG did not interfere with
the ability of a proteasome inhibitor, bortezomib, to suppress neuroblastoma and
medulloblastoma cell lines and xenografts in nude mice (116). Our results showed that HNG had
synergistic/additive action with CP on metastatic melanoma. Therefore, our novel results in
metastatic lung tumors as a model for cancer paves the way to the development of humanin
analogues as an adjuvant therapy to enhance chemotherapy-induced tumor suppression while
protecting against chemotherapy-induced adverse effects.
The mechanisms of action of humanin and its analogues on enhancing chemotherapy-induced
tumor suppressive effects while simultaneously protecting normal cells and tissues against
chemotherapy-induced adverse effects are divergent and suggest different signaling pathways.
Humanin has cytoprotective, anti-inflammatory, antioxidative properties, that may involve a
systemic normalization of stress-response homeostasis in many cell types (105,272). The

85
cytoprotective effects of humanin on normal cells are mediated through at least 2 receptors, one
of which is the immuno-modulating protein-coupled formyl peptide receptor-like 1 (100), and
the other is the IL-like heterotrimeric cell membrane receptor composed of a
CNTFRα/gp130/WSX-1 (299). Upon receptor binding, humanin activates Signal Transducer and
Activator of Transcription 3 (STAT3) intracellular signaling pathways to promote cell survival
(120). Humanin also binds IGFBP-3 and modulates IGF bioactivity to regulate cell growth,
survival, and apoptosis (78). In addition, within cells humanin binds to Bcl-2-associated X
protein and BH3 domain proteins preventing their entry to the mitochondria inhibiting initiation
of apoptosis induced by stress (300–302). Our previous studies on humanin indicates that the
protective effect of humanin against stress-induced (including chemotherapy) germ cell
apoptosis is through interaction with the putative membrane receptors enhancing STAT3
signaling and decrease p38 MAPK as well as through sequestration of Bcl-2-associated X
protein in the cytoplasm and preventing its entry to the mitochondria to initiate the apoptosis
cascade (303).
The mechanisms of the novel finding of divergent but beneficial effects of HNG on cancer-
bearing animals are intriguing. This phenomenon is reminiscent of the concept of differential
protection, which was first proposed by Longo and coworkers using fasting and caloric-
restriction regimens in rodent aging and cancer models (304). They demonstrated that fasting
selectively protected normal cells and organs, but not cancer cells against oxidants and
chemotherapeutic agents (305). There are many possible mechanisms of action of HNG on
cancers. For example, binding of IGF-I to its receptor (IGF1-R) activates the IGF-1 signaling
pathway through Akt/mTOR promoting cancer growth (306,307). Studies have shown that
circulating humanin has an inverse relationship with IGF-1 and GH in mouse models and in men.

86
Humanin levels are high in GH and IGF-I liver-specific knockout mice and in GH receptor-
deficient children. On the other hand, humanin levels are suppressed in transgenic mice
overexpressing GH and in children after GH treatment (308). We showed in this study that HNG
not only inhibited plasma IGF-I levels but also enhanced the CP-induced suppression of IGF-I in
tumor-bearing mice. The decrease of IGF-I by HNG may inhibit IGF-I bio-availability to its
receptor leading to the suppression of melanoma growth (309). It is worth noting that strategies
using blocking antibodies against IGF-I receptor were not successful because of lack of efficacy
and increased side effects; these were likely related to the increase in GH levels with IGF-I
receptor blockade which is not induced by fasting or humanin treatment (306,310). Thus, the
HNG-induced suppression of IGF-I may play a role in the protection of normal cells and
suppression of cancer cells.
Intriguingly, the IGF-I suppression by humanin cannot be recapitulated in normal young mice. It
is noteworthy that the melanoma mice have significantly higher IGF-I levels than normal wild-
type mice, and the injection of CP or co-injection of CP and HNG significantly decreased
circulating IGF-I levels in mice with tumor. Furthermore, co-treatment of CP and HNG at
0.5mg/kg or 5mg/kg body weight brought IGF-I down to the levels in normal healthy mice.
These findings suggest that the dietary-mimetic effects of humanin are selective. It has been
reported that multiple types of cancers have elevated levels of circulating IGF-I (311,312).
Higher IGF-I levels are also associated with increased risk of breast cancer, prostate cancer and
colorectal cancer (313–315), however, maintaining IGF-I at an optimal level is required to
promote growth, cytoprotection and anti-inflammation (316). Based on the results, we found that
unlike CR or fasting, humanin suppressed excess IGF-I in tumor-bearing mice but not in young
healthy mice.

87
CR in animal models and men are associated with longevity and reduced inflammation, oxidative
stress and risks of developing many types of cancer. High humanin levels are associated with
longevity mimicking the effects of CR on lifespan (183,308). Short term fasting in mice lowers
IGF-I and increases IGFBP-1 levels protecting against chemotherapy-induced cytotoxicity (317).
HNG may influence tumor growth by altering nutrient levels, because humanin analogues have
been proposed as a potential agent for treating patients with diabetes (105,272). By regulating the
metabolism of the host, HNG may sensitize the cancer cell making them more responsive to
chemotherapy. HNG-induced suppression of IGF-I mimicked the homeostatic conditions created
by CR or fasting in the host suggesting that HNG may be a CR mimetic with a promising
therapeutic potential for patients with cancer.
Humanin Acts As a CR-Mimetic and Extends Healthspan
To date, a number of studies focused on the beneficial effects of humanin on one age-related
disease at a time. However, in order to better understand the role of humanin in aging, a wide
spectrum of diseases needs to be considered together. The effects of long-term humanin
administration on the aging process have never been carefully elucidated. Based on the fact that
humanin and its analogs ameliorated a range of age-related pathologies, and the previous finding
of humanin-induced IGF-I suppression, we hypothesized that humanin can act as a CR-mimetic
and cause broad changes in metabolism, signaling events and stress response. The most
prominent feature of CR is that it retards the aging process, therefore, we decided to investigate
how humanin influences adverse outcomes of aging in mice starting from middle-aged mice (18-
month-old) onwards. As seen with CR, HNG treatment resulted in loss of body weight.
Nonetheless, the reduction in body weight is modest, and the difference between weights of
HNG mice and control mice gradually diminished as they approached extremely old age. For

88
being considered as old, C57BL/6 mice and many other genotypes should be at least 18 months
old and 24 months is generally the upper limit (318). The descending trend in body weight after
around 670 days (approx. 23 months old) is most likely due to loss of muscle mass at advanced
age. Loss of muscle mass after 18 months of age was also accompanied by a significant increase
in bone resorption (319). The fact that HNG did not further reduce body weight might be a
beneficial effect. To confirm whether HNG treatment can help the old mice maintain healthy
body composition, we used microCT to analyze each mouse’s visceral fat percentage,
subcutaneous fat percentage and lean body mass at 28 months. From the results, we gathered that
long-term HNG administration led to decreased volume of visceral fat deposits, while the
subcutaneous adipose tissue was not affected. Visceral fat accumulation is a hallmark of aging,
the surgical removal of visceral fat improved mean of maximal lifespan in rats (320). There are
several mechanisms suggested to support causative link between visceral adipose tissue and
healthspan/lifespan: 1) because of its proximity to the major organs, visceral fat accrual results in
greater free fatty acids and glycerol release (321). 2) adipose tissue can behave like an active
endocrine organ, which is capable of secreting many cytokines, often referred to as adipokines,
that can promote inflammation and interfere with insulin action (322). Although insulin levels
were not changed in the HNG cohort, glucose levels were significantly decreased, and IGF-I was
reduced by 30% by the end of HNG treatment. Reduced signaling of the GH/IGF-I axis extends
health-and lifespan in rodents (323,324). These results suggest that humanin can maintain
glucose homeostasis and healthy body composition.
Aging is also accompanied with chronic, low-grade inflammation. While the acute immune
response and inflammation are essential for defense against infection and can facilitate the repair,
persistent production of pro-inflammatory factors can eventually damage adjacent tissues, or

89
result in microenvironment changes that are unfavorable for normal cell signaling (325). The
primary source of inflammation is the senescent cells: the secretion of SASP factors. More
recently, these data were further strengthened as levels of IL-6 and soluble tumor necrosis factor-
α receptor 1 (TNF-α R1) were identified as predictors of 10-year all-causes mortality (326). As
discussed before, visceral fact deposits can also produce adipokines and trigger inflammatory
response. IL-6 can be secreted from infiltrated macrophages in adipose tissue, the lower levels of
circulating IL-6 in HNG mice may be the result of smaller visceral fat volume (327). It was also
reported that pretreatment with humanin decreased the level of IL-6, IL-1β and tumor necrosis
factor α (TNFα) released from astrocytes induced by liposaccharide (LPS), moreover, humanin
partially antagonized inflammation injury induced by LPS (278). Based on this in vitro finding,
we can also hypothesize that humanin has direct suppressive effect on the production of pro-
inflammatory cytokines from astrocytes and macrophages. This hypothesis is also supported by
the reduced Iba-1 staining in cortex. Iba-1 is a marker for activated microglia and there is a
consistent linkage between aging and upregulation of microglial activation that could be
interpreted as aging-related neuroinflammation. Taken together, HNG extends healthspan
through suppression of chronic inflammation. The systemic reduction in inflammation burden
may contribute to the mitigation of neurodenegeration, cardiovascular diseases symptoms by
humanin treatment.
To test the effects of HNG on cognitive performance we carried out working memory test. The
spontaneous alternative behavior (SAB) describes the tendency of rodents to alternate arm
choices on successive trials. Animals are normally willing to explore novel environment, and the
SAB is indicative of animals’ spatial working memory. At 23-month old and 24-month old, there
was no difference in SAB between the two groups. The difference is noticeable and significant

90
when mice were 28-month old. According to previous study, significant impairment on working
memory could not be detected up to 28 months of age (328). As for hippocampal-dependent
spatial learning and memory testing, Barnes maze was applied to control and HNG mice (329).
Mice at age of 23-month were allowed to learn from visual cues and received training for 7 trials.
While the control mice did not seem to learn during the training period, HNG had better success
rate at finding the escape box after trial 4. The improvement on hippocampal learning and
memory is usually related to neurogenesis (330), however, neurogenesis did not exhibit any
changes in response to HNG injection as shown by BrdU
+
staining (data not shown). This result
suggests that the improvement on hippocampal learning and memory by humanin is independent
of inducing neurogenesis. It has been widely accepted that inflammation can lead to impairments
in learning, memory and other cognitive functions (331,332). Based on the finding that humanin
had profound effect on neuroinflammation, we can also speculate the neuroprotection and
prevention of cognitive decline could be partially attributed to suppression of inflammation.
Humanin Activates AMPK Pathway and Improves Muscle Bioenergetics
The rotarod test is often used to assess animal’s motor function. At 23 months of age, HNG mice
were able to consistently outperform control mice at 6 successive trials by staying on the
accelerating rod longer. Rotarod latency is negatively associated with advancing age (333), as a
progressive loss of muscle mass and strength is accompanied with aging (334). Motor function
improvement and AMPK activation were often observed together: AMPK has been identified as
a critical regulator of muscle fiber contractile gene expression (335), treatment with AMPK
agonist AICAR in mice also increased rotarod latency (336). We found that long-term
administration of HNG activated AMPK in skeletal muscle and heart muscle. Moreover, acute
HNG treatment also enhanced AMPK phosphorylation in C2C12 myotubes. One of AMPK’s

91
primary functions is promoting mitochondrial biogenesis to facilitate cell’s adaptation to energy
deficiency. Many correlative studies also demonstrated that AMPK activation is associated with
increased mitochondrial enzyme content (337) and mitochondrial biogenesis (338) in rat skeletal
muscle. Subsequent analysis of skeletal muscle isolated from the long-term mouse study
revealed that HNG-treated mice had higher expression of PGC-1α. PGC-1α is a key regulator of
mitochondrial biogenesis, and the levels of mitochondrial content as shown by mtDNA copy
number were also increased (339). The role of PGC-1a in regulating the key mitochondrial
proteins in skeletal muscle was strongly supported by the changes stimulated by overexpressing
PGC-1α both in myotubes (339) and in mice (340).TFAM, the mitochondrial transcription factor
that is responsible for mtDNA transcription and replication, also had higher expression level.
TFAM levels have also been positively correlated with mtDNA copy number. Other
mitochondrial enzymes including mitochondrial Lon protease, mitochondrial solute carrier and
ubiquinol-cytochrome c reductase also had increased expression. The promoters of the LonP1,
TFAM and Slc25A5 genes all have functional binding sites for NRF-1 (341,342). Therefore,
NRF-1 is highly likely involved in HNG induced mitochondrial biogenesis. It has been shown
that redox conditions can cause NRF-1 nuclear translocation via phosphorylation by Akt (342).
Humanin can bind to GP130 receptor and result in Akt activation, hence the speculation of HNG
induction of NRF-1 translocation was firstly tested. However, the results failed to support this
hypothesis. A number of studies also suggested that the chronic AMPK activation can lead to
increased NRF-1 binding affinity to gene promoters (338). The binding affinity of NRF-1 can be
enhanced by coactivation through physical interaction with PGC-1α (339). As demonstrated
here, PGC-1α was dramatically increased and powerfully induced mRNA for multiple target
genes such as LonP1 and TFAM. In addition, NRF-1 binding affinity to TFAM promoter was

92
enhanced, increased its transcriptional activity on target genes, including TFAM. Our results
demonstrate that PGC-1α can coactivate the transcription factor NRF-1 (nuclear respiratory
factor-1) to regulate TFAM, which is consistent with what has already been reported. Future
experiments are needed to validate direct interaction between PGC-1α and NRF-1 proteins.
Interestingly, unlike chronic HNG administration, short time HNG treatment is not able to
induce mitochondrial biogenesis and increase in mitochondrial proteins, such as TFAM, LonP1.
However, we did observe an increase in phosphorylated AMPK and PGC-1α. Moreover, the
activation of AMPK pathway is responsible for HNG-induced increase in ATP production. There
seems to be a discrepancy: how AMPK-PGC-1α pathway results in ATP elevation without
influencing mitochondrial biogenesis? We speculated that the short time HNG treatment and
acute AMPK activation lead to higher mitochondrial fatty acid oxidation activity, hence
increased ATP production; while the chronic AMPK activation gradually causes mitochondrial
adaptation and biogenesis. The metabolomic profile of plasma from short-term HNG-treated
DIO mice supported this hypothesis. Malonate is crucial metabolite for fatty acid synthesis,
moreover, the serum concentration of malonate is positively correlated with de novo fatty acid
synthesis (343). The decreased levels of acylcarnitines and malonate in plasma are indicative of
higher activity of fatty acid oxidation and lower activity of lipogenesis. In addition, elevated
glucose also suggests that humanin leads to fuel selection from glucose to fat. In summary, our
findings, are consistent with the known regulatory effects of PGC-1α on cellular energy
metabolism (344), suggest a mechanism by which humanin activates AMPK-PGC-1α to
suppress muscle glucose oxidation while increasing the expression of genes involved in
mitochondrial fatty acid oxidation.

93
In this chapter, we demonstrated humanin, a mitochondrial-derived micropeptide that acts as a
“mitokine”, can be supplemented to tumor-bearing or aged mice and ameliorate symptoms. We
found that humanin and its potent analog HNG suppressed GH/IGF-I axis and enhanced AMPK
pathways, which are effectors of pro-longevity intervention, CR. As for the biological outcomes,
similar to CR, humanin administration exhibited differential chemotherapy protection,
improvement on metabolic markers and suppression of chronic inflammation, leading to better
cognitive function. Accordingly, approaches involving partial pathways implicated in CR might
show limited effects as a CR mimetic.

Figure 2-11 Humanin Acts as A Caloric Restriction (CR) Mimetic and Improves Healthspan

Improved metabolic
markers
Suppression of
chronic
inflammation
Delayed motor
function decline
Differential
chemo-protection

94
Materials and Methods
Animals and Reagents
Young adult (10-week-old; body weight, 24–28 g) male mice (C57BL/6J) and Old adult (18-
month-old; body weight, 28-29 g), purchased from The Jackson Laboratory, were housed in a
standard animal facility under controlled temperature (22°C) and photoperiod (12 h light, 12 h
dark) with free access to water and mouse chow. Animal handling and experimentation were in
accordance with the recommendation of the American Veterinary Medical Association. All
animal protocols were approved by the Institutional Animal Care and Use Committee (IACUC)
of the University of Southern California. CP monohydrate was obtained from Sigma and
dissolved in saline at concentration of 20 mg/mL before use. Humanin analog HNG was
synthesized by CPC Scientific and dissolved in saline at concentration of 1 or 2 mg/mL before
use.
Old mice were housed in clear shoebox-cages in groups of three animals per cage. Animals were
randomly divided (by cage to avoid fighting) into the control group and the HNG group.
Bodyweight of individual animals was measured routinely every week and recorded. Food intake
was measured daily. Upon indication of progressing dermatitis, animals were treated with a triple
antibiotic ointment (Fougera Pharmaceuticals Inc.) and were euthanized if the condition
progressed. To reduce subjective bias, mice were randomly assigned (using the online Random
Number Calculator from GraphPad) to any behavioral and physiological assessments shortly
before any experiment. Mice that appeared weak and/or showed signs of illness were not
included into any experiment. HNG (a potent analogue of humanin with a glycine substitution,
S14G) were synthesized and received from Genscript (Piscataway, NJ, USA) and dissolved in
double-distilled water (ddH2O).

95
Longevity Study
110 eighteen-month-old mice were randomized into control group and HNG group. Control mice
were injected intraperitoneally twice every week with vehicle (ddH2O), HNG mice received
twice weekly intraperitoneal injection with HNG (5mg/kg BW). Upon death, bodies were stored
in formalin. The study was terminated at 32-month-old, and all the remaining alive mice were
sacrificed after isoflurane anesthesia.
Behavioral Studies
Y-maze
11 mice per treatment group were tested at 23 months of age. Spontaneous alternation behavior
(SAB) score was calculated as the proportion of alternations (an arm choice differing from the
previous two choices) to the total number of alternation opportunities.
Accelerating rotarod
At 23 months of age, 18 mice/group were evaluated using an accelerating rotarod. The speed and
time after which the mice fell off were recorded. On two consecutive days, the mice were given
three successive trials, for a total of six trials. Performance was measured with two variables: the
mean of the individual best performance over the two consecutive trial days and the mean time
the mice of each treatment group remained in balance over the six-trial session as an index of
training.
Barnes Maze
12 mice/group were tested twice daily for 7 days at 23 months of age. Success rate (100%,
finding the escape box (EB) within 2 minutes or 0%, not finding the EB within 2 minutes),
latency (time to enter the EB), number of errors (nose pokes and head deflections over false

96
holes), deviation (how many holes away from the EB was the first error) and strategies used to
locate the EB were recorded and averaged from two tests to obtain daily values. Search strategies
were classified as random (crossings through the maze center), serial (searches in clockwise or
counter-clockwise direction) or spatial (navigating directly to the EB with both error and
deviation scores of no more than 3). Retention was assessed by testing once on day 14.
Animal behavioral assays were performed by the Animal Phenotyping Core.
Blood Collection and Tissue Preparation
Mice were injected with heparin (1300 IU/kg BW, ip) 15 minutes before being killed by a lethal
injection of sodium pentobarbital (200 mg/kg BW ip). BW was recorded at autopsy. Blood
samples were collected from the right ventricle of each mouse immediately after death and used
for complete blood count using an automated cell counter (VetScanHM2; ABAXIS). Plasma was
separated and stored at −20°C for subsequent HNG, IGF-1, and IGFBP-1 measurements by
specific and sensitive ELISA assays as previously described (168,345).
Cell Culture and HNG Treatment
C2C12 myoblasts were differentiated into myotubes by growing them in DMEM supplemented
with 2% horse serum as described previously (292). At a seeding density of 2×104 cells/ml (1
and 2 ml per well in 24- or 12-well plates, respectively) C2C12 cells fully differentiate into
myotubes by day 7. Myotubes were deprived of serum for 3 h prior to all experimental
manipulations. HNG was prepared in ddH2O as a 1mM stock solution, and diluted in DMEM
before application to the cells. ddH2O was added to control cells in all experiments to obtain the
same final concentration as in HNG-treated conditions.

97
Western Blotting
About 50 mg of skeletal muscle (m. gastrocnemius) was homogenized in RIPA buffer
(ThermoFisher) added protease and phosphatase inhibitor cocktails (ThermoFisher), kept on ice
for 30 min, centrifuged at 15000 x g for 15 min at 4°C, and the supernatant was collected.
Protein concentration was assayed using BSA as working standard. Equal amounts of protein (30
µg) were heat-denaturized in sample-loading buffer (50 mM Tris-HCl, pH 6.8, 100 mM DTT,
2% SDS, 0.1% bromophenol blue, 10% glycerol), resolved by SDS-PAGE and transferred to
PVDF membranes (Bio-Rad, CA, USA). The filters were blocked with Tris-buffered saline
(TBS) containing 0.05%Tween and 5% bovine serum albumin (BSA) for 30 min and then
incubated overnight at 4°C with antibodies directed against total AMPK, phospho-AMPK
(Thr172), TFAM and PGC-1α (Cell Signaling Technologies, MA, USA), the antibody against
LonP1 and NRF-1 (Novus Biologicals, CO, USA). HRP-conjugated IgG (Cell Signaling
Technologies, MA, USA) was used as secondary antibody. Membrane-bound immune
complexes were detected by an enhanced chemiluminescence system (Bio-Rad, CA, USA) on a
ChemiDoc System (Bio-Rad, CA, USA). Protein loading was normalized according to GAPDH
expression. Quantification was performed by densitometric analysis using ImageJ software.
Nucleic Acid Extraction and Quantitative PCR
DNA was extracted from tissue by using Qiagen DNeasy kit following manufacturer’s protocol,
purity and concentration of DNA samples were examined using NanoDrop 2000 (ThermoFisher,
CA, USA). RNA was extracted by using Zymo Quick-RNA kit (Zymo Research, CA, USA) and
cDNA was made by SuperScript III reverse transcription system (Life Technologies, CA, USA)
Relative transcript expression levels were measured using a SYBR Greenbased method. Average
fold changes were calculated by differences in threshold cycles (Ct) between pairs of samples.

98
Subcellular Fractionation
Snap-frozen skeletal muscle tissue (50mg) were placed in a pre-chilled glass Petri dish and
minced on ice using razor. All samples were resuspended in 300-500 µl of STM buffer
comprising 250 mM sucrose, 50 mM Tris–HCl pH 7.4, 5 mM MgCl2, protease and phosphatase
inhibitor cocktails (all chemicals were from Sigma-Aldrich) and homogenized for 1 minute on
ice using a tight-fitting Teflon pestle by hand. The homogenate was then inspected under light
microscope to see if intact tissue still presents and whether nuclei were intact. The homogenate
was decanted into a 1.5mL Eppendorf tube and maintained on ice for 30 minutes, vortexed at
maximum speed for 15 seconds and then centrifuged at 800 x g for 15 minutes. The pellet was
labelled as P0 and kept on ice, the supernatant was labelled as S0 and used for subsequent
isolation of cytosolic fraction. The pellet P0 (containing nuclei and debris) was resuspended in
300-500 µl STM buffer, vortexed at maximum speed for 15 seconds and then centrifuged at
500 x g for 15 minutes. Following the above step, the nuclear pellet was labelled as P1 and kept
on ice, the supernatant S1 (cell debris) was discarded. Cytosolic was extracted from S0 by
centrifugation at 800 g for 10 minutes. The supernatant S2 was saved and the pellet (P2) was
discarded, S2 was then centrifuged at 11,000 x g for 10 minutes and the supernatant S3
(containing cytosol and microsomal fraction) was precipitated in cold 100% acetone at −20°C for
at least 1 hour followed by centrifugation at 12,000 g for 5 minutes, the pellet (P7) was then
resuspended in 100-300 µl STM buffer and labelled as “cytosolic fraction” (346).
Statistical Analysis
All data are expressed as the mean ± SEM. For mice, all statistical analyses were two-sided and
P values <0.05 were considered significant (* p<0.05, ** p<0.01, *** p<0.001). Differences

99
among groups were tested either by Student t-test comparison, one-way ANOVA followed by
Tukey’s multiple comparison, or 2way ANOVA using GraphPad Prism v.5.

100
CHAPTER 3 CHARACTERIZATION OF SMALL-HUMANIN-LIKE
PEPTIDE 2 (SHLP2)
Abstract
Mitochondria are key players in aging and in the pathogenesis of age-related diseases. Recent
mitochondrial transcriptome analyses revealed the existence of multiple small mRNAs
transcribed from mtDNA, including humanin, a peptide encoded in the mtDNA 16S ribosomal
RNA region that has neuroprotective effect, and MOTS-c, an exercise-mimetic that triggers
metabolomic changes and AMPK activation. An in silico search revealed six additional peptides
in the same region of mtDNA as humanin; we named these peptides small humanin-like peptides
(SHLPs). We identified the functional roles for these peptides and the potential mechanisms of
action. We focused on SHLP2 because it shared similar protective effects with humanin and the
availability of a SHLP2 antibody. Specifically, they significantly reduced apoptosis and the
generation of reactive oxygen species, and improved mitochondrial metabolism in vitro. SHLP2
also enhanced 3T3-L1 pre-adipocyte differentiation. Systemic hyperinsulinemic-euglycemic
clamp studies showed that intracerebrally infused SHLP2 increased glucose uptake and
suppressed hepatic glucose production, suggesting that it functions as an insulin sensitizer both
peripherally and centrally. Similar to humanin, the levels of circulating SHLP2 were found to
decrease with age.
We further characterized the physical and biochemical properties of SHLP2 by using a
combination of approaches, including metabolomic and biophysical analyses. MDP
cytoprotective functions are generally attributed to anti-apoptotic activity, however, little is
known about their capacity to facilitate the cell’s unfolded protein response via direct
interactions with amyloidogenic proteins. Here, we explored the effects of SHLP2, on the

101
misfolding of IAPP, a critical pathogenic step in T2DM in collaboration with the Langen lab.
Seeded fluorescence and co-sedimentation studies demonstrate SHLP2 blocks amyloid seeding
and directly binds misfolded, seeding-capable IAPP species. Furthermore, our electron
paramagnetic resonance spectroscopy and circular dichroism data indicate SHLP2 does not act
by binding IAPP monomers. Our results reveal a novel chaperone-like activity wherein these
MDPs specifically target misfolded amyloid seeds to inhibit IAPP misfolding which, along with
direct anti-apoptotic activity and beneficial metabolic effects, make HNG and SHLP2 exciting
prospects as T2DM therapeutics. These data also suggest that other mitochondrial stress response
factors within the MDP family may be amenable to development into therapeutics for protein-
misfolding diseases.
SHLP2, as a novel mitochondrial-encoded peptide and an important mitochondrial retrograde
signaling molecule, was found to be associated with prostate cancer risk in a race-specific way.
Additional DNA sequencing was carried out in the hope of identifying a genetic factor
responsible for the disparity between races. Based on the previous finding that both humanin and
SHLP are able to act as retrograde signaling molecules, we designed and conducted an open-
ended metabolomic study in which the plasma metabolic profiles of HNG- and SHLP2-treated
diet-induced obesity (DIO) mice were analyzed. SHLP2 administration significantly altered the
concentrations of amino acid and lipid metabolites in plasma. Among all the metabolic
pathways, the glutathione and sphingolipid metabolism responded most strongly to peptide
treatment.
These results suggest that mitochondrial play critical roles in metabolism and survival through
the synthesis of mitochondrial peptides and provide new insights into mitochondrial biology with
relevance to aging and human biology.

102
Background
Human mitochondrial DNA (mtDNA) is a double-stranded, circular molecule of 16,569 bp and
contains 37 genes encoding 13 proteins, 22 tRNAs, and 2 rRNAs. Recent mitochondrial
transcriptome analyses revealed the existence of small RNAs derived from mtDNA (347). In
2001, Nishimoto and colleagues identified humanin, a 24-amino-acid peptide encoded from the
16S rRNA region of mtDNA. humanin is a potent neuroprotective factor capable of antagonizing
AD-related cellular insults (75). Humanin is a component of a novel retrograde signaling
pathway from the mitochondria to the nucleus, which is distinct from mitochondrial signaling
pathways, such as the SIRT4-AMPK pathway (348). Humanin-dependent cellular protection is
mediated in part by interacting with and antagonizing pro-apoptotic Bax-related peptides (77)
and IGFBP-3 (78).
Because of their involvement in energy production and free radical generation, mitochondria
likely play a major role in aging and age-related diseases (349–351). In fact, improvement of
mitochondrial function has been shown to ameliorate age-related memory loss in aged mice
(352). Recent studies have shown that humanin levels decrease with age, suggesting that
humanin could play a role in aging and age-related diseases, such as AD, atherosclerosis, and
diabetes. Along with lower humanin levels in the hypothalamus, skeletal muscle, and cortex of
older rodents, the circulating levels of humanin were found to decline with age in both humans
and mice (272). Notably, circulating humanin levels were found to be (i) significantly higher in
long-lived Ames dwarf mice but lower in short-lived GH transgenic mice, (ii) significantly
higher in a GH-deficient cohort of patients with Laron syndrome, and (iii) reduced in mice and
humans treated with GH or IGF-I (160). Age-dependent declines in the circulating humanin
levels may be due to higher levels of ROS that contribute to atherosclerosis development. Using

103
mouse models of atherosclerosis, it was found that humanin-treated mice had a reduced disease
burden and significant health improvements (96,133). In addition, humanin improved insulin
sensitivity, suggesting clinical potential for mitochondrial peptides in diseases of aging (272).
The discovery of humanin represents a unique addition to the spectrum of roles that
mitochondria play in the cell (74,109). A second MDP, MOTS-c, has also been shown to have
metabolic effects on muscle and may also play a role in aging (72).
We further investigated mtDNA for the presence of other MDPs. Recent technological advances
have led to the identification of sORFs in the nuclear genomes of Drosophila (353,354) and
mammals (108,355). Therefore, we attempted to identify novel sORFs using the following
approaches: 1) in silico identification of potential sORFs; 2) determination of mRNA expression
levels; 3) development of specific antibodies against these novel peptides to allow for peptide
detection in cells, organs, and plasma; 4) elucidating the actions of these peptides by performing
cell-based assays for mitochondrial function, signaling, viability, and differentiation; and 5)
delivering these peptides in vivo to determine their systemic metabolic effects. Focusing on the
16S rRNA region of the mtDNA where the humanin gene is located, we identified six sORFs
and named them small humanin-like peptides (SHLPs) 1–6. While surveying the biological
effects of SHLPs, we found that SHLP2 and SHLP3 were cytoprotective; therefore, we
investigated their effects on apoptosis and metabolism in greater detail. Further, we showed that
circulating SHLP2 levels declined with age, similar to humanin, suggesting that SHLP2 is
involved in aging and age-related disease progression.
We found that SHLP2 was able to rescue primary neurons from amyloid-peptide induced
apoptosis, it is necessary for us to identify the underlying mechanism. Here we directly
investigate this notion using the 37-amino acid polypeptide, IAPP, which plays a critical role in

104
the pathogenesis of T2DM (356). The misfolding and subsequent aggregation of IAPP induces a
gain-of-function associated with β-cell apoptosis, oxidative damage, mitochondrial dysfunction
and endoplasmic reticulum (ER) stress (357–360). In the islets of Langerhans, the process of
IAPP misfolding and aggregation ultimately leads to the replacement of pancreatic β-cell mass
with deposits of fibrillary amyloid, which is the hallmark of T2DM pathology (361). Given the
general paucity of information available regarding MDPs and diabetes-associated protein-
misfolding, plus observations that multiple MDPs display cytoprotective phenotypes (73,105),
we sought to determine whether the potent humanin analog, HNG, as well as the MDP, SHLP2,
could function in a chaperone-like capacity to prevent the misfolding of IAPP. We therefore used
a combination of Thioflavin T (ThT) fluorescence studies in combination with electron
paramagnetic resonance spectroscopy (EPR), circular dichroism (CD) and transmission electron
microscopy to investigate the effect of HNG and SHLP2 on IAPP misfolding.
Prostate cancer is the most common cancer in men in the United States and the second leading
cause of death from cancer in men (362). Therefore, identification of disease-specific biomarkers
is critical to addressing current challenges of whom to biopsy and how to choose interventional
therapies. In spite of the clear involvement of androgens and other sex steroids in the
pathogenesis of prostate cancer, the levels of these hormones are not useful in screening or
diagnosis (363). Prostate-specific antigen (PSA) is an important tool for screening patients for
prostate cancer, but PSA has a number of limitations. The lack of specificity of PSA for prostate
cancer has resulted in unnecessary biopsies and an over-diagnosis of indolent prostate cancer. In
fact, only about 25% of men who undergo a prostate biopsy due to an elevated PSA level have
prostate cancer (364). Secondly, despite a “normal” PSA, men can still harbor prostate cancer

105
(365). Taken together, the development of novel biomarkers that can improve the accuracy of
screening is becoming increasingly important (366).
The vertebrate mitochondrion encodes 13 important OXPHOS proteins and serves as the cellular
center of energy production. In many cancer cells, “aerobic glycolysis” occurs instead of
maximal ATP generation by OXPHOS, suggesting mitochondrial dysfunction. One potential
source of new biomarkers may be derived from the mitochondria, given that many of the key
aspects of malignancy involve changes in mitochondrial energy metabolism and resistance to
mitochondrial apoptosis (367,368). Furthermore, specific mutation and deletion patterns in the
mtDNA have been associated with various types of cancer including prostate cancer (215,369).
However, mtDNA might not be a convenient biomarker for prostate cancer as very little
neoplastic DNA is shed from the prostate epithelium to the urine or blood, thus making the
mutated mtDNA dilute and difficult to detect (197). Moreover, mitochondrial cancer mutations
must lead to altered signaling in the nucleus to have biological impact, a process called
retrograde signaling (370). Hence, the ideal prostate cancer biomarker should be a mitochondrial
retrograde signaling molecule that is directly involved in neoplastic transformation. Our group
identified several peptides encoded by sORFs in the mtDNA (183) and each of them possesses
important biological functions (72,371). We recently reported the expression and biological
significance of six additional small peptides encoded within the mitochondrial 16S rRNA region,
referred to as SHLPs”. Among these, SHLP2 has potent retrograde signaling effects on reducing
ROS production and improving mitochondrial metabolism in vitro as well as protective actions
in vivo. In addition to their retrograde signaling function, these MDPs may act as hormones, for
all the currently discovered MDPs are transported by the circulatory system and actively
involved in metabolism. For example, humanin regulates glucose homeostasis via hypothalamic

106
STAT3 activation and is a new player in the GH/ IGF axis (251,272,372). MOTS-c appears to
regulate glucose metabolism and muscle insulin action (139). We believe SHLP2 also represents
a novel mitochondrial hormone as it belongs to the MDP family and can influence mitochondrial
respiration. Moreover, we developed a SHLP2 ELISA assay and showed that its levels decline
with age. Taken into account the importance of mitochondria in prostate cancer progression and
the role of SHLP2 as a mito-enhancing hormone, we hypothesized that SHLP2 may be
associated with prostate cancer risk and possibly serve as a novel biomarker for this disease.
SHLP 1-6 were reported as circulating peptides of approximately 20 amino acids in length and
are encoded within the same 16S rRNA in which humanin is located. SHLP2 exhibited similar
effects as humanin, in terms of anti-apoptosis, insulin sensitization and maintenance of glucose
homeostasis. Furthermore, in vitro SHLP2 treatment also promoted mitochondrial health by
increasing OCR and ATP generation. It is also reported that SHLP2 played a key role in the
development and racial disparity of prostate cancer, as low levels of SHLP2 were linked with
increased prostate cancer in white men. Since MDPs are made from the mitochondrial genome to
carry on crucial signaling tasks, they are involved in a broad range of metabolic events. So far,
most of the studies on MDPs had focused on one disease model or one biological effect at a time.
This targeted research approach which, although allows deep investigation, may nevertheless
provide limited information. Humanin and SHLP2 have been shown to enhance metabolic
fitness, and exhibit protection against metabolic syndromes such as cardiovascular diseases,
T2DM and prostate cancer. It is intriguing that many of these diseases are related to obesity,
therefore we speculate humanin and humanin-like peptides effect changes in obesity-induced
pathobiology. In this study, we administered potent humanin analog HNG or SHLP2 in DIO

107
mice. Blood samples were collected, and plasma were sent for metabolome analysis to obtain
comprehensive understanding of the changes in the metabolite profile and metabolic pathways.
In this chapter, we focus on the initial characterization of the mitochondrial-peptide SHLP2, the
similarities and dissimilarities of effects between it and humanin. Like humanin, SHLP2 has
anti-apoptotic activity and can physically interact with amyloidogenic peptide. Moreover, the
change in metabolite profile in response to SHLP2 treatment is similar to that of humanin. Given
the fact that they are both transcribed from the 16S rRNA, SHLP2 and humanin might belong to
the same protein family.

108
Results
Identification and Validation of Small Open Reading Frames (sORFs) Within
Mitochondrial 16S rRNA
An in silico search for potential sORFs within the 16S rRNA with cut-offs between 20 - 40
amino acids returned six potential sequences encoding for peptides ranged from 20-38 amino
acid (Figure 3-1 A, B), which we have named SHLP1-6. To test endogenous expression, we
generated antibodies for SHLPs 1, 2, 3, 4, and 6 and used them to detect the expression of
SHLPs in mouse tissues. Despite multiple attempts, we were not able to obtain a specific
antibody against SHLP5. The expressions of SHLP1-4 and 6 were detected in multiple mouse
tissues at varying levels (Figure 3-1 C). We observed an organ-specific pattern of individual
SHLP abundance; SHLP1 in heart, kidney and spleen, SHLP2 in liver, kidney and muscle,
SHLP3 in brain and spleen, SHLP4 in liver and prostate, and SHLP6 in liver and kidney. To
further confirm the mitochondrial origin of SHLPs, we compared SHLP1-4 and 6 expressions
between HeLa cells that are selectively devoid of mitochondrial DNA (HeLa-ρ0) and their
parental cells (373). SHLP1-4 and 6 as well as the other mitochondrial-encoded proteins COI
and COII were detected in parental HeLa cells but not in HeLa-ρ0 cells (Figure 3-1 D). This
suggests a mitochondrial origin for these peptides, although we cannot completely rule out the
possibility that these effects were secondary to the loss of mitochondria. To further support the
mitochondrial origin of the SHLP peptides, we next probed for the presence of total and
polyadenylated mRNA transcripts transcribed from the 16S rRNA region of the mtDNA using a
Northern blot of: (a) whole cell RNA, (b) whole cell mRNA, (c) mitochondrial RNA, and (d)
mitochondrial mRNA. We were able to detect several specific mitochondrial mRNA transcripts
of various sizes that are smaller than the full-length rRNA transcript indicating the presence of

109
mRNA transcripts within the 16S region (Figure 3.1 E).

Figure 3-1 Identification and Validation of Small-Open-Reading-Frames (sORFs) Within the
Mitochondrial 16S Ribosomal RNA (rRNA) Gene
(A) Assigned names, size, location, and predicted sequences of SHLPs. (B) Location of SHLP
ORFs within 16S rRNA. (C) Relative expression of SHLPs (1–4 and 6) in C57BL/6 mouse
tissues relative to GAPDH (used as a loading control). (D) Expression of SHLPs (1–4 and 6),
mitochondrial complexes (COI and COII) and nuclear GAPDH proteins in HeLa parental and
HeLa-ρ0 cells. (E) Northern blot of whole cell RNA, whole cell mRNA, mitochondrial RNA,

110
and mitochondrial mRNA using a SHLP2 probe identifies several bands smaller than the 16S
rRNA indicated by the arrows.
The in silico analysis of novel SHLP ORFs, generation of antibodies against SHLP1-6 and the
immunoblots were performed by other Cohen lab members. Mitochondria and RNA extraction,
Northern blot were conducted by me.
SHLP2 is an Insulin Sensitizer Acting both Peripherally and Centrally
SHLP2 and 3 treatment for 7 days in the presence of insulin accelerated the differentiation of
3T3-L1 murine pre-adipocytes, as measured by Oil red-O staining, suggesting that these peptides
promote cellular differentiation and enhance insulin sensitivity in adipose tissue. Adipose tissue
has been implicated in aging and age-related diseases (375). As SHLP2 and 3 increased insulin
sensitivity in vitro, we hypothesized that they may also regulate insulin action in vivo.
Furthermore, because of the likely connection between mitochondria, an energy generator, and
the hypothalamus, an energy regulator, we tested the metabolic consequences of centrally
delivered SHLPs. Intracerebroventricular delivery of humanin regulates peripheral (hepatic)
insulin action (272) and we tested the ability of SHLP2 and 3 to exert similar effects. We
performed systemic pancreatic insulin clamp and physiologic hyperinsulinemic-euglycemic
clamp studies to quantify the glucose flux following studies done with humanin (272). SHLP2
improved insulin responsiveness as reflected by a~50% increase in the exogenous GIR required
to maintain euglycemia during insulin stimulation. The effects of insulin on suppressing hepatic
glucose production as well as on promoting glucose disposal into peripheral tissues was
enhanced by SHLP2. Unlike SHLP2, SHLP3 did not have an effect on insulin action in vivo.
Refer to (186) for more information and figures.
SHLP2 is a Bioactive Peptide that Modulates Mitochondrial Function
Since SHLP2 consistently demonstrated bioactive effects, especially insulin sensitization and
glucose metabolism, we set to study where are the primary sites of SHLP2 action. It has been

111
shown that staurosporine (STS) was used to activate caspase-3, which causes impairment of
mitochondrial membrane potential (374). We have previously established that STS-induced
apoptosis was fully blocked by SHLP2 in NIT-1 b-cells. SHLP2 also increased mitochondrial
OCR and ATP production in 22Rv1 cells in real-time using an XF24 Extracellular Flux Analyzer
(Seahorse Bioscience). Therefore, we focused on studying how SHLP2 affect mitochondrial
membrane potential in b-cells. In fact, mitochondria play essential roles in pancreatic β-cell
function. Mitochondrial energy state, which is in the form of an electrochemical gradient,
commonly termed the mitochondrial membrane potential (ΔΨ). This gradient influences
ATP:ADP ratio, redox state, and ROS. The glucose-stimulated insulin secretion (GSIS) relies on
hyperpolarization of mitochondrial membrane potential (ΔψM) and the stronger protonmotive
force leads to increased mitochondrial production of ATP, which is largely responsible for
insulin secretion. We found that SHLP2 further enhanced glucose-stimulated mitochondrial
membrane hyperpolarization in INS-1 cells (Figure 3-2 A), which is consistent with increased
ATP production and basal respiration observed by using Seahorse Analyzer. Moreover, SHLP2
treatment also reduced H2O2 production from INS-1 cells (Figure 3-2 B). Taken together, these
data suggest that SHLP2 affects mitochondrial function and enhance mitochondrial metabolism.

112

Figure 3-2 SHLP2 Modulates Mitochondrial Membrane Potential and H2O2 release

113
(A) INS-1 β cells were treated with 3mM glucose with or without MDP (HNG or SHLP2, 36µM)
for 1 hour, followed by staining with JC-1 dye. Green fluorescence indicates depolarization, and
red fluorescence indicates hyperpolarization. (B) INS-1 β cells H
2
O
2
release was decreased when
treated with SHLP2. All data are presented as means ± SEM. *p <0.05; **p <0.01; ***p <0.001.
Cell culture and assays performed by me.
SHLP2 Inhibits IAPP Misfolding
Since both humanin and SHLP2 can enhance insulin sensitivity and antagonize amyloid-induced
toxicity, we hypothesize that the two MDPs have the capacity to inhibit the misfolding of islet
amyloid polypeptide. We chose to test the humanin analog HNG, where serine 14 is modified to
a glycine residue, because of its significantly enhanced neuroprotective activity (92,94,95). We
first monitored IAPP misfolding in the presence of HNG or SHLP2 using thioflavin T. For this
experiment the IAPP concentration was held constant (at 12.5 µM) and the MDP concentrations
were varied. To avoid complications from potential aggregation of HNG and SHLP2, we
followed the peptide handling protocol first described by Arakawa et al. 2011 (376) and only
used fresh stocks of MDPs. According to our ThT data, both MDPs inhibit IAPP aggregation.
We find that IAPP fibrilization, in the presence of either HNG or SHLP2, is reduced in a dose-
dependent manner (Figure 3-3 A–C). In addition, kinetic analysis of IAPP misfolding
demonstrates that both MDPs slow or entirely prevent the misfolding of IAPP within the
timeframe of the experiment in a similarly dose-dependent manner (Supplemental Table 1a,b).
HNG exhibited the greater potency, essentially completely inhibiting IAPP misfolding at
substoichiometric concentrations, whereas closer to stoichiometric concentrations of SHLP2
were required. The remarkable ability of HNG to perform at substoichiometric concentrations
(almost full inhibition at a molar ratio of 1:250, HNG: IAPP) implies that it is unlikely for HNG
to act on the bulk of the monomeric IAPP to inhibit aggregation. Although binding to monomers
could slow down aggregation by reducing the monomer pool available for misfolding, the

114
strongly substoichiometric ratios would only allow a small subset of the IAPP molecules to be
bound by HNG molecules. Such a minor reduction in available free IAPP would have negligible
effects on the misfolding kinetics. SHLP2 also inhibits at substoichiometric concentrations,
however the effect is not nearly as pronounced as in the case of HNG. Therefore, we performed
additional biophysical measurements to more clearly determine whether HNG, as well as
SHLP2, act on species other than monomeric IAPP.

Figure 3-3 Mitochondrial-Derived Peptides HNG and SHLP2 Inhibit IAPP Fibrilization

115
(A and B) Representative ThT kinetics traces of IAPP misfolding in the presence of a) HNG or b)
SHLP2. (C) End point analysis of IAPP misfolding by ThT fluorescence at 18 hours. Trend lines
are shown for clarity. Concentrations of MDPs used to inhibit IAPP misfolding are given in
micromolar below the figure. *p < 0.01. Error bars represent +/− 1 standard deviation from a
minimum of 3 experiments.
Biophysical experiments were performed by the Langen lab. I provided technical advice.
Low Circulating Levels of SHLP2 as Novel Biomarker for Prostate Cancer Risk
Baseline characteristics and serum SHLP2 levels are shown in Table 3-1. There were no
significant differences between cases and controls for age, body mass index (BMI), and digital
rectal examination (DRE). Of the cases, there were 63 low-grade (Gleason score ≤6) and 37
high-grade (Gleason score 7-10).
Table 3-1 Baseline Characteristics in Cases and Controls

Cases
(N=100)
Controls
(N=100)
Total
(N=200)
p value
Age 0.125
1

Median 63 61.5 62
Q1, Q3 60, 66 59, 65 59, 66

Race 0.888
2

White 49 (49%) 50 (50%) 99 (50%)
Black 51 (51%) 50 (50%) 101 (50%)

Year of consent <0.001
1

Median 2010 2008 2008
Q1, Q3 2009, 2011 2008, 2008 2008, 2010

BMI (kg/m2) 0.150
1

Median 28.3 29.8 28.9
Q1, Q3 25.4, 31.3 25.9, 32.9 25.6, 32.1

PSA (ng/ml) <0.001
1

Median 7.1 5.5 6.0
Q1, Q3 5.0, 13.4 4.4, 7.5 4.7, 9.1

116

Cases
(N=100)
Controls
(N=100)
Total
(N=200)
p value

Grade group
Missing 0 100 100
1 63 (63%) 0 (0%) 63 (63%)
1 14 (14%) 0 (0%) 14 (14%)
2-3 23 (23%) 0 (0%) 23 (23%)

DRE 0.404
2

Normal 75 (75%) 81 (81%) 156 (78%)
Suspicious 24 (24%) 19 (19%) 43 (22%)
Unknown 1 (1%) 0 (0%) 1 (1%)

Prostate Volume (c
c)
0.009
1

Median 36.5 45.5 40.0
Q1, Q3 27.0, 56.8 34.0, 69.5 28.5, 63.1

Family history of P
C
<0.001
2

No 54 (54%) 78 (78%) 132 (66%)
Yes 20 (20%) 17 (17%) 37 (19%)
Unknown 26 (26%) 5 (5%) 31 (16%)

SHLP2 <0.001
1

Median 241 361.5 268.5
Q1, Q3 205, 270 261, 421.5 218, 363

SHLP2 <0.001
2

<=350 99 (99%) 43 (43%) 142 (71%)
>350 1 (1%) 57 (57%) 58 (29%)
1
Wilcoxon
2
Chi-Square

117
Table 3-2 Association between SHLP2 and other variables (SHLP2 cut-off at 350-pg/ml)
<=350
(N=142)
>350
(N=58)
p value
Age 0.335
1

Median 63 61
Q1, Q3 59, 66 59, 65

Race 0.010
2

White 62 (44%) 37 (64%)
Black 80 (56%) 21 (36%)

Year of consent <0.001
1

Median 2009 2008
Q1, Q3 2008, 2011 2007, 2008

BMI (kg/m2) 0.709
1

Median 28.4 29.6
Q1, Q3 26.0, 32.0 25.2, 32.8

PSA (ng/ml) 0.004
1

Median 6.6 5.0
Q1, Q3 4.9, 9.5 4.4, 7.4

Grade group
No cancer 43 57
1 63 (64%) 0 (0%)
2-3 13 (13%) 1 (100%)
4-5 23 (23%) 0 (0%)

DRE 0.802
2

Normal 111 (78%) 45 (78%)
Suspicious 30 (21%) 13 (22%)
Unknown 1 (1%) 0 (0%)

Prostate Volume (cc) 0.180
1

118
<=350
(N=142)
>350
(N=58)
p value
Median 38.0 44.5
Q1, Q3 27.7, 62.4 31.4, 70.0

Family history of PC 0.077
2

No 87 (61%) 45 (78%)
Yes 29 (20%) 8 (14%)
Unknown 26 (18%) 5 (9%)

Diabetes 0.884
2
No 73 (52%) 32 (55%)
Yes 23 (16%) 9 (16%)
Unknown 46 (32%) 17 (29%)

Status <0.001
2

Controls 43 (30%) 57 (98%)
Cases 99 (70%) 1 (2%)
1
Wilcoxon
2
Chi-Square
In addition, we analyzed the associations between SHLP2 and other variables (Table 3-2). Men
with SHLP2 ≤350-pg/ml were more likely to be black (P =0.010), had more recent biopsies (P
<0.001), and higher PSA (P =0.004), compared to men with SHLP2 >350-pg/ml. The serum
levels of SHLP2 were higher in controls than cases (341 vs. 230-pg/ml, P <0.001). When
stratified by race (Figure 3-4 A), SHLP2 was higher in white controls vs. white cases (393 vs.
196-pg/ml, P <0.001), while the difference between black controls vs. black cases was smaller
and not significantly different (289 vs. 261-pg/ml, P =0.093). SHLP2 was similar between white
and black cases (296 vs. 276-pg/ml, P =0.17); however, SHLP2 was higher in white than black
controls (393 vs. 290-pg/ml, P <0.001) (Figure 3-4 B). There was no association between SHLP2
and prostate cancer grade (Figure 3-4 B).

119

Figure 3-4 The Distribution of SHLP2 Levels Stratified by Race or Outcome
(A) When stratified by race, mean SHLP2 was significantly higher in white controls vs. white
cases (393 vs. 196 pg/ml, p <0.001), while the difference in SHLP2 between black controls vs.
black cases was smaller and not statistically significant (289 vs. 261 pg/ml, p = 0.093). (B) When
stratified by outcome, average value of SHLP2 was significantly higher in white control than
black control (393 vs. 290, p <0.001). However, SHLP2 levels were not different between white
and black in all cancer groups.
The p-interaction between race and SHLP2 for predicting prostate cancer was 0.006. Thus,
analyses were stratified by race. On multivariable analysis, lower SHLP2 was linked with
increased risk of overall, low-, and high-grade prostate cancer in white men (all P <0.038), but
none of these associations even approached significance in black men (all P >0.37) (Table 3-3).
After adjusting for family history of prostate cancer and prostate volume, results were similar in
black men but stronger in magnitude among white men, albeit the associations had larger
confidence intervals and were not statistically significant with the exception of high-grade
prostate cancer (Supplemental Table 2). A SHLP2 cut-off of 350-pg/ml separated prostate cancer
from controls in both black and white men (Figure 3-5). Among men with SHLP2>350-pg/ml,
0/37 white (100% NPV) and only 1/20 black men had prostate cancer (95% NPV), which was a
Gleason 7. Using a SHLP2 cut-off>350-pg/ml to not biopsy would have avoided 57/100 negative
biopsies while missing one Gleason 7.

120
Table 3-3 Association Between SHLP2 and Overall Risk of Cancer and Risk of Cancer Grade,
Stratified by Race
White Black
OR 95% CI P-value OR 95% CI P-value
Univariable
No PC Ref. Ref.
Overall PC 0.95 0.93-0.97 <0.0001 0.996 0.99-1.00 0.095

No PC Ref. Ref.
Low-grade PC 0.95 0.93-0.97 <0.001 0.99 0.90-1.00 0.044
High-grade PC 0.95 0.93-0.97 <0.001 1.00 0.99-1.00 0.598
Multivariable*
No PC Ref. Ref.
Overall PC 0.92 0.84-0.995 0.037 0.998 0.99-1.01 0.706

No PC Ref. Ref.
Low-grade PC 0.91 0.83-0.99 0.038 1.00 0.98-1.01 0.377
High-grade PC 0.91 0.83-0.99 0.031 1.00 0.99-1.02 0.671

121

Figure 3-5 The Distribution of SHLP2 Levels and a Cut-Off at 350-pg/ml
A cut-off of 350-pg/ml SHLP2 differentiate between controls and PCa cases in both black and
white men. Among men with SHLP2>350-pg/ml, 0/37 white (100% NPV) and only 1/20 black
men had PCa (95% NPV).
The area under the curve (AUC) of the model including only age, DRE, race and PSA (standard
clinical variables) to predict prostate cancer risk was 0.67. This improved to 0.85 when SHLP2
was added to the model (p<0.001). Moreover, there was a substantial improvement in AUC for
white men (0.72 to 0.99, p<0.001) but only a slight increase in AUC for black men (0.70 to 0.72,
p=0.47) after SHLP2 was added to the model including age, DRE, and PSA (Figure 3-6).

122

Figure 3-6 ROC Curve and AUC Statistics Before and After Adding SHLP2 in the Model
The true positive rate (sensitivity) is plotted in function of the false positive rate (1−specificity)
for the model excluding or including SHLP2 levels. The AUC is a measure of how well a
quantitative test can distinguish between subjects with and without prostate cancer. The AUC of
the model including only age, DRE, race and PSA to predict PCa risk was 0.67. This improved to
0.85 when SHLP2 was added to the model (p <0.001).
Patients were recruited by Durham General Hospital; serum samples were sent to the Aging
Biomarker Core for SHLP2 measurement. Lauren Howard from Duke University provided
patient baseline characteristics. Biomarker analysis was done by me.
Multivariate Analysis of Metabolomic Profile Alterations in Response to SHLP2 Treatment
The present dataset comprises a total of 549 compounds of known identity (named
biochemicals). Plasma levels of 52 biochemicals changed significantly (14 rose and 38 fell) in
response to HNG treatment, while 77 biochemicals changed significantly (16 rose and 61 fell) in
response to SHLP2 treatment. Figure 3.12 A is the Principle Component Analysis (PCA) taking
into account the data acquired on the 549 biochemicals. PCA permits visualization of how

123
individual samples, within a group, cluster with respect to their “principle components”
calculated from the relative changes in the 549 metabolite levels. As such, this analysis tool aids
in determining if the plasmas from peptide-treated mice can be differentiated from control, or
from each other, based on differences in their overall metabolite signature. Comparing control to
all peptide-treated samples, there appears to be some degree of segregation on the x-axis
(component 1). There was substantial variability across all groups, which may reflect animal to
animal variability in response to peptide. However, Hierarchical clustering analysis displays
stronger evidence of peptide specific effects. Hierarchical clustering is used to compare
similarities and differences between metabolite profiles. When using the changes in all 549
biochemicals, the control mice were nicely clustered while the humanin and SHLP2 groups
exhibit more similar pattern of changes (Figure 3-7 B).

124

Figure 3-7 Principle Component Analysis (PCA) of Selected Metabolic Pathways and
Hierarchical Clustering of All Measured Metabolites
The four pathways responded most strongly to peptide treatment are methionine, cycsteine, SAM
and taurine metabolism, glutathione metabolism, gamma-glutamyl amino acid, and sphingolipid
metabolism. (A) the PCA taking account into the metabolites responded most strongly to MDP
treatment. (B) hierarchical clustering analysis of the whole metabolome, red color indicates the
abundance of metabolite was up-regulated compared to that of the control and green color
indicates down-regulation (see ratio color key).
Random Forest (RF) analysis attempts to bin individual samples in groups based on their
metabolite similarities and differences. Random Forest also defines which metabolites contribute
most strongly to the grouping process. (Figure 3-8 A) presents Random Forest results for the
control and peptide-treated plasmas. RF was able to properly bin 4 of 6 control samples

125
(predictive accuracy 67%), which is an indication that control sample profiles were unique from
the peptide-treated samples. (Figure 3-8 A) also displays a plot of the top 30 Mean Decrease
Accuracy values calculated for the comparison of all five sample groups. A higher Mean
Decrease Accuracy value indicates a greater group differentiating contribution. Since we did not
observe a clear distinction between HNG and SHLP2 treatment, we performed a second RF test
taking HNG and SHLP2 as a single “Peptide treated” group. (Figure 3-8 B) demonstrated that
RF properly bin 11 of 12 treatment samples. The metabolite with the highest mean-decrease-
accuracy value were 2-aminobutyrate and 1-(3-aminopropyl)-2-pyrrolidone. The former is a
metabolite generated in the process of transsulfuration to generate cysteine. Statistical
comparison of 2-aminobutyrate levels in control vs. each of the peptides generated the lowest p-
values among all the measured metabolites. Amino acid and lipid metabolites were highly
represented in the top 30. Furthermore, we found that the metabolites from sphingolipid
metabolism, glutathione and gamma glutamyl transferase (GGT) metabolism and methionine,
cysteine, S-adenosylmethionine (SAM) and taurine metabolism were overrepresented, which
suggests that these biological pathways are most influenced by humanin and SHLP2 (Figure 3-8
A). We also did a PCA including only the fold changes of metabolites in the aforementioned
metabolic pathways and found that the control animals were better separated from the peptide-
treated animals (Figure 3-8 B).

126

Figure 3-8 Random Forest Classification Using Named Metabolites in Plasma of Control
Compared to Plasma of HNG and SHLP2 Treated Mice
The dot chart displays the top 30 Mean Decrease Accuracy values calculated for (A) the
comparison of all three sample groups. A higher Mean Decrease Accuracy value indicates a
greater group differentiating contribution. Amino acid and lipid metabolites were highly
represented in the top 30. (B) the comparison between control group and peptide treatment group
(Humanin and SHLP2).

127
Changes in Metabolite Profile with HNG or SHLP2 Treatment
Overall, we can group the altered metabolites based on their biological pathways into 4
categories: (1) The methionine cycle and glutathione metabolism: The metabolites alpha-
ketobutyrate, 2-hydroxybutyrate (an isobar of with 2-hydroxyisobutyrate where 2-
hydroxybutyrate predominates) and 2-aminobutyrate (top metabolite in RF analysis above)
decreased in response to peptide administration, which may be indicative of reduced cysteine
synthesis from cystathionine. However, plasma cysteine levels were relatively consistent
between groups, as were methionine and cystathionine. The generation of cysteine from
methionine through the transulfuration pathway occurs primarily in liver and kidneys. It is
possible that the secretion of methionine, cystathionine and cysteine into plasma is not affected
by subtle changes in the syntheses. Betaine, which is utilized to convert homocysteine back to
methionine, was lower in all peptide-treated plasma, possibly consistent with greater recycling of
homocysteine and less homocysteine conversion to cystathionine to support cysteine and
glutathione synthesis. Oxidized glutathione was lower to a statistically significant degree in
SHLP2 treated mice but not in HNG treated mice, relative to control (Figure 3-9)

128

129

130
Figure 3-9 The Metabolite Profile of the Transmethylation, Transulfuration and Gamma-
Glutamyl Cycle
(A) the schematic diagram of the transmethylation, transulfuration pathway and gamma-glutamyl
cycle. Yellow box indicates metabolite not measured in the metabolomic discovery, green box
indicates metabolite measured but did not change in response to peptide treatment, blue box
indicates metabolite that was down-regulated compared to control. (B) box plots of selected
biochemicals. Values are normalized raw area counts rescaled to set the median equal to 1.
(2) Gamma-glutamyl-amino acid: the plasma levels of several gamma-glutamyl amino acids
were lowered, including gamma-glutamyl glutamate, gamma-glutamyl glutamine and gamma-
glutamyl histidine. In addition, the concentration of the downstream metabolite 5-oxoproline also
decreased. Gamma-glutamyl amino acids are made through adding a gamma-glutamyl functional
group from molecules such as glutathione to an amino acid and this reaction is catalyzed by the
enzyme GGT. This result suggests that humanin and SHLP2 suppress gamma-glutamyl-amino
acid cycle through (1) down-regulation of GGT activity; (2) reducing availability of glutathione.
However, the latter hypothesis is less likely, since the levels of cysteine, glutamate, and glycine
were not affected (Figure 3-9).

131

Figure 3-10 Major Intermediates in the Gamma-Glutamyl Cycle Decreased, Suggesting
Reduction in Oxidative Stress Burden
(A) The detailed pathway of the gamma-glutamyl cycle. Yellow box indicates metabolite not
measured in the metabolomic discovery, green box indicates metabolite measured but did not
change in response to peptide treatment, blue box indicates metabolite that was down-regulated
compared to control. Grey box indicates key enzyme responsible for the metabolite conversion.

132
(B) Mitochondrial H
2
O
2
release was decreased when treated with HNG or SHLP2. Values are
mean ± SEM. * p <0.05 *** p < 0.001.
(3) Sphingolipid metabolism: sphingolipids play significant roles in cell membrane and provide
many bioactive metabolites that regulate cell function. From this metabolome study, we observed
consistent modulation induced by these MDPs, which was a reduction in multiple components of
the sphingolipids pathway. Sphinganine, which is a precursor to ceramide, decreased
significantly in HNG and SHLP2 treated plasma, relative to control. This may suggest a
reduction in synthesis in tissues. Furthermore, two glycosylated ceramides, glycosyl-N-
palmitoyl-sphingosine and glycosyl-N-steroylsphingosine, displayed strong decreases with both
peptides. Multiple types of sphingomyelin also exhibited significant reduction. Sphingomyelin is
synthesized from the enzymatic transfer of a phosphocholine to a ceramide with diacylglycerol
being produced as a byproduct. We also observed trend of decrease in plasma diacylglycerol in
response to HNG/SHLP2 treatment, which suggests that the synthesis of sphingomyelin from
ceramide was impeded (Figure 3.11)

133

134
Figure 3-11 The Metabolite Profile of the Sphingolipid Metabolism
(A) the schematic diagram of the sphingolipid pathway. Yellow box indicates metabolite not
measured in the metabolomic discovery, green box indicates metabolite measured but did not
change in response to peptide treatment, blue box indicates metabolite that was down-regulated
compared to control. (B) box plots of selected biochemicals. Values are normalized raw area
counts rescaled to set the median equal to 1.
Animal experiment was led by Kelvin Yen, plasma samples were processed and metabolites
were quantified by Metabolom Inc. Data analyses, plotting of graphs were done by me.

135
Discussion
Our data unravel a novel function of the mitochondria: to produce specific MDPs that regulate
cellular processes and human disease in an age-dependent manner. Specifically, age-dependent
changes in the SHLP levels and other undiscovered MDPs might play a role in the development
of age-related diseases. It is well recognized that multiple signals, mostly in the form of imported
proteins, are delivered to the mitochondria to regulate its functions; however, the nature of
mitochondrial retrograde signaling remains controversial. Although mitochondrial ROS or
degraded protein products have been proposed to play such a role, especially in model
organisms, they do not fully account for the complexity of tissue-specific mito-chondrial
signaling in more complex organisms. We propose that mitochondria control the expression of
MDPs that have specific activities in response to three major mitochondrial functions: apoptosis,
metabolism, and oxidative stress. Our findings are consistent with recent reports (377,378) that
suggested that unknown factors termed ‘mitokines’ are produced from the mitochondria of
Caenorhabditis elegans and secreted to communicate with other cells. Further investigations into
the extracellular and intracellular roles of SHLPs are required. Furthermore, although SHLPs 1–6
all originate from 16S rRNA, the various effects of individual SHLPs suggests that their
expression is complex. We believe that each SHLP may activate its own unique receptor, leading
to differential effects.
Implications of SHLP2 in Cellular Function in Age-Related Diseases
SHLPs are derived from the mitochondrial genome within the 16S rRNA region, and we showed
that they can regulate cell growth and function in a SHLP-specific manner. Like humanin,
SHLP2 and SHLP3 improved cell survival in response to toxic insults and prevented apoptosis,
while SHLP6 had the opposite effect. Programmed cell death in response to extrinsic or intrinsic

136
death signals is critical for tissue homeostasis. Age-related accumulations in cellular damage
may lead to excessive cell death, limiting tissue function and life span. Increased apoptosis has
been observed in neurodegenerative diseases such as AD, Parkinson's disease, and Huntington's
disease (379). Furthermore, SHLP2 and SHLP3 showed similar protective effects as humanin,
and significantly blocked cell death induced by STS. SHLP2 treatment alone also protected
against Aβ1–42 induced cell death, which contributes to AD. Together this data suggests that
SHLP2 reduction with aging may participate in the pathogenesis of age-related
neurodegenerative diseases.
SHLP2 is a Central and Peripheral Insulin Sensitizer
Humanin has been reportedly a central regulator of insulin action (272). Continuous
intracerebroventricular CV infusion of humanin significantly improved insulin sensitivity, and a
single dose of the humanin analogue HNGF6A significantly decreased blood glucose levels in
Zucker diabetic rats. Similarly, in rats, continuous intracerebroventricular infusion of SHLP2
significantly improved insulin sensitivity by increasing the GIR, suppressing HGP, and
increasing peripheral glucose uptake in hyperinsulinemic-euglycemic clamp studies. This
activity of SHLP2 is particularly significant because many centrally acting peptides that affect
hepatic glucose metabolism (e.g., leptin, insulin, and IGF-I) have not been shown to increase
peripheral glucose uptake. These results suggest that SHLP2 is an MDP that can communicate
with the hypothalamus, under conditions that are yet to be determined, leading to a change in
peripheral metabolism. In the future, SHLP2 may have a role in the treatment of diabetes
because of its insulin-sensitizing effects.

137
SHLP2 in Aging
Mitochondria have been implicated in increased lifespan in several life-extending treatments
(380,381); however, it is not known whether the relationship is correlative or causative (381).
Additionally, it is well known that hormone levels change with aging. For example, levels of
aldosterone, calcitonin, growth hormone, and IGF-I decrease with age. Circulating humanin
levels decline with age in humans and rodents, specifically in the hypothalamus and skeletal
muscle of older rats. These changes parallel increases in the incidence of age-associated diseases
such as AD and T2DM. The decline in circulating SHLP2 levels with age, the anti-oxidative
stress function of SHLP2, and its neuroprotective effect indicate that SHLP2 has a role in the
regulation of aging and age-related diseases.
SHLP2 and IAPP Interaction
We demonstrated that the mitochondrial-derived peptides HNG and SHLP2 can prevent the
amyloid formation of IAPP. Multiple lines of evidence indicate that the mechanism underlying
this activity is the binding of misfolded IAPP seeds by the two MDPs. The strongly
substoichiometric inhibition of IAPP misfolding by HNG rules out the necessity for bulk capture
of monomeric IAPP by HNG. Moreover, the EPR data do not detect any binding interaction
between free monomeric IAPP and either MDP, while the CD data reveal that mixing of IAPP
and MDPs does not result in detectable changes in secondary structure for either of the peptides.
These data are inconsistent with any substantive binding of the free pool of naïve bulk IAPP by
either MDP. Binding to a small subset of seeding competent monomers could, at least in
principle be possible. However, there is no evidence that such monomers exist for IAPP and it is
well-known that multimeric species, like the ones used here, can be potent seeds. In fact, we
observe direct interactions between the MDPs and multimeric IAPP misfolding seeds by co-

138
sedimentation. This interaction is sufficient to robustly prevent seeded misfolding reactions
indicating that these MDPs have the capacity to prevent such seeds from functioning as a
template for the misfolding of naïve IAPP. Consistent with this notion, our CD and EPR data
further demonstrate that the hallmark cooperativity normally observed in IAPP amyloid
formation reactions can be blocked by MDP treatments.
The chaperone-like function of the MDPs observed here also provides an additional likely
mechanistic explanation to findings from studies showing that humanin protects in the setting of
Aβ toxicity and AD pathology. Indeed, it was recently found that humanin and HNG interact
with Aβ oligomers (385). We speculate that such chaperone-like activity, when taken in context
with the known neuroprotective effects of humanin, raises the possibility that humanin and other
MDPs serve a role as a sensor of, or a response to, aberrant protein or amyloid misfolding. If so,
this would expand the function of mitochondria in amyloid disease beyond its apoptotic role to
that of an early component of the cellular defense system.
The data presented here also highlight the therapeutic potential of HNG, SHLP2 or related
molecules in the treatment of T2DM. Our data reveal that interactions of MDPs with IAPP have
a marked degree of specificity for non-monomeric IAPP species. This quality has two beneficial
aspects. First, it is likely to reduce interactions that prevent IAPP monomers from performing
their regular physiological function, and second, by inhibiting seeding only a small amount of
misfolding inhibitor is needed. Finally, it will be important to investigate the degree to which all
the members of the MDP family have similar chaperone-like qualities to protect against various
forms of protein misfolding and which oligomerization state mediates their chaperone-like
activities.

139
SHLP2 as A Novel Prostate Cancer Biomarker
Cancer cell metabolism has attracted increasing attention as a promising area of cancer study.
One of the first pieces of evidence suggesting a link between mitochondrial metabolism and
cancer is the excessive production of lactate and glycolytic metabolism of tumor tissues (388).
Later studies were carried out to determine whether this aerobic glycolysis was due to
mitochondrial OXPHOS dysfunction, which generally yielded negative results (389). In addition
to mtDNA mutations/deletions detected in cancer cells, mutations in nuclear-encoded
mitochondrial enzymes are also associated with tumorigenesis (390). These findings suggest
neoplastic transformation taking place in mitochondria do not only involve defects in
mitochondrial energy production but also altered mitochondrial bioenergetics and metabolomics.
As for prostate cancer, the aggressive phenotypes of cancer cells are associated with metabolic
transformation to aerobic glycolysis, citrate oxidizing, and loss of ability to accumulate zinc
(391,392). Therefore, identifying the key signaling events leading to reprogrammed metabolism
in cancer cells and the tumor environment might be more useful for developing cancer
therapeutics and reliable biomarkers. Our group has been focusing on the mitochondrial
retrograde signaling and how it can impact metabolism. Previously reported mitochondrial
retrograde signals include small molecules such as Ca
2+
, reactive oxygen species and NADH.
We further investigated mtDNA for the possibility of translating small peptides from sORFs and
identified several hormone-like peptides with retrograde signaling functions. These MDPs are
secreted in circulation and play important roles in the cell.
SHLP2 activates ERK and STAT3 signaling pathways via an unknown receptor. It is a regulator
for mitochondrial respiration and ROS production. Moreover, administration of SHLP2
improved insulin sensitivity in rodent models (73). In this case-control study, we demonstrated

140
that serum levels of the mitochondrial protein SHLP2 >350 pg/ml are strongly predictive of a
negative biopsy in both black and white men undergoing prostate biopsy. While we have no
direct explanation for the relationship between SHLP2 and cancer development, and our results
require confirmation in other studies, if validated the observed association between SHLP2 and
prostate cancer risk suggests a possible role for SHLP2 in the development of prostate cancer.
SHLP2 is both mitochondrially encoded and regulates mitochondrial function; therefore, its
levels may be an indicator of mitochondrial integrity and health, preventing carcinogenesis. As
such, low SHLP2 levels may represent a more generalized mitochondrial dysfunction that could
represent a pre-malignant state.
There are several alternative hypothetical explanations for the relationship between low SHLP2
and prostate cancer that we can speculate about. First, during cancer development, the prostate
microenvironment co-evolves with the tumor and facilitates the acquisition of its malignant
phenotype. It is possible that the low levels of SHLP2 contribute to a favorable milieu for tumor
growth by shifting cells towards aerobic glycolysis (393–395). Moreover, the shift in
mitochondrial metabolism and energetics may contribute to the cancer-initiating characteristic of
prostate cancer stem cells (396). Previous study also demonstrated that the mtDNA-depleted
prostate cancer cells exhibit cancer stem cell features (397). Second, chronic inflammation is
closely associated with prostate cancer etiology. SHLP2 promotes mitochondrial function and
may support healthy immune surveillance as well as suppress pro-inflammatory signaling. Since
SHLP2 has been shown to reduce mitochondrial ROS (73), lower SHLP2 levels may lead to
elevated oxidative stress, which could potentially be a major source of mitochondrial genome
instability leading to mitochondrial dysfunction. As a result, mitochondrial dysfunction could
activate the redox-sensitive NF- κB and trigger the release of inflammatory cytokines (398).

141
Alternatively, SHLP2 may lower prostate cancer risk through its effects on enhancing insulin
sensitivity. Insulin resistance is known to be associated with a higher risk of prostate cancer, and
may exacerbate cancer through upregulation of inflammation, oxidative stress and unbound
testosterone (399). Therefore, the insulin sensitizing action of SHLP2 may contribute to its
protective effects against prostate cancer. The continuous intra-cerebro-ventricular (ICV)
infusion of SHLP2 improved hepatic glucose metabolism and peripheral glucose uptake, which
indicates that this centrally acting peptide also triggers a change in peripheral insulin
responsiveness (73). Although the signaling events linking SHLP2 and glucose metabolism are
unknown, it is possible that this humanin-like molecule activates STAT3 and AKT pathways
(372). Furthermore, SHLP2 maintains mitochondrial integrity and may prevent insulin resistance
by reducing mitochondrial ROS generation. prostate cancer risk has been reported to be related
to circulating IGF-I levels (400), which are inversely related to the levels of the mitochondrial
peptide humanin (160). As SHLP2 and humanin share similarities, it is also possible that higher
SHLP2 levels are correlated with lower IGF-I levels, thus predicting lower prostate cancer risk.
Our observation of differential regulation of SHLP2 levels in black and white men coupled with
the known differential risk for prostate cancer incidence and mortality rates between races is
intriguing.
Endocrine involvement in prostate cancer has been well recognized (401), with sex steroids,
growth factors, and cytokines playing a role. The involvement of mitochondria in this disease is
also recognized (402), and the current study implicates mitochondrial-derived peptides. An
intriguing finding is that the black controls have lower SHLP2 levels than white controls. It is
possible that the lifestyle differences between races, such as exercise and diet composition, can
affect mitochondrial functions and influence metabolism. The lower SHLP2 levels of black

142
controls might be due to impaired mitochondrial activities. As for genetic factors, the different
mitochondrial haplogroups not only identify major branches points on the phylogenetic tree, but
also demonstrate differences in OXPHOS activities, mtDNA copy number and mitochondrial-
encoded gene expression as evidenced by a number of cybrid studies (213). Moreover, mtDNA
of black men may harbor SNPs that directly suppress SHLP2 expression. In fact, health disparity
among black and white men with prostate cancer has always been an issue. The racial difference
of SHLP2 levels in control men may provide partial explanation to prostate cancer ethnic
disparity. Furthermore, evidence for the contribution of mtDNA to prostate cancer has been
shown to involve mutations and variations in the region of the mtDNA RNR2/16SrRNA which is
near the site of the SHLP2 ORF (105). Since the causes of prostate cancer involve genetic,
hormonal and environmental elements, additional studies are underway to further discern the
relationship of SHLP2 levels with other prostate cancer risk factors, including obesity and
diabetes, that have been linked to mitochondrial peptides (105). Further investigations are needed
to understand the biological role of SHLP2 in prostate cancer development and progression.
Our study is not without limitations. The number of men included was small and thus validation
is needed. While SHLP2 was not significantly related to prostate cancer risk as a continuous
variable for black men, using a cut-off of >350-pg/ml it had a 95% NPV. If validated, this
suggests examining SHLP2 using cut-off values may have more clinical utility than as a
continuous variable, though we lacked sufficient power to explore multiple cut-points. As such,
future studies are needed to both validate our findings and explore potentially alternative cut-
points. Future studies are needed to explore how SHLP2 correlates with other hormonal levels
thought to be important for prostate cancer such as testosterone. Our outcome was prostate
cancer on biopsy. As some men with a negative biopsy may still harbor prostate cancer, there

143
will be some misclassification. Though SHLP2 did not correlate with grade, further evaluation of
SHLP for other end-points including progression, which was not available in our dataset, is
needed. Future studies are needed to explore the role of SHLP2 in races other than white and
black men.
In conclusion, our data suggest the mitochondrial peptide SHLP2 may be a novel prostate cancer
biomarker. Future studies are needed to confirm these findings and understand the mechanistic
link between low SHLP2 and prostate cancer as well as the link between SHLP2 and race and
whether it is involved in explaining racial disparities in prostate cancer.
SHLP2 Treatment Reduced γ-Glutamyl Cycle and ROS
The γ-glutamyl cycle is an important process in the cell to maintain normal redox state; it
generates glutathione to scavenge ROS, and thereby avoiding detrimental oxidative stress (403).
The thiol group of cysteine in glutathione is able to donate electron to other molecules, such as
reactive oxygen species to neutralize them. With contributing an electron, glutathione can be
converted to its oxidized form, GSSG. Therefore, the ratio of reduced glutathione to oxidized
glutathione within cells is often used as a measure of cellular oxidative stress (404). In our
metabolomic study, the levels of glutathione are not available, because the non-targeted approach
we utilized is not an ideal way to measure labile biochemicals that are very rapidly oxidized in
blood. Although the ratio of glutathione: GSSG cannot be calculated, the reduction of GSSG in
plasma still implicated reduced oxidative stress in tissues (405).
2-hydroxybutyrate is released as a byproduct when cystathionine is cleaved to cysteine that is
incorporated into glutathione, which can be further catalyzed to form 2-aminobutyrate. When
there is increased oxidative burden and glutathione consumption, ophthalmate can be synthesized

144
from glutamate and 2-aminobutyrate as a tripeptide analogue of glutathione (Figure 3.15). For
this reason, the elevation of ophthalmate and intermediates in this pathway is considered
potential biomarker for glutathione depletion following oxidative stress (406). It was also
reported that ophthalmate, 2-aminobutyrate and 2-hydroxybutyrate increased significantly in
various cancer types including breast cancer, ovarian cancer, and oral cancer (407,408).
Moreover, the elevated levels of 5-oxoproline, an amino acid derivative in the glutathione cycle,
is strongly correlated with increased glutathione consumption in colorectal cancer (409). These
aforementioned metabolites were reduced in plasma from HNG and SHLP2 treated DIO mice,
which is indicative of lower ROS levels in tissues and a decrease in demand for glutathione. This
metabolic signature also agrees with previous studies demonstrating the anti-ROS benefits of
humanin and SHLP2 (91). From the experiment done on isolated mitochondria from INS-1 cells,
we also observed direct effect of SHLP and HNG on reducing H2O 2 released from mitochondria
(Figure 3.15), suggesting HNG and SHLP2 may affect the mitochondrial glutathione.
Furthermore, suppression of the key intermediates in glutathione synthesis pathway may help
explain why high SHLP2 levels are associated with lower risk of prostate cancer, as oxidative
stress is involved in all aspects of tumor development.
Sphingolipids are important structural components of cell membranes and play essential roles in
cell signaling. Sphingolipids are primary components in lipid rafts, which are stabilized areas on
the plasma membrane that are enriched in sphingomyelin and cholesterol. The lipid rafts stabilize
the membrane, serve as attachment points for proteins and provide structural scaffolding (410).
With their structural and signaling roles in cell and plasma membranes, sphingolipids are
involved in insulin-action and inflammation (411,412). In addition, the sphingolipids ceramide
and sphingosine 1-phosphate (S1P) are mediators of apoptotic signaling (413). Several reports

145
have linked sphingolipids to normal and abnormal aging and their levels are elevated in diabetes
and obesity (414,415). Sphingomyelin is one of the most abundant sphingolipid and acts as both
membrane component and pool for rapid ceramide generation. Several studies showed that HFD
increased sphingomyelin levels in liver, adipose tissue and plasma (416). High concentration of
sphingomyelin has also been correlated with coronary artery disease in obese patients. S1P, a
potent bioactive sphingolipid, was found to be significantly higher in liver, skeletal muscle, and
plasma in response to several experimental models of HFD. While S1P in liver and skeletal
muscle was unchanged by HFD or palmitate in some studies, S1P in plasma was consistently
elevated, which suggests that circulating S1P may mediate some of the systemic effects induced
by HFD and obesity (417). Furthermore, there is a positive correlation between plasma S1P and
body fat percentage, body mass index, waist circumference, and fasting plasma insulin in obese
humans (418).
The ability of humanin to suppress ceramide synthesis has been demonstrated in in vitro models
(133). However, the two ceramides (N-palmitoyl-sphingosine (C16:0) and N-stearoyl-
sphingosine (C18:0)) measured in current study did not exhibit significant decrease in plasma.
This result does not exclude the possibility that synthesis of C16:0 and C18:0 ceramides were
indeed suppressed by humanin, because circulating ceramide levels might not promptly reflect
tissue levels. Moreover, the reduction in S1P and sphingomyelin further suggests that the peptide
treatments restored the HFD-induced dysregulation of sphingolipids and reduced sphingolipid
levels might partially explain their anti-apoptosis and insulin-sensitizing effects. It has also been
reported that AMPK activation inhibits ceramide synthesis at the serine palmitoyl transferase
step (419). Consistent with this notion, MDPs have been shown to activate AMPK in vivo. The
role of sphingolipids in obesity and metabolic aging is still emerging, our findings of reduced

146
circulating levels of sphingolipids in humanin and SHLP2 treated mice provide a new putative
biochemical explanation for their importance in maintaining metabolic fitness during aging.
From the current knowledge, we know that humanin and SHLP2 are both made from the
mitochondrial 16S rRNA and have some common biological functions. For example, they both
activate ERK and STAT3 signaling pathways, and exhibit insulin sensitization as well as anti-
apoptosis effects (73). As was seen in the PCA analysis and hierarchical clustering, clustering of
control samples is observed, however, the clustering and separation between humanin and
SHLP2 treated samples are limited. Limited hierarchical clustering or PCA group separations
can be an indication that the overall metabolite profiles are somewhat similar, but this does not
exclude the possibility that individual metabolite levels may be significantly different between
groups.
Humanin is known to exert its diverse functions through extracellular membrane receptors and
intracellular binding partners. Binding of humanin to the CNTF, WSX1 and GP130 trimeric
receptor leads to STAT3 phosphorylation and activation of downstream pathways (98,372). As
for SHLP2, it has been reported that SHLP2 also activated STAT3 pathways in a time-dependent
manner but with different kinetics than those induced by humanin, however the mechanism prior
to STAT3 action remains unknown (73). Through this study, we established more metabolic
similarities between humanin and SHLP2, and these metabolic similarities may be the result of
one or two common pathways.

147
Materials and Methods
Reagents
HNG peptide was synthesized by Genscript (Piscataway, NJ), All SHLP peptides were
synthesized by CPC Scientific (Sunnyvale, CA). Wild-type human IAPP was purchased from
Bachem America (Torrance, CA). Hexafluoroisopropanol was purchased from Sigma-Aldrich. 1-
oxyl-2,2,5,5-tetramethyl-Δ3-pyrroline-3-methyl methanethiosulfonate (MTSL), was purchased
from Toronto Research Chemicals (Toronto, Ontario, Canada). Human IAPP cysteine mutants
with alanine substitutions for the native cysteines at positions 2 and 7 were purchased from
Biomer Technology (Pleasanton, CA). The SHLP antibodies were custom synthesized by Harlan
(Indianapolis, IN) and YenZym (San Francisco, CA); all other antibodies were purchased from
Cell Signaling Technology (Danvers, MA). Western blotting reagents were purchased from
Biorad (Hercules, CA). Cell culture reagents, primary cortical neurons, dihydroethidine (DHE),
TRIzol®, Oil Red-O, and calcein-AM were purchased from Life Technologies (Grand Island,
NY). LINCOplex™ was purchased from Millipore (Billerica, MA). All primers and chemicals
were purchased from Sigma (St Louis, MO). Luminescent ATP assay kits were purchased from
Promega (Madison, WI). Amplex Red hydrogen peroxide/peroxidase assay kits were purchased
from Invitrogen (Carlsbad, CA).
Cell Culture
All cell lines (22RV1, NIT-1 murine β-cells, INS-1 cells and 3T3-L1 murine pre-adipocytes)
were purchased from ATCC (American tissue culture collection) (Manassas, VA). The human
prostate carcinoma cell line 22RV1 was maintained in RPMI 1640 medium supplemented with
10% fetal bovine serum (FBS) and 1% penicillin/streptomycin. NIT-1 murine β-cells were
cultured in F-12K medium supplemented with 10% FBS and 1% penicillin/streptomycin. INS-1

148
cells were maintained in RPMI1640 medium supplemented with 10% FBS, 100 µM 2-
mercaptoethanol and 1mM sodium pyruvate. 3T3-L1 murine pre-adipocytes were maintained in
DMEM supplemented with 10% calf serum and 1% penicillin/streptomycin. For individual
experiments, cells were seeded at a density of 1 × 105 cells/cm2 in individual wells of 96-, 6-
well plates, or 10-cm plates and grown to 80% confluence in a humidified atmosphere of 5%
CO2 at 37°C before treatment. All treatments were carried out as indicated in serum-free media.
HeLa-ρ0 cells were derived from parental HeLa cells by culturing HeLa cells in the presence of
100 ng/mL ethidium bromide (EtBr) for more than 20 generations in DMEM (373).
Cell Death, Proliferation, and Viability Assays
NIT-1 β-cells and 22Rv1 cells were cultured in complete media with control or SHLP peptides.
To assess cell viability, cells growing in 96-well plates were analyzed using a CellTiter 96®
AQueous One Solution MTS Assay kit (Promega) following the manufacturer's instructions. In
24-well plates, apoptosis was assessed by cell death detection using an ELISA (enzyme-linked
immuno-sorbent assay) (Roche, Branchburg, NJ) according to the manufacturer's instructions.
The cell proliferation rate was assessed by BrdU incorporation using ELISA (Roche) according
to the manufacturer's instructions (420).
Rabbit Anti-SHLP1-6 Antibody Generation
Custom rabbit anti-SHLP 1, 3, and 5 were ordered from Yenzym and anti-SHLP 2, 4, and 6 were
ordered from Harlan. High-titer polyclonal anti-sera against SHLPs 1–4 and 6 were obtained.
Anti-sera against SHLP5 could not be generated due to technical issues.

149
Mitochondrial Extraction
PC3 cells were maintained in RMPI 1640 media with 10% FBS and subjected to RNA or
mitochondrial isolation when the cell confluency reached 70–80%. The mitochondria were
prepared from 4 × 108 cells grown in 150-mm dishes. Cells were scraped in PBS (phosphate
buffered saline), centrifuged (500 g for 5 min at 4°C), and then resuspended in 10 mL of ice-cold
1× mannitol sucrose homogenization buffer (210 mM mannitol, 70 mM sucrose, 5 mM Tris-HCl
[pH 7.5], and 1 mM EDTA (ethylenediaminetetraacetic acid) [pH 7.5]). Cells were then
homogenized 15 times using a 5-mL Teflon homogenizer, and the degree of homogenization was
checked under a microscope. Homogenized cells were centrifuged (1,300 g for 5 min at 4°C) to
sediment the nuclei. The centrifugation step was repeated to ensure that all nuclei were in the
pellet fraction. Mitochondria were isolated from the supernatant by centrifugation (15,000 g for
15 min at 4°C), washed again by resuspending the pellet in 1× MS buffer, and centrifuged at
7,000–17,000 g.
RNA Isolation and Northern Blotting
Total cell RNA and mitochondrial RNA were extracted using the TRIzol® method followed by
column purification (Direct-zol™ RNA MiniPrep; Zymo Research, Irvine, CA). mRNA was
further purified by polyA selection using a Dynabeads® mRNA Purification Kit (Thermo
Scientific, Waltham, MA). The RNA concentration and purity were checked using a
NanoDrop™ (Thermo Scientific) and agarose gel electrophoresis. Then, 130 ng of each RNA
sample was separated on a 10% precast polyacrylamide tris-borate-EDTA(TBE)-urea gel (Bio-
Rad, Hercules, CA) and transferred onto a positively charged nylon transfer membrane
(Ambion®; Life Technologies). The blot was hybridized with 200 ng/mL digoxigenin (DIG)-
labeled SHLP6 RNA probe in Easy Hyb™ hybridization buffer (Roche) for 16 h at 60°C. After

150
two stringent washes in 2× SSC (saline sodium citrate buffer) (1× SSC contains 0.15 M NaCl
and 0.015 M sodium citrate)-0.1% SDS for 5 min at room temperature and 0.2× SSC-0.1% SDS
for 15 min at 60°C, the membrane was incubated for 30 min at room temperature with an anti-
DIG antibody conjugated to alkaline phosphatase (Roche). Subsequent visualization was
achieved using the BCIP-NBT substrate as instructed using a nucleic acid detection kit (Roche).
Western Blotting
Cell lysates were separated by SDS-PAGE (polyacrylamide gel electrophoresis) on tris-tricine or
TGXTM gels and transferred to polyvinyl-idenedifluoride membranes (Bio-Rad). Membranes
were blocked in 0.2% I-Block (Applied Biosystems, Foster City, CA) in PBS containing 0.1%
Tween 20 followed by incubation with the appropriate primary and secondary antibodies.
Antibody-antigen complexes were visualized using a ChemiLucent ECL detection system
(Millipore), followed by autoradiography.
Peptide Handling and Storage
Lyophilized wild type human IAPP was dissolved in HFIP, aliquoted into individual tubes and
flash frozen in N2 (l) prior to lyophilization. UV absorbance at 280 nm was used to determine
IAPP protein concentrations in denaturing conditions (8 M guanidinium chloride) using an ε280
of 1405 M−1cm−1 and verified by CD spectroscopy upon resolublization. Lyophilized IAPP
stocks were stored in N2 (g) under vacuum. MTSL labeled IAPP was stored at −20 °C in HFIP.
HNG was obtained lyophilized from Genscript and stored lyophilized at −80 °C until
solubilization. Solubilization and storage of HNG was performed according to the method
described by Arakawa et al. 2011 (376). Briefly, HNG was solubilized in water at 1 mg/mL and
aliquoted. Aliquots were stored at −20 °C until use. SHLP2 was obtained lyophilized and stored
at −80 °C until used. SHLP2 was solubilized in water at 1 mg/mL, aliquoted and stored at −20 °C

151
until use. We observed that prolonged storage (>1 month) of solubilized HNG stocks at −20 °C
markedly attenuated the potency of HNG and special care was taken to work with fresh stocks of
HNG. With SHLP2 this effect was less pronounced, but similar care was nevertheless taken to
work with fresh stocks of SHLP2.
Peptide Labeling
Spin labelling was performed as before (423). Briefly, single cysteine mutants of IAPP were
incubated with MTSL (>5 molar excess) for ~1 h at room temperature. Excess MTSL was
removed via cation exchange using a Toyopearl cation exchange column and subsequently
desalted on a C18 reverse phase SpinColumn (Harvard Apparatus, Holliston, MA), and
ultimately eluted in HFIP. Spin labelled peptide was stored in HFIP at −20 °C. Peptide
concentration was verified at the beginning of each experiment by comparing central line
amplitudes and double integral values against a standard concentration curve on the EPR
apparatus.
Thioflavin T Fluorescence Studies
Thioflavin T (ThT) was stored at a 5 mM stock concentration in water at −20 °C. ThT was used
at a 25 µM final concentration to monitor IAPP misfolding. IAPP aliquots were prepared as
above. Individual samples of IAPP were solubilized in appropriate buffer with ThT to a
concentration of 12.5 µM from a dry powder in 10 mM potassium phosphate buffer, pH 7.4.
MDPs were prepared to a stock concentration of 1 mg/mL as described above and added to
appropriate reactions. The mixtures were monitored for fluorescence in a 2 mm quartz cuvette
and a Jasco FP-6500 spectrofluorometer at room temperature. Fluorescence was monitored under
the following settings and conditions: excitation wavelength = 450 nm, emission
wavelength = 482 nm, excitation slit width = 1 nm, emission slit width = 10 nm, and pH = 7.4. t50

152
values were determined as before (424,425) using a sigmoidal model to fit our data. Each
experiment was normalized to an appropriate IAPP control by dividing fluorescence intensities
by the maximal IAPP control intensity. Statistical analysis for comparison was performed using
the student t-test.
For seeding experiments IAPP was fibrilized in 10 mM phosphate buffer, pH 7.4 for 2 weeks at
55 µM. Fibrils were sonicated using a titanium tip sonicator 4 × 30 seconds each and placed on
ice in between sonications. ThT, IAPP and MDPs were prepared as described above.
Fluorescence was measured in an Eppendorf AF2200 96-well fluorescence plate reader. Reaction
volumes were ~100 µL. Fluorescence was monitored under the following settings and conditions:
excitation wavelength = 440 nm, emission wavelength = 484 nm, 25 flashes per measurement and
a gain setting of 75.
Animals and Sample Collection
Male dietary-induced obesity (DIO) C57BL/6 mice at 12 weeks were obtained from Jackson
Laboratory (Bar Harbor, ME USA). The mice were singly housed under standard 12-hr light-
dark cycle with access to water and rodent food ad libitum (LabDiet, MO). Mice were randomly
assigned to one of three experimental groups (n = 6 per group): a control group receiving daily
Intraperitoneal (IP) injection of vehicle (sterilized water); a HNG-treated group receiving daily
IP injection of 5mg HNG per kg body weight; or a SHLP2-treated group receiving daily IP
injection of 5mg SHLP2 per kg body weight. Mice were euthanized after 3 days of treatment.
Prior to euthanasia, these mice were fasted for 9 hours. Blood were collected in blood collection
tubes with spray-coated K2EDTA (Becton Dickinson, USA) and cells were removed by
centrifugation at 1,500 x g for 10 minutes at 4°C. The resulting supernatant (plasma) was
transferred and aliquoted into Eppendorf tubes, then immediately stored at -80°C. The plasma

153
samples were then shipped on dry ice to Metabolon (NC, USA) for subsequent fractionation,
mass-spectrometry and analysis.
Plasma Sample Preparation
Each sample was accessioned into the Metabolon LIMS system and was assigned by the LIMS a
unique identifier that was associated with the original source identifier only. This identifier was
used to track all sample handling, tasks, results, etc. Samples were then prepared using the
automated MicroLab STAR® system (Hamilton, NV USA). Several recovery standards were
added prior to the first step in the extraction process for quality control (QC) purposes. To
remove protein, dissociate small molecules bound to protein or trapped in the precipitated protein
matrix, and to recover chemically diverse metabolites, proteins were precipitated with methanol
under vigorous shaking for 2 min using GenoGrinder 2000 (Glen Mills, NJ USA) followed by
centrifugation. The resulting extract was divided into five fractions: two for analysis by two
separate reverse phase (RP)/UPLC-MS/MS methods with positive ion mode electrospray
ionization (ESI), one for analysis by RP/UPLC-MS/MS with negative ion mode ESI, one for
analysis by HILIC/UPLC-MS/MS with negative ion mode ESI, and one sample was reserved for
backup. Samples were placed briefly on a TurboVap® (Zymark) to remove the organic solvent.
The sample extracts were stored overnight under nitrogen before preparation for analysis.
QA/QC
Several types of controls were analyzed in concert with the experimental samples: a pooled
matrix sample generated by taking a small volume of each experimental sample served as a
technical replicate throughout the data set; extracted water samples served as process blanks; and
a cocktail of QC standards that were carefully chosen not to interfere with the measurement of
endogenous compounds were spiked into every analyzed sample, allowed instrument

154
performance monitoring and aided chromatographic alignment. The QC standards include an
aliquot of solvents used in extraction, a recovery standard to assess variability and verify
performance of extraction and an internal standard to ensure performance of instrument.
Instrument variability was determined by calculating the median relative standard deviation
(RSD) for the standards that were added to each sample prior to injection into the mass
spectrometers. Overall process variability was determined by calculating the median RSD for all
endogenous metabolites (i.e., non-instrument standards) present in 100% of the pooled matrix
samples.
Plasma Metabolome Analysis Using UPLC-MS/MS
All methods utilized a Waters ACQUITY ultra-performance liquid chromatography (UPLC) and
a Thermo Scientific Q-Exactive high resolution/accurate mass spectrometer interfaced with a
heated electrospray ionization (HESI-II) source and Orbitrap mass analyzer operated at 35,000
mass resolution. The sample extract was dried then reconstituted in solvents compatible to each
of the four methods. Each reconstitution solvent contained a series of standards at fixed
concentrations to ensure injection and chromatographic consistency. One aliquot was analyzed
using acidic positive ion conditions, chromatographically optimized for more hydrophilic
compounds. In this method, the extract was gradient eluted from a C18 column (Waters UPLC
BEH C18-2.1x100 mm, 1.7 µm) using water and methanol, containing 0.05% perfluoropentanoic
acid (PFPA) and 0.1% formic acid (FA). Another aliquot was also analyzed using acidic positive
ion conditions, however it was chromatographically optimized for more hydrophobic
compounds. In this method, the extract was gradient eluted from the same afore mentioned C18
column using methanol, acetonitrile, water, 0.05% PFPA and 0.01% FA and was operated at an
overall higher organic content. Another aliquot was analyzed using basic negative ion optimized

155
conditions using a separate dedicated C18 column. The basic extracts were gradient eluted from
the column using methanol and water, however with 6.5mM Ammonium Bicarbonate at pH 8.
The fourth aliquot was analyzed via negative ionization following elution from a HILIC column
(Waters UPLC BEH Amide 2.1x150 mm, 1.7 µm) using a gradient consisting of water and
acetonitrile with 10mM Ammonium Formate, pH 10.8. The MS analysis alternated between MS
and data-dependent MSn scans using dynamic exclusion. The scan range varied slighted between
methods but covered 70-1000 m/z.
Metabolite Quantification and Statistical Analysis
Raw data was extracted, peak-identified and QC processed using Metabolon’s hardware and
software. Compounds were identified by comparison to library entries of purified standards or
recurrent unknown entities. Metabolon maintains a library based on authenticated standards that
contains the retention time/index (RI), mass to charge ratio (m/z), and chromatographic data
(including MS/MS spectral data) on all molecules present in the library. Furthermore,
biochemical identifications are based on three criteria: retention index within a narrow RI
window of the proposed identification, accurate mass match to the library +/- 10 ppm, and the
MS/MS forward and reverse scores between the experimental data and authentic standards. The
MS/MS scores are based on a comparison of the ions present in the experimental spectrum to the
ions present in the library spectrum. While there may be similarities between these molecules
based on one of these factors, the use of all three data points can be utilized to distinguish and
differentiate biochemicals.
Peaks were quantified using area-under-the-curve. For studies spanning multiple days, a data
normalization step was performed to correct variation resulting from instrument inter-day tuning
differences. For studies that did not require more than one day of analysis, no normalization is

156
necessary. Missing values (if any) are assumed to be below the level of detection. However,
biochemicals that were detected in all samples from one or more groups, but not in samples from
other groups, were assumed to be near the lower limit of detection in the groups in which they
were not detected. In this case, the lowest detected level of these biochemicals was imputed for
samples in which that biochemical was not detected. Biochemical fold differences were
generated on the basis of group means after imputation with ratios greater than one representing
fold increases, whereas ratios less than one show fold decreases. One-way ANOVA (analysis of
variance) and Tukey & Dunnett multiple comparisons were conducted to identify biochemicals
that differed between control and peptide treatment when comparing the metabolic profiles of
plasma samples (ArrayStudio). P values<0.05 were considered statistically significant. The level
of 0.05 is the false positive rate when there is one test. However, for a large number of tests we
need to account for false positives. The FDR was estimated using the q-value (ArrayStudio). The
additional tests were conducted by using RStudio version 1.0.143.
Human Subjects
We used samples from an ongoing case-control study of men undergoing prostate biopsies at the
Durham Veterans Affairs Medical Center. The study was approved by the Institutional Review
Board and written informed consent was obtained from all subjects before enrollment. Subjects
were recruited between January 2007 and September 2015 from the urology clinic. Eligible
subjects were men with no prior history of prostate cancer who were undergoing a prostate
needle biopsy because of abnormal PSA and/or suspicious digital rectal exam (DRE). We
selected 200 men with available serum samples for our study equally divided by biopsy outcome
and race (100 negative biopsies; 100 prostate cancer cases; 101 black and 99 white). Diabetes
status and race were self-reported.

157
Biochemical Analysis of Serum
Serum was collected from all patients prior to the biopsy. Most patients (n=148; 74%) were
fasting at the time of blood draw, but as SHLP2 levels were similar in those fasting vs. not
fasting (p=0.078), fasting was not considered in the models. Endogenous serum SHLP2 levels
were measured using our developed SHLP2 ELISA using total immunoglobin (Ig)G and ligand
purified antibodies with a detection limit of 50 pg/ml. Custom rabbit anti-SHLP2 antibody was
ordered from Harlan (Indianapolis, IN). The intra- and inter-assay coefficient variations (CV) of
the SHLP2 ELISA were less than 10%. Prior to the SHLP2 ELISA, from each sample, 100 µL of
serum was extracted using an acid solution (90% acetonitrile and 10% 1 N HCl). The supernatant
was dried using a SpeedVac™ (Thermo Fisher, Waltham, MA). The dried samples were
reconstituted with assay buffer (50 mM PBS containing 0.5% Tween 20). 96-well microtiter
plates were coated with SHLP2 capture antibody at a concentration of 0.5 µg/well in 200 µL of
50 mM sodium bicarbonate buffer (pH 9.5). The plates were incubated for 3–4 h at room
temperature on a shaker, washed with wash buffer, and then washed twice with Superblock™
buffer (Pierce Chemicals, Rockford, IL). Standards, controls, or extracted samples were added to
the appropriate wells with pre-tittered detection antibody and incubated overnight. After
washing, streptavidin-HRP (horse radish peroxidase) was added and further incubated for 30 min
at room temperature. After four washes with wash buffer, 200 µL/well of OPD (o-
phenylenediamine dihydro-chloride) substrate (1 mg/ml in hydrogen peroxide) was added and
incubated for 10–20 minutes. Reactions were terminated with 50 µL/well 2 N H2SO4, and
absorbance values were measured on a plate spectrophotometer (Molecular Designs, Sunnyvale,
CA) at 490 nm.

158
Statistical Analyses
Wilcoxon or chi-squared tests were used for univariable comparisons of subject characteristics
between cases and controls, and also between those with SHLP2 ≥350 pg/ml vs. <350 pg/ml.
The interaction between SHLP2 and race in predicting prostate cancer was tested by including a
cross product term in a logistic regression model. Logistic regression models to test the
association between SHLP2 and prostate cancer risk were then stratified by race. Multinomial
logistic regression was used to test the link between SHLP2 and low-grade prostate cancer
(Gleason <7) vs. no prostate cancer and high-grade prostate cancer (Gleason 7-10) vs. no
prostate cancer. Models were adjusted for age, BMI, PSA (logarithmically transformed), DRE
and year of biopsy. In a secondary analysis, we also adjusted for prostate volume and family
history of prostate cancer.
A SHLP2 cut-off of 350 pg/ml (~upper tertile) was used to assess the negative predictive value
(NPV) of prostate cancer by race. ROC curves were plotted as false-positive rate (1–specificity)
vs sensitivity for adjusted SHLP2 values. The diagnostic performance of SHLP2 to predict
prostate biopsy results (our primary outcome, prostate cancer vs no cancer) in each race group
was assessed using the area under the ROC curve (AUC). Statistical analyses were performed
using SAS® 9.3 (SAS Institute Inc.; Cary, NC) and Stata® 13.1 (StataCorp.; College Station,
TX), with p<0.05 defined as statistical significance.
For other biological assays, all values shown are presented as mean ± SEM (standard error of the
mean). Independent two-tailed t-tests were used to compare the differences between two groups
for analysis, such as cell viability, apoptosis, and proliferation. P < 0.05 was considered
statistically significant.

159
Future Directions
The field of mitochondrial-derived peptides (MDPs) is emerging, and there is still much to learn
and explore. As for the already identified MDPs, especially humanin and SHLP2, there are
knowledge gaps that yet to be researched. For instance, the AMPK activation by short-term
humanin treatment results in increase in ATP production without apparent changes in
mitochondrial biogenesis. We hypothesized that the short-term humanin administration may first
induce ATP production via fatty acid oxidation, while the mitochondrial biogenesis happens
after chronic stimulation of AMPK. Clearly, the hypothesis needs to be carefully investigated
through more in vitro and in vivo experiments. The association between SHLP2 levels and
prostate cancer risk is very intriguing, but also requires more thorough study on the molecular
mechanism underlying the etiology.

160
Supplemental Figures and Tables

Supplemental Figure S1
Treatment with HNG over long-term did not change the percentage of subcutaneous fat in female
mice.

Supplemental Figure S2
Treatment with 30µM HNG peptide in C2C12 myotubes for 24 hours failed to increase the levels
of mitochondrial proteins including LonP1, TFAM and SLC25A5.
Control HNG
0
50
100
150
Subcutaneous Fat %
% of control
LonP1
Slc
Tfam
0.0
0.5
1.0
1.5
Expression of key mitochondrial genes
Relative expression
Control
HNG
LonP1
GAPDH
TFAM
Contro
l
HNG
A B

161
Supplementary Table 1
MDPs HNG and SHLP2 inhibit the misfolding of IAPP. A and B) Kinetics study of IAPP
misfolding by ThT fluorescence in the presence of varying concentrations of (a) HNG or (b)
SHLP2. Mean t50 values from the experiments performed in Figure 3.5 are shown in the table.
MDP concentrations are given in µM starting at 0, in the absence of MDPs, and t50 averages are
given in hours. Data are presented as an average t50 ± 1 standard deviation. For conditions where
no fibrilization was observed over the course of the experiment, a >20h is given. Refer to Figure
3.5 for methods.

Supplementary Table 2
Association between SHLP2 and overall risk of cancer and risk of cancer grade, stratified by
race.

162
References
1. Friedman JR, Nunnari J. Mitochondrial form and function. Nature. 2014;505(7483):335-
343. doi:10.1038/nature12985.
2. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organization of the human
mitochondrial genome. Nature. 1981;290(5806):457-465. doi:10.1038/290457a0.
3. Asin-Cayuela J, Gustafsson CM. Mitochondrial transcription and its regulation in
mammalian cells. Trends Biochem Sci. 2007;32(3):111-117.
doi:10.1016/j.tibs.2007.01.003.
4. Shutt TE, Shadel GS. A compendium of human mitochondrial gene expression machinery
with links to disease. Environ Mol Mutagen. 2010;51(5):360-379. doi:10.1002/em.20571.
5. Sproule DM, Kaufmann P. Mitochondrial encephalopathy, lactic acidosis, and strokelike
episodes: Basic concepts, clinical phenotype, and therapeutic management of MELAS
syndrome. Ann N Y Acad Sci. 2008;1142:133-158. doi:10.1196/annals.1444.011.
6. Kaguni LS. DNA polymerase gamma, the mitochondrial replicase. Annu Rev Biochem.
2004;73(1):293-320. doi:10.1146/annurev.biochem.72.121801.161455.
7. Bogenhagen D, Clayton DA. Mouse L cell mitochondrial DNA molecules are selected
randomly for replication throughout the cell cycle. Cell. 1977;11(4):719-727.
doi:10.1016/0092-8674(77)90286-0.
8. Birky CW. Relaxed and Stringent Genomes: Why Cytoplasmic Genes Don’t Obey
Mendel’s Laws. J Hered. 1994;85(5):355-365.
doi:10.1093/oxfordjournals.jhered.a111480.
9. Longley MJ, Ropp P a, Lim SE, Copeland WC. Characterization of the native and
recombinant catalytic subunit of human DNA polymerase gamma: identification of

163
residues critical for exonuclease activity and dideoxynucleotide sensitivity. Biochemistry.
1998;37(29):10529-10539. doi:10.1021/bi980772w.
10. Lim SE, Longley MJ, Copeland WC. The mitochondrial p55 accessory subunit of human
DNA polymerase γ enhances DNA binding, promotes processive DNA synthesis, and
confers N- ethylmaleimide resistance. J Biol Chem. 1999;274(53):38197-38203.
doi:10.1074/jbc.274.53.38197.
11. Longley MJ, Prasad R, Srivastava DK, Wilson SH, Copeland WC. Identification of 5’-
deoxyribose phosphate lyase activity in human DNA polymerase gamma and its role in
mitochondrial base excision repair in vitro. Proc Natl Acad Sci U S A. 1998;95(21):12244-
12248. doi:10.1073/pnas.95.21.12244.
12. Alam TI, Kanki T, Muta T, et al. Human mitochondrial DNA is packaged with TFAM.
Nucleic Acids Res. 2003;31(6):1640-1645. doi:10.1093/nar/gkg251.
13. Ngo HB, Kaiser JT, Chan DC. The mitochondrial transcription and packaging factor Tfam
imposes a U-turn on mitochondrial DNA. Nat Struct {&} Mol Biol. 2011;18(11):1290-
1296. doi:10.1038/nsmb.2159.
14. Hallberg BM, Larsson NG. TFAM forces mtDNA to make a U-turn. Nat Struct Mol Biol.
2011;18(11):1179-1181. doi:10.1038/nsmb.2167.
15. Taylor RW, Turnbull DM. Mitochondrial DNA mutations in human disease. Nat Rev
Genet. 2005;6(5):389-402. doi:10.1038/nrg1606.
16. Wong LJC, Naviaux RK, Brunetti-Pierri N, et al. Molecular and clinical genetics of
mitochondrial diseases due to POLG mutations. Hum Mutat. 2008;29(9).
doi:10.1002/humu.20824.
17. Zeviani M, Di Donato S. Mitochondrial disorders. Brain. 2004;127(Pt 10):2153-2172.

164
doi:10.1093/brain/awh259.
18. Harman D. Aging: a theory based on free radical and radiation chemistry. J Gerontol.
1956;11(3):298-300.
19. Barja G. Mitochondrial free radical production and aging in mammals and birds. In:
Annals of the New York Academy of Sciences. Vol 854. ; 1998:224-238.
doi:10.1111/j.1749-6632.1998.tb09905.x.
20. Trifunovic A. Mitochondrial DNA and ageing. Biochim Biophys Acta - Bioenerg.
2006;1757(5-6):611-617. doi:10.1016/j.bbabio.2006.03.003.
21. Balaban RS, Nemoto S, Finkel T. Mitochondria, oxidants, and aging. Cell.
2005;120(4):483-495. doi:10.1016/j.cell.2005.02.001.
22. Trifunovic A, Wredenberg A, Falkenberg M, et al. Premature ageing in mice expressing
defective mitochondrial DNA polymerase. Nature. 2004;429(6990):417-423. doi:DOI
10.1038/nature02517.
23. Kujoth GC, Hiona A, Pugh TD, et al. Mitochondrial DNA mutations, oxidative stress, and
apoptosis in mammalian aging. Science (80- ). 2005;309(5733):481-484.
doi:10.1126/science.1112125.
24. Trifunovic A, Hansson A, Wredenberg A, et al. Somatic mtDNA mutations cause aging
phenotypes without affecting reactive oxygen species production. Proc Natl Acad Sci U S
A. 2005;102(50):17993-17998. doi:10.1073/pnas.0508886102.
25. Krtolica a, Parrinello S, Lockett S, Desprez PY, Campisi J. Senescent fibroblasts promote
epithelial cell growth and tumorigenesis: a link between cancer and aging. Proc Natl Acad
Sci U S A. 2001;98(21):12072-12077. doi:10.1073/pnas.211053698.
26. Muñoz-Espín D, Serrano M. Cellular senescence: from physiology to pathology. Nat Rev

165
Mol Cell Biol. 2014;15(7):482-496. doi:10.1038/nrm3823.
27. Campisi J, d’Adda di Fagagna F. Cellular senescence: when bad things happen to good
cells. Nat Rev Mol Cell Biol. 2007;8(9):729-740. doi:10.1038/nrm2233.
28. Baker DJ, Wijshake T, Tchkonia T, et al. Clearance of p16Ink4a-positive senescent cells
delays ageing-associated disorders. Nature. 2011;479(7372):232-236.
doi:10.1038/nature10600.
29. Janzen V, Forkert R, Fleming HE, et al. Stem-cell ageing modified by the cyclin-
dependent kinase inhibitor p16INK4a. Nature. 2006;443(7110):421-426. doi:nature05159
[pii] 10.1038/nature05159.
30. Correia-Melo C, Birch J, Passos JF. Powering senescence: The ugly side of mitochondria.
Cell Cycle. 2016;15(19):1-2. doi:10.1080/15384101.2016.1204852.
31. Ziegler D V., Wiley CD, Velarde MC. Mitochondrial effectors of cellular senescence:
Beyond the free radical theory of aging. Aging Cell. 2015;14(1):1-7.
doi:10.1111/acel.12287.
32. Moiseeva O, Bourdeau V, Roux A, Deschênes-Simard X, Ferbeyre G. Mitochondrial
dysfunction contributes to oncogene-induced senescence. Mol Cell Biol.
2009;29(16):4495-4507. doi:10.1128/MCB.01868-08.
33. Wiley CD, Velarde MC, Lecot P, et al. Mitochondrial dysfunction induces senescence
with a distinct secretory phenotype. Cell Metab. 2016;23(2):303-314.
doi:10.1016/j.cmet.2015.11.011.
34. Correia-Melo C, Marques FD, Anderson R, et al. Mitochondria are required for pro-
ageing features of the senescent phenotype. EMBO J. 2016;35(7):724-742.
doi:10.15252/embj.201592862.

166
35. van Deursen JM. The role of senescent cells in ageing. Nature. 2014;509(7501):439-446.
doi:10.1038/nature13193.
36. WEINHOUSE S, WARBURG O, BURK D, SCHADE AL. On Respiratory Impairment in
Cancer Cells. Science (80- ). 1956;124(3215):267-272. doi:10.1126/science.124.3215.267.
37. Hsu PP, Sabatini DM. Cancer Cell Metabolism: Warburg and Beyond. Cell.
2008;134(5):703-707. doi:10.1016/j.cell.2008.08.021.
38. Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg Effect: The
Metabolic Requirements of Cell Proliferation. Science (80- ). 2009;324(5930):1029-1033.
doi:10.1126/science.1160809.
39. Parr RL, Mills J, Harbottle A, et al. Mitochondria, prostate cancer, and biopsy sampling
error. Discov Med. 2013;15(83):213-220.
http://www.ncbi.nlm.nih.gov/pubmed/23636138.
40. Neiman M, Taylor DR. The causes of mutation accumulation in mitochondrial genomes.
Proc R Soc B Biol Sci. 2009;276(1660):1201-1209. doi:10.1098/rspb.2008.1758.
41. Ishikawa K, Takenaga K, Akimoto M, et al. ROS-generating mitochondrial DNA
mutations can regulate tumor cell metastasis. Science. 2008;320(5876):661-664.
doi:10.1126/science.1156906.
42. Ferraresi R, Troiano L, Pinti M, et al. Resistance of mtDNA-depleted cells to apoptosis.
Cytom Part A. 2008;73(6):528-537. doi:10.1002/cyto.a.20544.
43. Moro L, Arbini a a, Yao JL, di Sant’Agnese P a, Marra E, Greco M. Mitochondrial DNA
depletion in prostate epithelial cells promotes anoikis resistance and invasion through
activation of PI3K/Akt2. Cell Death Differ. 2009;16(4):571-583.
doi:10.1038/cdd.2008.178.

167
44. Lin MT, Beal MF. Mitochondrial dysfunction and oxidative stress in neurodegenerative
diseases. Nature. 2006;443(7113):787-795. doi:10.1038/nature05292.
45. Moreira PI, Carvalho C, Zhu X, Smith MA, Perry G. Mitochondrial dysfunction is a
trigger of Alzheimer’s disease pathophysiology. Biochim Biophys Acta - Mol Basis Dis.
2010;1802(1):2-10. doi:10.1016/j.bbadis.2009.10.006.
46. Corral-Debrinski M, Horton T, Lott MT, Shoffner JM, Beal MF, Wallace DC.
Mitochondrial DNA deletions in human brain: regional variability and increase with
advanced age. Nat Genet. 1992;2:324-329. doi:10.1038/ng1292-324.
47. Ekstrand MI, Terzioglu M, Galter D, et al. Progressive parkinsonism in mice with
respiratory-chain-deficient dopamine neurons. Proc Natl Acad Sci. 2007;104(4):1325-
1330. doi:10.1073/pnas.0605208103.
48. Dai Y, Kiselak T, Clark J, et al. Behavioral and metabolic characterization of
heterozygous and homozygous POLG mutator mice. Mitochondrion. 2013;13(4):282-291.
doi:10.1016/j.mito.2013.03.006.
49. Luoma P, Melberg A, Rinne JO, et al. Parkinsonism, premature menopause, and
mitochondrial DNA polymerase γ mutations: Clinical and molecular genetic study.
Lancet. 2004;364(9437):875-882. doi:10.1016/S0140-6736(04)16983-3.
50. Trancikova A, Ramonet D, Moore DJ. Genetic mouse models of neurodegenerative
diseases. Prog Mol Biol Transl Sci. 2011;100:419-482. doi:10.1016/B978-0-12-384878-
9.00012-1.
51. Martin LJ, Pan Y, Price AC, et al. Parkinson’s disease alpha-synuclein transgenic mice
develop neuronal mitochondrial degeneration and cell death. J Neurosci. 2006;26(1):41-
50. doi:10.1523/JNEUROSCI.4308-05.2006.

168
52. Tsika E, Moore DJ. Mechanisms of LRRK2-mediated neurodegeneration. Curr Neurol
Neurosci Rep. 2012;12(3):251-260. doi:10.1007/s11910-012-0265-8.
53. Manczak M, Anekonda TS, Henson E, Park BS, Quinn J, Reddy PH. Mitochondria are a
direct site of A beta accumulation in Alzheimer’s disease neurons: implications for free
radical generation and oxidative damage in disease progression. Hum Mol Genet.
2006;15(9):1437-1449. doi:10.1093/hmg/ddl066.
54. Praticò D, Uryu K, Leight S, Trojanoswki JQ, Lee VM-Y. Increased Lipid Peroxidation
Precedes Amyloid Plaque Formation in an Animal Model of Alzheimer Amyloidosis. J
Neurosci. 2001;21(12):4183-4187. doi:21/12/4183 [pii].
55. Raimondi A, Mangolini A, Rizzardini M, et al. Cell culture models to investigate the
selective vulnerability of motoneuronal mitochondria to familial ALS-linked G93ASOD1.
Eur J Neurosci. 2006;24(2):387-399. doi:10.1111/j.1460-9568.2006.04922.x.
56. Rizzuto R, Brini M, Murgia M, Pozzan T. Microdomains with high Ca2+ close to IP3-
sensitive channels that are sensed by neighboring mitochondria. Science (80- ).
1993;262(5134):744-747. doi:10.1126/science.8235595.
57. Butow RA, Avadhani NG. Mitochondrial signaling: The retrograde response. Mol Cell.
2004;14(1):1-15. doi:10.1016/S1097-2765(04)00179-0.
58. Liu X, Kim CN, Yang J, Jemmerson R, Wang X. Induction of apoptotic program in cell-
free extracts: Requirement for dATP and cytochrome c. Cell. 1996;86(1):147-157.
doi:10.1016/S0092-8674(00)80085-9.
59. Finkel T. Signal transduction by mitochondrial oxidants. J Biol Chem. 2012;287(7):4434-
4440. doi:10.1074/jbc.R111.271999.
60. Chandel NS, Maltepe E, Goldwasser E, Mathieu CE, Simon MC, Schumacker PT.

169
Mitochondrial reactive oxygen species trigger hypoxia-induced transcription. Proc Natl
Acad Sci U S A. 1998;95(20):11715-11720. doi:10.1073/pnas.95.20.11715.
61. Martinus RD, Garth GP, Webster TL, et al. Selective induction of mitochondrial
chaperones in response to loss of the mitochondrial genome. Eur J Biochem.
1996;240(1):98-103. doi:10.1111/j.1432-1033.1996.0098h.x.
62. Biswas G, Adebanjo OA, Freedman BD, et al. Retrograde Ca2+ signaling in C2C12
skeletal myocytes in response to mitochondrial genetic and metabolic stress: A novel
mode of inter-organelle crosstalk. EMBO J. 1999;18(3):522-533.
doi:10.1093/emboj/18.3.522.
63. Nargund AM, Pellegrino MW, Fiorese CJ, Baker BM, Haynes CM. Mitochondrial import
efficiency of ATFS-1 regulates mitochondrial UPR activation. Science (80- ).
2012;337(6094):587-590. doi:10.1126/science.1223560.
64. Nargund AM, Fiorese CJ, Pellegrino MW, Deng P, Haynes CM. Mitochondrial and
nuclear accumulation of the transcription factor ATFS-1 promotes OXPHOS recovery
during the UPRmt. Mol Cell. 2015;58(1):123-133. doi:10.1016/j.molcel.2015.02.008.
65. Quirós PM, Mottis A, Auwerx J. Mitonuclear communication in homeostasis and stress.
Nat Rev Mol Cell Biol. 2016;17(4):213-226. doi:10.1038/nrm.2016.23.
66. Augustin S, Nolden M, Müller S, Hardt O, Arnold L, Langer T. Characterization of
peptides released from mitochondria: Evidence for constant proteolysis and peptide efflux.
J Biol Chem. 2005;280(4):2691-2699. doi:10.1074/jbc.M410609200.
67. Basrai MA, Hieter P, Boeke JD. Small Open Reading Frames: Beautiful Needles in
the Haystack. Genome Res. 1997;7(8):768-771. doi:10.1101/gr.7.8.768.
68. Andrews SJ, Rothnagel JA. Emerging evidence for functional peptides encoded by short

170
open reading frames. Nat Rev Genet. 2014;15(3):193-204. doi:10.1038/nrg3520.
69. Vanderperre B, Lucier J-F, Bissonnette C, et al. Direct detection of alternative open
reading frames translation products in human significantly expands the proteome. PLoS
One. 2013;8(8):e70698. doi:10.1371/journal.pone.0070698.
70. Anderson S, Bankier AT, Barrell BG, et al. Sequence and organization of the human
mitochondrial genome. Nature. 1981;290(5806):457-465.
http://www.ncbi.nlm.nih.gov/pubmed/7219534. Accessed January 5, 2015.
71. Taanman J-W. The mitochondrial genome: structure, transcription, translation and
replication. Biochim Biophys Acta - Bioenerg. 1999;1410(2):103-123. doi:10.1016/S0005-
2728(98)00161-3.
72. Lee C, Zeng J, Drew BG, et al. The mitochondrial-derived peptide MOTS-c promotes
metabolic homeostasis and reduces obesity and insulin resistance. Cell Metab.
2015;21(3):443-454. doi:10.1016/j.cmet.2015.02.009.
73. Cobb LJ, Lee C, Xiao J, et al. Naturally occurring mitochondrial-derived peptides are age-
dependent regulators of apoptosis, insulin sensitivity, and inflammatory markers. Aging
(Albany NY). 2016;8(4):1-14. doi:10.18632/aging.100943.
74. Lee C, Yen K, Cohen P. Humanin: A harbinger of mitochondrial-derived peptides?
Trends Endocrinol Metab. 2013;24(5):222-228. doi:10.1016/j.tem.2013.01.005.
75. Hashimoto Y, Niikura T, Tajima H, et al. A rescue factor abolishing neuronal cell death
by a wide spectrum of familial Alzheimer’s disease genes and Abeta. Proc Natl Acad Sci
U S A. 2001;98(11):6336-6341. doi:10.1073/pnas.101133498.
76. Hashimoto Y, Ito Y, Niikura T, et al. Mechanisms of neuroprotection by a novel rescue
factor humanin from Swedish mutant amyloid precursor protein. Biochem Biophys Res

171
Commun. 2001;283(2):460-468. doi:10.1006/bbrc.2001.4765.
77. Guo B, Zhai D, Cabezas E, et al. Humanin peptide suppresses apoptosis by interfering
with Bax activation. Nature. 2003;423(6938):456-461. doi:10.1038/nature01627.
78. Ikonen M, Liu B, Hashimoto Y, et al. Interaction between the Alzheimer’s survival
peptide humanin and insulin-like growth factor-binding protein 3 regulates cell survival
and apoptosis. Proc Natl Acad Sci USA. 2003;100(22):13042-13047.
doi:10.1073/pnas.2135111100.
79. Firth SM, Baxter RC. Cellular actions of the insulin-like growth factor binding proteins.
Endocr Rev. 2002;23(6):824-854. doi:10.1210/er.2001-0033.
80. Rajah R, Valentinis B, Cohen P. Insulin-like growth factor (IGF)-binding protein-3
induces apoptosis and mediates the effects of transforming growth factor-beta1 on
programmed cell death through a p53- and IGF-independent mechanism. J Biol Chem.
1997;272(18):12181-12188. doi:10.1074/jbc.272.18.12181.
81. Niikura T, Hashimoto Y, Okamoto T, et al. Insulin-Like Growth Factor I (IGF-I) Protects
Cells from Apoptosis by Alzheimer’s V642I Mutant Amyloid Precursor Protein through
IGF-I Receptor in an IGF-Binding Protein-Sensitive Manner. J Neurosci.
2001;21(6):1902-1910. http://www.jneurosci.org/content/21/6/1902.short. Accessed
December 4, 2015.
82. Kuliawat R, Klein L, Gong Z, et al. Potent humanin analog increases glucose-stimulated
insulin secretion through enhanced metabolism in the β cell. FASEB J. 2013;27(12):4890-
4898. doi:10.1096/fj.13-231092.
83. Gong Z, Tas E, Muzumdar R. Humanin and age-related diseases: a new link? Front
Endocrinol (Lausanne). 2014;5:210. doi:10.3389/fendo.2014.00210.

172
84. Thummasorn S, Apaijai N, Kerdphoo S, Shinlapawittayatorn K, Chattipakorn SC,
Chattipakorn N. Humanin exerts cardioprotection against cardiac ischemia/reperfusion
injury through attenuation of mitochondrial dysfunction. Cardiovasc Ther.
2016;34(6):404-414. doi:10.1111/1755-5922.12210.
85. Muzumdar RH, Huffman DM, Calvert JW, et al. Acute humanin therapy attenuates
myocardial ischemia and reperfusion injury in mice. Arterioscler Thromb Vasc Biol.
2010;30(10):1940-1948. doi:10.1161/ATVBAHA.110.205997.
86. Zhao S-T, Huang X-T, Zhang C, Ke Y. Humanin protects cortical neurons from ischemia
and reperfusion injury by the increased activity of superoxide dismutase. Neurochem Res.
2012;37(1):153-160. doi:10.1007/s11064-011-0593-0.
87. Murakami M, Nagahama M, Maruyama T, Niikura T. Humanin ameliorates diazepam-
induced memory deficit in mice. Neuropeptides. 2017;62:65-70.
doi:10.1016/j.npep.2016.10.008.
88. Zhao H, Sonada S, Yoshikawa A, Yoshikawa M, Ohinata K. Rubimetide, humanin, and
MMK1 exert anxiolytic-like activities via the formyl peptide receptor 2 in mice followed
by the successive activation of DP1, A2A, and GABAA receptors. Peptides. 2016;83:16-
20. doi:10.1016/j.peptides.2016.07.001.
89. Gidlund E, von Walden F, Venojärvi M, et al. Humanin skeletal muscle protein levels
increase after resistance training in men with impaired glucose metabolism. Physiol Rep.
2016;4(23):e13063. doi:10.14814/phy2.13063.
90. Nikolakopoulos P, Tzimagiorgis G, Goulis DG, Chatzopoulou F, Zepiridis L, Vavilis D.
Serum humanin concentrations in women with pre-eclampsia compared to women with
uncomplicated pregnancies. Journal of Maternal-Fetal and Neonatal Medicine. 2017:1-7.

173
91. Paharkova V, Alvarez G, Nakamura H, Cohen P, Lee K-W. Rat Humanin is encoded and
translated in mitochondria and is localized to the mitochondrial compartment where it
regulates ROS production. Mol Cell Endocrinol. 2015;413:96-100.
doi:10.1016/j.mce.2015.06.015.
92. Yamagishi Y, Hashimoto Y, Niikura T, Nishimoto I. Identification of essential amino
acids in Humanin, a neuroprotective factor against Alzheimer’s disease-relevant insults.
Peptides. 2003;24(4):585-595. http://www.ncbi.nlm.nih.gov/pubmed/12860203. Accessed
December 23, 2015.
93. Maftei M, Tian X, Manea M, et al. Interaction structure of the complex between
neuroprotective factor humanin and Alzheimer’s β-amyloid peptide revealed by affinity
mass spectrometry and molecular modeling. J Pept Sci. 2012;18(6):373-382.
doi:10.1002/psc.2404.
94. Hashimoto Y, Niikura T, Ito Y, et al. Detailed characterization of neuroprotection by a
rescue factor humanin against various Alzheimer’s disease-relevant insults. J Neurosci.
2001;21(23):9235-9245. http://www.ncbi.nlm.nih.gov/pubmed/11717357. Accessed
December 23, 2015.
95. Terashita K, Hashimoto Y, Niikura T, et al. Two serine residues distinctly regulate the
rescue function of Humanin, an inhibiting factor of Alzheimer’s disease-related
neurotoxicity: functional potentiation by isomerization and dimerization. J Neurochem.
2003;85(6):1521-1538. http://www.ncbi.nlm.nih.gov/pubmed/12787071. Accessed
December 23, 2015.
96. Oh YK, Bachar AR, Zacharias DG, et al. Humanin preserves endothelial function and
prevents atherosclerotic plaque progression in hypercholesterolemic ApoE deficient mice.

174
Atherosclerosis. 2011;219(1):65-73. doi:10.1016/j.atherosclerosis.2011.06.038.
97. Jia Y, Ohanyan A, Lue Y-H, et al. The effects of humanin and its analogues on male germ
cell apoptosis induced by chemotherapeutic drugs. Apoptosis. 2015;20(4):551-561.
doi:10.1007/s10495-015-1105-5.
98. Hashimoto Y, Kurita M, Aiso S, Nishimoto I, Matsuoka M. Humanin inhibits neuronal
cell death by interacting with a cytokine receptor complex or complexes involving CNTF
receptor alpha/WSX-1/gp130. Mol Biol Cell. 2009;20(12):2864-2873.
doi:10.1091/mbc.E09-02-0168.
99. Harada M, Habata Y, Hosoya M, et al. N-Formylated humanin activates both formyl
peptide receptor-like 1 and 2. Biochem Biophys Res Commun. 2004;324(1):255-261.
doi:10.1016/j.bbrc.2004.09.046.
100. Ying G, Iribarren P, Zhou Y, et al. Humanin, a newly identified neuroprotective factor,
uses the G protein-coupled formylpeptide receptor-like-1 as a functional receptor. J
Immunol. 2004;172(11):7078-7085. http://www.ncbi.nlm.nih.gov/pubmed/15153530.
Accessed December 23, 2015.
101. Hardie DG, Ross FA, Hawley SA. AMPK: A nutrient and energy sensor that maintains
energy homeostasis. Nat Rev Mol Cell Biol. 2012;13(4):251-262. doi:10.1038/nrm3311.
102. Ming W, Lu G, Xin S, et al. Mitochondria related peptide MOTS-c suppresses
ovariectomy-induced bone loss via AMPK activation. Biochem Biophys Res Commun.
2016;476(4):412-419. doi:10.1016/j.bbrc.2016.05.135.
103. Zhai D, Ye Z, Jiang Y, et al. MOTS-c peptide increases survival and decreases bacterial
load in mice infected with MRSA. Mol Immunol. 2017;92:151-160.
doi:10.1016/j.molimm.2017.10.017.

175
104. Fuku N, Pareja-Galeano H, Zempo H, et al. The mitochondrial-derived peptide MOTS-c:
A player in exceptional longevity? Aging Cell. 2015;14(6):921-923.
doi:10.1111/acel.12389.
105. Hoang PT, Park P, Cobb LJ, et al. The neurosurvival factor Humanin inhibits beta-cell
apoptosis via signal transducer and activator of transcription 3 activation and delays and
ameliorates diabetes in nonobese diabetic mice. Metabolism. 2010;59(3):343-349.
doi:10.1016/j.metabol.2009.08.001.
106. Lopez J V., Yuhki N, Masuda R, Modi W, O’Brien SJ. Numt, a recent transfer and
tandem amplification of mitochondrial DNA to the nuclear genome of the domestic cat. J
Mol Evol. 1994;39(2):174-190. doi:10.1007/BF00163806.
107. Bodzioch M, Lapicka-Bodzioch K, Zapala B, Kamysz W, Kiec-Wilk B, Dembinska-Kiec
A. Evidence for potential functionality of nuclearly-encoded humanin isoforms.
Genomics. 2009;94(4):247-256. doi:10.1016/j.ygeno.2009.05.006.
108. Slavoff SA, Mitchell AJ, Schwaid AG, et al. Peptidomic discovery of short open reading
frame–encoded peptides in human cells. Nat Chem Biol. 2012;9(1):59-64.
doi:10.1038/nchembio.1120.
109. Yen K, Lee C, Mehta H, Cohen P. The emerging role of the mitochondrial-derived peptide
humanin in stress resistance. J Mol Endocrinol. 2012;50(1). doi:10.1530/JME-12-0203.
110. White MC, Holman DM, Boehm JE, Peipins LA, Grossman M, Jane Henley S. Age and
Cancer Risk. Am J Prev Med. 2014;46(3):S7-S15. doi:10.1016/j.amepre.2013.10.029.
111. Maximov V, Martynenko A, Hunsmann G, Tarantul V. Mitochondrial 16S rRNA gene
encodes a functional peptide, a potential drug for Alzheimer’s disease and target for
cancer therapy. Med Hypotheses. 2002;59(6):670-673. doi:10.1016/S0306-

176
9877(02)00223-2.
112. Maximov V V, Martynenko A V, Arman IP, Tarantul VZ. Humanin binds MPP8:
mapping interaction sites of the peptide and protein. J Pept Sci. 2013;19(5):301-307.
doi:10.1002/psc.2500.
113. Hartmann TB, Mattern E, Wiedemann N, et al. Identification of selectively expressed
genes and antigens in CTCL. Exp Dermatol. 2008;17(4):324-334. doi:10.1111/j.1600-
0625.2007.00637.x.
114. Lu X, Walker T, MacManus JP, Seligy VL. Differentiation of HT-29 human colonic
adenocarcinoma cells correlates with increased expression of mitochondrial RNA: effects
of trehalose on cell growth and maturation. Cancer Res. 1992;52(13):3718-3725.
115. Corral M, Paris B, Baffet G, et al. Increased level of the mitochondrial ND5 transcript in
chemically induced rat hepatomas. Exp Cell Res. 1989;184(1):158-166. doi:10.1016/0014-
4827(89)90374-1.
116. Eriksson E, Wickström M, Perup LS, et al. Protective role of humanin on bortezomib-
induced bone growth impairment in anticancer treatment. J Natl Cancer Inst.
2014;106(3):djt459. doi:10.1093/jnci/djt459.
117. Wu S, De Luca F. Inhibition of the proteasomal function in chondrocytes down-regulates
growth plate chondrogenesis and longitudinal bone growth. Endocrinology.
2006;147(8):3761-3768. doi:10.1210/en.2005-1672.
118. Zaman F, Menendez-Benito V, Eriksson E, et al. Proteasome inhibition up-regulates p53
and apoptosis-inducing factor in chondrocytes causing severe growth retardation in mice.
Cancer Res. 2007;67(20):10078-10086. doi:10.1158/0008-5472.CAN-06-3982.
119. Liang J, Mills GB. AMPK: A contextual oncogene or tumor suppressor? Cancer Res.

177
2013;73(10):2929-2935. doi:10.1158/0008-5472.CAN-12-3876.
120. Hashimoto Y, Suzuki H, Aiso S, Niikura T, Nishimoto I, Matsuoka M. Involvement of
tyrosine kinases and STAT3 in Humanin-mediated neuroprotection. Life Sci.
2005;77(24):3092-3104. doi:10.1016/j.lfs.2005.03.031.
121. Jin H, Liu T, Wang W-X, et al. Protective effects of [Gly14]-Humanin on beta-amyloid-
induced PC12 cell death by preventing mitochondrial dysfunction. Neurochem Int.
2010;56(3):417-423. doi:10.1016/j.neuint.2009.11.015.
122. Zhang W, Du Y, Bai M, Xi Y, Li Z, Miao J. S14G-humanin inhibits Aβ1-42 fibril
formation, disaggregates preformed fibrils, and protects against Aβ-induced cytotoxicity
in vitro. J Pept Sci. 2013;19(3):159-165. doi:10.1002/psc.2484.
123. Tajima H, Kawasumi M, Chiba T, et al. A humanin derivative, S14G-HN, prevents
amyloid-beta-induced memory impairment in mice. J Neurosci Res. 2005;79(5):714-723.
doi:10.1002/jnr.20391.
124. Miao J, Zhang W, Yin R, et al. S14G-Humanin ameliorates Abeta25-35-induced
behavioral deficits by reducing neuroinflammatory responses and apoptosis in mice.
Neuropeptides. 2008;42(5-6):557-567. doi:10.1016/j.npep.2008.08.004.
125. Zhang W, Zhang W, Li Z, et al. S14G-humanin improves cognitive deficits and reduces
amyloid pathology in the middle-aged APPswe/PS1dE9 mice. Pharmacol Biochem Behav.
2012;100(3):361-369. doi:10.1016/j.pbb.2011.09.012.
126. Niikura T, Sidahmed E, Hirata-Fukae C, Aisen PS, Matsuoka Y. A humanin derivative
reduces amyloid beta accumulation and ameliorates memory deficit in triple transgenic
mice. PLoS One. 2011;6(1):e16259. doi:10.1371/journal.pone.0016259.
127. Park T-Y, Kim S-H, Shin Y-C, et al. Amelioration of neurodegenerative diseases by cell

178
death-induced cytoplasmic delivery of humanin. J Control Release. 2013;166(3):307-315.
doi:10.1016/j.jconrel.2012.12.022.
128. Xu X, Chua CC, Gao J, et al. Neuroprotective effect of humanin on cerebral
ischemia/reperfusion injury is mediated by a PI3K/Akt pathway. Brain Res.
2008;1227:12-18. doi:10.1016/j.brainres.2008.06.018.
129. Xu X, Chua K-W, Chua CC, Liu C-F, Hamdy RC, Chua BHL. Synergistic protective
effects of humanin and necrostatin-1 on hypoxia and ischemia/reperfusion injury. Brain
Res. 2010;1355:189-194. doi:10.1016/j.brainres.2010.07.080.
130. Caricasole A, Bruno V, Cappuccio I, Melchiorri D, Copani A, Nicoletti F. A novel rat
gene encoding a Humanin-like peptide endowed with broad neuroprotective activity.
FASEB J. 2002;16(10):1331-1333. doi:10.1096/fj.02-0018fje.
131. Bachar AR, Scheffer L, Schroeder AS, et al. Humanin is expressed in human vascular
walls and has a cytoprotective effect against oxidized LDL-induced oxidative stress.
Cardiovasc Res. 2010;88(2):360-366. doi:10.1093/cvr/cvq191.
132. Widmer RJ, Flammer AJ, Herrmann J, et al. Circulating humanin levels are associated
with preserved coronary endothelial function. Am J Physiol Heart Circ Physiol.
2013;304(3):H393-7. doi:10.1152/ajpheart.00765.2012.
133. Bachar AR, Scheffer L, Schroeder AS, et al. Humanin is expressed in human vascular
walls and has a cytoprotective effect against oxidized LDL-induced oxidative stress.
Cardiovasc Res. 2010;88(2):360-366. doi:10.1093/cvr/cvq191.
134. Maedler K, Schumann DM, Schulthess F, et al. Aging correlates with decreased β-cell
proliferative capacity and enhanced sensitivity to apoptosis: A potential role for fas and
pancreatic duodenal homeobox-1. Diabetes. 2006;55(9):2455-2462. doi:10.2337/db05-

179
1586.
135. Kahn SE. The relative contributions of insulin resistance and beta-cell dysfunction to the
pathophysiology of Type 2 diabetes. Diabetologia. 2003;46(1):3-19. doi:10.1007/s00125-
002-1009-0.
136. Muzumdar RH, Huffman DM, Atzmon G, et al. Humanin: a novel central regulator of
peripheral insulin action. PLoS One. 2009;4(7):e6334. doi:10.1371/journal.pone.0006334.
137. Voigt A, Jelinek HF. Humanin: A mitochondrial signaling peptide as a biomarker for
impaired fasting glucose-related oxidative stress. Physiol Rep. 2016;4(9):1-5.
doi:10.14814/phy2.12796.
138. Kariya S, Hirano M, Furiya Y, Sugie K, Ueno S. Humanin detected in skeletal muscles of
MELAS patients: a possible new therapeutic agent. Acta Neuropathol. 2005;109(4):367-
372. doi:10.1007/s00401-004-0965-5.
139. Lee C, Kim KH, Cohen P. MOTS-c: A novel mitochondrial-derived peptide regulating
muscle and fat metabolism. Free Radic Biol Med. 2016;100:182-187.
doi:10.1016/j.freeradbiomed.2016.05.015.
140. Mookerjee SA, Goncalves RLS, Gerencser AA, Nicholls DG, Brand MD. The
contributions of respiration and glycolysis to extracellular acid production. Biochim
Biophys Acta - Bioenerg. 2015;1847(2):171-181. doi:10.1016/j.bbabio.2014.10.005.
141. Ashrafi G, Schwarz TL. The pathways of mitophagy for quality control and clearance of
mitochondria. Cell Death Differ. 2013;20(1):31-42. doi:10.1038/cdd.2012.81.
142. Sreekumar PG, Ishikawa K, Spee C, et al. The mitochondrial-derived peptide humanin
protects RPE cells from oxidative stress, senescence, and mitochondrial dysfunction.
Investig Ophthalmol Vis Sci. 2016;57(3):1238-1253. doi:10.1167/iovs.15-17053.

180
143. Kariya S, Takahashi N, Hirano M, Ueno S. Humanin improves impaired metabolic
activity and prolongs survival of serum-deprived human lymphocytes. Mol Cell Biochem.
2003;254(1-2):83-89. doi:10.1023/A:1027372519726.
144. Klein LE, Cui L, Gong Z, Su K, Muzumdar R. A humanin analog decreases oxidative
stress and preserves mitochondrial integrity in cardiac myoblasts. Biochem Biophys Res
Commun. 2013;440(2):197-203. doi:10.1016/j.bbrc.2013.08.055.
145. Bohovych I, Khalimonchuk O. Sending Out an SOS: Mitochondria as a Signaling Hub.
Front Cell Dev Biol. 2016;4. doi:10.3389/fcell.2016.00109.
146. Wellen KE, Hatzivassiliou G, Sachdeva UM, Bui T V, Cross JR, Thompson CB. ATP-
Citrate Lyase Links Cellular Metabolism to Histone Acetylation. Science (80- ).
2009;324(May):1076-1080.
147. Muoio DM, Newgard CB. Mechanisms of disease: molecular and metabolic mechanisms
of insulin resistance and beta-cell failure in type 2 diabetes. Nat Rev Mol Cell Biol.
2008;9(3):193-205. doi:10.1038/nrm2327.
148. Suzuki Y, Taniyama M, Muramatsu T, et al. Diabetes mellitus associated with 3243
mitochondrial tRNA(Leu(UUR)) mutation: clinical features and coenzyme Q10 treatment.
Mol Aspects Med. 1997;18 Suppl:S181-S188. ISBN:0098-2997 (Print)n0098-2997
(Linking).
149. Maechler P, Wollheim CB. Mitochondrial function in normal and diabetic β-cells. Nature.
2001;414(6865):807-812. doi:10.1038/414807a.
150. Maechler P, Kennedy ED, Pozzan T, Wollheim CB. Mitochondrial activation directly
triggers the exocytosis of insulin in permeabilized pancreatic β-cells. EMBO J.
1997;16(13):3833-3841. doi:10.1093/emboj/16.13.3833.

181
151. Maechler P, Wollhelm CB. Mitochondrial glutamate acts as a messenger in glucose-
induced insulin exocytosis. Nature. 1999;402(6762):685-689. doi:10.1038/45280.
152. Kim J-A, Wei Y, Sowers JR. Role of mitochondrial dysfunction in insulin resistance. Circ
Res. 2008;102(4):401-414. doi:10.1161/CIRCRESAHA.107.165472.
153. Oropeza D, Jouvet N, Bouyakdan K, et al. PGC-1 coactivators in β-cells regulate lipid
metabolism and are essential for insulin secretion coupled to fatty acids. Mol Metab.
2015;4(11):811-822. doi:10.1016/j.molmet.2015.08.001.
154. Boden G, Chen X, Ruiz J, White J V., Rossetti L. Mechanisms of fatty acid-induced
inhibition of glucose uptake. J Clin Invest. 1994;93(6):2438-2446.
doi:10.1172/JCI117252.
155. Shah P, Vella A, Basu A, et al. Effects of free fatty acids and glycerol on splanchnic
glucose metabolism and insulin extraction in nondiabetic humans. Diabetes.
2002;51(2):301-310. ISBN:0012-1797 (Print)r0012-1797 (Linking).
156. Boden G. Free fatty acids and insulin secretion in humans. Curr Diab Rep. 2005;5:167-
170. doi:10.1007/s11892-005-0004-5.
157. Pagel-Langenickel I, Bao J, Joseph JJ, et al. PGC-1α integrates insulin signaling,
mitochondrial regulation, and bioenergetic function in skeletal muscle. J Biol Chem.
2008;283(33):22464-22472. doi:10.1074/jbc.M800842200.
158. Auger C, Alhasawi A, Contavadoo M, Appanna VD. Dysfunctional mitochondrial
bioenergetics and the pathogenesis of hepatic disorders. Front Cell Dev Biol. 2015;3.
doi:10.3389/fcell.2015.00040.
159. Yang J, Anzo M, Cohen P. Control of aging and longevity by IGF-I signaling. Exp
Gerontol. 2005;40(11):867-872. doi:10.1016/j.exger.2005.08.001.

182
160. Lee C, Wan J, Miyazaki B, et al. IGF-I regulates the age-dependent signaling peptide
humanin. Aging Cell. 2014;13(5):958-961. doi:10.1111/acel.12243.
161. Bartke A. Can growth hormone (GH) accelerate aging? Evidence from GH-transgenic
mice. Neuroendocrinology. 2003;78(4):210-216. doi:10.1159/000073704.
162. Bartke A, Masternak MM, Al-Regaiey K a, Bonkowski MS. Effects of dietary restriction
on the expression of insulin-signaling-related genes in long-lived mutant mice. Interdiscip
Top Gerontol. 2007;35:69-82. doi:10.1159/000096556.
163. Yakar S, Liu J-L, Stannard B, et al. Normal growth and development in the absence of
hepatic insulin-like growth factor I. Dev Biol. 1999;96(13):7324-7329.
doi:10.1073/pnas.96.13.7324.
164. Yakar S, Rosen CJ, Bouxsein ML, et al. Serum complexes of insulin-like growth factor-1
modulate skeletal integrity and carbohydrate metabolism. FASEB J. 2009;23(3):709-719.
doi:10.1096/fj.08-118976.
165. Guevara-Aguirre J, Balasubramanian P, Guevara-Aguirre M, et al. Growth hormone
receptor deficiency is associated with a major reduction in pro-aging signaling, cancer,
and diabetes in humans. Sci Transl Med. 2011;3(70). doi:10.1126/scitranslmed.3001845.
166. Chistiakov DA, Sobenin IA, Revin V V., Orekhov AN, Bobryshev Y V. Mitochondrial
aging and age-related dysfunction of mitochondria. Biomed Res Int. 2014;2014.
doi:10.1155/2014/238463.
167. Colón E, Strand ML, Carlsson-Skwirut C, et al. Anti-apoptotic factor humanin is
expressed in the testis and prevents cell-death in Leydig cells during the first wave of
spermatogenesis. J Cell Physiol. 2006;208(2):373-385. doi:10.1002/jcp.20672.
168. Chin Y-P, Keni J, Wan J, et al. Pharmacokinetics and tissue distribution of humanin and

183
its analogues in male rodents. Endocrinology. 2013;154(10):3739-3744.
doi:10.1210/en.2012-2004.
169. Juul A. Serum levels of insulin-like growth factor I and its binding proteins in health and
disease. Growth Horm IGF Res. 2003;13(4):113-170. doi:10.1016/S1096-6374(03)00038-
8.
170. Lue Y, Swerdloff R, Wan J, et al. The potent humanin analogue (HNG) protects germ
cells and leucocytes while enhancing chemotherapy-induced suppression of cancer
metastases in male mice. Endocrinology. 2015;156(12):4511-4521.
171. Zhou CK, Check DP, Lortet-Tieulent J, et al. Prostate cancer incidence in 43 populations
worldwide: An analysis of time trends overall and by age group. Int J Cancer.
2016;138(6):1388-1400. doi:10.1002/ijc.29894.
172. Crawford ED, Black L, Eaddy M, Kruep EJ. A retrospective analysis illustrating the
substantial clinical and economic burden of prostate cancer. Prostate Cancer Prostatic
Dis. 2010;13(2):162-167. doi:10.1038/pcan.2009.63.
173. Hatcher D, Daniels G, Osman I, Lee P. Molecular mechanisms involving prostate cancer
racial disparity. Am J Transl Res. 2009;1(3):235-248.
http://www.ncbi.nlm.nih.gov/pubmed/19956434.
174. Kimura T. East meets west: Ethnic differences in prostate cancer epidemiology between
East Asians and Caucasians. Chin J Cancer. 2012;31(9):421-429.
doi:10.5732/cjc.011.10324.
175. Wong MCS, Goggins WB, Wang HHX, et al. Global Incidence and Mortality for Prostate
Cancer: Analysis of Temporal Patterns and Trends in 36 Countries. Eur Urol.
2016;70(5):862-874. doi:10.1016/j.eururo.2016.05.043.

184
176. Bell KJL, Del Mar C, Wright G, Dickinson J, Glasziou P. Prevalence of incidental
prostate cancer: A systematic review of autopsy studies. Int J Cancer. 2015;137(7):1749-
1757. doi:10.1002/ijc.29538.
177. Cazares LH, Drake RR, Esquela-Kirscher A, Lance RS, Semmes OJ, Troyer DA.
Molecular pathology of prostate cancer. Cancer Biomarkers. 2011;9(1-6):441-459.
doi:10.3233/CBM-2011-0181.
178. Hsing AW, Sakoda LC, Chua S. Obesity, metabolic syndrome, and prostate cancer. Am J
Clin Nutr. 2007;86(3):s843-57. http://www.ncbi.nlm.nih.gov/pubmed/18265478.
179. Wallace DC. Mitochondrial Diseases in Man and Mouse. Science (80- ).
1999;283(5407):1482-1488. doi:10.1126/science.283.5407.1482.
180. López-Otín C, Blasco MA, Partridge L, Serrano M, Kroemer G. The hallmarks of aging.
Cell. 2013;153(6). doi:10.1016/j.cell.2013.05.039.
181. Trifunovic A, Wredenberg A, Falkenberg M, et al. Premature ageing in mice expressing
defective mitochondrial DNA polymerase. Nature. 2004;429(6990):417-423.
doi:10.1038/nature02517.
182. Hashimoto Y, Niikura T, Tajima H, et al. A rescue factor abolishing neuronal cell death
by a wide spectrum of familial Alzheimer’s disease genes and A beta. Proc Natl Acad Sci
U S A. 2001;98(11):6336-6341.
183. Lee C, Yen K, Cohen P. Humanin: A harbinger of mitochondrial-derived peptides?
Trends Endocrinol Metab. 2013;24(5):222-228. doi:10.1016/j.tem.2013.01.005.
184. Kim S-J, Guerrero N, Wassef G, et al. The mitochondrial-derived peptide humanin
activates the ERK1/2, AKT, and STAT3 signaling pathways and has age-dependent
signaling differences in the hippocampus. Oncotarget. 2014;7(30):46899-46912.

185
doi:10.18632/oncotarget.10380.
185. Cohen P. New Role for the Mitochondrial Peptide Humanin: Protective Agent Against
Chemotherapy-Induced Side Effects. JNCI J Natl Cancer Inst. 2014;106(3):dju006.
doi:10.1093/jnci/dju006.
186. Cobb LJ, Lee C, Xiao J, et al. Naturally occurring mitochondrial-derived peptides are age-
dependent regulators of apoptosis, insulin sensitivity, and inflammatory markers. Aging
(Albany NY). 2016;8(4):1-14. doi:10.18632/aging.100943.
187. Shiao SL, Chu GC-Y, Chung LWK. Regulation of prostate cancer progression by the
tumor microenvironment. Cancer Lett. 2016;380(1):340-348.
doi:10.1016/j.canlet.2015.12.022.
188. Landskron G, De La Fuente M, Thuwajit P, Thuwajit C, Hermoso MA. Chronic
inflammation and cytokines in the tumor microenvironment. J Immunol Res. 2014;2014.
doi:10.1155/2014/149185.
189. GALTIER N, NABHOLZ B, GLÉMIN S, HURST GDD. Mitochondrial DNA as a marker
of molecular diversity: a reappraisal. Mol Ecol. 2009;18(22):4541-4550.
doi:10.1111/j.1365-294X.2009.04380.x.
190. Mandavilli BS, Santos JH, Van Houten B. Mitochondrial DNA repair and aging. Mutat
Res. 2002;509(1-2):127-151. doi:10.1016/S0027-5107(02)00220-8.
191. Zhao YB, Yang HY, Zhang XW, Chen GY. Mutation in D-loop region of mitochondrial
DNA in gastric cancer and its significance. World J Gastroenterol. 2005;11(21):3304-
3306. doi:10.3748/wjg.v11.i21.3304.
192. Sharma H, Singh A, Sharma C, Jain SK, Singh N. Mutations in the mitochondrial DNA
D-loop region are frequent in cervical cancer. Cancer Cell Int. 2005;5:34.

186
doi:10.1186/1475-2867-5-34.
193. Chang SC, Lin PC, Yang SH, Wang HS, Liang WY, Lin JK. Mitochondrial D-loop
mutation is a common event in colorectal cancers with p53 mutations. Int J Colorectal
Dis. 2009;24(6):623-628. doi:10.1007/s00384-009-0663-9.
194. Payne BAI, Wilson IJ, Yu-Wai-Man P, et al. Universal heteroplasmy of human
mitochondrial DNA. Hum Mol Genet. 2013;22(2):384-390. doi:10.1093/hmg/dds435.
195. Stoneking M. Hypervariable Sites in the mtDNA Control Region Are Mutational Hotspots.
Vol 67.; 2000. doi:10.1086/303092.
196. Chen JZ, Gokden N, Greene GF, Mukunyadzi P, Kadlubar FF. Extensive somatic
mitochondrial mutations in primary prostate cancer using laser capture microdissection.
Cancer Res. 2002;62(22):6470-6474. ISBN:0008-5472 (Print)r0008-5472 (Linking).
197. Jerónimo C, Nomoto S, Caballero OL, et al. Mitochondrial mutations in early stage
prostate cancer and bodily fluids. Oncogene. 2001;20(37):5195-5198.
doi:10.1038/sj.onc.1204646.
198. Wang Y, Liu VWS, Ngan HYS, Nagley P. Frequent occurrence of mitochondrial
microsatellite instability in the D-loop region of human cancers. In: Annals of the New
York Academy of Sciences. Vol 1042. ; 2005:123-129. doi:10.1196/annals.1338.012.
199. Kulawiec M, Owens KM, Singh KK. mtDNA G10398A variant in African-American
women with breast cancer provides resistance to apoptosis and promotes metastasis in
mice. J Hum Genet. 2009;5489(10):647-654. doi:10.1038/jhg.2009.89.
200. Canter JA, Kallianpur AR, Parl FF, Millikan RC. Mitochondrial DNA G10398A
polymorphism and invasive breast cancer in African-American women. Cancer Res.
2005;65(17):8028-8033. doi:10.1158/0008-5472.CAN-05-1428.

187
201. Singh KK, Kulawiec M. Mitochondrial DNA polymorphism and risk of cancer. Methods
Mol Biol. 2009;471:291-303. doi:10.1007/978-1-59745-416-2_15.
202. Petros JA, Baumann AK, Ruiz-Pesini E, et al. mtDNA mutations increase tumorigenicity
in prostate cancer. Proc Natl Acad Sci U S A. 2005;102(3):719-724.
doi:10.1073/pnas.0408894102.
203. Kloss-Brandstätter A, Schäfer G, Erhart G, et al. Somatic mutations throughout the entire
mitochondrial genome are associated with elevated PSA levels in prostate cancer patients.
Am J Hum Genet. 2010;87(6):802-812. doi:10.1016/j.ajhg.2010.11.001.
204. Jerónimo C, Nomoto S, Caballero OL, et al. Mitochondrial mutations in early stage
prostate cancer and bodily fluids. Oncogene. 2001;20(37):5195-5198.
doi:10.1038/sj.onc.1204646.
205. Álvarez-Cubero MJ, Saiz Guinaldo M, Martínez-González LJ, Álvarez Merino JC, Cózar
Olmo JM, Acosta JAL. Mitochondrial haplogroups and polymorphisms reveal no
association with sporadic prostate cancer in a southern European population. PLoS One.
2012;7(7). doi:10.1371/journal.pone.0041201.
206. Takasugi M, Yagi S, Hirabayashi K, Shiota K. DNA methylation status of nuclear-
encoded mitochondrial genes underlies the tissue-dependent mitochondrial functions.
BMC Genomics. 2010;11:481. doi:10.1186/1471-2164-11-481.
207. Biswas G, Anandatheerthavarada HK, Zaidi M, Avadhani NG. Mitochondria to nucleus
stress signaling: a distinctive mechanism of NFκB/Rel activation through calcineurin-
mediated inactivation of IκBβ. J Cell Biol. 2003;161(3):507-519.
doi:10.1083/jcb.200211104.
208. Costello LC, Feng P, Milon B, Tan M, Franklin RB. Role of zinc in the pathogenesis and

188
treatment of prostate cancer: critical issues to resolve. Prostate Cancer Prostatic Dis.
2004;7(2):111-117. doi:10.1038/sj.pcan.4500712.
209. Kang D, Lee K-M, Park SK, et al. Functional variant of manganese superoxide dismutase
(SOD2 V16A) polymorphism is associated with prostate cancer risk in the prostate, lung,
colorectal, and ovarian cancer study. Cancer Epidemiol Biomarkers Prev.
2007;16(8):1581-1586. doi:10.1158/1055-9965.EPI-07-0160.
210. Chaudhary AK, Bhat TA, Kumar S, et al. Mitochondrial dysfunction-mediated apoptosis
resistance associates with defective heat shock protein response in African–American men
with prostate cancer. Br J Cancer. 2016;114(April):1-11. doi:10.1038/bjc.2016.88.
211. Kelly RDW, Mahmud A, McKenzie M, Trounce IA, St John JC. Mitochondrial DNA
copy number is regulated in a tissue specific manner by DNA methylation of the nuclear-
encoded DNA polymerase gamma A. Nucleic Acids Res. 2012;40(20):10124-10138.
doi:10.1093/nar/gks770.
212. Zhao H, Zhao Y, Ren Y, et al. Epigenetic regulation of an adverse metabolic phenotype in
polycystic ovary syndrome: the impact of the leukocyte methylation of PPARGC1A
promoter. Fertil Steril. 2017;107(2):467-474.e5. doi:10.1016/j.fertnstert.2016.10.039.
213. Kenney MC, Chwa M, Atilano SR, et al. Molecular and bioenergetic differences between
cells with African versus European inherited mitochondrial DNA haplogroups:
Implications for population susceptibility to diseases. Biochim Biophys Acta - Mol Basis
Dis. 2014;1842(2):208-219. doi:10.1016/j.bbadis.2013.10.016.
214. Xiao J, Howard L, Wan J, et al. Low Circulating Levels of the Mitochondrial-Peptide
Hormone SHLP2: Novel Biomarker for Prostate Cancer Risk.; 2017.
doi:10.18632/oncotarget.20134.

189
215. Kalsbeek AMF, Chan EFK, Grogan J, et al. Mutational load of the mitochondrial genome
predicts pathological features and biochemical recurrence in prostate cancer. Aging
(Albany, NY). 2016;8(11):2702-2711. doi:10.18632/aging.101044.
216. Arnold RS, Sun CQ, Richards JC, et al. Mitochondrial DNA mutation stimulates prostate
cancer growth in bone stromal environment. Prostate. 2009;69(1):1-11.
doi:10.1002/pros.20854.
217. Booker LM, Habermacher GM, Jessie BC, et al. North American white mitochondrial
haplogroups in prostate and renal cancer. J Urol. 2006;175(2):468-472.
doi:10.1016/S0022-5347(05)00163-1.
218. Cano D, Gomez CF, Ospina N, et al. Mitochondrial DNA haplogroups and susceptibility
to prostate cancer in a colombian population. ISRN Oncol. 2014;2014:530675.
doi:10.1155/2014/530675.
219. Kim W, Yoo TK, Shin DJ, et al. Mitochondrial DNA haplogroup analysis reveals no
association between the common genetic lineages and prostate cancer in the Korean
population. PLoS One. 2008;3(5). doi:10.1371/journal.pone.0002211.
220. Grupp K, Jedrzejewska K, Tsourlakis MC, et al. High mitochondria content is associated
with prostate cancer disease progression. Mol Cancer. 2013;12(1):145. doi:10.1186/1476-
4598-12-145.
221. Reznik E, Miller ML, Şenbabaoğlu Y, et al. Mitochondrial DNA copy number variation
across human cancers. Elife. 2016;5(FEBRUARY2016). doi:10.7554/eLife.10769.
222. Bai R-K, Wong L-JC. Simultaneous detection and quantification of mitochondrial DNA
deletion(s), depletion, and over-replication in patients with mitochondrial disease. J Mol
diagnostics JMD. 2005;7(5):613-622. doi:10.1016/S1525-1578(10)60595-8.

190
223. Koochekpour S, Marlowe T, Singh KK, Attwood K, Chandra D. Reduced Mitochondrial
DNA Content Associates with Poor Prognosis of Prostate Cancer in African American
Men. PLoS One. 2013;8(9). doi:10.1371/journal.pone.0074688.
224. Pelicano H, Xu RH, Du M, et al. Mitochondrial respiration defects in cancer cells cause
activation of Akt survival pathway through a redox-mediated mechanism. J Cell Biol.
2006;175(6):913-923. doi:10.1083/jcb.200512100.
225. Kolenko V, Teper E, Kutikov A, Uzzo R. Zinc and zinc transporters in prostate
carcinogenesis. Nat Rev Urol. 2013;10(4):219-226. doi:10.1038/nrurol.2013.43.
226. Desouki MM, Geradts J, Milon B, Franklin RB, Costello LC. hZip2 and hZip3 zinc
transporters are down regulated in human prostate adenocarcinomatous glands. Mol
Cancer. 2007;6(1):37. doi:10.1186/1476-4598-6-37.
227. Rishi I, Baidouri H, Abbasi JA, et al. Prostate cancer in African American men is
associated with downregulation of zinc transporters. Appl Immunohistochem Mol
Morphol. 2003;11:253-260. doi:10.1097/00129039-200309000-00009.
228. Rico-bautista E, Zhu W, Kitada S, Ganapathy S. Small Molecule-Induced Mitochondrial
Disruption Directs Prostate Cancer Inhibition via Unfolded Protein Response Signaling
ABSTRACT : Oncotarget. 2013;4(8):1212-1229.
229. Agrimi G, Di Noia MA, Marobbio CMT, Fiermonte G, Lasorsa FM, Palmieri F.
Identification of the human mitochondrial S-adenosylmethionine transporter: bacterial
expression, reconstitution, functional characterization and tissue distribution. Biochem J.
2004;379:183-190. doi:10.1042/bj20031664.
230. Shock LS, Thakkar P V., Peterson EJ, Moran RG, Taylor SM. DNA methyltransferase 1,
cytosine methylation, and cytosine hydroxymethylation in mammalian mitochondria. Proc

191
Natl Acad Sci. 2011;108(9):3630-3635. doi:10.1073/pnas.1012311108.
231. Chen H, Dzitoyeva S, Manev H. Effect of valproic acid on mitochondrial epigenetics. Eur
J Pharmacol. 2012;690(1-3):51-59. doi:10.1016/j.ejphar.2012.06.019.
232. Ghosh S, Sengupta S, Scaria V. Comparative analysis of human mitochondrial
methylomes shows distinct patterns of epigenetic regulation in mitochondria.
Mitochondrion. 2014;18:58-62. doi:10.1016/j.mito.2014.07.007.
233. Yu D, Du Z, Pian L, et al. Mitochondrial DNA Hypomethylation Is a Biomarker
Associated with Induced Senescence in Human Fetal Heart Mesenchymal Stem Cells.
Stem Cells Int. 2017;2017.
234. Chinnery PF, Elliott HR, Hudson G, Samuels DC, Relton CL. Epigenetics, epidemiology
and mitochondrial DNA diseases. Int J Epidemiol. 2012;41(1):177-187.
doi:10.1093/ije/dyr232.
235. Massie CE, Mills IG, Lynch AG. The importance of DNA methylation in prostate cancer
development. J Steroid Biochem Mol Biol. 2016;166:1-15.
doi:10.1016/j.jsbmb.2016.04.009.
236. Lee WH, Morton RA, Epstein JI, et al. Cytidine methylation of regulatory sequences near
the pi-class glutathione S-transferase gene accompanies human prostatic carcinogenesis.
Proc Natl Acad Sci U S A. 1994;91(24):11733-11737. doi:10.1073/pnas.91.24.11733.
237. Gong M, Dong W, Shi Z, Xu Y, Ni W, An R. Genetic Polymorphisms of GSTM1,
GSTT1, and GSTP1 with Prostate Cancer Risk: A Meta-Analysis of 57 Studies. PLoS
One. 2012;7(11). doi:10.1371/journal.pone.0050587.
238. Bellizzi D, D’Aquila P, Giordano M, Montesanto A, Passarino G. Global DNA
methylation levels are modulated by mitochondrial DNA variants. Epigenomics.

192
2012;4:17-27. doi:10.2217/epi.11.109.
239. Park S-Y, Murphy SP, Wilkens LR, Henderson BE, Kolonel LN. Fat and meat intake and
prostate cancer risk: the multiethnic cohort study. Int J Cancer. 2007;121(6):1339-1345.
doi:10.1002/ijc.22805.
240. Raina K, Blouin MJ, Singh RP, et al. Dietary feeding of silibinin inhibits prostate tumor
growth and progression in transgenic adenocarcinoma of the mouse prostate model.
Cancer Res. 2007;67(22):11083-11091. doi:10.1158/0008-5472.CAN-07-2222.
241. Ngo TH, Barnard RJ, Cohen P, et al. Effect of isocaloric low-fat diet on human LAPC-4
prostate cancer xenografts in severe combined immunodeficient mice and the insulin-like
growth factor axis. Clin Cancer Res. 2003;9(7):2734-2743. ISBN:1078-0432 (Print) 1078-
0432 (Linking).
242. Allott EH, Arab L, Su LJ, et al. Saturated fat intake and prostate cancer aggressiveness:
results from the population-based North Carolina-Louisiana Prostate Cancer Project.
Prostate Cancer Prostatic Dis. 2017;20(1):48-54. http://dx.doi.org/10.1038/pcan.2016.39.
243. di Sebastiano KM, Mourtzakis M. The role of dietary fat throughout the prostate cancer
trajectory. Nutrients. 2014;6(12):6095-6109. doi:10.3390/nu6126095.
244. Sadowski MC, Pouwer RH, Gunter JH, Lubik AA, Quinn RJ, Nelson CC. The fatty acid
synthase inhibitor triclosan: repurposing an anti-microbial agent for targeting prostate
cancer. Oncotarget. 2014;5(19):9362-9381. doi:10.18632/oncotarget.2433.
245. Schadler KL, Thomas NJ, Galie PA, et al. Tumor vessel normalization after aerobic
exercise enhances chemotherapeutic efficacy. Oncotarget. 2016;7(40):65429-65440.
doi:10.18632/oncotarget.11748.
246. Pin F, Busquets S, Toledo M, et al. Combination of exercise training and erythropoietin

193
prevents cancer-induced muscle alterations. Oncotarget. 2015;6(41):43202-43215.
doi:10.18632/oncotarget.6439.
247. Kenfield SA, Stampfer MJ, Giovannucci E, Chan JM. Physical activity and survival after
prostate cancer diagnosis in the health professionals follow-up study. J Clin Oncol.
2011;29(6):726-732. doi:10.1200/JCO.2010.31.5226.
248. Bonn SE, Sjölander A, Lagerros YT, et al. Physical activity and survival among men
diagnosed with prostate cancer. Cancer Epidemiol Biomarkers Prev. 2015;24(1):57-64.
doi:10.1158/1055-9965.EPI-14-0707.
249. Barnard RJ, Leung PS, Aronson WJ, Cohen P, Golding L a. A mechanism to explain how
regular exercise might reduce the risk for clinical prostate cancer. Eur J Cancer Prev.
2007;16(5):415-421. doi:10.1097/01.cej.0000243851.66985.e4.
250. Aronson WJ, Barnard RJ, Freedland SJ, et al. Growth inhibitory effect of low fat diet on
prostate cancer cells: results of a prospective, randomized dietary intervention trial in men
with prostate cancer. J Urol. 2010;183(1):345-350. doi:10.1016/j.juro.2009.08.104.
251. Xiao J, Kim SJ, Cohen P, Yen K. Humanin: Functional Interfaces with IGF-I. Growth
Horm IGF Res. 2016;29:21-27. doi:10.1016/j.ghir.2016.03.005.
252. Fontana L, Partridge L, Longo VD. Extending healthy life span-from yeast to humans.
Science (80- ). 2010;328(5976):321-326. doi:10.1126/science.1172539.
253. Lee C, Longo V. Dietary restriction with and without caloric restriction for healthy aging.
F1000Research. 2016. doi:10.12688/f1000research.7136.1.
254. Blagosklonny M V. Calorie restriction: Decelerating mTOR-driven aging from cells to
organisms (including humans). Cell Cycle. 2010;9(4):683-688. doi:10.4161/cc.9.4.10766.
255. Canto C, Auwerx J. Calorie Restriction: Is AMPK a Key Sensor and Effector? Physiology.

194
2011;26(4):214-224. doi:10.1152/physiol.00010.2011.
256. Guarente L. Calorie restriction and sirtuins revisited. Genes Dev. 2013;27(19):2072-2085.
doi:10.1101/gad.227439.113.
257. Speakman JR, Mitchell SE. Caloric restriction. Mol Aspects Med. 2011;32(3):159-221.
doi:10.1016/j.mam.2011.07.001.
258. Klass MR. A method for the isolation of longevity mutants in the nematode
Caenorhabditis elegans and initial results. Mech Ageing Dev. 1983;22(3-4):279-286.
doi:10.1016/0047-6374(83)90082-9.
259. Kenyon C, Chang J, Gensch E, Rudner A, Tabtiang R. A C. elegans mutant that lives
twice as long as wild type. Nature. 1993;366:461-464.
260. Apfeld J, Kenyon C. Regulation of lifespan by sensory perception in Caenorhabditis
elegans. Nature. 1999;402(6763):804-809. doi:10.1038/45544.
261. Clancy DJ, Gems D, Harshman LG, et al. Extension of life-span by loss of CHICO, a
Drosophila insulin receptor substrate protein. Science (80- ). 2001;292(5514):104-106.
doi:10.1126/science.1057991.
262. Bartke A. Minireview: Role of the Growth Hormone/Insulin-Like Growth Factor System
in Mammalian Aging. Endocrinology. 2005;146(9):3718-3723. doi:10.1210/en.2005-
0411.
263. Brown-Borg HM, Borg KE, Meliska CJ, Bartke A. Dwarf mice and the ageing process.
Nature. 1996;384(6604):33-33. doi:10.1038/384033a0.
264. Coschigano KT, Clemmons D, Bellush LL, Kopchick JJ. Assessment of growth
parameters and life span of GHR/BP gene-disrupted mice. Endocrinology.
2000;141(7):2608-2613. doi:10.1210/en.141.7.2608.

195
265. Tsuchiya T, Dhahbi JM, Cui X, Mote PL, Bartke A, Spindler SR. Additive regulation of
hepatic gene expression by dwarfism and caloric restriction. Physiol Genomics.
2004;17(3):307-315. doi:10.1152/physiolgenomics.00039.2004.
266. Bartke A, Wright JC, Mattison JA, Ingram DK, Miller RA, Roth GS. Extending the
lifespan of long-lived mice. Nature. 2001;414(6862):412. doi:10.1038/35106646.
267. Turcotte B, Liang XB, Robert F, Soontorngun N. Transcriptional regulation of
nonfermentable carbon utilization in budding yeast. FEMS Yeast Res. 2010;10(1):2-13.
doi:10.1111/j.1567-1364.2009.00555.x.
268. Schulz TJ, Zarse K, Voigt A, Urban N, Birringer M, Ristow M. Glucose Restriction
Extends Caenorhabditis elegans Life Span by Inducing Mitochondrial Respiration and
Increasing Oxidative Stress. Cell Metab. 2007;6(4):280-293.
doi:10.1016/j.cmet.2007.08.011.
269. Greer EL, Dowlatshahi D, Banko MR, et al. An AMPK-FOXO Pathway Mediates
Longevity Induced by a Novel Method of Dietary Restriction in C. elegans. Curr Biol.
2007;17(19):1646-1656. doi:10.1016/j.cub.2007.08.047.
270. Funakoshi M, Tsuda M, Muramatsu K, Hatsuda H, Morishita S, Aigaki T. A gain-of-
function screen identifies wdb and lkb1 as lifespan-extending genes in Drosophila.
Biochem Biophys Res Commun. 2011;405(4):667-672. doi:10.1016/j.bbrc.2011.01.090.
271. Stenesen D, Suh JM, Seo J, et al. Adenosine nucleotide biosynthesis and AMPK regulate
adult life span and mediate the longevity benefit of caloric restriction in flies. Cell Metab.
2013;17(1):101-112. doi:10.1016/j.cmet.2012.12.006.
272. Muzumdar RH, Huffman DM, Atzmon G, et al. Humanin: A novel central regulator of
peripheral insulin action. PLoS One. 2009;4(7). doi:10.1371/journal.pone.0006334.

196
273. Zhang X, Urbieta-Caceres VH, Eirin A, et al. Humanin prevents intra-renal microvascular
remodeling and inflammation in hypercholesterolemic ApoE deficient mice. Life Sci.
2012;91(5-6):199-206. doi:10.1016/j.lfs.2012.07.010.
274. Zacharias DG, Kim SG, Massat AE, et al. Humanin, a cytoprotective peptide, is expressed
in carotid artherosclerotic plaques in humans. PLoS One. 2012;7(2).
doi:10.1371/journal.pone.0031065.
275. Mamiya T, Ukai M. [Gly(14)]-Humanin improved the learning and memory impairment
induced by scopolamine in vivo. Br J Pharmacol. 2001;134(8):1597-1599.
doi:10.1038/sj.bjp.0704429.
276. Matsuoka M, Hashimoto Y, Aiso S, Nishimoto I. Humanin and colivelin: Neuronal-death-
suppressing peptides for Alzheimer’s disease and amyotrophic lateral sclerosis. CNS Drug
Rev. 2006;12(2):113-122. doi:10.1111/j.1527-3458.2006.00113.x.
277. Xu X, Chua CC, Gao J, Hamdy RC, Chua BHL. Humanin is a novel neuroprotective agent
against stroke. Stroke. 2006;37(10):2613-2619.
doi:10.1161/01.STR.0000242772.94277.1f.
278. Zhao ST, Zhao L, Li JH. Neuroprotective peptide humanin inhibits inflammatory response
in astrocytes induced by lipopolysaccharide. Neurochem Res. 2013;38(3):581-588.
doi:10.1007/s11064-012-0951-6.
279. Tannenbaum A. The Dependence of Tumor Formation on the Composition of the Calorie-
Restricted Diet as Well as on the Degree of Restriction. Cancer Res. 1945;5(11):616-625.
doi:10.1016/S0899-9007(96)00179-7.
280. Weindruch R, Walford R. Dietary restriction in mice beginning at 1 year of age: effect on
life-span and spontaneous cancer incidence. Science (80- ). 1982;215(4538):1415-1418.

197
doi:10.1126/science.7063854.
281. Dunn SE, Kari FW, French J, et al. Dietary restriction reduces insulin-like growth factor I
levels, which modulates apoptosis, cell proliferation, and tumor progression in p53-
deficient mice. Cancer Res. 1997;57(21):4667-4672. ISBN:0008-5472 (Print)r0008-5472
(Linking).
282. Drumond AL, Weng CC, Wang G, Chiarini-Garcia H, Eras-Garcia L, Meistrich ML.
Effects of multiple doses of cyclophosphamide on mouse testes: Accessing the germ cells
lost, and the functional damage of stem cells. Reprod Toxicol. 2011;32(4):395-406.
doi:10.1016/j.reprotox.2011.09.010.
283. Au WW, Finch M, da Cunha MF. Damaging effects of fourteen chemotherapeutic drugs
on mouse testis cells. Cancer Res. 1982;42(1):122-131. ISBN:0008-5472 (Print)r0008-
5472 (Linking).
284. Huyan X-H, Lin Y-P, Gao T, Chen R-Y, Fan Y-M. Immunosuppressive effect of
cyclophosphamide on white blood cells and lymphocyte subpopulations from peripheral
blood of Balb/c mice. Int Immunopharmacol. 2011;11(9):1293-1297.
doi:10.1016/j.intimp.2011.04.011.
285. D’Agostini C, Pica F, Febbraro G, Grelli S, Chiavaroli C, Garaci E. Antitumour effect of
OM-174 and Cyclophosphamide on murine B16 melanoma in different experimental
conditions. Int Immunopharmacol. 2005;5(7-8):1205-1212.
doi:10.1016/j.intimp.2005.02.013.
286. De Felice FG, Ferreira ST. Inflammation, defective insulin signaling, and mitochondrial
dysfunction as common molecular denominators connecting type 2 diabetes to Alzheimer
Disease. Diabetes. 2014;63(7):2262-2272. doi:10.2337/db13-1954.

198
287. Barnes CA. Spatial learning and memory processes: the search for their neurobiological
mechanisms in the rat. Trends Neurosci. 1988;11(4):163-169. doi:10.1016/0166-
2236(88)90143-9.
288. Beninger RJ, Jhamandas K, Boegman RJ, El-Defrawy SR. Effects of scopolamine and
unilateral lesions of the basal forebrain on T-maze spatial discrimination and alternation in
rats. Pharmacol Biochem Behav. 1986;24(5):1353-1360. doi:10.1016/0091-
3057(86)90195-4.
289. Shiotsuki H, Yoshimi K, Shimo Y, et al. A rotarod test for evaluation of motor skill
learning. J Neurosci Methods. 2010;189(2):180-185. doi:10.1016/j.jneumeth.2010.03.026.
290. Viollet B. AMPK: Lessons from transgenic and knockout animals. Front Biosci.
2009;Volume(14):19. doi:10.2741/3229.
291. Greer EL, Banko MR, Brunet A. AMP-activated protein kinase and FoxO transcription
factors in dietary restriction-induced longevity. In: Annals of the New York Academy of
Sciences. Vol 1170. ; 2009:688-692. doi:10.1111/j.1749-6632.2009.04019.x.
292. Yaffe D, Saxel O. Serial passaging and differentiation of myogenic cells isolated from
dystrophic mouse muscle. Nature. 1977;270(5639):725-727. doi:10.1038/270725a0.
293. BAAR K. Adaptations of skeletal muscle to exercise: rapid increase in the transcriptional
coactivator PGC-1. FASEB J. 2002;16(14):1879-1886. doi:10.1096/fj.02-0367com.
294. Longo N, Frigeni M, Pasquali M. Carnitine transport and fatty acid oxidation. Biochim
Biophys Acta - Mol Cell Res. 2016;1863(10):2422-2435.
doi:10.1016/j.bbamcr.2016.01.023.
295. Valdivieso M, Kujawa AM, Jones T, Baker LH. Cancer survivors in the united states: A
review of the literature and a call to action. Int J Med Sci. 2012;9(2):163-173.

199
doi:10.7150/ijms.3827.
296. Kopp LM, Gupta P, Pelayo-Katsanis L, Wittman B, Katsanis E. Late Effects in Adult
Survivors of Pediatric Cancer: A Guide for the Primary Care Physician. Am J Med.
2012;125(7):636-641. doi:10.1016/j.amjmed.2012.01.013.
297. Surampudi P, Chang I, Lue Y, et al. Humanin protects against chemotherapy-induced
stage-specific male germ cell apoptosis in rats. Andrology. 2015;3(3):582-589.
doi:10.1111/andr.12036.
298. Colman RJ, Anderson RM, Johnson SC, et al. Caloric restriction delays disease onset and
mortality in rhesus monkeys. Science (80- ). 2009;325(5937):201-204.
doi:10.1126/science.1173635.
299. Hashimoto Y, Kurita M, Matsuoka M. Identification of soluble WSX-1 not as a dominant-
negative but as an alternative functional subunit of a receptor for an anti-Alzheimer’s
disease rescue factor Humanin. Biochem Biophys Res Commun. 2009;389(1):95-99.
doi:10.1016/j.bbrc.2009.08.095.
300. Guo B, Zhai D, Cabezas E, Welsh K, Nouraini S. Humanin peptide suppresses apoptosis
by interfering with Bax activation. Nature. 2003;423(May):456-461.
doi:10.1038/nature01608.1.
301. Zhai D, Luciano F, Zhu X, Guo B, Satterthwait AC, Reed JC. Humanin binds and nullifies
Bid activity by blocking its activation of Bax and Bak. J Biol Chem. 2005;280(16):15815-
15824. doi:10.1074/jbc.M411902200.
302. Luciano F, Zhai D, Zhu X, et al. Cytoprotective peptide humanin binds and inhibits
proapoptotic Bcl-2/Bax family protein BimEL. J Biol Chem. 2005;280(16):15825-15835.
doi:10.1074/jbc.M413062200.

200
303. Jia Y, Lue YH, Swerdloff R, et al. The cytoprotective peptide humanin is induced and
neutralizes Bax after pro-apoptotic stress in the rat testis. Andrology. 2013;1(4):651-659.
doi:10.1111/j.2047-2927.2013.00091.x.
304. Raffaghello L, Lee C, Safdie FM, et al. Starvation-dependent differential stress resistance
protects normal but not cancer cells against high-dose chemotherapy. Proc Natl Acad Sci.
2008;105(24):8215-8220. doi:10.1073/pnas.0708100105.
305. Lee C, Longo VD. Fasting vs dietary restriction in cellular protection and cancer
treatment: From model organisms to patients. Oncogene. 2011;30(30):3305-3316.
doi:10.1038/onc.2011.91.
306. Heidegger I, Pircher A, Klocker H, Massoner P. Targeting the insulin-like growth factor
network in cancer therapy. Cancer Biol Ther. 2011;11(8):701-707.
doi:10.4161/cbt.11.8.14689.
307. Tognon CE, Sorensen PHB. Targeting the insulin-like growth factor 1 receptor (IGF1R)
signaling pathway for cancer therapy. Expert Opin Ther Targets. 2012;16(1):33-48.
doi:10.1517/14728222.2011.638626.
308. Lee C, Wan J, Miyazaki B, et al. IGF-I regulates the age-dependent signaling peptide
humanin. Aging Cell. 2014:958-961. doi:10.1111/acel.12243.
309. Capoluongo E. Insulin-like growth factor system and sporadic malignant melanoma. Am J
Pathol. 2011;178(1):26-31. doi:10.1016/j.ajpath.2010.11.004.
310. Arnaldez FI, Helman LJ. Targeting the Insulin Growth Factor Receptor 1. Hematol Oncol
Clin North Am. 2012;26(3):527-542. doi:10.1016/j.hoc.2012.01.004.
311. Tricoli J V, Rall LB, Karakousis CP, et al. Enhanced levels of insulin-like growth factor
messenger RNA in human colon carcinomas and liposarcomas. Cancer Res. 1986;46(12

201
Pt 1):6169-6173. ISBN:0008-5472 (Print)r0008-5472 (Linking).
312. Harman SM, Metter EJ, Blackman MR, Landis PK, Carter HB. Serum levels of insulin-
like growth factor I (IGF-I), IGF-II, IGF-binding protein-3, and prostate-specific antigen
as predictors of clinical prostate cancer. J Clin Endocrinol Metab. 2000;85(0021-
972X):4258-4265. doi:10.1210/jcem.85.11.6990.
313. Chan JM, Stampfer MJ, Giovannucci E, et al. Plasma insulin-like growth factor-I and
prostate cancer risk: a prospective study. Science (80- ). 1998;279(5350):563-566.
doi:10.1126/science.279.5350.563.
314. Hormones TE, Cancer B, Group C. Insulin-like growth factor 1 (IGF1), IGF binding
protein 3 (IGFBP3), and breast cancer risk: Pooled individual data analysis of 17
prospective studies. Lancet Oncol. 2010;11(6):530-542. doi:10.1016/S1470-
2045(10)70095-4.
315. Chen W, Wang S, Tian T, et al. Phenotypes and genotypes of insulin-like growth factor 1,
IGF-binding protein-3 and cancer risk: Evidence from 96 studies. Eur J Hum Genet.
2009;17(12):1668-1675. doi:10.1038/ejhg.2009.86.
316. Puche JE, Castilla-Cortázar I. Human conditions of insulin-like growth factor-I (IGF-I)
deficiency. J Transl Med. 2012;10(1). doi:10.1186/1479-5876-10-224.
317. Lee C, Safdie FM, Raffaghello L, et al. Reduced levels of IGF-I mediate differential
protection of normal and cancer cells in response to fasting and improve chemotherapeutic
index. Cancer Res. 2010;70(4):1564-1572. doi:10.1158/0008-5472.CAN-09-3228.
318. Flurkey K, M. Currer J, Harrison DE. The Mouse in Biomedical Research. Vol III.; 2007.
doi:10.1016/B978-012369454-6/50074-1.
319. Hamrick MW, Ding KH, Pennington C, et al. Age-related loss of muscle mass and bone

202
strength in mice is associated with a decline in physical activity and serum leptin. Bone.
2006;39(4):845-853. doi:S8756-3282(06)00416-9 [pii] 10.1016/j.bone.2006.04.011.
320. Muzumdar R, Allison DB, Huffman DM, et al. Visceral adipose tissue modulates
mammalian longevity. Aging Cell. 2008;7(3):438-440. doi:10.1111/j.1474-
9726.2008.00391.x.
321. Hajer GR, Van Haeften TW, Visseren FLJ. Adipose tissue dysfunction in obesity,
diabetes, and vascular diseases. Eur Heart J. 2008;29:2959-2971.
doi:10.1093/eurheartj/ehn387.
322. Trayhurn P, Wood IS. Signalling role of adipose tissue: adipokines and inflammation in
obesity. Biochem Soc Trans. 2005;33(Pt 5):1078-1081. doi:10.1042/BST20051078.
323. Guarente L, Kenyon C. Genetic pathways that regulate ageing in model organisms.
Nature. 2000;408(6809):255-262. doi:10.1038/35041700.
324. Junnila RK, List EO, Berryman DE, Murrey JW, Kopchick JJ. The GH/IGF-1 axis in
ageing and longevity. Nat Rev Endocrinol. 2013;9(6):366-376.
doi:10.1038/nrendo.2013.67.
325. Rodier F, Campisi J. Four faces of cellular senescence. J Cell Biol. 2011;192(4):547-556.
doi:10.1083/jcb.201009094.
326. Varadhan R, Yao W, Matteini A, et al. Simple biologically informed infammatory index
of two serum cytokines predicts 10 year all-cause mortality in older adults. Journals
Gerontol - Ser A Biol Sci Med Sci. 2014;69 A(2):165-173. doi:10.1093/gerona/glt023.
327. Kershaw EE, Flier JS. Adipose tissue as an endocrine organ. J Clin Endocrinol Metab.
2004;89(6):2548-2556. doi:10.1210/jc.2004-0395.
328. Fahlström A, Yu Q, Ulfhake B. Behavioral changes in aging female C57BL/6 mice.

203
Neurobiol Aging. 2011;32(10):1868-1880.
doi:10.1016/J.NEUROBIOLAGING.2009.11.003.
329. Barnes CA. Memory deficits associated with senescence: A neurophysiological and
behavioral study in the rat. J Comp Physiol Psychol. 1979;93(1):74-104.
doi:10.1037/h0077579.
330. Raber J, Rola R, LeFevour A, et al. Radiation-Induced Cognitive Impairments are
Associated with Changes in Indicators of Hippocampal Neurogenesis. Radiat Res.
2004;162(1):39-47. doi:10.1667/RR3206.
331. Frank MG, Barrientos RM, Hein AM, Biedenkapp JC, Watkins LR, Maier SF. IL-1RA
blocks E. coli-induced suppression of Arc and long-term memory in aged F344 × BN F1
rats. Brain Behav Immun. 2010;24(2):254-262. doi:10.1016/j.bbi.2009.10.005.
332. Terrando N, Rei Fidalgo A, Vizcaychipi M, et al. The impact of IL-1 modulation on the
development of lipopolysaccharide-induced cognitive dysfunction. Crit Care. 2010;14(3).
doi:10.1186/cc9019.
333. Shoji H, Takao K, Hattori S, Miyakawa T. Age-related changes in behavior in C57BL/6J
mice from young adulthood to middle age. Mol Brain. 2016;9(1). doi:10.1186/s13041-
016-0191-9.
334. Peterson CM, Johannsen DL, Ravussin E. Skeletal muscle mitochondria and aging: A
review. J Aging Res. 2012;2012. doi:10.1155/2012/194821.
335. Narkar VA, Downes M, Yu RT, et al. AMPK and PPARδ Agonists Are Exercise
Mimetics. Cell. 2008;134(3):405-415. doi:10.1016/j.cell.2008.06.051.
336. Kobilo T, Guerrieri D, Zhang Y, Collica SC, Becker KG, van Praag H. AMPK agonist
AICAR improves cognition and motor coordination in young and aged mice. Learn Mem.

204
2014;21(2):119-126. doi:10.1101/lm.033332.113.
337. Winder WW, Holmes BF, Rubink DS, Jensen EB, Chen M, Holloszy JO. Activation of
AMP-activated protein kinase increases mitochondrial enzymes in skeletal muscle. J Appl
Physiol. 2000;88(6):2219-2226. ISBN:8750-7587 (Print)r0161-7567 (Linking).
338. Bergeron R, Ren JM, Cadman KS, et al. Chronic activation of AMP kinase results in
NRF-1 activation and mitochondrial biogenesis. Am J Physiol Endocrinol Metab.
2001;281(6):E1340-6. doi:10.1016/j.bbabio.2012.02.033.
339. Wu Z, Puigserver P, Andersson U, et al. Mechanisms controlling mitochondrial
biogenesis and respiration through the thermogenic coactivator PGC-1. Cell.
1999;98(1):115-124. doi:10.1016/S0092-8674(00)80611-X.
340. Lin J, Wu H, Tarr PT, et al. Transcriptional co-activator PGC-1α drives the formation of
slow-twitch muscle fibres. Nature. 2002;418(6899):797-801. doi:10.1038/nature00904.
341. Plaisier CL, O’Brien S, Bernard B, et al. Causal Mechanistic Regulatory Network for
Glioblastoma Deciphered Using Systems Genetics Network Analysis. Cell Syst.
2016;3(2):172-186. doi:10.1016/j.cels.2016.06.006.
342. Piantadosi C a, Suliman HB. Mitochondrial transcription factor A induction by redox
activation of nuclear respiratory factor 1. J Biol Chem. 2006;281(1):324-333.
doi:10.1074/jbc.M508805200.
343. Honda A, Yamashita K, Ikegami T, et al. Highly sensitive quantification of serum
malonate, a possible marker for de novo lipogenesis, by LC-ESI-MS/MS. J Lipid Res.
2009;50(10):2124-2130. doi:10.1194/jlr.D800054-JLR200.
344. Wende AR, Huss JM, Schaeffer PJ, Giguere V, Kelly DP. PGC-1 Coactivates PDK4
Gene Expression via the Orphan Nuclear Receptor ERR : a Mechanism for

205
Transcriptional Control of Muscle Glucose Metabolism. Mol Cell Biol.
2005;25(24):10684-10694. doi:10.1128/MCB.25.24.10684-10694.2005.
345. Hwang DL, Lee PDK, Cohen P. Quantitative ontogeny of murine insulin-like growth
factor (IGF)-I, IGF-binding protein-3 and the IGF-related acid-labile subunit. Growth
Horm IGF Res. 2008;18(1):65-74. doi:10.1016/j.ghir.2007.07.007.
346. Baghirova S, Hughes BG, Hendzel MJ, Schulz R. Sequential fractionation and isolation of
subcellular proteins from tissue or cultured cells. MethodsX. 2015;2:e440-e445.
doi:10.1016/j.mex.2015.11.001.
347. Mercer TR, Neph S, Dinger ME, et al. The human mitochondrial transcriptome. Cell.
2011;146(4):645-658. doi:10.1016/j.cell.2011.06.051.
348. Ho L, Titus AS, Banerjee KK, et al. SIRT4 regulates ATP homeostasis and mediates a
retrograde signaling via AMPK. Aging (Albany NY). 2013;5(11):835-849. doi:100616
[pii].
349. Hur JH, Cho J, Walker DW. Aging: Dial M for Mitochondria. Aging (Albany NY).
2010;2(1):69-73.
350. Edgar D, Trifunovic A. The mtDNA mutator mouse: Dissecting mitochondrial
involvement in aging. Aging (Albany NY). 2009;1(12):1028-1032. ISBN:1945-4589
(Electronic)r1945-4589 (Linking).
351. Marcinek DJ, Siegel MP. Targeting redox biology to reverse mitochondrial dysfunction.
Aging (Albany NY). 2013;5(8):588-589.
352. Thomas RR, Kha SM, Smigrodzki RM, et al. RhTFAM treatment stimulates
mitochondrial oxidative metabolism and improves memory in aged mice. Aging (Albany
NY). 2012;4(9):620-635. doi:10.18632/aging.100488.

206
353. Galindo MI, Pueyo JI, Fouix S, Bishop SA, Couso JP. Peptides encoded by short ORFs
control development and define a new eukaryotic gene family. PLoS Biol.
2007;5(5):1052-1062. doi:10.1371/journal.pbio.0050106.
354. Magny EG, Pueyo JI, Pearl FMG, et al. Conserved regulation of cardiac calcium uptake
by peptides encoded in small open reading frames. Science (80- ). 2013;341(6150):1116-
1120. doi:10.1126/science.1238802.
355. Frith MC, Forrest AR, Nourbakhsh E, et al. The abundance of short proteins in the
mammalian proteome. PLoS Genet. 2006;2(4):515-528.
doi:10.1371/journal.pgen.0020052.
356. Hull RL, Westermark GT, Westermark P, Kahn SE. Islet amyloid: A critical entity in the
pathogenesis of type 2 diabetes. J Clin Endocrinol Metab. 2004;89(8):3629-3643.
doi:10.1210/jc.2004-0405.
357. Haataja L, Gurlo T, Huang CJ, Butler PC. Islet amyloid in type 2 diabetes, and the toxic
oligomer hypothesis. Endocr Rev. 2008;29(3):303-316. doi:10.1210/er.2007-0037.
358. Konarkowska B, Aitken JF, Kistler J, Zhang S, Cooper GJS. Thiol reducing compounds
prevent human amylin-evoked cytotoxicity. In: FEBS Journal. Vol 272. ; 2005:4949-
4959. doi:10.1111/j.1742-4658.2005.04903.x.
359. Ma ZA, Zhao Z, Turk J. Mitochondrial dysfunction and beta-cell failure in type 2 diabetes
mellitus. Exp Diabetes Res. 2012:1-11. doi:10.1155/2012/703538.
360. Huang C, Lin C, Haataja L, et al. High Expression Rates of Human Islet Amyloid
Polypeptide Induce Endoplasmic Reticulum Stress–Mediated β-Cell Apoptosis, a
Characteristic of Humans With Type 2 but Not Type 1 Diabetes. Diabetes.
2007;56(8):2016-2027. doi:10.2337/db07-0197.

207
361. Alejandro EU, Gregg B, Blandino-Rosano M, Cras-Meneur C, Bernal-Mizrachi E. Natural
history of beta-cell adaptation and failure in type 2 diabetes. Mol Asp Med.
2014;1(734):19-41. doi:10.1016/j.mam.2014.12.002.
362. Siegel RL, Miller KD, Jemal A. Cancer statistics, 2016. CA Cancer J Clin. 2016;66(1):7-
30. doi:10.3322/caac.21332.
363. Crawford ED, Ventii K, Shore ND. New biomarkers in prostate cancer. Oncology
(Williston Park). 2014;28(2):135-142. http://www.ncbi.nlm.nih.gov/pubmed/24701701.
364. Barry MJ. Clinical practice. Prostate-specific-antigen testing for early diagnosis of
prostate cancer. N Engl J Med. 2001;344(18):1373-1377.
doi:10.1056/NEJM200105033441806.
365. Thompson IM, Ankerst DP, Chi C, et al. Operating Characteristics of Prostate-Specific
Antigen in Men With an Initial PSA Level of 3 . 0 ng / mL or Lower. JAMA.
2005;294(1):66-70. doi:10.1001/jama.294.1.66.
366. Prensner JR, Rubin MA, Wei JT, Chinnaiyan AM. Beyond PSA: The Next Generation of
Prostate Cancer Biomarkers. Sci Transl Med. 2012;4(127):127rv3-127rv3.
doi:10.1126/scitranslmed.3003180.
367. Herrmann PC, Gillespie JW, Charboneau L, et al. Mitochondrial proteome: Altered
cytochrome c oxidase subunit levels in prostate cancer. In: Proteomics. Vol 3. ;
2003:1801-1810. doi:10.1002/pmic.200300461.
368. Raffo AJ, Perlman H, Chen MW, Streitman JS, Buttyan R, Day ML. Overexpression of
bcl-2 Protects Prostate Cancer Cells from Apoptosis in Vitro and Confers Resistance to
Androgen Depletion in Vivo. Cancer Res. 1995;55(19):4438-4445. ISBN:0008-5472
(Print) 0008-5472 (Linking).

208
369. Petros JA, Baumann AK, Ruiz-Pesini E, et al. mtDNA mutations increase tumorigenicity
in prostate cancer. Proc Natl Acad Sci U S A. 2005;102(3):719-724.
doi:10.1073/pnas.0408894102.
370. Wallace DC. Mitochondrial function and cancer. Nat Rev Cancer. 2012;12(10):1.
doi:10.1038/nrc3365.
371. Hashimoto Y, Niikura T, Tajima H, et al. A rescue factor abolishing neuronal cell death
by a wide spectrum of familial Alzheimer’s disease genes and Abeta. Proc Natl Acad Sci
USA. 2001;98:6336-6341.
372. Kim S-J, Guerrero N, Wassef G, et al. The mitochondrial-derived peptide humanin
activates the ERK1/2, AKT, and STAT3 signaling pathways and has age-dependent
signaling differences in the hippocampus. Oncotarget. 2016;7(30):46899.
373. Park SY, Choi B, Cheon H, et al. Cellular aging of mitochondrial DNA-depleted cells.
Biochem Biophys Res Commun. 2004;325(4):1399-1405. doi:10.1016/j.bbrc.2004.10.182.
374. Shin JH, Kim H wook, Rhyu IJ, Song KJ, Kee SH. Axin expression reduces
staurosporine-induced mitochondria-mediated cell death in HeLa cells. Exp Cell Res.
2012;318(16):2022-2033. doi:10.1016/j.yexcr.2012.06.014.
375. Vernochet C, Kahn CR. Mitochondria, obesity and aging. Aging (Albany NY).
2012;4(12):859-860. doi:100518 [pii].
376. Arakawa T, Niikura T, Kita Y. The biological activity of Humanin analogs correlates with
structure stabilities in solution. Int J Biol Macromol. 2011;49(1):93-97.
doi:10.1016/j.ijbiomac.2011.04.003.
377. Durieux J, Wolff S, Dillin A. The cell-non-autonomous nature of electron transport chain-
mediated longevity. Cell. 2011;144(1):79-91. doi:10.1016/j.cell.2010.12.016.

209
378. Woo DK, Shadel GS. Mitochondrial stress signals revise an old aging theory. Cell.
2011;144(1):11-12. doi:10.1016/j.cell.2010.12.023.
379. Friedlander RM. Apoptosis and Caspases in Neurodegenerative Diseases. N Engl J Med.
2003;348(14):1365-1375. doi:10.1056/NEJMra022366.
380. Powers RW, Kaeberlein M, Caldwell SD, Kennedy BK, Fields S. Extension of
chronological life span in yeast by decreased TOR pathway signaling. Genes Dev.
2006;20(2):174-184. doi:10.1101/gad.1381406.
381. Regmi SG, Rolland SG, Conradt B. Age-dependent changes in mitochondrial morphology
and volume are not predictors of lifespan. Aging (Albany NY). 2014;6(2):118-130.
doi:10.18632/aging.100639.
382. Arosio P, Michaels TCT, Linse S, et al. Kinetic analysis reveals the diversity of
microscopic mechanisms through which molecular chaperones suppress amyloid
formation. Nat Commun. 2016;7. doi:10.1038/ncomms10948.
383. Hatters DM, Lindner RA, Carver JA, Howlett GJ. The molecular chaperone, alpha-
crystallin, inhibits amyloid formation by apolipoprotein C-II. J Biol Chem.
2001;276(36):33755-33761. doi:10.1074/jbc.M105285200.
384. Darrow MC, Sergeeva OA, Isas JM, et al. Structural mechanisms of mutant huntingtin
aggregation suppression by the synthetic chaperonin-like CCT5 complex explained by
cryoelectron tomography. J Biol Chem. 2015;290(28):17451-17461.
doi:10.1074/jbc.M115.655373.
385. Romeo M, Stravalaci M, Beeg M, et al. Humanin Specifically Interacts with Amyloid-β
Oligomers and Counteracts Their in vivo Toxicity. J Alzheimer’s Dis. 2017;57(3):857-
871. doi:10.3233/JAD-160951.

210
386. Pistolesi S, Rossini L, Ferro E, Basosi R, Trabalzini L, Pogni R. Humanin structural
versatility and interaction with model cerebral cortex membranes. Biochemistry.
2009;48(22):5026-5033. doi:10.1021/bi900187s.
387. Hashimoto Y, Terashita K, Niikura T, et al. Humanin antagonists: mutants that interfere
with dimerization inhibit neuroprotection by Humanin. Eur J Neurosci. 2004;19(9):2356-
2364. doi:10.1111/j.0953-816X.2004.03298.x.
388. Warburn O, Dickens F. The metabolism of tumors. Am J Med Sci. 1931;182(1):123.
doi:10.1097/00000441-193107000-00022.
389. Vyas S, Zaganjor E, Haigis MC. Mitochondria and Cancer. Cell. 2016;166(3):555-566.
doi:10.1016/j.cell.2016.07.002.
390. Thompson CB. Metabolic enzymes as oncogenes or tumor suppressors. N Engl J Med.
2009;360:813-815. doi:10.1056/NEJMe0810213.
391. Shiraishi T, Verdone JE, Huang J, et al. Glycolysis is the primary bioenergetic pathway
for cell motility and cytoskeletal remodeling in human prostate and breast cancer cells.
Oncotarget. 2015;6(1):130-143. doi:10.18632/oncotarget.2766.
392. Costello LC, Franklin RB, Feng P. Mitochondrial function, zinc, and intermediary
metabolism relationships in normal prostate and prostate cancer. Mitochondrion.
2005;5(3):143-153. doi:10.1016/j.mito.2005.02.001.
393. Risbridger GP, Taylor RA. Minireview: Regulation of prostatic stem cells by stromal
niche in health and disease. Endocrinology. 2008;149(9):4303-4306. doi:10.1210/en.2008-
0465.
394. Tuxhorn JA, Ayala GE, Smith MJ, Smith VC, Dang TD, Rowley DR. Reactive stroma in
human prostate cancer: Induction of myofibroblast phenotype and extracellular matrix

211
remodeling. Clin Cancer Res. 2002;8(9):2912-2923. ISBN:1078-0432 (Print)n1078-0432.
395. Pavlides S, Whitaker-Menezes D, Castello-Cros R, et al. The reverse Warburg effect:
Aerobic glycolysis in cancer associated fibroblasts and the tumor stroma. Cell Cycle.
2009;8(23):3984-4001. doi:10.4161/cc.8.23.10238.
396. Rybak AP, Bristow RG, Kapoor A. Prostate cancer stem cells: deciphering the origins and
pathways involved in prostate tumorigenesis and aggression. Oncotarget. 2015;6(4):1900-
1919. doi:10.18632/oncotarget.2953.
397. Li X, Zhong Y, Lu J, et al. MtDNA depleted PC3 cells exhibit Warburg effect and cancer
stem cell features. Oncotarget. 2016;7(26). doi:10.18632/oncotarget.9610.
398. van der Burgh R, Boes M. Mitochondria in autoinflammation: Cause, mediator or
bystander? Trends Endocrinol Metab. 2015;26(5):263-271.
doi:10.1016/j.tem.2015.03.004.
399. Hsing AW, Gao Y-T, Chua S, Deng J, Stanczyk FZ. Insulin Resistance and Prostate
Cancer Risk. JNCI J Natl Cancer Inst. 2003;95(1):67-71. doi:10.1093/jnci/95.1.67.
400. Johansson M, McKay JD, Wiklund F, et al. Implications for prostate cancer of insulin-like
growth factor-I (IGF-I) genetic variation and circulating IGF-I levels. J Clin Endocrinol
Metab. 2007;92(12):4820-4826. doi:10.1210/jc.2007-0887.
401. Quagliariello V, Rossetti S, Cavaliere C, et al. Metabolic syndrome, endocrine disruptors
and prostate cancer associations: biochemical and pathophysiological evidences.
Oncotarget. 2015. doi:10.18632/oncotarget.16725.
402. Chaudhary AK, O’Malley J, Kumar S, et al. Mitochondrial dysfunction and prostate
cancer racial disparities among American men. Front Biosci (Schol Ed). 2017;9:154-164.
http://www.ncbi.nlm.nih.gov/pubmed/27814581.

212
403. Pompella A, Corti A, Paolicchi A, Giommarelli C, Zunino F. γ-Glutamyltransferase,
redox regulation and cancer drug resistance. Curr Opin Pharmacol. 2007;7(4):360-366.
doi:10.1016/j.coph.2007.04.004.
404. Griffith OW, Bridges RJ, Meister A. Transport of gamma-glutamyl amino acids: role of
glutathione and gamma-glutamyl transpeptidase. Proc Natl Acad Sci. 1979;76(12):6319-
6322. doi:10.1073/pnas.76.12.6319.
405. Harris C, Hansen JM. Oxidative stress, thiols, and redox profiles. Methods Mol Biol.
2012;889:325-346. doi:10.1007/978-1-61779-867-2_21.
406. Soga T, Baran R, Suematsu M, et al. Differential metabolomics reveals ophthalmic acid as
an oxidative stress biomarker indicating hepatic glutathione consumption. J Biol Chem.
2006;281(24):16768-16776. doi:10.1074/jbc.M601876200.
407. Fong MY, McDunn J, Kakar SS. Identification of metabolites in the normal ovary and
their transformation in primary and metastatic ovarian cancer. PLoS One. 2011;6(5).
doi:10.1371/journal.pone.0019963.
408. Tan B, Qiu Y, Zou X, et al. Metabonomics identifies serum metabolite markers of
colorectal cancer. J Proteome Res. 2013;12(6):3000-3009. doi:10.1021/pr400337b.
409. Qiu Y, Cai G, Zhou B, et al. A distinct metabolic signature of human colorectal cancer
with prognostic potential. Clin Cancer Res. 2014;20(8):2136-2146. doi:10.1158/1078-
0432.CCR-13-1939.
410. Gault CR, Obeid LM, Hannun YA. An overview of sphingolipid metabolism: From
synthesis to breakdown. Adv Exp Med Biol. 2010;688:1-23. doi:10.1007/978-1-4419-
6741-1_1.
411. Li Z, Zhang H, Liu J, et al. Reducing Plasma Membrane Sphingomyelin Increases Insulin

213
Sensitivity. Mol Cell Biol. 2011;31(20):4205-4218. doi:10.1128/MCB.05893-11.
412. Kang SC, Kim BR, Lee SY, Park TS. Sphingolipid metabolism and obesity-induced
inflammation. Front Endocrinol (Lausanne). 2013;4(JUN).
doi:10.3389/fendo.2013.00067.
413. Ponnusamy S, Meyers-Needham M, Senkal CE, et al. Sphingolipids and cancer: ceramide
and sphingosine-1-phosphate in the regulation of cell death and drug resistance. Futur
Oncol. 2010;6(10):1603-1624. doi:10.2217/fon.10.116.
414. Samad F, Hester KD, Yang G, Hannun YA, Bielawski J. Altered adipose and plasma
sphingolipid metabolism in obesity: A potential mechanism for cardiovascular and
metabolic risk. Diabetes. 2006;55(9):2579-2587. doi:10.2337/db06-0330.
415. Shah C, Yang G, Lee I, Bielawski J, Hannun YA, Samad F. Protection from high fat diet-
induced increase in ceramide in mice lacking plasminogen activator inhibitor 1. J Biol
Chem. 2008;283(20):13538-13548. doi:10.1074/jbc.M709950200.
416. Russo SB, Ross JS, Cowart LA. Sphingolipids in Obesity, Type 2 Diabetes, and Metabolic
Disease. Handb Exp Pharmacol. 2013;216:373-401. doi:10.1007/978-3-7091-1511-4.
417. Choi S, Snider AJ. Sphingolipids in High Fat Diet and Obesity-Related Diseases.
Mediators Inflamm. 2015;2015. doi:10.1155/2015/520618.
418. Kowalski GM, Carey AL, Selathurai A, Kingwell BA, Bruce CR. Plasma Sphingosine-1-
Phosphate Is Elevated in Obesity. PLoS One. 2013;8(9).
doi:10.1371/journal.pone.0072449.
419. Erickson KA, Smith ME, Anthonymuthu TS, et al. AICAR inhibits ceramide biosynthesis
in skeletal muscle. Diabetol Metab Syndr. 2012;4(1). doi:10.1186/1758-5996-4-45.
420. Seco-Cervera M, Spis M, García-Giménez JL, et al. Oxidative stress and antioxidant

214
response in fibroblasts from Werner and Atypical Werner Syndromes. Aging (Albany NY).
2014;6(3):231-245. doi:10.18632/aging.100649.
421. Gesta S, Bezy O, Mori MA, Macotela Y, Lee KY, Kahn CR. Mesodermal developmental
gene Tbx15 impairs adipocyte differentiation and mitochondrial respiration. Proc Natl
Acad Sci. 2011;108(7):2771-2776. doi:10.1073/pnas.1019704108.
422. Rogers GW, Brand MD, Petrosyan S, et al. High throughput microplate respiratory
measurements using minimal quantities of isolated mitochondria. PLoS One. 2011;6(7).
doi:10.1371/journal.pone.0021746.
423. Bedrood S, Li Y, Isas JM, et al. Fibril structure of human islet amyloid polypeptide. J Biol
Chem. 2012;287(8):5235-5241. doi:10.1074/jbc.M111.327817.
424. Jayasinghe SA, Langen R. Lipid membranes modulate the structure of islet amyloid
polypeptide. Biochemistry. 2005;44(36):12113-12119. doi:10.1021/bi050840w.
425. Okada AK, Teranishi K, Isas JM, Bedrood S, Chow RH, Langen R. Diabetic Risk Factors
Promote Islet Amyloid Polypeptide Misfolding by a Common, Membrane-mediated
Mechanism. Sci Rep. 2016;6. doi:10.1038/srep31094.
426. Margittai M, Langen R. Spin Labeling Analysis of Amyloids and Other Protein
Aggregates. Methods Enzymol. 2006;413:122-139. doi:10.1016/S0076-6879(06)13007-4.
427. Langen R, Isas JM, Luecke H, Haigler HT, Hubbell WL. Membrane-mediated assembly
of annexins studied by site-directed spin labeling. J Biol Chem. 1998;273(35):22453-
22457. doi:10.1074/jbc.273.35.22453.

Abstract (if available)

Abstract The discovery of humanin nearly two decades ago has ushered in a new interest in mitochondrial biology and has combined several nascent fields of research. Over the past decade, with advancement of the current proteomic technologies and progress in the field of small open-reading-frames (ORFs), the concept of mitochondrial small ORFs has been established. This innovative concept opened up the possibility that the mitochondrial genome encodes for more than 13 proteins and contains molecular instructions for biological functions besides energy production. Humanin is the first of several mitochondrial-derived-peptides (MDPs) that are originated within small, alternative ORFs of the mitochondrial genome. Humanin has a number of different cytoprotective and metabolo-protective effects while the recently discovered small humanin-like peptides (SHLP) 1-6 have similar and distinct properties compared to humanin. The discovery of another MDP called MOTS-c as an exercise mimetic and activator of AMP-activated protein kinase (AMPK) suggests that these peptides will have an important role in metabolism and could be used as future therapeutics. Humanin and other MDPs are small circulating peptides that have important signaling functions and participate in the mitochondrial retrograde signaling events. While nearly 200 papers were published on humanin and other MDPs since their discovery, most of them have focused on the protective effects of this molecule in vitro and in vivo, showing its ability to ameliorate damage induced by multiple stressors and disease conditions. Yet little is known about their fundamental mechanisms: for example, the enigma of humanin expression in the mitochondria is still largely unsolved. Similarly, the actual mechanism of action of humanin and the target organs involved remain unclear. To fill this gap, I utilized rodent models and tissue culture to examine the systemic mechanism involved, particularly for the caloric-restriction (CR)-mimetic effects. Chapter 2 summarized the how humanin modulated three major pathways closely associated with CR: the insulin-like growth factor-I (IGF-I) pathway, the AMPK pathway and the oxidative stress to prolong healthspan and confer differential protection, which are established benefits of dietary interventions. My next main project involves the characterization of a family of humanin-like peptides called SHLPs. They exhibit potent signaling effects, such as regulation of mitochondrial metabolism and glucose homeostasis. My research interest mainly revolved around SHLP2, since it resembles humanin in terms of signaling and biological functions. I demonstrated the presence of polyadenylated mRNA transcripts of SHLP2 from the 16S rRNA region of mitochondrial genome. Furthermore, I found it regulated mitochondrial oxidative stress and improved multiple metabolic biomarkers such as circulating sphingolipids and key intermediates in the gamma-glutamyl cycle, which provides mechanistic explanation of my finding that low circulating levels of SHLP2 can serve as a novel biomarker for prostate cancer risk. Similarities between humanin and SHLP2 were characterized: they both interact with islet amyloid polypeptide (IAPP) structurally and interfere IAPP aggregation

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Role of mitochondrially derived peptides in the inhibition of IAPP misfolding

PDF

The mitochondrial energy – redox axis in aging and caloric restriction: role of nicotinamide nucleotide transhydrogenase

PDF

A small molecule protease inhibitor induces senescence phenotypes that are reversible upon drug removal

PDF

The intersection of mitochondrial biology and cancer: insights from mitochondrial microproteins and mtDNA alterations

PDF

Signaling mechanisms governing intestinal regeneration and gut-glia cross-talk in Drosophila

PDF

Mitonuclear communication in metabolic homeostasis during aging and exercise

PDF

Extracellular matrix regulation of mitochondrial function in engineered cardiac myocytes

PDF

Discovery of mitochondria-DNA-encoded microproteins in neurodegeneration

PDF

To adapt or not to adapt: the age-specific and sex-dependent differences in the adaptive stress response

PDF

Mitochondrial dynamics regulate Leydig cell health and integrity

PDF

The human Lon protease in mitochondrial stress protection and aging

PDF

The overlap between mTOR signaling, rapamycin and cellular senescence

PDF

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

PDF

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

PDF

Understanding prostate cancer genetic susceptibility and chromatin regulation

PDF

Development of dihydromyricetin (DHM) as a novel therapy for alcoholic liver disease (ALD) and alcohol use disorder (AUD)

PDF

Statistical algorithms for examining gene and environmental influences on human aging

PDF

Cisplatin activates mitochondrial oxphos leading to acute treatment resistance in bladder cancer

PDF

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

PDF

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

Asset Metadata

Creator Xiao, Jialin (author)

Core Title The regulation, roles, and mechanism of action of mitochondrial-derived-peptides (MDPs) in aging

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Biology of Aging

Publication Date 10/11/2018

Defense Date 05/14/2018

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

Tag aging,metabolic syndromes,mitochondria,OAI-PMH Harvest,prostate cancer

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Cohen, Pinchas (committee chair), Davies, Kelvin (committee member), Schilling, Birgit (committee member)

Creator Email jialinxi@usc.edu,joyce901211@hotmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-88103

Unique identifier UC11676638

Identifier etd-XiaoJialin-6812.pdf (filename),usctheses-c89-88103 (legacy record id)

Legacy Identifier etd-XiaoJialin-6812.pdf

Dmrecord 88103

Document Type Dissertation

Format application/pdf (imt)

Rights Xiao, Jialin

Type texts

Source 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 a...

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