University of Southern California Dissertations and Theses

Air pollution neurotoxicity throughout the lifespan: studies on the mechanism of toxicity and interactions with effects of sex and genetic background

(USC Thesis Other)

Air pollution neurotoxicity throughout the lifespan: studies on the mechanism of toxicity and interactions with effects of sex and genetic background

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Air pollution neurotoxicity throughout the lifespan: studies on the
mechanism of toxicity and interactions with effects of sex and genetic
background

By

Amin Haghani

A Dissertation Presented to the
FACULTY OF THE
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial fulfillment of the
Requirement for the Degree
DOCTOR OF PHILOSOPHY
(BIOLOGY OF AGING)

May 2020

Copyright 2020 Amin Haghani

ii
Table of Contents
Dedication ....................................................................................................................... iv
Acknowledgment .............................................................................................................. v
Abstract .......................................................................................................................... viii
Chapter 1. Background ................................................................................................. 1
Chapter 2. Cell-based assays that predict in vivo neurotoxicity of urban ambient
nano-sized particulate matter ........................................................................................... 9
Abstract ................................................................................................................ 11
1. Introduction ................................................................................................... 12
2. Methods and Materials ................................................................................. 14
3. Results .......................................................................................................... 20
4. Discussion .................................................................................................... 30
Supplementary data ............................................................................................. 35
Chapter 3. Toxicity of urban air pollution particulate matter in developing and adult
mouse brain: comparison of total and filter-eluted nanoparticles ................................... 36
Abstract ................................................................................................................ 38
1. Introduction ................................................................................................... 39
2. Methods ........................................................................................................ 41
3. Results .......................................................................................................... 48
4. Discussion .................................................................................................... 58
Supplementary data ............................................................................................. 64
Chapter 4. Air pollution alters Caenorhabditis elegans development and lifespan:
responses to traffic-related nanoparticulate matter (nPM) ............................................. 70
Abstract ................................................................................................................ 71
1. Introduction ................................................................................................... 72
2. Method .......................................................................................................... 74
3. Results .......................................................................................................... 79
4. Discussion .................................................................................................... 89
Supplementary data ............................................................................................. 96
Chapter 5. Sex-specific Cerebral Cortex and Blood Transcriptome Changes in Mouse
Neonates Prenatally Exposed to Air Pollution .............................................................. 106
Abstract .............................................................................................................. 107
1. Introduction ................................................................................................. 108
2. Methods ...................................................................................................... 110
3. Results ........................................................................................................ 114
4. Discussion .................................................................................................. 122

iii
Chapter 6. Hippocampal transcriptome changes associated with long-term
neurological changes of gestational air pollution toxicity in mouse .............................. 127
Abstract .............................................................................................................. 128
1. Introduction ................................................................................................. 130
2. Methods ...................................................................................................... 132
3. Results ........................................................................................................ 137
4. Discussion .................................................................................................. 146
Chapter 7. Mouse brain transcriptome responses to inhaled nanoparticulate matter
differed by sex and ApoE in Nrf2 - NF-kB interactions ................................................. 152
Abstract .............................................................................................................. 154
1. Introduction ................................................................................................. 155
2. Methods ...................................................................................................... 157
3. Results ........................................................................................................ 161
4. Discussion .................................................................................................. 174
Supplementary data ........................................................................................... 180
Chapter 8. Female vulnerability to the effects of smoking on health outcomes in older
people ................................................................................................................ 185
Abstract .............................................................................................................. 186
1. Introduction ................................................................................................. 187
2. Methods ...................................................................................................... 188
3. Results ........................................................................................................ 191
4. Discussion .................................................................................................. 198
Chapter 9. Summary and Synthesis ......................................................................... 203
Bibliography .................................................................................................................. 206

iv
Dedication

To my family, most especially, my wife,
for their unwavering support and unbridled enthusiasm in all aspects of life.

v
Acknowledgment
I would like to extend my utmost gratitude to Dr. Caleb Finch, who guided me
throughout my Ph.D. He has an exemplary enthusiasm and depth of understanding for
science. He pushed me to develop strong critical thinking and overcome my weaknesses
to become a competent independent scientist. Dr. Finch also inspired me to build a strong
and multidisciplinary scientific network. I learned that a strong science can only be made
by communication with others. I am grateful to Dr. Todd Morgan, who guided me patiently
on a daily basis to overcome the facing experimental challenges. Our routine scientific
discussions helped me to develop several novel research ideas.
I was honored to work with a large network of scientists during my Ph.D. I like to thank
Dr. Sean Curran, who supported me with my research idea on using C. elegans in air
pollution research. He has been approachable, enthusiastic, and inspiring for doing
molecular mechanism research to resolve scientific questions. I am grateful to Dr. Eileen
Crimmins, who trusted me to do epidemiological analysis in my research. She is an
exceptionally supportive and patient mentor. Her lecture was an inspiration for me to take
a multidisciplinary approach in aging research by combining human population analysis
with experimental research. Dr. Henry Forman was a great inspiration throughout my
research. He is an outstanding biochemist who thought me the importance of mechanistic
experiments and signaling transduction pathways. His former trainee, Dr. Hongqiao
Zhang, is my friend and daily mentor to design and perform several mechanistic
experiments. I am really grateful for their training and inspiration.
I like to acknowledge some other key individuals who helped me during my research.
Dr. Thalida Em Arpawong guided me in human data analysis and thought me to think like

vi
an epidemiologist. Dr. Juan Pablo Lewinger thought me the basics of statistical learning
in health sciences and inspired me with his deep understanding of mathematics. He
guided me in selecting proper courses that could advance my career. I am grateful to Dr.
Morgan Levine, who was my role model as a student. She could succeed in an
independent research career in a really short period. I like to extend my gratitude to Dr.
Steve Horvath, from UCLA. He supported my research idea and agreed to collaborate in
one of my projects. He patiently thought me some of the analytical approaches that I can
take for high dimensional data analysis. His networking and project management skills
was also an inspiration for my research career. I was also privileged to collaborate and
learn from several known scientists. Drs. Constantinos Sioutas, Heather Volk, and Daniel
Campbell guided and supported me in some of my projects.
I am grateful to my lab mates whose efforts and expertise allowed for great research.
I extend my heartfelt thanks to Dr. Hans Dalton, who worked with me throughout the C.
elegans project. I thank my other lab mates Richard Johnson, Nicholas Woodward,
Amirhosein Mousavi, Mafalda Cacciotolo, Carla D’Agostino, Max Thorwald, Hang Cheng,
and those who have worked with me over the years; allowing me to teach and learn from
them.
I like to extend my deepest and heartfelt thanks to my wife, Dr. Nikoo Safi, who was
my support, lab mate, and closest friend over the years. Her encouragement gave me the
energy and motivation to move forward and reach for the top.
Finally, I am grateful to USC and Davis School of Gerontology. This institute has
every opportunity to learn and is a breeding ground for success. The gerontology

vii
department is a multidisciplinary community and like a family to me. I am excited to keep
in touch with all these great people and collaborate on numerous future projects.

viii
Abstract
Air pollution (AirP) is a leading global environmental risk factor of mortality and
morbidity in humans and many other organisms. AirP affects the lives of more than 90%
of the humans from all ages and unfortunately has an increasing trend in developing
countries. Depends on the life stage, exposure to AirP can increase the risk of
neurodevelopmental or neurodegenerative diseases. Alarmingly, lack of understanding
of the AirP toxicity mechanisms likely led to a severe underestimation of the global health
burden of these complex toxicants. AirP toxicity is shaped by a complex interface of
chemical properties of toxicants mixture and the biological features of the exposed
individuals (e.g. genetic structure, sex, and life-stage). The current dissertation is
comprised of a series of projects to resolve some of these complex interactions. In
chapters 2-3, we discussed how the chemical characterization of AirP can alter the
neurotoxicity in vivo and in vitro. Moreover, we developed a cell-based assay that can
predict the neurotoxicity of the AirP particulate matter (PM) in the brain. Chapters 4-7
approaches the nuances in the molecular toxicity of AirP between different life stages
(development vs adult), sexes (males vs females), and some genetic variants (ApoE3 vs
ApoE4). We also corroborated our findings in Caenorhabditis elegans, a known nematode
model for aging research. In the last chapter, we described the gender differences in
vulnerability to life-time cigarette smoke exposure for some aging-associated health
outcomes. Throughout the dissertation, we discussed the facing challenges in the AirP
toxicology field, and also identified potential molecular mechanisms that underlay each
complex research question. We hope that our findings lead to several novel studies and
translate into the regulation soon.

1
Chapter 1. Background
Air pollution (AirP) and mortality: It is estimated that ambient PM2.5 contributes to
3-4 million deaths annually (around 7.6% of total global death) (1). The main causes of
AirP-mediated death include cerebrovascular disease and ischemic heart disease,
followed by chronic obstructive pulmonary disease, acute lower respiratory illness and
lung cancer (2). This number is apart from 3.5 million deaths attributed to indoor AirP
such as solid fuels for cooking and heating (2). A recent study in 1 million births across
sub-Saharan Africa using satellite-based estimates suggested 10 µg/m
3
PM2.5
concentration is associated with a 9% increase of infant mortality, which was responsible
for 22% of infant deaths (449,000 deaths in 2015) in these countries (3). This risk is almost
double of the Global Burden of Disease on ambient AirP estimates (4), and also higher
than 7.3% increase of mortality observed with 10 µg/m
3
PM2.5 in USA Medicare
beneficiaries’ study (5). Terrifyingly, as opposed to the studies in developing countries,
this study showed that long-term AirP effects were independent of the wealth or
socioeconomic status of the household in African countries (5). All these studies affirm
PM2.5 is still harmful below the current US national ambient air quality standard (NAAQS,
12 µg/m
3
) (3, 5). It seems like the only way to diminish these harmful effects is by
implementing enforced global action to improve the overall air quality. The current air
quality alert program had limited public health benefits with no success in reducing the
burden of this toxicant on mortality and almost all of the chronic diseases (6). In a
simulation model, a modest reduction in AirP (only 5 µg/m
3
across sub-Saharan Africa)

2
could project into vast health benefits comparable with many key child health
interventions such as vaccines or nutritional supplements in Africa (3).
Complexity of AirP Toxicity: In addition to mortality, AirP has many harmful effects on
morbidity including respiratory diseases (e.g. asthma, chronic obstructive pulmonary
disease, lung cancer), cardiovascular disorders (e.g. myocardial infarction, heart failure)
and neurodegeneration (cerebrovascular issues, brain tumor, cognitive decline, and
Alzheimer’s disease) (7-11). While AirP is a risk factor for many aging-associated chronic
diseases, the biology behind AirP toxicity remains ambiguous. AirP is a complex and
ubiquitous mixture of pollutants derived from a variety of sources and is currently a global
issue. It is suggested that carbonaceous particles such as fossil fuel combustion and
traffic-related sources are more toxic compared to crustal, nitrate or sulfate materials (12).
Thus, most research focuses on traffic-related AirP (TRAP) and its hazardous health
effects.
TRAP is a mixture of particulate matter (PM), gases (e.g. ozone, carbon monoxide,
sulfur oxide, nitrogen oxides), organic compounds (e.g. polycyclic aromatic hydrocarbons
and endotoxins) and elements (e.g. iron, lead, manganese, nickel). PM is categorized by
size, which includes coarse particles with aerodynamic diameter (2.5 to 10 µm; PM10),
fine particles (<2.5 µm; PM2.5), and ultrafine particles (usually <0.2 µm, PM0.2). While
PM0.2 is widely implicated in PM-associated pathology, particularly in the nervous system
compared to larger PM sizes, it is still unregulated and unmonitored by U.S.
Environmental Protection Agency regulation (13). Since the majority of the
epidemiological studies are based on PM10 and PM2.5 toxicity, it is highly probable that
epidemiologists are underestimating the AirP global burden on health and mortality.

3
PM is also composed of a mixture of toxic components. Unfortunately, several studies
interpret the magnitude of PM toxicity based on individual toxicants, rather than a mixture.
This viewpoint has misled the field to ignore some important PM characteristics in their
thinking workflow. In PM neurotoxicology research, the findings are biased toward
reporting the positive findings and overlooking the complexity of PM characteristics on
their biological effects. Thus, the findings are often contradictory and irreproducible. In
my deduction, understanding the effects of PM chemistry as a mixture on the toxicity is
the first step to resolve the mechanisms that underlie different biological responses.
AirP and development: Several meta-analyses showed that gestational exposure to
AirP would lead to premature birth (14), low birth weight (15), risk of autism spectrum
disorder (ASD) (16-20), and children hyperactivity (21). Other long-term health hazards
of prenatal AirP exposure include childhood cardiovascular problems (22), blood pressure
(23), childhood cancer (24), and cognitive changes (25). AirP has considerable
neurotoxicity that can disrupt neurodevelopment and cause behavioral changes. Several
studies reported that high AirP levels during pregnancy would alter cerebral cortex
morphology (26, 27), decrease intelligence, increase depression, cause attention deficits
and hyperactivity disorders, and also increase the probability of receiving academic
support in later childhood (27-29). Thus, AirP has a strong effect on human development
and adversely alters the risk of several chronic diseases. A large gap in most of the
epidemiological studies is the lack of stratification by infants’ sex in their analysis.
Numerous epidemiological and mouse model studies reported that gestational AirP
exposure would have sex-specific toxicities (30-33). For example, while boys show higher
AirP mediated cognitive decline in attention domains (31, 34), girls show more

4
vulnerability in memory domains (31). ASD is also diagnosed 4 times more often in males
than in females and AirP exposure is a major contributor to ASD risk (16-20). Lack of
proper inclusion of AirP interaction with sex has misguided the field in underestimating
the global burden of these airborne toxicants on human health. This dissertation will
discuss some of the molecular mechanisms involved in AirP developmental toxicity and
also the differences between male and female responses.
AirP and neurodegeneration: Many studies showed that AirP is associated with
cognitive decline and also increased the epidemiological risk for many CNS diseases
such as ischemic stroke, Alzheimer's (AD) and Parkinson's diseases (PD) (35). Residents
in high polluted cities (age: 34-84) had a greater neuronal and astrocytic accumulation of
protein Ab42 (a hallmark of AD) in the frontal cortex, hippocampus and olfactory bulbs
(10). Early childhood exposure to high AirP concentration was also associated with
excessive accumulation of Ab42 and a-synuclein (a hallmark of PD) and white matter
hyperintensities in the brain (36, 37). Importantly, these pathologies are not exclusive to
humans. Feral dogs living in Mexico City, which represent a highly polluted area, showed
signs of oxidative damage, premature diffuse amyloid plaques, and DNA damage in the
olfactory bulb, frontal cortex, and hippocampus (37). The dogs in polluted areas had a
gradual decrease of metals accumulated (nickel and vanadium) from olfactory mucosa to
the olfactory bulb and frontal cortex, which suggested an olfactory exposure route might
be accessible to PM0.2 (37). In recent years, our lab examined the neurotoxic effects of
nPM in rodent and cell models to further understand the underlying mechanism. Similar
to humans, mice exposed to nPM, particularly APOE4+/+ animals, induced Ab
production. In addition, nPM reduced AMPA1 receptors (GluR1) in the hippocampus and

5
specifically decreased the neurite density in the CA1 region (13, 38). CA1, along with the
subiculum is the first hippocampal region affected in AD (39). Some of these effects were
age-dependent. Old control animals (18 months) had a lower baseline of neurite density
and GluR1 levels with no additional nPM effect (40). In contrast to observed changes, old
mice (18 months) were more susceptible to nPM mediated behavioral alteration (-36%)
particularly in a total number of explorations during initial novel object recognition (NOR)
test compared to young exposed mice (3 months) (40). We and others have thoroughly
characterized this as a useful model for AirP neurotoxicity.
AirP mediated neuroinflammation: AirP is one of the most prevalent systemic and
localized pro-inflammatory stimuli. Many studies reported that particulate matter is
associated with increased systemic inflammatory markers such as IL6, TNFa and C-
reactive protein (CRP) (41-44). In the human brain, AirP neurotoxicity is associated with
microglial activation and inflammatory markers such as IL1b and COX2 (37). Our rodent
and cell models also showed nPM can induce robust inflammatory responses. Some of
the affected markers included CD14, CD68, GFAP, IL1a, and TNFa (13). The microglial
activation marker IBA1 also increased following chronic exposure (10 weeks) to nPM,
particularly in hippocampal CA1 and DG regions of young mice (3 months old) (40). In
contrast, old mice (18 months old) had a higher baseline level of IBA1 with no nPM effect.
Transcriptome analysis of rat primary mixed glial culture identified the TLR pathway as
one of the main affected biological processes during nPM toxicity (45). The silencing of
TLR4 in these cells led to decreased nPM-mediated inflammation. Comparison of
prenatal exposure of TLR-/- and TLR+/- animals to diesel exhaust particles diverged in
the inflammatory response, cortical volume, number of round microglia, hippocampal

6
volume and number of stout microglia (mainly in male) (46). This dissertation examined
the transcriptome responses of PM in both developing and adult brains. We identified
some the upstream regulators of AirP mediated neuroinflammation and also discussed
the differences between sexes and ApoE genotype. Moreover, our studies added new
evidence that linked chronic AirP mediated neuroimmune changes with behavioral and
systemic metabolism following prenatal exposure to PM.
AirP mediated Nrf2 and NF-kB responses: We further focused on genomic pathways
mediated by Nrf2 and NF-kB, which responded to AirP in our prior studies (45, 47). These
transcription factors control hundreds of genes that mediate cellular responses to
oxidative stress and immunity. They respond to oxidative stress, tobacco smoke,
traumatic brain injury, and ischemic stroke and are altered by aging and AD (48). NRF2
downstream genes include anti-oxidants (e.g. glutathione, thioredoxin), anti-inflammatory
cytokines (e.g. IL-10), phase I & II xenobiotic detoxifying enzymes (e.g. CYP450) and free
radical scavengers (49). NF-kB family of transcription factors regulate the expression of
immune-related proteins including cytokines (e.g. TNFa, IL1a/b), antigen-presenting
proteins (e.g. MHCI, b2M), chemokines (e.g. MCP-1, MIP-1), adhesion molecules (e.g.
ICAM-1, VCAM-1), inducible nitric oxide synthase (iNOS), and proapoptotic (e.g. BIM,
BAX) or antiapoptotic proteins (e.g. XIAP, BCL-2) (50). The complex interplay of NRF2
and NF-kB signaling pathways can determine cell fate by coordinating the balance of anti-
oxidative or inflammatory responses. The mechanism of this interaction is dependent on
the type of stress, target cell or tissue (48). In addition to the rodent models, we
corroborated the AirP mediated Nrf2 and inflammatory responses in Caenorhabditis

7
elegans. This novel study expanded our understanding of AirP mediated Nrf2 responses
and the relationship to the development and lifespan of the organism.
Biomarkers of AirP toxicity: Currently, there is no available biomarker of AirP
neurotoxicity. Unfortunately, lack of a standard method to assess AirP exposure in human
studies halted the epidemiologists to find a reliable biomarker that can predict long term
health hazards of these toxicants. Depends on the availability, scientists use personal
monitoring of AirP components, Land use regression, distance to highway, and recently
satellite-based images to estimate AirP concentration. For example, a research group
used a land-use regression model to estimate individual AirP exposure and identified 186
interacting genetic markers that can increase the risk of childhood asthma in polluted
areas (51). Further evaluation of these genetic variants in the Children Health Study
cohort only confirmed interaction in 8 of these single-nucleotide polymorphisms. This
cohort uses a completely different method to assess AirP exposure by central site
monitors in the study communities. Such diversity in the methods warrants a controlled
experiment to identify the biological changes that can be applied in humans as a
biomarker of AirP neurotoxicity. In chapter 4, we propose some novel blood genes that
are associated with prenatal AirP neurotoxicity. Changes of these genes in human core
blood or placenta might have biomarker applications for epidemiological studies.
AirP and epidemiological research: Estimating accumulating AirP exposure dosage
throughout the lifespan is a challenge that is remained to be resolved. Currently, scientists
rely on PM mass and some individual chemical concentrations during a limited number
of years to assess the AirP toxicity. Unfortunately, several PM characteristics such as
chemical composition, surface chemistry, surface reactivity, PM morphology, and acidity

8
are also missing in large scale data. Thus, all the reported hazards of AirP exposure are
confounded by these missing variables. Another gap is the potential interaction of ambient
AirP with cigarette smoke exposure. Cigarette smoke and AirP has a comparable
chemical composition (52), but epidemiologists rarely include both of these variables in
their analysis.
Tobacco smokes, particularly cigarettes are also among the leading causes of death
and many aging-associated in the United States (53, 54). From 2005 to 2015, around
20.9-15.1% of the adults in the U.S were cigarette smokers (55), particularly daily
smokers. Smoking is a global hazard that ranked as the second cause of mortality and
disability-adjusted life years in the 2015 global burden of diseases (1). In chapter 7, we
will discuss the hazards of cigarette smoke on aging-associated health conditions and
also the gender-specificity of the results. The findings extend our results on gender
differences in AirP responses of rodents.
In sum, this dissertation uses a multidisciplinary approach to address several critical
gaps of knowledge in the AirP toxicology field. We will discuss some of the underlying
mechanisms behind AirP neurotoxicity at different life-stages, sex, and ApoE genotype.
The findings were also corroborated in other model organisms such as Caenorhabditis
elegans. We hope that our finding facilitates several research opportunities that can be
translated into regulation and clinical research.

9
Chapter 2. Cell-based assays that predict in vivo neurotoxicity of
urban ambient nano-sized particulate matter
(Adapted from Free Radical Biology and Medicine, Doi:
10.1016/j.freeradbiomed.2019.09.016)

Hongqiao Zhang
1#
, Amin Haghani
1#
, Amirhosein H Mousavi
2
, Mafalda Cacciottolo
1
,
Carla D’Agostino
1
, Nikoo Safi
3
, Mohammad H Sowlat
1
, Constantinos Sioutas
2
, Todd E
Morgan
1
, Caleb E Finch
1,4
, Henry Jay Forman
1
*

1
Leonard Davis School of Gerontology and, University of Southern California
2
Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089
3
Center for Cancer Prevention and Translational Genomics at the Samuel Oschin
Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, CA.
4
Dept. Neurobiology, Dornsife College

# Equal contribution, co-first authors, *Corresponding author

Corresponding author
Henry Jay Forman, PhD
Email: hforman@usc.edu
Address:
3715 McClintock Avenue GER306A
Los Angeles 90089

10
Highlights
• Ambient PMs were assessed in cell-based assays for NF-kB activation, lipid
peroxidation and NO production
• Batches of ambient PM diverged in potency of cellular responses and neurotoxicity
• In vitro NF-kB reporter assay correlated best with ambient particle-mediated
neurotoxicity in mice

Graphical abstract

11
Abstract
Exposure to urban ambient particulate matter (PM) is associated with risk of
Alzheimer’s disease and accelerated cognitive decline. Assessment of the neurotoxic
effects caused by urban PM is complicated by variations of composition from source,
location, and season. Several in vitro cell-based assays were compared in relation to the
in vivo neurotoxicity of urban ambient PM: NF-kB responses, nitric oxide induction, and
lipid peroxidation. We focused on nPM, a nanosized subfraction of PM2.5 (i.e., ultrafine
PM), extracted as an aqueous suspension, as used in prior studies. In vitro activities were
compared with in vivo responses of mice chronically exposed to the same batch of PM.
The potency of nPM varied widely between nPM batches for NF-kB activation,
analyzed with an NF-kB reporter in human monocytes. Three independently collected
batches of nPM had corresponding differences to responses of mouse cerebral cortex to
chronic nPM inhalation, for induction of pro-inflammatory cytokines, microglial activation
(Iba1), and soluble Ab40 & 42 peptides. The in vitro responses of BV2 microglia for NO-
production and lipid peroxidation also differed by nPM batch, but without correlation to in
vivo responses. In conclusion, these data confirm that batches of nPM can differ widely
in toxicity. The in vitro NF-kB reporter assay offers a simple, high throughput screening
method to predict the in vivo neurotoxic effects of nPM exposure.
Keywords: air pollution, particulate matter, ultrafine particles, NF-kB, inflammation,
neurotoxicity, in vitro screening

12
1. Introduction
Epidemiological studies consistently associate exposure to urban ambient air
pollution particulate matter (PM) with accelerated cognitive aging and increased risk of
Alzheimer disease and other neurodegenerative disorders (38, 56, 57). Correspondingly,
rodent air pollution exposure models show cognitive impairments, brain inflammation,
loss of CA1 synapses, decreased glutamate receptor subunit GluR1 receptor, and
increased Ab oligomers (38, 58-60).
Ambient air pollution is globally monitored by the WHO as PM2.5 (aerodynamic
diameter of particles below 2.5 µm) (61). The epidemiological associations of PM2.5 with
aging and diseases are based mainly on mass concentration µg/m
3
, with little
consideration of chemical composition. However, ambient urban PM is a heterogeneous
mixture of particles with different sizes that vary widely in chemical composition.
Moreover, PM2.5 bioactivities depend on how active chemicals are distributed on particle
surfaces. The composition varies greatly among batches due to the variety of local
sources, as well as daily variations in temperature, sunlight, humidity, among other
environmental factors, (62-66).
Several studies tried to predict PM toxicity in populations by composition and
oxidative activities in cell-free systems (67, 68). The weak correlations of biochemical
assays with epidemiological and rodent responses (52) has stimulated the development
of in vitro assays that may predict in vivo responses; these assays include for
mutagenicity (69), oxidative potential (68), cytotoxicity (70) and cytokine production (70-
74).

13
Inflammation is induced in rodent brains by chronic exposure to air pollution PM (38,
58-60), and oxidative stress is postulated to underlie the neurotoxicity of ambient PM2.5
(35). Therefore, we assessed the activation of NF-kB, a master transcription factor in pro-
inflammatory responses (75), for comparison with lipid peroxidation and nitric oxide (NO)
production in cell-based assays. We tested two forms of PM collected at different times
near an urban highway in Los Angeles: PM2.5 collected directly as an aqueous slurry
(sPM), and nPM, a nano-scaled subfraction of PM2.5 (i.e., ultrafine PM), collected on
filters, and eluted by sonication. We also examined diesel exhaust particles (DEP) widely
used to model PM for air pollution. We showed wide differences on nPM batches for NF-
kB activation, lipid peroxidation, NO production. The in vitro potency of NF-kB activation
by nPM strongly predicted in vivo neurotoxic responses.

14
2. Methods and Materials
Most reagents were from Sigma-Aldrich (St. Louis, MO, USA). Diphenyl-1-
pyrenylphosphine (DPPP), Cayman Chemical Company (Ann Arbor, MI, USA); RPMI
1640 and DMEM cell culture media (Thermal Fisher, Rockford, IL, USA); QUANTI-Blue,
InvivoGen (San Diego, CA, USA); diesel particles (DEP), Environment Protection Agency
(EPA, USA), and Dr. Staci Bilbo, Harvard University. Antibodies for NF-kBp65, LaminA/C
were from Santa Cruz Biotechnologies (Sata Cruz, CA, USA).
2.1. Collection of nPM and sPM
Ambient nanoscale particulate matter (nPM; particles with aerodynamic diameters
less than 0.25 µm) were collected on Zeflour PTFE filters (Pall Life Sciences, Ann Arbor,
MI) by a High-Volume Ultrafine Particle (HVUP) Sampler (76) at 400 L/min flow rate within
150 m downwind of the urban freeway (I-110). These aerosols represent a mix of fresh
and aged ambient PM, mostly from vehicular traffic. The filter-deposited dried nPM were
eluted by sonication into deionized water, referred to as nPM (59, 77).
The slurry PM (sPM) was collected using the aerosol-into-liquid collector, which
utilizes the saturation–condensation, particle-to-droplet growth system developed for the
versatile aerosol concentration enrichment system (VACES) (78, 79). This sampler
operates with a major flow rate of 300 l/min and collects ambient PM2.5 directly as
concentrated a slurry (80). The PM samples used in this study are listed in table 1.

15
Table 1. PM samples in this study
Extraction method Sample ID
Collection
Date(s)
Water-extracted (nPM)
1 Apr-Jun2016
2 June-Aug 2016
3 May-Sep2016
4
Nov2016-
Jan2017
5 May-Jun2017
6 Jan-Mar 2018
7 May-July2018
Direct suspension
(sPM)
1 May-Jun2017
2 Mar 2019
Diesel exhaust
particles (DEP)
1

2

2.2. Chemical profiles of nPM and sPM
Total elemental composition of the nPM samples was quantified by digestion of a
section of the filter-collected nPM using a microwave aided, sealed bomb, mixed acid
digestion (nitric acid, hydrofluoric acid, hydrochloric acid). Digests were analyzed by mass
spectrometry (SF-ICPMS) (81). WSOC analyses used Sievers 900 Total Organic Carbon
Analyzer (82).
Endotoxin (83) was assayed by Limulus based Pierce LAL chromogenic endotoxin
kit (Thermo Fisher Scientific). LPS neutralization was done by pre-incubating PM samples
with 10 ng/ml polymyxin B for 20 min at RT with vertexing.

16
2.3. Cell culture
THP1-Blue NF-kB monocytes from InvivoGen (San Diego, CA, USA) were
maintained in RPMI1640 medium containing 10% FBS, 1% pen/strep, 100 µg/ml
Normocin, and 10 µg/ml Blasticidin at 4x10
5
-1x10
6
cells/ml with 5% CO2 at 37
o
C.
BV-2 cells (mouse microglia line) were cultured in DMEM/F12 medium containing
10% FBS, 1% pen/strep and 1% glutamine with 5% CO2 at 37
o
C.
2.4. NF-kB activity
THP1-Blue monocytes were seeded in 96-well plates at 8x10
5
cells/ml for PM
treatment. After PM exposure, cells were centrifuged at 2,000g for 10 min and
supernatants were assayed for SEAP activity with QUANTI-Blue (InvivoGen). In brief,
supernatant was added to QUANTI-Blue solution in 96-well assay plates, incubated at 37
o
C/1h, followed by spectrophotometry at 652 nm.
2.5. Lipid peroxidation assay with DPPP assay
THP1-Blue monocytes/8x10
5
cells/ml were incubated with 4.5 µM DPPP for 20 min
at 37
o
C. Cells were washed twice, re-suspended in media, and seeded in 96-well plates.
Fluorescence was followed kinetically for 20 min (excitation, 351 nm; emission, 380 nm).
Lipid peroxidation was assessed 20 min posttreatment.
2.6. Nitrogen oxide (NO)
NO and NO2-/NO3- in cell media were measured by Griess reagent (84), with NaNO2
as standards.

17
2.7. Isolation of nuclear protein and Western blotting
Nuclear protein was extracted by using NE-PER™ Nuclear and Cytoplasmic Extraction
Reagents, and the Western blotting was performed as previously described (85).
2.8. Quantitative PCR
Total RNA was extracted with TRIzol and cDNA was synthesized using reverse
transcription kit from Thermal Fisher Scientific. Then mRNA was determined using
quantitative PCR (qPCR) assay in ABI7500 Real Time PCR machine (Thermal Fisher
Sceintific). The primers and analysis method were as reported before (85).
2.9. Animals
Animal procedures were approved by the University of Southern California (USC)
Institutional Animal Care and Use Committee (IACUC). Young adult C57BL/6NJ male
mice (6-8 wks, 27 g mean weight; n=10/group) were purchased from Jackson
Laboratories. After acclimation, mice were exposed to re-aerosolized nPM at 300 µg/m
3

(nPM group) or filtered air (controls) for 5 hrs/d, 3 d/wk, for 3 or 8 weeks (59, 77). Mice
were euthanized under anesthesia; brains were stored at -80 °C.
2.10. Inhalation exposure
To provide a stable source of concentrated PM for in vivo exposure studies, the
collected nPM slurries were re-aerosolized using the commercially available HOPE
nebulizers (B&B Medical Technologies, USA). As presented in Fig. S2, a pump (Model
VP0625-V1014-P2-0511, Medo Inc., USA) was generating HEPA-filtered compressed air
introduced into the nebulizer's suspension to re-aerosolize the PM liquid suspension and
produced the concentrated aerosol. Using a vacuum pump (Model VP0625-V1014-P2-

18
0511, Medo Inc., USA) at 5lpm dilution flow rates, the re-aerosolized PM is drawn through
a diffusion dryer (Model 3620, TSI Inc., USA) filled with silica gel to remove excess water,
and through Po-210 neutralizers (Model 2U500, NRD Inc., USA) to remove their electrical
charges. The particles are collected in parallel on 37 mm PTFE (Teflon) and Quartz (Pall
Life Sciences, 2-μm pore, Ann Arbor, MI) filters for chemical analysis and compared their
chemical composition (77).
In addition, a scanning mobility particle sizer (SMPS 3936, TSI Inc., USA) connected
to a condensation particle counter (CPC 3022A, TSI Inc., USA) was used to evaluate the
physical properties (i.e., number and mass size distribution) of the re-aerosolized PM at
100 inch H2O pressures and 5 lpm dilution flow rate.
2.11. Inflammatory cytokines
Cerebral cortex lysates were assayed for select cytokines by V-PLEX
proinflammatory panel 2 immunoassay (Mesoscale Diagnostics, Rockville, MD, USA).
2.12. Iba1 and Ab peptides
Brain sections were stained for Iba1 (Abcam, Cambridge, MA, USA) as marker of
microglia activation. Hippocampal Iba1 fluorescence was calculated by imageJ software.
Ab40/42 peptides were measured in supernatants of cortex using Peptide Panel 1 (4G8)
Kit V-plex (Meso Scale Discovery, Rockville, MD, USA) (86).
2.13. Statistical analysis
Data were analyzed by GraphPad Prism v.7. Depending on the number of classes in
the categorical variable, mean differences were analyzed by t-test or ANOVA, with Tukey

19
post hoc with Bonferroni multiple test correction. Values were expressed as mean ± SE
for p<0.05 as significant. Hierarchical clustering analysis used Rstudio. Interquartile
ranges were calculated from scaled data with mean of 0 and SD of 1, which compares
the variability in each outcome regardless of baseline average.

20
3. Results
3.1. Urban ambient PM induced NF-kB activation
The potency of nPM (PM with aerodynamic diameter ≤ 0.25 µm) and sPM (slurry PM
with aerodynamic diameter ≤ 2.5 µm) for activating NF-kB signaling was assayed with
human THP1 monocytes transfected with a NF-kB reporter (THP1-Blue monocytes). The
reporter is driven by IkB cis elements (NF-kB binding sites) that increase secretion of
embryonic alkaline phosphatase (SEAP) into cell media (Fig. 1A). At 5 µg/ml, the maximal
NF-kB induction for several batches was 10-fold or more above control (Fig. 1B). The
batches of nPM and sPM differed 4-fold in NF-kB activation. Diesel particles (DEP) had
weaker activity by mass, inducing NF-kB by 2-fold at 100 µg/ml (Fig.1B), but not at 5
µg/ml (not shown).
Since lipopolysaccharide can activate NF-kB (Fig. 1B), we examined its possible
contribution to NF-kB activation by nPM. Neutralization of LPS activity by pretreatment
with polymyxin B attenuated nPM-induced NF-kB activation by 10-30%, depending on
the batch (Fig. 1C); batches nPM-2 and sPM-2 retained activity after polymyxin treatment.
Polymyxin B (PMB) is a cyclic cationic polypeptide that can bind the negatively charged lipid
A, the toxic component of LPS, and neutralize the activity by reduction of binding with LPS
receptor TLR4 (87). Thus, these data suggest that LPS carried by nPM only partly
contributes to its NF-kB activation.

21

Fig. 1. NF-kB activation by air pollution PMs. (A) Schema of NF-kB activation in the THP1-
Blue monocyte assay; (B) NF-kB activation by 24 h (nPM and sPM, 5 µg/ml; DEP, 100
µg/ml; LPS, 10 ng/ml). (C) LPS neutralization by pretreatment with polymyxin B (10 ng/ml,
20 min) before NF-kB activation by nPM or sPM (5 µg/ml/ 24 h). * Adjusted p-value<0.05,
N=4.
To further analyze NF-kB activation, we showed increased nuclear levels of the NF-
kBp65 subunit (Fig.2A). Corresponding to NF-kB activation, nPM induced cytokines TNF-
Control
LPS
1
2
3
4
5
6
7
1
2
1
2
0
5
10
15
NFkB activation/Ctrl
NF-κB induction
*
nPM sPM DEP
Control
LPS
1
2
3
4
6
7
2
0
5
10
15
NFkB activation/Ctrl
LPS neutralization
Polymyxin B -
Polymyxin B +
*
*
*
*
*
*
nPM sPM
A
B C

22
a, IL-1b, IL-6 and IL-8 (Fig.2B). The link of cytokine responses to NF-kB induction was
shown by SN50, a specific NF-kB inhibitor, which suppressed cytokine induction by 65%
(Fig. 2B).

Fig. 2. nPM induced cytokines through NF-kB activation in THP1 monocytes. (A) nPM
increased nuclear NF-kB p65 (Western blot) by 5 µg/ml nPM (B) Cytokine induction by
nPM (RT-PCR) was suppressed by pretreatment with 40 µg/ml of SN50, a NF-kB
inhibitor. * P<0.05, ** P<0.01 vs control, #, P<0.05, nPM alone, N=4.

3.2. Urban ambient PM induced nitric oxide (NO) production by microglia
Since THP1 monocytes has low NO, nitritrite/nitrate production and induction
capacity (data is not shown) (88), NO was assayed in BV2 microglia (10 µg/ml for 6 h of
nPM, sPM or DEP). The nPM and sPM batches differed hugely in NO induction (Fig. 3),
ranging from no induction (nPM-7) to 8-fold induction (DEP-2), over controls. DEP
#
*
*
#
*
#
#
*
*
*
*
0
5
10
15
20
Expression/control
Control
nPM
SN50+nPM
TNFα IL-8
mRNA changes
6 h post treatment
IL-1β IL-6
0
1
2
3
4
5
OD (Intensity relative control)
NF-κB P65 nuclear localization
1 h post treatment
**
A B

23
induced slightly more NO than nPM at the same dose after 6 h (Fig. 3) and 24 h (not
shown).

Fig.3 NO induction by urban PMs. BV2 cells, a mouse microglia cell line, were exposed
to various PMs (10 µg/ml) or LPS (10 ng/ml) for 6 h; nitrite/nitrate in media was assayed
by Greiss reagent. * Adjusted p-value<0.05, N=4.

3.3. Urban ambient PM caused lipid peroxidation
Peroxidation of lipids (fatty acid moieties) in the plasma membrane results from direct
action of PM components, or secondarily from oxidants generated by cells interacting with
PM. We assayed fluorescence of DPPP (2, 2-diphenyl-1-picrylhydrazyl), a product of
oxidation in human THP1 monocytes. Pre-incubation with DPPP embeds DPPP into
plasma membranes (89), where it can be oxidized to the fluorescent product, DPPP oxide
(90). Again, we observed major batch differences at 5 µg/ml (Fig. 4), with lipid
peroxidation ranging from 2-fold (nPM-1) to 12-fold (sPM-2) above vehicle control.
Control
LPS
1
2
3
4
5
6
7
1
2
1
2
0.5
1
2
4
8
16
NO/control
Nitric oxide production
*
*
nPM sPM DEP

24

Fig. 4 Lipid peroxidation potency measured by DPPP assay. After incubation with DPPP
(4.5 µM/15 min), THP1-Blue cells were exposed to air pollution PM (10 µg/ml), LPS (10
ng/ml), or H2O2 (200 µM) for 20 min, and fluorescence was measured (excitation, 351
nm; emission, 380 nm). * p-value <0.05, N=4.

3.4. Comparisons of in vitro potency to induce NF-kB, NO, and lipid
peroxidation
The levels of induction for NF-kB, NO, and lipid peroxidation did not correlate significantly
by batch at p<0.05 (Fig S1); we note a marginal correlation between NF-kB activation
and NO production (p=0.08). These data suggest that batch differences of in vitro cellular
responses arise from independent processes from undefined compositional differences
between air pollution batches.

Control
H2O2 200 µM
LPS
1
2
3
4
5
6
7
1
2
1
2
0
100
200
300
400
Fluorescence : Ctl
Lipid peroxidation
nPM sPM DEP
*

25
3.5. In vivo neurotoxic effects are associated with in vitro NF-kB activity
of PM
To examine if in vitro activity can predict in vivo responses, we exposed the mice to
batches of urban nPM and sPM used for the in vitro assays: nPM-2, 4, 5, and 6, and sPM-
1, which differed in NF-kB activation (Fig. 1A). After 3 weeks exposure, nPM-4, nPM-5,
and sPM-1 induced cerebral cortex responses, including increased IFNg and decreased
IL2 (protein level measured by ELISA), with the magnitude differing by batch (Fig. 5A).
These batches were collected at the same site, but at different times (see Methods, Table
1).
Principle components (PC) were calculated for each mouse based the changes in 9
inflammatory cytokines in the cerebral cortex. PCs are considered as latent variable of
inflammation and is a linear combination of change in the 9 inflammatory cytokines. The
NF-kB responses in vitro were significantly associated with PC3, which explained 17% of
in vivo differences by batch (Fig. 5B). In contrast, lipid peroxidation and NO production
were poorly correlated with in vivo responses (Fig. 5C-D). This suggests that PM potency
of NF-kB activation could predict the long-term neurotoxic trajectories of the exposure.
The role of NF-kB was further tested by examining in vivo neurotoxic responses of
mice exposed to nPMs with different in vitro NF-kB activation potency for 8 weeks.
Exposure to nPM-2 that induced NF-kB in vitro by 12-fold also increased cerebral cortex
soluble Ab42 and Ab40 peptides, and hippocampal Iba1. In contrast, exposure to nPM-

26
6, which induced NF-kB activity by 5-fold at the same dose, did not alter soluble Ab42
and Ab40 peptides, and hippocampal Iba1 (Fig.5E).
Since oxidative stress is postulated in mechanisms of air pollution-caused
neurotoxicity (35), we assayed protein adducts of 4-hydroxynonenal (HNE), an
established marker of oxidative stress produced by lipid peroxidation (91). HNE-protein
adducts did not change in cerebral cortex and hippocampus of mice exposed to either
nPM batch for 3 or 8 weeks (data not shown).

27

Figure 5. In vivo neurotoxic responses of nPM paralleled in vitro NF-kB induction activity.
Young male mice were exposed for 3 or 8 weeks to nPM batches with different in vitro
NF-kB activity for neurotoxic responses. A) Inflammatory cytokines after 3 weeks of
exposure to 300 µg/m
3
of PM0.2. Heatmap shows fold changes of exposed animals
relative to control animals that were exposed to filtered air in parallel. B) In vitro NF-kB
activation potency of PM was associated with the variance of cerebral cortex inflammatory
A B
E
In vivo cerebral cortex
proinflammation
after 3 weeks exposure
Association of in vitro NFκB assay
and cerebral cortex neuroinflammation
2 6
0.0
0.5
1.0
1.5
2.0
2.5
Batch ID
Concentration relative to control
Aβ40 peptide
*
2 6
0.0
0.5
1.0
1.5
2.0
2.5
Batch ID
Concentration relative to control
Aβ42 peptide
*
2 6
0.0
0.5
1.0
1.5
2.0
2.5
Batch ID
Integrated density /control
Microglial Iba1, hippocampus
*
2 6
0
5
10
15
Batch ID
NFkB activation/Ctrl
In vitro NFκB induction
*
*
*
Cerebral cortex changes after 8 weeks exposure
C In vitro DPPP vs cerebral cortex proinflammation
In vitro NO vs cerebral cortex proinflammation
D
5 10 15
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
In vivo
neuroinflammation (PC3)
R
2
=0.26
P=0.006
In vitro NFκB
50 100 150 200
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
In vivo
neuroinflammation (PC3)
In vitro DPPP/control
P=0.74
1.4 1.6 1.8 2.0
-1.5
-1.0
-0.5
0.0
0.5
1.0
1.5
In vivo
neuroinflammation (PC3)
In vitro NO/control
P=0.39
Control
nPM

28
responses. Principle component analysis of the 9 inflammatory cytokines were used to
calculate principal components (PCs) of inflammation in the brain of each mouse. In vitro
NF-kB response was associated with PC3 that explained 17% of the variance brain
inflammatory response. Cerebral cortex inflammatory profile of mice exposed to different
batches of nPM or sPM did not correlate with in vitro (C) lipid peroxidation or (D) NO
production. E) nPM with stronger NF-kB activation in vitro caused more increase in
cerebral cortex amyloid-β40 &-42 peptides and hippocampal Iba1. Mean±SEM. *
adjusted P-values<0.05. N=5-10/group.

3.6. Chemical composition of nPM and sPM batches
We examined the chemical composition of different batches for clues to explain the
many-fold differences in activity. Metals, inorganic elements, water-soluble organic
carbons (WSOC), and LPS were assayed in 7 batches of nPM and 2 batches of sPM,
each collected at different times from the same site. The chemical composition varied
greatly among the batches (Fig. 6), e.g. for nPM, iron varied 80-fold, and WSOC up to
300-fold. The components with highest variance were LPS, and alkali and alkaline-earth
metals.
The nPM components of in vitro biological responses were identified by hierarchical
clustering, using a dissimilarity matrix (generated from pairwise Pearson correlation) of
biological responses and chemical subgroups of different nPM batches. This statistical
method clusters objects by their degree of association (Fig. 6H). Weak associations were
found between nPM components and lipid peroxidation or NO production (Fig.6H), which

29
confirms bulk chemical composition does not strongly predict the bioactivities of urban
ambient PM.

Fig.6 Chemical composition varied among nPM batches collected from the same site
at different times (Methods, Table 1). A) Water soluble organic carbons (WSOC); B) Alkali
and alkaline-earth metals; C) Non-metals; and D) Post transition metals; E) Transition
metals; F) endotoxin (83); G) Normalized interquartile range of components and metal
categories in nPM batches, showing magnitude of variation in different components. H)
Hierarchical clustering of biological responses and chemical components of nPM batches,
showing the degree of correlation. The dissimilarity matrix was calculated from Pearson
1
2
3
4
5
6
7
0
50
100
150
200
250
Alkali &
alkaline earth metals
Na
Ca
Mg
K
nPM
1
2
3
4
5
6
7
0
50
100
150
Non-metals
S
P
Se
nPM
1
2
3
4
5
6
7
0
10
20
30
40
ng/µg PM
Transition metals
Fe
Zn
Cu
Ti
Mn
Ni
nPM
1
2
3
4
5
6
7
0
5
10
15
Post-transition metals
Al
Pb
Sn
nPM
1
2
3
4
5
6
7
0
100
200
300
400
ng/µg PM
Water soluble
organic carbon
nPM
1
2
3
4
5
6
7
1
2
0.000
0.005
0.010
0.015
0.020
EU/ug PM
Endotoxin
nPM sPM
K
Cu
Endotoxin
Sn
Mg
Ti
Ni
Na
Al
Sum Alkali and alkali earth metals
P
Sum post transition metals
Zn
Mn
Ca
Pb
Pr
S
Sum non-metals
Fe
Sb
Sum transition metals
WSOC
Se
0.0
0.5
1.0
1.5
2.0
Standard deviation
Interquartile range normalized for nPM batches
A B C D E
G F H

30
correlation; clustering used average linkage method, which calculates the mean distance
between elements of each cluster for further merging and degree of correlation.

4. Discussion
Batches of urban ambient PM had wide differences NF-kB activation in a cell-based
assay that correlated best with in vivo neurotoxicity. To our knowledge this is the first
attempt to explore in vitro cell-based assays to predict particle-caused in vivo neurotoxic
effects. Our recent review (52) discussed how epidemiologic and animal studies using
ambient PM exposures fail to show correlations between bulk composition and in vitro
oxidative potential and neurotoxicity.
For cell-based assays, we selected a human monocyte cell line (THP1 cells) to
assess the pro-inflammatory potency and oxidative potential of PMs. Monocytes and
macrophages are major cells that scavenge particles and produce inflammatory
mediators in the body. The THP1 monocytes are widely used to study inflammatory
responses. Compared to human bronchial epithelial cells and THP1-derived
macrophages, the THP1 monocytes had stronger pro-inflammatory responses to nPM for
NF-kB activation and induction of pro-inflammatory cytokines (data not shown). For NO
production, we assayed microglia, the resident immune cell type in brain. NO production
is increased in BV2 microglia by PM2.5 (92). Most comparisons of cell-based and animal
responses have focused on acute pulmonary toxicity of particle exposure, using
macrophages and alveolar epithelial cells, which are directly contacted by inhaled PM

31
(70-74). There is considerable cell-type specificity for mechanisms of phagocytosis and
response to PM (93, 94).
The inflammatory potency was assayed by NF-kB, a master transcription factor that
regulates the induction of inflammatory mediators by air pollution PM (Fig.1). Compared
to assessing mRNA or protein of inflammatory mediators, this NF-kB reporter assay is
more accurate, faster, and cost saving. NF-kB activation occurs in inflammatory
responses common to all cells, whereas some pro-inflammatory mediators vary by cell
type and animal species.
Membrane oxidation was assessed as lipid peroxidation in THP1 monocytes by the
DPPP assay, which measures combined effects of the interaction of PM with extracellular
molecules, the cell membrane, and the surface chemistry of PMs. The DPPP assay
minimizes confounds of other methods based on the oxidation of dithiothreitol, 2',7'-
dichlorodihydrofluorescein and other fluorescent dyes by particles in a test tube (52). NO
production was measured in BV2 microglia using Greiss assay [55], a simple, fast and
widely used assay in NO studies. BV2 microglia were chosen for this assay because PM-
mediated NO production was not detected in THP1 monocytes and because microglia
are a major source of NO in the brain (95) and NO production was induced by urban
ambient PM in BV2 microglia (92).
In vivo neurotoxic effects were assessed by cytokines (Fig.5A) (59, 96, 97), microglial
activation (Iba1) (96-98), and a neurodegenerative peptide (Ab40/42), which are
increased in mouse brain by chronic exposure to nPM (38, 99-101).

32
As expected, the current data confirm that urban PM samples can vary widely in
bioactivities relevant to neurotoxicity for NF-kB activation, lipid peroxidation, and NO-
production (Fig.1, 3, 4), and for in vivo neurotoxic responses (Fig.5). This unequal
bioactivity/toxicity of particles collected from different sources, sites, and/or time has been
observed by many groups (102-111). It is generally thought that varying toxicity is due to
composition differences (62-66) that alter particle-cell interaction, phagocytosis, and
subsequent responses (93, 94, 112). However, there is little definition of how each
component may contribute to biological endpoints. The best correlations were for NF-kB
activation with LPS and WSOC in the particles for in vitro cell-based responses (Fig. 6H).
Specific components contributing to different cellular responses will be sought in future
studies. Because ambient particles vary widely in composition, we may consider
developing synthetic ultrafine PM with defined chemical components. Because the
surface components of particles cause initial cell responses in vitro and in vivo, surface
composition rather than bulk analysis needs to be evaluated for its biological activity. For
example, NF-kB induction in THP1 monocytes was dependent on iron on the surface of
natural silica particles, a small fraction of the total iron (113).
Cell-based inflammatory responses to predict in vivo lung responses to ultrafinePM
have been examined by several groups (70-74). These studies with different cell models
were focused on acute responses of lung toxicity. The nPM batches with a higher potency
of NF-kB activation caused stronger neurotoxic responses (Fig.5), suggesting that the
cell-based NF-kB activation assay is a better predictor than in vivo responses to chronic

33
nPM exposure. Because THP1 monocytes are human derived, we suggest that the NF-
kB assay could be useful for estimating neurotoxicity of human urban PM exposure.
The potency of urban PM for membrane lipid peroxidation or NO production was
poorly correlated with in vivo neurotoxic responses (Fig.5), suggesting that neither DPPP
oxidation nor NO production is predictive for the in vivo neurotoxicity. Although we don’t
know how the inhalation of nanoparticle causes neurotoxicity, some evidence suggests
indirect mechanisms via neurotoxicity carried by blood or lymph, lung-to-brain [36]. Unlike
cytokines, oxidants and NO produced by airway cells are usually short-lived and
detoxified quickly at production loci, and thus are unlikely to reach the brain at sufficient
concentrations for direct responses [15]. DPPP oxidation in THP1 cells results from direct
interaction of the particles with these H2O2-producing cells. These products are very low
in quantity and unlikely to travel far from the membrane site of production. Although the
lipid peroxidation measured by HNE-protein adducts was increased in some regions of
mouse brain upon acute exposure to urban particles (97, 114), we did not see 4-HNE
elevations in mouse brain after chronically exposed to nPM (data not shown).
In sum, these three cell-based assays for biological responses to air pollution
particles had limited correlation with in vivo responses with each other and with in vivo
neurotoxic responses. The in vitro NF-kB activation was the most robust correlation with
in vivo neurotoxic responses to urban ambient PM. A caveat is the small number of nPM
batches available for comparison for in vivo neurotoxic effects, which cannot rule out false
correlation between the cell-based assays and in vivo responses. Another limitation is the
time-dependent changes of in vivo endpoint makers, which were limited to 8 weeks for

34
Ab40/42 peptides and Iba1. The NF-kB cell-based assay hold promise for predicting
neurotoxicity in vivo.

Acknowledgements: This study is funded by NIH grants T32 AG052374 (A.H.), R01
ES023864 and P01 AG055367.
Author contribution: Conceptualization, H.F., C.E.F.; Methodology: H.Z., A.H.,
M.C., C.D.; nPM collection and characterization: A.H.M., M.H.S., C.S.; Writing: H.Z., A.H.,
C.E.F., T.E.M, H.F.; Supervision, Project Administration, and Funding Acquisition: H.F.

35
Supplementary data

Figure S1. NF-kB, lipid peroxidation and NO are not strongly correlated among different
nPM batches.

Figure S2. Aerosolization system for nPM inhalation exposure of mice. A) Schematic of
aerosol generation setup for filter collection and inhalation exposure; and B) Example of
nPM particle number distribution by size.
0 5 10 15
0
100
200
300
NF-κB induction
Lipid peroxidation
R
2
=0.07
P=0.47
0 5 10 15
0
2
4
6
NF-κB induction
NO/control
R
2
=0.36
P=0.08
0 100 200 300
0
2
4
6
Lipid peroxidation
NO/control
nPM
sPM
R
2
=0.001
P=0.9
A B C
0
50000
100000
150000
200000
250000
300000
350000
400000
450000
500000
550000
600000
650000
10 100 1000
Particle number concentration
(#/cm3)
Dp(nm)
B
A

36
Chapter 3. Toxicity of urban air pollution particulate matter in
developing and adult mouse brain: comparison of total and filter-
eluted nanoparticles
(Adapted from Environment International Journal, doi: 10.1016/j.envint.2020.105510)

Authors: Amin Haghani
1
, Richard Johnson
1
, Nikoo Safi
2
, Hongqiao Zhang
1
, Max
Thorwald
1
, Amirhosein Mousavi
3
, Nicholas C. Woodward
4
, Farimah Shirmohammadi
3
,
Valerio Coussa
1
, John P. Wise Jr
5
, Henry Jay Forman
1
, Constantinos Sioutas
3
, Hooman
Allayee
4
, Todd E Morgan
1
, Caleb E Finch
1*

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Center for Cancer Prevention and Translational Genomics at the Samuel Oschin
Comprehensive Cancer Institute, Cedars-Sinai Medical Center, Los Angeles, CA.
3, Viterbi School of Engineering, University of Southern California, Los Angeles, CA.
4, Department of Preventive Medicine, University of Southern California, Los Angeles,
CA.
5, School of Medicine, University of Louisville, Louisville, KY
*Corresponding author. Email cefinch@usc.edu.

Highlights:
• Filter-eluted nanoparticles (nPM) are depleted in PAH and transition metals compared to total
nanoparticles (sPM)
• Inhalation of nPM and sPM caused comparable neurotoxic reactions in adult and developing
brain
• Gestational exposure to nPM or sPM caused depressive behaviors in adults
• Gestational exposure to sPM caused glutamatergic changes in adults

37
Abbreviations:

Term Abbreviation
air pollution
particulate matter
AirP
PM
polycyclic aromatic hydrocarbons PAHs
nano-sized particulate matter, filter extraction nPM
slurry particulate matter sPM
ultrafine particulate matter, diameter < 0.2µm PM0.2
i.p. glucose tolerance test IPGTTs
water-soluble organic carbon WSOC
versatile aerosol concentration enrichment system VACES

Graphical abstract:

Gestational exposure:
Metabolic dysfunction
Depressive behavior
Adult exposure:
NF-κB mediated inflammation
Glutamatergic response
Gestational exposure:
Glutamatergic deficits
Through systemic or local effects
PM0.2
nPM
sPM
PAH
organic carbons,
transition metals,
endotoxins,…

38
Abstract
Air pollution (AirP) is associated with many neurodevelopmental and neurological disorders
in human populations. Rodent models show similar neurotoxic effects of AirP particulate matter
(PM) collected by different methods or from various sources. However, controversies continue on
the identity of the specific neurotoxic components and mechanisms of neurotoxicity. We collected
urban PM by two modes at the same site and time: direct collection as an aqueous slurry (sPM)
versus a nano-sized sub-fraction of PM0.2 that was eluted from filters (nPM). The nPM lacks
water-insoluble PAHs (polycyclic aromatic hydrocarbons) and is depleted by >50% in bioactive
metals (e.g., copper, iron, nickel), inorganic ions, black carbon, and other organic compounds.
Three biological models were used: in vivo exposure of adult male mice to re-aerosolized nPM
and sPM for 3 weeks, gestational exposure, and glial cell cultures. In contrast to larger
inflammatory responses of sPM in vitro, cerebral cortex responses of mice to sPM and nPM
largely overlapped for adult and gestational exposures. Adult brain responses included induction
of IFNg and NF-kB. Gestational exposure to nPM and sPM caused equivalent depressive
behaviors. Responses to nPM and sPM diverged for cerebral cortex glutamate receptor mRNA,
systemic fat gain and insulin resistance. The shared toxic responses of sPM with nPM may arise
from shared transition metals and organics. In contrast, gestational exposure to sPM but not nPM,
decreased glutamatergic mRNAs, which may be attributed to PAHs. We discuss potential
mechanisms in the overlap between nPM and sPM despite major differences in bulk chemical
composition.

Keywords: PM0.2, inhaled pollutants, neurodevelopment, neuroinflammation

39
1. Introduction
Air pollution (AirP) is associated with adverse impact on human brain functions throughout
life, from development, e.g., myelin deficits and risk of autism (27, 115), into later ages, e.g.,
accelerated cognitive decline and risk of dementia (38, 116-118). Rodent models show similar
neurotoxic effects of AirP particulate matter (PM). Adult brain responses to various AirP models
included inflammatory responses with microglial activation (96, 119); induction of IFNg, NF-kB,
and TLR4 (45, 97); increased amyloidogensis (38, 120); and altered glutamatergic receptors (13)
. Gestational exposure to AirP activated microglia (32), impaired myelinogenesis (40, 121);
decreased adult neural stem cells (96), and increased depressive behaviors (96, 122).
Particulate matter (PM) is a heterogeneous mixture that varies widely by time and place in its
numerous toxic compounds (e.g., polycyclic aromatic hydrocarbons (PAHs), organic and
elemental carbon (OC/EC) and transition metals). However, it is unclear which of these potentially
neurotoxic components of AirP are responsible for injury. While cognitive impairments in several
human populations varied in proportion to ambient density of AirP (117), no study to our
knowledge has compared the neurotoxicity of AirP from different sources in rodent or cell models.
The identification of causative neurotoxic components is challenged by the complexity of the
mixtures, the overlapping toxicity of individual components, and their interactions. Our approach
to identify some chemical-specific responses in PM mixture is to compare experimental responses
of a total PM sample (i.e. collecting all PM components) with a filter-eluted fraction of ultrafine PM
that lacks PAHs (13), and is depleted in a wide range of other toxic components (e.g., metals,
black carbon, and some organics) (13, 45). Using of a high-volume aerosol-into-liquid collector
allows the direct collection of ambient PM into aqueous suspension (slurry, sPM) (80). The sPM
is equivalent in PAH and transition metal composition to ambient PM (77).

40
Toxicological studies of PAHs suggest their importance to neurotoxicity of AirP in human
and rodent models. Epidemiological and clinical studies show correlations of PAH with
myelination deficits (27), autistic behavior (20), and childhood obesity (123). Gestational exposure
of rodents to specific PAHs, e.g., benzo(a)pyrene at high dosage, caused adult impairments of
the hippocampus, including long-term potentiation, with altered glutamate receptors such as
GluN1 (124-126), oxidative stress and lipid peroxidation in several brain regions (127). Systemic
effects include greater post-weaning weight gain, glucose dysregulation, and decreased lipolysis
(128, 129). These toxic effects of gestational PAH exposure resemble those caused by
gestational exposure to AirP-PM or diesel exhaust particles (Table S1) (32, 33, 130-132).
Contrary to expectations, the gestational exposure to nPM, which lacks PAHs, also caused similar
long-lasting neurotoxic effects (96, 122, 133). These observations necessitate direct comparison
of nPM with total PM0.2.

41
2. Methods
2.1. Air pollution sample collection:
Ambient nanoscale particulate matter (nPM; particles with aerodynamic diameters
less than 0.20 µm) were collected on an 8 × 10 inch-Zeflour PTFE filter (Pall Life
Sciences, Ann Arbor, MI) by a High-Volume Ultrafine Particle (HVUP) Sampler (76) at
400 L/min flow rate at the Particle Instrumentation Unit of University of Southern California
located within 150 m downwind of a major freeway (I-110). These aerosols represent a
mixture of primary emissions ambient PM from vehicular traffic on this freeway.
Gravimetric mass (nPM mass concentration) was determined from pre- and post-
weighing the filters at 22-24 ºC/relative humidity 40-50% by high precision (±0.001 mg)
microbalance (MT5, Mettler Toledo Inc., Columbus, OH). The nPM fraction of filter-
deposited PM0.2 was obtained by 30 min aqueous sonication (13). The resuspended
nPM mass was quantified as the difference between the total pre-extraction and the post-
extraction weight of filters.
The second set of samples (sPM) was collected using the aerosol-into-liquid collector tandem
that utilizes the particle-to-droplet growth system via supersaturation of ultrapure water vapor of
the versatile aerosol concentration enrichment system (VACES) (78, 79). This sampler operates
at a flow rate of 200 l/min and collects ambient PM2.5 (i.e, particles with aerodynamic diameters
< 2.5 µm) directly as concentrated slurry samples (80). Briefly, sampled air is drawn into a
saturator tank for mixture with ultrapure water (Milli-Q integral system (Resistivity: 18.2 MΩ•cm,
< 1 particulate/mL, Millipore and Sigma-Aldrich, North Carolina, USA) vapor at 30 °C. The
particle–vapor mixture then passes through a condensational growth section where it is cooled to

42
about 20 °C; the resulting super-saturation condenses ultrapure water vapor onto incoming
particles, which grow to about 3.5-4 mm droplets. Grown droplets are separated from the air
stream by inertial impaction and accumulated as concentrated slurry samples. (sPM). Both nPM
and sPM samplings were concurrent in May of 2017.
2.2. Characterization of PM components:
The re-aerosolized ultrafine nPM and sPM were collected on filters in animal
exposure chambers for chemical characterization (134). The total elemental composition
of the nPM samples was quantified by digestion of a section of the nPM deposited filter
using a microwave aided, sealed bomb, mixed acid digestion (nitric acid, hydrofluoric
acid, and hydrochloric acid). Digests were subsequently analyzed by high-resolution
mass spectrometry (SF-ICPMS) (81). Total and Water-Soluble Organic Carbon
(TOC/WSOC) analysis was conducted utilizing the Sievers 900 Total Organic Carbon
Analyzer (82). PAHs were quantified using gas chromatography-mass spectrometry (GC-
MS) (135). The isotopically-labeled internal recovery standards used as spikes include
pyrene-D10, benz(a)anthracene-D12, coronene-D12. For PAH assay, AirP filters were
extracted by acetone/dichloromethane and derivatized by diazomethane. Black carbon
was assayed in re-aerosolized nPM and sPM by an Aethalometer (Model AE51,
AethLabs, California, USA). These PM samples were compared to ambient PM0.2
collected on filters from the same site in parallel with the collection of nPM and sPM in
2017. Chemical composition was analyzed by SF-ICPMS and GCMS.

43
2.3. Animal ethics statement
All animal procedures of this study were approved by the University of Southern
California (USC) Institutional Animal Care and Use Committee (IACUC).
2.4. Exposure of adult mice
Young C57BL/6NJ male mice (6-8 weeks; 27.5 ± 2 gm mean weight; n=10/group)
were purchased from Jackson Laboratories. After acclimation for 1 week, mice were
exposed to different concentrations of re-aerosolized PM for 45 hours over 3 weeks (5 h
per day, 3 days per week). Mice were transferred from home cages into sealed exposure
chambers that were well ventilated (13). One chamber was the “control” filtered air. Other
cages were exposed to three PM levels of approximately 100, 200, and 300 µg/m
3
. For
100 and 200 µg/m
3
, PM was diluted by filtered air. Mass concentration of re-aerosolized
nPM was measured gravimetrically by filters parallel to the exposure stream, before and
after exposure. Nebulizer pressure was adjusted to yield similar distributions of particle
number size, equivalent to ambient PM0.2 for re-aerosolized nPM and sPM (Fig. S1) (77).
Thus, sPM for in vivo exposure excluded PM > 0.2 µm.
2.5. Gestational exposure
9-week-old C57BL/6NJ mice were housed as breeding trios (1 male, 2 females) and
randomly assigned to each treatment group (nPM, sPM, FA), 6 breeding trios per group.
Mice were exposed to nPM or sPM at 300 µg/m
3
throughout gestation for 5 h/day, 3
days/week for 3 weeks. The five breeding trios per treatment group had viable litters;
numbers of offspring were 35 pups (nPM), 30 (sPM), 33 (FA).

44
2.6. Forced swim
Stress coping strategies were assessed at 11 weeks of age. Mice were placed in a
cylindrical water bath at 24-25°C and recorded for five minutes with latency to the first
period of immobility and total time immobile. For glutamatergic functions, at 16 weeks of
age, male sPM-exposed mice underwent a second round of forced swim testing 30
minutes after i.p. MK-801 (0.06mg/kg) or saline (control).
2.7. Glucose tolerance test
At age 16 weeks, mice underwent i.p. glucose tolerance tests (IPGTTs) (136). After
overnight fasting, mice were given i.p glucose (1mg/g body weight; 10% wt/vol in sterile
water). Tail vein blood was sampled at 0, 15, 30, 60, 90, and 120 min post-injection.
2.8. Body weight and composition analyses
Bodyweight was measured every 1-2 weeks after weaning. Body composition for fat,
lean mass, body fluids, and total body water was assessed by NMR (LF90, TD-NMR;
Bruker, USA) at 3 ages in weeks 4 to 16.
2.9. Cell fractions
The frontal cerebral cortex (20 mg) was mechanically homogenized in 1x RIPA buffer
supplemented with 1mM Na3VO2, 1mM phenylmethylsulfonyl fluoride (PMSF), 10 mM
NaF, phosphatase inhibitor cocktail (Sigma), and Complete Mini EDTA-free Protease
Inhibitor Cocktail Tablet (Roche). For biochemical assays, supernatants were obtained
by centrifugation 10,000g/10 min. For NF-kB assays, nuclear and cytosolic fractions were
obtained from tissue homogenization in sucrose-Tris (STM) buffer with phosphatase and

45
protease inhibitors (137). After centrifugation (800 g/15 min, 4 °C), the cytosolic
supernatant was collected and the nuclear pellet was washed in STM buffer, then
resuspended in NET and sonicated. Fraction purity was validated by immunoblotting for
nuclear marker histone 3 (H3) and cytosolic glyceraldehyde 3-phosphate dehydrogenase
(GAPDH).
2.10. Western Blot
Protein lysates (20 µg) were electrophoresed on Criterion 4-15% TGX gels (Biorad)
and transferred to PVDF membranes. Following washing with TBS+0.05% Tween-20
(PBST), membranes were blocked (LiCOR) 1 h at ambient temperature, followed by
primary antibody incubation overnight at 4 °C: anti-GluR1 (1:1000, Rabbit polyclonal,
Abcam, ab31232), anti-actin (1:5000, mouse, Sigma), anti-NF-kB/p65 (1:750, Rabbit
polyclonal, Cell Signaling Technology, D14E12), anti-H3 (1:1000, Rabbit polyclonal, Cell
Signaling Technology, D1H2), and anti-GAPDH (1:500, Mouse monoclonal, Santa Cruz
Biotechnology, sc-32233). After 1 h incubation with 1:20,000 fluorochrome-conjugated
LICOR-antibodies (anti-mouse IRDye 800CW or anti-rabbit IRDye 700CW), blots were
scanned and band intensity analyzed by ImageJ.
2.11. Multiplex Immunoassay
Cerebral cortex lysates were analyzed by V-PLEX Proinflammatory Panel 2
immunoassay (Mesoscale Diagnostics, Rockville, MD).

46
2.12. PCR
Cerebral cortex RNA was extracted by TriZol; cDNA was prepared using qScript
cDNA Supermix (Quantabio); gene-specific primers, Table S2. Data were normalized to
GAPDH.
2.13. Cell culture
Primary cultures of mixed glia (astrocyte: microglia, 3:1) from rat cerebral cortex of
PN day 3-5 were grown in Dulbecco’s modified Eagle’s medium (DMEM)/F12 (Cellgro,
Mediatech, Herndon, VA) containing 10% fetal bovine serum, 1% penicillin/streptomycin
and 1% L-glutamine (45). BV2 microglia (mouse-derived) were similarly cultured.
Cytotoxicity of nPM and sPM was assessed by MTT and by CellTiter-Glo Luminescent
Cell Viability (Fig. S2-S3).
2.14. Cellular Nitrate/Nitrite production
Conditioned media were assayed for nitric oxide (NO) and NO2
-
/NO3
-
by the Griess
reagent (84).
2.15. NF-kB assay
THP1-BlueÔ NF-kB cells (InvivoGen, San Diego, CA) were grown in RPMI 1640
media with 2mM L-glutamine, 25 mM HEPES, 10% FBS, 100µg/ml Normocin, and 1%
penicillin/streptomycin. THP1-Blue NF-kB cells are derived from human THP1 monocytes
by stable transfection of an NF-kB-inducible SEAP reporter (secreted embryonic alkaline
phosphatase) (138). SEAP1 expression is driven by an interferon-g promoter fused with

47
five copies of NF-kB consensus response elements. SEAP in media were assessed by
enzymatic reaction with QUANTI-BlueÔ (InvivoGen).
2.16. Microbial screening
Bacterial growth media for PM samples included BBL Brain Heart Infusion,
MacConkey Agar, BD Difco™ LB Broth, and Difco R2A. Samples (100 µl were incubated
at 37°C/72 h.
2.17. Endotoxin
Levels were assessed by a Limulus assay (Pierce LAL chromogenic endotoxin,
Thermo Fisher). For LPS neutralization, PM samples were preincubated with Polymyxin
B (0, 1, 10 ng/ml) for 20 min/ambient temperature.
2.18. Statistical analysis
Data were analyzed by GraphPad Prism v.7 for ANOVA with multiple test correction.
Multivariate modeling was used for cerebral cortex cytokines. Multivariate analysis,
heatmaps, Pearson correlations, and univariate linear regression used Rstudio.

48
3. Results
3.1. Chemical characterization of PM samples
AirP particulate matter (PM) was collected in two modes. The nPM, as used in our prior
studies, was collected as nanosized PM0.2 on the filter, and eluted into the water by sonication
as an aqueous suspension. The sPM was PM2.5 directly collected from ambient air pollutants
into water suspension. Upon re-aerosolization, the size of sPM is fractionated for exposure to
PM0.2. Data are shown as analytical results of re-aerosolized (nPM, sPM) to compare with total
ambient PM0.2 (Fig 1).
The chemical composition (per PM mass) of nPM and sPM was equivalent for total organic
carbon, whereas nPM had 4-fold less black carbon and no detectible PAHs (Fig 1A). The highest
PAH levels in sPM were benzo(g,h,i)perylene, benzo(b)fluoranthene, and coronene (0.023,
0.016, and 0.015 ng/µg PM, respectively). Ambient PM0.2 and re-aerosolized sPM had equivalent
levels of the major PAHs.
The concentration of iron and other metals was much lower in nPM than sPM (Fig 1B): iron,
30-fold lower; copper, 6-fold lower; nickel, 2-fold lower; total transition metals, 8-fold lower; post-
transition metals, 2-fold lower. nPM had 50% higher levels of other metals and non-metal
elements (Fig 1C) (i.e. Na, Mg, S, Mg). The chemical composition and concentration of nPM agree
with prior analyses (13, 45).
An independent sample of nPM and sPM suspensions showed equivalence for levels of
organic acids, n-alkanes, hopanes and steranes, which represent less than 0.1% of total organic
carbon (Fig. S4).

49

Figure 1. Chemical composition of re-aerosolized nPM and sPM compared with total ambient
air PM0.2 collected in parallel on a filter. A) Total organic carbon, black carbon, and PAH with
concentration > 0.005 ng/µg PM. Indeno(1,2,3-cd)pyrene with 0.015 ng/µg PM concentration and
EPA toxicity equivalent factor of 1 was the most toxic PAH species in sPM and ambient PM
samples. B) Metals, transition, and post-transition, alkali and alkaline earth; C) non-metals. Only
elements with > 0.1 ng/µg PM are shown.
3.2. Inflammatory responses of cerebral cortex of adult mice to nPM and
sPM
Adult mice were exposed for 45 total hours to re-aerosolized nPM and sPM. Both types of
PM, with the dose range of 100-300 µg/m
3
, induced inflammatory responses in the cerebral
cortex, consistent with the responses upon nPM exposure in prior studies (114). Bodyweight was
not altered by exposures to nPM or sPM (Fig. S5). Protein levels of IFNg in the cerebral cortex
nPM
sPM
0
50
100
150
ng/µg PM
Alkali &
alkaline earth metals
K
Mg
Ca
Na
Re-
aerosolized
nPM
sPM
0
20
40
60
80
ng/µg PM
Se
P
S
Re-
aerosolized
Non-metals
nPM
sPM
PM0.2
0
2
4
6
8
10
ng/µg PM
Transition metals
Ni
Mn
Ti
Cu
Fe
Re-
aerosolized
Ambient
nPM
sPM
PM0.2
0
5
10
15
ng/µg PM
Post-transition metals
Sn
Pb
Al Re-
aerosolized
Ambient
Zn
nPM
sPM
PM0.2
0.00
0.05
0.10
0.15
ng/µg PM
Polycyclic aromatic hydrocarbons
benzo(k)fluoranthene
benz(a)anthracene
benzo(e)pyrene
indeno(1,2,3-cd)pyrene
coronene
benzo(b)fluoranthene
benzo(g,h,i)perylene Re-
aerosolized
Ambient
A. Organic compounds
B. Metals
nPM
sPM
0
100
200
300
400
ng/µg PM Total organic carbon
Re-
aerosolized
nPM
sPM
0
20
40
60
80
100
ng/µg PM
Black carbon
Re-
aerosolized
C. Non-metals

50
were induced with similar dose-dependence by nPM and sPM, while nPM caused increased IL1b
and decreased IL2 (Fig. 2A). IL6 and TNFa did not respond to nPM or sPM. For mRNA, nPM
and sPM decreased TLR4 and MyD88 (Fig. 2B). sPM exposure decreased TLR4 levels more
than nPM. GluA1 mRNA was decreased at 300 µg/m
3
by nPM, but not by sPM. GluA1 protein
was slightly lowered by nPM (-10%, P=0.2) (Fig. S6). Nuclear localization of NF-kB/p65 was
increased by sPM (Fig. 2C). The cytosolic NF-kB/p65 was increased by nPM. These data suggest
that adult exposure to nPM and sPM caused inflammatory responses in the brain, and sPM had
a stronger effect.

51

Figure 2. Cerebral cortex inflammatory responses of nPM and sPM in adult mice. A) Dose-
response of nPM and sPM showed an increase of IFNg. nPM increased IL1b and decreased IL2.
B) mRNA determination showed decreased TLR4 and MyD88 upon exposure to nPM and sPM.
GluA1 was decreased by nPM, but not by sPM. C) NF-kB/p65 nuclear localization. Mean ± SE.
Adjusted p-values: < 0.05 (*), < 0.01(**).

52
3.3. Gestational impact of nPM and sPM on behaviors and neuro-
inflammation in later life
To compare the neurotoxic effects of gestational exposure to nPM and sPM, the dams were
exposed to 300 µg/m
3
of nPM or sPM during the three weeks of pregnancy, followed by an
examination of their offspring. Compared to controls, the postnatal body weight gains were greater
in nPM and sPM by week 4 (juvenile) (Fig. 3A). Body fat was greater only for nPM males by 4
weeks (Fig. 3B). At 16 weeks, glucose tolerance was impaired in nPM males, but not females
(Fig. 3C). Changes in glucose tolerance and fat content were correlated (Fig. 3D).
Young adults (age > 10 weeks) after gestational exposure to nPM and sPM showed similar
depressive changes (forced swim cognitive test), but the effect was greater on males than females
(Fig. 3E). The depression caused by gestational sPM exposure was ameliorated by MK-801, an
NMDA glutamate receptor antagonist (Fig. 3F), suggesting altered NMDA pathway may underlie
depression caused by gestational exposure. Brains from sPM males were not further analyzed to
avoid confounds from MK-801 injection and the second round of forced swim behavior test.
Expression of inflammatory and glutamate associated genes in the cerebral cortex was
altered by gestational exposure only in female offspring exposed to sPM (Fig. 3G). Among the
examined inflammatory mediators, decreased Cox2 (-30%) and iNos (-70%) were the largest
responses. The mRNA levels of genes involved in glutamatergic pathway were decreased by 25-
35% (Fig. 3G): glutamate receptors (Nmda3a, Nmda2c, GluA1, GluA2, and Glun1), glutamine
synthase (Glul), glutaminase (Gls), and glutamate transporters (Glt1, and Glast).
In brief, gestational exposure to nPM and sPM caused depression similarly in later life, but distinct
effects were observed, including effects of nPM on metabolism and effects of sPM on glutamate
pathways. These effects were sex-dependent.

53

Figure 3. Effects of gestational exposure to nPM and sPM. A) Weight difference at age of4
weeks after gestational exposure to nPM or sPM. B) Body fat content at the age of 4 weeks,
measured by NMR (N=6-25/group). C) Glucose tolerance test at the age of 16 weeks after
gestational exposure to nPM or sPM (N=7-13/group). Data are shown as AUC (‘area under the
curve’ for blood glucose changes (mg/dL) during 2 h post glucose challenge). D) Correlation of
glucose tolerance in gestationally nPM exposed offspring at week 16 with body fat content of
postnatal week 4. Multiple test correction by Benjamini two-stage Adjusted p-values: < 0.05 (*), <

54
0.01(**). E) Depressive behavior of mice assessed by forced swim cognitive test at the age of 11-
15 weeks (N=5-16 mice/group). F) Attenuation of depression indicator by MK-801 (i.p. 0.06mg/kg,
30 min before forced swim (N=5-7/group). G) RT-qPCR analysis of the selected glutamatergic
and inflammatory genes in the cerebral cortex (N=7-17/group). Heatmap represents log2 fold
changes of mRNA relative to non-exposed controls of each group. * significance at 5% FDR
rate, exposed vs control.
3.4. In vitro inflammatory responses of glia cells to nPM and sPM
To further compare the toxic effects of nPM and sPM, the expression of inflammation-related
genes was measured in cultured cells upon exposure. The effects of nPM and sPM on the TLR4
pathway (Fig. 4A), which were activated by AirP in lung (139) and brain (45), were examined in
two glial cell models. In BV2 microglia, MyD88 and NF-kB1 were equally induced by nPM and
sPM, while the expression of TLR4 was not changed (Fig.4B). The induction of downstream
inflammatory mediators, including IL6, IL1b, TNFa, Cox2, and iNOS, was 50% greater by sPM
vs nPM (Fig. 4B). These five inflammatory genes shared similar correlations of response to sPM
and nPM (heatmap of pairwise Pearson correlations) (Fig. 4C), e.g., IL1b-iNOS and IL1b-Cox2
(Fig. 4D). Nitric oxide (NO) induction was evaluated in mixed glia (Fig. 4E). NO production was
induced by an equal scale in response to nPM and sPM (Fig. 4E), similarly, iNOS mRNA was
induced 25-fold by both nPM and sPM (Fig. 4F). These data indicate that both nPM and sPM can
induce the expression of inflammatory genes, with a stronger effect oft sPM on microglia.

55

Figure 4. in vitro inflammatory gene responses to nPM and sPM in glia. A) Schema of PM
mediated TLR4 pathway activation and inflammatory gene responses. B) mRNA levels of
downstream inflammatory genes after 6 hr exposure to 5 µg/ml nPM or sPM in BV2 microglial

56
cells (RT-qPCR; N=6/group´2). C) Heatmap of pairwise Pearson correlations of mRNA response.
D) IL1b-iNOS and IL1b-Cox2 as examples of inflammatory genes with positively correlated
response to nPM and sPM. E) Dose-response of nitric oxide (NO) production after 24 h of
exposure to nPM and sPM; Greiss assay (N=6-8/group, an average of 3 independent replicates).
F) Induction of iNOS mRNA (RT-qPCR) by 5 µg/ml nPM and sPM in mixed glia (N=6/group,
duplicate average). Mean ± SE. Adjusted P-values: < 0.05(*), < 0.01(**), < 0.001(***), <
0.0001(****), compared with vehicle control.
3.5. In vitro NF-kB induction by nPM and sPM
The time- and dose-dependent response of NF-kB activity to nPM and sPM was studied with
a human monocyte reporter for NF-kB (Fig. 5A-B). At 5 µg/ml, sPM had 50% steeper dose-
response slope than nPM (Fig. 5B), paralleling BV2 cell responses; at 1 µg/ml, neither nPM nor
sPM activated NF-kB. The stronger NF-kB activation by sPM matches the greater mRNA
induction of 5 genes regulated by NF-kB in BV2 cells (Fig.4B).
Because sPM and nPM are collected from ambient air, we investigated their microbial
content. On four growth media (Brain Heart infusion, MacConkey, LB, R2A), the sPM samples
had at least 10-fold greater microbial growth with more species than nPM (Table S3).
In vitro cell cultures included antibiotics to minimize bacterial growth, for up to 24 h incubation
with sPM or nPM, cell culture medium did not manifest turbidity and no bacterial contamination
was observed (data not shown). Endotoxin (LPS), a trigger of inflammatory responses, was also
assayed. Both nPM and sPM had 0.014 EU/µg PM of LPS (Fig. 5C), or approximately 1.4 pg
LPS/µg PM (1.4´10
-4
pmol LPS/µg PM) (140). Neutralization of LPS activity by polymyxin B
attenuated 40% of the nPM mediated NF-kB response (Fig. 5D). In contrast, NF-kB activation by
sPM was not altered by polymyxin B with a concentration up to 10 ng/ml.

57

Figure 5. NF-kB response to nPM and sPM in THP1-blue monocytes. A) Dose- and time
response of NF-kB induction by nPM, sPM, and LPS at 3-h intervals during 6-24 h exposure. B)
Slopes of NF-kB responses during 24 h of exposure to nPM, sPM, and LPS calculated by linear
regression (N=4/group). C) Endotoxin levels in nPM and sPM, measured by Limulus amoebocyte
lysate assay (LAL). EU ~ 100 pg LPS (10,000 g/mol) (140, 141). D) Contribution of LPS to NF-kB
responses of THP1-blue monocytes. LPS contributed up to 40% of NF-kB response in nPM and
not sPM. Mean ± SE. Adjusted P-values: < 0.05 (*), < 0.01(**), < 0.001(***), < 0.0001(****),
compared with vehicle control.
0
5
10
15
20
25
Hours post treatment
NF-κB activation/Ctrl
6 24
1
µg/ml
5
µg/ml
10
µg/ml
1
µg/ml
5
µg/ml
1
ng/ml
10
ng/ml
6 24 6 24 6 24 6 24 6 24 6 24 6 24
NF-κB activation in THP1-Blue cells
LPS nPM sPM Control
0.0
0.2
0.4
0.6
0.8
1.0
Slope
Time slope of NF-κB induction
Control
nPM
sPM
LPS
****
**** ****
****
****
**
1 5 10 1 5 1 10
µg/ml ng/ml
Concentration
A B
0.000
0.005
0.010
0.015
0.020
EU/ug PM
Endotoxin
Samples
0.5
1
2
4
8
16
Polymixin B concentration (ng/ml)
NF-κB activation / control
LPS contribution in NF-κB activation
Control
LPS (10ng/ml)
sPM (5µg/ml)
nPM (5µg/ml)
0 1 10 0 1 10 0 1 10 0 1 10
*
****
****
**
****
C D

58

Figure 6. Summary of the chemical composition and biological effects of nPM and sPM.
Underline indicates distinct effects or chemical components of nPM and sPM.

4. Discussion
Despite the compositional differences of nPM and sPM, they shared neurotoxic and systemic
effects for adult and gestational exposure, and cell model responses (Fig. 6). In contrast with
sPM, nPM lacked detectable PAHs; the metal content of nPM was lower, differing more for the
transition than non-transition metals. Moreover, sPM had at least 10-fold higher microbial load
than nPM. Thus, the shared neurotoxicity of nPM and sPM may be attributed to transition metals,
and unidentified organic carbon compounds among many other components. Below, we discuss
potential mechanisms underlying shared neurotoxic responses of nPM and sPM. We also discuss
the distinct responses that could be attributed to compositional differences of nPM and sPM.
Additionally, several PM characteristics such as surface chemistry and reactivity that are currently
missing in the experimental neuroscience framework could play a predominant role in AirP
neurotoxicity.

59
Shared responses to nPM and sPM were observed in all three models. Three-week exposure
of adult mice to nPM or sPM (45 hours total) induced a dose-dependent increase of IFNg, IL1b,
and activated the TLR4-NFkB pathway in the cerebral cortex. Also, gestational exposure of nPM
and sPM caused similar depressive behaviors in adult mice with sex specificity for males, which
extends prior findings by us (96, 122) and others (32, 121). In cell models, both nPM and sPM
samples induced NF-kB mediated inflammation and nitric oxide production. These results suggest
that the components and particle characteristics of nPM sufficed to cause neurotoxicity in the 3-
week gestational and adult exposures, and in cell models. We provisionally conclude that PAHs,
which are absent from nPM, have a minor contribution to neuroinflammation for the 3-week
duration of exposure.
Metal content was consistently lower in nPM than sPM: two transition metals differed greatly
(Cu, 6-fold; iron, 30-fold), while Ni, Mn, and Zn were within 50%. As noted for PAH, responses to
sPM and nPM were shared extensively. Nickel may be a shared neurotoxicant. In a mouse model,
inhalation exposure to 1 mg/m
3
nickel nanoparticles caused a rapid doubling of brain Ab40 and
42 (this high level is within OSHA limits) (142). However, the content of iron, nickel, or other metals
in nPM and sPM did not correlate with the induction of NF-kB, nitric oxide, or lipid membrane
oxidation in cell-based assays, or in the brain with microglial activation, proinflammatory
cytokines, or Ab42 (143).
Besides the total metal content, we must also consider the distribution of particle surfaces
which contact cells and fluids. The surface chemistry of the particles can largely determine the
adsorption of opsonizing proteins such as phospholipids and acute phase proteins in broncho-
alveolar lavage fluids (144). Surface chemistry also alters PM distribution in different organs,
biotransformation, and clearance of the particles (145). Non-organic PM0.2 vary widely in surface
reactivity, e.g., SiO2 (quartz) and black carbon, assayed as inflammatory cell responses in lungs

60
(146, 147). The bioactivity of metals depends upon their surface availability, rather than total
content per particle, which is not equivalent to their molarity in solution. For example, NF-kB
induction in THP1 monocytes was dependent on the surface iron of natural silica particles, while
equal amounts of iron in solution failed to activate NF-kB (113). This might explain the broad
similarity of nPM and sPM activity despite having 5-fold less transition metals. Iron and other
metals on particle surfaces, while a minute fraction of the total, could contribute to the in vivo
neurotoxicicity of nPM if it is upon the surface of particles.
The toxicity of PM is sensitive to aerosol acidity and consequently metal solubility (148). For
example, the sulfate content of the PM can produce acidic aerosols, dissolve primary transition
metals, and increase the oxidative potential of PM (148). Future studies should consider the
acidity of nPM, sPM, and other PM samples collected at different sites or by different methods.
Our study presents the first direct neurotoxic comparison of two kinds of ambient PM with
large compositional differences. While the responses of nPM and sPM largely overlapped, there
were some differences. sPM caused higher NF-kB response in glial and THP1 cells, induced
nuclear localization of NF-kB/P65, and higher downregulation of TLR4 mRNA in the cerebral
cortex of adult animals. Larger sPM response was stronger in vitro, which may be attributed to
direct physical exposure to PM at shorted duration. Glutamatergic effects of PM after an adult or
gestational exposure differed between nPM and sPM: GluA1 mRNA was only decreased in nPM
exposed adults; several glutamatergic genes were only downregulated in animals gestationally
exposed to sPM. Fat gain and also an increase of glucose tolerance were unique to males that
were prenatally exposed to nPM.
The present data do not allow identification of the chemical classes underlying the different
responses to nPM and sPM. While nPM samples are depleted in PAHs and several transition
metals, they are enriched in non-metals, alkali and alkaline earth metals (Fig. 1B-C). The unique

61
responses of nPM can be due to lack of PAHs, or enrichment of specific classes of other
chemicals. For example, depending on the route of the exposure, endpoint, and dosage, PAHs
can be potent immune/inflammation activators or suppressors (149-153). Thus, depletion in PAHs
could increase the inflammatory responses of nPM that lead to a higher fat gain and increase of
glucose tolerance. On the other hand, gestational exposure to specific PAHs, e.g.,
benzo(a)pyrene at high dose altered hippocampal glutamatergic function and receptors such as
Glun1 (124-126). The similar decrease of Glun1 from gestational exposure to sPM might indicate
PAH contribution.
Because endotoxins induce inflammatory responses, we titrated endotoxin activity with
polymyxin B, a polycationic antibiotic that binds to diverse bacterial lipopolysaccharides.
Polymyxin B attenuated NF-kB induction in THP-1 monocytes for nPM, but did not alter sPM
activity. This suggests that nPM and sPM have different endotoxin compositions, which is
consistent with the greater density and diversity of viable bacteria in sPM than nPM on four growth
media. The binding stoichiometry of polymyxin differs widely between LPS chemotypes by length
and chemistry of sugar chains on the bacterial outer walls (154). Gram-negative bacteria have an
additional target of polymyxin B in their secreted outer membrane vesicles which enhance LPS
inflammatory activity (155). Because the THP-1 cell assay includes antibiotics, the greater
inflammatory activity of sPM in vitro could represent its higher levels of iron and other transition
metals, and potential interactions with endotoxins.
AirP neurotoxicity may be mediated by direct impact of particles that reach the brain by
inhalation or ingestion, and indirectly by systemic responses. Thus, we do not know the pathways
leading to elevated brain levels of copper, iron-rich magnetite, and zinc in highly polluted cities
(156, 157). While some PM may be transported directly into the brain by olfactory neurons (52,
158, 159), the bulk of inhaled particles and gases are received by the lung (160) . The ‘lung-to-

62
brain’ path may include some direct transfer of PM to the brain by the circulation, but also includes
systemic inflammatory responses (101, 161). A systemic effect of AirP on the developing brain
seems likely because most PM is trapped by maternal lungs and placenta (162). Other systemic
effects are due to inhalation of nitrogen dioxide (NO2), ozone (O3), and sulfur dioxide (SO2), which
are rapidly quenched in contact with respiratory tract fluids yet cause neurotoxicity in the brain.
Some of these responses include brain glutamatergic responses (e.g., decrease of hippocampal
GluA1, GluA2, and Grin2a proteins) and activated MAPK signaling by NO2 (163), decrease of
mitochondrial activity in cerebellum and increase of Ab by O3 (164, 165), and a dose-response
increase of iNOS, Cox2, and ICAM1 proteins in the brain by SO2 (166). Thus, indirect effects of
lung-to-brain by circulatory factors may underlie the similar neurotoxicity of exposure to gases
and PM (101).
In sum, the current study demonstrated that nPM neurotoxicity highly overlapped with sPM
during short term exposure in adulthood or development. A critical unknown is the amount of
inhaled PM and its subcomponents that reaches the brain, PM surface chemistry, and particle
structure, which may vary widely by PM sources. Further analysis might consider the use of
synthetic nanoparticles containing characterized surface components and longer exposure times.

Acknowledgment
We thank Dr. Steven E. Finkel and Dr Christopher H. Corzett for providing the bacterial
growth media and advising on the microbial screen.

Author contribution
Design research: A.H., T.E.M., C.E.F., N.S.; Performed research: Adult exposure, A.H., N.S.;
Gestational exposure, R.J., N.W., N.S., A.H. Methodology, A.H., R.J., N.S., H.J.Z., M.T., N.W.,

63
V.C.; nPM collection and characterization, A.H.M., F.S., C.S. J.W. Jr. Analyzed data: A.H. Writing:
A.H., C.E.F., T.E.M, R.J., M.T., H.J.Z., H.F. Supervision: Project Administration, and Funding
Acquisition: C.E.F.

Funding
We are grateful for support by the National Institute on Aging: CEF (R01-AG051521, P50-
AG05142, P01-AG055367); HJF (ES023864); AH (T32- AG052374).

Conflict of interest
The authors have no conflicts of interest to declare.

64
Supplementary data

Table S1. Neurodevelopmental effects of gestational air pollution exposure in rodent models
Prenatal
PM
Exposure Concentration Gestation
exposure
Animal Litter
size
Effects
(130) PM2.5
urban
China
Tracheal 7.5-103 µg every 3 d Mouse
Kunming
Adult: spatial memory
deficits, mitochondrial
damage in hippocampus,
decreased neurogenesis,
increased apoptosis &
inflammatory cytokines

(131) PM2.5
urban
China
(water
dispersed,
nPM2.5)

tracheal 3.75,7.5,15
mg/kg
every 3 d Mouse
ICR
None Adult: increased IL1, IL6,
TNF in serum &
hippocampus; lower
hippocampal BNDF,
CREB,TrkB

(132) PM0.2
urban
US
inhalation 93 µg/m
3
4 hr/d
E0.5-E16
Mouse
B6C3F1
None Postnatal d11-15:
ventriculomegaly,
enlarged corpus callosum
(CC), smaller
hippocampus, microglial
activation, iron deposits in
female CC.

(122) nPM,
urban US
inhalation 350 µg/m
3
5hr/d, 3d/wk,
7 wk before
gestation to
2 d pre-birth

Mouse
C57BL/6
None Adult: tail suspension
depression index in males;
no change in open field or
novel object.

(96)

(133)

(33)

nPM,
urban US

nPM,
urban US

diesel
exhaust
particles

inhalation

inhalation

oropharyngeal

350 µg/m
3

350 µg/m
3

50 µg

ibid,
gestation to
postnatal 25
wk

ibid,
gestation to
postnatal 25
wk

every 3 d

Rat,
Sprague-
Dawley

Mouse
C57BL/6

Mouse
C57BL/6

None

Adult: impaired
neurogenesis (-70%),
Blood-brain barrier ZO1,
2-fold more microbleeds

Adult: male excess body
wt, impaired glucose
tolerance, hypothalamic
neuropeptides

Adult: cognitive & anxiety
impairment, high brain
cytokines

(32) diesel
exhaust
particles
inhalation 2 mg/m
3
4 hr/d
E9-E17
Mouse
C57BL/6
None E18: high cytokines, Adult:
excess weight on fat diet

65
Table S2: Primers for qPCR

Gene Forward Reverse
NOS2 (iNOS) GTCTTGGTGAAAGTGGTGTT GTGCTTGCCTTATACTGGTC
IL6 TGCCTTCTTGGGACTGATGCT GCATCCATCATTTCTTTGTAT
IL1B CTAAAGTATGGGCTGGACTG GGCTCTCTTTGAACAGAATG
TNFa CGTCAGCCGATTTGCTATCT CGGACTCCGCAAAGTCTAAG
GAPDH CCAATGTGTCCGTCGTGGATCT GTTGAAGTCGCAGGAGACAACC
Cox-2 CCCCCACAGTCAAAGACACT CTCATCACCCCACTCAGGAT
NFkB1 CCAGAAGAGGGTGTCAGAGC ACATTTGCCCAGTTCCGTAG
TLR4 AGCCGGAAGGTTATTGTGGT CAGCAGGGACTTCTCAACCT
MyD88 CTGCTACTGCCCCAACGATA ATTAGCTCGCTGGCAATGGA
GLT1 GCCTTCCTGGATCTCATTAG GCCCTTCTTGATAACGATCT
GLAST TTGAACAGACCTTAATCGCA ACAAAGTGAAGTTTCCCAGT
GLS GCGGGCGACAATAAAATAAA TTTCAACCTGGGATCAGATG
Glul TGCCTGTTTGTTTCAATGTC GGTGGGGGTGTGTTTATTAT
GluA2 GCCAAACATTGTGGATTCAA CTCCAACCATACCATTCCAA
GluA3 CAAAGCCATTTATGAGCCTG AGACAATGCACATCCAGATT
GluA4 GTAATGGCTGAAACTTTCCG CAGTGTTCTCTCCATGTCAA
GluN1 TACCCGAATGTCCATCTACT CTGACCAGCAGGATGATATG
NMDA2A(Grin2a) ATGACTATTCTCCGCCTTTC TCCAACTTCCCAGTTTTCAA
NMDA2B
(Grin2b) AGTAAGGCTGAAGGGAAATG CCTCTCTCACCCTACTGTAA
NMDA2C
(Grin2c) ATAACGAGAAGAACGAGGTG GAAGTTTCCAGTAGACCAGG
NMDA2D
(Grin2d) CTTCTTTGCCGTCATCTTTC GATATTCTTCTCCGTGGACC
NMDA3A
(Grin3a) ATCAATACTGCAAGAAGCCA GTACATAGTCCTCAGCACTG
NMDA3B
(Grin3b) ATTTGTGTTCACCAGGGAAT CAGTAGCCATAACAGCATCT

66
Table S3. Microbial load in nPM and sPM, as % area of plate surface covered by
colonies

Growth media Sterile
water
nPM
200 µg/ml
sPM
50
µg/ml
Potential species of bacterial growth by
medium in 72 hr
BBL Brain Heart
Infusion
0 11% 37% Nonenterococcal group D streptococci,
CLSI, Escherichia coli, Staphylococcus
aureus, Pseudomonas aeruginosa,
Enterococcus faecalis, Streptococcus
pyogenes,

MacConkey Agar 0 0 43 Campylobacter, Enterobacteriaceae,
Enterococcus, Escherichia coli and
Coliforms, Proteus, Salmonella,
Shigella, Staphylococcus.
BD Difco™ LB
Broth
0 1.7 29 E. coli

Difco R2A

0

1.2

40

Enterococcus faecalis, E. coli,
Pseudomonas aerginosa, Staphylococcus
aureus.

We did not specifically examine fungal growth.

Figure S1. Size distribution of re-aerosolized particles for mouse exposure.
0 200 400 600 800 51
57
0
200000
400000
600000
PM diameter (nm)
Particle frequency
Size distribution of reaerosolized PM
sPM
nPM
Mode diameter

67
Mitochondrial reductase activity:
By MTT assay, nPM and sPM have similar mitochondrial reductase activity. The data
indicate that at concentrations higher than 10 µg/ml, nPM0.2 samples might bea little
more active compared to sPM. Based on this result, 5µg/ml concentration was chosen
for further mechanistic experiments.

Figure S2. Dose response MTT analysis of nPM0.2, sPM0.2 and sPM2.5 in BV2 and
mixed glial culture. Concentrations above 10 µg/ml only were tested in sPM2.5 and
nPM0.2 due to limitation in collecting sPM0.2 with concentration higher than 10 µg/ml.
Adjusted P-values: <0.05 (*), <0.01(**), <0.001(***), <0.0001(****).
Cell viability at 6 hrs post treatment:
CellTiter-Glo Luminescent cell viability assay showed that there is no cell death at 5
ug/ml treatment of BV2 cells with either of nPM and sPM samples during first 6 hrs. In
contrast, the viability of the cells treated with sPM slightly (<10%) increased compared to
control group.

0 10 20 30 40
0
50
100
150
Mitochondrial reductase activity %
BV2 24 hpt
**
Concentraiton (µg/ml)
*
0 1 5 10
0
50
100
150
Mitochondrial reductase activity %
Mixed Glial Culture 24 hpt
nPM
sPM
Concentration (µg/ml)

68
A B

Figure S3. Particulate matter samples do not cause BV2 cell death at 5 µg/ml during
first 6 hrs of exposure. A) Efficacy of CellTiter-Glo assay in estimating BV2 viable cell
number. B) BV2 cells viability after 6 hrs treatment with 5 µg/ml of nPM and sPM. *
indicates adjusted p-value<0.05.

Figure S4. GC-MS analysis of organic compounds in nPM and sPM suspensions
collected on spring 2019.
0 50000 100000 150000 200000
0
5000
10000
15000
BV2 Cell number/well
Luminescent
BV2 cell number association with luminescent output
R
2
=0.99
0
50
100
150
% Cell Viability
Cell viability, BV2 cells
*
Control
nPM0.2 (5µg/ml)
sPM2.5 (5µg/ml)
6 hours post treatment
A B
nPM
sPM
0.000
0.002
0.004
0.006
0.008
ng/µg PM
Polycyclic aromatic hydrocarbons
Phenanthrene
Fluoranthene
Pyrene
Chrysene
Suspension
nPM
sPM
0.00
0.05
0.10
0.15
0.20
0.25
ng/µg PM
Organic acids
Suspension
decanoic Acid
dodecanoic Acid
succinic Acid
tetradecanoic Acid
octanoic Acid
octadecanoic Acid
hexadecanoic Acid
phthalic Acid
nPM
sPM
0.000
0.002
0.004
0.006
0.008
0.010
ng/µg PM
N-Alkanes
Suspension
hexatriacontane
n-Pentacosane
n-Heptacosane
n-Docosane
n-Tetracosane
nonacosane
n-Tricosane
hentriacontane
nPM
sPM
0.0000
0.0005
0.0010
0.0015
ng/µg PM
Hopanes and Steranes
Suspension
17A(H)-21B(H)-30-Norhopane
22S-Homohopane
17A(H)-21B(H)-Hopane
22S-Bishomohopane
22R-Bishomohopane

69
Animal weight
The mouse exposure was done in 3 weeks’ time course. Regardless of treatment,
the weight of the mice tends to increase during this time course. The average weight of
the all mice in the start of the experiment was 27.54(±2.02) gr.

Figure S5. Weights of the mice exposed to nPM0.2 and sPM2.5 in 4 weeks period.

Figure S6. Western blot analysis of GluA1 in the cerebral cortex of the exposed
animals.

0 7 14 21
90
95
100
105
110
115
120
Percentage weight / day 0
Control
nPM0.2 100ug/m3
nPM0.2 200ug/m3
nPM0.2 300ug/m3
Days
0 7 14 21
90
95
100
105
110
115
120
Percentage weight / day 0
Control
sPM0.2 100ug/m3
sPM0.2 200ug/m3
sPM0.2 300ug/m3
Days
0
0.5
1
1.5
PM concentration (µg/m
3
)
OD (Intensity relative control)
GluA1 protein
0 300 0 300

70
Chapter 4. Air pollution alters Caenorhabditis elegans development
and lifespan: responses to traffic-related nanoparticulate matter (nPM)
(Adapted from Journal of Gerontology Series A: Biological Sciences, doi:
10.1093/gerona/glz063)

Authors: Amin Haghani
1#
, Hans M Dalton
1#
, Nikoo Safi
2
, Farimah Shirmohammadi
3
,
Constantinos Sioutas
3
, Todd E Morgan
1
, Caleb E Finch
1
, Sean P Curran
1*

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Department of Biomedical Sciences, Center for Bioinformatics and Genomics, Cedars-
Sinai Medical Center, Los Angeles, CA.
3, Viterbi School of Engineering, University of Southern California, Los Angeles, CA.
# Co-first/equal authorship
*Corresponding author. Email: spcurran@usc.edu

71
Abstract
Air pollution is a heterogeneous environmental toxicant that impacts humans
throughout their life. We introduce Caenorhabditis elegans as a valuable air pollution
model with its short lifespan, medium-throughput capabilities, and highly conserved
biological pathways that impact healthspan. We exposed developmental and adult life
stages of C. elegans to airborne nano-sized particulate matter (nPM) produced by traffic
emissions and measured biological and molecular endpoints that changed in response.
Acute nPM did not cause lethality in C. elegans, but short-term exposure during larval
stage 1 caused delayed development. Gene expression responses to nPM exposure
overlapped with responses of mouse and cell culture models of nPM exposure in previous
studies. We showed further that the skn-1/Nrf2 antioxidant response has a role in the
development and hormetic effects of nPM. This study introduces the worm as a new
resource and complementary model for mouse and cultured cell systems to study air
pollution toxicity across the lifespan.

Keywords: Caenorhabditis elegans, air pollution, nPM, skn-1, development

72
1. Introduction
Epidemiological studies show that air pollution is associated with multiple chronic
health hazards of older age including Alzheimer’s disease (AD), ischemic heart disease
and stroke, lung cancer, and chronic obstructive pulmonary diseases - all of which
decrease life expectancy (4, 38, 167-169). Understanding the underlying mechanisms
between air pollution and these diseases requires modeling both air pollution and the
resulting biological responses.
Traffic related air pollution (TRAP) particles are a complex environmental toxicant
consisting of a variety of inflammogens and toxicants derived from vast heterogeneous
sources. It is estimated that ambient particulate matter (PM <2.5 micron diameter)
contributes to 3-4 million deaths (around 7.6% of total global death) and 103.1 million
(4.2%) global disability-adjusted life year annually (4). The mortality number is apart from
3.5 million deaths attributed to indoor air pollution such as solid fuels for cooking and
heating (2). It is essential that gerontological research understand the effects of
environmental toxicants over organismal lifespan, the association with common aging
phenotypes, and also identify the biomarkers that can predict these effects (reviewed in
(170)).
The nanosized subfraction of TRAP (nPM) has consistent toxic effects in rodent
and cell models (171). Biochemical and cell assays of air pollution toxicity, while widely
used, are not good predictors of in vivo responses for multicellular organisms (e.g.
Dithiothreitol [DTT], ascorbic acid [AA]-glutathione [GSH], and MTT (52). Exposure of
mouse models to reaerosolized nPM results diverse systemic and organ-specific
localized effects that involve different biological networks such as oxidative stress and

73
antioxidant responses, innate immunity, and the nervous system (13, 40, 114, 122).
These responses were dependent on the dosage of PM samples and the developmental
stage of the exposure. Mouse models are a valuable tool to study air pollution toxicity, but
they are limited as a biological model; low reproduction yield, long lifespan, expense, and
ethical considerations in these animals can reduce the feasibility of this model and also
the statistical power of any experiments.
Caenorhabditis elegans is a valuable model for TRAP toxicology with a potential
for much higher throughput than rodents (172). Humans and worms share several basic
physiological and stress response processes with homologues in most human genes (60-
80%), including multiple signal transduction pathways (173). Easy maintenance, large
scale production, small size, body transparency, full genomic characterization, complete
cell lineage map, and mutant libraries make the worm an ideal model for gene network
and environment interactions (174). While there are limitations such as lack of specific
organ equivalents, a smaller immune response system, and large differences in overall
lifespan, C. elegans allows medium-throughput whole organism-level assays with
multiple end points (e.g. development, reproduction, feeding, lifespan, locomotion) (173).
Worms have been used to assess the toxicity of terrestrial environmental samples (e.g.
soils, sludge, river sediments, (175, 176), pesticides (e.g. Glyphosate, Paraquat,
Endosulfan and Dichlorvos for neurotoxicity, DNA damage, sterility and embryonic
lethality) (177), metal toxicity (e.g. Ag, Cd, Pb, Fe), lifespan, fertility, growth (178),
nanoparticles (179), drugs (180), toxins (e.g. nicotine) (181-183), as well as other
bioreactive molecules including NaAsO2, NaF, caffeine and DMSO (184). While airborne
bacteria have been tested in C. elegans (175), its responses to TRAP have not been

74
studied. Considering the usefulness of C. elegans in diverse toxicology models, this study
introduces C. elegans as a multicellular model organism for air pollution toxicity.
Our prior mouse and cell culture studies show that nPM can induce oxidative
stress, systemic inflammation, and selective neuroinflammation in different brain regions
as well as exacerbate amyloidogenesis processes in AD mouse models (13, 38, 40, 114,
122). We also noticed that there is an interaction between age and air pollution mediated
toxicity (40). Here we examined different nPM dosages, as well as developmental and
lifespan effects in C. elegans to evaluate the similarities of this model to higher organisms
in several cell survival pathways and Alzheimer Ab-related genes. In contemplating the
response of cytoprotective transcription factor SKN-1/Nrf2 in C. elegans to nPM, we also
investigated the association of SKN-1 in an nPM-mediated developmental delay
phenotype as well as changes in lifespan.

2. Method
2.1. C. elegans strain maintenance
C. elegans were maintained at 20°C unless otherwise noted. Strains used were
Bristol N2 (wild type), LG335 skn-1(zu135)/nT1[qIs51(myo-2::GFP;pes-
10::GFP;F22B7.9::GFP)], and CL2166 dvIs19 [(pAF15)gst-4p::GFP::NLS] III. Some
strains were provided by the CGC, funded by NIH Office of Research Infrastructure
Programs (P40 OD010440). LG335 was a gift from the Leonard Guarente laboratory. E.
coli strain OP50 was used for all non-RNAi experiments and for general C. elegans
maintenance. For synchronization prior to plating, gravid adults were treated with a

75
solution of bleach and hypochlorite to harvest eggs; then, eggs were washed and rocked
overnight in M9 solution, allowing the hatching and L1 arrest/synchronization (185).
2.2. Airborne nano-sized particle collection:
Ambient nano-sized particles (diameter<0.18 µm) were collected on 8x10 inch
commercially available Zeflour PTFE filters (Pall Life Sciences, Ann Arbor, MI) using a
High-Volume Ultrafine Particle (HVUP) Sampler (186) operating at a sampling flow rate
of 400 liters/min flow rate at the Particle Instrumentation Unit (PIU) of University of
Southern California located within 150m downwind of a major freeway (I-110).
Gravimetric mass (nPM mass concentration) was determined from pre- and post-
weighing the filters under controlled temperature (22–24 ºC) and relative humidity (40–
50%) conditions. The filter-deposited nPM was eluted by sonication into ultrapure
deionized (milli-Q) water (13) providing the concentrated slurry suspension used for these
exposures. A portion of the aqueous suspension was chemically characterized. After acid
digestion, samples were analyzed by high resolution inductively coupled plasma sector
field mass spectrometry (SF-ICPMS). Another portion was analyzed using a Sievers 900
Total Organic Carbon Analyzer to determine total organic carbon (TOC) content (Fig. S1).
This characterized suspension was used in all the experiments of this study.
2.3. RNA interference
The E. coli strain HT115 (DE3), harboring either the empty L4440 plasmid ("Control
RNAi") or the skn-1 RNAi plasmid (dsRNA production of skn-1 sequence - Ahringer
Library), was grown 16-18hrs at 37
o
C overnight. Cultures were seeded onto RNAi plates
(normal NGM plates with 5mM isopropyl-β-D-thiogalactoside (IPTG) and 50 µg/ml

76
carbenicillin) and left overnight to generate dsRNA for experiments (maintained at 20
o
C
during and after dsRNA generation). To optimize RNAi of skn-1 in offspring, P0 worms
were plated on bacteria expressing skn-1 RNAi for 12, 18, 24, or 48 hrs (Fig. S2); this
was done to further reduce maternally-deposited skn-1 mRNA transcripts as well as to
deposit skn-1 RNAi in the F1 generation prior to hatching. To control developmental and
RNAi timing, these adults were placed in 15°C for the duration of skn-1 RNAi exposure.
Adult exposure for 24 or 48 hr caused 90-100% of dead F1 eggs. 18 hrs was chosen as
an acceptable 20-50% egg death while decreasing skn-1 transcripts in living animals to
~70% (Fig. S2).
2.4. Optimization of air pollution exposure model
Two routes of nPM exposure (liquid or chronic exposure on growth medium plates
[chronic exposure data not shown]) and duration (1, 2, 4, 8, 24 hours) were tested in larval
stage 1 or 4 (L1 or L4) C. elegans, as well as solvents (M9 buffer or K medium). 1 hr
exposure in diluted nPM at different dosages (1-200 µg nPM/ml) with M9 was chosen for
further experiments as it was the fastest exposure in inducing phenotypes without
compromising developmental timing. For treatments, worms were washed into an
Eppendorf tube, brought to a known volume, and nPM was added to each tube to achieve
the listed concentration. Worms were incubated at 20°C for all experiments, unless
otherwise indicated. Worms were gently rocked 1 hr for even distribution of nPM. After
exposure, worms were washed once before plating at time 0 in post exposure time.
Worms treated as "L1s" were treated immediately at the synchronized developmental
stage (prior to feeding).

77
2.5. Size analysis
Following exposure of L1 stage for 1 hr to nPM and re-plating, worms were incubated
at 20 °C for 72 hrs. Body size was analyzed by area using ImageJ (average of width ´
length).
2.6. Pharyngeal Pumping:
Worms were treated with 200 µg/ml nPM for 1 hr at L1 stage, then allowed to recover
for 24hrs on regular agar food plates described above (~100-150 worms/plate). These
plates were then directly placed under a 10x objective, allowed to settle for 1 minute to
prevent changes in pumping due to plate movement/vibrations, and individual worms
were then followed and recorded using the Movie Recorder in the ZEN 2 software at 6-
8ms exposure (Zeiss Axio Imager) for 10-15 seconds (making sure to keep the pharynx
in frame). Slowed-down (~4x) movies were later analyzed for pharyngeal pumping rate.
Worms without pharyngeal pumping (dead, lethargus) during recording were excluded
(estimated <5% of total worms observed).
2.7. Fluorescent analysis of gst-4p::GFP labeled C. elegans:
gst-4p::GFP animals were treated with the listed nPM concentration (0-200 µg/ml) for
1 hr and then imaged 4 or 24 hrs after exposure. Recovered worms were then plated into
10-12 ul droplets of M9 solution directly onto slides and 10mM sodium azide was added.
Once animals were paralyzed, cover slips were slowly placed at a diagonal angle on top
of the M9-immersed animals to prevent bursting/flattening. Slides were then immediately
taken to image whole worms using an oil-based 40x objective (Numerical aperture 1.4)
with DIC and GFP filters (Zeiss Axio Imager, an epifluorescence microscope). Animals

78
were imaged to have as much of the pharynx, germline, and intestine in focus as possible.
As prolonged exposure to sodium azide can eventually induce gst-4:GFP (187), each
plate was imaged within 20-30 minutes, and a new plate was made after this time if more
worms were needed - both done to avoid inflation of the GFP signal. Later, fluorescence
was measured (the "measure" command) on the imaged worms via ImageJ using the
polygon selection tool surrounding both the whole animal and a section containing no
worm(s) next to the animal (within the same image) as background (Whole worm
autofluorescence background was not considered in the analysis). Corrected total cell
fluorescence (CTCF) was then determined using Microsoft Excel. CTCF = Integrated
Density – (Area of selected cell X Mean fluorescence of background readings).
2.8. Lifespan
C. elegans were treated with 50 µg/ml nPM at either L1 stage or at L1 and Day 2
adult (~96 hrs). Synchonized L1 animals (after nPM exposure) were plated on control
RNAi or skn-1 RNAi. Worms were re-plated each day of reproduction (Day 1-5) to avoid
overcrowding from progeny. Individuals were checked each day for survival by prodding
with a platinum wire to verify touch responsivity. Individual worms were censored from
survival analysis for gross morbidity (bursting, vulval protrusion, crawling off plate, etc.).
2.9. Statistics
Size and gst-4p::GFP analysis results were compared using ANOVA followed by
Tukey post hoc with calculating multiplicity adjusted P values by Graphpad Prism 7. In
qPCR time points and the experiments with only two groups, pairwise comparison was
done with T distribution tests in Graphpad Prism 7. Lifespan was analyzed by log-rank

79
and Wilcoxon tests in JMP software. Heatmap and dendrogram analysis was done in R
by the CompexHeatmap package of Bioconductor using Euclidean distance matrix and
complete linkage for clustering of the genes. Factorial models were made from size, gst-
4p::GFP, and skn-1 RNAi data to analyze the significance of interaction terms using R.
2.10. Quantitative real-time PCR:
L1 or L4 worms were exposed to 50 µg/ml nPM for 1 hr, and then worms were
collected and washed in M9 solution at 0 (immediately after 1hr exposure), 1, 2, 4, or 8
hrs after exposure. "Populations" in figure legends refer to biological replicates of 500-
1000 worms. Worms were pelleted at collection, 500 µl of TriZol reagent (Zymo) was
added. RNA was extracted using TriZol method (13) with an additional step of treatment
with DNase, lysis by T&C lysis and protein precipitation by MPC reagent (Epicenter, USA)
for purification. RNA concentration and purity were assessed by O.D. 230, 260, and 280
by spectrophotometry before qRT-PCR with gene specific primers in Table S1. The data
was normalized to ama-1 as housekeeping gene. For Figure 2, skn-1 RNAi was
administered to P0 adults, and F1 progeny were analyzed (as "RNA interference" above).

3. Results
3.1. Acute nPM exposure drives a reduction in C. elegans body size
Different exposure methods were tested to determine the acute lethal dosage 50
(LD50) of nPM samples. Larval stage 1 and 4 (L1 and L4, chosen to represent near the
beginning and end of development) wild type N2-Bristol worms (~100 per replicate; at
least 3 replicates/group) were exposed to a range of nPM concentrations (0.5-200 µg/ml)

80
for 1, 4, or 24 hrs in liquid culture or on growth medium plates. C. elegans did not show
any acute mortality at the highest dose in any of tested conditions during 3 days after
exposure.
However, nPM exposure did affect development in a dose-dependent manner (Fig.
1A). Exposure of L1 stage animals to nPM in M9 buffer (10, 50 and 200 µg/ml) for only 1
hr showed dose dependent size reduction (13-26% of the worm’s area) in day 1. This size
difference was not due to lack of food intake because pharyngeal pumping rates were not
impaired (Fig. 1B). Since nPM at 50 µg/ml caused the same size reduction as 200 µg/ml,
the lower dose was used for further experiments.

Figure 1. Acute nPM exposure of L1 worms results in adult worm size
reduction. A) Dose dependent size changes (mean±SEM) of Day 1 adult animal body
area following 1 hr exposure to nPM in L1 stage (n=3*15 to 22 /group). The average area
of the control group was 0.06 mm
2
with average length of 0.97 mm. B) Pharyngeal
pumping rate of animals (n=3*~7/group) analyzed at 24 hrs post-exposure of the L1 stage
wildtype animals to 200 µg/ml nPM. Statistical tests include ANOVA followed by Tukey

81
post hoc with correction for multiple statistical hypothesis testing for size and t test for
pharyngeal pumping. Adjusted p-values: <0.05 (*), <0.01(**), <0.001(***), <0.0001(****).

3.2. skn-1 is activated in response to nPM exposure
Given the size impairment under nPM exposure, we were interested in the
underpinning biological response - including any defense mechanisms that were
activated. TRAP nPM is a heterogeneous environmental toxicant consisting of multiple
elements and water-soluble organic carbons with high oxidant activity (Fig. S1). We chose
to examine SKN-1/Nrf2, a cytoprotective transcription factor, due to its established anti-
oxidant defense response (188) and role in development (189); specifically, we used the
used the reporter gst-4p::GFP (GFP linked to Glutathione S-transferase, a target of SKN-
1) as an estimate of activation of SKN-1 in animals treated with nPM. In the gst-4p::GFP
strain, both L1 and L4 had dose dependent increases in response as early as 4 hr post-
exposure to nPM. By 18 hr post-exposure of L1 stage and 24 hr post-exposure of L4
stage, we observed reduction of GFP signal in most of the nPM exposed animals (all
groups except for 10 µg/ml nPM exposed L1s) (Fig. 2A, Fig. S3) suggesting a rapid return
of reversal of the gene expression changes (perhaps compensatorily) post-SKN-1
activation.
To determine the role of SKN-1 activation in response to nPM stress, we down-
regulated SKN-1 by RNAi before exposure to nPM. First, we tested the effects of short
term nPM exposure on lifespan at different life stages to try to understand if certain
exposure windows were more important for any lifespan effects. Contrary to the negative
survival effects observed in cell lines (52), dual nPM exposure (50 µg/ml) during

82
development (L1 and L3), or triple nPM exposure (50 µg/ml) during reproduction stage
(D1, D3 and D5) had a modest increase of the lifespan. Increase of exposure intervals by
including of both developmental and reproduction stages removed this lifespan
difference. We then examined if dual nPM exposure (50 µg/ml) would have the same
outcome in skn-1 knockdown worms. While worms fed with control RNAi had a modest
increase in mean lifespan (1.1-day, p=0.015 Log rank test, analysis in 90% of total
lifespan), reducing skn-1 expression by RNAi ablated this increase in lifespan (Fig 2B,
Table S2). In general, skn-1 knockdown lead to decrease of total lifespan regardless of
nPM exposure (Fig S4). While our confidence in the observed lifespan change is limited
by the number of animals used (n=200-300 per replicate), we are confident in our finding
that the high nPM concentrations utilized do not negatively impact C. elegans lifespan.
We investigated skn-1 knockout (skn-1(zu135)) for its response to nPM. While L1
exposure of wild type worms to 50 µg/ml nPM reduced day 1 adult size by 20%, skn-
1(zu135) worms did not show this decrease in size (Fig. 2C). In both conditions, skn-
1(zu135) worms were smaller overall compared to wild type.

83

Figure 2. nPM mediated SKN-1 response underlies developmental and lifespan
effects of TRAP particulate matter. A) Dose response gst-4 protein responses to nPM
in L1 and L4 gst-4p::GFP strain (n=3´3 to 7/group). Representative images of L1 animals
are presented in front of the graph. B) Examining the interaction of different life stages on
nPM (50 µg/ml) mediated hormesis effects (n=3´ 52-101/group) and the role of skn-1
response in this process until 90% survival in wild type non-exposed animals (n=2-3´ 47-
53/group) (Summary statistics in Table S2). L: Larval stage, D: Day adult. Multivariate

84
model: GST-4 ~ nPM´Life stage´Time post exposure, F=21.36, adjusted R
2
=0.42,
p<0.0001****. (nPM):(Time) interaction, p=0.003**, b=-0.0002 C) Analysis of adult skn-
1(zu135) mutants and wild type worms exposed to nPM during development (n=3´13 to
27/group). Multivariate model: size ~ skn-1´nPM, F=37.72, adjusted R
2
=0.35,
p<0.0001****, skn-1:nPM interaction: p=0.03*, b=0.24. Statistical tests: Survival data was
analyzed by Log-Rank and Wilcoxon tests. Size: Univariate ANOVA followed by Tukey
post hoc with correction for multiple statistical hypothesis testing and also Multivariate
ANOVA for interaction was used for size and gst-4p::GFP results. Adjusted p-values:
<0.05 (*), <0.01(**), <0.001(***), <0.0001(****). Please refer to Table S3 for summary
statistics of multivariate models.
3.3. Gene expression changes induced by nPM exposure
To understand molecular responses to nPM, selected mRNAs were measured by
qPCR to analyze a panel of genes involved in C. elegans stress responses, development,
vitellogenesis, innate immunity, amyloid processing, and TGF-b signaling pathway in L1
or L4 (Fig. 3, S5-S8). In general, animals exposed at L1 had larger nPM mediated mRNA
changes, particularly in the first hours of exposure compared to L4 animals; moreover,
the genes with largest nPM responses in L1 stage did not change in L4 stage exposed
animals (e.g. gst-4, daf-2, apl-1, sel-12).
In L1 stage animals, 1 hr exposure to nPM immediately (0 hr post-exposure)
changed the expression of several genes (Fig. 3, S6-8). Some of the observed changes
at this time included heat shock responses (e.g. hsp-4 [expression/control=0.5 fold]),
hormone receptor and development (e.g. daf-2 [0.58], daf-12 [0.61]), vitellogenin (e.g. vit-

85
6 [0.64]) and Alzheimer amyloid processing genes (e.g. apl-1 [0.64], lrp-1 [0.68] and sel-
12 [0.55]) and TGF-b signaling pathway (e.g. daf-7 [0.53]). At 1 hr post-exposure, the
expression of most genes returned to baseline (pre-exposure) levels (exceptions were
hsf-1 [0.6], hsp-4 [0.47] and daf-2 [0.82]). This shift continued and lead to upregulation of
several genes at 2 hr post-exposure of L1s in response to nPM. These responses
consisted of skn-1 antioxidant target genes (e.g. gst-4 [3.5]), metal response genes (e.g.
cdr-1 [5.7]), innate immune response (e.g. abf-2 [2.8]) and amyloid processing genes (e.g.
sel-12 [2.9]). At 8 hr post-exposure, these changes returned to base line or were
decreased compared to controls. Metal-sensing genes (e.g. aip-1 [0.54]) and heat shock
responses (e.g. hsf-1 [0.46]) were among the genes with lower mRNA levels compared
to controls at 8 hr post-exposure (Fig. 3).
In L4 stage animals, nPM exposure did not significantly alter the expression of
most genes examined. However, the gene expression followed the trend observed in L1
exposed animals with increased expression at 1 hr post-exposure and return to baseline
at 4 hr post-exposure (Fig. 3). nPM caused decrease of tol-1 mRNA [-0.86] at this time.
These data suggest that transcriptional responses to nPM exposure are dependent on
developmental stage.

86

Figure 3. mRNA response to nPM varies with developmental stage. A) Heatmap
showing the mRNA changes of the selected genes in L1 and L4 stage animals exposed
to 50 µg/ml nPM for 1 hr (n=4-5 replicates, ~500 animals/replicate). RNA levels were
followed for 8 hr post-exposure in L1 stage animals and 4 hr post-exposure in L4 stage
animals. Responses are clustered based on Euclidian distance using complete linkage
method. The heatmap is annotated with the time point, treatment, life stage, average of

87
mRNA changes and the function of target genes. Significant changes are shown in the
heatmap. B) mRNA changes of Alzheimer homologue genes in C. elegans to nPM. The
t-test was used to compare nPM and controls at each time. p-values: <0.05 (*), <0.01(**),
<0.001(***), <0.0001(****).
3.4. The role of skn-1 transcriptional activity in response to nPM
exposure.
Given the responsiveness of gst-4 to nPM, we further studied the role of SKN-1
activity in nPM mediated toxicity. The expression of skn-1 and its downstream genes was
targeted by skn-1 RNAi at L1, which blocked gst-4 response in first 2 hours after nPM
exposure (Fig. 4); gst-4 was the only gene with significant nPM and the associated
interaction with skn-1 or time based on multivariate regression analysis. At 0 hr post-
exposure, all SKN-1 targets were lower (skn-1 [0.7], gst-4 [0.58], gcs-1 [0.47] and ugt-11
[0.25]) vs negative control RNAi. Control animals had significant gst-4 [1.56] mRNA
increase at 0 hr post-exposure to nPM, which was not observed for skn-1 RNAi. Contrary
to skn-1-related genes after RNAi, nPM still induced some innate immune responses, e.g.
abf-2 [0.7 nPM/control at 0 hr post-exposure]. Moreover skn-1 RNAi slightly increased
abf-2 response (+30%) at 2 hr post-exposure regardless of nPM. For the amyloid
processing genes, skn-1 RNAi decreased sel-12 (-40%) with no further response to nPM
at 0 hr post-exposure. At 2 hr post-exposure, nPM still decreased sel-12 mRNA in worms
fed Control RNAi (-20%), but not when fed skn-1 RNAi. Taken together with the gst-
4::GFP expression data, it appears that SKN-1 may have a role in mediating the
transcriptional responses to nPM exposure.

88

Figure 4. SKN-1 mediates some transcriptional responses of nPM. RNA
responses to 1 hr nPM (50 µg/ml nPM) after skn-1 RNAi (n=3-5 replicates, ~500

89
animals/replicate). Pairwise t-test was used to compare Control RNAi vehicle controls
with skn-1 RNAi or nPM vs controls at each time. Multivariate coefficients of the gst-4
expression regressed on nPM, skn-1 RNAi, time and the associated interactions: gst-4
was the only gene with significant nPM and nPM:skn-1 interaction. p-values: <0.05 (*),
<0.01(**), <0.001(***), <0.0001(****).
4. Discussion
This study introduces C. elegans as a valuable short-lived model with potential for
medium-throughput for genetic and toxicological studies of air pollution toxicity in humans.
Most notable are the transcriptomic responses to nPM, a toxic subfraction of the ultrafine
air pollution particulate matter. We identified several conserved genes that show
developmental sensitivity to nPM. While nPM does not cause lethality in adult C. elegans
up to 200 µg/ml, developmental exposure to nPM does reduce worm size and alter mRNA
levels, dependent on SKN-1. Despite the observed reduction in worm size under nPM
exposure, the time from which an animal hatches from an egg to laying its first egg is
roughly the same (egg to egg time) (Fig. S3). As such, development is not markedly
perturbed. Genes responsive to nPM include skn-1, the C. elegans homologue of the
mammalian NRF2, which has fundamental roles in air pollution toxicity (52). In addition,
50 µg/ml of nPM given twice early in the C. elegans life cycle caused a mild increase in
lifespan [1.2 days in mean lifespan], again dependent on SKN-1.
Although C. elegans lacks a respiratory system, exposing animals to cigarette
smoke in smoking chambers for 3hrs impaired intestinal bacterial clearance and gene
expression changes (190); we observe similar gene expression changes following nPM

90
exposure (e.g. tir-1: -3.24, and hso-16.2: 2.14 fold change). In the insect Drosophila,
chronic exposure to air pollution decreased lifespan by 50% (191), and a similar exposure
chamber was also used for exposing C. elegans to cigarette smoke (190). Future studies
using this methodology will allow comparisons between liquid and aerosolized delivery of
nPM. It is possible that suspended particles could limit the toxic effects of nPM, and
exposure to aerosolized particles should be considered for future experiments. Using
lifelong liquid cultures to maintain worms could also be used to increase exposure of nPM.
While most collected nPM has a cell cytotoxicity 50 (CC50) of 10-20 µg/ml (52),
our data show that there is no acute lethality to nPM even at concentrations as high as
200 µg/ml, which have no obvious counterpart in human exposure. This lack of toxicity at
high levels of nPM suggests that C. elegans may be highly resistant against nPM, through
a currently unknown mechanism. It is possible that molecular defense pathways, such as
those we observe, are important for this resistance, or perhaps its outer cuticle simply
provides a strong barrier against the toxicity. While 200µg/ml is an extremely high dose
compared to normal human exposure, it may be prudent in the future to test "mega-doses"
(e.g. >500µg/ml) to arrive at a true LD50 in C. elegans. However, our collection protocol
hits a technical limitation around 200µg/ml, and even with those concentrations, the
volume becomes limiting in that it is difficult to do many follow-up experiments using the
same nPM batch (as we have done here), requiring constant re-characterization of
different nPM batches. We are interested in attempting higher concentrations to find a
true LD50 if we find a way to make it technically feasible in the future.

91
Our study examined the role of skn-1/Nrf2 in the observed physiological and
molecular changes of nPM. SKN-1/Nrf2 is a transcription factor well-known for affecting
both development (189) and cytoprotection/detoxification (188), among other genes.
Exposure of young adult mice (3 mo), but not older adult mice (18 mo), to nPM induced
Nrf2-dependent phase II detoxifying enzymes such as GCLC and GCLM in lung, liver and
brain samples (47). These age differences further show the different susceptibility of life
stages to air pollution toxicity. Down-regulation of skn-1/Nrf2 in C. elegans did not cause
lethality in response to nPM. However, skn-1 knockout animals no longer had a reduction
in size under nPM exposure (albeit were smaller overall). It is possible this developmental
effect may involve sel-12, as sel-12 mRNA was decreased by skn-1 RNAi - similar to size
reduction in these animals. In view of SKN-1 targets, nPM specifically activated gst-4 with
minimal change in gcs-1 expression. Genome wide screening of the genes associated
with skn-1-mediated detoxifying responses showed involvement of alternative pathways
(e.g. apb-2 and csn-2) in activation of gcs-1, regardless of skn-1 activation (192). Thus
gcs-1 may be regulated independently of other phase 2 genes. It is worth noting that while
animals lacking skn-1 mRNA transcripts experience embryonic lethality, it is unclear
whether progeny that do survive the transgenerational RNAi exposure are "sicker" than
wild type, or if their survival indicates that enough skn-1 mRNA was around to reach a
"survival threshold" where they are healthy enough (observationally, they do not seem
"sick"). In addition, while we focused on the effects of skn-1, it is likely that other defense-
associated transcription factors could be implicated in these gene expression changes or
otherwise. While each individual experiment on skn-1-nPM interactions yielded modest,

92
albeit significant changes, in sum, our data strongly suggests a role for skn-1 under nPM
exposure.
We must consider if C. elegans is an appropriate model for human air pollution
toxicity, because humans have more complex circulatory and immune systems and 100-
fold longer lifespans. Our previous studies of air pollution toxicity in mouse and cell culture
models nonetheless show overlap in oxidative stress, inflammation and amyloidogenesis
pathways with these C. elegans findings (38). This study investigates our previous
findings of air pollution toxicity based on prior mouse and cell culture models. Specifically,
we used TRAP ultrafine particulate matter (<0.2 µm dia.) in this study. UltrafinePM may
be more toxic compared to PM2.5 (193, 194), but is not currently regulated or monitored
by the United States Environmental Protection Agency.
Air pollution and oxidative stress are strongly associated (195-197). Previously,
we showed only 5 hr exposure to air pollution sufficed for oxidative damage of membrane
lipids in olfactory epithelium of exposed mice, assayed as 4-HNE (114). Oxidative stress
(e.g. gst-4) was a prominent response to nPM. The other consistent air pollution response
in cell cultures, mice, and humans is inflammation. Microarray analysis of the primary
mixed glial culture responses to nPM showed MyD88 dependent activation of TLR4,
suggesting this pathway as an important upstream sensors of air pollution leading to
inflammation (45). Exposed mice had higher levels of brain TNFa depending on the
duration of nPM exposure (13, 171). Humans exposed to diesel exhaust also show rapid
systemic inflammatory responses (198). Notably, TLR-associated gene homologues (tol-
1: TLR4, abf-2: downstream of tol-1) were among the earliest responses of C. elegans to
nPM.

93
Air pollution is recently recognized as an environmental risk factor for Alzheimer’s
disease (AD) and accelerated cognitive decline (38, 199). In mouse and cell models,
exposure to nPM increased production of the Ab peptide (38). C. elegans also shows
similar effects of air pollution in the responsiveness of its homologous amyloid processing
genes to nPM. Several amyloid-related genes are associated with C. elegans
development; for example, inactivation of apl-1/APP results in penetrant lethality during
the L1 to L2 transition due to molting defects (200). On the other hand, overexpression of
apl-1/APP results in penetrant L1 lethality, shortened body length and morphological,
locomotive, and reproductive effects (200, 201). sel-12/PSEN is one of the genes
regulating apl-1/APP cleavage and trafficking (200). The expression of both apl-1/APP [0
hr: 0.64] and sel-12/PSEN [0 hr: 0.55, 2 hrs: 2.88] is significantly changed in L1 stage
following nPM exposure, suggesting the importance of apl-1 and sel-12 expression for
the developmental responses to nPM. In addition, air pollution is linked with
developmental changes in humans; for example, childhood obesity is associated with
maternal exposure to ambient air polycyclic aromatic hydrocarbons during pregnancy
(202). Moreover, we showed that prenatal exposure of mice to nPM alters neuronal
differentiation and depression-like responses (122). Lack of sel-12 lead to elevated
endoplasmic reticulum (ER) mitochondrial Ca
2+
signaling and oxidative stress due to
mitochondrial superoxide production (203). Thus, sel-12/PSEN might be a part of
antioxidant response to nPM.
We initially hypothesized exposure to nPM might be fatal for C. elegans, especially
during development. Instead, we found that short term exposure to air pollution could
cause a hormetic increase of lifespan. This effect is apparently dose dependent, since

94
exposing the worms for multiple times during development and reproduction ablated this
effect of air pollution. Hormesis, wherein small doses of a toxin or stress can induce
increases in health and lifespan, is well established in eukaryotes (204), including C.
elegans (205). While modest, our data suggest the hormetic effect of air pollution might
be dependent on a functional skn-1 response, indicating its importance on later lifespan
in addition to its importance in physical development. If exposure to nPM is administered
more chronically, perhaps by air exposure chambers, longer or more constant doses may
eventually cause decreases in lifespan.
Future studies will further identify signaling pathways in responses to air pollution
and hopefully help to identify drug and/or diet intervention strategies to counteract these
ambient toxins. Systematic study of the life stages of C. elegans will identify critical
periods of vulnerability to acute and chronic air pollution exposure. C. elegans could
productively complement rodent models for studies of air pollution toxicity throughout the
life cycle.

Acknowledgements
We thank the CGC, funded by NIH Office of Research Infrastructure Programs (P40
OD010440) for some strains, and WormBase. This work was supported by NIH grants
T32AG052374 (A.H.); R01AG051521 (C.E.F.); R21AG05020 (C.E.F.); Cure Alzheimer’s
Fund, (C.E.F.); R01GM109028 (S.P.C.), F31AG051382 (H.M.D.) and T32AG000037
(H.M.D).

Author contributions

95
Conceptualization, S.P.C., C.E.F., T.E.M.; Methodology, A.H., H.M.D. and S.P.C.;
nPM collection and characterization: F.S., C.S.; Investigation, A.H., H.M.D., N.S. and
S.P.C.; Writing, A.H., H.M.D. and S.P.C.; Supervision, Project Administration, and
Funding Acquisition, S.P.C. and C.E.F.

96
Supplementary data
Table S1. List of qPCR primers used.
snb-1 F CCGGATAAGACCATCTTGACG
snb-1 R GACGACTTCATCAACCTGAGC
ama-1 F CGGAGGAGATTAAACGCATG
ama-1 R CCAACTTTGGCTTTCCGTTC
hsf-1 F TTGACGACGACAAGCTTCCAGT
hsf-1 R AAAGCTTGCACCAGAATCATCCC
aip-1 F AAGCAAGAACAGAGGGAGATG
aip-1 R CACTAAATTGGATGCTATGAGAC
hsp-4 F CTAAGATCGAGATCGAGTCACTC
hsp-4 R GCTTCAATGTAGCACGGAAC
gcs-1 F CCAATCGATTCCTTTGGAGA
gcs-1 R TCGACAATGTTGAAGCAAGC
tir-1 F GGACAACTTCTTGATGGGAT
tir-1 R GGTTTCAAATGCTTGTGTCA
lrp-1 F CACCAAACAGACCATCAACG
lrp-1 R CTTCGAGATTTCCGCTTTTG
tol-1 F CCAAAGGTTCTCATTCAGGA
tol-1 R CCGTATTGACAGCAGATACA
apl-1 F TGGTGGAAACATCAGTACAA
apl-1 R ACTTCTGGTGATTGGATGAG
daf-2 F GCCCGAATGTTGTGAAAACT
daf-2 R CCAGTGCTTCTGAATCGTCA
cdr-1 F TCTTCTCTCAATTGGCAACTG
cdr-1 R TTTGGGTAAACTTCATGACGA
gst-4 F GATGCTCGTGCTCTTGCTG
gst-4 R CCGAATTGTTCTCCATCGAC
daf-7 F AAAGAGGCACCAAAGGGATT
daf-7 R TCAAACTTGGCAACAAGCTG
daf-12 F GAGGCAATGATTCCAAAGGA
daf-12 R CTTTAAGCTCAGCGGCATTC
abf-2 F TCGACTTTAGTACTTGTGCC
abf-2 R AGTGGAATATCTCCTCCTCC
vit-6 F CAATCAATGTTGAACCACGC
vit-6 R CTCCTCCATTTGTGGTTGGT
sel-12 F TCTGGAGTAAGGGTGGAACG
sel-12 R TGGCCACATAACAAGCGATA
skn-1 F CCACTTCAATCCCCACAAAG
skn-1 R CCGGGCTCAAATGAAAAAC
gcs-1 F CCAATCGATTCCTTTGGAGA
gcs-1 R TCGACAATGTTGAAGCAAGC
ugt-11 F CCGATTTCTGGGACTCTCAA
ugt-11 R GGACTCCCAGGAAGTGTGAC

97
Table S2. Summary of the animal numbers and replicated in the survival experiments
FOOD RNAI NPM STAGE OF
EXPOSURE
TOTAL
(REPLICATE
´ NUMBER
OF
ANIMALS)
NUMBER
OF
CENSORED
MEAN LIFE
TILL 90%
SURVIVAL
OF NON-
EXPOSED
CONTROLS
OF EACH
REPLICATE
OP50 - - 3´~79 88 12.81,
12.69, 12.05
OP50 - + L1, L3 3´~81 71 12.62,
12.59, 12.85
OP50 - + D1, D3, D5 3´~52 61 13.71,
12.59, 12.86
OP50 - + L1, L3, D1,
D3, D5
3´~101 100 11.84,
13.17, 12.09
HT115 Control
RNAi
- L1, D1 3´ ~53 33 14.04, 13.4,
14.29
HT115 Control
RNAi
+ L1, D1 2´~53 55 15.89, 14.34
HT115 skn-1
RNAi
- L1, D1 3´~47 24 15, 14.47,
14.19
HT115 skn-1
RNAi
+ L1, D1 3´~53 29 14.28, 15,
15.14

Note: Animals were censored due to any sign of physical damage because of gravity
washing, during picking for transfer (e.g. bagging and bursting) or contamination. The
comparison of individual replicates was not significant. The data became significant after
pooling the data and reanalyzing the data.

98
Table S3. Summary statistics of the multivariate models including the interaction
terms.
GST-4 ~ nPM ´ Life stage ´ Time post exposure

Estimate Std. Error t value Pr(>|t|)

(Intercept) 1.26E+00 9.51E-02 13.202 < 2e-16 ***
nPM 3.21E-03 9.11E-04 3.517 5.46E-04 ***
Life stage (L4 vs L1) 1.15E-01 1.21E-01 0.95 0.343235

Time post exposure -2.10E-02 7.08E-03 -2.957 0.003502 **
nPM:Life stage -1.04E-03 1.21E-03 -0.864 0.388929

nPM:Time -2.48E-04 6.93E-05 -3.577 0.000442 ***
Life stage:Time -3.53E-03 8.20E-03 -0.43 0.667829

nPM:Life.stage:Time 1.48E-04 8.18E-05 1.807 0.072317 .

F-statistic: 21.36 on 7 and 189 DF, p-value: < 2.2e-16

Multiple R-squared: 0.4417, Adjusted R-squared: 0.421

Size ~ skn-1 ´ nPM

Estimate Std. Error t value Pr(>|t|)

(Intercept) 100.01923 2.87004 34.849 < 2e-16 ***
skn-1 -33.05769 4.05885 -8.145 3.89E-14 ***
nPM -0.39231 0.08118 -4.833 2.66E-06 ***
skn-1:nPM 0.24588 0.11537 2.131 0.0343 *

F-statistic: 37.72 on 3 and 202 DF, p-value: < 2.2e-16

Multiple R-squared: 0.359, Adjusted R-squared: 0.3495

Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

99
Table S4. Abbreviations
PM particulate matter
TRAP Traffic related air pollution
nPM nanosized subfraction of TRAP
RNAi RNA interference
IPTG isopropyl-β-D-thiogalactoside
GFP green fluorescent protein
TOC total organic carbon
CTCF Corrected total cell fluorescence
IPTG isopropyl-β-D-thiogalactoside
ICPMS
inductively coupled plasma sector field
mass spectrometry
HVUP High-Volume Ultrafine Particle
AD Alzheimer’s disease
PIU Particle Instrumentation Unit

Figure S1. Chemical composition of the nPM0.2 sample used in this study. For
additional characterization of nPM samples, please refer to (13).
S
Na
Ca
Mg
K
Zn
Ba
P
Al
Cu
B
Mn
Ni
Sb
Fe
Mo
Se
V
Cr
Li
As
Sn
Co
Ti
Rb
0.001
0.01
0.1
1
10
100
1000
Elements >0.005 ng/ug nPM
Pb
W
Cd
Tl
Ce
Cs
Y
Pd
U
La
Nd
Ag
Eu
Sc
Hf
Pt
Nb
Th
Dy
Rh
Sm
Yb
Pr
Ho
Lu
0.000001
0.00001
0.0001
0.001
0.01
Elements <0.005 ng/ug nPM
0.001
0.01
0.1
1
10
100
1000
mass fraction (ng/µg PM)
Total
Organic carbon

100

Figure S2. Optimization of the duration of skn-1 RNAi exposure prior to nPM
exposure experiment. 18 hrs treatment was selected for the experiments. During
optimization only 2-3 biological replicates per group (2-3, ~500 animals/replicate) were
tested.

Figure S3. Representative images of gst-4::GFP animals exposed to nPM at L1
or L4 stage. (As the size of each picture is modified for better representation of the
0.5
1.0
1.5
RNAi treatment duration (hrs)
Expression relative to control
skn-1 mRNA
12 18 24
*
0.0
0.5
1.0
1.5
RNAi treatment duration (hrs)
Expression relative to control
gst-4 mRNA
12 18 24
Control RNAi
skn-1 RNAi
**
***
*

101
fluorescent changes, images are not to scale but the scalebar in pictures indicate
the size).

Figure S4. skn-1 RNAi decreased the lifespan of the animals regardless of nPM
exposure (n= 12´~50/group)

A

0 1 2 4 8 0 1 2 4 8
0.0625
0.125
0.25
0.5
1
2
4
8
16
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
gst-4
*
L1 L4
0 1 2 8
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
gcs-1
L1
Control
nPM
SKN-1 targeted genes

102
B

Figure S5. Changes in SKN-1 pathway targets following nPM exposure at L1 or
L4 stage of C. elegans development. A) Changes in skn-1 target genes from time point
0,1,2,8 L1 and 0,1,2,8 L4 following nPM exposure (n=4-5 replicates, ~500
animals/replicate). B) gst-4, gcs-1 changes at 4 hrs and 24 hrs post exposure of L1
animals (n=9´~500 animals/groups). Expression data was normalized to ama-1 Ct values
as a housekeeping gene. Statistics: Pairwise t-test was performed to compare vehicle
controls with nPM at each time point. p-values: <0.05 (*), <0.01(**), <0.001(***),
<0.0001(****).

0
50
0
50
0.0625
0.125
0.25
0.5
1
2
Expression relative to control
gst-4
L1
***
nPM Concentration (µg/ml)
4 hrs 24 hrs
0
50
0
50
0.25
0.5
1
2
4
Expression relative to control
gcs-1
L1
*
4 hrs 24 hrs
nPM Concentration (µg/ml)

103

Figure S6. Changes in genes associated with innate immune responses
following exposure to 50 µg/ml nPM at L1 or L4 stage of C. elegans development.
Expression data was normalized to ama-1 Ct values as a housekeeping gene (n=4-5
replicates, ~500 animals/replicate). Statistics: Pairwise t-test was performed to compare
vehicle controls with nPM at each time point. p-values: <0.05 (*), <0.01(**), <0.001(***),
<0.0001(****).

0 1 2 4 8 0 1 2 4 8
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
tol-1
**
L1 L4
0 1 2 8
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
tir-1
L1
0 1 2 8
0.25
0.5
1
2
4
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
abf-2
L1
*
Control
nPM
Innate immune response genes

104

Figure S7. Additional analysis of gene expression in response to air pollution
after nPM exposure at L1 or L4 stage of C. elegans development. Expression data
was normalized to ama-1 Ct values as a housekeeping gene (n=4-5 replicates, ~500
animals/replicate). Statistics: Pairwise t-test was performed to compare vehicle controls
with nPM at each time point. p-values: <0.05 (*), <0.01(**), <0.001(***), <0.0001(****).
0 1 2 8
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
hsf-1
L1
*
*
0 1 2 8
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
hsp-4
L1
*
*
Heat shock genes
0 1 2 8
0.03125
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
aip-1
L1
*
0 1 2 8
0.0625
0.125
0.25
0.5
1
2
4
8
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
cdr-1
L1
**
Metal and alcohol response genes
Control
nPM
0 1 2 8
0.015625
0.03125
0.0625
0.125
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
asr-1
L1
*

105

Figure S8. Analysis of the genes from hormone receptor, vitellogenin, and
TGFb pathways in response to air pollution exposure at L1 or L4 stage of C. elegans
development. Expression data was normalized to ama-1 Ct values as a housekeeping
gene (n=4-5 replicates, ~500 animals/replicate). Statistics: Pairwise t-test was performed
to compare vehicle controls with nPM at each time point. p-values: <0.05 (*), <0.01(**),
<0.001(***), <0.0001(****).
0 1 2 8
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
daf-12
L1
*
0 1 2 4 8 0 1 2 4 8
0.25
0.5
1
2
4
8
16
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
daf-7
*
L1 L4
0 1 2 8
0.25
0.5
1
2
4
8
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
vit-6
L1
**
0 1 2 4 8 0 1 2 4 8
0.25
0.5
1
2
Hours after 1 hr exposure
(time of dropping)
Expression relative to control
daf-2
*
L1 L4
Hormone receptor, development or vitellogenin
TGFβ pathway

106
Chapter 5. Sex-specific Cerebral Cortex and Blood Transcriptome
Changes in Mouse Neonates Prenatally Exposed to Air Pollution

Authors: Amin Haghani
1
, Jason I. Feinberg
2,3
, Kristy C. Lewis
4
, Christine Ladd-
Acosta
3,5
, Richard G. Johnson
1
, Andrew E. Jaffe
2,6-9,
, Constantinos Sioutas
10
, Caleb E.
Finch
1
, Daniel B. Campbell
4*
, Todd E. Morgan
1*
, Heather E. Volk
2,3,11*

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Department of Mental Health, Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD.
3, Wendy Klag Center for Autism and Developmental Disabilities, Johns Hopkins
Bloomberg School of Public Health, Baltimore, MD.
4, Department of Pediatrics and Human Development, Michigan State University, East
Lansing, MI.
5, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD.
6, Lieber Institute for Brain Development, Baltimore, MD.
7, Department of Biostatistics, Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD
8, Center for Computational Biology, Johns Hopkins University, Baltimore, MD
9, Lieber Institute for Brain Development, Johns Hopkins Medical Campus, Baltimore, MD
10, Department of Civil and Environmental Engineering, Viterbi School of Engineering,
University of Southern California, Los Angeles, CA.
11, Department of Environmental Health and Engineering, Johns Hopkins Bloomberg
School of Public Health, Baltimore, MD.

*Contributed equally
Corresponding author: temorgan@usc.edu

107

Abstract
Prenatal exposure to air pollutants is associated with increased risk for
neurodevelopmental and neurodegenerative disorders. Sex-specific effects of air
pollution exposure are hypothesized, but few studies have identified exposure-related
transcriptional changes. In this study we used RNA sequencing (RNAseq) to examine
transcriptomic changes in blood and cerebral cortex of three male and female mouse
neonates prenatally exposed to traffic-related nano-sized particulate matter (nPM). We
identified 19 differentially expressed genes in blood and 124 in brain, at a significance
level of p < 0.005. Only three genome-wide significant expression changes were
identified in brain. A subset of 25 showed differential response to prenatal nPM exposure
across blood and cerebral cortex. Across tissues, shared nPM responses were related to
DNA damage (e.g. NER), oxidative stress (e.g. Nrf2), and immune responses (e.g.
CD28). The potential upstream regulators of nPM responses included several Alzheimer
associated genes (e.g. App, Psen1, Tau), Rictor, and Nfe2l1 (Nrf1). Sex stratified
analyses suggest and increased number of differentially expressed genes among both
tissues in female mice, though at nominal significance. The differential expressed gene
profile of females also suggested an increase in the risk of mortality, morbidity, seizure,
and impaired learning after prenatal nPM exposure, which parallels epidemiologic
findings. Our findings suggest that prenatal nPM exposure induces transcriptional
changes in the brain, which may be mirrored in the blood and differ by sex. Further
research is needed to replicate sex-specific changes in larger sample sizes and additional
biologically relevant time points for brain development.

108
Keywords: nPM, RNA sequencing, cerebral cortex, blood

1. Introduction
Poor air quality remains a leading global risk factor of mortality and disability in
humans.(3) The gestational period is one of the most vulnerable life-stages for air
pollution exposure, with potential long-term impacts on human health.(29) Several meta-
analyses have showed that prenatal air pollution exposure is associated with premature
birth (14), low birth weight (15), as well as other longer-term health outcomes including
cardiovascular disease (22, 206), blood pressure (23, 207), and childhood cancer (24,
208). A growing body of research further indicates that prenatal air pollution exposure,
and particulate matter (PM) specifically, may affect the brain throughout life. Prenatal
exposure to PM has been associated with increased risk of autism spectrum disorder
(ASD) (16-20), childhood hyperactivity (21, 209), and cognitive impairments (25, 210,
211). Moreover, prenatal PAH exposure relates to the decrease of brain white matter
surface and increase of BMI in later childhood (27, 212). These kids show attention
deficits and hyperactivity syndromes (27, 209).
Studies from animal models similarly show the effects of air pollution on the
developing brain. Prenatal exposure to urban particulate matter (PM2.5 or PM0.2) caused
impaired neurogenesis, blood-brain barrier leakage, hippocampal mitochondrial damage,
ventriculomegaly, and neuroinflammation during adulthood (96, 122, 130, 132, 133).
These changes accompanied by spatial memory deficits and depressive behaviors (122,
130).

109
Here, we seek to identify the initial transcriptome changes that occur following
prenatal exposure to nano-size particulate matter (nPM), in both brain and blood.
Transcriptome analysis of air pollution is limited particularly in brain, rodent models, and
prenatal exposure. Prior microarray analysis of frontal cortex of Mexican cities with
different air pollution levels identified around 134 differentially expressed genes related
to inflammation and antioxidant responses in children and adults (213). In a study of adult
rat brain, chronic exposure to PM0.2 changed the expression of some genes related to
inflammation, calcium channels, and glutamate receptors (214). Chronic diesel exhaust
inhalation caused gene expression changes related to inflammation of adult mice
olfactory bulbs (215). Thus, as it is evidenced, there is a large gap of knowledge in the
brain and even blood transcriptome responses of gestational air pollution exposure.
Numerous epidemiological and mouse model studies report that gestational AirP
exposure have sex-specific toxicities (30-33). For example, while boys show higher AirP
mediated cognitive decline in attention domains (31, 34), girls show more vulnerability in
memory domains (31). ASD is also diagnosed 4 times more often in males than in females
and AirP exposure is a major contributor to ASD risk (16-20). Our experiment includes
both male and female neonates to identify sex-specific responses to prenatal AirP
exposure in early developmental stages.
These novel results can guide us to find the potential mechanism of the air pollution
neurodevelopmental toxicity. Studying the transcriptome changes of the blood and
cerebral cortex of the same animals gave us the opportunity to find the blood genes
associated with air pollution neurotoxicity. The identified genes should be examined in
human core or peripheral blood of the polluted areas to determine the translatability of

110
the findings. Moreover, this study is a valuable resource for a selective GxE analysis of
the gene variants interacting with air pollution to increase or decrease the risk of
neurodevelopmental disorders.
2. Methods
2.1. Experimental design
A total of five breeding trios of 9-week-old C57BL/6J mouse (1 male, 2 females) were
randomly assigned to per nPM or filtered air (FA) groups. Prenatal transcriptome changes
of air pollution were examined by exposing pregnant C57BL/6 dams to 300 µg/m
3
nano-
sized particulate matter (nPM) for 5 hr/day, 3 days/weeks beginning at gestational day 2
until birth. The five breeding trios per treatment group had viable litters with 35 pups for
nPM and 33 for FA. Prior to euthanasia and sample collection, the sex of the pups was
determined by the distance of the genital papilla and the anal opening. Sex was also
confirmed by PCR (see below). A random number of 3 pups per sex of each group were
euthanized at day 5 age for collection of blood and cerebral cortex.
2.2. Animals and ethics statements
Wildtype Male and Female C57BL/6 mice were purchased from Jackson Laboratory.
The experimental protocols were approved by the University of Southern California
Institutional Animal Care and Use Committee. The maintenance condition was according
to standard NIH and IACUC guidelines.
2.3. Particulate matter collection and exposure
nPM is a nano-sized (diameter <200nm) particulate matter sub-fraction of ambient
AirP collected near CA-110 Freeway in Los Angeles following prior protocols (13). Briefly,
nPM was collected on Teflon filters and resuspended in water using sonication. For

111
animal exposure, the suspended nPM was re-aerosolized at 300 µg/m
3
concentration.
The nPM composition and size distribution are characterized in our previous studies (13,
40, 216).
2.4. RNA extraction and determination of sex by qPCR
Samples were homogenized and disrupted with pestles, 22-gauge needles, pipetting
and QIA shredders (Qiagen #79654). DNA was extracted using QIAzol (Qiagen #79306)
from the organic layer for genotyping. RNA was extracted following the QIAGEN RNeasy
plus universal protocol (Qiagen #73404). Extracted RNA was used to determine sex of
mouse pups by qPCR for the Y-chromosome Sry gene. Following the determination of
sex, three males and females per group were randomly selected for further RNA
sequencing.
2.5. Library preparation and RNA sequencing of blood and cerebral
cortex
RNA (600ng) was made into RNA-seq libraries using the Illumina RiboZero Gold
library preparation kit and sequenced on an Illumina HiSeq 3000 sequencer at the Lieber
Institute for Brain Development.
2.6. Preprocessing of the RNAseq data
Raw sequencing reads were quality checked with FastQC (Babraham Bioinformatics,
2016) and, where needed, adapter sequences were trimmed from the reads using
Trimmomatic (217). Reads were aligned to the mm10 genome using the HISAT2 splice-
aware aligner (218) and alignments overlapping genes were counted using featureCounts
version 1.5.0-p3 (219) relative to Gencode version M11 (118,925 transcripts across
48,709 genes, March 2016).

112
Following alignment and mapping of the sequences to the mouse genome, the data
was converted to Count per Million (CPM) for data visualization and preliminary
assessment using EdgeR package in R. The CPM values were normalized using TMM
method (weighted trimmed mean of the log expression ratios) (220). One of the blood
samples was excluded from the analysis due to distinct count distribution compared to
others. The genes were further filtered for blood and cortex separately by omitting the
duplicates and genes with zero CPM for any sample. The data was next converted to
Log2 expression using Voom package in R for further linear modeling.
2.7. Differential expression analysis of blood and cerebral cortex based
on nPM exposed and non-exposed groups
The expression differences of each gene were calculated using Empirical Bayes
Statistics (eBayes) in the Limma package (221). In the large model, the nPM effect was
studied after adjustment for sex as a co-variate. We assessed statistical significance
using a False Discovery Rate (FDR) of 5% (q-value < 0.05) (222). We set a nominal
significant level at p-value<0.005, where no genes was detected at q < 0.05.
2.8. Sex-stratified differential expression analysis of nPM effects in
blood and cerebral cortex
The analysis was done as described before but in the sex-stratified data. The
overlapped differentially expressed genes (DEGs) were identified between blood and
cerebral cortex of male and female neonates

113
2.9. Weighted gene co-expression analysis (WGCNA) of the sex-
stratified data
WGCNA is an unsupervised clustering method to pinpoint the modules with a shared
pattern of expression (223). Briefly, the adjacency matrix of the data was converted to
scale-free networks using soft threshold power of the signed matrix. The dissimilarity
matrix was calculated from the topological overlap matrix (TOM) for hierarchical
clustering. A dynamic tree cut algorithm was used to identify the modules with a minimum
of 30 genes. The association of the modules with nPM was examined by correlation with
the module Eigengene (ME), which was calculated from the singular value decomposition
method. Hub genes of each module were defined by the highest eigengene connectivity
(kME). The maximum 150 hub genes of each module were selected for further analysis.
2.10. Ingenuity pathway analysis (IPA) of identified gene sets
The identified genes were further studies by Ingenuity Pathway Analysis (IPA)
software. The enriched canonical pathways and candidate upstream regulators were
calculated based on Right-Tailed Fisher’s Exact test of the overlap of observed genes
with the database. Enrichment of the diseases and bio-functions was done by calculating
z-scores based on the direction of the observed expression profile. Z-score is a statistical
measure that compares the direction of observed changes to the expression signature of
a specific disease constructed from prior kinds of literature.

114
2.11. Comparison of sex-specific nPM effects across blood and cerebral
cortex
We used two approaches to identify the potential blood marker for the air pollution-
mediated transcriptome changes in the cerebral cortex. Linear modeling identified shared
nPM responses between cerebral cortex and blood in stratified analysis. Sparse
canonical correlation analysis using penalized matrix decomposition (sparse CCA)
selected blood genes with maximum canonical correlation with the selected genes in the
cerebral cortex (224). This analysis was done using PMA package in R.

3. Results
3.1. Air pollution transcriptome changes in blood and cerebral cortex of
prenatally exposed pups
The total number of detected transcripts in cerebral cortex and blood of the neonates
was around 21000 and 10000 respectively. At 5% FDR, only a few differentially
expressed genes (DEGs) were identified in blood and cerebral cortex. At p<0.005,
cerebral cortex had higher DEGs compared to blood (Table 1). In cerebral cortex, the 124
nPM associated DEGs were associated with immune responses (e.g. neuroinflammation
and PI3K signaling), neurodevelopment (e.g. axonal guidance), and some metabolic
pathways (e.g. glutamate, arginine, and histamine metabolism) (Fig.1A). The upstream
regulators of these DEGs included cyclic-AMP response element binding protein (CREB),
brain drived neurotropic factor (BDNF), and interferon gamma (IFNg). In blood, the top
canonical pathways of 19 nPM DEGs included immune responses (e.g. Bcell
development, immunodeficiency signaling), cell cycle regulation, and inhibition of matrix

115
metalloproteases (Fig. 1B). The top upstream regulator of blood nPM DEGs was
mesesnchyme homeobox 2 (MEOX2), which regulates vertebrate limb myogenesis, and
also involved in neurovascular dysfunction in Alzheimer disease.
To examine if there are any potential sex-specific responses for further analysis, the
changes were studied by two models. Adding nPM: Sex interaction term increased the
number of DEGs in both cerebral cortex and blood. Thus, the data were stratified by sex
for further downstream analysis.

Table 1. Number of DEGs in cerebral cortex and blood of pups prenatally exposed to nPM
Factors Cerebral Cortex Blood
Model 1 Model 2 Model 1 Model 2
q-
value
<0.05
p-
value
<0.005
q-
value
<0.05
p-
value
<0.005
q-
value
<0.05
p-
value
<0.005
q-
value
<0.05
p-
value
<0.005
nPM vs
control
Up
Down
3
0
73
51
0
1
140
242
0
0
7
12
0
1
48
60

Sex (Male
vs Female)
Up
Down
3
4
38
71
18
30
299
273

0
0

11
16

0
1
41
37

nPM:Sex
interaction
0 267 0 106
Note: The models were adjusted for RNA integrity number (RIN). The models are
multivariate linear regression analysis of log2 gene expression. Covariates in model 1:
nPM, sex; model 2: nPM, sex, nPM-sex interaction.

116

Figure 1. Ingenuity pathway analysis of nPM associated genes. A) Canonical
pathways and potential upstream regulators of 124 nPM DEGs in cerebral cortex. B)
Canonical pathways and potential upstream regulators of 19 nPM DEGs in blood. The
DEGs were selected at p < 0.005 significance level. The models were adjusted for sex
and RIN quality of the input RNA.

3.2. Sex-specific transcriptome changes in blood and cerebral cortex of
neonates prenatally exposed to nPM
The stratified data were analyzed by linear modeling and WGCNA method. In the
cerebral cortex, only females showed nPM responses at 5% FDR with 322 DEGs (Fig.
2A). Most of these genes were downregulated following nPM exposure (259 vs. 63).
When nPM responses in males were analyzed at p<0.005 significance, 64 DEGs were
identified. Blood gene responses were lower compared to the cerebral cortex in both male
and females. Only one DEG following nPM exposure survived correction for multiple
comparisons (Gm23444 gene in females). At p<0.005 significance, 87 DEGs were
detected in female and 26 male DEGs were observed in blood samples following
exposure of pregnant dams to nPM (Fig. 2A). In both tissues, females had higher nPM
0
2
4
6
1.3
TGFB1
FEV
IFNG
BDNF
HNRNPA2B1
ACVR1
CTNNB1
CREB1
PI3K Signaling in B Lymphocytes
Protein Citrullination
Glutamate Degradation III (via 4-aminobutyrate)
Axonal Guidance Signaling
Arginine Degradation I (Arginase Pathway)
Neuroinflammation Signaling Pathway
Glutamate Dependent Acid Resistance
Histamine Biosynthesis
-log(p-value)
nPM effects, cerebral cortex
Canonical
pathways
Upstream
regulators
0
1
2
3
4
1.3
APBB2
SERPINB3
CHST15
HTR6
ZNF608
MEOX2
Primary Immunodeficiency Signaling
Inhibition of Matrix Metalloproteases
Cell Cycle Regulation by BTG Family Proteins
B Cell Development
-log(p-value)
nPM effects, blood
Canonical
pathways
Upstream
regulators
A B

117
responses compared to males, which highlights sex-specific gene expression changes.
Around 2.6% of changes (10 DEGs) were shared between male and female cerebral
cortex (Fig. 2B). Blood and cerebral cortex only shared two DEGs in females (Gm37532,
and Rgl2).
WGCNA identified several nPM associated modules in different groups of this study
(Fig. 2C). The top 150 hub genes of these modules based on connectivity (kME) were
selected for further overlap analysis (225, 226). This method could identify a subset of
25 DEGs that were shared between blood and cerebral cortex (Fig. 2D).

Figure 2. Prenatal exposure of mice to nPM caused sex and tissue-specific gene
expression changes in male and female pups. A) Differential expression analysis of the
cerebral cortex and blood transcriptome responses to nPM. Only female cerebral cortex

118
had DEGs at q<0.05 significance. B) Venn diagram showing the overlapped DEGs
between brain and blood of male and female neonates. C) WGCNA modules associated
with nPM in the cerebral cortex and blood. ** p<0.01, * p<0.05. Mean±SEM. D) Overlap
of the top 150 hub genes of identified nPM associated modules. The models and
associations were adjusted for RIN values as co-variate. WGCNA soft thresholds:
cerebral cortex: F 110, M 158; blood: F 60, M 158.

3.3. IPA analysis of the sex-specific transcriptional changes in blood and
cerebral cortex of neonates prenatally exposed to nPM
IPA was performed on the nPM associated genes in males and females cerebral
cortex and blood. The large modules were limited to the top 150 genes based on
connectivity (kME) for this analysis (225, 226). The results of different gene subsets in
each group were put together to identify the shared, tissue-, and sex-specific canonical
pathways, and potential upstream regulators of the observed changes. The nPM affected
pathways in the cerebral cortex were enriched for genes involved in mitochondria,
metabolism, protein ubiquitination, and EI2F signaling (Fig. 3A). In females, nPM affected
some nervous system pathways such as calcium, and opioid signaling. Besides, nPM
altered nitric oxide synthesis in the female cerebral cortex. In contrast, males had some
specific pathways such as Gap junction and immune responses (i.e. phagosome
maturation, RhoGDI, and CD40 signaling). In blood, nPM changes enriched for some
stress and inflammatory-related pathways in both male and females (i.e. G2/M DNA
damage checkpoint, p38 MAPK signaling). Some of the sex-specific blood pathways
included B-cell development, stem cell pluripotency, and BRCA1 DNA damage response

119
in females; hepatic fibrosis, GP6, and growth hormone signaling in males. The shared
canonical pathways between blood and cerebral cortex of male or female pups were
associated with DNA damage (i.e. NER signaling), oxidative stress (i.e. Nrf2 signaling),
and immune responses (i.e. cross-talk of dendritic and NK cells, CD28 signaling).
Several potential upstream regulators were enriched in cerebral cortex and blood of
the prenatally exposed animals (Fig. 3B). Some of the cerebral cortex regulators included
Psen1, and Mapt (Tau) which are associated with Alzheimer disease (AD); PolG and
Nfe2l1 (Nrf1) that are related to DNA damage and oxidative stress. The blood specific
upstream regulators of nPM responses included Myod1 (myogenic differentiation 1,
related to muscle regeneration), Bcyrn1 (Brain cytoplasmic RNA 1), and Ddr1 (Discoidin
domain receptor tyrosine kinase 1, involved in cell growth, differentiation and
metabolism). The shared regulators between blood and cerebral cortex included App
(amyloid precursor protein, a known AD-associated gene), Egfr (epidermal growth factor
receptor, widely recognized for the role in cancer), and Rictor (Rptor independent
companion of mTOR complex 2, an aging-related gene).
We also compared the observed nPM gene responses to gene expression signatures
of diseases in the IPA database. Only the changes in the female cerebral cortex could
significantly enrich the diseases in IPA. The results suggested that nPM would increase
the risk of morbidity, mortality, seizure, tumor formation, and learning impairments in
females (Fig. 3C).

120

Figure 3. Ingenuity pathway analysis of nPM responsive genes in the blood and
cerebral cortex of male and female pups. A) Comparison analysis of enriched canonical
pathways in all groups. The heatmap shows the top pathways shared between blood and
cerebral cortex. B) Potential upstream regulators of nPM responses in blood and cerebral
cortex. The heatmaps are sorted based on the sum of -log10(p-values) in each row. P-
values below 10
-6
were converted to 10
-6
for better visualization. C) Enriched diseases in
the cerebral cortex of females that were prenatally exposed to nPM. Z-score is a statistical

121
measure that matches between expected relationship direction built from previous studies
and observed gene expression. Z-scores > 2 or < -2 is considered as significant.

3.4. Blood nPM responses as a biomarker of cerebral cortex gene
expression changes
Comparison of the blood and cerebral cortex results revealed that nPM effects are
mostly tissue specific. The overlap analysis identified 25 genes that were shared between
blood and cerebral cortex (Fig. 2). To further select the blood genes that highly correlates
with nPM mediated brain transcriptome changes, we used sparse canonical correlation
analysis (CCA) of tissue-specific nPM DEGs. At p<0.005, a subset of 382 genes in the
cerebral cortex and 108 genes in blood showed nPM responses. CCA of these subsets
identified a group of 3 blood genes that canonically correlated (r=0.98) with 14 DEGs in
the cerebral cortex (Fig. 4). The blood genes included Id3 (inhibitor of DNA binding 3,
involved in several pathways such as adipogenesis, Wnt, Hedgehog, and Notch),
hist2h2ac (histone cluster 2, involved in meiosis and Rho GTPases), and Myom1
(Myomesin 1, involved in striated muscle contraction). Some example DEGs in the brain
included Arg1 (arginase 1, associated with innate immune responses), and Col22a1
(collagen type XXII alpha 1).

122

Figure 4. Sparse canonical correlation of nPM associated gene in blood and cerebral
cortex of mouse neonates. Heatmap showing the mean expression changes of the
selected genes in blood and cerebral cortex. Gene expressions were adjusted for RIN
values of each sample. These gene sets had a canonical correlation of 0.98.

4. Discussion
This is a novel transcriptome analysis of cerebral cortex and blood of neonates
prenatally exposed to nPM. Prenatal nPM exposure induced 124 DEGs in the cerebral
cortex of both sexes. On the contrary, only 19-87 blood DEGs and four modules were
identified in blood, which had low overlap with cerebral cortex. Sex-specific effects were
noted with the greatest numbers of changes (322 DEGs, 3 modules of genes) occurring

123
in females. Simultaneous analysis of blood and cerebral cortex of the same animals gave
us the opportunity to identify blood genes as potential biomarkers for the cerebral cortex
gene expression changes.
Prenatal nPM exposure affected 124 DEGs in the cerebal cortex that were related to
immune system (e.g. neuroinflammation), brain development (e.g. axonal guidance), and
some metabolism pathways. Prior studies on developmental effects of PM also showed
an increase of neuroinflammation (32, 227), microglial activation (96, 132), and
neurodevelopmental changes such as ventriculomegaly and hypermyelination (227). Our
study highlighted the potential role of the Creb and Bdnf in the observed changes in the
cerebral cortex. A recent study showed that prenatal PM2.5 exposure would also lead to
CREB/BDNF signaling activation in the hippocampus of one-month male and female
neonates (131). CREB is a transcriptional factor that regulates cell differentiation,
proliferation, and survival in the nervous system. This gene is a target for cancer therapy
through inhibition of phosphorylation, CREB-DNA, and CREB-CBP interactions (228).
Understanding the role of CREB during AirP toxicity and neurodevelopmental changes
requires further experiments.
Moreover, several other enriched pathways of nPM responsive genes in cerebral
cortex of males and female neonates are supported experimentally. For example, in adult
nPM exposed male and female mice selective cortical glutamatergic nPM effects were
reported (13, 38, 40). Another example is nitric oxide signaling and the potential
regulatory role of nitric oxide synthase during nPM mediated neurotoxicity. Our prior
studies showed that nPM induces iNOS, nitric oxide and/or nitrosylation in cell culture
(92, 114), hippocampal slice (122) and in vivo (114).

124
Female cerebral cortex had the highest nPM responses compared to males. A subset
of 322 DEGs and 3 gene modules were identified in the female neonatal cerebral cortex
after gestational exposure to nPM. Many epidemiological and mouse models document
that prenatal AirP exposure have sex-specific neurodevelopmental trajectories. In mice,
while prenatal PM2.5 increased the anxiety (measured by open field) in both sexes, only
females showed impaired spatial learning and memory deficits. In humans, prenatal AirP
exposure affect the attention domains of boys and memory domains of girls (31). Prior
measurement of mitochondrial DNA copy number (mtDNAcn) as a marker of
mitochondrial dysfunction in human cord blood and placenta showed off-spring sex-
specific responses to gestational AirP exposure (229, 230). Mitochondrial dysfunction and
oxidative phosphorylation were among the top nPM responsive pathways in the cerebral
cortex of both sexes of neonates, which parallels these studies.
Some of the potential upstream regulators of prenatal nPM:sex interaction included
RICTOR, Psen1, App, MAPT (Tau) and NFE2L1 (Nrf1). Recently we introduced
Caenorhabditis elegans as a model of AirP toxicity (231). We showed initial nPM
mediated skn-1/Nrf homolog responses in the developmental stage can lead to long term
developmental and lifespan changes. This study also showed sel-12/Psen homolog is
among the first larval stage nPM responses. This gene is a gamma secretase that is
involved in amyloidogenesis and Alzheimer disease. Exposure of adult AD mouse models
to nPM lead to increase of cerebral cortex amyloid b levels (38). The current results
suggest gestational AirP exposure might lead to sex-specific AD risk.
AirP is a global risk factor of mortality and morbidity. The observed nPM mediated gene
expression profile suggested that prenatally exposed animals are at a higher risk of

125
mortality, morbidity, tumor formation, seizure, and learning deficits, particularly in females.
A recent study estimated that high PM2.5 is responsible for 22% of infant death, which
was around 449,000 death excess in more than 30 countries (3). The sex differences in
AirP mediated mortality is still not clear. Cancer and neurodevelopmental effects of
gestational AirP are also documented in human studies (24, 232). Potential hazards of
gestational AirP exposure on brain tumors warrants further epidemiological investigation.
Prenatal nPM exposure had fewer gene responses in the blood compared to the
cerebral cortex.
Despite this lower response, our results identified several shared pathways between
these two tissues. This includes DNA damage (i.e. NER), oxidative stress response (i.e.
Nrf2) and some immune responses (i.e. CD28 signaling). This suggests that the nPM
systemic response impacts the brain transcriptome changes.
Our study used the mice transcriptome data to identify blood genes that can be tested
in human studies as possible biomarkers of AirP neurotoxicity. Studies show that
gestational AirP exposure leads to sex-specific telomere shortening (233, 234) and
increase of PAH-DNA adducts in cord blood (29) but these have not been validated as
potential biomarkers for brain effects. For the first time, we identified 27 blood genes (25
shared DEGs, 3 genes through CCA) associated with gene expression changes in the
cerebral cortex of the offspring. The application of these genes in human core blood or
placenta should be tested in future studies.
There are still some large gaps of knowledge to comprehend the mechanism of AirP
neurotoxicity. The first gap is unclear neurotoxic components of PM that contributes to
the observed neurodevelopmental changes. For example, our recent study showed that

126
polycyclic aromatic hydrocarbons (PAHs) of PM has limited contribution to PM
neurotoxicity (235). Prenatal exposure to PAHs is associated with neurodevelopmental
and cognitive changes during the childhood (27), however, their contribution to PM
neurotoxicity is still not clear. Another parameter not included in our study is maternal
stress. Maternal resource deprivation interacts with gestational exposure to diesel
exhaust particles (DEP) by inducing long-term offspring anxiety-like behavior and sex-
specific gene expression changes; e.g. only male offspring with prenatal DEP and
maternal stress showed increased Tlr4 and Casp1 (33). Furthermore, in humans, a three-
way interaction between PM2.5., maternal trauma and sex was shown for placental
mitochondrial DNA copy number (229). Future studies can elucidate the contribution of
maternal stress to AirP mediated transcriptome changes of neonates.
In conclusion, this novel study describes the transcriptome changes in cerebral cortex
and blood of male and female neonates after prenatal AirP exposure. Since we studied
the changes in early postnatal stages, the observed changes can be applied as
biomarkers or further studied as the biological responses that contribute to long-term
effects of AirP toxicity. We identified some blood genes that correlated with cortical
responses to AirP. Future analysis of these genes in human cord blood will determine
their relevance.

Acknowledgement
ES023780; ES022845; Dr. Steve Horvath for advices for the analysis.

Data availability: NCBI GEO Accession, GSE142453.

127
Chapter 6. Hippocampal transcriptome changes associated with long-
term neurological changes of gestational air pollution toxicity in
mouse

Authors: Amin Haghani
1#
, Richard G. Johnson
1#
, Nicholas C. Woodward
1
, Jason I.
Feinberg
2,3
, Kristy Lewis
4
, Christine Ladd-Acosta
3,5
, Nikoo Safi
1
, Andrew E. Jaffe
6
,
Constantinos Sioutas
7
, Caleb E. Finch
1
, Hooman Allayee
8
, Daniel B. Campbell
4&
, Heather
E. Volk
3,5,9&
, Todd E. Morgan
1&*

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD.
3, Wendy Klag Center for Autism and Developmental Disabilities, Johns Hopkins
Bloomberg School of Public Health, Baltimore, MD.
4, Department of Pediatrics and Human Development, Michigan State University, East
Lansing, MI.
5, Department of Mental Health, Johns Hopkins Bloomberg School of Public Health,
Baltimore, MD.
6, Lieber Institute of Brain Development, Johns Hopkins Medical Campus, Baltimore, MD.
7, Department of Civil and Environmental Engineering, Viterbi School of Engineering,
University of Southern California, Los Angeles, CA.
8, Department of Preventive Medicine, University of Southern California, Los Angeles,
CA.
9, Department of Environmental Health and Engineering, Johns Hopkins Bloomberg
School of Public Health, Baltimore, MD.

# Co-first authors
& Contributed equally
*Corresponding authors. Email: temorgan@usc.edu

128
Abstract
Air pollution adversely alters the early phases of human development and increase
the risk of autism, insulin resistance, and other diseases. The molecular mechanisms for
this toxicity are still not clear. This study exposed the pregnant dams to urban derived
nanosized particulate matter (nPM). Young adult male and female offspring were studied
for behavioral changes using forced swim test, fat gain, glucose tolerance, and
hippocampal transcriptome changes. Male offspring showed depressive behaviors,
decrease of neurogenesis in the dentate gyrus, and increase of glucose tolerance. Both
sexes showed a gain in fat and body weight. Prenatal exposure to nPM induced 29
differentially expressed genes (DEGs, p < 0.005) related to cytokine production, IL17a
signaling and dopamine degradation in both sexes. Male-specific DEGs were larger than
females (69 vs 37) in pathways such as serotonin signaling and some immune responses.
WGCNA also identified a module of 43 genes that showed divergent responses to nPM
between males and females. Chronic changes in 14 DEGs (e.g. miR9-1), mainly from
acute immune responses (e.g. HMGB1 and TLR4 signaling), were associated with
depressive behaviors and metabolic effects of nPM. Our results also showed the effects
of gestational nPM exposure on three genes (Tgfb2, Ngfr, Pik3r6) from HMGB1 signaling
in the cerebral cortex of neonates at 3-5 post-natal days. In mixed glial culture, nPM
caused a reduction in HMGB1 protein and an increase of proinflammatory cytokines after
24 hr of exposure. This response was ameliorated with the TLR4 knockdown. In
conclusion, long-lasting neuroimmune effects of nPM might underlie the depressive
behaviors and systemic metabolic changes. We also discussed several gene expression
changes that are involved in social behaviors and the risk of the Autism spectrum.

129
Understanding the causal relationship of the identified genes with depressive and social
behaviors may have important implications for therapeutic and intervention development
research.

Keywords: nPM, RNA sequencing, neurodevelopment, sex difference

130
1. Introduction
Air pollution (AirP) affects the lives of more than 90% of the world population and
cause a wide range of toxicity throughout the lifespan, particularly during developmental
stages. Gestational exposure to high level of AirP can affect brain development (27),
weight gain (212), cognition (27), behavior (21), and also increase the risk of several
diseases such as Autism spectrum (115), cancer (236), and diabetes (237, 238).
Experimental studies validated several of these relationships by exposing the rodent
models to a known concentration of AirP particulate matter (PM) during gestation and
assessing the phenotypic changes till adulthood. Short term gestational exposure of
pregnant rodents to PM2.5 or PM0.2 caused several long-lasting effects in the offspring
including changes in brain morphology (hypermyelination and ventriculomegaly) (132),
decreased neurogenesis (96), impaired cognition (239), increased depression symptoms
(122) and territorial aggressive behaviors (240), and weight gain (32). Unfortunately, AirP
is getting worst in developing countries, and the mechanism of toxicity is still too
ambiguous for efficient intervention measures. Some encouraging pieces of evidence
suggested a modest protective effect of some drugs or supplements (e.g. ascorbic acid
(241), pioglitazone (242), probiotics (243), IMD-0354 (244) and omega 3 fatty acids (245))
against AirP toxicity in rodents. Thus, resolving the molecular mechanism of AirP toxicity
that underlies chronic phenotypic changes is essential for developing more efficient
intervention strategies. We approached this gap by the hippocampal transcriptome
analysis adult mice the were prenatally exposed to nano-sized urban particulate matter
(nPM) and identified the transcriptional changes that are associated with long-lasting
phenotypic effects of nPM. This novel study allowed us to identify several new targets for

131
future mechanic and intervention experiments. Moreover, the identified genes are a great
resource for GxE analysis of AirP effects on the risk of health outcomes such as autism
spectrum and childhood asthma.
AirP developmental hazards seem to be sex-specific, which is supported by both
epidemiological and experimental research. Some studies reported that boys are more
vulnerable to developmental AirP exposure for risk of cognitive decline and attention
problems (30, 31). In mouse, prenatal exposure to PM0.2 or DEP led to male-specific
hypermyelination (121), depressive behavior (122), and insulin resistance (133). Thus,
sex seems to alter AirP toxicity and potentially gene expression responses. We included
both sexes in our analysis and described sex-specific transcriptional changes that could
lead to the observed phenotypic changes.
We further explored the changes in high mobility group box 1 (HMGB1) signaling in
the cerebral cortex of neonates and also the hippocampus of adults that were prenatally
exposed to nPM. HMGB1 is an intracellular molecule protein that is a key participant in
DNA recombination, replication (246), remodeling and repair (247, 248), transcription
(249) and autophagy (250). Interestingly, HMGB1 can be released into extracellular
space through the cytosolic compartments and mediates inflammatory or other responses
by binding to pattern recognition receptors (e.g. RAGE, TLR4, and TLR2), LPS, IL1b,
CXCL12, DNA, or nucleosome (251). This protein is considered as a potential therapeutic
target in several diseases including diabetes, Alzheimer, Parkinson, and cancer (252,
253). Our study examined the association of the HMGB1 signaling in the brain with
phenotypic changes following gestational exposure to nPM. Moreover, the initial changes
in HMGB1 protein and the correlation with pro-inflammatory cytokines were studied in the

132
mixed glial culture of rat neonates. This study supports the significance of neuroimmune
changes in the long-lasting developmental effects of gestational exposure AirP.
2. Methods
2.1. nPM Collection
nPM was collected on Teflon filters (20625.4 cm, PTFE, 2 mm pore; Pall Life
Sciences) from urban air in Los Angeles, California near the CA-110 Freeway using a
High-Volume Ultrafine Particle (HVUP) Sampler at 400L/min. The collected nPM was then
resuspended in deionized water by sonication and stored at -20°C. The aqueous
suspension is then aerosolized to an approximate concentration of 300 µg/m
3
during
exposure.
2.2. Ethics Statement
The Institutional Animal Care and Use Committee at USC approved of the animal
experiments reported here (protocol #11992). Also, experiments followed the
recommendations in the Guide for the Care and Use of Laboratory Animals of the National
Institutes of Health. All rodents were treated humanely with attention to the alleviation of
suffering.
2.3. Prenatal Exposure
9-week-old CB57BL/6NJ mice were obtained from Jackson Laboratory. Females
were housed together for a week to suppress the ovulation cycle due to the Whitten Effect.
Mice were weighed, and ear-tagged for identification. Shavings from the male’s cage
were introduced three days before the formation of the breeding pairs to restart the
ovulation cycle. Breeding trios were formed by placing two females and one male into a

133
fresh cage just before the dark cycle. Males were then removed three days later. The five
most productive breeding trios were randomly assigned into two treatment groups: nPM
and filter air-exposed. Mice were exposed to re-aerosolized nPM (300 µg/m3) for 5hr/day,
3day/week from 10:00 to 15:00 each day. Exposure stopped with the birth of the first pup.
No difference in the number of viable litters was observed in these groups. Mice were
weaned (control: Male 15, female 24; nPM exposed: male 17, female 7), and earmarked
at 4 weeks old. Animals were housed 4-5 to a cage at 25°C on a 12hr light/dark cycle.
At week 19, Mice were euthanized via cardiac puncture and the brain was perfused
with 0.9% Saline and hemisected. The left hemisphere was fixed overnight in 4%
paraformaldehyde in 0.1M borate buffer (pH 8.5) and cryoprotected in 12% sucrose, plus
flash-frozen in isopentane for sagittal sections (30 µm). The hippocampus of the right
hemisphere was used for RNA sequencing analysis.
2.4. Metabolic Profile
Total body weight was recorded every two weeks after weaning. Body compositional
analysis was conducted at weeks 6, 12, and 16 using the minSpec NMR machine.
2.5. Intraperitoneal glucose tolerance test (IPGTT)
At 16 weeks old, mice underwent IPGTT using standard protocols. Briefly, mice
fasted 10h overnight before being injected with a bolus (1mg/g body weight) of glucose
(10% wt/vol in sterile H2O) into the peritoneal cavity. Blood samples were obtained from
conscious mice through the tail vein at 0, 15, 30, 60, 90 and 120min post-injection. Plasma
glucose levels were measured using standard assays.

134
2.6. Forced Swim
The Forced Swim Test is a behavioral assay to measure stress coping strategies.
Mice were placed in a clear cylindrical water bath at 24-25 °C and recorded for five
minutes with latency to the first period of immobility and total time immobile recorded.
2.7. Immunohistochemistry
Male mice were injected I.P with 41.1 ug/g 5-ethynyl-2’-deoxyuridine (EdU) for 7 times
over three days, ending 18 days before tissue collection. Floating sections were
permeabilized in triton-100, blocked in 3% NDS, incubated in primary antibody (4°C
overnight), and secondary for four hours (1:500, Thermo Scientific). Primary antibodies
were NeuN (ab104224, Abcam), and GFAP (ab7260, Abcam). After primary staining,
slides were stained for EdU (Click-iT Plus EdU Alexa Flour 555 Imaging Kit C10639,
Thermo Fisher Scientific). EdU positive cells were identified with fluorescence microscopy
and colocalized with NeuN and GFAP to determine neurogenesis and astrogenesis.
Images were analyzed with Image J software.
2.8. RNA sequencing of the hippocampus
Five samples per group were randomly selected for RNA seq analysis. The
hippocampal RNA was extracted using QIAGEN RNeasy plus universal protocol (Qiagen
#73404). 600 ng RNA was used to prepare RNA-seq libraries using the Illumina RiboZero
Gold library preparation kit. The libraries were sequenced by Illumina HiSeq 3000
sequencer at the Lieber Institute for Brain Development. The data preprocessing included
quality check by FastQC (254) and where needed, trimming of the adapter sequences by
Trimmomatic (217). The raw reads were aligned to the mm10 genome using the HISAT2

135
splice-aware aligner (218). The overlapping gene alignments were counted using
featureCounts version 1.5.0-p3 (219) relative to Gencode version M11 (118,925
transcripts across 48,709 genes, March 2016).
The counts were then converted to count per million (CPM) for data visualization and
preliminarily assessment in the EdgeR package in R. Then, the data were normalized by
the TMM method (220) and converted to Log2 expression using the Voom package in R
for further linear modeling. Differential expression analysis was done by Empirical Bayes
Statistics (eBayes) in the Limma package. In the first model, sex was considered as a
covariate to identify the nPM effects in all groups. Next, the data were stratified by sex,
and the nPM effects were studied separately. We set a nominal significance level at p-
value<0.005, where no genes were detected at 5% False Discovery Rate (FDR).
The Log2 expression data were also analyzed by weighted gene co-expression
analysis (WGCNA). Briefly, the scale-free network was produced by giving the unsigned
matrix a soft threshold power. This network was used to calculate the topological overlap
and dissimilarity matrices. Following hierarchical clustering, the gene modules were
identified by a dynamic tree cut algorithm. Using the singular value decomposition
method, the module Eigengene was calculated and tested for association with nPM
exposure.
The association of RNA expression of the genes was tested for phenotypic changes
as an outcome using linear modeling. Log2 expression of the genes and sex were
included in the model as an independent variable. The significance of the association was
studied at 5% FDR.

136
2.9. Ingenuity pathway analysis (IPA)
The identified gene sets were analyzed by IPA (Qiagen) for the enrichment of
canonical pathways and candidate upstream regulators. The significance was calculated
by Right-Tailed Fisher’s Exact test of the overlap of observed genes with the database.
2.10. Cell culture
Primary mixed glial culture, including microglia and astrocytes, were prepared from
the cerebral cortex of postnatal day 3 of rat (mixed-sex, Sprague Dawley, Envigo,
Livermore, CA) as described before (ref). The ratio of astrocytes to microglia in these
cultures was 3:1. The cells were treated with 10 µg/ml nPM for 24 h. For TLR4
knockdown, the cells were transfected with TLR4 siRNA (Thermo fisher scientific, ID
198667) using lipofectamine RNAiMAX reagent (Thermo fisher scientific).
2.11. Protein analysis
Protein extraction was done by 1x RIPA buffer supplemented with 1mM Na3VO2,
1mM phenylmethylsulfonyl fluoride (PMSF), 10 mM NaF, phosphatase inhibitor cocktail
(Sigma), and Complete Mini EDTA-free Protease Inhibitor Cocktail Tablet (Roche).
Protein lysate supernatant was obtained by centrifugation 10,000g/10 min and the
concentration was estimated by Bradford assay. HMGB1 protein was detected by
Western blot using anti-HMGB1 (abcam, ab18256) primary antibody. The inflammatory
cytokines were analyzed by V-plex proinflammatory panel 2 immunoassays (Mesoscale
Diagnostics, Rockville, MD).

137
3. Results
3.1. Prenatal exposure to nPM caused transcriptional changes in the
adult hippocampus
No differentially expressed genes (DEGs) were detected at a 5% FDR rate, thus, the
nPM effects were studied at the nominal significance of p < 0.005. A total of 29 nPM
DEGs (15 upregulated, 14 downregulated) were detected in the hippocampus of adult
mice (age 19 weeks) after controlling for sex differences (Fig. 1A). These genes were
related to the immune system (e.g. cytokine production and IL17A signaling), metabolic
pathways that involve fatty acids, histamine, dopamine, ethanol, and estrogen. Some of
the upstream regulators of these changes include suppressor of cytokine signaling 6
(Socs6, a classical negative feedback system regulating cytokine signal transduction),
interferon regulatory factor 4 (IRF4, involved in interferon-gamma signaling), and tumor
protein P53 (TP53, a well-known tumor suppressor gene).
Weighted gene co-expression analysis (WGCNA) of the data identified a module of
43 genes with divergent sex responses to nPM (Fig. 1B). This module involved some
nervous pathways (e.g. opioid, and GABA receptors signaling), chemotaxis (e.g. G-
coupled receptor and CCR5 signaling), and metabolism. The regulators of this module
include growth differential factor 7 (GDF7, a secreted ligand of TGF-b family), and SRY-
Box transcriptional factor 2 (SOX2, involved in the stem cell maintenance in the brain).
Next, we stratified the data to understand the sex-specific effects of prenatal nPM
exposure. The results showed that males have almost 1.8 folds higher DEGs (69 vs. 37)
compared to females at p < 0.005 significance level (Fig. 1C). Male DEGs were involved

138
in G-coupled protein receptor signaling (e.g. serotonin receptor, and Gai) and some
immune pathways (e.g. endocytosis, and cAMP-mediated signaling). In contrast, female
DEGs were related to integrin signaling and some metabolic pathways such as estrogen
biosynthesis. Top potential regulators of these DEGs were zinc finger protein SNAI1 in
males and dipeptidyl peptidase like 10 (DPP10) in females.
The overlap analysis revealed that the identified four gene-sets have low overlap (Fig.
1E). Thus, each statistical approach captured a specific aspect of nPM effects and in total
affected the expression of 153 genes in a sex-specific manner. These genes were further
analyzed for association with phenotypic changes.

139

Figure 1. Prenatal exposure to nPM cause long-lasting hippocampal transcriptional
changes. A) Number of differentially expressed genes (DEGs) and Ingenuity Pathway

140
Analysis of the gene set in the hippocampus of 19 weeks mice prenatally exposed to
nPM. The nPM effects were studied at p < 0.005 significance after adjustment for sex
differences. B) WGCNA identified one module associated with nPM. C) Sex-stratified
differential expression analysis of nPM effects in the hippocampus. The significance was
studied at p < 0.005 significance. D) Venn diagram of the identified gene sets. The top
ten genes of each analysis were listed in the figure. N = 5 / group.
3.2. Long-lasting phenotypic changes after prenatal exposure to nPM:
We assessed the phenotypic changes of male and female offspring that were
exposed to nPM during three-weeks of gestational development till young adulthood (Fig.
2A). At 11-13 weeks of age, behavioral changes were assessed by forced swim assay.
Only nPM exposed male animals showed depressive behavior (~ 50% increase in the
total time of immobility) compared to females (Fig. 2B). Similarly, nPM caused an 80%
decrease of neurogenesis (Edu+ / NeuN) in the hippocampal dentate gyrus of adult males
at age 19 weeks (Fig. 2C). Other kinds of differentiated cell types such as astrocytes did
not show change by nPM (data not shown). Mice of both sexes showed weight (Fig. 2D,
5-20% slope, 5-10% intercept at age 16 weeks) and fat gain (Fig. 2E, 30-400% slope, 20-
60% intercept at age 16 weeks). While females had lower baseline in weight and fat
compared to males, prenatal nPM exposure caused greater weight and fat gain compared
to males. In the glucose tolerance test, only male mice showed an increase in insulin
resistance (20%) after being exposed to nPM.

141

Figure 2. Prenatal exposure to nPM cause long-lasting phenotypic changes in adult
mice. A) Experimental schedule. Pregnant dams were exposed to nPM during gestation.
The phenotypic assessment started after weaning in week 4. NMR: nuclear magnetic
resonance. B) Forced swim test suggest depressive behaviors only in male adult mice.
0 5 10 15 20
0
10
20
30
40
Weeks
Weight, fitted values (g)
Weight
Ctrl female
nPM females
Ctrl male
nPM males
0 5 10 15 20
0
2
4
6
8
Weeks
Fat, fitted values (g)
Fat
Ctrl Female
nPM Females
Ctrl Male
nPM Males
0
10000
20000
30000
40000
Glucose tolerance
AUC
Male Female
*
0.0
0.2
0.4
0.6
0.8
1.0
Slope
Male Female
**
*
0
10
20
30
40
Intercept at 16 weeks (g)
Male Female
**
*
0.0
0.1
0.2
0.3
Slope
Male Female
**
**
0
2
4
6
8
Intercept at 16 weeks (g)
Male Female
**
**
0
50
100
150
200
Total time Immobile (s)
Forced swim
Female Male
*
0.0
0.1
0.2
0.3
0.4
0.5
Edu / NeuN Male dentate gyrus
*
0
50
100
150
First Immobile (s)
Forced Swim
Female Male
*
Control
nPM
A B
D
E
F
Weight 4-16 weeks age
Fat 6-16 weeks age
0 50 100 150
0
100
200
300
400
Time (m)
blood glucose (mg/dl)
Weight
Ctrl female
nPM females
Ctrl male
nPM males
C
Age weeks -3 0
Birth
nPM
exposure
6 12 16 19
Tissue
collection
NMR
Weight
Forced swim
Glucose
tolerance
test
Edu injections
Wean

142
C) Neurogenesis was decreased by 70% in the dentate gyrus of male mice. The analysis
was done by t-test. Prenatal exposure to nPM cause (D) weight and (E) fat gain during
development and adulthood of both sexes. The slope and intercept at week 16 of weight
and fat were calculated from the mixed effect model including the random effects of the
mice in each group. F) Glucose tolerance test suggests an increase in insulin resistance
in male mice with prenatal exposure to nPM. Statistical analysis for outcomes with four
groups was done by ANOVA with multiple comparisons at a 5% FDR rate. * p < 0.05, **
p < 0.01. N=7-17 / group.
3.3. Hippocampal transcriptional changes are associated with
phenotypic changes:
The association of the identified 153 genes was tested with phenotypic changes at a
5% FDR rate, using linear modeling and sex as co-variate. Only the expression of
MicroRNA 9-1 (Mir9-1) was associated with changes in all phenotypes including forced
swim changes (Fig. 3A). This RNA gene is mentioned in diseases such as oral squamous
cell carcinoma and Schizophrenia.
Glucose tolerance showed some unique associations with genes such as Brevican
(Bcan, involved in chondroitin sulfate/dermatan sulfate metabolism, and adult nervous
system differentiation), and hemoglobin subunit theta 1 (Hbq1a, involved in iron ion and
oxygen-binding). Weight and fat gain were associated with a shared set of genes such
as interleukin 12A (IL12A, which is associated with atypical autism), and Ras homology
family member U (Rhou, an innate immune gene).
IPA analysis of the 14 identified genes highlighted some immune pathways such as
high mobility group box 1 (HMGB1) signaling in the phenotypic changes (Fig. 3A). Some

143
of the upstream regulators of these changes included NF-kB complex (RelA-cRel), which
is a transcriptional factor that regulates inflammation.
Phenotypes and gene expression changes showed a sex-specific correlation (Fig.
3B). While weight and fat were correlated in both sexes, forced swim and glucose
tolerance was correlated with fat gain only in males. For genes, a subset of 11 genes in
males and 7 in females were correlated with phenotypes. Some examples of sex-specific
correlation with phenotypic changes include Hbq1a in males and Cyp4f15 in females.

Figure 3. Hippocampal transcriptional changes are associated with long-lasting
phenotyping effects of prenatal nPM exposure. A) A subset of 14 / 153 nPM DEGs was

144
associated with different phenotypes in male and female adult mice. The models were
adjusted for sex differences. The heatmap represents the -log10 (FDR adjusted p-values)
of the associations. B) Pearson correlation heatmaps of the measured phenotypes and
gene expression changes in male and female mice. * p < 0.05.
3.4. HMGB1 signaling is among the initial responses to nPM:
Changes in HMGB1 signaling was assessed in the cerebral cortex of the neonates
(age 3-5 days) and the hippocampus of adults (age 19 weeks) that were prenatally
exposed to nPM. Principal component analysis (PCA) of 158 genes in the pathway
revealed that changes in HMGB1 signaling start in early development. Exposure to nPM
was associated with changes in the cerebral cortex PC4 that explained 12% of the
variance in the expression of these 158 genes (Fig. 4A). Some example genes that were
affected by nPM in the cerebral cortex of male and female neonates included 20%
increase in transforming growth factor-beta 2 (Tgfb2, involved in development and
immune system function), 50% increase in nerve growth factor receptor (Ngfr, a member
of tumor necrosis factor receptor superfamily), and 50% decrease in phosphoinositide-3-
kinase gamma adapter (Pik3r6, roles in immune system and inflammatory responses).
In contrast to the neonate cortex, nPM exposure was negatively associated with
HMGB1 PC3, which explained 10% of expression variance in this pathway (Fig. 4B). Two
genes that were chronically downregulated by gestational nPM exposure in these animals
included IL12a (> -200%) and Rhou (-100% in females).
We further examined the changes in HMGB1 protein in a mixed glial culture that were
exposed to 10 mg/ml nPM for 24 h. The results revealed that nPM exposure causes a
27% decrease in HMGB1 protein. The knockdown of TLR4 in these cells caused a

145
reduction in the HMGB1 baseline and ameliorated the nPM response. TLR4 knockdown
partially reduced pro-inflammatory responses of nPM (ref). This experiment also showed
that 30% decrease in HMGB1 protein is associated increase of proinflammatory cytokines
(KC/GRO 150 folds, IL6 100 folds, TNFa 25 folds, IL1b 6 folds, and IL5 4 folds), and
decrease of anti-inflammatory cytokines such as 50% decrease in IL4 levels.

Figure 4. Prenatal exposure to nPM alters HMGB1 signaling from early development
till adulthood. A) Changes in HMGB1 signaling and some representative genes in the
cerebral cortex of neonates that are prenatally exposed to nPM. Principal component
M F M F
-0.4
-0.2
0.0
0.2
0.4
HMGB1 PC3 (10.17%)
adjusted for sex
HMGB1 signaling
*
M F M F
-0.4
-0.2
0.0
0.2
0.4
HMGB1 PC4 (11.95%)
adjusted for sex
HMGB1 signaling
*
M F M F
-3
-2
-1
0
Log2 Expression
IL12a
*
M F M F
5.0
5.5
6.0
Log2 Expression
Rhou
*
0.0
0.5
1.0
1.5
Intensity relative to control
*
*
*
TLR4 siRNA: - - + +
HMGB1 protein
0.6 0.7 0.8 0.9 1.0 1.1
0
50
100
150
200
HMGB1 protein
Protein concentration / control
IL-6 (r = -0.96*)
KC/GRO (r = -0.96*)
TNF-α (r = -0.91*)
0.6 0.7 0.8 0.9 1.0 1.1
0
2
4
6
8
HMGB1 protein
Protein concentration / control
IFN-γ (r = -0.92*)
IL-1β (r = -0.98*)
IL-4 (r = 0.98*)
IL-5 (r = -0.93*)
C. Mixed glial culture, 24 h exposure
A. Cerebral cortex of pups, age 3-5 days
B. Hippocampus of adults, age 19 weeks
Control
nPM
M F M F
3.0
3.5
4.0
4.5
Log2 Expression
Tgfb2
*
M F M F
0.0
0.5
1.0
1.5
2.0
Log2 Expression
Ngfr
*
M F M F
0.0
0.5
1.0
1.5
2.0
2.5
Log2 Expression
Pik3r6
*

146
analysis of 158 genes from the HMGB1 signaling pathway showed nPM is associated
with changes in PC4 of the neonate cerebral cortex. B) Principal component analysis and
representative changes in 158 genes from HMGB1 signaling pathway in the hippocampus
of adult mice prenatally exposed to nPM. C) HMGB1 is among the initial responses in rat
mixed glial culture after exposure to 10 µg/ml nPM for 24 h. HMGB1 response was TLR4
dependent, which was shown by siRNA knockdown (N = 3 / group, representative of two
independent experiments). * FDR adjusted p-value < 0.05. The decrease of HMGB1
protein was significantly correlated with the increase of inflammatory cytokines such as
IL6, KC/GRO, and TNFa. The correlation analysis does not include groups with TLR4
knockdown. * p < 0.05.
4. Discussion
This study identified several long-term transcriptional changes of prenatal nPM
exposure in the hippocampus of mice. The changes included immune responses and
several metabolic pathways related to dopamine, estrogen and fatty acid oxidation.
WGCNA and also stratified analysis of the data highlighted sex-specific responses of nPM
with higher gene responses in males. This observation paralleled with male excess in
phenotypic changes such as depressive behavior and insulin resistance by prenatal nPM
exposure. We identified a set of genes mainly related to immune responses (e.g. HMGB1
signaling) that were associated with behavioral and phenotypic changes. Mir9-1 was the
only gene that was associated with all the phenotypes. Further analysis revealed that
HMGB1 signaling is among the initial response of nPM in the cerebral cortex of pups and
mixed glial culture.

147
Only a few studies described the brain transcriptome effects of prenatal exposure to
PM. A recent study identified around 11 PM associated DEGs in the cerebellum of male
and female mice at age 11-12 post-natal days (PND) (227). These genes were related to
acute immune responses. Another study analyzed the cerebral cortex transcriptome
responses of prenatal exposure to carbon black nanoparticles (CNP) at 6 weeks and 12
weeks of age (241). This study used a relaxed criterion (p < 0.01) and identified around
652 and 775 CNP associated DEGs at 6- and 12-weeks age respectively. These genes
were related to a wide range of pathways such as dopamine receptor signaling,
proinflammation and immune responses, and growth factor activities. Around 25% of
these gene expression changes could be prevented by intraperitoneal injection of
ascorbic acid (500 mg/kg, an antioxidant chemical) of pregnant dams on days 5 and 9 of
gestation, which suggest the involvement of oxidative stress and some other acute
responses in CNP toxicity.
Our study is the first that describe the hippocampal transcriptome responses of
gestational nPM exposure at later ages (19 weeks). At p < 0.005, nPM affected 29 DEGs
related to immune responses such as IL17a signaling and also metabolic pathways such
as dopamine degradation. These pathways overlap with prior reported transactional
changes of the brain at earlier ages. Moreover, dopaminergic and inflammatory changes
are among the reported long-lasting effects of gestational exposure to PM0.2, PM2.5,
diesel exhaust ultrafine particles (DEP0.2) (239, 240, 255, 256). Thus, our study extended
prior findings by identifying some new transcriptional changes and also potential
upstream regulators such as IRF4 and TP53 for future mechanistic experiments.
Changes in IL17a signaling support the hypothesis that prenatal AirP exposure can

148
increase the risk of Autism spectrum. A recent study showed that delivering IL17a, but no
other cytokines, to primary somatosensory cortex dysregular zone can rescue social
behavior deficits of several Autism spectrum mutant mouse models (257). This recent
evidence necessitates further experiments to understand the effects of AirP exposure on
brain IL17a signaling and social behaviors at different life stages.
Both of the epidemiological and experimental studies suggested that sex can alter
some of the effects of gestational exposure to AirP. Our study also showed a male excess
response in depressive behaviors, an increase of glucose tolerance and also a higher
number of gene expression changes. This observation suggests higher damage and also
a less efficient protective mechanism in developing males. Several genes related to
xenobiotic metabolism (e.g. Nrf2 components) are specifically suppressed in male
embryos, studied at gestational day (GD) 19 (258). On the contrary, the expression of
some detoxifying enzymes spikes at post-natal day (PND) 7, only in females. Thus,
potential sex-differences in oxidative stress responses start in the embryonic stage.
Resolving these initial responses is of interest for future studies.
Male-specific transcriptional responses of nPM included serotonin receptor signaling
and some immune-related responses. A study showed that gestational DEP0.2 exposure
is associated with decrease of serotonin in nucleus accumbens, amygdala, and
hypothalamus, decrease of dopamine, increase of serum testosterone, and increase of
social isolation-induced territorial aggressive behavior in male mice at 12 weeks age
(240). We did not assess the territorial aggressive behavior of our mice; however, we
observed a male-specific depressive behavior. Understanding the relationship between

149
changes in serotonin and dopamine neurotransmitters with depressive and aggressive
behaviors merits further experiments for AirP research.
We also identified a module of 43 genes that showed divergent nPM responses by
sex. These genes had functions in some neuronal pathways (e.g. opioid, GABA receptor,
and G-coupled receptor signaling), and immune responses such as CCR5 signaling. Sex-
specific responses of nPM in these pathways are unresolved in prior studies. Eomes,
known as T-box brain protein 2 (Tbr2), is identified as a potential regulator of this module
of genes. This protein transiently expresses during embryonic subventricular zone
neurogenesis and regulates neuronal development including differentiation of upper
cortical layer of neurons and also neurogenesis processes in adult dentate gyrus (259).
Developmental effects of nPM on Eomes/Tbr2 might underlie the adulthood neurogenesis
decline. Our study only analyzed the male mice for neurogenesis, however, other studies
showed a male excess in neurogenesis decline in subgranular and subventricular zone
after a single dosage of exposure to DEP0.2 during adulthood (242). Thus, sex-specific
changes in Eomes/Tbr2 during developmental exposure to nPM merits additional
experiments.
Systemic metabolic changes (e.g. grain in fat, weight, and male-specific glucose
tolerance) were among the effects of gestational exposure to nPM. These results
extended prior findings by us (133) and others (32). Exposure to AirP during adulthood
also causes an increase of glucose tolerance (119, 244, 260), and gain in fat (244). Intra-
cerebroventricular administration of an anti-inflammatory drug (IMD-0354, IKK2 inhibitor)
in adult mice ameliorated the observed glucose tolerance and fat gain by a decrease of
inflammation specifically in the brain and not systemically (119, 244). This observation

150
highlights the underlying role of hypothalamic neuroinflammation in the observed
metabolic changes (244). Our study identified a set of 14 genes in the hippocampus that
are associated with these systemic changes. Interestingly, the assessed phenotypes,
including the depressive behavior, were correlated to each other, which suggests a
shared regulatory mechanism.
MicroRNA9-1 (miR9-1) was the gene that positively correlated with both depressive
behaviors and systemic metabolic changes in male and female mice. miR9 is highly
expressed in developing and adult vertebrate brain (261), and regulate a large gene
network involved in proliferation and differentiation of neural progenitor cell population in
both embryonic (262) and adult stages (263). Changes in miR9 is reported in several
neurodegenerative and chronic diseases including Alzheimer (264), Huntington (265),
ALS (266), Parkinson (267), and several cancer types (268). Moreover, miR9-1 can have
some sex-specific responses. A study on obese pigs showed an increase in miR9 in
subcutaneous adipose tissue of both sexes but in the liver of only males (269). miR9
regulates several key transcriptional factors such NF-kB1 (270, 271), which is among the
initial responses of nPM in both in vitro and in vivo (45).
HMGB1 signaling in the hippocampus was associated with nPM mediated phenotypic
changes. This pathway was also affected in the cerebral cortex of neonates by an
upregulation in Tgfb2, Ngfr, and downregulation of Pi3kg. Notably, Pi3kg was identified
as a key regulator of IL17 cytokine production in both mouse and cell models (272). As
discussed earlier, IL17 signaling is a novel link between neuro-immunity and social
behaviors in the Autism spectrum (257). HMGB1 is a damage-associated protein that
reacts with some pattern recognition receptors (e.g. TLR4) and initiates a wide range of

151
inflammatory responses. In mixed glia, nPM induced a decline in intracellular HMGB1
protein and an increase of pro-inflammatory cytokines. TLR4 knockdown reduced the
HMGB1 baseline, eliminated the nPM HMGB1 and cytokine response (45). TLR4
knockdown mouse also has a lower level of HMGB1 protein and inflammatory responses
(273). Thus, our study suggests that acute and chronic changes in HMGB1 and TLR4
signaling might underlie the systemic and behavioral effects of gestational exposure to
nPM. Future studies could include mutant mice in HMGB1, TLR4 and IL17 to understand
the mechanism of nPM mediated toxicity.
In sum, our study discussed several potential mechanisms of gestational AirP toxicity.
The findings extended prior research and identified some targets of interest such as
HMGB1, miR9, and IL17 for further mechanistic and also intervention studies. Mutations
in these genes should be screened for GxE interaction with air pollution and the risk of
neurodevelopmental disorders and the Autism spectrum.

152
Chapter 7. Mouse brain transcriptome responses to inhaled
nanoparticulate matter differed by sex and ApoE in Nrf2 - NF-kB
interactions

Authors: Amin Haghani
1
, Mafalda Cacciottolo
1
, Kevin R Doty
2
, Carla D’Agostino
1
, Max
Thorwald
1
, Nikoo Safi
1
, Morgan E Levine
3
, Constantinos Sioutas
4
, Terrence C Town
2
,
Henry Jay Forman
1
, Hongqiao Zhang
1
, Todd E Morgan
1
, Caleb E Finch
1,5*

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Zilkha Neurogenetic Institute, Department of Physiology and Neuroscience, Keck
School of Medicine of the University of Southern California, Los Angeles, CA.
3, Department of Pathology, Yale School of Medicine, New Haven, CT.
4, Department of Civil and Environmental Engineering, Viterbi School of Engineering,
University of Southern California, Los Angeles, CA.
5, Dornsife College, University of Southern California, Los Angeles, CA.

*Corresponding author. Email: cefinch@usc.edu

Running title: Air pollution and brain transcriptome

153
Graphical abstract

154
Abstract
The neurotoxicity of air pollution (AirP) is undefined for sex and ApoE alleles. These major
risk factors of Alzheimer disease (AD) were examined in mice given chronic exposure to
nPM, a nano-sized subfraction of urban AirP. In cerebral cortex, female mice had 2-fold
more genes responding to nPM than males. Transcriptomic responses to nPM had sex-
ApoE interactions in AD-relevant pathways. Only ApoE3 mice responded to nPM in genes
related to Ab deposition and clearance (Vav2, Vav3, S1009a). Other responding genes
included axonal guidance, inflammation (AMPK, NF-kB, APK/JNK signaling), and
antioxidant signaling (Nrf2, HIF1a). Genes downstream of NF-kB and Nrf2 responded in
opposite directions to nPM. Nrf2 knockdown in microglia augmented NF-kB responses to
nPM, suggesting a critical role of Nrf2 in AirP neurotoxicity. These findings give a rationale
for epidemiologic studies of AirP to consider sex interactions with ApoE alleles and other
AD-risk genes.

Keywords: Air pollution, ApoE, Sex, Nrf2, NF-kB, transcriptome

155
1. Introduction
Air pollution (AirP) is a major global environmental risk factor of morbidity and
mortality across the human lifespan (274, 275). AirP exposure is also associated with
neurodegeneration: accelerated cognitive decline of aging and increased risk of
Alzheimer’s disease (AD) (27-29, 116, 276). However, little is known of interaction of AirP
neurotoxicity for sex and apoE alleles and other AD risk factors (277).
Epidemiological studies of AirP neurotoxicity have not identified interactions of
gender by ApoE alleles. Findings are typically ‘adjusted or controlled’ for gender
differences (56, 278-281). A recent study from China suggested greater verbal test
deficits in men than women (282). Sex-ApoE interactions for AirP neurotoxicity remain
undefined. In a small sample from a polluted Mexican City, ApoE4 heterozygous females
with high BMI had higher risk of severe cognitive deficit than other groups (283). This
finding supported our recent study of the WHIMS cohort of elderly women, in which apoE4
homozygotes had a greater risk of dementia and accelerated cognitive decline (38)
Developmental AirP exposure has received greater attention for gender because of the
consistent male excess vulnerability for behavioral and cognitive deficits in the pre-
adolescent and young adult (284, 285).
Mouse models have not addressed sex and ApoE in responses to AirP. Our initial
study examined female EFAD mice carrying transgenes for familial AD genes together
with human apoE alleles by targeted replacement (ApoE-TR). Consistent with WHIMS
findings, E4FAD female mice accumulated more brain amyloid in response to nPM than
the E3FAD (38). However, for ozone exposure, male ApoE-TR showed the converse of

156
greater behavioral impairments in E3 than E4 (286). For further study of both sexes, we
examined the cerebral cortex transcriptomic responses of ApoE-TR and wildtype mice
(C57BL/6J) by RNAseq for the main regulators of AirP toxicity in AD pathways.
We focused on genomic pathways mediated by Nrf2 and NF-kB, which responded to
AirP in our prior studies (45, 47). These redox-sensitive transcription factors control
hundreds of genes that mediate cellular responses to oxidative stress and immunity. They
respond to oxidative stress, tobacco smoke, traumatic brain injury, and ischemic stroke,
and are altered by aging and AD (48). NRF2 downstream genes include anti-oxidants
(e.g. glutathione, thioredoxin), anti-inflammatory cytokines (e.g. IL-10), phase I & II
xenobiotic detoxifying enzymes (e.g. CYP450) and free radical scavengers (49). The NF-
kB family transcriptionally regulates the expression of immune related proteins including
cytokines (e.g. TNFa, IL1a/b), antigen presentation proteins (e.g. MHCI, b2-
microglobulin), chemokines (e.g. MCP-1, MIP-1), adhesion molecules (e.g. ICAM-1,
VCAM-1), inducible nitric oxide synthase (iNOS), and proapoptotic (e.g. BIM, BAX) or
antiapoptotic proteins (e.g. XIAP, BCL-2) (50). The complex interplay of NRF2 and NF-
kB signaling pathways can alter the balance of anti-oxidative or inflammatory responses,
depending on the type of stress, and target cell or tissue (48).
Sex and ApoE alleles can also alter NRF2 and NF-kB activities, as shown for the
larger response of female mice for hepatic Nrf2 activation by phenobarbital and
oxazepam and other xenotoxins (287). Nrf2 downstream genes including GSTA2, HO1,
and NQO1 showed lower hepatic expression in ApoE4-TR than ApoE3-TR mice (288).

157
We therefore examined both sex and ApoE allele for interactions with NRF2 / NF-kB
responses of AirP neurotoxicity.
2. Methods
2.1. Animals
Husbandry and experimental procedures were approved by the USC Institutional
Animal Care and Use Committee. The C57BL/6J and ApoE-TR mice were aged 2 months
at exposure onset (289, 290). For long-term nPM exposures (8-15 weeks), 4 mice for
each sex, genotype (C57BL/6J, ApoE3-TR, ApoE4-TR), and exposure (48 mice total)
were randomly assigned to nPM exposure or control. The dose-response experiment was
done with 10 male C57BL/6J mice per group. After exposure, mice were euthanized by
isoflurane anesthesia and perfused transcardially with phosphate-buffered saline. Brains
were hemisected at midline; total cerebral cortex was frozen on dry ice and stored at -
80ºC. Investigators were blinded to exposure groups during data measurement and
analyses.
2.2. Air pollution nPM collection and exposure
Mice were exposed to nPM, a nano-sized subfraction of airP particulate matter of 2.5
microns diameter (PM2.5) collected from a local urban freeway corridor (40). Chemical
composition and size distribution of re-aerosolized nPM is characterized in previous
studies (13). Re-aerosolized nPM or filtered air (control) was delivered to the sealed
exposure chambers at approximately 300 µg/m3 concentration to model chronic
exposure: 5 h/day, 3 days/week, for 8 (C57BL/6J) or 15 weeks (ApoE-TR). For dose-

158
response experiment, 8 weeks male C57BL/6J mice were exposed to approximately 100,
200, and 300 µg/m
3
for 3 weeks.
2.3. RNA sequence (RNAseq) analysis of mouse cerebral cortex
RNA was extracted with TRIzol (Invitrogen) and RNAeasy Mini Kit (Qiagen) with
DNase digestion. Libraries were made with the TRUseq Stranded mRNA Kit (Illumina)
with 1 mg of RNA. For Illumina NextSeq500sequencing, a single end-sequencing length
of > 50 nt was used. Reads were aligned and quantified to the mouse reference genome
RefSeq mm10 with Tophat2 (v2.0.8b), restricted to uniquely mapping reads with 3
possible mismatches using the Partek flow software platform (291).
Differential gene expression was calculated by linear modeling (Limma package in
R). Significance was calculated at 5% FDR rate or p-value, 0.005. Gene responses were
analyzed by Qiagen Ingenuity Pathway Analysis (IPA) software.
2.4. Weighted Gene Co-Expression Network Analysis (WGCNA)
The co-expression network, based on WGCNA, was constructed from RNAseq data.
WGCNA is an unsupervised clustering approach, which assigns groups of genes with
shared expression patterns into modules (223). Briefly, the adjacency matrix (correlations
between genes) was converted to a scale-free network using soft threshold power (tuned
in each group) of the signed matrix. The result was converted to a topological overlap
matrix (TOM). Hierarchical clustering used 1-TOM distance measure (dissimilarity). A
dynamic tree-cut algorithm was used to assign modules containing at least 30 genes.
Module eigengenes (MEs) were calculated as the maximum amount of the variance of
the model for each module, based on the Singular Value Decomposition method. Linear

159
regression models estimated association of nPM or ApoE with the MEs. The top 150 hub
genes of the modules were selected for IPA analysis by the highest eigengene
connectivity (kME).
2.5. Cell culture and Nrf2 siRNA
BV2 microglia (mouse-derived) were grown in Dulbecco’s modified Eagle’s medium
(DMEM)/F12 (Cellgro, Mediatech, Herndon, VA) containing 10% fetal bovine serum, 1%
penicillin/streptomycin, and 1% L-glutamine (45). Nfe2l2 (Nrf2) siRNA (156499,
Thermofisher Scientific) was delivered by Lipofectamine RNAiMAX reagent
(Thermofisher Scientific).
2.6. Quantitative real-time PCR
Extracted RNA was converted to cDNA using qScript cDNA Supermix (Quantabio).
qPCR was done using Taq master mix (Biopioneer) and gene-specific primers (Table S1;
Fig. S1).
2.7. Protein extraction
Frontal cerebral cortex (anterior to Bregma, excluding olfactory bulbs) was
homogenized (20 mg in 0.2 ml) in 1x RIPA buffer supplemented with 1 mM Na3VO2, 1
mM phenylmethane sulfonyl fluoride (PMSF), 10 mM NaF, phosphatase inhibitor cocktail
(Sigma), and Complete Mini EDTA-free Protease Inhibitor Cocktail Tablet (Roche).
Supernatants were obtained by centrifugation at 12,000g/15 min.
2.8. Nrf2 localization
Nuclear and cytosolic fractions were separated after tissue homogenization in
sucrose-Tris-MgCl2 (STM) buffer with phosphatase and protease inhibitors and

160
centrifuged 800g x 15 min (137). After removing supernatant, the nuclear pellet was
washed in STM buffer, resuspended in HEPESpH 7.9 buffer (20 mM HEPES 1.5 mM
MgCl2, 0.5 M NaCl, 0.2 mM EDTA, 20% glycerol, 1% Triton-X-100, protease and
phosphatase inhibitors) and sonicated. Cell fraction purity was validated by
immunoblotting for nuclear histone 3 (H3) and cytosolic glyceraldehyde 3-phosphate
dehydrogenase (GAPDH).
2.9. Protein analysis
Nrf2 was detected by Western blot using anti-NRF2 primary antibody (1:1000, rabbit
polyclonal, ab137550). Proteins (20 µg) were electrophoresed on Criterion 4-15% TGX
gels (Biorad) and transferred to PVDF membranes. After washing with TBS+0.05%
Tween-20 (PBST), membranes were blocked (LiCOR) 1 h/ambient, then incubated with
primary antibody overnight at 4 °C: anti-NRF2 (1:1000, rabbit polyclonal, ab137550), anti-
H3 (1:1000, Rabbit polyclonal, Cell Signaling Technology, D1H2), and anti-GAPDH
(1:500, Mouse monoclonal, Santa Cruz Biotechnology, sc-32233). After 1 hr incubation
with 1:20,000 fluorochome-conjugated LICOR-antibodies (anti-mouse IRDye 800CW or
anti-rabbit IRDye 700CW), blots were scanned, and band intensity analyzed by ImageJ.
GCLC was assayed by capillary electrophoresis (12-230kDa range, Jess ProteinSimple,
California, USA). Total lysate 1 µg/µl was electrophoresed and treated with anti-GCLC
(1:100) and HRP-labeled secondary antibody. Results were normalized to total protein
(PN module, ProteinSimple). IL2 was assayed by V-PLEX proinflammatory panel 2
immunoassay (Mesoscale Diagnostics, Rockville, MD).

161
3. Results
To define brain transcriptional responses of AirP and interactions with sex and ApoE
alleles, we examined responses of adult C57BL/6J (wild type, ‘B6’) and B6 mice carrying
human ApoE alleles by targeted replacement (ApoE-TR) to nPM, a nanosized subfraction
of urban air pollution PM2.5. Three independent exposures used different batches of nPM
at specified durations of exposure. In vitro studies on BV2 microglia examined the role of
Nrf2 and NF-kB in responding inflammatory pathways.
3.1. Cerebral cortex transcriptome responses to nPM
Differentially expressed genes (DEGs) were analyzed by RNAseq for nPM
responses. Stratification by ApoE and sex was done subsequently to establishing general
effects. The multivariate model of combined B6 and ApoE-TR data was adjusted
statistically for sex, ApoE genotype, and different nPM batches of the two exposures. For
p = 0.005, there were 140 DEGs (118 increased, 22 decreased) responses to nPM (Fig.
1A). Ingenuity pathway analysis (IPA) of responding pathways included synapse function
(e.g. axonal guidance, calcium signaling, endocannabinoid neuronal synapse),
inflammation (e.g. AMPK, SAPK/JNK), circadian rhythm, NRF2 mediated antioxidant
response, and hypoxia-inducible factor 1-a (HIF1-a) (Fig 1B).
RNAseq data were stratified to identify sex- and ApoE-specific nPM responses by
linear models and by weighted gene co-expression (WGCNA). Females had more DEG
than males for both B6 and ApoE-TR mice, by to 2-fold (Fig. 1C). Female ApoE3 and B6
had the most nPM responding genes (153 vs 189, respectively). Gene modules identified
by WGCNA also had more female responses to nPM for B6 and ApoE-TR (Fig. 1C).

162
Modules were constrained to a maximum of 150 hub genes, based on connectivity
(Methods). Both analyses (DE, WGCNA) showed more nPM-responding gene responses
for ApoE3 than E4.
Upstream regulators and canonical pathways were identified by IPA for sex-specific
and shared nPM responses. The top upstream candidate was NFE2L2 (NRF2), a
regulator of Phase II detoxification (Fig. 1D, S2, S3), which had the strongest associations
for B6 and ApoE4. Sexes differed in immune-related upstream regulators of gene
responses. Female-specific responses included PPAR-g (peroxisome proliferator
activated receptor gamma), SP1 (specificity protein1 transcription factor), and TNFSF11
(TNF superfamily 11). Male-specific responses included TRAF6 (TNF associated
receptor factor 6), CAMK2 (regulator of synaptic plasticity and AMPA glutamate
receptors), and CCDC22 (regulator of NF-kB signaling by interaction with COMMD
proteins). These results paralleled the enrichment of Nrf2 and immune response
pathways in the combined multivariate model above.
Stratified analysis by ApoE and sex for canonical pathways showed nPM responses
of neuronal pathways; e.g. G-coupled protein receptors, axonal guidance, ephrin
receptors, synaptic long-term depression, and endocannabinoid development neuron
pathway (Fig 1E). Other nPM responding genes were related to relaxin, GM-CSF, and c-
AMP mediated signaling. Female-specific responses include genes associated with the
following pathways: AMPK, dopamine-PARPP32 feedback in cAMP, gap junction
signaling, and nitric oxide production. Male-specific responses in both mouse strains were
enriched for ‘cardiac hypertrophy’ signaling; e.g., ELK1 (transcription factor) and HSP27.
ApoE3 and E4 had different inflammatory responses for NF-kB, IL6, CREB, and IL22 (Fig.

163
S3). Cell-type deconvolution analysis of RNAseq also showed apoE and sex-specificity
for glial responses to nPM (Fig. S3B).

Figure 1. Cerebral cortex transcriptome responses to nPM in B6 and ApoE-TR mice.
A) Multivariate differential expression analysis of nPM responses in combined data from

164
the independent C57BL/6J and ApoE-TR experiments. Batch differences of nPM (Zhang
et al. 2018) were adjusted by the COMBAT method (292). Covariates included sex, ApoE
genotype, and nPM. DEGs identified at p-value, 0.005. B) Canonical pathways associated
with nPM DEGs. C) Stratified DE and WGCNA modules associated with nPM responses,
male (M) and female (F). The top 150 genes of modules (kME inter-module connectivity)
were used for IPA analysis. Significance was calculated from the Pearson correlation of
eigengene of the modules with nPM. D) Upstream regulators and E) canonical pathways
associated with nPM transcriptome responses in B6 and ApoE-TR mice. Solid horizontal
lines separate responses that are shared and sex-specific. Heatmaps were sorted by the
sum of -log10 (p-values) in each row. p-values < 10
-6
were converted to 10
-6
for better
visualization; grey, not significant.
3.2. Baseline effects of ApoE4 allele and the overlap with nPM responses
Baseline differences by ApoE in non-exposed controls were analyzed in both sexes
in two steps. The combined multivariate model showed 133 DEGs differed in baseline
ApoE allele effect (5% FDR) (Fig. 2A). These DEGs were enriched for immune-related
pathways including rheumatoid arthritis, granulocyte adhesion, IL10, and NF-kB
signaling. E4 baseline differences included pathways of glutamate metabolism, and
production of nitric oxide, superoxide and hydrogen peroxide, the LXL/RXR pathway of
cholesterol efflux, and atherosclerosis.
In stratified analysis, males had 60% more DEGs differing by ApoE alleles (male, 75
genes; female, 45 genes (Fig. 2B, Fig. S4). For WCGNA modules differing by ApoE
alleles, males had 3-fold excess (male 12 modules; female, 4 modules). Subsets of DEG
(females, 70 DEGs; males, 37) had about 7% overlap with nPM responding genes (Fig.

165
2C-D), suggesting convergent effects of E4 allele and nPM. In females, the shared
responding genes involved metabolic pathways (glycolysis, oxidative phosphorylation)
and DNA repair (HMGN1). For males, in contrast, the overlap involved a different set of
genes related to iron homeostasis, telomere extension, and immune response (TREM1,
IL11, and JAK signaling).
AD-related pathways differed by sex and ApoE alleles for nPM responses (Fig 3A-
B). Only female E3 responded to nPM in five AD pathway genes for amyloid precursor
protein (APP) processing and for tau: App, Bace1, Psen1 (Fig. 3A); Tau and Gsk3b (Fig
3B). About 10% of genes related to amyloid clearance (5/46) differed by ApoE or nPM.
For amyloid clearance genes, only E3 carriers in both sexes responded to nPM (Fig 3C).
Three genes mediating amyloid clearance responded to nPM with sex differences only in
E3 of both sexes. For E3 females, Vav2 (+50%); Vav3 (decreased -50%); the Vav
guanine nucleotide exchange factors mediate phagocytosis of fibrillar Ab (293). For E3
males, S100a9 (-60%), also known as Mrp14, regulates microglial phagocytosis of fibrillar
Ab (294). Baseline expression of two complement genes was higher in E3 than E4: C3
(10-fold), C3ar1 (+50%).

166

Figure 2. ApoE allele baseline differences of RNA in cerebral cortex. A) Differential
expression analysis of ApoE4-TR vs -E3, at 5% FDR and p-value, 0.005. B) WGCNA
modules associated with E4 allele. IPA of the top 150 genes of the modules identified by
kME (inter-module connectivity). Significance was calculated from the Pearson
correlation of eigengenes for modules with E4 allele. * p < 0.05; ** p < 0.01; ***p < 0.001;

167
****p < 0.0001. IPA analysis of overlapped genes between baseline differences by ApoE
allele and nPM response in females (C) and males (D). The genes in each group are a
combination of identified genes based on DE and WGCNA.

Figure 3. AD-associated gene responses to nPM in cerebral cortex. A) Amyloidogenic
pathway responses. Female E3 had largest nPM response. B) Tau and its kinase, Gsk3b.
C) Aβ-amyloid clearance pathway responses to nPM. A small subset (10%, 5/46) of
amyloid clearance genes differed by ApoE or nPM (genes identified in the IPA database
for phagocytosis, proteolysis, degradation, deposition). Only ApoE3 responded to nPM.
Mean ± SEM. ANOVA; FDR multiple test correction. * Adj. p-value, 0.05.

0.0
0.5
1.0
1.5
2.0
CPM/ E3F Ct
App
*
F M F M
ApoE3-TR ApoE4-TR
*
Bace1
F M F M
ApoE3-TR ApoE4-TR
*
*
Psen1
F M F M
ApoE3-TR ApoE4-TR
*
*
Tau
F M F M
ApoE3-TR ApoE4-TR
*
*
Gsk3β
F M F M
ApoE3-TR ApoE4-TR
*
*
Control
nPM
A. Amyloidogenesis genes
0.0
0.5
1.0
1.5
2.0
CPM/ E3F Ct
C3
F M F M
ApoE3-TR ApoE4-TR
*
Vav3
F M F M
ApoE3-TR ApoE4-TR
*
C3ar1
F M F M
ApoE3-TR ApoE4-TR
*
S100a9
F M F M
ApoE3-TR ApoE4-TR
*
Vav2
F M F M
ApoE3-TR ApoE4-TR
*
C. Genes associated with amyloid clearance
B. Tau associated genes

168
3.3. Sex- and ApoE- specific nPM mediated NF-kB responses
Next, we examined genes of the NF-kB pathway, which regulate pro-inflammatory
responses to nPM, as shown for responses of wildtype mice (B6 males) cerebellum and
lung to nPM (45). In cerebral cortex, ApoE-TR mice responded to nPM with a subset of
genes downstream of NF-kB (13%, 8/133) that differed by sex and ApoE (Fig. 4A). Two
clusters of nPM responses were identified by Principle Component Analysis for these 133
NF-kB downstream genes: PC2 (20% of variance, nPM: sex interaction) and PC4 (2.5%
of variance nPM:E4 interaction) (Fig. 4B).
Inflammatory cytokine responses were examined as protein levels for these PCs
in cerebral cortex. PC2, but not PC4, was correlated with TNFa (b= -0.002, p = 0.02,
R
2
=0.15), IL1b (b=0.004, p = 0.002, R
2
=0.27), and CXCL1 (b=0.005, p = 0.01, R
2
=0.19)
proteins (Fig. 4C). Only females responded to nPM for these cytokines. These RNA and
protein responses are notable for consistent sex-specific inflammatory responses to nPM.

169

Figure 4. nPM induced inflammatory responses with sex- and ApoE specificity.
A) Stratified analysis of NF-kB downstream genes responses to nPM. The combined IPA
datasets included 133 NF-kB downstream genes. B) Principal component analysis of 133
NF-kB downstream genes in ApoE-TR: PC2 (20% of total variance) was associated with
nPM: sex interaction; PC4 (2.5% of variance), associated with nPM:ApoE interaction. C)
Protein levels of genes downstream of NF-kB were correlated with PC2: positive
correlations for CXCL1 and IL1b; inverse correlation with TNFa. Only females responded
to nPM.

170

3.4. Sex- and ApoE allele-specific Nrf2 responses
Nrf2 downstream responses to nPM differed by sex and ApoE (Fig. 5A). The 60
responding genes included Gpx3, and Gstp1, Jun, NFe2l1 (Nrf1), and several Maf family
transcription factors. A subset of gene responses was validated by qPCR, e.g. Nrf1 (Fig.
5B; Fig. S1). Female B6 and ApoE-TR had 2-fold or more Nrf2 downstream genes
responding to nPM. PC2 is associated with nPM for interactions with sex (p = 0.01) and
ApoE (p = 0.02), 6.4% of the variance, mainly associated with E3 females (Fig. 5C). The
strong inverse correlation of Nrf2 PC2 with NF-kB PC2 (Fig 5E, R
2
= - 0.91, p = 0.0001)
suggests crosstalk between these transcriptional factors during exposure to nPM.

171

Figure 5. Nrf2 responses to nPM in B6 and ApoE-TR mice. A) Heatmap of log2 fold
changes of nPM responses, showing altered expression of at least 60 genes downstream
of Nrf2, differing by sex or ApoE genotype. B) Validation by qPCR of Nfe2l1 (Nrf1)
changes in RNAseq. C) Principal component analysis of 513 Nrf2 downstream genes in
ApoE-TR: Only PC2 (6.4% of variance) had nPM-sex interaction (p = 0.01) and ApoE (p

172
= 0.02). E3 females had the highest nPM response. D) PC2 of Nrf2 downstream genes
varied inversely (R2 = 0.91, p = 0.0001) with the PC2 of NF-kB downstream genes.

3.5. Inhibitory effects of Nrf2 on NF-kB response to nPM
The relationship of Nrf2 and NF-kB responses of nPM was further explored in an
independent dose-response experiment. The duration of inhalation exposure was only
three weeks to assess the initial responses of nPM by male C57BL/6 mice. After a 3week
exposure to 300 µg/m
3
nPM, the cerebral cortex had nuclear translocation of Nrf2 protein
(+50%) and increased cytosolic NF-kB/P65 (+25%) (Fig. 6A). Downstream of Nrf2, the
rate limiting enzyme of glutathione synthetase (GCLC) had dose-dependent increase
correlated with Nrf2 mRNA (r

= 0.6, p = 0.005) (Fig. 6B). Nrf2 and NF-kB responses of
B6 male mice at three nPM doses had opposing changes of increased Nrf2 mRNA, but
decreased NF-kB/P65, NF-kB/P50 mRNA and IL2 protein, all with dose-dependence
(Fig. 6C).
Nrf2 and NF-kB interactions during acute (6 h) exposure to nPM were examined in
BV2 microglia in vitro. Partial knockdown of Nrf2 (-40%) by siRNA increased the NF-kB
mediated responses of nPM (Fig. 6D), with 30% higher NF-kB/P50 mRNA, and 200%
higher mRNA of iNOS, IL1b, and IL6.

173

Figure 6. Nrf2 and NF-kB interact with nPM toxicity in cerebral cortex of male C57BL/6
mice and in mouse microglia (BV2 cells). A) Increased nuclear translocation of Nrf2 and
cytosolic NF-kB/P65 of B6 mice exposed to 300 µg/m
3
nPM for 3 wks. B) nPM exposure
dose-dependent increase of Nrf2 and positive correlation with increase of GCLC protein.
C) nPM dose-dependent decrease of NF-kB/P65 and NF-kB/P50 mRNA, and IL2 protein
levels. Inhalation exposure to nPM at 100, 200, and 300 µg/m
3
nPM (N=10/group,
0
300
0
300
0
300
0
300
0.0
0.5
1.0
1.5
2.0
2.5
OD (Intensity relative control)

*
Nuc Cyt
PM concentration (µg/m
3
)
Control
nPM
*
Nuc Cyt
NRF2
protein
NF-κB/p65
protein
Cerebral cortex of young adult male B6 mice
B A
BV2 microglia D
0.0
0.5
1.0
1.5
2.0
2.5
Expression / control
Nrf2
mRNA
Control
*
nPM
*
*
0.0
0.5
1.0
1.5
2.0
2.5
NF-κB/P50
mRNA
*
*
Control nPM
Ctr media
ScrRNA
Nrf2 siRNA
0
2
4
6
IL1β
mRNA
*
*
Control nPM
0
2
4
6
8
10
Expression / control
iNOS
mRNA
*
*
Control nPM
0
2
4
6
IL6
mRNA
*
*
Control nPM
0 100 200 300
1
2
PM concentration (µg/m3)
Expression / control
Nrf2 mRNA***
(β= 0.001, R2=0.37)
0 100 200 300
1
2
PM concentration (µg/m3)
Expression / control
NF-κB/P65 mRNA*
(β= -0.0004, R
2
=0.15)
0 100 200 300
1
2
PM concentration (µg/m3)
Expression / control
NF-kB/P50 mRNA**
(β= -0.001, R
2
=0.25)
C
0 100 200 300
0.0005
0.0010
0.0015
0.0020
0.0025
PM concentration (µg/m3)
Concentration (pg/µg protein)
IL2 protein**
(β= -0.0007, R
2
=0.24)
0 100 200 300
1
2
3
PM concentration (µg/m3)
OD (Intensity relative control)
GCLC protein**
(β= 0.002, R
2
=0.37)

174
Exposure: 5 h/d, 3 d/wk, 3 wks). **p = 0.001, ***p = 0.0001. D) BV2 microglia response
to nPM at 5 µg/ml nPM for 6 h after partial knockdown of Nrf2 in (N = 6/group). Nrf2 mRNA
knockdown was > 60% at time 0. ANOVA with FDR multiple test correction. Mean ± SEM.
*Adj. p = 0.05.
4. Discussion
These finding address a gap in how sex and ApoE alleles alter neurodegenerative
responses to AirP. Cerebral cortex genomic responses to nPM, a subfraction of ultrafine
PM, were examined in C57BL/6 mice (wildtype) and with human transgenes for ApoE
alleles. Female mice had 2-fold more genes that responded to nPM, further enhanced for
ApoE3. Responding genes included neuronal pathways (e.g. axonal guidance; glutamate
genes), inflammation (e.g. AMPK, and APK/JNK signaling), and antioxidant and hypoxic
signaling (e.g. Nrf2, HIF1a signaling). Genes in pathways downstream of NF-kB and Nrf2
responded oppositely to nPM. Interactions of Nrf2 and NF-kB underlie sex and ApoE risk
for AD during AirP exposure.
The nPM responding genes may be useful for identifying GxE in
neurodegenerative risks from AirP and cigarette exposure. For example, a combination
of AirP and a specific IL1b variant increased the risk of Parkinson disease (295). Our
findings extend microarray analysis of frontal cortex of children and young adults from
Mexican cities differing in AirP levels: the 134 responding genes include inflammation
(e.g. NF-kB) and antioxidant responses (e.g. GPX2, GPX3) (213). Microarray analysis of
rat brain chronically exposed to PM0.2 also overlapped with our results: S100a9, calcium
channels (e.g. Cacna1i), and glutamate receptors (214).

175
These glutamatergic gene responses extend findings that hippocampal neurites
are selectively sensitive to nPM (13, 40). The increased levels of the ionotropic receptor
NMDA type subunit 1 (GRIN1) is notable: a GRIN polymorphism is associated with the
risks of Parkinson (296), schizophrenia (297), and also interacts with ApoE alleles for
earlier AD onset (298), GRIN variants might also alter AirP neurodegenerative
responses, e.g. mutations in GRIN2a and GRIN2b increased the risk of cognitive
impairments for lead poisoning of children (299). The nPM responses of Grin2a and
Grin2b mRNA, while modest (p = 0.05-0.06), merit further study among xenobiotics.
AD-associated genes responded differently to nPM by sex and ApoE allele. E3
females had the lowest baseline and the highest nPM response in genes associated with
amyloidogenesis and tau. The smaller responses of ApoE4 mice to nPM may be a ceiling
effect, because about 7% of the nPM responsive genes also had E4 baseline differences
in both sexes. For non-exposed controls (baseline), E4 vs E3 differed in 300 genes related
to known ApoE pathways including LXR/RXR (300), Atherosclerosis (301), and
Rheumatoid arthritis (302).
A similar ApoE allele specificity was found in responses to ozone (O3), which
impaired memory in E3 and not E4 mice (286). The ApoE3-specific gene expression
suggested that amyloid clearance, phagocytosis, and proteolysis differ by ApoE allele.
The ApoE3 & 4 proteins and Ab compete differentially for uptake by astrocytes via the
lipoprotein receptor-related protein 1 (LRP1)-dependent pathway (303). Moreover,
ApoE4 astrocytes have less efficient Ab clearance because of pathogenic acidification of
endosomes (303, 304). Some other nPM-responding genes that mediate amyloid
clearance differed by ApoE allele. The complement factor C3, which mediates amyloid

176
clearance (305), had a higher baseline, but lacked nPM response, whereas Vav and S100
calcium-binding protein A9 (S100A9) responses to nPM were restricted to ApoE3. Vav
and S100A9 regulate phagocytosis of fibrillary Ab by microglia (293, 294).
Similar to the amyloid pathway, E4 had higher baseline level of inflammation, while
their response to nPM was less than the E3. This allele specificity was also shown for
responses to ozone (286). In contrast, responses to LPS endotoxin by intra-peritoneal
injection were greater in E4 male mice for microglial activation and systemic inflammatory
cytokines (306). In vitro, ApoE4 macrophages also had higher induction of NF-kB, TNFa,
and heme oxygenase 1 in response to LPS (307). While AirP can include endotoxins, the
comparisons with injected LPS for sex and ApoE are limited because the systemic
responses are downstream of the lung, which receives most inhalants.
Female mice had greater response to nPM than males for immune and antioxidant
pathways, for wildtype and ApoE-TR mice. Sex hormones and early life gonadal
programming during brain development could underly these differences. Ex-vivo
microglial cultures from male and female brains had divergent inflammatory response to
estradiol (E2) and LPS (308). A gene expression microglial developmental index showed
a sex difference in maturation and immune reactivity, which correlated with the risk of AD
and autism spectrum (309). In SH-SY5Y neuroblastoma cells, E2 increased cell survival
and Nrf2 antioxidant defense against homocysteine (310). Similarly, E2 treatment of
postnatal rats ameliorated acute ethanol-induced oxidative stress, neuroinflammation,
and neuronal cell death through increase of sirt1, P53 acetylation inhibition, and reduction
in phospho-NF-kB nuclear localization (311). Higher adaptive genomic response by
females might allow faster detoxification or recovery from AirP. Metformin mediated NRF2

177
activation could ameliorate tight junction proteins, blood brain barrier (BBB) integrity,
reduce inflammation and oxidative stress, and also normalize the levels of BBB glucose
transporter Glut1 protein after cigarette smoke exposure in mice (312). Future studies
should examine sex differences in the recovery after nPM exposure, including
gonadectomized and older mice to identify the role of steroid hormones.
NF-kB and NRF2 had opposite responses to nPM, that included downstream
genes in wildtype B6 and ApoE-TR. This divergence was also shown in a short term (3
weeks) exposure of B6 male mice. The NRF2 and NF-kB crosstalk was validated in BV2
microglia. This is the first evidence for NRF2 and NF-kB interactions in response to AirP
of both in vivo and in vitro models. These results parallel LPS responses of monocytes,
which showed redox-mediated transcriptional cross-talk between Nrf2 and NF-kB
responses to LPS (85). Concurrent increase of nuclear Nrf2 and cytosolic NF-kB/p65 in
in cerebral cortex after nPM exposure suggest that Nrf2 activation can attenuate NF-kB
nuclear localization. We hypothesize involvement of Keap1, the Nrf2 repressing protein,
which can mediate IKKb degradation and inhibit NF-kB nuclear localization (313, 314).
Other mechanisms could be mediated by direct protein-protein interaction, and by
secondary messengers. For example, NRF2 can inhibit NF-kB through reduction of
reactive species and suppress RAC1 mediated NF-kB activation (315, 316). In contrast,
NF-kB can also inhibit NRF2 activity through enhancing the recruitment of histone
deacetylase (HDAC3) to ARE region (317), or competing with NRF2 for binding to CH1-
KIX domain of CBP protein inside the nucleus (318). These complexities are also shown

178
in nematodes: the C. elegans Nrf2 homologue skn-1 and the anti-bacterial factor 2 (abf-
2) responded rapidly to nPM, with persisting developmental effects (231).
The statistical power to identify all responding genes is intrinsically limited by the high
dimensionality of RNAseq data with >20,000 genes, together with the necessarily small
sample size of animals. Nonetheless small changes in some proteins could be critical for
AirP toxicity. For example, while RNAseq analysis did not detect responses of Nrf2, NF-
kB, and Gclc mRNA at p<0.005, the nPM dose-response experiment showed a 50%
dose-dependent change in Nrf2, NF-kB mRNA, and GCLC protein, which confirmed
findings on the cerebellum (47). There may be other false-negative results in the RNAseq
analysis for sex and ApoE allele.
In summary, AirP neurotoxicity was shown to have sex- and ApoE allele-specificity,
which are main risk factors for AD. These findings give a rationale for including apoE-
gender interactions in epidemiological studies of cognitive aging and dementias.

Acknowledgment
We are grateful to Drs. Arian Saffari and Farimah Shirmohammadi, who helped to
suring nPM collection and animal exposures.

Funding
We are grateful for support from the Cure Alzheimer’s Fund (CEF) and the National
Institute on Aging: CEF (R01-AG051521, P50-AG05142, P01-AG055367); AH (PI: Kelvin
Davis T32- AG052374); MT (PI: Eileen Crimmins T32-AG000037), T.T. (1RF1AG053982-
01A1); M.E.L. (1R01AG057912-01 ; 4R00AG052604-02).

179

Author contributions
Conceptualization, A.H., C.E.F.; Methodology, A.H., M.C., K.R.D., C.D., N.S., M.T.,
H.Z.; nPM collection and characterization: A.S., F.S., C.S.; Investigation, A.H.; Writing,
A.H., M.E.L., T.E.M., H.Z., H.F. and C.E.F.; Supervision, Project Administration, and
Funding Acquisition, C.E.F., T.C.T.

Data availability
NCBI Geo accession number: GSE142066.

Conflict of interest
The authors have no conflict of interest to declare.

180
Supplementary data
Table S1. qPCR Primers used in this study.
Gene Forward Reverse
Nfe2l1 TCTGGCAGTATCTGGAACTT GGGGTTTATAGGCTTTTGTT
mTor ACCTGGATGACTATCTGCAC AACTATTGGGTGAATGATGC
B2m TATGCCAAACCCTCTGTACT AAAAGCAGAAGTAGCCACAG
Foxo3 GGAACTTCACTGGTGCTAAG CTCTGTAGGTCTTCCGTCAG
Apoc1 GCAAAGTGAAGGAGAAGTTG ATTGGTCTGTGATGAAGAGG
Sirt1 GTTTGTACCACCAAATCGTT ATGTGCCACTGTCACTGTTA
Nqo2 GTGAAACACAGGGATTAGGA GGGAAGGTCTCATGTAACAA
Tap1 CTATCAGTTATGTGGCAGCA CAAGGCAAGAGAGAATCAAG
Eomes GCTTCAACATAAACGGACTC GCCAGTGTTAGGAGATTCTG
Ddx3y GACATGATGGAAAGAGGAAA GAGCAAGCATCTGTATCTCC
Xist CAAGTGTGAAAGTGTTGGTG TCCTTATGGGACAGTGACTC
Erdr1 TGCCCTAATTATTCTTGTCC GGTTAGACTTTCCATTCACG
Gapdh CCAATGTGTCCGTCGTGGATCT GTTGAAGTCGCAGGAGACAACC
NF-
kB1
CCAGAAGAGGGTGTCAGAGC ACATTTGCCCAGTTCCGTAG
Nfe2l2 CATAGAGCAGGACATGGAGCAA TCCATTTCTGTCAGTGTGGCTT
IL1b CTAAAGTATGGGCTGGACTG GGCTCTCTTTGAACAGAATG
IL6 TGCCTTCTTGGGACTGATGCT GCATCCATCATTTCTTTGTAT
NOS2 GTCTTGGTGAAAGTGGTGTT GTGCTTGCCTTATACTGGTC
RelA GCGTACACATTCTGGGGAGT ACCGAAGCAGGAGCTATCAA

181

Figure S1. qPCR validation of selected genes from RNAseq, with GAPDH as
reference gene, showing similar direction and scale of response to nPM by qPCR and.
Three-way interaction of qPCR and RNAseq shows high overlap of significant factors.
Mean±SE.
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
Nfe2l1
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
2.0
0.0
0.5
1.0
1.5
2.0
Expression relative to control
B2m
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
E3 E4 E3 E4
Female Male
qPCR RNAseq
0
500
1000
1500
0
2000
4000
6000
8000
Expression relative to control
Xist
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
E3E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
Foxo3
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
Ddx3y
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0
1
2
3
4
5
Expression relative to control
Erdr1
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
mTor
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
Sirt1
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
0.0
0.5
1.0
1.5
Expression relative to control
Nqo2
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0
1
2
3
4
0
1
2
3
4
Expression relative to control
Apoc1
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0.0
0.5
1.0
1.5
2.0
0.0
0.5
1.0
1.5
2.0
Expression relative to control
Tap1
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
0
2
4
6
8
0
2
4
6
8
Expression relative to control
Eomes
E3 E4 E3 E4
Female Male
RPKM expression/E3M control
E3 E4 E3 E4
Female Male
qPCR RNAseq
Control
nPM

182

Figure S2. Top canonical pathways and potential upstream regulators of nPM
associated genes in C57BL/6J mouse

183

Figure S3. nPM associated changes in male and female ApoE-TR mouse. A) Venn
diagram of the nPM DEGs in each group. B) Brain RNAseq cell type deconvolution using
BRETICEA package in R. (C) top upstream regulators and (D) canonical pathways of
nPM associated genes in ApoE-TR mouse.

184

Figure S4. Upstream regulators and canonical pathways associated with E4 baseline
difference in male and female ApoE-TR mouse.

185
Chapter 8. Female vulnerability to the effects of smoking on health
outcomes in older people

Authors: Amin Haghani
1*
, Thalida Em Arpawong
1
, Jung Ki Kim
1
, Juan Pablo
Lewinger
2
, Caleb E Finch
1
, Eileen Crimmins
1

1, Leonard Davis School of Gerontology, University of Southern California, Los Angeles,
CA.
2, Department of Preventive Medicine, University of Southern California, Los Angeles,
CA.

*Corresponding author. Email ahaghani@usc.edu

186
Abstract
Cigarette smoking is among the leading risk factors for mortality and morbidity. While
men have higher smoking prevalence, mechanistic experiments suggest that women are
at higher risk for health problems due to smoking. To address whether women are more
vulnerable to biological effects of smoking, we assess differences in smoking effects on
health outcomes by gender using data from the U.S. Health and Retirement Study (HRS).
The HRS is a population-representative cohort of individuals aged 50+ (n=22,708, 1992-
2014). Men had more pack-years of active smoking than women (22.0 vs 15.6 average
pack-years). Age of death, onset of lung disorders, heart disease, stroke, and cancer
showed dose-dependent effects of smoking for both genders. Among heavy smokers,
>28 pack-years, women were more vulnerable to smoke exposure for earlier age of death
(HR=1.3, 95%CI:1.03-1.65) and stroke (HR=1.37, 95%CI:1.02-1.83). Risk of cancer and
heart disease did not differ by gender for smokers. Women smokers were at lower risk
for lung disorders (HR=0.44-0.56, P<0.02) compared to men smokers across the pack-
year range, relative to non-smokers of each gender. Passive smoke exposure for non-
smokers increased risk of earlier heart disease (HR=1.33, 95%CI:1.07-1.65) and stroke
(HR:1.54, 95%CI:1.07-2.22), mainly in men. In sum, our results suggest that women are
more vulnerable to active smoking for earlier death and risk of stroke, but they were less
vulnerable for lung disorders. From an epidemiological perspective, gender differences in
smoking effects are important considerations that could underlie gender differences in
health outcomes. These findings also inform future experiments that could be designed
to test causality and resolve potential mechanisms for sex-specific cigarette smoke
toxicity.

187

Keywords: Cigarette, active smoke, passive smoke, gender, aging diseases

1. Introduction
Smoking is a leading global cause of mortality (6.5 million excess deaths) (319). While
men exhibit higher prevalence of smoking compared to women, studies have shown that
women smokers have worse outcomes. For instance, women smokers with lung cancer
show higher DNA adducts and mutation in P53 gene (320, 321). Sex-specific biological
effects of cigarette smoke are supported by experimental studies in mice. Chronically
exposed female mice showed significantly greater deficits than males in lung airway
remodeling, increases in biomarkers for oxidative stress, inflammation, (322) and allergic
reactivity (323). At an epidemiological level, active female smokers have shown higher
risk for coronary heart disease (324), stroke (325), lung cancer (320, 321), bladder cancer
(326), and chronic obstructive pulmonary disease (327). However, there are mixed
findings in epidemiological studies (328, 329). Such discrepancies may be attributable to
differences in the method for quantification of smoking exposure, use of cross-sectional
data, small sample sizes, and use of younger cohorts when studying aging-related
diseases.
Smoking hazards are dose-dependent (330), but we lack clear information on gender
differences in the dose-dependence (326). Some studies have found excess risk for men
in all-cause mortality when cigarette dosage was quantified as the number of current

188
cigarettes smoked per day (331). In contrast, other reports on both genders did not show
a clear gender difference (332, 333).
Gender-specific health outcomes are even more obscure for individuals exposed to
second-hand smoke, which is estimated to cause 650,000 deaths globally (330). The few
studies examining gender differences in passive smoking effects are inconsistent. For
example, men were more vulnerable than women to passive smoking effects on risk of
stroke in one study (334), whereas the opposite was reported in others (335, 336).
The current study examined the gender-specific smoking effects on aging-related
health outcomes in the U.S. Health and Retirement Study (HRS), a large nationally
representative U.S. aging cohort that has surveyed participants for more than 22 years.
To more precisely estimate smoking dosage (vs. number of current cigarettes smoked
per day), we calculated a pack-years index to represent the lifetime smoking exposure
for each individual. We also examined passive smoking impacts on non-smokers with
active smoking spouses. We focus on how gender alters smoking hazards on the age of
death and the onset of lung disorders, heart disease, stroke, and cancer. Evaluating these
multiple outcomes let us to examine the disease-specificity of smoking and the gender
interaction. Findings on gender and disease specificity are discussed for potential
biological mechanisms.
2. Methods
2.1. Study population
Participants were a part of the 1992-2014 waves of the HRS, which is a nationally
representative, longitudinal study of health and aging in the United States including adults

189
(50+) and their spouses (337). The current study used 12 waves of data, from 1992
through 2014. Respondent information was obtained from the RAND 2014 HRS datafile,
in addition to HRS Core data files for each wave.
2.2. Outcome variables
Age of death was computed from the year of death variable in the RAND file (radyear),
which is based on the National Death Index and exit interviews with proxy respondents.
Two variables were constructed for each of the health conditions: prevalence, which is a
binary variable for having been diagnosed with the condition; and age of onset, which is
the earliest reported age of the health problem. The incidence of specific diseases was
based on the question “whether or not a doctor has told the respondent that s/he had
these conditions”. The age of onset was extracted from the responses to “In what year
(when) did you have or were diagnosed with the condition”. For individuals with no prior
history of the condition, the age at the wave of incidence was considered as the first age
of onset for the condition. The health conditions include: 1) Lung disorders except for
asthma such as chronic bronchitis and emphysema; 2) Heart disease including heart
attack, coronary heart disease, angina, congestive heart failure or other heart problems;
3) Stroke or transient ischemic attacks (TIA); and 4) Cancer which included any kind of
cancer or malignant tumor, except for skin cancer.
2.3. Predictor variables
Lifetime exposure to smoking is indexed as lifetime pack-years smoked. The pack-
year variable is calculated as the multiplicand of reported average daily cigarette packs
by lifetime years of smoking. Briefly, the earliest age of smoking was extracted from

190
responses to “How many years ago”, “what year”, or “what age did you start smoking?”
The age of smoking cessation was extracted from questions on “How many years ago”,
“what year”, or “what age did you stop smoking?” The earliest age reported for starting
and the latest age for cessation were used for each individual to calculate total years of
smoking. Cigarettes smoked per day were calculated from both the average of the
reported number of cigarettes per day at each wave for each individual, and the maximum
number of cigarettes smoked during the time in which the individual reported smoking the
most. The average of the variables was imputed for smokers with no reported age of start,
former smokers with unknown age of cessation, and active smokers who did not report
the number of cigarettes per day. The continuous pack-year variable (multiplicand of
average daily packs and years of smoking) was then classified into dosage quartiles for
the analysis. Passive smokers are defined as never smoking who lived with at least one
smoker spouse.
The demographic characteristics for gender, race (White/Caucasian, Black/African
American, Other), and ethnicity (Hispanic/non-Hispanic) were extracted from the HRS
RAND file. The ethnicity variable was constructed from the self-reported race and
ethnicity as follows: White (non-Hispanic White), African American (non-Hispanic African
American), Hispanic, and Other (non-Hispanic others).
2.4. Statistical analysis
Hazard ratios (HRs) of sex, smoking pack-years, passive smoking were calculated
using Cox proportional hazard modeling for age of death, and onset of health conditions.
The models estimated time after age 50 to event, and included an interaction term for
gender and smoking to evaluate differences between men and women. All models were

191
adjusted for ethnicity. The HRs were also calculated in sex-stratified models for active
and passive smoking effects by gender. The restricted mean survival time (RMST) of
each group was calculated from the Cox proportional hazard model. Survival curves were
fitted for the gender stratified data on the Cox models that included a gender-smoking
interaction term to estimate RMST of each group. The RMST can be interpreted as the
average of event-free survival time from 50 to 85 years old age that is adjusted for
ethnicity (338-340). The analysis was done in R (version 3.5.3), using the survival
package. The Cox-proportional hazard formula is:
ℎ(𝑡) = ℎ
!
(𝑡)exp (b1𝑋1…+b𝑝𝑋𝑝)
where h(t) represents expected hazard at age t; the h0(t) is the baseline hazard when
all of the predictors are 0; b, coefficients; X, the predictors which included gender,
ethnicity, different categories of pack-years (or passive smoke), and gender interaction
with each pack-year categories (or passive smoke).
3. Results
Demographics of the HRS sample with pack-year categories are in Table 1. The 22
years of the study included 22,708 age-eligible individuals, ages 50-85 years. Men and
women had similar age (mean baseline age, 66) and were balanced for most variables,
with exceptions of a female excess for passive smokers (10.7% of women vs. 6.3% of
men), non-smokers (26.8% of women vs. 19.7% of men), and medium smokers with 15-
20 pack-years history (20.0% of women vs. 15.0% of men). Men had a greater percentage
of very high smokers with >28 pack-years (25.7%) than women (13.3%). The health
conditions with the highest and lowest prevalence were heart disease (26.0% in men,
22.0% in women) and lung disorders (5.9% in men, 7.0% in women).

192

Table 1. Demographic characteristics of the HRS sample, 1992-2014

Men Women
N

10945 (48.2%) 11763 (51.8%)
Age, years (range)

66.58 (50-85) 66 (50-85)
Ethnicity

White/Caucasian 7186 (65.7%) 7211 (61.3%)

African American 1983 (18.1%) 2636 (22.4%)

Hispanic 1409 (12.9%) 1546 (13.1%)

Other 367 (3.4%) 370 (3.1%)
Smoking

Never smoker 1458 (13.3%) 1901 (16.2%)

Passive smoker 695 (6.3%) 1255 (10.7%)

Active smoker 8792 (80.3%) 8607 (73.2%)
Pack years Category Range

Non-smoker (0) 2153 (19.7%) 3156 (26.8%)

Low (0.03-15.0) 1904 (17.4%) 2312 (19.7%)

Medium (15.1-20.0) 1640 (15.0%) 2358 (20.0%)

High (20.1-28.0) 2438 (22.3%) 2377 (20.2%)

Very high (28.1-258.0) 2810 (25.7%) 1560 (13.3%)
Disease prevalence

Lung disorders 642 (5.9%) 826 (7.0%)

Heart disease 2848 (26.0%) 2603 (22.1%)

Stroke 1115 (10.2%) 1089 (9.3%)

Cancer 1952 (17.8%) 2116 (18.0%)
Note: reported percentages for each variable are within gender. Pack-years quartiles were calculated
among smokers only.
3.1. Dose-dependent smoking hazard ratios (HR)
Active smokers had consistent dose-dependent HR for earlier death, and earlier
onset of lung disorders, heart disease, and stroke for both men and women. Smoking-
related HR for risk of death and lung disorders were elevated even at the lowest smoking
levels (0.03-15 pack-years) (Table 2). The highest HR from active smoking was observed

193
for lung disorders, which ranged from 3.0 (95%CI 1.85-4.96) to 7.0 (95%CI 4.42-10.92),
with HRs progressively increasing with higher smoking dosage. Increase in smoking
dosage from 15 to >28 pack-years caused a 20% increase in the HR of earlier onset for
heart disease and stroke: an HR=1.2 (95%CI 1.06-1.41) to 1.5 (95%CI 1.34-1.73). While
smoking increased the risk of earlier onset of cancer (1.2, 95%CI 1.07-1.45), there was
no clear pattern of dose-dependence.

Table 2. Hazard ratios of age of death, and age of onset of lung disorders, heart disease,
stroke and cancer according to active smoking and the interaction with gender.
variable level HR (95%CI) Age of death Lung disorders Heart disease Stroke Cancer
gender Men (ref)

Women 0.65 (0.52,0.8)*** 2.83 (1.74,4.6)*** 0.82 (0.71,0.95)** 0.81 (0.64,1.02) 1.14 (0.97,1.34)
ethnicity White/Caucasian (ref)

African American 1.64 (1.53,1.77)*** 0.93 (0.8,1.07) 1.04 (0.97,1.12) 2.06 (1.87,2.27)*** 0.86 (0.79,0.94)**
Hispanic 1.04 (0.93,1.15) 0.79 (0.65,0.97)* 0.8 (0.73,0.88)*** 1.34 (1.16,1.54)*** 0.73 (0.66,0.82)***
other 1.05 (0.86,1.28) 1.34 (0.99,1.81) 1.03 (0.87,1.21) 1.17 (0.89,1.54) 0.73 (0.59,0.91)**
Pack years Non-smokers (ref)

Low 1.38 (1.14,1.67)** 3.03 (1.85,4.96)*** 1.1 (0.95,1.27) 1.21 (0.96,1.53) 1.17 (0.99,1.39)
Medium 1.45 (1.2,1.75)*** 3.07 (1.88,5)*** 1.22 (1.06,1.41)** 1.24 (0.98,1.56) 1.22 (1.03,1.44)*
High 1.43 (1.2,1.71)*** 3.34 (2.09,5.35)*** 1.33 (1.16,1.51)*** 1.24 (1,1.54)* 1.12 (0.96,1.31)
Very high 2.24 (1.89,2.66)*** 6.95
(4.42,10.92)***
1.52 (1.34,1.73)*** 1.52 (1.23,1.87)*** 1.25 (1.07,1.45)**
Women x
Smoking
interaction
Low 1.1 (0.85,1.43) 0.44 (0.25,0.78)** 1.19 (0.98,1.45) 0.96 (0.7,1.31) 0.87 (0.69,1.08)
Medium 1.02 (0.79,1.31) 0.42 (0.24,0.73)** 0.95 (0.79,1.15) 1.01 (0.75,1.38) 0.79 (0.64,0.98)*
High 1.17 (0.92,1.49) 0.45 (0.26,0.76)** 0.9 (0.75,1.08) 1.1 (0.83,1.47) 0.86 (0.7,1.06)
Very high 1.3 (1.03,1.65)* 0.56 (0.34,0.94)* 1.07 (0.89,1.28) 1.37 (1.02,1.83)* 0.98 (0.79,1.2)
Total N 22708 21486 22708 22695 22689
Note: * p < 0.05, ** p < 0.01, *** p < 0.001

3.2. Gender-specific active smoking hazards
Overall, women died at older ages than men and had later onset of specific health
conditions (Table 2). Women had a lower HR for earlier death (HR=0.65, 95%CI 0.52-
0.8) and earlier diagnosis with heart disease (HR=0.82, 95%CI 0.71-0.95). In contrast,

194
women had a higher HR for earlier onset of lung disorders than men (HR=2.83, 95%CI
1.74-4.6).
Gender interactions with active smoking varied by outcome and smoking dosage. For
very heavy smoking (> 28 pack-years) women had a higher HR for earlier death (HR=1.3,
95%CI 1.03-1.65) and earlier stroke (HR=1.37, 95%CI 1.02-1.83) than men (Fig. 1, 2A).
In contrast, women smokers showed a lower risk of lung disorders diagnosed at an earlier
age than men smokers, particularly in low, medium and high smokers. As noted above,
women had a higher baseline risk of lung disorders (HR=2.8, 95%CI 1.74-4.6) than men.
Active smoking and gender did not show a strong interaction on the outcomes of earlier
onset for heart disease or cancer.

Figure 1. Kaplan-Meier survival curves for all-cause mortality by very high level of
active smoking in men and women between 50-85 years old age in HRS.

195

Figure 2. Smoking hazards are modified by gender for earlier age of death, and earlier
onset of lung disorders, and stroke. The hazard ratio from smoking pack-years (A) and
passive smoke (B) are calculated using Cox proportional hazard models separately by
gender. *p-value <0.05 based on Wald test in the models. The significant gender
differences are based on the interaction terms in the full model and are indicated in each
figure. The stratified models were adjusted for ethnicity. The baseline effects of gender
are reported in table 1 and 2.

196
The order of conditions based on earlier RMST age (an indicator of average event-
free age), is as follows: heart disease (men, RMST age 74.2; women, 75.3), cancer (men,
77.8; women, 77.1), death (men, 77.7; women, 79.3), stroke (men, 80.5; women, 82), and
lung disorders (men, 81.1; women, 79.2) (Fig. 3A). RMST analysis showed that very high
smoking (>28 pack-years) has greater impact on women than men for earlier age of death
(4.7 years earlier for women, 4.1 years earlier for men), and stroke (2 years earlier for
women, 1.1 years earlier for men). Men showed greater vulnerability to smoking for the
earlier age of onset for lung disorders, particularly in smokers with less than 28 pack-
years (women, -0.8 years; men, -1.3 years). Very heavy smoking had similar effects on
the earlier onset of heart diseases and cancer in both genders (Heart diseases: women,
-3.3 years, men, -3.1 years; Cancer: women -1.6 years, men -1.5 years).
3.3. Passive smoking hazards
Passive smoking increased the risk of earlier onset of heart disease (HR=1.3, 95%CI
1.07-1.65) and stroke (HR=1.5, 95%CI 1.07-2.22) (Table 3). Age of death and age of
onset of lung disorders, and cancer were not affected by passive smoking. Based on Cox
proportional hazard models, passive smoking effects were modified by gender only for
earlier age of stroke. Only passive smoking males showed an increase in the risk of earlier
age of stroke compared to females (Fig. 2B). The RMST analysis showed 2 years earlier
onset for heart diseases (at around age 75), and 1.1 years earlier onset for stroke (at age
81) only in male passive smokers (Fig. 3B). Heart disease and stroke onset did not show
any baseline differences between male and female never smokers with non-smoker
spouses.

197
Table 3. Hazard ratios of age of death, and age of onset of lung disorders, heart
disease, stroke and cancer according to passive smoke and the interaction with gender.

HR (95%CI) Age of death Lung disorders Heart disease Stroke Cancer
Gender
Men (ref)

Women 0.73 (0.54,0.98) 2.92 (1.51,5.65)** 0.89 (0.73,1.08) 1.02 (0.74,1.41) 1.33 (1.07,1.65)*
Ethnicity
White/Caucasian
(ref)

African American 1.37 (1.05,1.8)* 1.5 (0.95,2.39) 1.2 (1,1.43)* 2.93 (2.24,3.82)*** 0.8 (0.64,0.99)*
Hispanic 1.03 (0.75,1.42) 1.22 (0.72,2.05) 0.72 (0.57,0.9)** 1.48 (1.05,2.09)* 0.76 (0.6,0.96)*
other 0.73 (0.37,1.42) 1.03 (0.37,2.83) 0.61 (0.39,0.96)* 1.16 (0.59,2.28) 0.5 (0.3,0.84)**
Smoking
Never smokers (ref)

Passive Smokers 1 (0.72,1.38) 1.22 (0.5,2.95) 1.33 (1.07,1.65)* 1.54 (1.07,2.22)* 1.26 (0.97,1.64)
Women x passive smoking
interaction
0.71 (0.46,1.1) 0.86 (0.32,2.29) 0.78 (0.58,1.05) 0.52 (0.32,0.84)** 0.69 (0.5,0.96)*
Total N 5309 5201 5309 5305 5298
Note: * p < 0.05, ** p < 0.01, *** p < 0.001

198

Figure 3. Restricted mean survival time (RMST) of death and disease onset in men
and women smokers. The calculated RMST is based on Cox proportional hazard model
of (A) smoking pack-years, and (B) passive smoke. The results are reported as
RMST±SEM. The lines are showing the number of years that was reduced by smoking
from the baseline RMST age in each gender.
4. Discussion
Our findings show dose-dependent gender differences in smoking for premature
mortality and morbidity among a nationally-representative sample of U.S. adults. Women
were more vulnerable to active smoking for premature death and stroke incidence. In

199
contrast, for <28 pack-years, men smokers had earlier lung disorders than women. The
onset age of cancer and heart disease in smokers did not differ by gender, even for very
heavy smokers. Passive smoke exposure indicated excess risk for men for earlier heart
disease and stroke.
These findings on women are consistent with experimental studies. Mice chronically
exposed to cigarette smoke for 6 months had greater responses among females than
males for lung small airway remodeling and increased distal airway resistance with
corresponding differences for biomarkers of oxidative stress (2-fold increase in 3-
nitrotyrosine), and inflammation (1.5-fold increase in Tgfb1) (322). Ovariectomy
ameliorated lung airway remodeling, show the role of sex steroids. In contrast, male mice
of this study had higher induction of genes mediating oxidative stress responses (Nrf2,
Nqo1), and detoxification (Cyp1a1, Cyp1b1).
Prior analysis of older birth cohorts from 1800-1935 in US and European counties
showed increasing excess mortality for men born after 1900
(341)
. Around 30% of this
excess in men’s mortality was attributed to smoking. The data available were restricted
to smoking status (smokers vs non-smokers), for which pack-years were not known. From
this analysis of HRS, we may infer that more of the excess male mortality observed in the
earlier historical cohorts reflects the smoking behaviors of men. As shown for HRS, men
smoked 30% more pack-years than women (Table S1, an average of 22.0 vs 15.6). Of
key relevance, the current findings now provide a characterization of individual smoking
exposure and mortality for those born after 1934.
Others have shown for lung cancer patients that women smokers had 2-fold more
DNA adducts and frameshift mutations in the P53 gene than men smokers (320, 321).

200
This higher DNA damage among women smokers is also associated with a more
accelerated risk of cancer and other age-related chronic diseases (342, 343). While the
current study showed that smoking increased risk for earlier age of cancer, it did not show
a clear gender interaction. In stratified analysis, the risk of an earlier age of cancer onset
was significant only for women smokers with very high pack-years, which is consistent
with the women’s excess of mutations in lung cancer cited above (320, 321). Future
studies should examine gender-smoking interactions for specific cancer types.
Lung disorders in the current study included all chronic lung conditions other than
asthma, such as chronic bronchitis and emphysema. Thus, the assessment of the
outcome includes chronic obstructive pulmonary disease (COPD). North American
regional studies have shown men excess of COPD from cigarette smoking (344-346),
which parallels our finding of greater risk for earlier lung disorders onset in men smokers.
These findings contradict those in mouse models, that showed solely female responses
in airway remodeling due to chronic exposure (322). However, these mouse models do
not represent older human ages. One concern in interpreting these findings on lung
conditions is the potential under-diagnosis of COPD in North America, particularly for
women (347). Further experimental studies may resolve the gender-specific cigarette
smoke effects for the onset of lung disorders at later ages.
Hazards of passive smoke were close to very high smoking exposure for earlier onset
for heart disease and stroke. Importantly, the chemical composition of side-stream
(passive) smoke differs from main-stream (inhaled) smoke, as well as by its 10-100 times
lower density of particles and gasses (348). Thus, it is not surprising that main-stream
and second-hand smoke would diverge in toxicity, and have different gender interactions.

201
In this study, men exposed to passive smoke had a higher risk of heart disease and stroke
than women. In a mouse model of prenatal exposure to passive smoke, males had
greater alteration in adult lung tidal volume (349).
Despite the well-documented magnitude of cigarette smoking hazards and decades
of research on potential carcinogens and other toxins, we have still a surprisingly limited
understanding of gender interactions. This analysis revealed that sex-specific smoking
effects depend on the aging condition. The parallel experimental findings for sex
differences in mice highlight the possibility of broadly shared biological mechanisms.
Further population-level analyses are needed on gender differences in cigarette toxicity
that may be shared with air pollution, including diseases of arteries, lungs, and brain (52).
For example, lung cancer risk scales with pack-years and air pollution levels of PM2.5
independently, while the combination has multiplicative synergies (350). Further
experimental studies of developmental and adult exposure to cigarette smoke could
include mice with transgenes for detoxification gene variants associated with vulnerability
to air pollution, e.g. alleles of the glutathione S-transferase gene GSTP1 (351) and MET
receptor of tyrosine kinase (352).
The current study examined the association of smoking and several aging-related
conditions in a nationally representative dataset with follow-up over two decades and a
robust measure for exposure to different smoking doses over a lifetime. Future analyses
of smoking effects on aging-related conditions should also consider interactions with age,
and also air pollution, which includes both ambient (outdoors) and within households.
Despite the successful decrease of cigarette smoking in developed countries, Asian and

202
African markets for tobacco are still growing, which anticipates the need for new
therapeutic and preventive measures with gender-specificity.

Funding: A.H. is supported by NIH grant T32AG052374 (Kelvin Davis). C.E.F is
grateful for support from the Cure Alzheimer’s Fund and NIA (R01AG051521, P01-
AG055367, and P50AG05142-31).
Acknowledgment: None.
Author contributions: Conceptualization, A.H., T.E.A., E.C., C.E.F.; Statistical
analysis, A.H., T.E.A., E.C., J.K.K., J.P.L.; Writing, A.H., T.E.A. C.E.F. and E.C.;
Supervision, E.C.
Conflict of interest: The authors have no conflict of interest to declare.

203
Chapter 9. Summary and Synthesis
This dissertation comprises of a series of multidisciplinary projects to resolve some
of the complexities of AirP toxicity throughout the lifespan. We showed that AirP hazards
are shaped by the interface of chemical characteristics of the toxicant mixture with the
biological features of the individual. Overlooking these complex interactions misled the
field in underestimating the global burden of AirP on health and mortality. Epidemiologists
rely on only PM mass in their analysis and do not include any other PM characteristics
(e.g. chemical composition, morphology, surface chemistry and reactivity, and PM acidity)
in their analysis due to lack of data availability. Moreover, the interactions with gender
and genetic structure are also not resolved. We showed that sex and genetic background,
particularly ApoE, can largely change the AirP toxicity. Thus, AirP interaction with sex and
ApoE genotype can explain some portion of reported baseline health differences in the
literature. We hope that these novel findings encourage of AirP neurotoxicology field to
improve their framework of thoughts and properly include these complexities in their
analysis. AirP research should be directed for fast translation into regulation and clinical
research.
The chapters 2-3 of this dissertation revolved around the PM chemical characteristics
and the importance of studying PM toxicity as a mixture of toxicants. We developed a
cell-based assay to screen the inflammatory activity of different PM batches and predict
the brain neuroinflammatory responses. Moreover, we showed that PAHs have minimal
contribution to neuroinflammation after a short-term exposure period during gestation or
adulthood. This research should be continued by studying the neurotoxicity of synthetic
nanoparticles with known surface chemistry. Besides, nPM and sPM should be further

204
characterized for surface reactivity and chemistry, PM acidity, and morphology. It is
essential to resolve the contribution of different PM characteristics to AirP neurotoxicity.
This finding can be translated into regulation for widespread data collection and
epidemiological analysis.
Chapters 4-6 discussed the mechanisms that contributed to long-lasting AirP toxicity
from developmental exposure. In C. elegans, we showed that skn-1/Nrf homolog initial
responses at the L1 stage can lead to chronic developmental and lifespan changes. Nrf2
was also identified as a potential regulator of transcriptional changes in the blood and
cerebral cortex of pups after nPM gestational exposure. In chapter 7, we highlighted the
role of Nrf2 in sex and ApoE allele differences in responses to nPM. Resolving the
interactions of Nrf2 and NF-kB during AirP toxicity and also the changes by sex,
developmental stage, and ApoE allele is an interesting research question for further
studies. These transcriptional factors are therapeutic targets for several chronic diseases
such as cancer and AD. Thus, Nrf2 and NF-kB could also serve as intervention targets
for AirP neurotoxicity.
AirP responses were largely sex-specific. Gestational exposure to nPM caused male
excess in long-term changes in depressive behavior and also systemic metabolic
changes. This result paralleled with a higher number of chronic gene expression changes
in the hippocampus of males that had gestational nPM exposure. On the contrary,
females showed consistent higher acute gene responses to both developmental and adult
nPM exposure regardless of the ApoE genotype. Higher acute Nrf2 responses in females
and larger chronic changes in males suggest a female excess in the recovery after AirP
exposure. This might seem contradictory to the higher vulnerability of women smokers

205
for health outcomes. However, two points should be considered in such a comparison.
Firstly, the HRS data is mainly in old and menopaused women, and secondly, most
people in the study are active smokers with limited time for recovery. Thus, sexual
hormones might play an important role in the observed sex differences. Future
experiments should include ovariectomized mice and examine the responses to
developmental and adult nPM exposure.
A key finding of our studies was the link between chronic neuroimmune responses
and long-term phenotypic changes (e.g. behavior and systemic metabolic alterations)
after gestational nPM exposure. Chronic changes in IL17, HMGB1 and TLR4 signaling
could underlie the neurodevelopmental effects of nPM. Experimental knockdown of these
genes can inform us about the mechanisms of AirP mediated neurodevelopmental
changes. It is also important to differentiate between acute and chronic responses of
these pathways. Exposure to nPM can cause epigenetic changes and long-term
expression changes in these inflammatory pathways. In an ongoing experiment, we are
examining the DNA-methylome of nPM exposure during development and adulthood.
This study can inform us about the DNA methylation changes that correlated with
transcription and phenotypic changes. DNA methylation is often stable and can serve as
a good biomarker of AirP neurotoxicity.
In sum, we hope that our finding facilitates several research opportunities that can be
translated into regulation and clinical framework. AirP crisis is worsening in developing
countries. Thus, scientists must find biomarkers that can detect the sub-population at
higher risk of toxicity, and also develop the counteracting measures to reduce the health
burden of these global toxicants.

206
Bibliography
1. Cohen AJ, Brauer M, Burnett R, Anderson HR, Frostad J, Estep K, et al. Estimates
and 25-year trends of the global burden of disease attributable to ambient air
pollution: an analysis of data from the Global Burden of Diseases Study 2015.
Lancet. 2017;389:1907-1918.
2. Lelieveld J, Evans JS, Fnais M, Giannadaki D, Pozzer A. The contribution of outdoor
air pollution sources to premature mortality on a global scale. Nature. 2015;525:367-
371.
3. Heft-Neal S, Burney J, Bendavid E, Burke M. Robust relationship between air quality
and infant mortality in Africa. Nature. 2018;559:254-258.
4. Burnett RT, Pope CA, 3rd, Ezzati M, Olives C, Lim SS, Mehta S, et al. An integrated
risk function for estimating the global burden of disease attributable to ambient fine
particulate matter exposure. Environ Health Perspect. 2014;122:397-403.
5. Di Q, Wang Y, Zanobetti A, Wang Y, Koutrakis P, Choirat C, et al. Air Pollution and
Mortality in the Medicare Population. N Engl J Med. 2017;376:2513-2522.
6. Chen H, Li Q, Kaufman JS, Wang J, Copes R, Su Y, et al. Effect of air quality alerts
on human health: a regression discontinuity analysis in Toronto, Canada. Lancet
Planet Health. 2018;2:e19-e26.
7. Andersen ZJ, Pedersen M, Weinmayr G, Stafoggia M, Galassi C, Jorgensen JT, et
al. Long-term exposure to ambient air pollution and incidence of brain tumor: the
European Study of Cohorts for Air Pollution Effects (ESCAPE). Neuro Oncol.
2018;20:420-432.
8. Weuve J, Puett RC, Schwartz J, Yanosky JD, Laden F, Grodstein F. Exposure to
particulate air pollution and cognitive decline in older women. Arch Intern Med.
2012;172:219-227.
9. Cordier S, Monfort C, Filippini G, Preston-Martin S, Lubin F, Mueller BA, et al.
Parental exposure to polycyclic aromatic hydrocarbons and the risk of childhood
brain tumors: The SEARCH International Childhood Brain Tumor Study. Am J
Epidemiol. 2004;159:1109-1116.
10. Calderon-Garciduenas L, Reed W, Maronpot RR, Henriquez-Roldan C, Delgado-
Chavez R, Calderon-Garciduenas A, et al. Brain inflammation and Alzheimer's-like
pathology in individuals exposed to severe air pollution. Toxicol Pathol.
2004;32:650-658.
11. Mannucci PM, Harari S, Martinelli I, Franchini M. Effects on health of air pollution: a
narrative review. Intern Emerg Med. 2015;10:657-662.

207
12. Tuomisto JT, Wilson A, Evans JS, Tainio M. Uncertainty in mortality response to
airborne fine particulate matter: Combining European air pollution experts. Reliability
Engineering & System Safety. 2008;93:732-744.
13. Morgan TE, Davis DA, Iwata N, Tanner JA, Snyder D, Ning Z, et al. Glutamatergic
neurons in rodent models respond to nanoscale particulate urban air pollutants in
vivo and in vitro. Environ Health Perspect. 2011;119:1003-1009.
14. Sun X, Luo X, Zhao C, Chung Ng RW, Lim CE, Zhang B, et al. The association
between fine particulate matter exposure during pregnancy and preterm birth: a
meta-analysis. BMC Pregnancy Childbirth. 2015;15:300.
15. Sun X, Luo X, Zhao C, Zhang B, Tao J, Yang Z, et al. The associations between
birth weight and exposure to fine particulate matter (PM2.5) and its chemical
constituents during pregnancy: A meta-analysis. Environ Pollut. 2016;211:38-47.
16. Lam J, Sutton P, Kalkbrenner A, Windham G, Halladay A, Koustas E, et al. A
Systematic Review and Meta-Analysis of Multiple Airborne Pollutants and Autism
Spectrum Disorder. PLoS One. 2016;11:e0161851.
17. Ladd-Acosta C, Feinberg JI, Brown SC, Lurmann FW, Croen LA, Hertz-Picciotto I,
et al. Epigenetic marks of prenatal air pollution exposure found in multiple tissues
relevant for child health. Environ Int. 2019;126:363-376.
18. Kerin T, Volk H, Li W, Lurmann F, Eckel S, McConnell R, et al. Association Between
Air Pollution Exposure, Cognitive and Adaptive Function, and ASD Severity Among
Children with Autism Spectrum Disorder. J Autism Dev Disord. 2018;48:137-150.
19. Volk HE, Lurmann F, Penfold B, Hertz-Picciotto I, McConnell R. Traffic-related air
pollution, particulate matter, and autism. JAMA Psychiatry. 2013;70:71-77.
20. Volk HE, Hertz-Picciotto I, Delwiche L, Lurmann F, McConnell R. Residential
proximity to freeways and autism in the CHARGE study. Environ Health Perspect.
2011;119:873-877.
21. Fuertes E, Standl M, Forns J, Berdel D, Garcia-Aymerich J, Markevych I, et al.
Traffic-related air pollution and hyperactivity/inattention, dyslexia and dyscalculia in
adolescents of the German GINIplus and LISAplus birth cohorts. Environ Int.
2016;97:85-92.
22. Breton CV, Mack WJ, Yao J, Berhane K, Amadeus M, Lurmann F, et al. Prenatal Air
Pollution Exposure and Early Cardiovascular Phenotypes in Young Adults. PLoS
One. 2016;11:e0150825.
23. van Rossem L, Rifas-Shiman SL, Melly SJ, Kloog I, Luttmann-Gibson H, Zanobetti
A, et al. Prenatal air pollution exposure and newborn blood pressure. Environ Health
Perspect. 2015;123:353-359.

208
24. Ghosh JK, Heck JE, Cockburn M, Su J, Jerrett M, Ritz B. Prenatal exposure to traffic-
related air pollution and risk of early childhood cancers. Am J Epidemiol.
2013;178:1233-1239.
25. Ha S, Yeung E, Bell E, Insaf T, Ghassabian A, Bell G, et al. Prenatal and early life
exposures to ambient air pollution and development. Environ Res. 2019.
26. Guxens M, Lubczynska MJ, Muetzel RL, Dalmau-Bueno A, Jaddoe VWV, Hoek G,
et al. Air Pollution Exposure During Fetal Life, Brain Morphology, and Cognitive
Function in School-Age Children. Biol Psychiatry. 2018;84:295-303.
27. Peterson BS, Rauh VA, Bansal R, Hao X, Toth Z, Nati G, et al. Effects of prenatal
exposure to air pollutants (polycyclic aromatic hydrocarbons) on the development of
brain white matter, cognition, and behavior in later childhood. JAMA Psychiatry.
2015;72:531-540.
28. Stingone JA, McVeigh KH, Claudio L. Early-life exposure to air pollution and greater
use of academic support services in childhood: a population-based cohort study of
urban children. Environ Health. 2017;16:2.
29. Jedrychowski WA, Perera FP, Camann D, Spengler J, Butscher M, Mroz E, et al.
Prenatal exposure to polycyclic aromatic hydrocarbons and cognitive dysfunction in
children. Environ Sci Pollut Res Int. 2015;22:3631-3639.
30. Cowell WJ, Bellinger DC, Coull BA, Gennings C, Wright RO, Wright RJ. Associations
between Prenatal Exposure to Black Carbon and Memory Domains in Urban
Children: Modification by Sex and Prenatal Stress. PLoS One. 2015;10:e0142492.
31. Chiu YH, Hsu HH, Coull BA, Bellinger DC, Kloog I, Schwartz J, et al. Prenatal
particulate air pollution and neurodevelopment in urban children: Examining
sensitive windows and sex-specific associations. Environ Int. 2016;87:56-65.
32. Bolton JL, Smith SH, Huff NC, Gilmour MI, Foster WM, Auten RL, et al. Prenatal air
pollution exposure induces neuroinflammation and predisposes offspring to weight
gain in adulthood in a sex-specific manner. FASEB J. 2012;26:4743-4754.
33. Bolton JL, Huff NC, Smith SH, Mason SN, Foster WM, Auten RL, et al. Maternal
stress and effects of prenatal air pollution on offspring mental health outcomes in
mice. Environ Health Perspect. 2013;121:1075-1082.
34. Lertxundi A, Andiarena A, Martínez MD, Ayerdi M, Murcia M, Estarlich M, et al.
Prenatal exposure to PM2.5 and NO2 and sex-dependent infant cognitive and motor
development. Environmental Research. 2019.
35. Block ML, Calderon-Garciduenas L. Air pollution: mechanisms of neuroinflammation
and CNS disease. Trends Neurosci. 2009;32:506-516.

209
36. Calderon-Garciduenas L, Solt AC, Henriquez-Roldan C, Torres-Jardon R, Nuse B,
Herritt L, et al. Long-term air pollution exposure is associated with
neuroinflammation, an altered innate immune response, disruption of the blood-
brain barrier, ultrafine particulate deposition, and accumulation of amyloid beta-42
and alpha-synuclein in children and young adults. Toxicol Pathol. 2008;36:289-310.
37. Calderon-Garciduenas L, Mora-Tiscareno A, Ontiveros E, Gomez-Garza G,
Barragan-Mejia G, Broadway J, et al. Air pollution, cognitive deficits and brain
abnormalities: a pilot study with children and dogs. Brain Cogn. 2008;68:117-127.
38. Cacciottolo M, Wang X, Driscoll I, Woodward N, Saffari A, Reyes J, et al. Particulate
air pollutants, APOE alleles and their contributions to cognitive impairment in older
women and to amyloidogenesis in experimental models. Transl Psychiatry.
2017;7:e1022.
39. Small SA, Schobel SA, Buxton RB, Witter MP, Barnes CA. A pathophysiological
framework of hippocampal dysfunction in ageing and disease. Nat Rev Neurosci.
2011;12:585-601.
40. Woodward NC, Pakbin P, Saffari A, Shirmohammadi F, Haghani A, Sioutas C, et al.
Traffic-related air pollution impact on mouse brain accelerates myelin and neuritic
aging changes with specificity for CA1 neurons. Neurobiol Aging. 2017;53:48-58.
41. Ruckerl R, Ibald-Mulli A, Koenig W, Schneider A, Woelke G, Cyrys J, et al. Air
pollution and markers of inflammation and coagulation in patients with coronary
heart disease. Am J Respir Crit Care Med. 2006;173:432-441.
42. Hoffmann B, Moebus S, Dragano N, Stang A, Mohlenkamp S, Schmermund A, et
al. Chronic residential exposure to particulate matter air pollution and systemic
inflammatory markers. Environ Health Perspect. 2009;117:1302-1308.
43. Chuang KJ, Chan CC, Su TC, Lee CT, Tang CS. The effect of urban air pollution on
inflammation, oxidative stress, coagulation, and autonomic dysfunction in young
adults. Am J Respir Crit Care Med. 2007;176:370-376.
44. Anderson JO, Thundiyil JG, Stolbach A. Clearing the air: a review of the effects of
particulate matter air pollution on human health. J Med Toxicol. 2012;8:166-175.
45. Woodward NC, Levine MC, Haghani A, Shirmohammadi F, Saffari A, Sioutas C, et
al. Toll-like receptor 4 in glial inflammatory responses to air pollution in vitro and in
vivo. J Neuroinflammation. 2017;14:84.
46. Bolton JL, Marinero S, Hassanzadeh T, Natesan D, Le D, Belliveau C, et al.
Gestational Exposure to Air Pollution Alters Cortical Volume, Microglial Morphology,
and Microglia-Neuron Interactions in a Sex-Specific Manner. Front Synaptic
Neurosci. 2017;9:10.

210
47. Zhang H, Liu H, Davies KJ, Sioutas C, Finch CE, Morgan TE, et al. Nrf2-regulated
phase II enzymes are induced by chronic ambient nanoparticle exposure in young
mice with age-related impairments. Free Radic Biol Med. 2012;52:2038-2046.
48. Sivandzade F, Prasad S, Bhalerao A, Cucullo L. NRF2 and NF-B interplay in
cerebrovascular and neurodegenerative disorders: Molecular mechanisms and
possible therapeutic approaches. Redox Biol. 2019;21:101059.
49. Sandberg M, Patil J, D'Angelo B, Weber SG, Mallard C. NRF2-regulation in brain
health and disease: implication of cerebral inflammation. Neuropharmacology.
2014;79:298-306.
50. Hayden MS, Ghosh S. NF-kappaB, the first quarter-century: remarkable progress
and outstanding questions. Genes Dev. 2012;26:203-234.
51. Gref A, Merid SK, Gruzieva O, Ballereau S, Becker A, Bellander T, et al. Genome-
Wide Interaction Analysis of Air Pollution Exposure and Childhood Asthma with
Functional Follow-up. Am J Respir Crit Care Med. 2017;195:1373-1383.
52. Forman HJ, Finch CE. A critical review of assays for hazardous components of air
pollution. Free Radic Biol Med. 2018;117:202-217.
53. Gellert C, Schottker B, Brenner H. Smoking and all-cause mortality in older people:
systematic review and meta-analysis. Arch Intern Med. 2012;172:837-844.
54. Health UDo, Services H. The health consequences of smoking—50 years of
progress: a report of the Surgeon General. Atlanta, GA: US Department of Health
and Human Services, Centers for Disease …; 2014.
55. Jamal A, King BA, Neff LJ, Whitmill J, Babb SD, Graffunder CM. Current Cigarette
Smoking Among Adults - United States, 2005-2015. MMWR Morb Mortal Wkly Rep.
2016;65:1205-1211.
56. Gatto NM, Henderson VW, Hodis HN, St John JA, Lurmann F, Chen JC, et al.
Components of air pollution and cognitive function in middle-aged and older adults
in Los Angeles. Neurotoxicology. 2014;40:1-7.
57. Chen H, Kwong JC, Copes R, Tu K, Villeneuve PJ, van Donkelaar A, et al. Living
near major roads and the incidence of dementia, Parkinson's disease, and multiple
sclerosis: a population-based cohort study. Lancet. 2017;389:718-726.
58. Fonken LK, Xu X, Weil ZM, Chen G, Sun Q, Rajagopalan S, et al. Air pollution
impairs cognition, provokes depressive-like behaviors and alters hippocampal
cytokine expression and morphology. Mol Psychiatry. 2011;16:987-995, 973.

211
59. Morgan TE, Davis DA, Iwata N, Tanner JA, Snyder D, Ning Z, et al. Glutamatergic
Neurons in Rodent Models Respond to Nanoscale Particulate Urban Air Pollutants
in Vivo and in Vitro. Environmental Health Perspectives. 2011;119:1003-1009.
60. Bhatt DP, Puig KL, Gorr MW, Wold LE, Combs CK. A pilot study to assess effects
of long-term inhalation of airborne particulate matter on early Alzheimer-like changes
in the mouse brain. PLoS One. 2015;10:e0127102.
61. Landrigan PJ. Air pollution and health. Lancet Public Health. 2017;2:e4-e5.
62. Daher N, Hasheminassaba S, Shafer MM, Schauer JJ, Sioutas C. Seasonal and
spatial variability in chemical composition and mass closure of ambient ultrafine
particles in the megacity of Los Angeles. Environ Sci Process Impacts. 2013;15:283-
295.
63. Heo J, de Foy B, Olson MR, Pakbin P, Sioutas C, Schauer JJ. Impact of regional
transport on the anthropogenic and biogenic secondary organic aerosols in the Los
Angeles Basin. Atmos Environ. 2015;103:171-179.
64. Heo JB, Dulger M, Olson MR, McGinnis JE, Shelton BR, Matsunaga A, et al. Source
apportionments of PM2.5 organic carbon using molecular marker Positive Matrix
Factorization and comparison of results from different receptor models. Atmospheric
Environment. 2013;73:51-61.
65. Canagaratna MR, Jayne JT, Ghertner DA, Herndon S, Shi Q, Jimenez JL, et al.
Chase studies of particulate emissions from in-use New York City vehicles. Aerosol
Science and Technology. 2004;38:555-573.
66. Charron A, Harrison RM. Primary particle formation from vehicle emissions during
exhaust dilution in the roadside atmosphere. Atmospheric Environment.
2003;37:4109-4119.
67. Kunzli N, Mudway IS, Gotschi T, Shi T, Kelly FJ, Cook S, et al. Comparison of
oxidative properties, light absorbance, total and elemental mass concentration of
ambient PM2.5 collected at 20 European sites. Environ Health Perspect.
2006;114:684-690.
68. Rushton EK, Jiang J, Leonard SS, Eberly S, Castranova V, Biswas P, et al. Concept
of assessing nanoparticle hazards considering nanoparticle dosemetric and
chemical/biological response metrics. J Toxicol Environ Health A. 2010;73:445-461.
69. Seagrave J, McDonald JD, Gigliotti AP, Nikula KJ, Seilkop SK, Gurevich M, et al.
Mutagenicity and in vivo toxicity of combined particulate and semivolatile organic
fractions of gasoline and diesel engine emissions. Toxicol Sci. 2002;70:212-226.

212
70. Sayes CM, Reed KL, Warheit DB. Assessing toxicity of fine and nanoparticles:
comparing in vitro measurements to in vivo pulmonary toxicity profiles. Toxicol Sci.
2007;97:163-180.
71. Kim YH, Boykin E, Stevens T, Lavrich K, Gilmour MI. Comparative lung toxicity of
engineered nanomaterials utilizing in vitro, ex vivo and in vivo approaches. J
Nanobiotechnology. 2014;12:47.
72. Teeguarden JG, Mikheev VB, Minard KR, Forsythe WC, Wang W, Sharma G, et al.
Comparative iron oxide nanoparticle cellular dosimetry and response in mice by the
inhalation and liquid cell culture exposure routes. Part Fibre Toxicol. 2014;11:46.
73. Donaldson K, Schinwald A, Murphy F, Cho WS, Duffin R, Tran L, et al. The
biologically effective dose in inhalation nanotoxicology. Acc Chem Res.
2013;46:723-732.
74. Loret T, Rogerieux F, Trouiller B, Braun A, Egles C, Lacroix G. Predicting the in vivo
pulmonary toxicity induced by acute exposure to poorly soluble nanomaterials by
using advanced in vitro methods. Part Fibre Toxicol. 2018;15:25.
75. Lawrence T. The nuclear factor NF-kappaB pathway in inflammation. Cold Spring
Harbor perspectives in biology. 2009;1:a001651.
76. Misra C, Kim S, Shen S, Sioutas C. A high flow rate, very low pressure drop impactor
for inertial separation of ultrafine from accumulation mode particles. J Aerosol Sci.
2002;33:735-752.
77. Taghvaee S, Mousavi A, Sowlat MH, Sioutas C. Development of a novel aerosol
generation system for conducting inhalation exposures to ambient particulate matter
(PM). The Science of the total environment. 2019;665:1035-1045.
78. Kim S, Jaques PA, Chang MC, Barone T, Xiong C, Friedlander SK, et al. Versatile
aerosol concentration enrichment system (VACES) for simultaneous in vivo and in
vitro evaluation of toxic effects of ultrafine, fine and coarse ambient particles - Part
II: Field evaluation. J Aerosol Sci. 2001;32:1299-1314.
79. Kim S, Jaques PA, Chang MC, Froines JR, Sioutas C. Versatile aerosol
concentration enrichment system (VACES) for simultaneous in vivo and in vitro
evaluation of toxic effects of ultrafine, fine and coarse ambient particles - Part I:
Development and laboratory characterization. J Aerosol Sci. 2001;32:1281-1297.
80. Wang DB, Pakbin P, Saffari A, Shafer MM, Schauer JJ, Sioutas C. Development
and Evaluation of a High-Volume Aerosol-into-Liquid Collector for Fine and Ultrafine
Particulate Matter. Aerosol Sci Tech. 2013;47:1226-1238.
81. Herner JD, Green PG, Kleeman MJ. Measuring the trace elemental composition of
size-resolved airborne particles. Environ Sci Technol. 2006;40:1925-1933.

213
82. Stone EA, Hedman CJ, Sheesley RJ, Shafer MM, Schauer JJ. Investigating the
chemical nature of humic-like substances (HULIS) in North American atmospheric
aerosols by liquid chromatography tandem mass spectrometry. Atmos Environ.
2009;43:4205-4213.
83. Puente BN, Kimura W, Muralidhar SA, Moon J, Amatruda JF, Phelps KL, et al. The
oxygen-rich postnatal environment induces cardiomyocyte cell-cycle arrest through
DNA damage response. Cell. 2014;157:565-579.
84. Ignarro LJ, Fukuto JM, Griscavage JM, Rogers NE, Byrns RE. Oxidation of nitric
oxide in aqueous solution to nitrite but not nitrate: comparison with enzymatically
formed nitric oxide from L-arginine. Proc Natl Acad Sci U S A. 1993;90:8103-8107.
85. Zhang H, Liu H, Zhou L, Yuen J, Forman HJ. Temporal changes in glutathione
biosynthesis during the lipopolysaccharide-induced inflammatory response of THP-
1 macrophages. Free Radic Biol Med. 2017;113:304-310.
86. Cacciottolo M, Christensen A, Moser A, Liu J, Pike CJ, Smith C, et al. The APOE4
allele shows opposite sex bias in microbleeds and Alzheimer's disease of humans
and mice. Neurobiology of aging. 2016;37:47-57.
87. Duff GW, Atkins E. The inhibitory effect of polymyxin B on endotoxin-induced
endogenous pyrogen production. J Immunol Methods. 1982;52:333-340.
88. Weinberg JB, Misukonis MA, Shami PJ, Mason SN, Sauls DL, Dittman WA, et al.
Human mononuclear phagocyte inducible nitric oxide synthase (iNOS): analysis of
iNOS mRNA, iNOS protein, biopterin, and nitric oxide production by blood
monocytes and peritoneal macrophages. Blood. 1995;86:1184-1195.
89. Di Pietro A, Visalli G, Munao F, Baluce B, La Maestra S, Primerano P, et al. Oxidative
damage in human epithelial alveolar cells exposed in vitro to oil fly ash transition
metals. Int J Hyg Environ Health. 2009;212:196-208.
90. Okimotoa Y, Watanabea A, Nikia E, Yamashitab T, Noguchia N. A novel fluorescent
probe diphenyl-1-pyrenylphosphine to follow lipid peroxidation in cell membranes.
FEBS Lett. 2000;474:137-140.
91. Zhang H, Forman HJ. Signaling by 4-hydroxy-2-nonenal: Exposure protocols, target
selectivity and degradation. Archives of biochemistry and biophysics. 2017;617:145-
154.
92. Lovett C, Cacciottolo M, Shirmohammadi F, Haghani A, Morgan TE, Sioutas C, et
al. Diurnal variation in the proinflammatory activity of urban fine particulate matter
(PM 2.5) by in vitro assays. F1000Res. 2018;7:596.
93. Nakayama M. Macrophage Recognition of Crystals and Nanoparticles. Front
Immunol. 2018;9:103.

214
94. Ovrevik J, Refsnes M, Lag M, Brinchmann BC, Schwarze PE, Holme JA. Triggering
Mechanisms and Inflammatory Effects of Combustion Exhaust Particles with
Implication for Carcinogenesis. Basic Clin Pharmacol Toxicol. 2017;121 Suppl 3:55-
62.
95. Chao CC, Hu S, Molitor TW, Shaskan EG, Peterson PK. Activated microglia mediate
neuronal cell injury via a nitric oxide mechanism. J Immunol. 1992;149:2736-2741.
96. Woodward NC, Haghani A, Johnson RG, Hsu TM, Saffari A, Sioutas C, et al.
Prenatal and early life exposure to air pollution induced hippocampal vascular
leakage and impaired neurogenesis in association with behavioral deficits. Transl
Psychiatry. 2018;8:261.
97. Cole TB, Coburn J, Dao K, Roque P, Chang YC, Kalia V, et al. Sex and genetic
differences in the effects of acute diesel exhaust exposure on inflammation and
oxidative stress in mouse brain. Toxicology. 2016;374:1-9.
98. Nway NC, Fujitani Y, Hirano S, Mar O, Win-Shwe TT. Role of TLR4 in olfactory-
based spatial learning activity of neonatal mice after developmental exposure to
diesel exhaust origin secondary organic aerosol. Neurotoxicology. 2017;63:155-
165.
99. Durga M, Devasena T, Rajasekar A. Determination of LC50 and sub-chronic
neurotoxicity of diesel exhaust nanoparticles. Environmental toxicology and
pharmacology. 2015;40:615-625.
100. Di Natale C, Scognamiglio PL, Cascella R, Cecchi C, Russo A, Leone M, et al.
Nucleophosmin contains amyloidogenic regions that are able to form toxic
aggregates under physiological conditions. FASEB J. 2015;29:3689-3701.
101. Mumaw CL, Levesque S, McGraw C, Robertson S, Lucas S, Stafflinger JE, et al.
Microglial priming through the lung-brain axis: the role of air pollution-induced
circulating factors. FASEB J. 2016;30:1880-1891.
102. Gavett SH, Madison SL, Dreher KL, Winsett DW, McGee JK, Costa DL. Metal and
sulfate composition of residual oil fly ash determines airway hyperreactivity and lung
injury in rats. Environ Res. 1997;72:162-172.
103. Kodavanti UP, Hauser R, Christiani DC, Meng ZH, McGee J, Ledbetter A, et al.
Pulmonary responses to oil fly ash particles in the rat differ by virtue of their specific
soluble metals. Toxicol Sci. 1998;43:204-212.
104. Prahalad AK, Soukup JM, Inmon J, Willis R, Ghio AJ, Becker S, et al. Ambient air
particles: effects on cellular oxidant radical generation in relation to particulate
elemental chemistry. Toxicol Appl Pharmacol. 1999;158:81-91.

215
105. Hetland RB, Refsnes M, Myran T, Johansen BV, Uthus N, Schwarze PE. Mineral
and/or metal content as critical determinants of particle-induced release of IL-6 and
IL-8 from A549 cells. J Toxicol Environ Health A. 2000;60:47-65.
106. Jiang N, Dreher KL, Dye JA, Li Y, Richards JH, Martin LD, et al. Residual oil fly ash
induces cytotoxicity and mucin secretion by guinea pig tracheal epithelial cells via
an oxidant-mediated mechanism. Toxicol Appl Pharmacol. 2000;163:221-230.
107. Xia T, Kovochich M, Brant J, Hotze M, Sempf J, Oberley T, et al. Comparison of the
abilities of ambient and manufactured nanoparticles to induce cellular toxicity
according to an oxidative stress paradigm. Nano Lett. 2006;6:1794-1807.
108. Vidrio E, Phuah CH, Dillner AM, Anastasio C. Generation of hydroxyl radicals from
ambient fine particles in a surrogate lung fluid solution. Environ Sci Technol.
2009;43:922-927.
109. Saffari A, Daher N, Shafer MM, Schauer JJ, Sioutas C. Global perspective on the
oxidative potential of airborne particulate matter: a synthesis of research findings.
Environ Sci Technol. 2014;48:7576-7583.
110. Guastadisegni C, Kelly FJ, Cassee FR, Gerlofs-Nijland ME, Janssen NA, Pozzi R,
et al. Determinants of the proinflammatory action of ambient particulate matter in
immortalized murine macrophages. Environ Health Perspect. 2010;118:1728-1734.
111. Sun X, Wei H, Young DE, Bein KJ, Smiley-Jewell SM, Zhang Q, et al. Differential
pulmonary effects of wintertime California and China particulate matter in healthy
young mice. Toxicol Lett. 2017;278:1-8.
112. Murugan K, Choonara YE, Kumar P, Bijukumar D, du Toit LC, Pillay V. Parameters
and characteristics governing cellular internalization and trans-barrier trafficking of
nanostructures. Int J Nanomedicine. 2015;10:2191-2206.
113. Premasekharan G, Nguyen K, Contreras J, Ramon V, Leppert VJ, Forman HJ. Iron-
mediated lipid peroxidation and lipid raft disruption in low-dose silica-induced
macrophage cytokine production. Free Radic Biol Med. 2011;51:1184-1194.
114. Cheng H, Saffari A, Sioutas C, Forman HJ, Morgan TE, Finch CE. Nanoscale
Particulate Matter from Urban Traffic Rapidly Induces Oxidative Stress and
Inflammation in Olfactory Epithelium with Concomitant Effects on Brain. Environ
Health Perspect. 2016;124:1537-1546.
115. Becerra TA, Wilhelm M, Olsen J, Cockburn M, Ritz B. Ambient air pollution and
autism in Los Angeles county, California. Environ Health Perspect. 2013;121:380-
386.
116. Paul KC, Haan M, Mayeda ER, Ritz BR. Ambient Air Pollution, Noise, and Late-Life
Cognitive Decline and Dementia Risk. Annu Rev Public Health. 2019;40:203-220.

216
117. Peters R, Ee N, Peters J, Booth A, Mudway I, Anstey KJ. Air Pollution and Dementia:
A Systematic Review. J Alzheimers Dis. 2019;70:S145-S163.
118. Calderon-Garciduenas L, Reynoso-Robles R, Gonzalez-Maciel A. Combustion and
friction-derived nanoparticles and industrial-sourced nanoparticles: The culprit of
Alzheimer and Parkinson's diseases. Environ Res. 2019;176:108574.
119. Liu C, Fonken LK, Wang A, Maiseyeu A, Bai Y, Wang TY, et al. Central IKKbeta
inhibition prevents air pollution mediated peripheral inflammation and exaggeration
of type II diabetes. Part Fibre Toxicol. 2014;11:53.
120. Levesque S, Surace MJ, McDonald J, Block ML. Air pollution & the brain: Subchronic
diesel exhaust exposure causes neuroinflammation and elevates early markers of
neurodegenerative disease. J Neuroinflammation. 2011;8:105.
121. Klocke C, Allen JL, Sobolewski M, Blum JL, Zelikoff JT, Cory-Slechta DA. Exposure
to fine and ultrafine particulate matter during gestation alters postnatal
oligodendrocyte maturation, proliferation capacity, and myelination.
Neurotoxicology. 2018;65:196-206.
122. Davis DA, Bortolato M, Godar SC, Sander TK, Iwata N, Pakbin P, et al. Prenatal
exposure to urban air nanoparticles in mice causes altered neuronal differentiation
and depression-like responses. PLoS One. 2013;8:e64128.
123. Kim HW, Kam S, Lee DH. Synergistic interaction between polycyclic aromatic
hydrocarbons and environmental tobacco smoke on the risk of obesity in children
and adolescents: The U.S. National Health and Nutrition Examination Survey 2003-
2008. Environ Res. 2014;135:354-360.
124. Brown LA, Khousbouei H, Goodwin JS, Irvin-Wilson CV, Ramesh A, Sheng L, et al.
Down-regulation of early ionotrophic glutamate receptor subunit developmental
expression as a mechanism for observed plasticity deficits following gestational
exposure to benzo(a)pyrene. Neurotoxicology. 2007;28:965-978.
125. Wormley DD, Chirwa S, Nayyar T, Wu J, Johnson S, Brown LA, et al. Inhaled
benzo(a)pyrene impairs long-term potentiation in the F1 generation rat dentate
gyrus. Cell Mol Biol (Noisy-le-grand). 2004;50:715-721.
126. McCallister MM, Maguire M, Ramesh A, Aimin Q, Liu S, Khoshbouei H, et al.
Prenatal exposure to benzo(a)pyrene impairs later-life cortical neuronal function.
Neurotoxicology. 2008;29:846-854.
127. Saunders CR, Das SK, Ramesh A, Shockley DC, Mukherjee S. Benzo(a)pyrene-
induced acute neurotoxicity in the F-344 rat: role of oxidative stress. J Appl Toxicol.
2006;26:427-438.

217
128. Irigaray P, Ogier V, Jacquenet S, Notet V, Sibille P, Mejean L, et al. Benzo[a]pyrene
impairs beta-adrenergic stimulation of adipose tissue lipolysis and causes weight
gain in mice. A novel molecular mechanism of toxicity for a common food pollutant.
FEBS J. 2006;273:1362-1372.
129. Irigaray P, Lacomme S, Mejean L, Belpomme D. Ex vivo study of incorporation into
adipocytes and lipolysis-inhibition effect of polycyclic aromatic hydrocarbons.
Toxicol Lett. 2009;187:35-39.
130. Zheng X, Wang X, Wang T, Zhang H, Wu H, Zhang C, et al. Gestational Exposure
to Particulate Matter 2.5 (PM2.5) Leads to Spatial Memory Dysfunction and
Neurodevelopmental Impairment in Hippocampus of Mice Offspring. Front Neurosci.
2018;12:1000.
131. Zhang M, Liu W, Zhou Y, Li Y, Qin Y, Xu Y. Neurodevelopmental toxicity induced by
maternal PM2.5 exposure and protective effects of quercetin and Vitamin C.
Chemosphere. 2018;213:182-196.
132. Klocke C, Allen JL, Sobolewski M, Mayer-Proschel M, Blum JL, Lauterstein D, et al.
Neuropathological Consequences of Gestational Exposure to Concentrated
Ambient Fine and Ultrafine Particles in the Mouse. Toxicol Sci. 2017;156:492-508.
133. Woodward NC, Crow AL, Zhang Y, Epstein S, Hartiala J, Johnson R, et al. Exposure
to Nanoscale Particulate Matter from Gestation to Adulthood Impairs Metabolic
Homeostasis in Mice. Sci Rep. 2019;9:1816.
134. Pirhadi M, Mousavi A, Taghvaee S, Sowlat MH, Sioutas C. An aerosol
concentrator/diffusion battery tandem to concentrate and separate ambient
accumulation mode particles for evaluating their toxicological properties. Atmos
Environ. 2019;213:81-89.
135. Sheesley RJ, Schauer JJ, Chowdhury Z, Cass GR, Simoneit BR. Characterization
of organic aerosols emitted from the combustion of biomass indigenous to South
Asia. Journal of Geophysical Research: Atmospheres. 2003;108.
136. Dinger K, Mohr J, Vohlen C, Hirani D, Hucklenbruch-Rother E, Ensenauer R, et al.
Intraperitoneal Glucose Tolerance Test, Measurement of Lung Function, and
Fixation of the Lung to Study the Impact of Obesity and Impaired Metabolism on
Pulmonary Outcomes. J Vis Exp. 2018.
137. Dimauro I, Pearson T, Caporossi D, Jackson MJ. A simple protocol for the
subcellular fractionation of skeletal muscle cells and tissue. BMC Res Notes.
2012;5:513.
138. Yang TT, Sinai P, Kitts PA, Kain SR. Quantification of gene expression with a
secreted alkaline phosphatase reporter system. Biotechniques. 1997;23:1110-1114.

218
139. He M, Ichinose T, Yoshida Y, Arashidani K, Yoshida S, Takano H, et al. Urban
PM2.5 exacerbates allergic inflammation in the murine lung via a
TLR2/TLR4/MyD88-signaling pathway. Sci Rep. 2017;7:11027.
140. Reich J, Tamura H, Nagaoka I, Motschmann H. Investigation of the kinetics and
mechanism of low endotoxin recovery in a matrix for biopharmaceutical drug
products. Biologicals. 2018;53:1-9.
141. Hochstein HD, Mills DF, Outschoorn AS, Rastogi SC. The processing and
collaborative assay of a reference endotoxin. J Biol Stand. 1983;11:251-260.
142. Kim SH, Knight EM, Saunders EL, Cuevas AK, Popovech M, Chen LC, et al. Rapid
doubling of Alzheimer's amyloid-beta40 and 42 levels in brains of mice exposed to
a nickel nanoparticle model of air pollution. F1000Res. 2012;1:70.
143. Zhang H, Haghani A, Mousavi AH, Cacciottolo M, D'Agostino C, Safi N, et al. Cell-
based assays that predict in vivo neurotoxicity of urban ambient nano-sized
particulate matter. Free Radic Biol Med. 2019;145:33-41.
144. Kendall M, Guntern J, Lockyer NP, Jones FH, Hutton BM, Lippmann M, et al. Urban
PM2.5 surface chemistry and interactions with bronchoalveolar lavage fluid. Inhal
Toxicol. 2004;16 Suppl 1:115-129.
145. Stepien G, Moros M, Perez-Hernandez M, Monge M, Gutierrez L, Fratila RM, et al.
Effect of Surface Chemistry and Associated Protein Corona on the Long-Term
Biodegradation of Iron Oxide Nanoparticles In Vivo. ACS Appl Mater Interfaces.
2018;10:4548-4560.
146. Duffin R, Tran L, Brown D, Stone V, Donaldson K. Proinflammogenic effects of low-
toxicity and metal nanoparticles in vivo and in vitro: highlighting the role of particle
surface area and surface reactivity. Inhal Toxicol. 2007;19:849-856.
147. Lison D, Lardot C, Huaux F, Zanetti G, Fubini B. Influence of particle surface area
on the toxicity of insoluble manganese dioxide dusts. Arch Toxicol. 1997;71:725-
729.
148. Fang T, Guo H, Zeng L, Verma V, Nenes A, Weber RJ. Highly Acidic Ambient
Particles, Soluble Metals, and Oxidative Potential: A Link between Sulfate and
Aerosol Toxicity. Environ Sci Technol. 2017;51:2611-2620.
149. Abdel-Shafy HI, Mansour MSM. A review on polycyclic aromatic hydrocarbons:
Source, environmental impact, effect on human health and remediation. Egyptian
Journal of Petroleum. 2016;25:107-123.
150. Burchiel SW, Luster MI. Signaling by environmental polycyclic aromatic
hydrocarbons in human lymphocytes. Clin Immunol. 2001;98:2-10.

219
151. Ward EC, Murray MJ, Lauer LD, House RV, Irons R, Dean JH. Immunosuppression
following 7,12-dimethylbenz[a]anthracene exposure in B6C3F1 mice. I. Effects on
humoral immunity and host resistance. Toxicol Appl Pharmacol. 1984;75:299-308.
152. Dean JH, Ward EC, Murray MJ, Lauer LD, House RV, Stillman W, et al.
Immunosuppression following 7,12-dimethylbenz[a]anthracene exposure in
B6C3F1 mice--II. Altered cell-mediated immunity and tumor resistance. Int J
Immunopharmacol. 1986;8:189-198.
153. Blanton RH, Lyte M, Myers MJ, Bick PH. Immunomodulation by polyaromatic
hydrocarbons in mice and murine cells. Cancer Res. 1986;46:2735-2739.
154. Brandenburg K, David A, Howe J, Koch MH, Andra J, Garidel P. Temperature
dependence of the binding of endotoxins to the polycationic peptides polymyxin B
and its nonapeptide. Biophys J. 2005;88:1845-1858.
155. Pfalzgraff A, Correa W, Heinbockel L, Schromm AB, Lubow C, Gisch N, et al. LPS-
neutralizing peptides reduce outer membrane vesicle-induced inflammatory
responses. Biochim Biophys Acta Mol Cell Biol Lipids. 2019;1864:1503-1513.
156. Calderon-Garciduenas L, Serrano-Sierra A, Torres-Jardon R, Zhu H, Yuan Y, Smith
D, et al. The impact of environmental metals in young urbanites' brains. Exp Toxicol
Pathol. 2013;65:503-511.
157. Maher BA, Ahmed IA, Karloukovski V, MacLaren DA, Foulds PG, Allsop D, et al.
Magnetite pollution nanoparticles in the human brain. Proc Natl Acad Sci U S A.
2016;113:10797-10801.
158. Oberdorster G, Oberdorster E, Oberdorster J. Nanotoxicology: an emerging
discipline evolving from studies of ultrafine particles. Environ Health Perspect.
2005;113:823-839.
159. Ibanez C, Suhard D, Elie C, Ebrahimian T, Lestaevel P, Roynette A, et al. Evaluation
of the Nose-to-Brain Transport of Different Physicochemical Forms of Uranium after
Exposure via Inhalation of a UO4 Aerosol in the Rat. Environ Health Perspect.
2019;127:97010.
160. Li D, Morishita M, Wagner JG, Fatouraie M, Wooldridge M, Eagle WE, et al. In vivo
biodistribution and physiologically based pharmacokinetic modeling of inhaled fresh
and aged cerium oxide nanoparticles in rats. Part Fibre Toxicol. 2016;13:45.
161. Aragon MJ, Topper L, Tyler CR, Sanchez B, Zychowski K, Young T, et al. Serum-
borne bioactivity caused by pulmonary multiwalled carbon nanotubes induces
neuroinflammation via blood-brain barrier impairment. Proc Natl Acad Sci U S A.
2017;114:E1968-E1976.

220
162. Ho D, Leong JW, Crew RC, Norret M, House MJ, Mark PJ, et al. Maternal-placental-
fetal biodistribution of multimodal polymeric nanoparticles in a pregnant rat model in
mid and late gestation. Sci Rep. 2017;7:2866.
163. Yan W, Ku T, Yue H, Li G, Sang N. NO2 inhalation causes tauopathy by disturbing
the insulin signaling pathway. Chemosphere. 2016;165:248-256.
164. Valdez MC, Freeborn D, Valdez JM, Johnstone AFM, Snow SJ, Tennant AH, et al.
Mitochondrial Bioenergetics in Brain Following Ozone Exposure in Rats Maintained
on Coconut, Fish and Olive Oil-Rich Diets. Int J Mol Sci. 2019;20:6303.
165. Tyler CR, Noor S, Young TL, Rivero V, Sanchez B, Lucas S, et al. Aging
Exacerbates Neuroinflammatory Outcomes Induced by Acute Ozone Exposure.
Toxicol Sci. 2018;163:123-139.
166. Sang N, Yun Y, Li H, Hou L, Han M, Li G. SO2 inhalation contributes to the
development and progression of ischemic stroke in the brain. Toxicol Sci.
2010;114:226-236.
167. Casanova R, Wang X, Reyes J, Akita Y, Serre ML, Vizuete W, et al. A Voxel-Based
Morphometry Study Reveals Local Brain Structural Alterations Associated with
Ambient Fine Particles in Older Women. Front Hum Neurosci. 2016;10:495.
168. Wolf K, Popp A, Schneider A, Breitner S, Hampel R, Rathmann W, et al. Association
Between Long-term Exposure to Air Pollution and Biomarkers Related to Insulin
Resistance, Subclinical Inflammation, and Adipokines. Diabetes. 2016;65:3314-
3326.
169. Hamra GB, Guha N, Cohen A, Laden F, Raaschou-Nielsen O, Samet JM, et al.
Outdoor particulate matter exposure and lung cancer: a systematic review and meta-
analysis. Environ Health Perspect. 2014;122:906-911.
170. Sorrentino JA, Sanoff HK, Sharpless NE. Defining the toxicology of aging. Trends
Mol Med. 2014;20:375-384.
171. Cheng H, Davis DA, Hasheminassab S, Sioutas C, Morgan TE, Finch CE. Urban
traffic-derived nanoparticulate matter reduces neurite outgrowth via TNFalpha in
vitro. J Neuroinflammation. 2016;13:19.
172. Leung MC, Williams PL, Benedetto A, Au C, Helmcke KJ, Aschner M, et al.
Caenorhabditis elegans: an emerging model in biomedical and environmental
toxicology. Toxicol Sci. 2008;106:5-28.
173. Kaletta T, Hengartner MO. Finding function in novel targets: C. elegans as a model
organism. Nat Rev Drug Discov. 2006;5:387-398.

221
174. Antoshechkin I, Sternberg PW. The versatile worm: genetic and genomic resources
for Caenorhabditis elegans research. Nat Rev Genet. 2007;8:518-532.
175. Duclairoir Poc C, Groboillot A, Lesouhaitier O, Morin JP, Orange N, Feuilloley MJ.
Caenorhabditis elegans: a model to monitor bacterial air quality. BMC Res Notes.
2011;4:503.
176. Anbalagan C, Lafayette I, Antoniou-Kourounioti M, Haque M, King J, Johnsen B, et
al. Transgenic nematodes as biosensors for metal stress in soil pore water samples.
Ecotoxicology. 2012;21:439-455.
177. Anbalagan C, Lafayette I, Antoniou-Kourounioti M, Gutierrez C, Martin JR,
Chowdhuri DK, et al. Use of transgenic GFP reporter strains of the nematode
Caenorhabditis elegans to investigate the patterns of stress responses induced by
pesticides and by organic extracts from agricultural soils. Ecotoxicology.
2013;22:72-85.
178. Hunt PR, Olejnik N, Sprando RL. Toxicity ranking of heavy metals with screening
method using adult Caenorhabditis elegans and propidium iodide replicates toxicity
ranking in rat. Food Chem Toxicol. 2012;50:3280-3290.
179. Zhao Y, Wu Q, Tang M, Wang D. The in vivo underlying mechanism for recovery
response formation in nano-titanium dioxide exposed Caenorhabditis elegans after
transfer to the normal condition. Nanomedicine. 2014;10:89-98.
180. Liu S, Saul N, Pan B, Menzel R, Steinberg CE. The non-target organism
Caenorhabditis elegans withstands the impact of sulfamethoxazole. Chemosphere.
2013;93:2373-2380.
181. Saul N, Chakrabarti S, Sturzenbaum SR, Menzel R, Steinberg CE. Neurotoxic action
of microcystin-LR is reflected in the transcriptional stress response of
Caenorhabditis elegans. Chem Biol Interact. 2014;223:51-57.
182. Smith MA, Jr., Zhang Y, Polli JR, Wu H, Zhang B, Xiao P, et al. Impacts of chronic
low-level nicotine exposure on Caenorhabditis elegans reproduction: identification
of novel gene targets. Reprod Toxicol. 2013;40:69-75.
183. Moore CE, Lein PJ, Puschner B. Microcystins alter chemotactic behavior in
Caenorhabditis elegans by selectively targeting the AWA sensory neuron. Toxins
(Basel). 2014;6:1813-1836.
184. Sprando RL, Olejnik N, Cinar HN, Ferguson M. A method to rank order water soluble
compounds according to their toxicity using Caenorhabditis elegans, a Complex
Object Parametric Analyzer and Sorter, and axenic liquid media. Food Chem
Toxicol. 2009;47:722-728.
185. Stiernagle T. Maintenance of C. elegans. WormBook. 2006:1-11.

222
186. Misra C, Kim S, Shen S, Sioutas C. Design and evaluation of a high-flow rate, very
low pressure drop impactor for separation and collection of fine from ultrafine
particles. J Aerosol Sci. 2002;33:736-752.
187. Shore DE, Carr CE, Ruvkun G. Induction of cytoprotective pathways is central to the
extension of lifespan conferred by multiple longevity pathways. PLoS Genet.
2012;8:e1002792.
188. An JH, Blackwell TK. SKN-1 links C. elegans mesendodermal specification to a
conserved oxidative stress response. Genes Dev. 2003;17:1882-1893.
189. Bowerman B, Eaton BA, Priess JR. skn-1, a maternally expressed gene required to
specify the fate of ventral blastomeres in the early C. elegans embryo. Cell.
1992;68:1061-1075.
190. Green RM, Gally F, Keeney JG, Alper S, Gao B, Han M, et al. Impact of cigarette
smoke exposure on innate immunity: a Caenorhabditis elegans model. PloS one.
2009;4:e6860.
191. Wang X, Chen M, Zhong M, Hu Z, Qiu L, Rajagopalan S, et al. Exposure to
Concentrated Ambient PM2.5 Shortens Lifespan and Induces Inflammation-
Associated Signaling and Oxidative Stress in Drosophila. Toxicol Sci. 2017;156:199-
207.
192. Crook-McMahon HM, Olahova M, Button EL, Winter JJ, Veal EA. Genome-wide
screening identifies new genes required for stress-induced phase 2 detoxification
gene expression in animals. BMC Biol. 2014;12:64.
193. Zhang X, Staimer N, Gillen DL, Tjoa T, Schauer JJ, Shafer MM, et al. Associations
of oxidative stress and inflammatory biomarkers with chemically-characterized air
pollutant exposures in an elderly cohort. Environ Res. 2016;150:306-319.
194. Stone V, Miller MR, Clift MJD, Elder A, Mills NL, Moller P, et al. Nanomaterials
Versus Ambient Ultrafine Particles: An Opportunity to Exchange Toxicology
Knowledge. Environ Health Perspect. 2017;125:106002.
195. Brown DM, Donaldson K, Borm PJ, Schins RP, Dehnhardt M, Gilmour P, et al.
Calcium and ROS-mediated activation of transcription factors and TNF-alpha
cytokine gene expression in macrophages exposed to ultrafine particles. Am J
Physiol Lung Cell Mol Physiol. 2004;286:L344-353.
196. Baulig A, Singh S, Marchand A, Schins R, Barouki R, Garlatti M, et al. Role of Paris
PM(2.5) components in the pro-inflammatory response induced in airway epithelial
cells. Toxicology. 2009;261:126-135.

223
197. Gilmour PS, Brown DM, Lindsay TG, Beswick PH, MacNee W, Donaldson K.
Adverse health effects of PM10 particles: involvement of iron in generation of
hydroxyl radical. Occup Environ Med. 1996;53:817-822.
198. Salvi S, Blomberg A, Rudell B, Kelly F, Sandstrom T, Holgate ST, et al. Acute
inflammatory responses in the airways and peripheral blood after short-term
exposure to diesel exhaust in healthy human volunteers. Am J Respir Crit Care Med.
1999;159:702-709.
199. Zhang X, Chen X, Zhang X. The impact of exposure to air pollution on cognitive
performance. Proc Natl Acad Sci U S A. 2018.
200. Hornsten A, Lieberthal J, Fadia S, Malins R, Ha L, Xu X, et al. APL-1, a
Caenorhabditis elegans protein related to the human beta-amyloid precursor
protein, is essential for viability. Proc Natl Acad Sci U S A. 2007;104:1971-1976.
201. Ewald CY, Raps DA, Li C. APL-1, the Alzheimer's Amyloid precursor protein in
Caenorhabditis elegans, modulates multiple metabolic pathways throughout
development. Genetics. 2012;191:493-507.
202. Rundle A, Hoepner L, Hassoun A, Oberfield S, Freyer G, Holmes D, et al.
Association of childhood obesity with maternal exposure to ambient air polycyclic
aromatic hydrocarbons during pregnancy. Am J Epidemiol. 2012;175:1163-1172.
203. Sarasija S, Laboy JT, Ashkavand Z, Bonner J, Tang Y, Norman KR. Presenilin
mutations deregulate mitochondrial Ca(2+) homeostasis and metabolic activity
causing neurodegeneration in Caenorhabditis elegans. Elife. 2018;7.
204. Gems D, Partridge L. Stress-response hormesis and aging: "that which does not kill
us makes us stronger". Cell Metab. 2008;7:200-203.
205. Cypser JR, Johnson TE. Multiple stressors in Caenorhabditis elegans induce stress
hormesis and extended longevity. J Gerontol A Biol Sci Med Sci. 2002;57:B109-114.
206. Gorr MW, Velten M, Nelin TD, Youtz DJ, Sun Q, Wold LE. Early life exposure to air
pollution induces adult cardiac dysfunction. Am J Physiol Heart Circ Physiol.
2014;307:H1353-1360.
207. Lee PC, Talbott EO, Roberts JM, Catov JM, Bilonick RA, Stone RA, et al. Ambient
air pollution exposure and blood pressure changes during pregnancy. Environ Res.
2012;117:46-53.
208. Lavigne E, Belair MA, Do MT, Stieb DM, Hystad P, van Donkelaar A, et al. Maternal
exposure to ambient air pollution and risk of early childhood cancers: A population-
based study in Ontario, Canada. Environ Int. 2017;100:139-147.

224
209. Perera FP, Chang HW, Tang D, Roen EL, Herbstman J, Margolis A, et al. Early-life
exposure to polycyclic aromatic hydrocarbons and ADHD behavior problems. PLoS
One. 2014;9:e111670.
210. Guxens M, Garcia-Esteban R, Giorgis-Allemand L, Forns J, Badaloni C, Ballester F,
et al. Air pollution during pregnancy and childhood cognitive and psychomotor
development: six European birth cohorts. Epidemiology. 2014;25:636-647.
211. Liu J, Lewis G. Environmental toxicity and poor cognitive outcomes in children and
adults. J Environ Health. 2014;76:130-138.
212. McConnell R, Shen E, Gilliland FD, Jerrett M, Wolch J, Chang CC, et al. A
longitudinal cohort study of body mass index and childhood exposure to secondhand
tobacco smoke and air pollution: the Southern California Children's Health Study.
Environ Health Perspect. 2015;123:360-366.
213. Calderon-Garciduenas L, Kavanaugh M, Block M, D'Angiulli A, Delgado-Chavez R,
Torres-Jardon R, et al. Neuroinflammation, hyperphosphorylated tau, diffuse
amyloid plaques, and down-regulation of the cellular prion protein in air pollution
exposed children and young adults. J Alzheimers Dis. 2012;28:93-107.
214. Ljubimova JY, Kleinman MT, Karabalin NM, Inoue S, Konda B, Gangalum P, et al.
Gene expression changes in rat brain after short and long exposures to particulate
matter in Los Angeles basin air: Comparison with human brain tumors. Exp Toxicol
Pathol. 2013;65:1063-1071.
215. Yokota S, Hori H, Umezawa M, Kubota N, Niki R, Yanagita S, et al. Gene expression
changes in the olfactory bulb of mice induced by exposure to diesel exhaust are
dependent on animal rearing environment. PloS one. 2013;8:e70145.
216. Cheng H, Davis DA, Hasheminassab S, Sioutas C, Morgan TE, Finch CE. Urban
traffic-derived nanoparticulate matter reduces neurite outgrowth via TNFalpha in
vitro. J Neuroinflammation. 2016;13:19.
217. Bolger AM, Lohse M, Usadel B. Trimmomatic: a flexible trimmer for Illumina
sequence data. Bioinformatics. 2014;30:2114-2120.
218. Kim D, Langmead B, Salzberg SL. HISAT: a fast spliced aligner with low memory
requirements. Nat Methods. 2015;12:357-360.
219. Liao Y, Smyth GK, Shi W. featureCounts: an efficient general purpose program for
assigning sequence reads to genomic features. Bioinformatics. 2014;30:923-930.
220. Robinson MD, Oshlack A. A scaling normalization method for differential expression
analysis of RNA-seq data. Genome Biol. 2010;11:R25.

225
221. Smyth GK. Limma: linear models for microarray data. Bioinformatics and
computational biology solutions using R and Bioconductor. Springer; 2005:397-420.
222. Diboun I, Wernisch L, Orengo CA, Koltzenburg M. Microarray analysis after RNA
amplification can detect pronounced differences in gene expression using limma.
BMC Genomics. 2006;7:252.
223. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network
analysis. BMC Bioinformatics. 2008;9:559.
224. Witten DM, Tibshirani R, Hastie T. A penalized matrix decomposition, with
applications to sparse principal components and canonical correlation analysis.
Biostatistics. 2009;10:515-534.
225. Voineagu I, Wang X, Johnston P, Lowe JK, Tian Y, Horvath S, et al. Transcriptomic
analysis of autistic brain reveals convergent molecular pathology. Nature.
2011;474:380-384.
226. Xue Z, Huang K, Cai C, Cai L, Jiang CY, Feng Y, et al. Genetic programs in human
and mouse early embryos revealed by single-cell RNA sequencing. Nature.
2013;500:593-597.
227. Klocke C, Sherina V, Graham UM, Gunderson J, Allen JL, Sobolewski M, et al.
Enhanced cerebellar myelination with concomitant iron elevation and ultrastructural
irregularities following prenatal exposure to ambient particulate matter in the mouse.
Inhal Toxicol. 2018;30:381-396.
228. Xiao X, Li BX, Mitton B, Ikeda A, Sakamoto KM. Targeting CREB for cancer therapy:
friend or foe. Curr Cancer Drug Targets. 2010;10:384-391.
229. Brunst KJ, Sanchez-Guerra M, Chiu YM, Wilson A, Coull BA, Kloog I, et al. Prenatal
particulate matter exposure and mitochondrial dysfunction at the maternal-fetal
interface: Effect modification by maternal lifetime trauma and child sex. Environ Int.
2018;112:49-58.
230. Kaali S, Jack D, Delimini R, Hu L, Burkart K, Opoku-Mensah J, et al. Prenatal
Household Air Pollution Alters Cord Blood Mononuclear Cell Mitochondrial DNA
Copy Number: Sex-Specific Associations. Int J Environ Res Public Health.
2018;16:26.
231. Haghani A, Dalton HM, Safi N, Shirmohammadi F, Sioutas C, Morgan TE, et al. Air
pollution alters Caenorhabditis elegans development and lifespan: responses to
traffic-related nanoparticulate matter (nPM). J Gerontol A Biol Sci Med Sci. 2019.
232. Perera F, Ashrafi A, Kinney P, Mills D. Towards a fuller assessment of benefits to
children's health of reducing air pollution and mitigating climate change due to fossil
fuel combustion. Environ Res. 2018;172:55-72.

226
233. Rosa MJ, Hsu HL, Just AC, Brennan KJ, Bloomquist T, Kloog I, et al. Association
between prenatal particulate air pollution exposure and telomere length in cord
blood: Effect modification by fetal sex. Environ Res. 2019;172:495-501.
234. Martens DS, Cox B, Janssen BG, Clemente DBP, Gasparrini A, Vanpoucke C, et al.
Prenatal Air Pollution and Newborns' Predisposition to Accelerated Biological Aging.
JAMA Pediatr. 2017;171:1160-1167.
235. Haghani A, Johnson R, Safi N, Zhang H, Thorwald M, Mousavi A, et al. Toxicity of
urban air pollution particulate matter in developing and adult mouse brain:
Comparison of total and filter-eluted nanoparticles. Environ Int. 2020;136:105510.
236. von Ehrenstein OS, Heck JE, Park AS, Cockburn M, Escobedo L, Ritz B. In Utero
and Early-Life Exposure to Ambient Air Toxics and Childhood Brain Tumors: A
Population-Based Case-Control Study in California, USA. Environ Health Perspect.
2016;124:1093-1099.
237. Hu CY, Gao X, Fang Y, Jiang W, Huang K, Hua XG, et al. Human epidemiological
evidence about the association between air pollution exposure and gestational
diabetes mellitus: Systematic review and meta-analysis. Environ Res.
2020;180:108843.
238. Moody EC, Cantoral A, Tamayo-Ortiz M, Pizano-Zarate ML, Schnaas L, Kloog I, et
al. Association of Prenatal and Perinatal Exposures to Particulate Matter With
Changes in Hemoglobin A1c Levels in Children Aged 4 to 6 Years. JAMA Netw
Open. 2019;2:e1917643.
239. Kulas JA, Hettwer JV, Sohrabi M, Melvin JE, Manocha GD, Puig KL, et al. In utero
exposure to fine particulate matter results in an altered neuroimmune phenotype in
adult mice. Environ Pollut. 2018;241:279-288.
240. Yokota S, Oshio S, Moriya N, Takeda K. Social Isolation-Induced Territorial
Aggression in Male Offspring Is Enhanced by Exposure to Diesel Exhaust during
Pregnancy. PLoS One. 2016;11:e0149737.
241. Onoda A, Takeda K, Umezawa M. Dysregulation of major functional genes in frontal
cortex by maternal exposure to carbon black nanoparticle is not ameliorated by
ascorbic acid pretreatment. The Science of the total environment. 2018;634:1126-
1135.
242. Coburn JL, Cole TB, Dao KT, Costa LG. Acute exposure to diesel exhaust impairs
adult neurogenesis in mice: prominence in males and protective effect of
pioglitazone. Arch Toxicol. 2018;92:1815-1829.

227
243. Li X, Sun H, Li B, Zhang X, Cui J, Yun J, et al. Probiotics Ameliorate Colon Epithelial
Injury Induced by Ambient Ultrafine Particles Exposure. Adv Sci (Weinh).
2019;6:1900972.
244. Sun Q, Zhang G, Chen R, Li R, Wang H, Jiang A, et al. Central IKK2 Inhibition
Ameliorates Air Pollution-Mediated Hepatic Glucose and Lipid Metabolism
Dysfunction in Mice With Type II Diabetes. Toxicol Sci. 2018;164:240-249.
245. Li XY, Hao L, Liu YH, Chen CY, Pai VJ, Kang JX. Protection against fine particle-
induced pulmonary and systemic inflammation by omega-3 polyunsaturated fatty
acids. Biochim Biophys Acta Gen Subj. 2017;1861:577-584.
246. Topalova D, Ugrinova I, Pashev IG, Pasheva EA. HMGB1 protein inhibits DNA
replication in vitro: a role of the acetylation and the acidic tail. Int J Biochem Cell
Biol. 2008;40:1536-1542.
247. Ugrinova I, Pashev IG, Pasheva EA. Nucleosome binding properties and Co-
remodeling activities of native and in vivo acetylated HMGB-1 and HMGB-2 proteins.
Biochemistry. 2009;48:6502-6507.
248. Ugrinova I, Zlateva S, Pashev IG, Pasheva EA. Native HMGB1 protein inhibits repair
of cisplatin-damaged nucleosomes in vitro. Int J Biochem Cell Biol. 2009;41:1556-
1562.
249. Travers AA. Priming the nucleosome: a role for HMGB proteins? EMBO Rep.
2003;4:131-136.
250. Tang D, Kang R, Livesey KM, Cheh CW, Farkas A, Loughran P, et al. Endogenous
HMGB1 regulates autophagy. J Cell Biol. 2010;190:881-892.
251. Musumeci D, Roviello GN, Montesarchio D. An overview on HMGB1 inhibitors as
potential therapeutic agents in HMGB1-related pathologies. Pharmacol Ther.
2014;141:347-357.
252. Nishibori M, Mori S, Takahashi HK. Anti-HMGB1 monoclonal antibody therapy for a
wide range of CNS and PNS diseases. J Pharmacol Sci. 2019;140:94-101.
253. Tang D, Kang R, Zeh HJ, 3rd, Lotze MT. High-mobility group box 1 and cancer.
Biochim Biophys Acta. 2010;1799:131-140.
254. Wingett SW, Andrews S. FastQ Screen: A tool for multi-genome mapping and quality
control. F1000Res. 2018;7:1338.
255. Allen JL, Conrad K, Oberdorster G, Johnston CJ, Sleezer B, Cory-Slechta DA.
Developmental exposure to concentrated ambient particles and preference for
immediate reward in mice. Environ Health Perspect. 2013;121:32-38.

228
256. Cui J, Fu Y, Lu R, Bi Y, Zhang L, Zhang C, et al. Metabolomics analysis explores
the rescue to neurobehavioral disorder induced by maternal PM2.5 exposure in
mice. Ecotoxicol Environ Saf. 2019;169:687-695.
257. Reed MD, Yim YS, Wimmer RD, Kim H, Ryu C, Welch GM, et al. IL-17a promotes
sociability in mouse models of neurodevelopmental disorders. Nature.
2020;577:249-253.
258. Lee JS, Ward WO, Liu J, Ren H, Vallanat B, Delker D, et al. Hepatic xenobiotic
metabolizing enzyme and transporter gene expression through the life stages of the
mouse. PloS one. 2011;6:e24381.
259. Arnold SJ, Huang GJ, Cheung AF, Era T, Nishikawa S, Bikoff EK, et al. The T-box
transcription factor Eomes/Tbr2 regulates neurogenesis in the cortical
subventricular zone. Genes Dev. 2008;22:2479-2484.
260. Xu MX, Ge CX, Qin YT, Gu TT, Lou DS, Li Q, et al. Prolonged PM2.5 exposure
elevates risk of oxidative stress-driven nonalcoholic fatty liver disease by triggering
increase of dyslipidemia. Free Radic Biol Med. 2019;130:542-556.
261. Coolen M, Katz S, Bally-Cuif L. miR-9: a versatile regulator of neurogenesis. Front
Cell Neurosci. 2013;7:220.
262. Coolen M, Thieffry D, Drivenes O, Becker TS, Bally-Cuif L. miR-9 controls the timing
of neurogenesis through the direct inhibition of antagonistic factors. Dev Cell.
2012;22:1052-1064.
263. Zhao C, Sun G, Li S, Shi Y. A feedback regulatory loop involving microRNA-9 and
nuclear receptor TLX in neural stem cell fate determination. Nat Struct Mol Biol.
2009;16:365-371.
264. Cogswell JP, Ward J, Taylor IA, Waters M, Shi Y, Cannon B, et al. Identification of
miRNA changes in Alzheimer's disease brain and CSF yields putative biomarkers
and insights into disease pathways. J Alzheimers Dis. 2008;14:27-41.
265. Buckley NJ, Johnson R, Zuccato C, Bithell A, Cattaneo E. The role of REST in
transcriptional and epigenetic dysregulation in Huntington's disease. Neurobiol Dis.
2010;39:28-39.
266. Haramati S, Chapnik E, Sztainberg Y, Eilam R, Zwang R, Gershoni N, et al. miRNA
malfunction causes spinal motor neuron disease. Proc Natl Acad Sci U S A.
2010;107:13111-13116.
267. Kim J, Inoue K, Ishii J, Vanti WB, Voronov SV, Murchison E, et al. A MicroRNA
feedback circuit in midbrain dopamine neurons. Science. 2007;317:1220-1224.

229
268. Nowek K, Wiemer EAC, Jongen-Lavrencic M. The versatile nature of miR-9/9(*) in
human cancer. Oncotarget. 2018;9:20838-20854.
269. Mentzel CM, Anthon C, Jacobsen MJ, Karlskov-Mortensen P, Bruun CS, Jorgensen
CB, et al. Gender and Obesity Specific MicroRNA Expression in Adipose Tissue
from Lean and Obese Pigs. PloS one. 2015;10:e0131650.
270. Wan HY, Guo LM, Liu T, Liu M, Li X, Tang H. Regulation of the transcription factor
NF-kappaB1 by microRNA-9 in human gastric adenocarcinoma. Mol Cancer.
2010;9:16.
271. Gu R, Liu N, Luo S, Huang W, Zha Z, Yang J. MicroRNA-9 regulates the
development of knee osteoarthritis through the NF-kappaB1 pathway in
chondrocytes. Medicine (Baltimore). 2016;95:e4315.
272. Harris SJ, Ciuclan L, Finan PM, Wymann MP, Walker C, Westwick J, et al. Genetic
ablation of PI3Kgamma results in defective IL-17RA signalling in T lymphocytes and
increased IL-17 levels. Eur J Immunol. 2012;42:3394-3404.
273. Lai CH, Wang KC, Lee FT, Tsai HW, Ma CY, Cheng TL, et al. Toll-Like Receptor 4
Is Essential in the Development of Abdominal Aortic Aneurysm. PloS one.
2016;11:e0146565.
274. Burnett R, Chen H, Szyszkowicz M, Fann N, Hubbell B, Pope CA, 3rd, et al. Global
estimates of mortality associated with long-term exposure to outdoor fine particulate
matter. Proc Natl Acad Sci U S A. 2018;115:9592-9597.
275. Lelieveld J, Klingmuller K, Pozzer A, Poschl U, Fnais M, Daiber A, et al.
Cardiovascular disease burden from ambient air pollution in Europe reassessed
using novel hazard ratio functions. Eur Heart J. 2019;40:1590-1596.
276. Chen H, Kwong JC, Copes R, Hystad P, van Donkelaar A, Tu K, et al. Exposure to
ambient air pollution and the incidence of dementia: A population-based cohort
study. Environ Int. 2017;108:271-277.
277. Finch CE, Kulminski AM. The Alzheimer's Disease Exposome. Alzheimers Dement.
2019;15:1123-1132.
278. Clifford A, Lang L, Chen R, Anstey KJ, Seaton A. Exposure to air pollution and
cognitive functioning across the life course--A systematic literature review. Environ
Res. 2016;147:383-398.
279. Chen JC, Schwartz J. Neurobehavioral effects of ambient air pollution on cognitive
performance in US adults. Neurotoxicology. 2009;30:231-239.

230
280. Ailshire J, Karraker A, Clarke P. Neighborhood social stressors, fine particulate
matter air pollution, and cognitive function among older U.S. adults. Soc Sci Med.
2017;172:56-63.
281. Ailshire JA, Clarke P. Fine particulate matter air pollution and cognitive function
among U.S. older adults. J Gerontol B Psychol Sci Soc Sci. 2015;70:322-328.
282. Zhang X, Chen X, Zhang X. The impact of exposure to air pollution on cognitive
performance. Proc Natl Acad Sci U S A. 2018;115:9193-9197.
283. Calderon-Garciduenas L, de la Monte SM. Apolipoprotein E4, Gender, Body Mass
Index, Inflammation, Insulin Resistance, and Air Pollution Interactions: Recipe for
Alzheimer's Disease Development in Mexico City Young Females. J Alzheimers Dis.
2017;58:613-630.
284. Chiu YH, Bellinger DC, Coull BA, Anderson S, Barber R, Wright RO, et al.
Associations between traffic-related black carbon exposure and attention in a
prospective birth cohort of urban children. Environ Health Perspect. 2013;121:859-
864.
285. Sunyer J, Esnaola M, Alvarez-Pedrerol M, Forns J, Rivas I, Lopez-Vicente M, et al.
Association between traffic-related air pollution in schools and cognitive
development in primary school children: a prospective cohort study. PLoS Med.
2015;12:e1001792.
286. Jiang C, Stewart LT, Kuo HC, McGilberry W, Wall SB, Liang B, et al. Cyclic O3
exposure synergizes with aging leading to memory impairment in male APOE
epsilon3, but not APOE epsilon4, targeted replacement mice. Neurobiol Aging.
2019;81:9-21.
287. Rooney J, Oshida K, Vasani N, Vallanat B, Ryan N, Chorley BN, et al. Activation of
Nrf2 in the liver is associated with stress resistance mediated by suppression of the
growth hormone-regulated STAT5b transcription factor. PloS one.
2018;13:e0200004.
288. Graeser A-C, Boesch-Saadatmandi C, Lippmann J, Wagner AE, Huebbe P, Storm
N, et al. Nrf2-dependent gene expression is affected by the proatherogenic apoE4
genotype—studies in targeted gene replacement mice. Journal of molecular
medicine. 2011;89:1027-1035.
289. Cacciottolo M, Morgan TE, Finch CE. Rust on the Brain from Microbleeds and Its
Relevance to Alzheimer Studies: Invited Commentary on Cacciottolo Neurobiology
of Aging, 2016. J Alzheimers Dis Parkinsonism. 2016;6.

231
290. Youmans KL, Tai LM, Nwabuisi-Heath E, Jungbauer L, Kanekiyo T, Gan M, et al.
APOE4-specific changes in Abeta accumulation in a new transgenic mouse model
of Alzheimer disease. J Biol Chem. 2012;287:41774-41786.
291. Trapnell C, Roberts A, Goff L, Pertea G, Kim D, Kelley DR, et al. Differential gene
and transcript expression analysis of RNA-seq experiments with TopHat and
Cufflinks. Nat Protoc. 2012;7:562-578.
292. Johnson WE, Li C, Rabinovic A. Adjusting batch effects in microarray expression
data using empirical Bayes methods. Biostatistics. 2007;8:118-127.
293. Wilkinson B, Koenigsknecht-Talboo J, Grommes C, Lee CY, Landreth G. Fibrillar
beta-amyloid-stimulated intracellular signaling cascades require Vav for induction of
respiratory burst and phagocytosis in monocytes and microglia. J Biol Chem.
2006;281:20842-20850.
294. Kummer MP, Vogl T, Axt D, Griep A, Vieira-Saecker A, Jessen F, et al. Mrp14
deficiency ameliorates amyloid beta burden by increasing microglial phagocytosis
and modulation of amyloid precursor protein processing. J Neurosci.
2012;32:17824-17829.
295. Lee P-C, Raaschou-Nielsen O, Lill CM, Bertram L, Sinsheimer JS, Hansen J, et al.
Gene-environment interactions linking air pollution and inflammation in Parkinson's
disease. Environmental research. 2016;151:713-720.
296. Wu SL, Wang WF, Shyu HY, Ho YJ, Shieh JC, Fu YP, et al. Association analysis of
GRIN1 and GRIN2B polymorphisms and Parkinson's disease in a hospital-based
case-control study. Neurosci Lett. 2010;478:61-65.
297. Demontis D, Nyegaard M, Buttenschon HN, Hedemand A, Pedersen CB, Grove J,
et al. Association of GRIN1 and GRIN2A-D with schizophrenia and genetic
interaction with maternal herpes simplex virus-2 infection affecting disease risk. Am
J Med Genet B Neuropsychiatr Genet. 2011;156B:913-922.
298. de Oliveira FF, Cardoso AF, Mazzotti DR, Faria TC, Souza GS, Teixeira DV, et al.
GRIN1 genotypes and APOE gene haplotypes affect the age at onset of alzheimer's
disease dementia but not cognitive or functional response to memantine.
Alzheimer's & Dementia: The Journal of the Alzheimer's Association. 2016;12:P454.
299. Rooney JPK, Woods NF, Martin MD, Woods JS. Genetic polymorphisms of GRIN2A
and GRIN2B modify the neurobehavioral effects of low-level lead exposure in
children. Environ Res. 2018;165:1-10.
300. Courtney R, Landreth GE. LXR Regulation of Brain Cholesterol: From Development
to Disease. Trends Endocrinol Metab. 2016;27:404-414.

232
301. Mahley RW, Weisgraber KH, Huang Y. Apolipoprotein E: structure determines
function, from atherosclerosis to Alzheimer's disease to AIDS. J Lipid Res. 2009;50
Suppl:S183-188.
302. Toms TE, Smith JP, Panoulas VF, Blackmore H, Douglas KM, Kitas GD.
Apolipoprotein E gene polymorphisms are strong predictors of inflammation and
dyslipidemia in rheumatoid arthritis. J Rheumatol. 2012;39:218-225.
303. Verghese PB, Castellano JM, Garai K, Wang Y, Jiang H, Shah A, et al. ApoE
influences amyloid-beta (Abeta) clearance despite minimal apoE/Abeta association
in physiological conditions. Proc Natl Acad Sci U S A. 2013;110:E1807-1816.
304. Prasad H, Rao R. Amyloid clearance defect in ApoE4 astrocytes is reversed by
epigenetic correction of endosomal pH. Proc Natl Acad Sci U S A. 2018;115:E6640-
E6649.
305. Maier M, Peng Y, Jiang L, Seabrook TJ, Carroll MC, Lemere CA. Complement C3
deficiency leads to accelerated amyloid beta plaque deposition and
neurodegeneration and modulation of the microglia/macrophage phenotype in
amyloid precursor protein transgenic mice. J Neurosci. 2008;28:6333-6341.
306. Zhu Y, Nwabuisi-Heath E, Dumanis SB, Tai LM, Yu C, Rebeck GW, et al. APOE
genotype alters glial activation and loss of synaptic markers in mice. Glia.
2012;60:559-569.
307. Jofre-Monseny L, Loboda A, Wagner AE, Huebbe P, Boesch-Saadatmandi C,
Jozkowicz A, et al. Effects of apoE genotype on macrophage inflammation and
heme oxygenase-1 expression. Biochem Biophys Res Commun. 2007;357:319-
324.
308. Loram LC, Sholar PW, Taylor FR, Wiesler JL, Babb JA, Strand KA, et al. Sex and
estradiol influence glial pro-inflammatory responses to lipopolysaccharide in rats.
Psychoneuroendocrinology. 2012;37:1688-1699.
309. Hanamsagar R, Alter MD, Block CS, Sullivan H, Bolton JL, Bilbo SD. Generation of
a microglial developmental index in mice and in humans reveals a sex difference in
maturation and immune reactivity. Glia. 2017;65:1504-1520.
310. Chen CS, Tseng YT, Hsu YY, Lo YC. Nrf2-Keap1 antioxidant defense and cell
survival signaling are upregulated by 17beta-estradiol in homocysteine-treated
dopaminergic SH-SY5Y cells. Neuroendocrinology. 2013;97:232-241.
311. Khan M, Shah SA, Kim MO. 17beta-Estradiol via SIRT1/Acetyl-p53/NF-kB Signaling
Pathway Rescued Postnatal Rat Brain Against Acute Ethanol Intoxication. Mol
Neurobiol. 2018;55:3067-3078.

233
312. Prasad S, Sajja RK, Kaisar MA, Park JH, Villalba H, Liles T, et al. Role of Nrf2 and
protective effects of Metformin against tobacco smoke-induced cerebrovascular
toxicity. Redox Biol. 2017;12:58-69.
313. Kim JE, You DJ, Lee C, Ahn C, Seong JY, Hwang JI. Suppression of NF-kappaB
signaling by KEAP1 regulation of IKKbeta activity through autophagic degradation
and inhibition of phosphorylation. Cell Signal. 2010;22:1645-1654.
314. Lee DF, Kuo HP, Liu M, Chou CK, Xia W, Du Y, et al. KEAP1 E3 ligase-mediated
downregulation of NF-kappaB signaling by targeting IKKbeta. Mol Cell.
2009;36:131-140.
315. Sanlioglu S, Williams CM, Samavati L, Butler NS, Wang G, McCray PB, Jr., et al.
Lipopolysaccharide induces Rac1-dependent reactive oxygen species formation
and coordinates tumor necrosis factor-alpha secretion through IKK regulation of NF-
kappa B. J Biol Chem. 2001;276:30188-30198.
316. Cuadrado A, Martin-Moldes Z, Ye J, Lastres-Becker I. Transcription factors NRF2
and NF-kappaB are coordinated effectors of the Rho family, GTP-binding protein
RAC1 during inflammation. J Biol Chem. 2014;289:15244-15258.
317. Wakabayashi N, Slocum SL, Skoko JJ, Shin S, Kensler TW. When NRF2 talks,
who's listening? Antioxid Redox Signal. 2010;13:1649-1663.
318. Liu GH, Qu J, Shen X. NF-kappaB/p65 antagonizes Nrf2-ARE pathway by depriving
CBP from Nrf2 and facilitating recruitment of HDAC3 to MafK. Biochim Biophys Acta.
2008;1783:713-727.
319. Lelieveld J, Munzel T. Air pollution, chronic smoking, and mortality. Eur Heart J.
2019;40:3204.
320. Kure EH, Ryberg D, Hewer A, Phillips DH, Skaug V, Baera R, et al. p53 mutations
in lung tumours: relationship to gender and lung DNA adduct levels. Carcinogenesis.
1996;17:2201-2205.
321. Ryberg D, Hewer A, Phillips DH, Haugen A. Different susceptibility to smoking-
induced DNA damage among male and female lung cancer patients. Cancer Res.
1994;54:5801-5803.
322. Tam A, Churg A, Wright JL, Zhou S, Kirby M, Coxson HO, et al. Sex Differences in
Airway Remodeling in a Mouse Model of Chronic Obstructive Pulmonary Disease.
Am J Respir Crit Care Med. 2016;193:825-834.
323. Seymour BW, Friebertshauser KE, Peake JL, Pinkerton KE, Coffman RL, Gershwin
LJ. Gender differences in the allergic response of mice neonatally exposed to
environmental tobacco smoke. Dev Immunol. 2002;9:47-54.

234
324. Carnethon MR, Lynch EB, Dyer AR, Lloyd-Jones DM, Wang R, Garside DB, et al.
Comparison of risk factors for cardiovascular mortality in black and white adults.
Arch Intern Med. 2006;166:1196-1202.
325. Thun MJ, Carter BD, Feskanich D, Freedman ND, Prentice R, Lopez AD, et al. 50-
year trends in smoking-related mortality in the United States. N Engl J Med.
2013;368:351-364.
326. Castelao JE, Yuan JM, Skipper PL, Tannenbaum SR, Gago-Dominguez M, Crowder
JS, et al. Gender- and smoking-related bladder cancer risk. J Natl Cancer Inst.
2001;93:538-545.
327. Sorheim IC, Johannessen A, Gulsvik A, Bakke PS, Silverman EK, DeMeo DL.
Gender differences in COPD: are women more susceptible to smoking effects than
men? Thorax. 2010;65:480-485.
328. Lam TH, Li ZB, Ho SY, Chan WM, Ho KS, Tham MK, et al. Smoking, quitting and
mortality in an elderly cohort of 56,000 Hong Kong Chinese. Tob Control.
2007;16:182-189.
329. Yusuf HR, Giles WH, Croft JB, Anda RF, Casper ML. Impact of multiple risk factor
profiles on determining cardiovascular disease risk. Prev Med. 1998;27:1-9.
330. Doll R, Peto R. Cigarette smoking and bronchial carcinoma: dose and time
relationships among regular smokers and lifelong non-smokers. J Epidemiol
Community Health. 1978;32:303-313.
331. Fujisawa K, Takata Y, Matsumoto T, Esaki M, Ansai T, Iida M. Impact of smoking on
mortality in 80-year-old Japanese from the general population. Gerontology.
2008;54:210-216.
332. Newman AB, Sachs MC, Arnold AM, Fried LP, Kronmal R, Cushman M, et al. Total
and cause-specific mortality in the cardiovascular health study. J Gerontol A Biol Sci
Med Sci. 2009;64:1251-1261.
333. Doll R, Peto R, Boreham J, Sutherland I. Mortality in relation to smoking: 50 years'
observations on male British doctors. BMJ. 2004;328:1519.
334. Bonita R, Duncan J, Truelsen T, Jackson RT, Beaglehole R. Passive smoking as
well as active smoking increases the risk of acute stroke. Tob Control. 1999;8:156-
160.
335. McGhee SM, Ho SY, Schooling M, Ho LM, Thomas GN, Hedley AJ, et al. Mortality
associated with passive smoking in Hong Kong. BMJ. 2005;330:287-288.

235
336. McElduff P, Dobson AJ, Jackson R, Beaglehole R, Heller RF, Lay-Yee R. Coronary
events and exposure to environmental tobacco smoke: a case-control study from
Australia and New Zealand. Tob Control. 1998;7:41-46.
337. Sonnega A, Faul JD, Ofstedal MB, Langa KM, Phillips JW, Weir DR. Cohort Profile:
the Health and Retirement Study (HRS). Int J Epidemiol. 2014;43:576-585.
338. Royston P, Parmar MK. Restricted mean survival time: an alternative to the hazard
ratio for the design and analysis of randomized trials with a time-to-event outcome.
BMC Med Res Methodol. 2013;13:152.
339. Uno H, Claggett B, Tian L, Inoue E, Gallo P, Miyata T, et al. Moving beyond the
hazard ratio in quantifying the between-group difference in survival analysis. J Clin
Oncol. 2014;32:2380-2385.
340. Kim DH, Uno H, Wei LJ. Restricted Mean Survival Time as a Measure to Interpret
Clinical Trial Results. JAMA Cardiol. 2017;2:1179-1180.
341. Beltran-Sanchez H, Finch CE, Crimmins EM. Twentieth century surge of excess
adult male mortality. Proc Natl Acad Sci U S A. 2015;112:8993-8998.
342. Lodato MA, Walsh CA. Genome aging: somatic mutation in the brain links age-
related decline with disease and nominates pathogenic mechanisms. Hum Mol
Genet. 2019.
343. Sun S, Osterman MD, Li M. Tissue specificity of DNA damage response and
tumorigenesis. Cancer Biol Med. 2019;16:396-414.
344. Anderson DO, Ferris BG, Jr., Zickmantel R. The Chilliwack Respiratory Survey,
1963. Iv. The Effect of Tobacco Smoking on the Prevalence of Respiratory Disease.
Can Med Assoc J. 1965;92:1066-1076.
345. Lebowitz MD. Smoking habits and changes in smoking habits as they relate to
chronic conditions and respiratory symptoms. Am J Epidemiol. 1977;105:534-543.
346. Forey BA, Thornton AJ, Lee PN. Systematic review with meta-analysis of the
epidemiological evidence relating smoking to COPD, chronic bronchitis and
emphysema. BMC Pulm Med. 2011;11:36.
347. Chapman KR, Tashkin DP, Pye DJ. Gender bias in the diagnosis of COPD. Chest.
2001;119:1691-1695.
348. Lofroth G. Environmental tobacco smoke: overview of chemical composition and
genotoxic components. Mutat Res. 1989;222:73-80.

236
349. Noel A, Xiao R, Perveen Z, Zaman H, Le Donne V, Penn A. Sex-specific lung
functional changes in adult mice exposed only to second-hand smoke in utero.
Respir Res. 2017;18:104.
350. Turner MC, Cohen A, Jerrett M, Gapstur SM, Diver WR, Pope CA, 3rd, et al.
Interactions between cigarette smoking and fine particulate matter in the Risk of
Lung Cancer Mortality in Cancer Prevention Study II. Am J Epidemiol.
2014;180:1145-1149.
351. Morales E, Julvez J, Torrent M, de Cid R, Guxens M, Bustamante M, et al.
Association of early-life exposure to household gas appliances and indoor nitrogen
dioxide with cognition and attention behavior in preschoolers. Am J Epidemiol.
2009;169:1327-1336.
352. Volk HE, Kerin T, Lurmann F, Hertz-Picciotto I, McConnell R, Campbell DB. Autism
spectrum disorder: interaction of air pollution with the MET receptor tyrosine kinase
gene. Epidemiology. 2014;25:44-47.

Abstract (if available)

Abstract Air pollution (AirP) is a leading global environmental risk factor of mortality and morbidity in humans and many other organisms. AirP affects the lives of more than 90% of the humans from all ages and unfortunately has an increasing trend in developing countries. Depends on the life stage, exposure to AirP can increase the risk of neurodevelopmental or neurodegenerative diseases. Alarmingly, lack of understanding of the AirP toxicity mechanisms likely led to a severe underestimation of the global health burden of these complex toxicants. AirP toxicity is shaped by a complex interface of chemical properties of toxicants mixture and the biological features of the exposed individuals (e.g. genetic structure, sex, and life-stage). The current dissertation is comprised of a series of projects to resolve some of these complex interactions. In chapters 2-3, we discussed how the chemical characterization of AirP can alter the neurotoxicity in vivo and in vitro. Moreover, we developed a cell-based assay that can predict the neurotoxicity of the AirP particulate matter (PM) in the brain. Chapters 4-7 approaches the nuances in the molecular toxicity of AirP between different life stages (development vs adult), sexes (males vs females), and some genetic variants (ApoE3 vs ApoE4). We also corroborated our findings in Caenorhabditis elegans, a known nematode model for aging research. In the last chapter, we described the gender differences in vulnerability to life-time cigarette smoke exposure for some aging-associated health outcomes. Throughout the dissertation, we discussed the facing challenges in the AirP toxicology field, and also identified potential molecular mechanisms that underlay each complex research question. We hope that our findings lead to several novel studies and translate into the regulation soon.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Statistical algorithms for examining gene and environmental influences on human aging

PDF

Neuroinflammation and the behavioral consequences of air pollution over the life course

PDF

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

PDF

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

PDF

Synaptic transmission, nutrient sensors, and aging in Drosophila melanogaster

PDF

Neuroinflammatory effects of urban traffic-derived nanoparticulate matter on neural systems

PDF

The overlap between mTOR signaling, rapamycin and cellular senescence

PDF

Defining the functional roles of neurotransmitters and neuropeptides in neural circuits

PDF

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

PDF

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

PDF

Investigating brain aging and neurodegenerative diseases through omics data

PDF

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

PDF

The effects of a fasting mimicking diet (FMD) on mouse models of Alzheimer's and Parkinson's disease

PDF

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

PDF

Toxicity of urban particulate matter: long-term health risks, influences of surrounding geography, and diurnal variation in chemical composition and the cellular oxidative stress response

PDF

The effects of prolonged fasting/ fasting mimicking diet (FMD) on CNS protection, regeneration, and treatment

PDF

Development of novel techniques for evaluating physical, chemical and toxicological properties of particulate matter in ambient air

PDF

Associations of ambient air pollution exposures with perceived stress in the MADRES cohort

PDF

Detecting joint interactions between sets of variables in the context of studies with a dichotomous phenotype, with applications to asthma susceptibility involving epigenetics and epistasis

Asset Metadata

Creator Haghani, Amin (author)

Core Title Air pollution neurotoxicity throughout the lifespan: studies on the mechanism of toxicity and interactions with effects of sex and genetic background

School Leonard Davis School of Gerontology

Degree Doctor of Philosophy

Degree Program Biology of Aging

Publication Date 03/23/2020

Defense Date 02/28/2020

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

Tag Air pollution,Alzheimer's,Inflammation,neurodevelopment,neurotoxicity,OAI-PMH Harvest,oxidative stress

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Finch, Caleb E. (committee chair), Crimmins, Eileen (committee member), Curran, Sean P. (committee member), Lithgow, Gordon (committee member), Morgan, Todd E. (committee member)

Creator Email ahaghani@usc.edu,haghani_amin@yahoo.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-275871

Unique identifier UC11675452

Identifier etd-HaghaniAmi-8224.pdf (filename),usctheses-c89-275871 (legacy record id)

Legacy Identifier etd-HaghaniAmi-8224.pdf

Dmrecord 275871

Document Type Dissertation

Rights Haghani, Amin

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