University of Southern California Dissertations and Theses

Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health

(USC Thesis Other)

Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content

PRENATAL AIR POLLUTION EXPOSURE, NEWBORN DNA METHYLATION, AND
CHILDHOOD RESPIRATORY HEALTH

by

Lu Gao

A Dissertation Presented to the
FACULTY OF THE GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(EPIDEMIOLOGY)

August 2018

ii
ACKNOWLEDGEMENTS
First and foremost, I would like to express my sincere gratitude to my mentor Dr. Carrie
Breton, for her invaluable mentorship throughout my time as a graduate student, for the countless
hours she spent reading and editing my manuscripts and providing me constructive feedback, and
for the many times she met with me to give me advice. I would also like to thank my other
committee members, Drs. Kimberly Siegmund, Joshua Millstein and Louis Dubeau, for their
time, guidance and support in the completion of my dissertation work.
I would like to take this opportunity to thank Drs. Sandrah Eckel and Rima Habre for their
insightful suggestions and guidance on my projects. In addition, I am thankful to Dr. Robert
Urman who provided me invaluable assistance, comments and encouragement on my projects.
Last but not least, I would like to thank my family and friends for their continued support,
understanding and encouragement.

iii
TABLE OF CONTENTS

ACKNOWLEDGEMENTS……………………………………………………………………….ii

LIST OF TABLES………………………………………………………………………………...v

LIST OF FIGURES……………………………………………………………………………vii

OVERVIEW………………………………………………………………………………………1

CHAPTER 1: INTRODUCTION AND BACKGROUND……………………………………….4

1.1 DNA METHYLATION AND HUMAN DISEASES……………………………………4
1.1.1 The role of DNA methylation in human disease………………………………...4
1.1.2 Overview of DNA methylation………………………………………………….6
1.2 OVERVIEW OF CHILDHOOD ASTHMA……………………………………………..8
1.2.1 Introduction and pathophysiology ………………………………………………8
1.2.2 Risk factors……………………………………………………………………..11
1.2.3 Epigenetic regulation………………….………………………………………..14
1.3 OVERVIEW OF CHILDHOOD LUNG DEVELOPMENT…………………………...15
1.3.1 Introduction ……………………………………………………………………15
1.3.2 Risk factors……………………………………………………………………..17
1.3.3 Epigenetic regulation…………………………………………………………...20
1.4 INTRODUCTION OF CANDIDATE GENE – AXL…………………………………...21
1.4.1 TAM family genes……………………………………………………………..21
1.4.2 Biological function of AXL…………………………………………………….26
1.5 OVERVIEW OF ENVIRONMENTAL EPIGENETICS……………………………….28

CHAPTER 2: EPIGENETIC REGULATION OF AXL AND CHILDHOOD ASTHMA
SYMPTOMS…………………………………………………………………………………….32

2.1 ABSTRACT…………………………………………………………………………….32
2.2 INTRODUCTION………………………………………………………………………33
2.3 MATERIALS AND METHODS……………………………………………………….35
2.4 RESULTS……………………………………………………………………………….41
2.5 DISCUSSION…………………………………………………………………………...45
2.6 TABLES AND FIGURES………………………………………………………………49

CHAPTER 3: ASSOCIATION BETWEEN AXL PROMOTER METHYLATION AND LUNG
FUNCTION GROWTH DURING ADOLESCENCE…………………………………………..71

3.1 ABSTRACT…………………………………………………………………………….71
3.2 INTRODUCTION………………………………………………………………………72
3.3 MATERIALS AND METHODS……………………………………………………….73
3.4 RESULTS……………………………………………………………………………….78
3.5 DISCUSSION…………………………………………………………………………...81
3.6 TABLES AND FIGURES………………………………………………………………85

iv
CHAPTER 4: APPLYING A DISTRIBUTED LAG MODEL TO THE ASSOCIATION
BETWEEN PRENATAL AIR POLLUTION AND NEWBORN DNA METHYLATION…...103

4.1 ABSTRACT…………………………………………………………………………...103
4.2 INTRODUCTION……………………………………………………………………..104
4.3 MATERIALS AND METHODS……………………………………………………...107
4.4 RESULTS……………………………………………………………………………...113
4.5 DISCUSSION………………………………………………………………………….116
4.6 TABLES AND FIGURES……………………………………………………………..121

CHAPTER 5: SUMMARY AND FUTURE RESEARCH…………………………………….139

5.1 SUMMARY AND CONCLUSION…………………………………………………...139
5.2 IMPLICATIONS AND FUTURE DIRECTIONS…………………………………….141

REFERENCES…………………………………………………………………………………145

v
LIST OF TABLES

Table 2.1 List of SNPs analyzed 49
Table 2.2 Spearman correlation between methylation at each AXL CpG site in the
primary population (N=246)
52
Table 2.3 Demographic characteristics of participants 53
Table 2.4 Association between average DNA methylation levels at AXL CpG
sites and risk of asthma and related symptoms in childhood in the
primary study population (N=246)
54
Table 2.5 Association between DNA methylation levels at selected AXL CpG
sites and risk of asthma and related symptoms in childhood in the
primary study population (N=246)

and replication population (N=1038)
56
Table 2.6 Association between DNA methylation levels at AXL CpG sites and
risk of asthma and related symptoms in childhood in the primary study
population (N=246)
58
Table 2.7 Association between DNA methylation levels at multiple CpG sites
and gene polymorphisms in AXL in the primary study population
(N=207)
60
Table 2.8 Association between DNA methylation levels at selected CpG sites and
gene polymorphisms in AXL in the replication population (N=728)
62
Table 2.9 Sensitivity analysis for adding admixture in testing the association
between AXL DNA methylation and risk of childhood asthma and
related symptoms in the primary population (N=231)
63
Table 2.10 Sensitivity analysis for adding the top 7 principal components (PCs) of
AXL SNPs in testing the association between AXL DNA methylation
and risk of childhood asthma and related symptoms in the primary
population (N=165)
64
Table 2.11 Association between gene polymorphisms in AXL and risk of asthma
and related symptoms in childhood in all CHS samples (N=3845)
65
Table 3.1 Spearman correlation between methylation at each AXL CpG site in the
primary study population (N=923)
85
Table 3.2 Spearman correlation between methylation at each AXL CpG site in the
replication population (N=237)
86
Table 3.3 Characteristics of participants with lung function testing 87
Table 3.4 DNA methylation levels (%) at AXL CpG sites in the primary study
population (N=923) and the replication population (N=237)
88
vi
Table 3.5 Mean levels of growth in lung function during the study period 89
Table 3.6 Association between 8-year lung function growth and AXL methylation
in the primary study population (N=923)
90
Table 3.7 Association between 4-year lung function growth and AXL promoter
methylation in the replication population (N=237)
93
Table 3.8 Sensitivity analyses in non-asthmatic subjects for association between
8-year lung function growth and AXL promoter methylation in the
primary study population
94
Table 3.9 Association between lung function growth and AXL promoter
methylation stratified by sex in the primary study population (N=923)
and replication population (N=237)
95
Table 4.1 Demographic characteristics of study participants (N=248) 121
Table 4.2 Pearson correlation between PM10 levels measured at each month 122
Table 4.3 Association between prenatal PM10 exposure and DNA methylation at
birth
123
Table 4.4 Association between a 2SD increase in prenatal PM10 exposure at each
month and DNA methylation at birth assessed by third-degree
polynomial distributed lag models
124
Table 4.5 Number of PM10-associated CpGs having significant lag-specific
results (P-value < 0.05) for each month
129
Table 4.6 Top 10 enriched Gene Ontology (GO) and Kyoto Encyclopedia of
Genes and Genomes (KEGG) terms for PM10-associated CpGs
130
Table 4.7 Association between DNA methylation at birth and cardiovascular
health outcomes in childhood
132
Table 4.8 Association between DNA methylation at birth and asthma and related
symptoms in childhood
133

vii

LIST OF FIGURES

Figure 2.1 Genomic location of AXL CpG sites and SNPs under investigation 67
Figure 2.2 The association between AXL mRNA and average methylation at
all 12 CpG sites
68
Figure 2.3 The association between cg10564498 methylation and genotype at
each tagging SNP in the replication population (N=728)
69
Figure 2.4 Illustration of epigenetic marks in AXL gene-body region (yellow
box) and CpG sites (red and green bars) in multiple cell lines
63
Figure 3.1 Genomic locations of AXL CpG sites under investigation 96
Figure 3.2 Lung function growth curves for boys and girls over the study
period for FEV1, FVC and MMEF in the primary study population
(Panel A-C) and the replication population (Panel D-F)
97
Figure 3.3 Mean lung function growth versus the average methylation level
of AXL promoter region
98
Figure 3.4 The association between AXL promoter methylation and lung
function growth from 10 to 18 years of age (primary study
population) and 11 to 15 years of age (replication population)
100
Figure 3.5 The association between AXL mRNA and average promoter
methylation
102
Figure 4.1 Distribution of PM10 exposure for each month under investigation. 134
Figure 4.2 Association between a 2SD increase in prenatal PM10 exposure at
each month and DNA methylation
135

1

OVERVIEW
The burden of childhood respiratory disease remains an important public health problem.
Asthma is the most common chronic disease in childhood, affecting approximately 1 in 11
children in the US [1]. The consequences associated with childhood asthma include missed
school days, poor sleep and fatigue, and symptoms that interfere with play, sports or other
activities. Severe asthma attacks even require emergency treatment or hospital care. According to
the Centers for Disease Control and Prevention (CDC), current asthma prevalence increased at a
rate of 1.4% per year among children aged 0-17 years [1], imposing an economic burden of two
billion dollars annually on healthcare costs for managing childhood asthma [2]. Airway
inflammation plays a central role in the pathogenesis of childhood asthma [3]. Although the
etiologies are complex, growing evidence has suggested that childhood asthma is determined by
the interplay between genetic, demographic and both in-utero and early-life environmental
factors [4-7]. In recent years, epigenetic modifications, including DNA methylation, have
emerged as one mechanism underlying the development of childhood asthma by altering
regulation of genes involved in airways development or immune-mediated inflammatory
pathways [8, 9].
In addition to the burden of childhood asthma and other respiratory diseases, normal lung
function development during childhood and adolescence is of great importance since it is a
prerequisite for optimal respiratory health across the life course. Deficits in pulmonary function
have been associated with multiple adverse health outcomes including cardiovascular diseases
and chronic obstructive pulmonary disease (COPD) in adults, and increased risk of developing
asthma in adolescents [10-12]. Healthy lung development is a complex process that can be
modified by genetic, pathological and environmental factors [13-15]. Although DNA
2
methylation has been suggested to play a role in the development of lung function, studies were
mostly conducted in elder subjects or under disease conditions such as COPD [16, 17].
The aim of this dissertation is to explore epigenetic marks associated with childhood asthma
symptoms and lung function development to identify the early origins of chronic diseases and
better understand the underlying mechanisms. In addition, since DNA methylation a dynamic
form of modification and alterations can be induced by environmental factors [18], I also aim to
investigate the effects of prenatal air pollution exposure on DNA methylation patterns at birth.
The candidate gene we selected to study association with childhood respiratory health is AXL, a
member of the TAM family receptor tyrosine kinases. AXL and other TAM family genes are
important mediators for effective phagocytosis and play a crucial role in innate immune
responses [19]. However, the epigenetic regulation of this gene has rarely been studied in
outcomes related to childhood respiratory health.
In Chapter 1, I briefly review the role of DNA methylation in the development of human
diseases, the molecular basis of DNA methylation and how it regulates gene expression. I then
provide a summary for the pathophysiology and known risk factors (demographic, genetic and
environmental factors and epigenetic regulation) of childhood asthma and normal pulmonary
function development. Lastly, I discuss how environmental exposures may affect DNA
methylation.
In Chapter 2, I examine the association between AXL methylation at birth and the risk of
childhood asthma symptoms at age 6 years in subjects from the Children’s Health Study (CHS),
taking into consideration the underlying genetic variation in AXL. Findings are evaluated for
replication in a separate population of CHS subjects using Pyrosequencing. This manuscript was
published in Clinical Epigenetics [20].
3
In Chapter 3, I further explore the association between AXL DNA methylation at birth and
lung function growth during adolescence. We assess the association between AXL methylation
measured by Pyrosequencing and 8-year lung function growth (10 to 18 years of age) in CHS
subjects. Findings are evaluated for replication in a separate population of 237 CHS subjects
using methylation data from the Illumina HumanMethylation450 (HM450) array when possible.
In Chapter 4, I examine the effects of prenatal PM10 exposure on DNA methylation patterns
at birth measured by the HM450 array. I use polynomial distributed lag models to identify CpG
loci associated with PM10 exposure during the 3-month preconception and 9-month pregnancy
window.
Lastly, in Chapter 5 I provide a summary of main findings and suggestions for future
research.

4
CHAPTER 1: INTRODUCTION AND BACKGROUND

1.1 DNA METHYLATION AND HUMAN DISEASE
1.1.1 The role of DNA methylation in human disease
Although we have a nearly complete list of the genes needed to produce a human with the
completion of the Human Genome Project, the development of human disease is far more
complex than a simple catalogue of genes. A second system of equal importance is called
epigenetic marks, which does not alter the DNA sequence but rather, provides an “extra” layer of
transcriptional control that regulates when and where a particular gene will be expressed during
development [21]. Epigenetic modifications include DNA methylation, histone modifications,
microRNAs, small interfering RNAs and long noncoding RNAs, of which DNA methylation has
been most widely studied and is a fundamental determinant of chromatin structure with potent
suppressive effects on gene expression. Epigenetic modification systems are heritable during cell
division and the disruption can lead to inappropriate expression or silencing of genes, leading to
“epigenetic diseases” [22]. Perturbations in the carefully arranged patterns of DNA methylation
and histone modifications have been found to be causative factors in cancer, genetic disorders
and pediatric syndromes as well as contributing factors in autoimmune diseases and aging [23].
The rapidly evolving field of epigenetic disease and therapy offers exciting new opportunities for
the diagnosis and treatment of complex clinical disorders.
DNA methylation is the process by which a methyl group is added to the 5’ position of the
cytosine ring and it predominantly occurs at the cytosine-guanine dinucleotide (CpG) site in
human genome. Properly established and stable DNA methylation patterns are essential for the
regulation of chromatin structure, gene expression and mammalian development. DNA
5
methylation is involved in processes such as X chromosome inactivation, imprinting, embryonic
development, and silencing of repetitive DNA elements [21].
A growing number of human diseases have been associated with alterations in the
establishment and/or maintenance of DNA methylation, of which cancer is the most extensively
studied one. A link between DNA methylation and cancer development was first established in
the 1980s, when it was found that the genomes of cancer cells are hypomethylated relative to
their normal counterparts [24]. Hypomethylation of genomic repetitive elements, a marker of the
global genomic hypomethylation, has been implicated in several malignancies including
urothelial carcinoma [25], prostate cancer [26] and colorectal cancer [27], and often correlates
with disease severity and metastatic potential in many tumor types [28]. Global demethylation
early in tumorigenesis might predispose cells to genomic instability, whereas gene-specific
demethylation also occurs and promotes aberrant expression of oncogenes and loss of imprinting
[29]. In contrast to global DNA hypomethylation, aberrant hypermethylation in cancer
development mainly occurs at CpG islands, most of which are unmethylated in normal somatic
cells [30]. The transcriptional inactivation caused by promoter hypermethylation affects genes
involved in the main cellular pathways: cell-cycle regulation, tumor cell invasion, DNA repair,
chromatin remodeling, cell signaling, transcription and apoptosis [31]. Therefore,
hypermethylated promoters of specific genes have been proposed as a new generation of
biomarkers with great diagnostic and prognostic potential for clinicians [32].
The rapid development in unraveling the epigenetic etiology of cancer has encouraged the
development of epigenetic therapy and several have been approved for specific cancer types [33].
Hypomethylating agents such as azacitidine were shown effective in clinical trials in treating
myelodysplastic syndrome and leukemias characterized by gene hypermethylation [34].
6
Epigenetic profiles can also be used to identify molecular pathways most sensitive to cancer
drugs as a means of prioritizing therapeutic strategies, and to measure treatment efficacy and
disease progression [35]. For example, methylation of PITX2 can be used to predict outcomes of
individuals with early-stage breast cancer after treatment with adjuvant tamoxifen therapy [36].
DNA methylation also plays a role in common non-neoplastic human diseases. Some of
these involve abnormal imprinting, including behavioral disorders like Rett syndrome, Prader-
Willi syndrome and Turner syndrome, and neurodevelopmental disorders like autism and bipolar
disorder [37, 38]. Another common disease with an epigenetic component is systemic
autoimmune disease, in which aberrant methylation in T cells can result in T-cell autoreactivity
and is associated with a range of autoimmune disorders including inflammatory arthritis and
systemic lupus erythematosus (SLE) [39].
Given the intriguing insights from recent findings on the role of DNA methylation in the
control of cell development and disease pathogenesis, research efforts studying the epigenetic
regulation of human disease will advance our understanding of gene regulation and epigenetic
programming, and fuel the development of new therapeutic targets for complex disease.
1.1.2 Overview of DNA methylation
The mammalian DNA methylation patterns are established and maintained by DNA
methyltransferases (DNMTs) and the methyl-CpG binding proteins (MBDs) are involved in
“reading” methylation marks. 70-80% of the CpGs in human genome are generally methylated
[40] and the rest non-methylated CpGs are mostly occurring in CG-rich regions, referred to as
CpG islands (CGI), that are prevalent at transcription start sites and housekeeping and
developmental regulator genes [41]. Epigenetic changes are heritable through cell division and
may be transferred to future generations. Despite the high correlation across somatic tissues,
7
genome-wide methylation profiling exhibits tissue-specific patterns in certain genes and regions
related to tissue-specific functions [42].
When located at gene promoters, DNA methylation is usually a mark for repression of
transcription. This is achieved mainly through directly inhibiting the binding of transcription
factors to their recognition sequences, or through recruiting MBDs and their associated
corepressors to change chromatin structure from active to inactive form [43]. However, recent
evidence has shown that despite the widely-accepted role as a ‘silencing’ epigenetic mark, the
function of DNA methylation varies with different genomic contexts [44]. For instance, DNA
methylation in the gene body is often correlated with active transcription and may have an
impact on splicing [45].
The DNA methylation profile itself is dynamic and alterations are induced by age-related
factors, environmental and nutritional factors, and pathogenic factors such as viruses [18, 46,
47]. The effects of environmental factors on DNA methylation pattern will be described in more
detail in section 1.5. Recent work also suggests that genetic variation, for example, single
nucleotide polymorphisms (SNPs) may have a substantial impact on local methylation patterns.
These SNPs are therefore defined as DNA methylation quantitative trait loci (metQTL) [48, 49].
Specifically, methylation levels showed strong correlation with genotypes in their proximity (cis-
acting) and trans-metQTLs are much rarer, mainly affecting promoter CGI regions possibly
through the mediation effect of proximal CpG sites [49, 50], suggesting the contribution of local
genomic environment to establishing these associations. Furthermore, the genetic regulation of
methylation showed high similarity across different somatic tissue types [50, 51].
The mechanisms by which genetic variation influences DNA methylation levels are still far
from clear, but may involve the disruption of a CpG methylation site [49] and the binding of
8
transcription factors, which in turn could influence DNA methylation [52, 53]. However,
whether the effects of genetic variation on DNA methylation may further affect transcription
remains controversial and can vary depending on the tissue and genomic region. Bell et al.
observed a significant overlap of SNPs that were associated with both methylation and gene
expression levels in lymphoblastoid cell lines [54]. On the other hand, several other studies
found the majority of metQTLs were not directly associated with expression changes in adipose
tissue and brain tissue [55, 56], or suggested the allele-specific expression was not driven by
differential DNA methylation [53]. Nonetheless, the regulation of gene expression involves
many factors other than DNA methylation and genetic variants, and studying metQTL still helps
to elucidate the mechanisms through which the methylation landscape is established, and
facilitates the functional interpretation of phenotype-associated genetic variations.

1.2 OVERVIEW OF CHILDHOOD ASTHMA
1.2.1 Introduction and pathophysiology
Asthma is the most common chronic disease in childhood with increasing incidence and
prevalence, affecting approximately 1 in 11 children in the US [1]. The consequences associated
with childhood asthma include missed school days, poor sleep and fatigue, and symptoms that
interfere with play, sports or other activities. Severe asthma attacks even require emergency
treatment or hospital care. The treatment and management of childhood asthma imposes great
economic burden to the society. It is estimated that two billion dollars in the US are spent
annually on healthcare costs for managing childhood asthma [2].
Asthma is an inflammatory disease with common symptoms like frequent coughing,
wheezing or whistling sound in the chest especially when breathing out, chest tightness,
9
shortness of breath and feelings of weakness or tiredness. Symptoms may worsen when the
asthma triggers are present, such as allergens (pollen or dust mites) or irritants in the air (smoke
or strong odors). Although the disease can begin at any age, most children who have asthma
develop their first symptoms by age 5 [57]. According to the National Asthma Education and
Prevention Program’s (NAEPP) Expert Panel Report 3 (EPR-3), the current recommended
guidelines for asthma diagnosis include presence of episodic symptoms of wheeze, cough,
shortness of breath, and chest tightness; determining reversible airflow obstruction (FEV1 < 80%
predicted; FEV1/FVC < 65% or below the lower limit of normal); and exclusion of alternative
diagnoses (e.g., vocal cord dysfunction, vascular rings, foreign bodies, or other pulmonary
diseases). Additionally, the EPR-3 recommends further testing for suspected asthma cases, such
as assessing diurnal variation in peak flow, chest X-ray, or determining allergic sensitization
[58]. In epidemiological studies, the use of parents-completed questionnaire-based report of
physician-diagnosed asthma has been widely accepted as a valid method of classifying asthma
status [59].
The major pathophysiological elements of asthma include chronic airway inflammation,
airway remodeling, and airway hyperresponsiveness. Airway inflammation is undoubtedly
playing a central role [3]. It can be triggered by many external factors such as exposure to
allergens, viruses, and indoor and outdoor air pollutants and involves the interaction of many cell
types and mediators [60-62]. Airway inflammation starts from an early phase when
immunoglobulin E (IgE) on mast cells in the airway is bound by allergen, causing degranulation
of mast cells and production of proinflammatory mediators such as histamine and reactive
oxygen species. This is followed by the recruitment of inflammatory cells and synthesis of
cytokines (e.g., Interleukin-3 (IL-3), IL-4, IL-5, IL-6, IL-8, IL-13 and tumor necrosis factor
10
(TNF)) that further promote type 2 T-helper cell(TH2)-mediated immune responses [63]. These
inflammatory responses, if occurring persistently, will lead to chronic airway inflammation and
potential structural changes, such as thickening of the basal membrane, increased mucous glands,
hyperplasia of airway smooth muscle and goblet cells, and changes in extracellular matrix
composition, a process commonly known as airway remodeling [64]. These changes may
increase airway wall thickness, possibly explaining the poor response to corticosteroids in
established asthma, and further lead to accelerated lung function decline and irreversible airflow
obstruction [65, 66].
Airway hyperresponsiveness (AHR) is another key feature of asthma which correlates with
severity of symptoms, decline in lung function and the need for treatment. It is characterized by
increased sensitivity of the airways to a variety of pharmacological, chemical and physical
stimuli, such as histamine, methacholine, AMP, sulphur dioxide, fog and cold air [67]. The
mechanisms of AHR are related to both airway inflammation and airway remodeling. The acute
and variable AHR reflects episodic increase in airway inflammation due to environmental
triggers such as allergen exposure [68, 69]. The underlying chronic component of AHR is more
persistent and is relatively refractory to environmental factors and inhaled corticosteroids. This
probably relates to structural changes and airway remodeling due to chronic recurrent airway
inflammation [70].
Asthma-related symptoms, such as recurrent wheezing is common in early childhood.
Approximately one-third of all children have suffered from wheezing during respiratory
infection before age 3 years, rising to almost one in two (50%) by age 6 [71]. The presence of
airway inflammation has also been implicated in young children who were experiencing
wheezing symptoms [72, 73]. The relationship between early wheezing and subsequent
11
development of asthma, however, is controversial. Whilst most early childhood wheezers have
transient symptoms, yet only a small proportion develop persistent symptoms and have, or
develop, asthma [71]. On the other hand, there is growing evidence suggesting long-term
outcome even years after cessation of wheezing. Studies have shown a tendency for relapsing
symptoms during late childhood and adolescence [74], and have linked wheezing episodes by
age 3 years to 75% of childhood asthma cases [75]. Early-life wheezing is also associated with
later lung function abnormalities and faster lung function decline from the age of 40 years
onwards [76, 77].
Bronchitis and associated chronic respiratory symptoms – cough and phlegm, are referred to
as bronchitic symptoms in this dissertation. Bronchitic symptoms are common yet
underappreciated aspects of clinically important morbidity [78]. They are suggestive of chronic
symptoms that may follow an illness or acute exacerbation of asthma, or chronic inflammation in
the airway. Therefore, wheezing and bronchitic symptoms may also serve as pathological
markers for airway inflammation, early signs of childhood asthma, or asthma-induced respiratory
symptoms.
1.2.2 Risk factors
Although the etiologies of childhood asthma are complex, growing evidence has suggested
that it is determined by the interplay between genetic, demographic and both in-utero and early-
life environmental factors.
Genetic factors
Family and twin studies have indicated that genetic predisposition is a strong determinant of
childhood asthma [79]. Since asthma is a complex and heterogeneous disease and requires the
activation of many different cell types (B and T cells, dendritic cells, macrophages, eosinophils,
12
smooth muscle cells, etc.) and molecules (cytokines, inflammatory molecules, growth factors
and intracellular mediators), it is not surprising that many genes have been reported to be
involved. Genome-wide linkage studies and case control studies identified multiple genomic
regions and more than 100 genes associated with allergy and asthma in different populations
[80]. Candidate genes involved in innate immunity, TH2-cell differentiation and airway
inflammation and remodeling (e.g., GSTP1, IL13, and NOS1) have also gained extensive
attention [81-83]. Meanwhile, gene-environment interaction studies suggested that the function
of a genetic variant may also be amenable to modification by environmental exposures, such as
second-hand tobacco smoke and air pollution [84, 85].
Social and demographic factors
Numerous social, demographic and psychological factors have been associated with asthma,
such as sex, socioeconomic status (SES) and race/ethnicity [86]. The association between sex
and childhood asthma follows a time-dependent manner. The incidence and prevalence of
asthma are greater among boys than girls in childhood, but a higher incidence is shown among
adolescent females after puberty [87]. The influence of some environmental risk factors may also
be modified by sex. For example, obesity showed a larger effect on the development of asthma
among women than among men and this was not dependent on caloric intake or physical activity
[88]. The association between SES and asthma remains unclear and complex. While findings
with respect to SES and the prevalence of asthma are mixed, children of parents with lower SES
have greater morbidity from asthma [89, 90]. Studies have also suggested that part of the
observed racial/ethnic differences in asthma prevalence can be explained by factors related to
income, area of residence, and level of education [90].
Environmental factors
13
Studies examining prenatal dietary factors and childhood asthma have identified protective
effects from foods with anti-inflammatory properties (e.g., fish or fish oil that contain omega-3
fatty acids) and antioxidants such as vitamin E and zinc [91, 92]. Prenatal and early life exposure
to environmental toxicants, such as pesticides and chemical household products [93, 94], tobacco
smoke and air pollution, have been associated with higher risk for childhood asthma [95, 96].
Evidence for the direct association between exposure to indoor allergens and the development of
childhood asthma, however, is not conclusive [97, 98].
Both prenatal maternal smoking and early-life exposure to environmental tobacco smoke
(ETS) have been consistently associated with increased incidence of wheezing, airway
hyperresponsiveness, and asthma in children [95, 99, 100]. There is some evidence indicating a
dose-dependent manner of increased asthma risk with higher consumption of cigarettes during
pregnancy [100]. Functional variants in genes in the oxidant stress pathway such as GSTM1
could also modify the association between maternal smoking and asthma occurrence by altering
genetic susceptibility [101].
With rapid urbanization and increasing amount of traffic exhausts, a number of studies have
emerged to address the effects of air pollution on asthma. There has been an established
association between ambient and traffic-related air pollution exposure and asthma symptoms and
morbidity [6, 96, 102]. Besides, indoor air pollutants are another source of factors that may affect
asthma development, such as nitrogen dioxide (NO2) and particulate matter (PM) that come from
cooking exhaust, wood-burning stoves and fireplaces, and penetration of outdoor particles [103].
Whether air pollution contributes to the development of new asthma cases, however, remains
controversial due to the inconsistency in current findings [104, 105]. Although animal studies
provide some insights into the potential effect of in-utero exposure to air pollutants on asthma
14
occurrence in the offspring [106], epidemiological studies are still warranted to evaluate this
association in human.
1.2.3 Epigenetic regulation
Although genetic component plays an important role in childhood asthma, genome-wide
association studies (GWAS) have demonstrated to account for only a modest proportion of the
total phenotypic variance in asthma [107]. In recent years, epigenetic regulation has emerged as
one mechanism underlying the development of childhood asthma. Altered DNA methylation
status may lead to differential gene expression of cytokines and transcription factors involved in
asthma pathogenesis [108]. There is a growing body of evidence suggesting epigenetics as a
mediating factor that associates environmental exposures to childhood asthma-related
phenotypes [109].
DNA methylation in candidate genes in peripheral blood cells, buccal cells, or nasal
epithelia has been related to childhood asthma [110-112]. Importantly, epigenetic mechanisms
are shown to affect the expression of transcription factors involved in the development of critical
immune cells in asthma pathogenesis, such as mature T lymphocytes (TH1, TH2, and regulatory T
cells), and the polarization towards the TH2 phenotype [113, 114]. Many of these studies,
however, focused either on the critical immune pathways in asthma such as T cell differentiation,
or asthma candidate genes harboring SNPs that showed associations with asthma [115, 116].
Few studies have addressed the potential role of environmental exposure-associated genes that
could be involved in airway inflammation and innate immune responses. Several epigenome-
wide association studies (EWAS) of DNA methylation in relation to childhood asthma identified
differential methylation at specific regions or genes, but the possibility of reverse causality could
not be ruled out due to the cross-sectional nature of these studies [117-119].
15
Despite the rapid progress in studying epigenetic regulation of asthma, there are, however,
many challenges and unanswered questions facing epidemiologists. First of all, there are limited
evidence relating childhood asthma to methylation patterns at birth, when much of the
epigenome has already been established and could be reflective of prenatal environmental
exposures such as maternal smoking, dietary pattern and environmental toxicants [120]. The
notion is consistent with the emerging paradigm that chronic non-communicable diseases have
their origins in early life through an epigenetic calibration of set points for later responsiveness
and function [120]. Secondly, since epigenetic changes are dynamic and tissue- and cell-specific,
choosing the optimal type and timing of DNA samples becomes a critical question. Undefined
mixture of cells such as whole blood samples are easy to obtain but need to be analyzed critically
in studies of airway diseases, although the existence of immune cell populations in whole blood
may lend some information. On the other hand, pathologically more relevant samples such as
lung are ideal but acquiring such samples in children is almost impossible. Thus, future studies
are warranted to assess the persistence of methylation patterns across time and different cell
types. Lastly, more work is also needed to evaluate how epigenetic profiles mediate the effects of
environmental exposure on risk of childhood asthma, and whether these epigenetic changes are
heritable to the next generation.

1.3 OVERVIEW OF CHILDHOOD LUNG DEVELOPMENT
1.3.1 Introduction
Pulmonary function development
Pulmonary function is frequently used as quantitative index of respiratory health. Deficits in
pulmonary function have been associated with multiple adverse health outcomes including
16
cardiovascular diseases, and chronic obstructive pulmonary disease (COPD) in adults, and
increased risk of developing asthma in children [10-12].
The normal development of lung starts in utero, continues through adolescence and early
adulthood and is dependent on age, sex, height and ethnicity [121, 122]. Lung function increases
linearly with age until adolescence when significant sex differences develop. Girls have an
earlier growth spurt at about age 10 years, while the growth spurt for boys are around 12 but is
faster in rate and longer in duration, making the difference between boys and girls continues to
widen until at least age 18 [123]. Lung function reaches a maximum plateau at around 22 years
of age in men and slightly earlier in women, and then starts a slow but steady annual decline due
to gradual loss of lung elasticity [122].
Normal lung function development during childhood and adolescence is of great importance
since it is a prerequisite for optimal respiratory health across the life course. Adverse factors
leading to declined lung function growth rates during this time may lead to reduced maximal
attained lung function and accelerated lung function decline in adulthood, and further respiratory
symptoms and increased risk of cardiovascular disease earlier in life than otherwise might be
expected [124, 125].
Pulmonary function testing and definition of measures
The pulmonary function testing is usually performed with forced maximal effort maneuvers
using a spirometer. Common measures of pulmonary function include:
(1) Forced vital capacity (FVC): the total amount of air exhaled forcefully after inhaling as
deeply as possible. This is often considered as the best measure of lung capacity.
(2) Forced expiratory volume (FEV1): the amount of air expired during the first second of the
FVC test. FEV1 is a useful measure of how quickly the lungs can be emptied and is often
17
considered as the best single measure of lung function. It captures information about both
lung volume and flow rate. The FEV1/FVC ratio also gives a clinically useful index of
airway limitation.
(3) Forced expiratory flow at 25%, 50%, 75% of FVC (FEF25, FEF50, FEF75): the flow rate at
which the 25%, 50%, 75% point of the total volume (FVC) has been exhaled. FEF25-75
(also referred to as maximum mid-expiratory flow (MMEF)) measures the middle half of
the FVC, and is considered by many physicians as an indicator of obstruction in the small
airways.
(4) Peak expiratory flow rate (PEFR): the fastest rate at which air is being exhaled forcefully.
This landmark is important in judging the strength of expiratory muscles and condition of
the large airways, such as the trachea and main bronchi.
1.3.2 Risk factors
Since lung development in childhood and adolescence is closely related to respiratory health
in adult life, understanding the factors affecting lung function in children is critical for
preventing subsequent diseases later in life. Healthy lung development is a complex process that
can be modified by genetic, pathological and environmental factors.
Genetic factors
Results from familial aggregation studies have suggested the role of genetic component in
pulmonary function development [13]. Several candidate genes involved in important pathways
for lung development have been identified as important factors in childhood pulmonary function
growth, including the glutathione-S-transferase (GST) family of genes [126, 127] and the NOS2A
region [128]. GSTM1 genotype was also found to modify the effects of in utero smoke exposure
on MMEF levels in a study of German school children aged 9 to 11 years [129]. Genome-wide
18
analyses have also identified novel genes harboring SNPs associated with pulmonary function,
including genes involved in the hedgehog (Hh) signaling pathway, inflammation, cell signaling,
and immune-mediated airway reactivity [130, 131].
Asthma and airway inflammation
Asthma exacerbation is known to associate with reduced lung function [14]. Studies on the
chronic effects of asthma have identified a progressive reduction in lung function in children
with mild to moderate asthma and suggested the deficits may persist into adulthood [132, 133].
In the CHS, larger deficits in MMEF and FEF75 were seen in males than in females with asthma
[133]. Children with in utero tobacco smoke exposure and early onset asthma (<=5 years of age)
showed largest deficits in lung function compared to either exposure alone [134]. The deficits in
lung function growth associated with childhood asthma may involve airway remodeling resulting
from persistent inflammatory effects of undertreated asthma [135].
An important pathological feature of asthma, airway hyperresponsiveness (AHR), is also
related to deficits in lung function growth but independently of asthmatic symptoms, the
mechanisms behind which are unclear [136, 137]. Harmsen et al. reported that AHR was
independently associated with reduced rates of lung function growth in both boys and girls and
declines in maximal attained level of lung function at age 18, after adjustment for asthma and
smoking. Furthermore, the declines persisted throughout the follow-up until the last examination
at age 27-37 years [138].
Environmental factors
Many studies have been conducted to evaluate environmental factors associated with lung
function growth during childhood and adolescence. Intrauterine growth restriction, for example,
is a risk factor for reduced lung function and respiratory morbidity during childhood and
19
adulthood [139, 140]. Maternal vitamin D deficiency and alcohol consumption during pregnancy
may also play a role in declined lung development, but epidemiological studies are lacking to
support direct associations in human [141, 142].
The effects of tobacco smoke exposure on lung development have been extensively studied.
Nicotine can cause abnormal airway branching, thicker alveolar walls, increased airway smooth
muscle and collagen deposition, and AHR with airflow restriction [15]. Exposure to maternal
smoking prenatally was associated with poor lung function during childhood and adulthood [143,
144]. In the CHS, exposure to environmental tobacco smoke (ETS) was associated with reduced
MMEF and FEF75 in childhood [144]. Further investigation suggested the association between
in-utero maternal smoking and lung function deficits was stronger among children with asthma,
and the effect of ETS exposure varied by children’s sex and asthma status [145].
Air pollution is considered a serious public health issue in the past few decades. Children are
considered more susceptible to the detrimental effects of air pollution than adults since their lung
are still under development and they are likely to have higher exposure to air pollutants due to
more outdoor activities [146]. The association between air pollution exposure and reduced lung
function in children has been well described. Cross-sectional studies showed associations of
regional particulate matter (PM10 and PM2.5), nitrogen dioxide (NO2) and O3 with lower levels
for a number of lung function measures [147-149]. Longitudinal analyses indicated that higher
concentrations of regional NO2, acid vapor, PM10, PM2.5 and elemental carbon were also
associated with lower lung function growth [150, 151]. Correspondingly, the adverse effects of
air pollution can be reversed with decreasing exposure, as reported in a recent CHS study, which
found long-term improvements in air quality were associated with statistically and clinically
significant positive effects of lung function growth in children [152].
20
The importance of near-roadway pollutants has been recently acknowledged as having
potentially different toxicological effects compared to regional pollutants. Truck density was
found to have negative association with lung function among children living within 300 meters
of motorways. In the same study lower measures of lung function were found to associate with
measured concentrations of black smoke and NO2 in school attended by these children [153]. In
the CHS, children living within 500 meters of a freeway showed significant deficits in FEV1 and
MMEF growth compared to children living farther than 1500 meters from a freeway, and the
effects were independent from regional air pollution [154].
1.3.3 Epigenetic regulation
Epigenetic changes represent a form of molecular integrator of different signals affecting
disease susceptibility, linking the environment and phenotype. DNA methylation has emerged as
one mechanism underlying the development of lung function. Many of the studies investigating
epigenetic regulation of lung development were conducted in the elderly population or under
disease conditions such as COPD. In the Normative Aging Study (NAS) of older and primarily
white men from North America, the researchers investigated the methylation status of Alu and
LINE-1, which can be representative of total genomic methylation content [155, 156]. Alu
hypomethylation was associated with lower lung function and LINE-1 hypomethylation was
associated with more rapid decline in FEV1 and FVC. The role of inflammatory genes was also
studied in this cohort and researchers found that decreased methylation in CRAT, F3 and TLR2
was significantly associated with lower lung function, while decreased IFN𝛾 and IL6
methylation was associated with better lung function [16]. The results suggested the role of DNA
methylation in lung function decline. Furthermore, methylation status of inflammation- and
immunity-related genes might influence the association between traffic-related air pollutants and
21
lung function in the elderly [157]. Methylation of one asthma-associated gene, IL-13 was found
to interact with its genotype to affect asthma-related lung function [17]. In COPD patients,
genome-wide scan identified significant and replicable associations between SERPINA1
hypomethylation with COPD and lower average lung function phenotypes [158].
To date, the epigenetic regulation of lung development is still a relatively new area and there
are several unanswered questions we will be facing. First of all, in light of the recently expanded
developmental origins of disease hypothesis that describes early events altering lung
development and disease susceptibility, animal studies have identified epigenetic changes in
several genes induced by prenatal factors and may be associated with lung development [159,
160]. However, epidemiological studies are lacking to support the effects of early-life epigenetic
changes on lung development, especially in healthy children and adolescents. Since lung
function growth during this time is closely related to the maximal achieved level in adulthood
and the risk of cardiorespiratory disease in later life, identifying epigenetic biomarkers of lung
growth helps us to understand the underlying molecular mechanisms of chronic diseases.
Secondly, in studies involving children, peripheral blood may be the only accessible tissue. Thus,
paired lung-blood samples should be analyzed to understand the epigenetic marks that carry over
from the lung to peripheral blood. Thirdly, due to the dynamic nature of epigenetic marks, such
studies should be interpreted in the context of other exposures, diet and age.

1.4 INTRODUCTION OF CANDIDATE GENE – AXL
1.4.1 TAM family genes
We selected a member of the TAM receptor tyrosine kinases family, AXL, as our candidate
for gene for investigating the associations with childhood respiratory health. We previously
22
reported that DNA methylation of AXL in childhood was associated with in utero exposure to
maternal smoking in a genome-wide analysis based on microarrays followed by replication with
pyrosequencing. The examined CpG site in AXL was located in the core promoter region
including a Sp1/Sp3 transcription factor binding site where DNA methylation level correlates
with gene expression of AXL, supporting a functional role of the DNA methylation status [161,
162]. Follow-up studies suggested the effects of in utero maternal smoking on DNA methylation
in AXL were magnified in girls as compared to boys [163]. We then investigated the potential
immunological role of AXL and other members of TAM family to evaluate whether this gene
may be involved in childhood asthma pathogenesis and lung development.
Overview of TAM receptors and ligands
The TAM receptors – TYRO3, AXL and MER – is a unique family of receptor tyrosine
kinases (RTK). They were first described as crucial mediators for the engulfment and phagocytic
clearance of apoptotic cells and membranes in adult tissues. RTKs are cell-surface
transmembrane receptors that function as sensors for extracellular ligands, the binding of which
triggers activation of the receptor’s kinase and multiple downstream signaling proteins [164].
Two closely related proteins, growth-arrest-specific 6 (GAS6) and protein S (PROS1), act as
ligands for TAM receptors. GAS6 can bind and activate all three receptors, but with different
affinities (AXL ≥ TYRO3 >> MER) and is often called the GAS6/AXL pathway, while protein S
was found to be a major agonist of TYRO3 and potentially MER, but little or no affinity for AXL
[165, 166].
TAM receptors are widely expressed in cells of the mature immune (dendritic cells,
macrophages, lymphocytes, natural killer cells, etc.), nervous, reproductive and vascular
systems, and usually there is expression of more than one TAM receptor in a given cell type
23
[164]. They regulate a wide variety of biological processes, including cell proliferations/survival,
cell adhesion and migration, release of inflammatory cytokines, blood coagulation and cancer
progression. The generation of mice lacking Tyro3, Axl and Mer has enabled us to make
significant progress in understanding the biological role of the TAM pathway [167, 168]. The
triple knockout mice were found to be viable at birth, but developed a plethora of debilitating
phenotypes such as loss of photoreceptors, male infertility, profound lymphoproliferation and
broad-spectrum autoimmune disease. These degenerative phenotypes were associated with
TAM’s regulation of two likely functionally-linked phenomena: the phagocytosis of apoptotic
cells, and the innate inflammatory response to pathogens by dendritic cells(DCs) and
macrophages, as discussed below.
TAM regulation of phagocytosis
TAM receptor signaling plays a critical role in mediating the engulfment and phagocytosis
of apoptotic cells and membranes in adult tissues [164, 169]. This process mediated by TAM
receptors is essential for phagocytosis that continuously occurs in adult organs to maintain
normal tissue homeostasis, including, but not restricted to the immune system [164]. The TAM
ligands, GAS6 or protein S, serves as a ‘bridging molecule’ that physically links a TAM receptor
expressed on the surface of the phagocyte to PtdSer that is displayed on the surface of the soon-
to-be engulfed apoptotic cell [170]. The male infertility in TAM triple knock-out (KO) mice was
explained by the inefficient removal of apoptotic germ cells and apoptotic cell debris generated
during spermatogenesis, causing accumulation of apoptotic cell corpses and eventually poisoning
of the tubule epithelium [167, 171]. A remarkably similar phenotype observed in TAM triple KO
mice, loss of photoreceptors (PRs), reflects the loss of TAM signaling in a specialized
phagocytic cell – the retinal pigment epithelia (RPE) cell. The failure of RPE cells to perform the
24
phagocytosis of outer segments of PRs led to the nonautonomous apoptotic death of all PRs in
the retina [166].
Parallel evidence for TAM’s role in apoptosis has been found in the immune system.
Macrophages and DCs are responsible for clearing a large number of apoptotic-cell corpses
generated by bacterial and viral infections. However this form of homeostatic phagocytosis is
also impaired in the TAM-deficient mice, which is likely to contribute to the development of
autoimmune diseases since an elevated steady-state number of apoptotic cells constitutes a
source of self-antigens [172, 173]. Recent evidence suggests that all three TAM receptors
contribute to the phagocytosis of apoptotic cells by DCs and macrophages, albeit to different
degrees [172].
TAM regulation of the innate immune response
The innate immune response is a dynamic system that must be carefully regulated. It needs
to initiate a rapid inflammatory response to combat infection from bacteria, viruses and other
disease-causing pathogens; yet unrestrained signaling by Toll-like receptors (TLRs) and
cytokines in DCs and macrophages will lead to a chronic inflammation milieu and further
autoimmune diseases. Recent evidence suggested the function of TAM receptors as pivotal
inhibitors that prevent this dysregulation from occurring [174]. In DCs and other sentinel cells,
the presence of invariant molecular patterns associated with bacteria, viruses and other pathogens
are recognized by TLRs and this further leads to the production of proinflammatory cytokines
such as tumor necrosis factor (TNF) 𝛼 , IL-6, and type I interferons [175]. The TAM receptors
inhibit unrestrained signaling from these cytokines by binding to the type I IFN receptor (IFNAR)
and switching the IFNAR signaling complex from proinflammatory to one that inhibits
inflammation. This is achieved by activation of the IFNAR-induced genes encoding the
25
suppressor of cytokine signaling (SOCS) 1 and 3. The SOCS proteins are well known to be
induced by cytokine receptor activation and involved in the classic negative-feedback loop of
both cytokine receptor and TLR pathways [174, 176, 177]. Thus, the induced SOCS proteins,
whose expression is very largely dependent on the activation of the TAM-IFNAR multimeric
complex, terminate the inflammatory response to pathogens in DCs [174].
Apart from inhibiting unrestrained signaling of cytokines induced by TLRs, the TAM
signaling system also has an important role in regulating the activity of natural killer (NK) cells
that recognize and destroy cells infected with pathogens [178]. All three TAM receptors are
expressed in NK cells in the bone marrow, and the maturation of NK cells is driven by stromal
cells of the bone marrow which express ligands for TAM receptors [178, 179]. Natural killer
cells isolated from TAM-deficient mice were found to have very poor killing ability against
target cells compared to wild-type NK cells [180].
TAM and autoimmune disease
Given the pivotal role of TAM receptors in regulating the immune response, it is therefore
not surprising that mouse mutants in TAM receptor genes eventually develop broad-spectrum
autoimmune disease having clinical features of both systemic lupus erythematosus (SLE) and
rheumatoid arthritis (RA) [168, 181]. TAM mutant mice also display relatively high titers of
antibodies and autoantigens [168, 182]. Both defects in the clearance of apoptotic cells and loss
of TAM receptor regulation in the innate inflammatory response are thought to contribute to
these autoimmune disorders. Correspondingly, there is evidence suggesting diminished TAM
signaling may be associated with human autoimmunity [181]. Levels of free protein S were
significantly lower in SLE patients with a history of serositis, neurologic, hematologic and
immunologic disorder [183]. Polymorphisms in MER gene have been tied to SLE and multiple
26
sclerosis [184, 185]. MER has also been identified as a potential therapeutic target in the context
of autoimmune disease [186].
TAM regulation of the vasculature system
TAM signaling is also involved in the homeostatic regulation of blood vessel integrity and
permeability. Vascular endothelial cells are a major source of circulated protein S and AXL and
TYRO3 are also expressed by the vascular smooth muscle cells surrounding these endothelia
[187, 188]. A complex pattern of differential regulation of AXL, MER, GAS6 and PROS1 has
been reported in human atherosclerotic plaque and possible mechanism includes deficits in the
clearance of apoptotic cells from these plaques [189]. TAM signaling also affects vascular
integrity through the regulation of platelet function [190].
1.4.2 Biological function of AXL
AXL, also called UFO, ARK, or TYRO7, was originally identified as a transforming gene in
human leukemia and other cancers [191, 192]. It is ubiquitously expressed and detectable in a
wide range of organs and cell lines of epithelial, mesenchymal and haematopoietic origin, but
absent from lymphocytes and granulocytes [192].
The AXL gene transcription and expression is regulated by a variety of transcription factors
and small molecules. The promoter region of AXL contains potential recognition sites for a
variety of transcription factors, including Sp1 (specificity protein 1), AP-2 (activating protein 2)
and CREB (cyclic AMP response element binding protein) [193]. Recent research in colon
cancer cells revealed a minimal GC-reach region (-556 to +7 upstream of the translational start
site) that contains five Sp-binding sites. Specifically, Sp1/Sp3 transcription factors can bind to
this region and regulate AXL expression. Methylation of CpG sites within specific Sp1 motifs is
negatively associated with AXL expression [162]. Other transcription factors, such as MZF1
27
(myeloid zinc finger 1), can bind to AXL promoter and activate its expression [194]. Moreover,
several microRNAs, including miR-34a, miR-199a and miR-199b target the 3’-UTR untranslated
region of AXL gene, and inhibit its expression in several cancer cell lines [195].
Overexpression and increased activity of AXL has been implicated in the progression of
various chronic pathological conditions. A large body of evidence suggests that AXL plays an
important role in cancer pathological conditions. Up-regulation of AXL protein is associated with
a vast majority of tumors such as breast cancer and mesothelioma, probably through increasing
tumor cell survival, migration and angiogenesis [196, 197]. The GAS6/AXL pathway also
regulates thrombosis and is crucial for cardiovascular events, particularly under pathological
conditions [198]. As described before, the GAS6/AXL pathway inhibits TLR and cytokine
receptor signaling in innate immune cells via suppressing proinflammatory signals [174, 199],
but the studies on the immune-regulatory role of AXL were mostly conducted in animal models
or in severe human autoimmune diseases such as SLE. Few studies have investigated the role of
AXL in inflammatory disease of the respiratory system, such as asthma, and the epigenetic
regulation of AXL. There is evidence showing that GAS6 exhibited higher expression in subjects
with severe asthma during exacerbation [200]. Moreover, genetic polymorphisms in TAM genes
and their ligands have also been implicated in inflammation and autoimmune diseases [201,
202]. The identification of epigenetic and genetic variation of AXL associated with childhood
asthma symptoms may shed light on the etiology of this complex disease and the biological role
of AXL.
Recent studies also suggest AXL’s role in regulating lung homeostasis. AXL was expressed
in both human and mouse airway/alveolar macrophages and was constitutively bound to its
ligand GAS6. AXL was critical for effective phagocytosis and reduced AXL expression may
28
contribute to persistent airway inflammation through inefficient clearance of apoptotic cells from
the inflamed lungs [203, 204]. Similar findings were also reported in human patients with
moderate-to-severe asthma, suggesting AXL’s role in preventing excessive inflammation in the
lung and regulating lung immune homeostasis [204].
Given the potential importance of AXL in regulating innate immune responses and
suppressing uncontrolled inflammatory signals, we hypothesized the epigenetic changes of AXL
may be associated with childhood asthma and related symptoms, which primarily involves
airway inflammation. Furthermore, in light of the findings from animal studies on AXL’s role in
effective phagocytosis in the lung and regulating lung immune homeostasis, we also sought to
investigate the epigenetic regulation of AXL in lung function growth during adolescence.

1.5 ENVIRONMENTAL EPIGENETICS
Since epigenetic modifications are dynamic and can be influenced by environmental
exposures such as chemicals, air pollutants, diet and aging, recent studies have explored the
epigenetic changes in various exposure scenarios, and how this is related to disease outcomes.
Studies in adults have demonstrated epigenetic changes in relation to environmental exposure to
metals, air pollution, benzene, pesticides and persistent organic pollutants [205]. For example,
low-dose exposure to airborne benzene in a cohort of police officers and gas-station attendants
was associated with hypomethylation of LINE1- and Alu repetitive elements in blood,
hypermethylation of p15 tumor suppressor gene, and hypomethylation of MAGEA1 (melanoma-
associated antigen 1) gene [206].
Among these environmental factors, particulate matter (PM) has been associated with a
variety of cardiovascular and respiratory diseases and epigenetic alterations induced by PM are
29
proposed as a potential mechanism [207, 208]. PM is a heterogeneous mixture of solid particles
and liquid droplets existing in the air and contains up to hundreds of different inorganic and
organic chemicals [209]. Ambient PM can come directly from sources such as construction sites,
unpaved roads, smokestacks or fires, or from complicated reactions in the atmosphere of
chemicals emitted from power plants, industries and automobiles. PM is often characterized by
size: PM10 refers to coarse particles with diameters smaller than 10 micrometers; PM2.5 are fine
particles with diameters smaller than 2.5 micrometers; and PM0.1 are ultrafine particles with
diameters less than 0.1 micrometers. PM exposure may influence DNA methylation because the
particle components can penetrate through pulmonary endothelium and travel to circulation, then
the induced reactive oxygen species (ROS) may mediate systemic oxidative stress which can
interfere with the ability of methyltransferases to interact with DNA, further leading to altered
methylation patterns [205, 210, 211].
In an elderly population, short term PM10 was associated with global DNA (LINE-1 and
Alu) and gene-specific iNOS methylation, and findings from CHS also documented that 7-day
cumulative average exposure to PM2.5 was associated with iNOS methylation in children [212-
214]. Findings from the same CHS study also suggested the joint effects of PM2.5, genetic
variants and promoter methylation in iNOS on exhaled nitric oxide (FeNO), indicating a role of
DNA methylation in acute inflammatory pathways in the airways. In animal models using mice
to mimic long-term exposure of humans to ambient particulate air pollution near steel mills and
major highways, sperm DNA hypermethylation was observed and persisted even after exposure
ceased [215]. It is likely that PM-mediated epigenetic effects involve multiple genes and
pathways. Thus, systemic and comprehensive profiling of DNA methylation would improve our
30
understanding of PM-mediated DNA methylation changes, and benefit identification of gene
targets/pathways involved in this process [215].
Although the epigenome is susceptible to alterations throughout life, the prenatal period,
which is a period of rapid cell division and epigenetic remodeling, is considered highly sensitive
to environmental factors [216]. Initial epigenetic reprogramming occurs during gametogenesis,
then the genome undergoes demethylation after fertilization and de novo genome-wide
methylation patterns are reestablished in the zygote, leading to tissue-specific methylation
patterns, which is largely maintained but can also be modified by environmental exposures
before and/or after birth [217, 218]. The fetuses are also more susceptible to environmental
toxicants than adults because of physiological immaturity. Thus, there is a growing body of
evidence studying how prenatal environmental pollutants alter epigenetic programming. In
human studies, in utero exposure to arsenic, lead, polycyclic aromatic hydrocarbon (PAH) and
bisphenol A (BPA) have all been related to altered gene methylation in the offspring [219, 220].
For the effects of prenatal tobacco smoke exposure, individual studies have related maternal
smoking during pregnancy to differential methylation both at global level and at individual CpG
loci [161, 221-223]. In a meta-analysis performed by the Pregnancy And Childhood Epigenetics
(PACE) consortium, the authors identified more than 6,000 CpG differentially methylated in
offspring in response to maternal smoking during pregnancy with persistence into later
childhood, suggesting long-term effects of these changes [224]. Although studies mostly focused
on the influence of maternal environment, recent evidence suggests that paternal factors
(nutritional, toxicological and phenotypic variation) can also affect epigenetic patterns in the
offspring [225]. With regards to the association between prenatal air pollution and methylation
patterns in newborns, the limited studies have utilized average exposure of either the individual
31
trimesters or the whole pregnancy [226, 227], but rarely conducted time series analyses of air
pollution to estimate the contribution of exposure at each month to the outcome.
Although current findings suggest the influence of environmental factors on DNA
methylation, there are still some major questions to be addressed in this filed. First of all,
although environmental exposure can alter epigenetic sates and similar alterations can be found
in patients with the disease of concern [213], few studies have directly linked the exposure-
induced changes in epigenetic profiles to the risk and development of disease. Future studies
with longitudinal evaluation of the effects of environmental exposures on epigenetic marks are
needed to determine whether exposed subjects develop epigenetic alterations over time, and
whether such alterations increase the risk of diseases. Epigenetic regulation should also be
studied in conjunction with other aspects of gene regulation such as genetic polymorphisms to
understand the contribution to altered gene expression. With the advancements of such studies,
epigenetic patterns associated with environmental exposure may serve as useful biomarkers to
investigate the long-term effects of exposure on human health. Secondly, as with most epigenetic
studies, DNA methylation profiles need to be assessed in specific cell types. Investigating
multiple cell types of interest will help to identify whether there are subsets of relatively stable or
dynamic epigenetic marks to better understand the mechanism of environmental exposure-
induced changes. Thirdly, animal studies have shown transgenerational effects of exposures on
DNA methylation [228, 229], but whether the such effects are present in humans remains to be
elucidated.

32
CHAPTER 2: EPIGENETIC REGULATION OF AXL AND CHILDHOOD ASTHMA
SYMPTOMS

Lu Gao, Joshua Millstein, Kimberly D. Siegmund, Louis Dubeau, Rachel Maguire,

Frank D.
Gilliland, Susan K. Murphy, Cathrine Hoyo, Carrie V. Breton

2.1 ABSTRACT
AXL is one of the TAM (TYRO3, AXL and MERTK) receptor tyrosine kinases and may affect
numerous immune-related health conditions. However, the role for AXL in asthma, including its
epigenetic regulation, has not been extensively studied. We investigated the association between
AXL DNA methylation at birth and risk of childhood asthma symptoms at age 6 years. DNA
methylation of multiple CpG loci across the regulatory regions of AXL was measured in newborn
bloodspots using the Illumina HumanMethylation450 array on a subset of 246 children from the
Children’s Health Study (CHS). Logistic regression models were fitted to assess the association
between asthma symptoms and DNA methylation. Findings were evaluated for replication in a
separate population of 1038 CHS subjects using Pyrosequencing on newborn bloodspot samples.
AXL genotypes were extracted from genome-wide data. Higher average methylation of CpGs in
the AXL gene at birth was associated with higher risk of parent-reported wheezing and the
association was stronger in girls than in boys. This relationship reflected the methylation status
of the gene-body region near the 5’ end, for which a 1% higher methylation level was
significantly associated with a 72% increased risk of ever having wheezed by 6 years. The
association of one CpG locus, cg00360107 was replicated using Pyrosequencing. Increased AXL
methylation was also associated with lower mRNA expression level of this gene in lung tissue
from the Cancer Genome Atlas (TCGA) dataset. Furthermore, AXL DNA methylation was
33
strongly linked to underlying genetic polymorphisms. These results indicate that AXL DNA
methylation at birth was associated with higher risk for asthma-related symptoms in early
childhood.

2.2 INTRODUCTION
Asthma is the most common chronic disease in childhood [230, 231]. It is a complex disease
determined by the interplay between genetic and environmental factors [4-7, 232]. The
pathogenesis of childhood asthma is characterized by both structural features in the airway wall
such as breach in epithelium integrity, and immunological features like airway inflammation
[233-236]. We and others have shown that early life exposure to tobacco smoke and air pollution
are associated with increased risk of childhood asthma and asthma symptoms [96, 100, 237-240].
In addition, epigenetic modifications, including DNA methylation, can alter regulation of genes
involved in airways development or immune-mediated inflammatory pathways and may play a
role in mediating the effects of environmental exposures [8, 9, 241-243]. Many studies have
now been conducted to investigate the effects of epigenetic variation on risk of asthma-related
phenotypes [110, 119]. Although alterations in DNA methylation patterns can occur throughout
life, important patterns in the methylome are established during embryogenesis and early life
[244]. However, few studies have examined the effects of DNA methylation in immune cells at
birth on asthma pathogenesis.
In previous work we identified a gene - AXL, one of the TAM (TYRO3, AXL and MERTK)
family receptor tyrosine kinases - in which methylation status of certain CpG loci varied based
on prenatal exposure to tobacco smoke [161, 163]. Given the known associations between
prenatal tobacco smoke exposure and asthma risk, as well as prenatal tobacco smoke and DNA
34
methylation in AXL, we sought to investigate whether methylation in AXL at birth was associated
with childhood asthma or asthma-related symptoms.
TAM genes are key signaling molecules in innate immune responses and may affect
numerous immune-related health conditions [164]. They act as pivotal inhibitors of immuno-
regulatory factors and prevent unrestrained signaling of inflammatory responses by these factors
[174, 199]. Growth-arrest-specific 6 (GAS6) and protein S are the ligands that bind and activate
the TAM receptors [245] and GAS6 showed higher expression in subjects with severe asthma
during exacerbation [200]. Moreover, genetic polymorphisms in TAM genes and their ligands
have also been implicated in inflammation and autoimmune diseases [201, 202]. Genetic variants
may affect DNA methylation of CpG sites in their genomic surroundings and gene expression
through altering the affinity of DNA binding factors, enhancer activity, or chromatin formation
[246, 247]. Increased CpG methylation in the promoter region may modulate expression or
silence AXL entirely and lead to overstimulation of the immune system. The identification of
epigenetic and genetic variation associated with childhood asthma symptoms may shed light on
the etiology of this complex disease and the biological role of AXL.
In this study, we investigated the association between methylation of multiple CpG sites
across the regulatory regions of AXL at birth and risk of childhood asthma symptoms, taking into
consideration the underlying genetic variation in AXL. We first assessed the association in a
subset of 246 subjects from the Children’s Health Study (CHS), then sought to replicate the
associations in a separate population of 1038 CHS subjects. Correlations between DNA
methylation and expression of AXL were also evaluated in two tissue types, cord blood from 235
subjects of the Newborn Epigenetic STudy (NEST) and lung tissue samples from the Cancer
Genome Atlas (TCGA) dataset [248-250]. To further investigate the epigenetic control of AXL
35
and address potential confounding effects from genetic polymorphisms, the association between
AXL genetic polymorphisms and methylation was also evaluated.

2.3 MATERIALS AND METHODS
Study population
This study was conducted in subsets of participants in the Children’s Health Study, a
longitudinal study of respiratory health of children in southern California [6, 146, 147, 251]. A
subset of 737 children were initially sampled to participate in a study of atherosclerosis [252] of
whom 689 could be linked to California birth records. Of these, we randomly selected 246
children from participants for whom at least 700 ng of DNA was available from a dried newborn
bloodspot. DNA methylation was assessed in the newborn bloodspots using the Infinium
HumanMethylation450 BeadChip (HM450) arrays. A separate subset of 1038 CHS subjects who
were not participants in the atherosclerosis study, but who were enriched in children with
asthma, was selected to have newborn bloodspot DNA methylation measured by
Pyrosequencing. By design, neither population had exposure to in utero tobacco smoke.
Personal, parental, socio-demographic characteristics including maternal smoking during
pregnancy, and medical history for all CHS subjects were obtained from parent-completed
questionnaires at study entry. Asthma and related symptoms at age 6 years were evaluated
through these questionnaires and included a) asthma (defined by a “yes” answer to the question
“Has a doctor ever diagnosed this child as having asthma?”); b) wheeze (defined by a “yes”
answer to the question “Has your child’s chest ever sounded wheezy or whistling?”); c) wheeze
in the previous 12 months; d) bronchitic symptoms in the previous 12 months (defined by the
36
parent’s report of a daily cough for 3 months in a row, congestion of phlegm other than when
accompanied by a cold, or bronchitis).
A subset of 235 Newborn Epigenetics STudy (NEST) subjects was evaluated for the
association between methylation at several AXL CpG loci and its mRNA level in cord blood. The
NEST is a prospective study of women and their children [253]. It was designed to identify
exposures during pregnancy and early life associated with stable epigenetic alterations in infants
that may alter chronic disease susceptibility later in life. Women were eligible if they were aged
18 years and older, pregnant and spoke English. The catchment area for Duke Maternal Fetal
Medicine prenatal care clinic largely includes three contiguous counties in central North Carolina
(NC); Durham, Orange and Wake. Women who met eligibility criteria were either consented and
interviewed in-person in consultation rooms during the visit, or were given the questionnaire to
self-administer and mail back to the study office. Smokers were preferentially enrolled to the
extent possible, identified through medical records.
DNA methylation
DNA methylation was measured in newborn bloodspots (NBS) that were obtained as part of
the routine California Newborn Screening Program from the California Department of Public
Health Genetic Disease Screening Program. The NBS were stored by the state of California at -
20 degrees Celsius. A single complete newborn bloodspot for each requested participant was
mailed to us and then stored in our lab at -80 degrees Celsius upon receipt. Laboratory personnel
performing DNA methylation analysis were blinded to study subject information. DNA was
extracted from whole blood cells using the QiaAmp DNA blood kit (Qiagen Inc, Valencia, CA)
and stored at -80 degrees Celsius. 700ng to 1 µg of genomic DNA from each sample was treated
with bisulfite using the EZ-96 DNA Methylation Kit™ (Zymo Research, Irvine, CA, USA),
37
according to the manufacturer’s recommended protocol and eluted in 18 µl. The Infinium
HM450 data was compiled for each locus and was expressed as beta (β) values. Minfi package
(version 1.16.0) in R was used to process the HM450 array data [254], applying a normal
exponential background correction to the raw intensities to reduce array-level background noise
followed by dye-bias correction [255]. We then normalized each sample’s methylation values to
the same quantiles to address sample to sample variability [256]. Seven cord blood cell sub-
populations (CD8+ T-lymphocytes, CD4+ T-lymphocytes, natural killer cells, B-lymphocytes,
monocytes, granulocytes and nucleated red blood cells) were estimated using regression
calibration approach algorithm described by Bakulski et al. [257, 258]. After preprocessing, CpG
loci containing SNPs were removed from analyses. DNA methylation was studied for a total of
12 features on the HM450 array spanning the AXL gene, identified according to their genomic
positions (Figure 2.1).
For Pyrosequencing assays, three CpG loci (cg10564498, cg12722469, and cg00360107)
were selected for replication based on results in the primary population. For NEST subjects,
genomic DNA from buffy coat specimens was extracted from umbilical cord blood using
Puregene Reagents (Qiagen, Valencia, CA). PCR primers were designed by EpigenDx Inc.
(http://www.epigendx.com) to cover the loci of interest and the specificity of the primer
sequences was confirmed using in-silico PCR. 500 ng of genomic DNA extracted from NEST
and CHS samples (randomized together) was bisulfite treated using the EZ DNA Methylation
Kit™ (Zymo Research, Irvine, CA, USA) and was purified according to the manufacturer’s
protocol. Methylation assays (assay ADS8097-FS) were performed by EpigenDx Inc. using the
PSQ96HS system (Pyrosequencing, Qiagen) according to standard procedures as described in
previous work [259, 260]. The methylation level was determined using QCpG software
38
(Pyrosequencing, Qiagen) and was reported as percent of DNA methylation at each CpG locus.
Each experiment included cytosines not part of a CpG dinucleotide as internal controls to
evaluate incomplete bisulfite conversion of the input DNA. A series of unmethylated and
methylated DNA were included as controls in each assay. Furthermore, PCR bias testing was
performed by mixing unmethylated control DNA with in vitro methylated DNA at different
ratios (0%, 5%, 10%, 25%, 50%, 75%, and 100%), followed by bisulfite modification, PCR, and
Pyrosequencing analysis.
mRNA expression in NEST
Origene’s qStar mRNA detection system (Origene, Rockville,MD) was used in the
quantification of AXL mRNA in cord blood in NEST subjects. qPCR primers for the major AXL
transcript (#HK228780) and its corresponding copy number standard (#HK201002) were
designed by qStar. All measurements of expression were conducted in duplicate in cord blood
samples from 235 participants in the NEST cohort. AXL mRNA was isolated from stored
PAXgene tubes of cord blood using the PAXgene blood miRNA isolation kit (Qiagen, Valencia,
CA). First strand cDNA conversion of mRNA was performed using Origene’s cDNA synthesis
kit (#NP100042). qPCR reactions were run with Kappa Sybr Fast qPCR kit (# KK4604;
KapaBiosystems, Boston, MA) in the ABI 7900HT thermocycler (Thermofisher). 10% repeats
were included to evaluate reproducibility.
In silico analyses in publicly available data
To assess the association between DNA methylation and gene expression in lung tissue, we
downloaded AXL methylation profiling data of 29 histologically normal tissue samples from
cases with lung adenocarcinoma (LUAD) or lung squamous cell carcinoma (LUSC) from the
TCGA dataset [248, 250]. All samples had both methylation profiling (Illumina Infinium
39
HumanMethylation450 Beadchip) and RNA-seq (Illumina HiSeq) data. The mean age was 65.9
years (SD: 12.39), and 75.9% of the subjects were male. 51.7% of the subjects were moderate to
heavy smokers.
To visualize epigenetics marks and regulatory regions of the AXL gene in relevant tissues
and cell types we used the WashU EpiGenome Browser [261, 262].
Genotyping
Buccal scrapes were collected from CHS subjects beginning in 1998 using standard
protocols [263]. A customized package including three buccal kits with instructions on buccal
cell collection was sent to each participant. Genomic DNA was isolated from buccal cells using a
Puregene
TM
DNA isolation Kit (Gentra Systems, Minneapolis, MN) and genotyping was
performed using the Illumina HumanHap550, HumanHap550-Duo or Human610-Quad
BeadChip microarrays as described previously [264]. Data was phased using SHAPEIT and
imputed using IMPUTE2 with 1000Genomes Phase 1 integrated variant v3 phased reference
(April 2012). Genotypes of single nucleotide polymorphisms (SNPs) in AXL and its surrounding
region (1kb upstream and downstream) were extracted from the CHS genome-wide genotypic
data. SNPs with minor allele frequency (MAF) less than 5% or missing in more than 5% of the
samples were removed, leaving 90 SNPs for analyses. RS numbers, minor allele frequencies, and
genomic locations of all 90 SNPs under investigation were shown in Table 2.1. 28 tagged SNPs
were identified with a pair tag r
2
>0.8 in Haploview using all available CHS samples (N=3845)
and were included in the analyses [265]. In addition, we performed principal component (PC)
analysis on the 28 AXL tagged SNPs and the top 7 PCs, which represented 80% of the total
variation cumulatively, were added as covariates in regression models to test confounding effects
from gene polymorphisms. Genotype data was available for 207 of the 246 subjects in the
40
primary population and for 728 of the 1038 subjects in the replication study. Admixture was
assessed using the program STRUCTURE from a set of ancestral informative markers that were
scaled to represent the proportion of African American, Asian, Native American and white
admixture [266].
Statistical analyses
Descriptive analyses were performed to examine the distribution of methylation and subject
characteristics. Spearman correlations of methylation between each CpG site were calculated and
shown in Table 2.2. We took the average of methylation at CpG sites in the same genomic region
to represent regional methylation status (Figure 2.1). To evaluate the association between AXL
methylation and asthma symptoms, we fitted logistic regression models for each outcome and
CpG individually, adjusted for child’s age, sex, ethnicity, city of residence at study entry, and
plate effect, while history of doctor-diagnosed asthma was additionally adjusted for wheezing
and bronchitic outcomes. Additional adjustment for genetic polymorphisms, methylation slide,
estimated cord blood cell type proportions, parental education level, allergy history, birth weight,
mode of delivery, gestational age, environmental exposures (pets, pests, cockroaches, mildew
and carpet), asthma medication use and admixture did not change the effect estimates by more
than 10% and were removed from final models. Confounding effects from each of the 28 AXL
tagged SNPs were also tested one by one and were found to be minimal (results not shown). To
estimate if the associations between AXL methylation and asthma symptoms were modified by
sex, we included an interaction term between sex and methylation in the regression models.
Wald tests were used to compute interaction p-values.
A similar logistic regression model was used in the replication population to evaluate the
association between asthma symptoms and methylation at each of the three CpG sites
41
(cg10564498, cg12722469, and cg00360107) measured by Pyrosequencing, with adjustment for
child’s age, sex, ethnicity, city of residence at study recruitment and asthma history (for
wheezing and bronchitic outcomes). Adjusting for methylation plate had no effect on results and
was not included in the final model. Effect modification by sex was also assessed.
Linear regression models were used to evaluate the associations between genetic
polymorphisms and DNA methylation at AXL CpG sites, adjusting for sex, admixture and
gestational age. All SNPs were coded additively by the number of minor alleles. Logistic
regression models were used to assess the association between AXL SNPs and asthma symptoms,
adjusting for child’s age, sex, ethnicity and admixture. We controlled the false discovery rate
(FDR) at the 0.05 level using the Benjamini-Hochberg procedure [267], accounting for multiple
tests across CpG sites and SNPs in AXL.
All tests assumed a two-sided alternative hypothesis and were conducted using the R
programming language, version 3.3.1.

2.4 RESULTS
DNA methylation of AXL and risk of asthma symptoms
Demographic characteristics of the primary and replication study populations were shown in
Table 2.3. The primary population had fewer males, more Hispanic subjects and lower parental
education level. There were more subjects having doctor-diagnosed asthma and related
symptoms in the replication population by design. Prevalence of asthma was 16% in the primary
study population and 28% in the replication population. Participants were 6 years old on average
in the primary population and 7 years old in the replication population at the time of asthma
42
symptoms assessment. Many of the 12 CpG loci were significantly correlated (Table 2.2), with
CpG sites closer in proximity showing stronger correlations.
We first investigated whether average DNA methylation in AXL was associated with
childhood asthma symptoms (Table 2.4). Average methylation of all 12 CpG sites was positively
associated with ever wheezing (OR: 1.46, 95% CI: 1.12-1.91), and the association remained
significant after adjusting for multiple testing at these genomic regions (FDR-adjusted p-value:
0.008). This was mainly driven by methylation status of the gene-body region near the 5’ end,
for which a 1% higher methylation level was significantly associated with a 72% higher risk of
ever wheezing (OR: 1.72, 95% CI: 1.30-2.28), and a 109% higher risk of wheezing in the
previous 12 months (OR: 2.09, 95% CI: 1.32-3.30). Moreover, the effects of average AXL
methylation on risk of wheezing in the previous 12 months were limited to girls (OR: 1.88, 95%
CI: 1.09-3.24) and not boys (OR=0.75, 95% CI: 0.40-1.39; pint=0.03). Increased AXL
methylation was also associated with higher risk for acute bronchitic symptoms, although effects
were not significant.
We sought to replicate results of four individual CpG loci in the primary analysis (Table 2.5
and 2.6). We chose the two loci in the gene body with the most consistent associations
(cg00360107 and cg19270050) and the two loci showing significant interactions with sex
(cg10564498 and cg12722469) (Table 2.6). These loci were evaluated in a separate population of
1038 CHS subjects using Pyrosequencing. A successful PCR primer design could not be found
for cg19270050 therefore only cg00360107, cg10564498 and cg12722469 were evaluated in the
replication population (Table 2.5). Consistent with primary results, methylation at cg00360107
was also negatively associated with asthma-related symptoms, especially the risk of ever
wheezing (Table 2.5, OR: 0.90, 95% CI: 0.82-0.99). The differences in associations by sex were
43
marginally replicated for cg10564498 (pint=0.06), but not for cg12722469 (Table 2.5). In both
populations higher cg10564498 methylation was associated with higher risk for ever wheezing
and wheezing in the previous 12 months in girls but lower risk in boys, with similar magnitudes
of sex-stratified effects.
AXL methylation and expression in cord blood and lung
Next, we sought to identify whether DNA methylation in AXL was associated with its
mRNA expression level in cord blood and in lung tissue. To do so, we evaluated the correlations
between paired data in NEST and TCGA datasets. While transcripts of AXL mRNA were
detectable in cord blood in NEST, overall expression was very low and we did not find evidence
to support a correlation. We then evaluated the correlations with gene expression for AXL
methylation using 29 histologically normal lung tissue samples based on HM450 array and RNA
sequencing data (Figure 2.2). Average methylation of the whole AXL gene as represented by 12
CpG loci, showed negative correlation with expression (r=-0.42, p-value=0.03). These data,
albeit in a population of adult males some of whom have a history of smoking, lend preliminary
support to the notion that increased methylation may lead to lower AXL expression level in the
lung, a more pathologically relevant tissue for asthma and related phenotypes than evaluation of
peripheral blood.
Genetic variants and DNA methylation of AXL
We also tested whether SNPs in AXL and the surrounding regions (1kb upstream and
downstream) were associated with average DNA methylation in the primary study population
(Table 2.7). A few tagging SNPs were significantly associated with average AXL DNA
methylation in the near-transcription start site (TSS) region and the whole gene after FDR
adjustment. We further tested if SNPs were associated with DNA methylation at individual CpG
44
sites in the replication population and found that AXL DNA methylation was strongly linked to
underlying genetic polymorphisms (Table 2.8). The associations between cg10564498
methylation and tagging SNPs were shown in Figure 2.3, suggesting that SNPs were having
stronger associations with CpG sites in closer proximity. The SNPs under investigation were
tagging SNPs, thus linkage disequilibrium (LD) was low by design (Figure 2.3). The results of
sensitivity analyses assessing the confounding effects of admixture and top 7 PCs from AXL
SNPs were shown in Table 2.9 and 2.10, respectively. None of these SNPs were confounders to
the association between AXL methylation and asthma-related symptoms (Table 2.10), or
statistically significantly associated with asthma and related symptoms in childhood (Table
2.11).
Lastly, we used the WashU EpiGenome Browser to conduct an in silico investigation of
epigenetic regulatory traits in tissues and cell types relevant to our fetal programming hypothesis
and potential involvement of immune cells. We contrasted histone marks, regulatory regions, and
transcription factor binding sites in the IMR90 fetal lung cell line, adult lung fibroblast cells and
adult CD4 naïve primary cells (Figure 2.4). The CpG loci evaluated in the AXL gene body region
are located adjacent to enhancers in both fetal and adult lung fibroblast cells, and adult CD4
naïve primary cells (light green and yellow in chromHMM tracks). The gene-body region under
investigation also contains transcription factor binding sites, indicated by peaks from ChIP-Seq
input, and marks for active transcription (green in chromHMM tracks). There are more
transcription factor binding sites and active histone marks associated with this region in fetal
lung than in adult lung and blood, suggesting the active transcription and regulation of AXL in
fetal lung. Additionally, the combination of H3K4me1 and H3K27ac, which marks active
enhancers [268, 269], was observed in the same genomic position (CpG sites cg00360107-
45
cg26521562) in both fetal and adult lung fibroblast cells, suggesting this region may act as an
enhancer throughout life. The epigenetic marks in CD4 naïve primary cells were fewer, although
the pattern of H3K4me1 in particular was generally consistent with lung cells, and suggests that
these epigenetic marks observed in blood may reflect patterns in fetal and adult lung.

2.5 DISCUSSION
Our results show that average methylation in AXL at birth was associated with higher risk
for asthma-related phenotypes in childhood, especially wheezing. The effects of average AXL
methylation on wheezing symptoms were magnified in girls compared to boys. One CpG locus,
cg00360107, which was inversely correlated with its nearest neighbors, was associated with
marginally significantly reduced wheezing risk and the result was replicated by Pyrosequencing
in a separate population of 1038 CHS subjects.
The AXL CpG region showing the strongest association with wheeze is located in a region of
the gene body harboring histone marks for active transcription and enhancers in fetal and adult
lung cells and CD4 immune cells. This region was predicted from the UCSC genome browser
[270, 271] to have binding sites for the transcription factor IRF7 (interferon regulatory factor 7),
which is involved in transcriptional activation of virus-inducible cellular genes, the
transcriptional activator ISGF3 (interferon-stimulated gene factor 3), and AP-1 (activator protein
1) that regulates gene expression in response to a variety of pathogenic stimuli [272-274].
Alterations in CpG methylation levels in this region during fetal development may modify the
transcriptional activity of AXL and the binding of transcription factors in response to stimuli,
particularly in the lung.
46
AXL has been well characterized in the pathogenesis of numerous cancers and
cardiovascular events [191, 275-279] but rarely addressed in asthma. Key elements in asthma
pathogenesis include the accumulation of polarized CD4
+
T helper (TH)2 cells and exaggeration
of pro-inflammatory TH2 cells over the infection-fighting TH1 cells in the T-cell repertoire,
accompanied by an up-regulation of the TH2 inflammatory cytokines [280]. In the key antigen-
presenting cells including dendritic cells (DCs) and macrophages, AXL and other TAM proteins
function to inhibit production of pro-inflammatory cytokines that are induced by Toll-like
receptors (TLRs), while activating the inflammation-inhibitory genes encoding the suppressor of
cytokine signaling (SOCS) 1 and 3 [174, 176]. Taken together, these concepts illuminate a
carefully regulated feedback control process that switches a pro-inflammatory signaling complex
to one that inhibits inflammation. Many of the genes inhibited or activated by AXL in this
process are involved in asthma pathogenesis.
This information suggests that AXL signaling may be associated with the suppression of
inflammatory responses and lower risk for asthma and related phenotypes. In our study, we
observed average DNA methylation at AXL was positively associated with wheezing symptoms.
The association was stronger in girls, where a 1% increase in average methylation was associated
with an 88% increase in risk of wheezing symptoms while no effect was seen in boys. Previous
research has reported sex-specific associations between DNA methylation and various health
outcomes including autoimmune diseases [281-283], although the mechanisms behind these are
unclear. The association between AXL methylation and higher risk for wheezing symptoms was
observed as early as the first few days of life, indicating that methylation status of AXL may be
reflecting epigenetic changes programmed in utero that make the child more susceptible to
symptoms in later childhood. Higher average AXL methylation was also associated with higher
47
risk for childhood bronchitic symptoms, which are suggestive of chronic symptoms that may
follow an illness or acute exacerbation of asthma, or chronic inflammation in the airway.
However, due to the small sample size in the primary population, we were not able to detect
significant associations.
Underlying genetic variants are known to influence epigenetic variation. Therefore, we
evaluated SNPs in AXL to understand whether genetic variation influenced DNA methylation
directly, to address potential confounding of observed associations between DNA methylation
and asthma and wheeze risk, and to test whether SNPs independently predicted asthma and
wheeze risk. Genome-wide studies have revealed quantitative trait loci (QTLs) for DNA
methylation, known as methylation QTLs (metQTLs) in multiple human tissues [50, 52, 55, 284-
287]. MetQTLs are usually located in intergenic or intragenic regions and affect DNA
methylation levels at nearby CpG sites [51]. In one study of metQTLs in human lung, the authors
identified 34,304 cis- and 585 trans-metQTLs, which were enriched in CTCF-binding sites,
DNaseI hypersensitivity regions and histone marks [50]. In this study, we found that average
DNA methylation in AXL was highly correlated with genetic variation in nearby sites acting in
cis.
The above evidence implies that alterations in the methylation landscape of AXL may be
attributable partially to genetic polymorphisms in nearby regions. Most of the SNPs under
investigation in this paper and the reported methylation-associated SNPs were located in gene-
body intragenic regions [51], suggesting an interaction between gene-body methylation and
proximal genetic variants. However, none of the SNPs under investigation were implicated in
asthma and related symptoms in childhood in this study and none of the SNPs acted as
48
confounders of the observed associated between DNA methylation and asthma-related
symptoms.
One of the strengths of this study is the temporal separation of DNA methylation assessment
(at birth) and respiratory health outcomes assessment (at 6-7 years of age), which enables the
investigation of fetal factors associated with asthma predisposition while overcoming the
concern for reverse causation. However, several limitations should also be noted. First, DNA
methylation of AXL was measured from newborn blood which is a mixed cell population. Since
AXL is expressed at very low levels in blood [288, 289], it may not be the ideal tissue to study
AXL gene activity. Nonetheless, methylation levels systemically altered during fetal development
ought to be reflected across multiple tissues, and therefore evaluating methylation in newborn
blood can serve as a useful biomarker of early life exposure relevant to the target tissue. Indeed,
the lack of AXL expression in cord blood but its presence in lung tissue, coupled with our in vitro
assessment of the epigenetic landscape in CD4 naïve primary cells compared to lung cells
supports this notion for AXL. Future evidence from human- or animal- based designs is
warranted to demonstrate the consistent pattern of AXL methylation across somatic tissues and in
which tissues the methylation correlates with expression. Second, characterization of asthma and
related phenotypes was based on parent-completed questionnaires, potentially introducing recall
bias or misclassification bias. Lastly, although we made every effort to control for potential
confounders, we cannot exclude the possibility of residual confounding by some unknown factor
associated with AXL DNA methylation levels and asthma-related phenotypes.
In conclusion, AXL DNA methylation at birth, which was strongly linked to underlying
genetic variation, was also associated with higher risk for asthma-related phenotypes in early
childhood. The effects on wheezing were stronger in girls than in boys.
49
2.6 TABLES AND FIGURES

Table 2.1. List of SNPs analyzed
RS number
Location
(hg19)
MAF Tag
a
Tagged SNPs eQTL tissue
b

rs2301235 41724671 0.18 Yes rs2301235, rs2301236

rs2569692 41724687 0.16 Yes rs2569692

rs2301236 41724820 0.18 No

rs28364580 41724885 0.19 Yes rs28364580

rs1654648 41725752 0.39 No

rs1709122 41725754 0.38 Yes rs1709122, rs1709121, rs1654648

rs1709121 41725790 0.41 No

rs4803446 41726167 0.27 Yes rs4803446

rs2271546 41727197 0.10 Yes rs2271546, rs10409443

rs10409443 41728539 0.11 No

rs186235601 41728703 0.06 Yes rs186235601

rs10409940 41728765 0.39 Yes rs10409940

rs11083613 41729505 0.22 Yes rs11083613

rs4803447 41731175 0.09 Yes
rs4803447, rs73043273, rs8109440,
rs12459929
rs59423102 41731749 0.10 Yes rs59423102

rs12462203 41732423 0.40 Yes
rs12462203, rs6508974,
rs12978098, rs12980248,
rs4802111, rs4803448, rs7246525,
rs4802113, rs12979764,
rs12459223, rs67033402

rs12459929 41732610 0.09 No

rs73043273 41732688 0.09 No

rs12984621 41732727 0.30 Yes rs12984621

rs6508974 41733145 0.39 No

rs8109440 41733355 0.09 No

rs4802111 41734059 0.40 No

rs4802112 41734490 0.28 Yes
rs12973055, rs4802112, rs4803451,
rs4802115
Tibial artery,
subcutaneous adipose
rs4803448 41734560 0.41 No

rs4803449 41734666 0.33 Yes rs4803449
Tibial artery,
subcutaneous adipose
rs12978098 41734845 0.41 No

rs12979764 41734958 0.41 No

rs12980248 41735115 0.41 No

rs79855742 41737344 0.08 No

rs76249126 41737410 0.08 Yes
rs73931459, rs79855742,
rs76249126, rs77033536
rs75955910 41737414 0.07 Yes rs75955910

rs12973055 41737851 0.26 No

rs12973061 41737854 0.34 No

50
rs7246525 41737996 0.41 No

rs7246896 41738212 0.07 Yes rs7246896, rs73043294

rs12459223 41738533 0.40 No

rs73043294 41738716 0.07 No

rs66841352 41739180 0.33 No

rs67033402 41739254 0.41 No

rs74816457 41739534 0.33 No

rs77287588 41739574 0.16 Yes rs77287588

rs73931459 41740639 0.08 No

rs4802113 41740895 0.41 No

rs4802114 41741278 0.33 Yes rs4802114

rs7256873 41742308 0.17 No

rs4637024 41743454 0.12 Yes rs4637024

rs3786556 41744831 0.17 No

rs2304235 41745414 0.17 No

rs12982872 41746535 0.17 No

rs3786555 41748153 0.21 Yes rs3786555
Tibial artery,
transformed fibroblast
cells
rs2304234 41748753 0.37 Yes
rs74816457, rs200156351,
rs116039364, rs2304234,
rs71337582, rs12973061,
rs66841352, rs201282430,
rs201384697, rs114806364
Tibial artery
rs55841050 41750550 0.08 Yes
rs12459996, rs55841050,
rs137991784, rs73045231,
rs139833223, rs17251589,
rs73045226, rs12461203,
rs55994820, rs73045223,
rs144025261

rs137991784 41751520 0.08 No

rs12972779 41752127 0.17 No

rs4803451 41752635 0.28 No

Tibial artery,
subcutaneous adipose,
esophagus, lung
rs201384697 41752731 0.36 No

rs200156351 41752732 0.32 No

rs201282430 41752733 0.32 No

rs71337582 41752736 0.32 No

rs116039364 41752738 0.32 No

rs114806364 41752739 0.32 No

51
rs12983027 41753634 0.16 Yes
rs12977563, rs35310790,
rs12972779, rs3786556,
rs11880729, rs11879429,
rs2304232, rs11879435,
rs2304235,rs4591267, rs12983027,
rs1946613, rs7256873, rs12982872,
rs2304231

rs4802115 41753782 0.28 No

Tibial artery,
subcutaneous adipose,
esophagus
rs11882467 41754217 0.27 No

Tibial artery,
subcutaneous adipose,
esophagus, lung
rs11880729 41755238 0.14 No

rs34772093 41755357 0.06 No

rs139833223 41755390 0.08 No

rs12978323 41756038 0.26 Yes
rs1946612, rs12980267, rs7250883,
rs11882467, rs12978323
Tibial artery, lung,
skeletal muscle,
esophagus,
subcutaneous adipose
rs17251589 41756085 0.08 No

rs12977563 41756503 0.14 No

rs144025261 41756563 0.08 No

rs12459996 41756906 0.08 No

rs11879429 41757444 0.14 No

rs11879435 41757603 0.14 No

rs55994820 41757707 0.08 No

rs35310790 41758047 0.14 No

rs73045223 41759474 0.08 No

rs116056574 41759637 0.36 Yes rs116056574

rs7250883 41760033 0.22 No

Tibial artery
rs73045226 41760913 0.08 No

rs77033536 41762000 0.07 No

rs2304232 41762525 0.14 No

rs2304231 41762670 0.14 No

rs12461203 41764568 0.08 No

rs35546772 41764758 0.06 Yes rs35546772, rs34772093

rs12980267 41765229 0.22 No

Tibial artery
rs4591267 41765407 0.14 No

rs73045231 41766978 0.08 No

rs1946613 41767785 0.14 No

rs1946612 41767987 0.22 No Tibial artery
a
Defined with a pair tag r
2
>0.8 in Haploview with all CHS samples (N=3845).
b
Data Source: GTEx Analysis Release V6p (dbGaP Accession phs000424.v6.p1)
52
Table 2.2. Spearman correlation between methylation at each AXL CpG site in the primary population (N=246)
Region CpG
cg1056
4498
cg0324
7049
cg1272
2469
cg0237
2201
cg1984
8291
cg1489
2768
cg2757
9501
cg0036
0107
cg1927
0050
cg2490
1063
cg2652
1562
cg2096
4856
Promoter
cg10564498 1.00 0.57* 0.55* 0.30* 0.15* 0.12 -0.01 -0.14* -0.01 0.01 -0.08 -0.17*
cg03247049 1.00 0.51* 0.30* 0.26* 0.11 -0.01 -0.11 -0.02 -0.01 -0.10 -0.16*
cg12722469 1.00 0.43* 0.24* 0.32* 0.08 0.01 -0.08 0.05 0.09 -0.07
cg02372201 1.00 0.15* 0.37* 0.17* 0.02 -0.18* -0.01 0.01 0.04
cg19848291 1.00 0.31* 0.16* -0.13* 0.27* 0.11 -0.35* -0.35*
cg14892768 1.00 0.11 0.03 -0.15* -0.06 0.04 0.04
Gene-
body
cg27579501 1.00 -0.14* 0.01 0.09 -0.15* -0.18*
cg00360107 1.00 -0.14* -0.14* 0.13* 0.16*
cg19270050 1.00 0.10 -0.30* -0.26*
cg24901063 1.00 -0.19* -0.15*
cg26521562 1.00 0.37*
3' UTR cg20964856 1.00
* p-value < 0.05
Definition of abbreviations: UTR=untranslated region
53

Table 2.3. Demographic characteristics of participants

Primary study
population (N=246)

Replication
population
(N=1038)
P-value
a

Male sex, n (%) 98 (39.8)

541 (52.1) 0.0005
Ethnicity, n (%)

0.05
Hispanic 147 (59.8)

531 (51.3)

Non-Hispanic White 72 (29.3)

381 (36.8)

Asian/Black/Other 27 (11.0)

124 (12.0)

Ever MD-diagnosed asthma, n (%) 39 (15.9)

295 (28.4) <0.0001
Ever wheezing, n (%) 66 (26.8)

455 (43.8) <0.0001
Wheezing in the previous 12
months, n (%)
37 (15.0)

250 (24.1) 0.001
Bronchitic symptoms in the previous
12 months, n (%)
40 (16.3)

229 (22.1) 0.04
Parental Education, n (%)

0.002
High school or less 89 (36.6)

316 (31.3)

Some college 76 (31.3)

438 (43.5)

Finished college/some graduate
school
78 (32.1)

254 (25.2)

Age years, mean (sd) 6.4 (0.6)

7.2 (1.3) <0.0001
Gestational age days, mean (sd) 277.5 (11.0) 272.7 (11.3) <0.0001
a
Derived from a Pearson's Chi-squared test for categorical variables and from an unequal
variance 2-sample t-test for continuous variables.

54
Table 2.4. Association between average DNA methylation levels at AXL CpG
sites and risk of asthma and related symptoms in childhood in the primary
study population (N=246)
a

Average of
near-TSS
CpG sites
b

Average of
gene-body
CpG sites
c

Average of
all 12 CpG
sites
d

OR P OR P OR P
Ever MD-diagnosed asthma

Overall 1.10 0.16 1.05 0.71 1.17 0.19
By sex

Boys 1.17 0.10 1.12 0.58 1.34 0.09
Girls 1.03 0.79 1.00 1.00 1.02 0.90
Interaction p-value 0.33 0.69 0.28
Ever wheezing

Overall 1.10 0.19 1.72 0.0002 1.46 0.005
By sex

Boys 1.06 0.60 2.70 0.001 1.42 0.12
Girls 1.12 0.20 1.51 0.01 1.48 0.02
Interaction p-value 0.69 0.08 0.88
Wheezing in the previous 12
months

Overall 0.97 0.79 2.09 0.002 1.26 0.25
By sex

Boys 0.73 0.10 1.85 0.11 0.75 0.35
Girls 1.16 0.31 2.20 0.005 1.88 0.02
Interaction p-value 0.05 0.70 0.03
Bronchitic symptoms in the
previous 12 months

Overall 0.99 0.90 1.23 0.14 1.04 0.74
By sex

Boys 0.94 0.51 1.18 0.46 0.92 0.64
Girls 1.05 0.63 1.26 0.19 1.18 0.35
Interaction p-value 0.41 0.82 0.32
Definition of abbreviations: TSS = transcription start site
55
Odds ratios are presented for an increase in 1% of DNA methylation level at
birth. For all comparisons the reference group is children not having the
corresponding outcome. Significant p-values (<0.05) are marked in bold.
a
Adjusted for child's age, sex, ethnicity, methylation plate and city of
residence at study recruitment; additionally adjusted for ever had MD-
diagnosed asthma for wheezing and bronchitic outcomes
b
Average of cg10564498, cg03247049, cg12722469, cg02372201,
cg19848291 and cg14892768
c
Average of cg27579501, cg00360107, cg19270050, cg24901063 and
cg26521562
d
Average of all 12 CpG sites

56
Table 2.5. Association between DNA methylation levels at selected AXL CpG sites and risk of asthma and
related symptoms in childhood in the primary study population (N=246)
a
and replication population (N=1038)
b

cg10564498 cg12722469 cg00360107

Primary
study
population
Replication
population
Primary
study
population
Replication
population
Primary
study
population
Replication
population
Distance to TSS (bp) -455 -455 -55 -55 6826 6826
Mean methylation (%) 26.71 19.60 17.05 11.32 6.68 4.35
OR P OR P OR P OR P OR P OR P
Ever MD-diagnosed
asthma
Overall 1.02 0.28 1.00 0.86 1.05 0.24 0.99 0.72 0.73 0.04 0.99 0.86
By sex

Boys 1.04 0.11 1.01 0.60 1.08 0.24 0.99 0.73 0.72 0.16 1.02 0.76
Girls 0.99 0.80 0.98 0.39 1.03 0.59 0.99 0.87 0.73 0.13 0.96 0.54
Interaction p-value 0.24 0.32 0.59 0.92 0.94 0.51
Ever wheezing

Overall 1.03 0.21 0.97 0.13 1.07 0.09 0.95 0.05 0.78 0.07 0.90 0.04
By sex

Boys 0.98 0.59 0.92 0.006 1.07 0.38 0.92 0.01 1.05 0.87 0.92 0.20
Girls 1.07 0.04 1.02 0.44 1.08 0.14 0.99 0.86 0.69 0.04 0.89 0.10
Interaction p-value 0.09 0.01 0.94 0.10 0.20 0.73
Wheezing in the
previous 12 months
Overall 1.02 0.60 0.96 0.03 0.99 0.93 0.93 0.007 0.55 0.04 0.95 0.34
By sex

Boys 0.92 0.13 0.92 0.004 0.82 0.10 0.92 0.02 0.74 0.37 0.90 0.12
Girls 1.15 0.02 1.00 0.94 1.11 0.22 0.95 0.16 0.40 0.03 1.03 0.68
Interaction p-value 0.009 0.06 0.04 0.58 0.23 0.19
Bronchitic symptoms in
the previous 12 months
Overall 0.99 0.66 0.95 0.02 1.06 0.19 0.95 0.05 0.75 0.07 0.93 0.15
By sex

Boys 0.96 0.24 0.95 0.04 1.02 0.74 0.94 0.07 0.80 0.31 0.93 0.24
Girls 1.03 0.46 0.97 0.25 1.09 0.14 0.97 0.37 0.70 0.12 0.93 0.40
Interaction p-value 0.18 0.60 0.47 0.60 0.67 0.96
57
Definition of abbreviations: TSS = transcription start site
Odds ratios are presented for an increase in 1% of DNA methylation level at birth. For all comparisons the
reference group is children not having the corresponding outcome. Significant p-values (<0.05) are marked in
bold.
a
Adjusted for child's age, sex, ethnicity, methylation plate and city of residence at study recruitment;
additionally adjusted for ever had MD-diagnosed asthma for wheezing and bronchitic outcomes
b
Adjusted for child's age, sex, ethnicity and city of residence at study recruitment; additionally adjusted for ever
had MD-diagnosed asthma for wheezing and bronchitic outcomes
58
Table 2.6. Association between DNA methylation levels at AXL CpG sites and risk of asthma and related symptoms in childhood in the primary study population (N=246)
a

cg10564498 cg03247049 cg12722469 cg02372201 cg19848291 cg14892768 cg27579501 cg00360107 cg19270050 cg24901063 cg26521562 cg20964856
Distance to TSS (bp)
-455 -210 -55 44 94 224 4549 6826 7015 7113 7360 42561
Mean methylation
(%) 26.71 21.44 17.05 10.13 13.55 45.11 89.36 6.68 11.96 9.61 13.16 65.00
OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P
Ever MD-diagnosed
asthma

Overall 1.02 0.28 1.10 0.06 1.05 0.24 0.92 0.65 1.24 0.04 0.99 0.81 0.88 0.21 0.73 0.04 1.05 0.18 0.92 0.50 1.03 0.59 0.89 0.19
By sex

Boys 1.04 0.11 1.10 0.20 1.08 0.24 1.16 0.55 1.35 0.05 0.99 0.91 0.95 0.73 0.72 0.16 1.04 0.44 0.81 0.28 1.05 0.38 0.91 0.39
Girls 0.99 0.80 1.11 0.16 1.03 0.59 0.75 0.26 1.13 0.38 0.99 0.83 0.83 0.14 0.73 0.13 1.05 0.24 1.00 0.99 0.99 0.92 0.88 0.25
Interaction p-value 0.24 0.94 0.59 0.22 0.37 0.95 0.48 0.94 0.84 0.38 0.46 0.84
Ever wheezing

Overall 1.03 0.21 1.04 0.44 1.07 0.09 0.98 0.88 1.16 0.15 0.98 0.58 1.11 0.27 0.78 0.07 1.10 0.005* 1.13 0.18 1.13 0.007* 1.00 1.00
By sex

Boys 0.98 0.59 1.10 0.28 1.07 0.38 1.33 0.22 1.30 0.15 1.02 0.83 1.38 0.07 1.05 0.87 1.07 0.26 1.20 0.38 1.19 0.008 0.88 0.32
Girls 1.07 0.04 1.01 0.87 1.08 0.14 0.78 0.25 1.10 0.42 0.96 0.44 1.03 0.77 0.69 0.04 1.11 0.007 1.11 0.27 1.09 0.10 1.06 0.54
Interaction p-value 0.09 0.43 0.94 0.08 0.43 0.56 0.15 0.20 0.63 0.75 0.23 0.20
Wheezing in the
previous 12 months

Overall 1.02 0.60 0.96 0.61 0.99 0.93 0.48 0.04 0.98 0.89 0.93 0.31 1.02 0.87 0.55 0.04 1.18 0.003* 1.20 0.14 1.13 0.11 1.14 0.32
By sex

Boys 0.92 0.13 1.03 0.84 0.82 0.10 0.26 0.04 1.01 0.98 0.80 0.10 0.98 0.92 0.74 0.37 1.35 0.002* 0.95 0.87 0.98 0.81 1.04 0.83
Girls 1.15 0.02 0.93 0.43 1.11 0.22 0.57 0.15 0.95 0.83 0.99 0.88 1.05 0.78 0.40 0.03 1.11 0.10 1.27 0.09 1.25 0.009 1.23 0.22
Interaction p-value 0.009 0.54 0.04 0.23 0.86 0.18 0.79 0.23 0.08 0.38 0.04 0.46
59
Bronchitic symptoms
in the previous 12
months

Overall 0.99 0.66 0.98 0.60 1.06 0.19 0.80 0.22 1.13 0.25 0.97 0.47 0.97 0.77 0.75 0.07 1.10 0.007 1.06 0.54 0.98 0.72 0.92 0.32
By sex

Boys 0.96 0.24 0.99 0.94 1.02 0.74 0.72 0.27 1.24 0.14 0.94 0.34 0.96 0.80 0.80 0.31 1.16 0.02 1.06 0.73 0.92 0.23 0.91 0.40
Girls 1.03 0.46 0.96 0.52 1.09 0.14 0.84 0.46 1.03 0.85 0.99 0.92 0.98 0.86 0.70 0.12 1.07 0.12 1.06 0.61 1.04 0.47 0.92 0.49
Interaction p-value 0.18 0.70 0.47 0.67 0.31 0.53 0.93 0.67 0.30 0.99 0.13 0.92
Definition of abbreviations: TSS = transcription start site
Odds ratios are presented for an increase in 1% of DNA methylation level at birth. For all comparisons the reference group is children not having the corresponding outcome. Significant raw p-values (<0.05) are
marked in bold. FDR was used to adjust for all tests performed at 12 CpG sites for each outcome. * indicates significant p-values (<0.05) after FDR adjustment.
a
Adjusted for child's age, sex, ethnicity, methylation plate and city of residence at study recruitment; additionally adjusted for ever had MD-diagnosed asthma for wheezing and bronchitic outcomes
b
Average of cg10564498, cg03247049, cg12722469, cg02372201, cg19848291 and cg14892768
c
Average of cg27579501, cg00360107, cg19270050, cg24901063 and cg26521562
d
Average of all 12 CpG sites
60
Table 2.7. Association between DNA methylation levels at multiple CpG sites and gene polymorphisms
in AXL in the primary study population (N=207)
a

Average of near-TSS CpG
sites
b

Average of gene-body
CpG sites
c

Average of all 12 CpG sites
d

RS Number Location  P
Adjusted
P
 P
Adjusted
P
 P
Adjusted
P
rs2301235 41724671 0.60 0.14 0.29 0.09 0.64 0.79 0.35 0.11 0.26
rs2569692 41724687 1.55 2.9E-04 0.01 0.24 0.24 0.44 0.84 2.1E-04 0.01
rs28364580 41724885 1.16 0.005 0.05 0.34 0.07 0.22 0.67 0.002 0.03
rs1709122 41725754 1.21 4.2E-04 0.01 0.05 0.76 0.87 0.59 0.001 0.03
rs4803446 41726167 0.66 0.09 0.25 0.22 0.23 0.44 0.41 0.05 0.19
rs2271546 41727197 0.87 0.12 0.28 0.46 0.08 0.24 0.63 0.03 0.16
rs186235601 41728703 1.34 0.05 0.22 0.52 0.11 0.26 0.89 0.02 0.10
rs10409940 41728765 0.78 0.03 0.16 -0.06 0.73 0.86 0.33 0.09 0.25
rs11083613 41729505 0.03 0.93 0.95 -0.33 0.06 0.22 -0.18 0.38 0.56
rs4803447 41731175 1.75 0.009 0.08 0.52 0.10 0.25 1.01 0.004 0.05
rs59423102 41731749 -0.13 0.81 0.89 0.22 0.38 0.56 0.07 0.81 0.89
rs12462203 41732423 0.18 0.59 0.75 -0.05 0.74 0.87 -0.02 0.91 0.95
rs12984621 41732727 -0.15 0.68 0.82 -0.14 0.41 0.60 -0.23 0.25 0.44
rs4802112 41734490 0.89 0.02 0.11 0.04 0.84 0.92 0.34 0.09 0.25
rs4803449 41734666 1.02 0.005 0.05 0.17 0.31 0.50 0.47 0.01 0.10
rs76249126 41737410 -1.22 0.06 0.22 -0.28 0.36 0.56 -0.71 0.04 0.18
rs75955910 41737414 -0.65 0.32 0.50 -0.52 0.08 0.24 -0.54 0.11 0.26
rs7246896 41738212 0.11 0.86 0.92 0.22 0.42 0.60 0.18 0.57 0.74
rs77287588 41739574 -0.48 0.29 0.49 -0.22 0.29 0.49 -0.27 0.27 0.46
rs4802114 41741278 0.36 0.32 0.50 0.07 0.66 0.81 0.14 0.47 0.64
rs4637024 41743454 -0.76 0.13 0.29 0.02 0.95 0.96 -0.36 0.17 0.34
rs3786555 41748153 0.70 0.10 0.25 -0.23 0.24 0.44 0.13 0.55 0.72
rs2304234 41748753 0.51 0.17 0.34 0.00 0.99 0.99 0.14 0.48 0.64
rs55841050 41750550 1.43 0.04 0.19 0.80 0.01 0.10 0.94 0.01 0.10
rs12983027 41753634 0.79 0.07 0.22 -0.18 0.36 0.56 0.18 0.43 0.61
rs12978323 41756038 0.89 0.03 0.15 0.09 0.63 0.79 0.33 0.13 0.28
rs116056574 41759637 0.10 0.77 0.88 0.12 0.45 0.63 0.02 0.92 0.95
rs35546772 41764758 1.12 0.07 0.22 0.05 0.86 0.92 0.43 0.18 0.34
Definition of abbreviations: TSS = transcription start site
61

a
SNP data was only available for a subset of subjects. SNPs were modeled as ordinal variables (0=major allele,
1=heterozygote, and 2=minor allele) and models were adjusted for child's sex, admixture and gestational age. Beta
values are showing the percent change in methylation per one unit increase in SNP. Tagging SNPs were defined
with a pair tag r
2
>0.8 in Haploview with all CHS samples (N=3845). FDR was used to adjust for all tests performed
at the 3 methylation averages (28 ☓ 3 tests). Significant FDR-adjusted p-values (<0.05) are marked in bold.
b
Average of cg10564498, cg03247049, cg12722469, cg02372201, cg19848291 and cg14892768
c
Average of cg27579501, cg00360107, cg19270050, cg24901063 and cg26521562
d
Average of all 12 CpG sites
62
Table 2.8. Association between DNA methylation levels at selected CpG sites and gene polymorphisms in
AXL in the replication population (N=728)
a

cg10564498
(Location: 41724653)
cg12722469
(Location: 41725053)
cg00360107
(Location: 41731934)
RS Number Location  P
Adjusted
P
 P
Adjusted
P
 P
Adjusted
P
rs2301235 41724671 0.94 0.002 0.004 0.39 0.11 0.15 -0.07 0.57 0.63
rs2569692 41724687 1.75 1.2E-07 7.7E-07 1.83 8.8E-12 3.7E-10 0.37 0.004 0.007
rs28364580 41724885 1.57 5.6E-07 2.9E-06 1.66 8.2E-11 2.3E-09 0.31 0.01 0.02
rs1709122 41725754 1.80 4.2E-12 3.5E-10 1.31 7.9E-10 1.7E-08 0.19 0.06 0.08
rs4803446 41726167 1.43 2.0E-07 1.2E-06 1.02 6.4E-06 2.5E-05 0.21 0.05 0.07
rs2271546 41727197 2.21 2.8E-08 2.3E-07 1.93 2.5E-09 4.1E-08 0.33 0.03 0.05
rs186235601 41728703 1.91 5.0E-04 0.001 2.13 1.9E-06 8.5E-06 0.57 0.007 0.01
rs10409940 41728765 1.46 2.1E-08 2.0E-07 1.18 3.0E-08 2.3E-07 0.18 0.07 0.10
rs11083613 41729505 0.10 0.74 0.77 0.06 0.79 0.79 0.08 0.46 0.53
rs4803447 41731175 0.57 0.22 0.27 1.04 0.006 0.01 0.38 0.03 0.05
rs59423102 41731749 0.57 0.16 0.20 0.58 0.08 0.10 0.06 0.70 0.75
rs12462203 41732423 0.77 0.002 0.004 0.81 5.1E-05 1.7E-04 0.47 5.1E-07 2.8E-06
rs12984621 41732727 0.55 0.03 0.05 0.58 0.006 0.01 0.36 1.9E-04 5.6E-04
rs4802112 41734490 0.87 9.8E-04 0.002 0.93 1.5E-05 5.6E-05 0.56 2.0E-08 2.0E-07
rs4803449 41734666 1.08 2.4E-05 8.4E-05 1.11 9.3E-08 6.5E-07 0.54 1.7E-08 2.0E-07
rs76249126 41737410 -0.93 0.03 0.05 -0.49 0.17 0.21 -0.08 0.62 0.67
rs75955910 41737414 -1.42 0.003 0.005 -1.34 5.7E-04 0.001 -0.07 0.72 0.76
rs7246896 41738212 0.79 0.10 0.13 0.22 0.57 0.63 -0.05 0.78 0.79
rs77287588 41739574 -0.46 0.17 0.21 -0.84 0.002 0.005 0.04 0.73 0.77
rs4802114 41741278 0.47 0.08 0.10 0.68 0.001 0.003 0.33 8.9E-04 0.002
rs4637024 41743454 -0.32 0.39 0.45 -0.04 0.90 0.90 -0.09 0.54 0.61
rs3786555 41748153 0.85 0.003 0.005 0.79 6.9E-04 0.002 0.50 3.2E-06 1.4E-05
rs2304234 41748753 0.37 0.13 0.16 0.66 9.6E-04 0.002 0.46 7.2E-07 3.6E-06
rs55841050 41750550 0.58 0.23 0.28 1.08 0.007 0.01 0.57 0.002 0.005
rs12983027 41753634 1.03 0.001 0.002 1.12 1.3E-05 5.0E-05 0.48 5.2E-05 1.7E-04
rs12978323 41756038 0.99 2.1E-04 5.8E-04 1.06 1.2E-06 5.6E-06 0.59 6.4E-09 8.9E-08
rs116056574 41759637 0.77 0.003 0.005 0.79 1.3E-04 3.9E-04 0.35 3.5E-04 9.2E-04
rs35546772 41764758 1.31 0.006 0.01 1.45 2.1E-04 5.8E-04 0.69 1.4E-04 4.2E-04
a
SNP data was only available for a subset of subjects. SNPs were modeled as ordinal variables (0=major allele,
1=heterozygote, and 2=minor allele) and models were adjusted for child's sex, admixture and gestational age. Beta values are
showing the percent changes in methylation at each CpG site per one unit increase in SNP. Tagging SNPs were defined with
a pair tag r
2
>0.8 in Haploview with all CHS samples (N=3845). FDR was used to adjust for all tests performed at the 3 CpG
sites (28 ☓ 3 tests) in each study population. Significant FDR-adjusted p-values (<0.05) are marked in bold.
63

Table 2.9. Sensitivity analysis for adding admixture in testing the association between AXL DNA methylation and risk of childhood asthma and related symptoms in the primary population
(N=231)

cg10564498 cg03247049 cg12722469 cg02372201 cg19848291 cg14892768 cg27579501 cg00360107 cg19270050 cg24901063 cg26521562 cg20964856
Average of
near-TSS
CpG sites
c

Average of
gene body
CpG sites
d

Average of
all 12 CpG
sites
e

Distance to TSS
(bp)
-455 -210 -55 44 94 224 4549 6826 7015 7113 7360 42561

OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P
Ever MD-diagnosed
asthma
Model 1
a
1.02 0.37 1.07 0.18 1.05 0.33 0.92 0.67 1.24 0.05 1.01 0.83 0.84 0.12 0.76 0.09 1.04 0.28 0.98 0.88 1.04 0.49 0.88 0.16 1.09 0.22 1.06 0.68 1.16 0.25
Model 2
b
1.02 0.32 1.07 0.19 1.05 0.30 0.92 0.68 1.24 0.05 1.01 0.80 0.85 0.13 0.76 0.09 1.04 0.28 0.97 0.84 1.04 0.46 0.88 0.17 1.09 0.20 1.06 0.67 1.17 0.22
Ever wheezing

Model 1
a
1.03 0.17 1.02 0.74 1.06 0.22 0.93 0.70 1.12 0.28 0.97 0.45 1.10 0.32 0.78 0.08 1.11 0.003 1.14 0.16 1.11 0.02 0.97 0.69 1.08 0.29 1.75 0.0002 1.42 0.01
Model 2
b
1.02 0.32 1.01 0.85 1.07 0.13 0.99 0.96 1.12 0.30 0.97 0.53 1.11 0.30 0.83 0.23 1.13 0.002 1.10 0.32 1.12 0.02 0.97 0.70 1.07 0.35 1.83 0.0002 1.41 0.02
Wheezing in the
previous 12 months
Model 1
a
1.02 0.61 0.96 0.52 0.99 0.86 0.47 0.04 0.98 0.90 0.92 0.31 1.01 0.94 0.56 0.04 1.17 0.005 1.20 0.15 1.12 0.11 1.13 0.38 0.97 0.75 1.99 0.003 1.23 0.30
Model 2
b
1.03 0.37 0.98 0.75 1.01 0.90 0.48 0.05 0.98 0.93 0.91 0.23 1.01 0.92 0.58 0.07 1.17 0.005 1.19 0.20 1.12 0.12 1.13 0.39 1.00 0.99 2.01 0.003 1.34 0.16
Bronchitic
symptoms in the
previous 12 months

Model 1
a
0.99 0.60 0.97 0.55 1.06 0.15 0.78 0.18 1.15 0.19 0.98 0.73 0.98 0.83 0.77 0.10 1.10 0.008 1.06 0.55 0.98 0.62 0.94 0.46 1.00 0.96 1.22 0.15 1.06 0.67
Model 2
b
0.99 0.58 0.96 0.45 1.06 0.16 0.76 0.14 1.13 0.27 0.98 0.62 0.98 0.82 0.74 0.08 1.10 0.008 1.04 0.67 0.98 0.68 0.92 0.35 0.99 0.86 1.22 0.15 1.04 0.78
Definition of abbreviations: TSS = transcription start site
Ancestry data was not available for all subjects. Odds ratios are presented for an increase in 1% of DNA methylation level at birth. For all comparisons the reference group is children not having the corresponding outcome. Significant raw p-values
(<0.05) are marked in bold.
a
Adjusted for child's age, sex, ethnicity, methylation plate and city of residence at study recruitment; additionally adjusted for ever had MD-diagnosed asthma for wheezing and BCP outcomes
b
Adjusted for child's age, sex, ethnicity, methylation plate, city of residence at study recruitment and admixture; additionally adjusted for ever had MD-diagnosed asthma for wheezing and BCP outcomes
c
Average of cg10564498, cg03247049, cg12722469, cg02372201, cg19848291 and cg14892768
d
Average of cg27579501, cg00360107, cg19270050, cg24901063 and cg26521562
e
Average of all 12 CpG sites
64
Table 2.10. Sensitivity analysis for adding the top 7 principal components (PCs) of AXL SNPs in testing the association between AXL DNA methylation and risk of childhood asthma and
related symptoms in the primary population (N=165)

cg10564498 cg03247049 cg12722469 cg02372201 cg19848291 cg14892768 cg27579501 cg00360107 cg19270050 cg24901063 cg26521562 cg20964856
Average of
near-TSS
CpG sites
c

Average of
gene body
CpG sites
d

Average of
all 12 CpG
sites
e

Distance to TSS
(bp)
-455 -210 -55 44 94 224 4549 6826 7015 7113 7360 42561

OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P OR P
Ever MD-
diagnosed asthma
Model 1
a
1.01 0.61 1.09 0.14 1.02 0.65 0.85 0.44 1.22 0.11 1.02 0.66 0.83 0.13 0.75 0.11 1.04 0.28 0.98 0.89 1.01 0.87 0.83 0.08 1.07 0.36 1.03 0.84 1.11 0.45
Model 2
b
1.01 0.66 1.08 0.22 1.01 0.89 0.87 0.52 1.21 0.16 1.02 0.73 0.81 0.11 0.64 0.04 1.05 0.22 0.97 0.82 1.03 0.69 0.83 0.14 1.05 0.49 1.05 0.78 1.10 0.53
Ever wheezing

Model 1
a
0.98 0.63 1.04 0.62 1.00 1.00 0.76 0.25 1.10 0.47 0.94 0.28 1.28 0.08 0.71 0.11 1.17 0.001 0.90 0.49 1.13 0.05 0.86 0.16 0.96 0.68 2.03 0.0006 1.19 0.30
Model 2
b
0.98 0.54 1.02 0.82 0.98 0.78 0.76 0.27 1.07 0.61 0.94 0.27 1.31 0.08 0.68 0.09 1.17 0.002 0.88 0.40 1.15 0.05 0.83 0.15 0.94 0.52 2.09 0.0008 1.14 0.46
Wheezing in the
previous 12 months
Model 1
a
0.94 0.26 0.89 0.21 0.95 0.52 0.36 0.07 1.09 0.73 1.01 0.94 1.23 0.37 0.48 0.10 1.56 0.007 0.70 0.23 1.12 0.30 0.88 0.55 0.86 0.29 3.20 0.007 1.01 0.97
Model 2
b
0.92 0.25 0.90 0.41 0.99 0.91 0.23 0.10 1.14 0.67 1.07 0.63 1.54 0.18 0.46 0.13 6.03 0.20 0.57 0.14 1.12 0.33 0.70 0.25 0.84 0.43 3.75 0.01 1.25 0.53
Bronchitic
symptoms in the
previous 12 months

Model 1
a
0.97 0.35 0.94 0.32 1.02 0.76 0.56 0.03 1.10 0.50 1.02 0.76 0.98 0.88 0.84 0.40 1.11 0.02 0.83 0.26 0.98 0.74 0.83 0.10 0.95 0.55 1.21 0.27 0.94 0.73
Model 2
b
0.95 0.19 0.88 0.10 0.99 0.87 0.49 0.02 1.08 0.60 1.01 0.84 0.98 0.90 0.76 0.23 1.11 0.04 0.78 0.15 0.98 0.72 0.83 0.12 0.90 0.27 1.16 0.41 0.84 0.36
Definition of abbreviations: PC = principal component; SNP = single nucleotide polymorphism; TSS = transcription start site
PC data was not available for all subjects. Odds ratios are presented for an increase in 1% of DNA methylation level at birth. For all comparisons the reference group is children not having the corresponding outcome. Significant raw p-values (<0.05) are
marked in bold.
a
Adjusted for child's age, sex, ethnicity, methylation plate and city of residence at study recruitment; additionally adjusted for ever had MD-diagnosed asthma for wheezing and BCP outcomes
b
Adjusted for child's age, sex, ethnicity, methylation plate, city of residence at study recruitment and the top 7 PCs of AXL SNPs; additionally adjusted for ever had MD-diagnosed asthma for wheezing and BCP outcomes
c
Average of cg10564498, cg03247049, cg12722469, cg02372201, cg19848291 and cg14892768
d
Average of cg27579501, cg00360107, cg19270050, cg24901063 and cg26521562
e
Average of all 12 CpG sites
65
Table 2.11. Association between gene polymorphisms in AXL and risk of asthma and related symptoms in childhood in all
CHS samples (N=3845)
a

Ever MD-diagnosed
asthma
Ever wheezing
Wheezing in the previous
12 months
Bronchitic symptoms in
the previous 12 months
RS Number OR P
Adjusted
P
OR P
Adjusted
P
OR P
Adjusted
P
OR P
Adjusted
P
rs2301235
1.02 0.85 0.90 0.98 0.77 0.86 0.98 0.79 0.86 1.01 0.92 0.98
rs2569692
1.01 0.90 0.90 0.96 0.57 0.77 0.89 0.14 0.38 0.99 0.87 0.98
rs28364580
1.03 0.73 0.90 0.96 0.47 0.73 0.91 0.21 0.45 1.00 0.98 0.98
rs1709122
0.98 0.81 0.90 1.01 0.88 0.91 0.89 0.08 0.38 0.91 0.13 0.50
rs4803446
0.98 0.78 0.90 0.97 0.61 0.77 0.91 0.16 0.38 0.95 0.47 0.72
rs2271546
1.04 0.66 0.90 0.92 0.34 0.60 0.86 0.12 0.38 0.99 0.91 0.98
rs186235601
0.97 0.81 0.90 0.90 0.33 0.60 0.88 0.35 0.52 0.81 0.12 0.50
rs10409940
1.03 0.66 0.90 1.03 0.63 0.77 0.93 0.26 0.46 0.92 0.18 0.50
rs11083613
1.05 0.53 0.90 1.08 0.23 0.60 1.14 0.08 0.38 0.99 0.93 0.98
rs4803447
0.94 0.57 0.90 0.89 0.19 0.60 1.00 0.98 0.98 0.94 0.56 0.78
rs59423102
1.03 0.76 0.90 1.02 0.77 0.86 0.93 0.46 0.61 0.97 0.78 0.98
rs12462203
0.90 0.09 0.37 0.97 0.52 0.77 0.94 0.29 0.47 0.92 0.18 0.50
rs12984621
0.93 0.24 0.61 1.01 0.86 0.91 0.94 0.38 0.53 0.96 0.49 0.72
rs4802112
0.88 0.06 0.31 0.89 0.04 0.52 0.88 0.06 0.38 0.90 0.10 0.50
rs4803449
0.88 0.05 0.31 0.94 0.24 0.60 0.91 0.14 0.38 0.90 0.12 0.50
rs76249126
1.04 0.75 0.90 1.09 0.35 0.60 1.13 0.26 0.46 1.08 0.48 0.72
rs75955910
0.98 0.85 0.90 0.91 0.34 0.60 0.96 0.75 0.86 1.21 0.08 0.50
rs7246896
1.02 0.88 0.90 1.09 0.36 0.60 0.99 0.92 0.95 0.92 0.46 0.72
rs77287588
1.06 0.49 0.90 0.97 0.63 0.77 0.98 0.78 0.86 1.12 0.16 0.50
rs4802114
0.92 0.18 0.57 0.93 0.18 0.60 0.94 0.34 0.52 0.89 0.06 0.50
rs4637024
1.02 0.85 0.90 1.10 0.19 0.60 1.13 0.16 0.38 1.03 0.75 0.98
66
rs3786555
0.85 0.04 0.31 0.94 0.32 0.60 0.84 0.02 0.38 0.94 0.41 0.72
rs2304234
0.90 0.08 0.37 0.93 0.18 0.60 0.93 0.23 0.45 0.95 0.37 0.72
rs55841050
0.93 0.55 0.90 0.83 0.06 0.52 0.97 0.79 0.86 0.88 0.28 0.68
rs12983027
0.84 0.04 0.31 0.93 0.26 0.60 0.89 0.15 0.38 0.93 0.39 0.72
rs12978323
0.85 0.03 0.31 0.87 0.02 0.51 0.87 0.05 0.38 0.91 0.17 0.50
rs116056574
0.90 0.12 0.42 0.91 0.10 0.60 0.87 0.04 0.38 0.93 0.29 0.68
rs35546772
0.85 0.22 0.61 1.00 0.98 0.98 0.93 0.58 0.74 0.99 0.96 0.98
a
SNPs were modeled as ordinal variables (0=major allele, 1=heterozygote, and 2=minor allele) and models were adjusted for child's sex,
age, ethnicity and admixture. Odds ratios are presented for one unit increase in SNP. For all comparisons the reference group is children
not having the corresponding outcome. Tagging SNPs were defined with a pair tag r
2
>0.8 in Haploview with all CHS samples
(N=3845). FDR was used to adjust for tests performed at the 28 tagged SNPs. Significant FDR-adjusted p-values (<0.05) are marked in
bold.

67
Figure 2.1. Genomic location of AXL CpG sites and SNPs under investigation.

Solid black box: CpG sites in the near-TSS region (cg10564498, cg03247049, cg12722469,
cg02372201, cg19848291 and cg14892768); dashed grey box: CpG sites in the gene-body region
(cg27579501, cg00360107, cg19270050, cg24901063 and cg26521562); dashed black box: CpG
site in the 3' untranslated region (cg20964856). Definition of abbreviations: TSS = transcription
start site.

68
Figure 2.2. The association between AXL mRNA and average methylation at all 12 CpG sites.

29 TCGA histologically normal lung tissue samples with HM450 methylation and RNA
sequencing data were used. Spearman correlation coefficients and p-values are shown.

69
Figure 2.3. The association between cg10564498 methylation and genotype at each tagging SNP
in the replication population (N=728).

Unadjusted -log(p-value) are displayed, with up-triangle and down-triangle indicating positive
and negative associations, respectively. Linkage disequilibrium (LD) heatmap of tagging SNPs is
shown with color scheme of the r-square.

70
Figure 2.4. Illustration of epigenetic marks in AXL gene-body region (yellow box) and CpG sites
(red and green bars) in multiple cell lines.

This region contains a putative enhancer in IMR90 fetal lung fibroblast cells (light green in
chromHMM track) and adult CD4 Naïve Primary cells (yellow in chromHMM track), and is
adjacent to a putative enhancer in NHLF adult lung fibroblast cells (light green in chromHMM
track). There are also transcription factor binding sites located within this region in all three cell
lines from ChIP-Seq input. This region is enriched with epigenetic marks for poised enhancer
(indicated by H3K4me1), active enhancer (indicated by H3K27ac) and active transcription
(green in chromHMM track). CpG site 1: cg27579501; CpG site 2: cg00360107; CpG site 3:
cg19270050; CpG site 4: cg24901063; CpG site 5: cg26521562. IMR90: fetal lung fibroblast
cell; NHLF: normal adult lung fibroblast cell; CD4 Naïve Primary cells: obtained from adult
blood.
71
CHAPTER 3: ASSOCIATION BETWEEN AXL PROMOTER METHYLATION AND
LUNG FUNCTION GROWTH DURING ADOLESCENCE

Lu Gao, Robert Urman, Joshua Millstein, Kimberly D. Siegmund, Louis Dubeau, Carrie V.
Breton

3.1 ABSTRACT
AXL is one of the TAM (TYRO3, AXL and MERTK) receptor tyrosine kinases and may be
involved in airway inflammation. Little is known about how epigenetic changes in AXL may
affect lung development during adolescence. We investigated the association between AXL DNA
methylation at birth and lung function growth from 10 to 18 years of age in 923 subjects from the
Children’s Health Study (CHS). DNA methylation from newborn bloodspots was measured at
multiple CpG loci across the regulatory regions of AXL using Pyrosequencing. Linear spline
mixed-effects models were fitted to assess the association between DNA methylation and 8-year
lung function growth. Findings were evaluated for replication in a separate population of 237
CHS subjects using methylation data from the Illumina HumanMethylation450 (HM450) array
when possible. A 5% higher average methylation level of the AXL promoter region at birth was
associated with a 61.6 ml decrease in mean FVC growth from 10 to 18 years of age in the
primary study population (95% CI: -117.5, -5.7), and a 44.6 ml decrease in mean FVC growth
from 11 to 15 years of age in the replication population (95% CI: -97.2, 8.1). One CpG locus in
the promoter region, cg10564498, was significantly associated with decreased growth in FEV1,
FVC and MMEF from 10 to 18 years of age and the negative associations were observed in a
72
similar age range in the replication population. These findings suggest an association between
AXL promoter methylation at birth and lower lung function growth during adolescence.

3.2 INTRODUCTION
Optimal lung function is important for health, as reduced lung function has been associated
with increased risks of coronary artery disease and respiratory disease in adults, and with the
onset of asthma in adolescents [10, 12]. Many factors, environmental, pathological and genetic,
may influence lung function [290-293]. These factors may exert their influence through
epigenetic mechanisms such as DNA methylation by altering the expression pattern and activity
of genes involved in airway development [223, 242]. Researchers have identified associations
between lung function decline and DNA methylation of the transposable elements LINE-1 [155],
and of several candidate genes such as CRAT, F3, TLR2 and SERPINA1, mostly among older
subjects [16, 158]. DNA methylation at birth may be an early surrogate marker of chronic
disease predisposition that is affected by prenatal environmental exposure [294]. However, few
studies have investigated the early-life epigenetic marks associated with lung function growth in
healthy adolescents.
Methylation of AXL, a member of the TAM (TYRO3, AXL and MERTK) family receptor
tyrosine kinases, was identified in our previous work as responsive to prenatal tobacco smoke
exposure [161, 163]. AXL has been implicated in various biological pathways including
clearance of apoptotic cells, natural killer cell differentiation, and inhibition of proinflammatory
cytokines induced by Toll-like receptors (TLR) [172, 174, 295]. Recent evidence suggests that
AXL is expressed in both human and mouse airway/alveolar macrophages and is critical for
effective phagocytosis. Reduced AXL expression may contribute to persistent airway
73
inflammation through inefficient clearance of apoptotic cells from inflamed lungs [203, 204].
Since defective phagocytosis and subsequent airway inflammation are closely related to
accelerated lung function decline and chronic lung diseases [290, 296], we hypothesized that
increased methylation in AXL promoter region may be associated with reduced lung function
growth even early in life, potentially through repressing its expression and promoting
inflammation in the airways.
In this study, we investigated the association between methylation of multiple CpG sites
across the regulatory regions of AXL at birth and lung function growth from 10 to 18 years of
age. Methylation was first assessed using Pyrosequencing in newborn bloodspots from a subset
of 923 subjects from the Children’s Health Study (CHS) [6, 147, 251]. We then sought to
replicate the association in a separate population of 237 CHS subjects using methylation levels
measured from the Infinium HumanMethylation450 BeadChip (HM450) array, and in whom
lung function was assessed from 11 to 15 years of age.

3.3 MATERIALS AND METHODS
Study population
This study was conducted in participants from the Children’s Health Study, a longitudinal
study of respiratory health of children in southern California [6, 147, 251]. Based on our ability
to link CHS subjects with California birth records and to obtain a newborn bloodspot, a subset of
923 children was selected for an epigenetic study in which DNA methylation at multiple CpG
loci on AXL was assessed using Pyrosequencing. Subjects were recruited either as fourth-grade
students in 1996 (cohort D) or as kindergarten and first-grade students in 2003 (cohort E). The
sample selected was enriched with subjects exposed to prenatal tobacco smoke. The replication
74
population included 237 CHS subjects from cohort E who had previously participated in a
substudy of childhood cardiovascular health [252], for whom California birth records could be
linked and at least 700ng of DNA was available from a dried newborn bloodspot. In these
subjects bloodspot DNA methylation was measured with the HM450 array.
Written questionnaires were completed by children’s parents at study entry and were
updated annually throughout the study thereafter. The questionnaires were used to obtain
personal, parental and socio-demographic characteristics with respect to the child, which
included the child’s age, sex, race, ethnic origin, level of parental education, and in-utero as well
as post-natal tobacco smoke exposure.
The study was approved by the Institutional Review Board of the University of Southern
California.
Pulmonary-function testing
Trained technicians measured subjects’ weight and height, and supervised performance of
pulmonary-function maneuvers. Details of the testing protocol have been published previously
[147]. Three measures of pulmonary-function were analyzed for each child: forced vital capacity
(FVC), forced expiratory volume in the first second (FEV1), and maximal midexpiratory flow
rate (MMEF). Pulmonary-function testing was performed annually from approximately 10 to 18
years of age with the use of rolling-seal spirometers in cohort D. Cohort E subjects were tested
for pulmonary-function every other year when they were approximately 11, 13 and 15 years of
age with the use of pressure transducer-based spirometers. Information on asthma status and
personal smoking status was collected at the time of each pulmonary-function testing.
DNA methylation
75
DNA methylation was measured in newborn bloodspots (NBS) obtained as part of the
routine California Newborn Screening Program from the California Department of Public Health
Genetic Disease Screening Program. The NBS were stored by the state of California at -20
degrees Celsius. A single complete newborn bloodspot for each requested participant was mailed
to us and stored in our lab at -80 degrees Celsius upon receipt. Laboratory personnel performing
DNA methylation analysis were blinded to study subject information. DNA was extracted from
whole blood cells using the QiaAmp DNA blood kit (Qiagen Inc, Valencia, CA) and stored at -
80 degrees Celsius. Eleven CpG loci (CpG 1-3, and 5-12) spanning the regulatory regions of
AXL were selected for Pyrosequencing assays (Figure 3.1); five (CpG 3, 5, 8, 10 and 12) were
located at positions corresponding to positions in the HM450 array. Methylation of CpG 1 was
previously reported to be associated with prenatal tobacco smoke exposure [161, 163].
Pyrosequencing assays were conducted as previously described [297]. PCR primers were
designed by EpigenDx Inc. (http://www.epigendx.com) to cover the loci of interest. 500 ng of
genomic DNA extracted from each sample was bisulfite treated using the EZ DNA Methylation
Kit™ (Zymo Research, Irvine, CA, USA) and was purified according to the manufacturer’s
protocol. The methylation level was determined using QCpG software (Pyrosequencing, Qiagen)
and was reported as percent of DNA methylation for each CpG locus.
For replication analyses, two CpG loci (CpG 3 (cg10564498) and CpG 5 (cg12722469))
were selected for replication based on results with the primary study population. We additionally
selected another HM450 array targeting locus in the promoter region (CpG 4 (cg03247049)) to
further investigate this region (Figure 3.1). For HM450 assays, 700 to 1000 ng of genomic DNA
from each sample was treated with bisulfite using the EZ-96 DNA Methylation Kit™ (Zymo
Research, Irvine, CA, USA), according to the manufacturer’s recommended protocol and eluted
76
in 18 µl. The Infinium HM450 data was processed as previously reported [297], compiled for
each locus and was expressed as beta (β) values. Methylation data was extracted for CpG 3, 4
and 5.
Genotyping
Details of genotyping assays, data processing and assessment of admixture were previously
described [298]. Genotypes of SNPs in AXL and its surrounding region (1 kb upstream and
downstream) were extracted from the CHS genome-wide genotypic data. 28 tagged SNPs were
identified with a pair tag r
2
>0.8 in Haploview using all available CHS samples (N=3845) and
were included in the analyses [265]. Admixture was assessed using the program STRUCTURE
from a set of ancestral informative markers that were scaled to represent the proportion of
African American, Asian, Native American and White [266].
In silico analyses in publicly available data
To evaluate the association between AXL promoter methylation and mRNA expression, we
downloaded AXL methylation profiling data in 29 histologically normal tissue samples from
cases with lung adenocarcinoma (LUAD) or lung squamous cell carcinoma (LUSC) from the
TCGA dataset [250]. All samples had both methylation profiling (HM450 array) and RNA-seq
(Illumina HiSeq) data. The mean age was 65.9 years (SD: 12.39) and 75.9% of the subjects were
males. 51.7% of the subjects were moderate to heavy smokers. Promoter methylation average
was calculated from average methylation of CpG 3 (cg10564498), CpG 4 (cg03247049) and
CpG 5 (cg12722469). Spearman correlation coefficient was calculated to evaluate the correlation
between promoter methylation and mRNA level.
77
R packages coMet and snp.plotter were used to graphically display additional information
about CpGs, including genomic location and functional annotation from the Roadmap
Epigenomics Project [299-301].
Statistical analyses
The methylation status of AXL promoter was defined as the average of CpG 1, 2, 3, 5 and 6
in the primary study population and the average of CpG 3, 4 and 5 in the replication population.
Although defined differently for the two populations, these CpG sites are located in the same
genomic region (Figure 3.1) and the correlations between them in the same population were
relatively high (Table 3.1 and 3.2).
To evaluate the association between AXL methylation and lung function growth from 10 to
18 years of age in the primary study population, all available pulmonary-function measurements
for each subject were used to estimate lung function growth curves. We used a previously-
reported linear spline model to account for the nonlinear pattern of growth during adolescence,
with knots placed at ages 12, 14 and 16 years [154, 302]. The model was fitted for each outcome
and CpG individually and adjusted for child’s age, sex, sex*age interaction, ethnicity, height,
height squared, body-mass index (BMI), BMI squared, city of residence at study entry, history of
asthma, parental history of asthma, and maternal smoking during pregnancy. Random effects
were included to account for multiple measurements contributed by each subject. Additional
adjustment for AXL genetic polymorphisms, methylation plate, admixture, cohort, parental
education level, second-hand smoke exposure, wheezing symptoms, personal smoking status and
field technician did not change the effect estimates by more than 10% and were removed from
final models. Visual inspection of residual values did not identify any departures from model
assumptions. Estimates of association between methylation with 8-year lung function
78
development (from 10 to 18 years of age) and with mean attained lung function at both 10 and 18
were obtained. Sensitivity analyses were conducted to evaluate the association in non-asthmatic
subjects, defined as subjects who were never diagnosed with asthma before the last pulmonary
function testing visit. To assess whether associations between lung function growth and
methylation are modified by sex, an interaction term between sex and methylation was included
in the regression models, and Wald tests were used to compute interaction p-values.
A similar mixed-effect linear spline model was used in the replication population, with knots
placed at ages 12 and 14 years and the same adjustment variables. Due to the limited availability
of lung function measurements, the model was constructed to yield estimates of the association
between methylation and 4-year lung function development (from 11 to 15 years of age). Effect
modification by sex was also tested.
All tests assumed a two-sided alternative hypothesis and were conducted using SAS
statistical analysis software (version 9.4). An alpha level of 0.05 was used to determine statistical
significance.

3.4 RESULTS
Characteristics of the study participants
The characteristics of the primary and replication populations are shown in Table 3.3. There
were fewer boys than girls in both the primary study population (46.7% male subjects) and
replication population (41.8% male subjects). The primary study population also had more white
subjects (46.3% versus 32.1%), slightly higher parental education level, more asthmatics (21.5%
versus 14.8%), more subjects exposed to prenatal tobacco smoke by design (22.9% versus 9.0%),
and more subjects exposed to environmental tobacco smoke (20.2% versus 6.8%). There were
79
very few smokers in both populations. Mean age at first pulmonary function testing was 10.8 and
11.4 years in the primary and replication populations, respectively. There were generally no
significant differences for mean height and BMI at age 11, 13 and 15 years between the two
populations.
Genomic locations of the AXL CpG sites under investigation are shown in Figure 3.1.
Methylation levels at many of the CpG sites were significantly correlated (Table 3.1 and 3.2),
with CpG sites closer to each other showing stronger correlations. The distribution of
methylation at each CpG site in both populations is shown in Table 3.4. The patterns of lung
function growth across the study period are shown in Table 3.5 and Figure 3.2.
DNA methylation of AXL and lung function growth
We investigated whether AXL DNA methylation was associated with lung function growth
in both the primary and replication populations. A 5% higher average promoter methylation level
was significantly associated with a 61.6 ml lower mean FVC growth from 10 to 18 years of age
in the primary study population (95% CI: -117.5, -5.7; P=0.03) (Figure 3.3-Panel B and Table
3.6), and a 44.6 ml lower mean FVC growth from 11 to 15 years of age in the replication
population (95% CI: -97.2, 8.1; P=0.10) (Figure 3.3-Panel E and Table 3.7). Similar negative
associations with FEV1 and MMEF growth were observed in the two populations (Figure 3.3,
Tables 3.6 and 3.7).
Methylation at several individual CpG sites in the promoter region was also associated with
lower lung function growth (Figure 3.4-Panel A). For example, methylation at CpG 3 was
significantly associated with lower growth in FEV1, FVC and MMEF from 10 to 18 years of age
in the primary study population (Table 3.6). Methylation at CpG 2, which is located near CpG 3,
was also significantly associated with lower mean 8-year FVC growth, and lower attained FEV1
80
and FVC levels at 18 years of age. In the replication sample, we observed associations between
CpG 3 methylation with lower growth in FEV1 and MMEF from 11 to 15 years of age (Table
3.7). No associations between lung function and methylation of specific CpG sites in the gene-
body region or 3’ untranslated region were observed (Table 3.6).
Since asthma itself is associated with lower lung function growth in children [133] and the
prevalence of asthma in the primary study population was higher, we also conducted sensitivity
analyses in non-asthmatic subjects (Table 3.8). The observed associations were generally
consistent with using all subjects.
We also investigated whether the relationship between AXL promoter methylation and lung
function growth differed by sex (Table 3.9). Although there was no significant interaction, a
stronger association was observed for boys than girls in both populations. For example, in the
primary study population, a 5% higher average promoter methylation was associated with a 69.4
ml decrease in mean 8-year FEV1 growth in boys (95% CI: -139.9, 1.1; P=0.05) but only a 33.0
ml decrease in girls (95% CI: -94.1, 28.0; P=0.29).
AXL promoter methylation and expression in lung
AXL was previously reported to be expressed at very low levels in blood [297]. Thus, we
evaluated the correlation between AXL promoter methylation and mRNA expression levels using
histologically normal lung tissue samples based on HM450 array and RNA sequencing data from
TCGA (Figure 3.5). Average promoter methylation showed negative correlation with expression
(r=-0.40, P=0.03). Although these data need to be interpreted with caution because they were
generated in an older population enriched with smokers, they showed preliminary evidence that
higher methylation might be associated with lower AXL expression level in the adult lung.
81
Lastly, we integrated chromHMM and RNA-seq data from the Roadmap Epigenomics
Project to display a functional annotation of CpG sites in the promoter region in relevant tissues
and cell types (Figure 3.4-Panel B). The CpG loci evaluated in the AXL promoter region are
located adjacent to enhancers in both fetal and adult lung fibroblast cells (yellow in chromHMM
tracks), but are located within transcriptionally-repressed region in cord blood primary T cells.
The observed chromatin state patterns are similar in fetal and adult lung cells. As indicated from
RNA-seq data, the CpG sites under investigation are located within or near the transcriptionally
active region in both fetal and adult lung, and a consistent pattern is observed in adult blood
progenitor cells, though at much lower levels. These data suggest that epigenetic alterations in
this region of AXL may be associated with active patterns for transcription and chromatin state of
this gene throughout life, supporting the use of blood as surrogate marker of AXL’s activity in
fetal and adult lung.

3.5 DISCUSSION
We showed that increased average methylation of the AXL promoter region at birth was
associated with lower lung function growth during adolescence, as measured by the growth in
FEV1, FVC and MMEF from the ages of 10 to 18 years. Similar associations were observed for
the average promoter methylation and specifically, CpG 3 in this region in a separate population
of 237 CHS subjects, despite differences in the age at which lung function was measured as well
as differences in some baseline characteristics of the two populations.
AXL and other TAM receptors are broadly expressed by cells of the vascular, nervous,
immune and reproductive systems [164]. Although the role of AXL in the pathogenesis of
cancers and cardiovascular events is well-characterized [275, 278], little is known about its role
82
in pulmonary homeostasis, which requires a balance between adequate responses to pathogens
and control of inflammatory processes arising from accumulation of apoptotic cells and cellular
products [303]. Inefficient clearance of apoptotic cells in the airways may lead to accumulation
of necrotic cell debris, subsequent uncontrolled inflammatory responses, and ineffective
resolution of lung inflammation upon microbial and viral infection [304]. Recent evidence from
mouse models suggests that Axl is expressed in airway macrophages [203] – the main cell
populations responsible for phagocytosis in the respiratory tract [304] – and is essential for
macrophage function through clearance of apoptotic cells and cell debris and inhibiting
influenza-induced inflammation [203]. Similar findings were reported in human patients with
moderate-to-severe asthma, suggesting a critical role of AXL in clearing apoptotic cells from the
inflamed lung [204].
This is the first paper investigating the role of epigenetic regulation of AXL in lung function
growth during adolescence. Understanding predictors of lung function growth is important
considering the maximal attained lung function predicts risk of subsequent cardiovascular and
respiratory conditions during adult life [124, 305]. Because we previously found that AXL
methylation was associated with risk of asthma-related symptoms in childhood [297], and
asthma may affect the rate of lung function growth [133], we performed sensitivity analyses in
subjects who were never diagnosed with asthma. Similar associations in non-asthmatics indicate
that the epigenetic regulation of AXL on lung function growth may act independently of asthma
status and probably through different biological pathways. The question of whether AXL
regulates lung homeostasis mainly through the apoptosis pathways as opposed to also exerting
negative control over pro-inflammatory cytokines and TLR signaling in the airway [164, 174]
remains unclear. Our results also support a stronger association between AXL promoter
83
methylation and lower function growth among boys than girls. Despite the non-significant
differences, the findings highlight the need for investigating whether there are sex-specific
effects.
The AXL promoter region under investigation in this paper overlapped with the core
promoter region (-556 to -182 upstream to translational start codon), which is a known Sp1/Sp3
transcription factor binding site. Methylation within or in close proximity to this core region is
negatively associated with AXL gene expression in cultured colon cancer cells [162]. Due to the
low expression of AXL in blood [288], we were not able to test whether AXL methylation
correlated with mRNA expression in our population. Nonetheless, average promoter methylation
of AXL was shown to correlate with lower mRNA level in normal lung tissue samples from the
TCGA database, suggesting a potential link between higher promoter methylation and lower
expression level of this gene. Moreover, in silico analyses using data from the Roadmap
Epigenomics Project suggest that this region may be associated with active transcription in the
lung since it is located adjacent to enhancers in both fetal and adult lung fibroblast cells.
The strengths of this study include the analysis of DNA methylation using Pyrosequencing,
a highly reproducible method that accurately quantifies DNA methylation [156], and the
replication of results in a separate population using a different assay. We also utilized the
longitudinal measurements of pulmonary function and covariates data, collected in a consistent
manner throughout the study period, to study the lung function growth prospectively. The
temporal separation of DNA methylation assessment (at birth) and lung function growth
measurement (during adolescence) removes the possibility of reverse causation.
Some limitations are noteworthy. As in any epidemiologic study, the observed effects could
be biased by some unknown confounders associated with both DNA methylation levels and lung
84
development. Since DNA methylation was only measured from newborn bloodspots in our study,
we were not able to evaluate if the methylation pattern in AXL still persists during adolescence,
or whether there is any correlation between AXL methylation in blood and other more
pathologically-relevant tissues such as lung or airway epithelial cells. Nonetheless, the epigenetic
changes at birth may reflect a summary of in-utero environmental exposures that affect different
tissues systematically. The subtle shifts in methylation occurring during fetal development
reflected in newborn blood may serve as early biomarkers for the disruption of pathways
involved in lung function growth later in life.
In conclusion, our results suggest that AXL promoter DNA methylation at birth is associated
with lower lung function growth during adolescence. Future studies on gene-specific methylation
will improve our understanding of the relationship between epigenetic changes and lung
development.
85
3.6 TABLES AND FIGURES
Table 3.1. Spearman correlation between methylation at each AXL CpG site in the primary study population (N=923)
Region CpG
CpG
1
CpG
2
CpG 3
(cg10564498)
CpG 5
(cg12722469)
CpG
6
CpG
7
CpG 8
(cg27579501)
CpG
9
CpG 10
(cg00360107)
CpG
11
CpG 12
(cg20964856)
Promoter
CpG 1 1.00 0.89 0.77 0.73 0.67 0.05 -0.05 0.18 0.18 0.18 -0.21
CpG 2 1.00 0.77 0.74 0.68 0.03 -0.04 0.15 0.18 0.18 -0.24
CpG 3
(cg10564498)
1.00 0.70 0.68 0.14 -0.03 0.26 0.31 0.33 -0.19
CpG 5
(cg12722469)
1.00 0.84 0.11 -0.03 0.28 0.29 0.28 -0.21
CpG 6 1.00 0.16 -0.04 0.36 0.37 0.4 -0.23
Gene-
body
CpG 7 1.00 0.18 0.35 0.38 0.37 0.03
CpG 8
(cg27579501)
1.00 0.08 0.06 0.03 0.06
CpG 9 1.00 0.74 0.76 -0.14
CpG 10
(cg00360107)
1.00 0.85 -0.15
CpG 11 1.00 -0.16
3' UTR
CpG 12
(cg20964856)
1.00
Bold: p-value < 0.05
UTR=untranslated region
Cg number in parenthesis: corresponding CpG locus in HM450 array.
86
Table 3.2. Spearman correlation between methylation at each
AXL CpG site in the replication population (N=237)
CpG
CpG 3
(cg10564498)
CpG 4
(cg03247049)
CpG 5
(cg12722469)
CpG 3
(cg10564498)
1.00 0.55 0.53
CpG 4
(cg03247049)
1.00 0.50
CpG 5
(cg12722469)
1.00
Bold: p-value < 0.05
Cg number in parenthesis: corresponding CpG locus in HM450 array.

87
Table 3.3. Characteristics of participants with lung function testing

Primary study
population
(N=923)

Replication
population
(N=237)
P-
value
a

Male sex, n (%) 431 (46.7) 99 (41.8) 0.17
Ethnicity, n (%) 0.0002
Hispanic 385 (41.7) 131 (55.3)
Non-Hispanic White 427 (46.3) 76 (32.1)
Asian/Black/Other 111 (12.0) 30 (12.7)
Parental Education, n (%) 0.11
High school or less 272 (30.4) 73 (33.0)
Some college 395 (44.2) 81 (36.7)
Finished college/some graduate school 227 (25.4) 67 (30.3)
Parental history of asthma at study entry, n (%) 229 (25.8) 31 (13.9) 0.0002
Ever MD-diagnosed asthma at first lung
function testing, n (%)
198 (21.5) 35 (14.8) 0.02
In-utero tobacco smoke exposure, n (%) 209 (22.9) 21 (9.0) <0.0001
Ever smoked cigarettes by age 15, n (%) 82 (8.9) 10 (4.2) 0.02
Passive tobacco smoke exposure by age 15, n
(%)
186 (20.2) 16 (6.8) <0.0001
Age (year) at first pulmonary function testing,
mean (sd)
10.8 (1.1) 11.4 (0.8) <0.0001
Height (cm), mean (sd)
Age 11 148.6 (7.0) 147.8 (7.0) 0.26
Age 13 160.2 (7.8) 159.1 (7.2) 0.12
Age 15 167.3 (8.4) 165.4 (8.2) 0.05
Body mass index (BMI, kg/m
2
), mean (sd)
Age 11 20.4 (4.2) 20.4 (4.1) 0.97
Age 13 21.9 (4.8) 21.7 (4.4) 0.68
Age 15 23.3 (4.6) 23.5 (4.5) 0.78
a
P-value for testing the difference between the two populations. Derived from a Pearson's Chi-
squared test for categorical variables and from an unequal variance 2-sample t-test for age, and an
equal variance 2-sample t-test for height and BMI.

88
Table 3.4. DNA methylation levels (%) at AXL CpG sites in the
primary study population (N=923) and the replication population
(N=237)
CpG
Distance to
TSS
Genomic
location
a

Mean %
methylation (sd)
Primary study population
CpG 1 -508 41724600 42.9 (8.2)
CpG 2 -492 41724616 50.1 (7.7)
CpG 3 (cg10564498) -455 41724653 19.9 (4.5)
CpG 5 (cg12722469) -55 41725053 11.6 (3.8)
CpG 6 -31 41725077 11.1 (3.4)
CpG 7 4518 41729626 36.6 (3.6)
CpG 8 (cg27579501) 4550 41729658 83.9 (3.1)
CpG 9 6813 41731921 2.3 (1.1)
CpG 10 (cg00360107) 6827 41731935 4.5 (1.8)
CpG 11 6834 41731942 5.7 (2.1)
CpG 12 (cg20964856) 42562 41767670 58.5 (3.2)
Average of promoter CpG
sites
b

27.1 (5.0)
Replication population
CpG 3 (cg10564498) -455 41724653 26.6 (9.4)
CpG 4 (cg03247049) -210 41724898 21.3 (4.1)
CpG 5 (cg12722469) -55 41725053 17.1 (4.9)
Average of promoter CpG
sites
c

21.7 (5.1)
a
Genomic location is showing the corresponding base pair number (hg19) on
chromosome 19.
b
The promoter average in the primary study population is taken by the average of
CpG 1, 2, 3, 5 and 6.
c
The promoter average in the replication population is taken by the average of CpG 3,
4 and 5.
TSS = transcription start site.
Cg number in parenthesis: corresponding CpG locus in HM450 array.
89
Table 3.5. Mean levels of growth in lung function during the study period

Primary study population (N=923) Replication population (N=237)
Girls Boys Girls Boys
Outcome
Age of
10 yr
Age of
18 yr
Average
8-yr
growth

Age of
10 yr
Age of
18 yr
Average 8-
yr growth

Age of
11 yr
Age of
15 yr
Average
4-yr
growth

Age of
11 yr
Age of
15 yr
Average
4-yr
growth
FEV 1 (ml) 2036 3395 1359 2107 4587 2480 2416 3309 893 2364 3962 1598
FVC (ml) 2322 3891 1569 2460 5414 2954 2733 3762 1029 2728 4581 1853
MMEF (ml/sec) 2457 3878 1421 2370 4847 2478 2953 3898 945 2776 4419 1642
FEV 1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal midexpiratory flow rate.
90
Table 3.6. Association between 8-year lung function growth and AXL methylation in the primary study population (N=923)
a

Lung function measurement
and CpG site

10 years of age 18 years of age Growth from 10 to 18 years of age
Coefficient (95% CI) P value Coefficient (95% CI) P value Coefficient (95% CI) P value
FEV 1 (ml)
CpG 1 -4.9 (-17.6 to 7.8) 0.45 -27.4 (-60.4 to 5.6) 0.10 -22.5 (-54.6 to 9.7) 0.17
CpG 2 -6.3 (-20.2 to 7.5) 0.37 -38.0 (-72.1 to -3.9) 0.03 -31.6 (-65.0 to 1.8) 0.06
CpG 3 (cg10564498) 8.3 (-14.7 to 31.2) 0.48 -51.4 (-110.1 to 7.3) 0.09 -59.7 (-116.6 to -2.7) 0.04
CpG 5 (cg12722469) -13.6 (-40.7 to 13.4) 0.32 -53.8 (-123.5 to 15.9) 0.13 -40.2 (-107.6 to 27.2) 0.24
CpG 6 3.4 (-27.7 to 34.6) 0.83 -78.2 (-157.0 to 0.5) 0.05 -81.7 (-158.0 to -5.3) 0.04
CpG 7 -11.2 (-39.6 to 17.2) 0.44 14.2 (-56.2 to 84.6) 0.69 25.4 (-43.0 to 93.7) 0.47
CpG 8 (cg27579501) -15.4 (-50.4 to 19.6) 0.39 -2.1 (-86.5 to 82.4) 0.96 13.4 (-69.8 to 96.6) 0.75
CpG 9 -29.9 (-131.3 to 71.6) 0.56 -39.4 (-288.3 to 209.5) 0.76 -9.5 (-251.2 to 232.1) 0.94
CpG 10 (cg00360107) -20.1 (-79.5 to 39.3) 0.51 -4.5 (-148.3 to 139.2) 0.95 15.6 (-123.0 to 154.2) 0.83
CpG 11 -15.4 (-65.5 to 34.7) 0.55 -44.6 (-168.3 to 79.1) 0.48 -29.2 (-149.1 to 90.7) 0.63
CpG 12 (cg20964856) -3.5 (-34.8 to 27.9) 0.83 -53.4 (-130.5 to 23.7) 0.17 -49.9 (-124.4 to 24.6) 0.19
Average of promoter CpG
sites
b

-4.9 (-25.9 to 16.1) 0.65 -53.3 (-106.6 to 0.0) 0.05 -48.4 (-100.2 to 3.4) 0.07
FVC (ml)
CpG 1 -3.5 (-17.1 to 10.1) 0.62 -37.1 (-72.1 to -2.2) 0.04 -33.6 (-68.1 to 0.8) 0.06
CpG 2 -5.5 (-20.4 to 9.4) 0.47 -44.9 (-81.2 to -8.7) 0.02 -39.4 (-75.5 to -3.4) 0.03
CpG 3 (cg10564498) 9.0 (-15.7 to 33.7) 0.47 -58.0 (-120.2 to 4.3) 0.07 -67.0 (-128.2 to -5.7) 0.03
CpG 5 (cg12722469) -9.4 (-38.6 to 19.8) 0.53 -71.5 (-145.5 to 2.6) 0.06 -62.0 (-134.8 to 10.7) 0.09
CpG 6 -0.6 (-34.1 to 32.9) 0.97 -95.6 (-179.4 to -11.9) 0.03 -95.0 (-177.5 to -12.5) 0.02
91
10 years of age 18 years of age Growth from 10 to 18 years of age
Coefficient (95% CI) P value Coefficient (95% CI) P value Coefficient (95% CI) P value
CpG 7 -10.9 (-41.3 to 19.6) 0.48 40.2 (-34.7 to 115.2) 0.29 51.1 (-22.9 to 125.1) 0.18
CpG 8 (cg27579501) -6.8 (-44.2 to 30.6) 0.72 11.4 (-78.1 to 100.8) 0.80 18.2 (-71.4 to 107.7) 0.69
CpG 9 -85.3 (-193.7 to 23.1) 0.12 -46.2 (-309.8 to 217.4) 0.73 39.1 (-220.8 to 299.0) 0.77
CpG 10 (cg00360107) -19.3 (-82.9 to 44.4) 0.55 -14.2 (-167.9 to 139.5) 0.86 5.1 (-145.6 to 155.8) 0.95
CpG 11 -25.1 (-78.9 to 28.6) 0.36 -53.0 (-184.3 to 78.4) 0.43 -27.8 (-157.2 to 101.6) 0.67
CpG 12 (cg20964856) -23.3 (-57.1 to 10.5) 0.18 -56.9 (-139.2 to 25.4) 0.18 -33.6 (-114.6 to 47.3) 0.42
Average of promoter CpG
sites
b

-3.2 (-25.8 to 19.5) 0.78 -64.7 (-121.4 to -8.1) 0.03 -61.6 (-117.5 to -5.7) 0.03
MMEF (ml/sec)
CpG 1 4.8 (-23.6 to 33.2) 0.74 -32.9 (-100.4 to 34.7) 0.34 -37.7 (-102.0 to 26.6) 0.25
CpG 2 4.9 (-26.2 to 35.9) 0.76 -56.1 (-125.8 to 13.5) 0.11 -61.0 (-127.6 to 5.5) 0.07
CpG 3 (cg10564498) 29.9 (-21.5 to 81.2) 0.25 -88.6 (-208.6 to 31.4) 0.15 -118.4 (-232.2 to -4.7) 0.04
CpG 5 (cg12722469) -12.5 (-73.2 to 48.2) 0.69 -74.2 (-216.5 to 68.1) 0.31 -61.7 (-196.2 to 72.7) 0.37
CpG 6 27.6 (-42.2 to 97.4) 0.44 -105.5 (-266.0 to 55.0) 0.20 -133.0 (-285.1 to 19.0) 0.09
CpG 7 -18.3 (-82.0 to 45.4) 0.57 -6.3 (-149.9 to 137.3) 0.93 12.0 (-123.8 to 147.7) 0.86
CpG 8 (cg27579501) -41.2 (-119.5 to 37.0) 0.30 -35.9 (-209.6 to 137.9) 0.69 5.4 (-162.0 to 172.7) 0.95
CpG 9 50.5 (-176.0 to 277.0) 0.66 -84.3 (-586.0 to 417.5) 0.74 -134.8 (-609.4 to 339.8) 0.58
CpG 10 (cg00360107) 2.4 (-130.6 to 135.4) 0.97 -81.6 (-373.6 to 210.4) 0.58 -84.0 (-358.5 to 190.5) 0.55
CpG 11 14.5 (-97.5 to 126.6) 0.80 -84.5 (-337.0 to 168.1) 0.51 -99.0 (-337.8 to 139.8) 0.42
CpG 12 (cg20964856) 7.5 (-63.9 to 78.9) 0.84 -26.5 (-184.1 to 131.2) 0.74 -34.0 (-183.4 to 115.4) 0.66
Average of promoter CpG
sites
b

11.4 (-35.7 to 58.5) 0.64 -78.2 (-186.9 to 30.5) 0.16 -89.6 (-192.8 to 13.7) 0.09
92
a
All models include adjustments for height and height
2
, body mass index (BMI) and BMI
2
, sex, age, sex*age interaction, ethnicity, city of residence at
study entry, history of asthma, parental history of asthma, and maternal smoking during pregnancy. Effect estimates are presented for an increase in 5%
of DNA methylation level at birth. Significant p-values (<0.05) are marked in bold. Cg number in parenthesis: corresponding CpG locus in HM450
array.
b
The promoter average is taken by the average of CpG 1, 2, 3, 5 and 6.
FEV 1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal midexpiratory flow rate.

93
Table 3.7. Association between 4-year lung function growth and
AXL promoter methylation in the replication population (N=237)
a

Lung function measurement and
CpG site
Growth from 11 to 15 years of age
Coefficient (95% CI) P value
FEV 1 (ml)
CpG 3 (cg10564498) -28.8 (-56.3 to -1.3) 0.04
CpG 4 (cg03247049) -57.2 (-118.5 to 4.1) 0.07
CpG 5 (cg12722469) -28.0 (-80.1 to 24.1) 0.29
Average of promoter CpG
sites
b

-53.9 (-104.3 to -3.5) 0.04
FVC (ml)
CpG 3 (cg10564498) -22.6 (-51.2 to 6.1) 0.12
CpG 4 (cg03247049) -54.6 (-118.6 to 9.4) 0.09
CpG 5 (cg12722469) -24.2 (-78.7 to 30.2) 0.38
Average of promoter CpG
sites
b

-44.6 (-97.2 to 8.1) 0.10
MMEF (ml/sec)
CpG 3 (cg10564498) -93.2 (-154.6 to -31.9) 0.003
CpG 4 (cg03247049) -100.5 (-238.1 to 37.1) 0.15
CpG 5 (cg12722469) -118.3 (-234.6 to -2.1) 0.05
Average of promoter CpG
sites
b

-166.2 (-278.4 to -54.1) 0.004
a
All models include adjustments for height and height
2
, body mass index (BMI)
and BMI
2
, sex, age, sex*age interaction, ethnicity, city of residence at study
entry, history of asthma, parental history of asthma, and maternal smoking
during pregnancy. Effect estimates are presented for an increase in 5% of DNA
methylation level at birth. Significant p-values (<0.05) are marked in bold. Cg
number in parenthesis: corresponding CpG locus in HM450 array.
b
The promoter average is taken by the average of CpG 3, 4 and 5.
FEV 1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF,
maximal midexpiratory flow rate.

94
Table 3.8. Sensitivity analyses in non-asthmatic subjects for association between 8-year lung function growth and AXL promoter methylation in
the primary study population
a

Lung function
measurement and CpG
site
10 years of age 18 years of age Growth from 10 to 18 years of age
All subjects
(N=923)
Non-asthmatic
subjects (N=640)
b

All subjects (N=923)
Non-asthmatic
subjects (N=640)
b

All subjects (N=923)
Non-asthmatic
subjects (N=640)
b

Coefficient
P
value
Coefficient
P
value
Coefficient
P
value
Coefficient
P
value
Coefficient
P
value
Coefficient
P
value
FEV 1 (ml)
CpG 1 -4.9 0.45 -7.5 0.32 -27.4 0.10 -26.2 0.19 -22.5 0.17 -18.8 0.33
CpG 2 -6.3 0.37 -7.0 0.40 -38.0 0.03 -42.7 0.04 -31.6 0.06 -35.7 0.07
CpG 3 (cg10564498) 8.3 0.48 5.1 0.71 -51.4 0.09 -69.4 0.05 -59.7 0.04 -74.5 0.03
CpG 5 (cg12722469) -13.6 0.32 -0.9 0.95 -53.8 0.13 -59.3 0.16 -40.2 0.24 -58.4 0.15
CpG 6 3.4 0.83 13.6 0.45 -78.2 0.05 -84.9 0.07 -81.7 0.04 -98.5 0.03
Average of promoter
CpG sites
c

-4.9 0.65 -5.4 0.66 -53.3 0.05 -62.5 0.06 -48.4 0.07 -57.0 0.07
FVC (ml)
CpG 1 -3.5 0.62 -6.8 0.40 -37.1 0.04 -30.4 0.15 -33.6 0.06 -23.6 0.25
CpG 2 -5.5 0.47 -7.5 0.40 -44.9 0.02 -49.3 0.02 -39.4 0.03 -41.9 0.05
CpG 3 (cg10564498) 9.0 0.47 3.3 0.82 -58.0 0.07 -62.3 0.10 -67.0 0.03 -65.7 0.07
CpG 5 (cg12722469) -9.4 0.53 2.6 0.88 -71.5 0.06 -61.0 0.17 -62.0 0.09 -63.6 0.14
CpG 6 -0.6 0.97 11.8 0.55 -95.6 0.03 -90.5 0.07 -95.0 0.02 -102.3 0.03
Average of promoter
CpG sites
c

-3.2 0.78 -5.4 0.69 -64.7 0.03 -67.2 0.05 -61.6 0.03 -61.8 0.07
MMEF (ml/sec)
CpG 1 4.8 0.74 7.4 0.67 -32.9 0.34 -28.0 0.51 -37.7 0.25 -35.4 0.38
CpG 2 4.9 0.76 9.9 0.60 -56.1 0.11 -51.8 0.23 -61.0 0.07 -61.7 0.14
CpG 3 (cg10564498) 29.9 0.25 29.8 0.34 -88.6 0.15 -123.3 0.10 -118.4 0.04 -153.1 0.03
CpG 5 (cg12722469) -12.5 0.69 15.5 0.67 -74.2 0.31 -88.5 0.32 -61.7 0.37 -104.0 0.21
CpG 6 27.6 0.44 54.0 0.19 -105.5 0.20 -120.7 0.22 -133.0 0.09 -174.7 0.06
Average of promoter
CpG sites
c

11.4 0.64 21.0 0.46 -78.2 0.16 -85.0 0.22 -89.6 0.09 -106.0 0.10
a
All models include adjustments for height and height
2
, body mass index (BMI) and BMI
2
, sex, age, sex*age interaction, ethnicity, city of residence at study entry, history of
asthma, parental history of asthma, and maternal smoking during pregnancy. Effect estimates are presented for an increase in 5% of DNA methylation level at birth.
Significant p-values (<0.05) are marked in bold. Cg number in parenthesis: corresponding CpG locus in HM450 array.
b
Non-asthmatic subjects were defined as subjects who were never diagnosed with asthma before the last pulmonary function testing.
c
The promoter average is taken by the average of CpG 1, 2, 3, 5 and 6.
FEV 1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal midexpiratory flow rate.
95
Table 3.9. Association between lung function growth and AXL promoter methylation stratified by
sex in the primary study population (N=923) and replication population (N=237)
a

Lung function
measurement
Primary study population Replication population
Growth from 10 to 18 years of
age

Growth from 11 to 15 years of
age
Coefficient (95% CI) P value Coefficient (95% CI) P value
FEV1 (ml)

Overall -48.4 (-100.2 to 3.4) 0.07 -53.9 (-104.3 to -3.5) 0.04
By sex

Girls -33.0 (-94.1 to 28.0) 0.29 -17.5 (-87.8 to 52.8) 0.62
Boys -69.4 (-139.9 to 1.1) 0.05 -93.6 (-166.1 to -21.1) 0.01
Interaction p-value
0.37 0.14
FVC (ml)

Overall -61.6 (-117.5 to -5.7) 0.03 -44.6 (-97.2 to 8.1) 0.10
By sex

Girls
-45.0 (-112.3 to 22.3) 0.19 -45.5 (-119.2 to 28.3) 0.22
Boys
-84.0 (-161.9 to -6.0) 0.03 -45.3 (-122.4 to 31.7) 0.25
Interaction p-value
0.40 0.99
MMEF (ml/sec)

Overall
-89.6 (-192.8 to 13.7)
0.09
-166.2 (-278.4 to -54.1)
0.004
By sex
Girls -75.3 (-194.4 to 43.8) 0.22 -87.0 (-244.6 to 70.5) 0.28
Boys -110.7 (-249.3 to 28.0) 0.12 -249.7 (-410.5 to -88.9) 0.003
Interaction p-value
0.65 0.16
a
The promoter average is taken by the average of CpG 1, 2, 3, 5 and 6 in the primary study
population and by the average of CpG 3, 4 and 5 in the replication population.
All models include adjustments for height and height
2
, body mass index (BMI) and BMI
2
, sex,
age, sex*age interaction, ethnicity, city of residence at study entry, history of asthma, parental
history of asthma, and maternal smoking during pregnancy. Effect estimates are presented for an
increase in 5% of DNA methylation level at birth. Significant p-values (<0.05) are marked in bold.
Cg number in parenthesis: corresponding CpG locus in HM450 array.
FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal
midexpiratory flow rate.

96
Figure 3.1. Genomic locations of AXL CpG sites under investigation.

Solid black box: CpG sites in the promoter region (CpG1-CpG 6); dashed gray box: CpG sites in
the gene-body region (CpG 7-CpG 11); dashed black box: CpG site in the 3' untranslated region
(CpG 12).
Cg number in parenthesis: corresponding CpG locus in HM450 array.

97
Figure 3.2. Lung function growth curves for boys and girls over the study period for FEV1, FVC
and MMEF in the primary study population (Panel A-C) and the replication population (Panel D-
F).

Points show individual lung function measures for the subjects. Growth curves (black lines) are
obtained from the linear mixed-effects spline model of lung function and age.
FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal
midexpiratory flow rate.

98
Figure 3.3. Mean lung function growth versus the average methylation level of AXL promoter
region.

AXL promoter methylation levels were first categorized into equal-sized groups based on deciles.
The mean growth in FEV1, FVC and MMEF from 10 to 18 years of age in the primary study
99
population (Panel A-C) and 11 to 15 years of age in the replication population (Panel D-F) are
plotted against the corresponding levels of methylation in each decile. The corresponding
coefficients and P-values are shown.
FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal
midexpiratory flow rate.

100
Figure 3.4. The association between AXL promoter methylation and lung function growth from
10 to 18 years of age (primary study population) and 11 to 15 years of age (replication
population).

Panel A: Coefficients and 95% confidence interval (CI) from mixed-effects model at each
individual CpG site, displayed according to their genomic locations. Effect estimates are
101
presented for an increase in 5% of DNA methylation level at birth. Solid error bars: results from
the primary population; dashed error bars: results from the replication population.
Panel B: Annotation tracks and mRNA expression level from the Roadmap Epigenomics Project
for the plotted genomic region. Black in chromHMM track: transcription repression; yellow:
enhancer; orange: flanking transcription start site (TSS); red: active TSS. Primary T cells:
obtained from cord blood; IMR90: fetal lung fibroblast cell; NHLF: normal adult lung fibroblast
cell.
Cg number in parenthesis: corresponding CpG locus in HM450 array.
CpG loci in the primary study population: CpG 1, 2, 3, 5 and 6; CpG loci in the replication
population: CpG 3, 4 and 5.
FEV1, forced expiratory volume in 1 s; FVC, forced vital capacity; MMEF, maximal
midexpiratory flow rate.

102
Figure 3.5. The association between AXL mRNA and average promoter methylation.

The data was obtained from 29 TCGA histologically normal lung tissue samples with HM450
methylation and RNA sequencing data. The average promoter methylation was taken by an
average of methylation at CpG 3 (cg10564498), CpG 4 (cg03247049) and CpG 5 (cg12722469).
Spearman correlation coefficient and p-value are shown.
103

CHAPTER 4: APPLYING A DISTRIBUTED LAG MODEL TO THE ASSOCIATION
BETWEEN PRENATAL AIR POLLUTION AND NEWBORN DNA METHYLATION

4.1 ABSTRACT
Particulate matter (PM) exposure has been associated with adverse health effects. Epigenetic
modifications such as DNA methylation may be important mechanisms in prenatal programming
and are sensitive to environmental stressors. Few studies have performed time-series analyses of
the effects of prenatal air pollution exposure on DNA methylation pattern in newborns. To
identify differentially-methylated genes in response to prenatal air pollution exposure, we
investigated the effects of PM10 exposure during the 3-month preconception and 9-month
pregnancy window on newborn DNA methylation (Illumina 450K) in subjects from the
Children’s Health Study (n=248). The polynomial distributed lag models (DLM) were used to
examine how the association between methylation and lagged exposure changes across each
month while controlling for collinearity. The results were also compared with individual lag
models which fit separate models for each lag. The PM10-associated CpG loci were further
evaluated for association with childhood cardiovascular and respiratory outcomes. We identified
13 PM10-associated loci using FDR-corrected p-values of less than 0.1. Examination of the
lagged structures of PM10 effects suggested the importance of the 3-month preconception
window. The results from DLM were generally consistent with individual lag models while
reducing the number of parameters tested. These results indicated that PM10 exposure during
preconception and pregnancy was associated with differential offspring DNA methylation in
several metabolism-related genes.

104
4.2 INTRODUCTION
Air pollutants are produced from many sources and have been linked to a spectrum of
adverse health effects. Exposure to both ambient air pollutants and near-roadway traffic-related
pollutants have been reported to associate with cardiovascular disease [207], airway systemic
inflammation [239, 306], development and/or exacerbation of asthma [96, 307], and deficits in
lung function growth [154]. Particularly, exposure to air pollution in the prenatal period, which is
a critical window in programming susceptibility to childhood and adult disease, is shown to
increase risks for adverse intrauterine conditions, adult cardiovascular disease and incidence of
childhood asthma in animal or human studies [308-310]. Particulate matter (PM) is a
heterogeneous mixture of solid particles and liquid droplets existing in the air and contains up to
hundreds of different inorganic and organic chemicals [209]. The particle component of PM can
penetrate through pulmonary endothelium and travel to circulation, further influencing a variety
of cardiovascular and respiratory health outcomes [207, 210, 311].
Epigenetic modifications, especially DNA methylation, are known to influence gene
expression and have been identified as one possible mechanism by which air pollution may exert
its adverse effects [312]. Exposure to PM or components of PM have been linked to alterations
of methylation in both Alu and LINE-1, which can be used estimate global methylation levels,
and other candidate genes involved in asthma, inflammation, and oxidative stress [156, 212, 213,
313]. However, most of these studies were conducted in adults. Only a few studies have
investigated the effects of prenatal air pollution on epigenetic changes in newborns, which could
serve as molecular signatures underlying the early life origins of chronic diseases [314]. The
prenatal/fetal period is also considered a highly sensitive window to epigenetic disruptions due to
the dramatic programming and remodeling occurring in the methylome during this time [217,
105
218]. Evaluating the epigenetic marks associated with prenatal air pollution helps to identify the
biological mechanisms underlying the adverse effects of air pollution and the potential
biomarkers for environmentally-related diseases. The existing studies investigating prenatal air
pollution and newborn methylation patterns only utilized average exposure of either the
individual trimesters or the whole pregnancy [226, 227, 315], but rarely explored the time series
analyses of prenatal air pollution to estimate the contribution of exposure at each month to the
outcome. Furthermore, few studies have evaluated the effects air pollution exposure during the
preconception period on differential methylation, which has also been suggested as a vulnerable
time window [316, 317].
Time series studies investigate the association between time-varying exposure with short-
term changes in acute health outcomes, and have been widely applied to air pollution studies.
The time series studies enable us to examine the same population repeatedly under various
exposure conditions [318]. The most common regression model choices for time series analyses
include generalized linear models (GLM) with parametric splines (e.g., natural cubic splines) or
generalized additive models (GAM) with nonparametric smoothers (e.g., smoothing splines or
lowess smoothers) [319].
The time series study design also offers advantages to estimate delayed effects using the
defined temporal structure of the data. The assumption is made that health effects associated with
exposure are spread out over time. The effects measured at a given time can be described as the
result of multiple exposure events of different intensities sustained in the past. This situation
occurs frequently when assessing the short-term effects with peak or chronic exposures to
environmental stressors, or occupational exposures to carcinogenic substances [320-322]. For
example, several time series studies have reported the health effects of high levels of pollution or
106
extreme temperatures persisting some days after exposure occurrence [323, 324]. The modeling
strategy to describe the delayed effects of exposure is called distributed lag model (DLM), where
multiple lags of pollution are included in the model simultaneously [325]. Allowing the effect of
a single exposure event to be distributed over a specific period of time, this methodology uses
several parameters to explain the contributions of different lags and provides a comprehensive
picture of the time-course of the exposure-response relationship [326].
DLM has been widely used in environmental epidemiologic studies, mostly focusing on the
effects of air pollution or temperature on mortality. Martins et al. investigated the daily levels of
air pollutants over 20 days in Brazil and found the increase in risk of cardiovascular disease in
people aged 64 years and older was highest at lag 0 [327]. In the CHS, several types of lag-based
models (unconstrained DLM, polynomial DLM, splines-based DLM, and cumulative lag models
(CLM)) were considered when modeling the association between ambient air pollution on
exhaled nitric oxide (FeNO), and short-term increases in PM2.5, PM10 and O3 were found to
associate with airway inflammation [328]. Only a few studies have applied DLM to study the
lagged effects of air pollution on DNA methylation. In the Normative Aging Study, the authors
used DLM to examine the intermediate-term associations between methylation at candidate
genes with air pollutants, temperature, or relative humidity [329, 330]. Janssen et al. investigated
the effects of prenatal PM2.5 exposure on placental global DNA methylation using a multi-lag
model with exposure of all three trimesters fitted as independent variables in the same model,
and found only PM2.5 exposure during trimester 1 was significantly associated with lower global
DNA methylation [315]. To date, no study has investigated the lagged structure of prenatal air
pollution by each month and how this would affect the methylome at birth to identify susceptible
genes and regions.
107
In this study, we investigated the effects of PM10 exposure during the 3-month
preconception and 9-month pregnancy window on DNA methylation profiles in newborns using
the Infinium HumanMethylation450 BeadChip (HM450) in a subset of the Children’s Health
Study. We used the polynomial distributed lag modeling approach to examine how the
association between methylation and lagged exposure changes across each month while
controlling multicollinearity. We also compared results with individual lag models that fit
separate models for exposure at each month. The CpG loci associated with PM10 was then
evaluated for association with childhood cardiovascular and respiratory outcomes, including
asthma and related symptoms, carotid intima-media thickness (CIMT), systolic and diastolic
blood pressure, and body mass index (BMI).

4.3 MATERIALS AND METHODS
Study population
This study was conducted in subsets of participants in the Children’s Health Study, a
longitudinal study of respiratory health of children in southern California [6, 147, 251]. A subset
of 737 children were initially sampled to participate in a study of atherosclerosis [252] of whom
689 could be linked to California birth records. Of these, we randomly selected 273 children
from participants for whom at least 700 ng of DNA was available from a dried newborn
bloodspot. DNA methylation was assessed in the newborn bloodspots using HM450 arrays.
Written questionnaires were completed by the children’s parents at study entry and were
updated annually throughout the study thereafter. The questions asked information about
personal, parental, socio-demographic and medical history of the subjects, including the child’s
age, sex, race, Hispanic ethnic origin, history of asthma and related symptoms, level of parental
108
education and parental smoking during pregnancy. Information on date of birth, gestational age,
birth weight and maternal age at delivery was obtained from California birth certificates.
Admixture was assessed using the program STRUCTURE from a set of ancestral
informative markers that were scaled to represent the proportion of African American, Asian,
Native American and white admixture [266].
DNA methylation
DNA methylation was measured in newborn bloodspots (NBS) as reported previously [20].
The HM450 data was compiled for each locus and was expressed as beta (β) values. Minfi
package (version 1.16.0) in R was used to process the HM450 array data [254], applying a
normal exponential background correction to the raw intensities to reduce array-level
background noise followed by dye-bias correction [255]. We then normalized each sample’s
methylation values to the same quantiles to address sample to sample variability [256]. Seven
cord blood cell sub-populations (CD8+ T-lymphocytes, CD4+ T-lymphocytes, natural killer
cells, B-lymphocytes, monocytes, granulocytes and nucleated red blood cells) were estimated
using regression calibration approach algorithm described by Bakulski et al. [258]. After
preprocessing, CpG loci on the HM450 array were removed from analysis if they were on sex
chromosomes, or if they contained SNPs, deletions, repeats, or if they had more than 10%
missing values, leaving 384,310 probes for analysis. The M values were also calculated based on
beta values using 𝑙𝑜𝑔 2
𝑏𝑒𝑡𝑎 1−𝑏𝑒𝑡𝑎 . Outlier DNA methylation values were identified as values that
were either greater than the median + 5*SD or less than the median – 5*SD and were removed
from analyses.
Residential air pollution assessment
109
Residential addresses were obtained from birth certificate and the questionnaire completed
by parents at study entry. Street-level residential addresses of participants were standardized and
batch geocoded at the parcel level and match codes were output using the Texas A&M Geocoder
(http://geoservices.tamu.edu/Services/Geocode/). Addresses that did not match to a parcel
centroid (e.g., address range interpolation or zip/city/state centroid) were corrected based on the
best available knowledge of the participant’s residence location. Sonoma Technology Inc.
provided monthly air pollution exposure data for up to 12 calendar months prior to the birth date
obtained from California birth records. Briefly, hourly air quality data from ambient monitoring
stations were downloaded from the U.S. Environmental Protection Agency’s Air Quality System
(AQS, http://www.epa.gov/ttn/airs/airsaqs) for the relevant time period and averaged to daily
level unless otherwise noted. PM data were restricted to using Federal Reference Method (FRM)
continuous monitors or Federal Equivalent Method (FEM) monitors. Monthly exposure values
were spatially interpolated from the air quality monitoring station’s locations to the finest
geographic resolution possible (usually parcel-level) based on the participant’s geocoded street-
level birth residence using inverse distance-squared weighting. Data from up to four stations
were included in each interpolation. Due to the regional nature of PM10 in Southern CA, a
maximum interpolation radius of 50 km was used for all pollutants. When a residence was
located within 5 km of one or more stations with valid observations, the interpolation was based
solely on the concentrations from those stations. When multiple addresses were reported, the
longest residence was used for exposure assessment. Prenatal and preconception air pollution
assignments were successfully made for 271 of the 273 participants.
Health measurements
110
Asthma and related symptoms were assessed using an annual follow-up questionnaire
completed the children’s parents, as previously described [78]. The asthma symptoms under
investigation in this study included the following outcomes evaluated at age 11 years: a) asthma
(defined by a “yes” answer to the question “Has a doctor ever diagnosed this child as having
asthma?”); b) wheeze (defined by a “yes” answer to the question “Has your child’s chest ever
sounded wheezy or whistling?”); c) wheeze in the previous 12 months; d) bronchitic symptoms
in the previous 12 months (defined by the parent’s report of a daily cough for 3 months in a row,
congestion of phlegm other than when accompanied by a cold, or bronchitis).
The procedures for measuring CIMT, blood pressure, weight and height were described
previously [252]. All subjects had systolic/diastolic blood pressure, supine heart rate, standing
height and weight measured during a classroom visit, and B-mode carotid artery ultrasound was
performed to assess CIMT. CIMT, heart rate, and blood pressure were assessed by a single
physician-imaging specialist from the USC Atherosclerosis Research Unit (ARU) Core Imaging
and Reading Center (CIRC). As described previously [331, 332], the jugular vein and carotid
artery were imaged transversely with the jugular vein stacked above the carotid artery. All
images contained internal anatomical landmarks for reproducing probe angulation and a single-
lead electrocardiogram was recorded simultaneously with the B-mode image to ensure that
CIMT was measured at the R-wave in the cardiac cycle. CIMT was measured along the far
(deep) wall of the distal common carotid artery (0.25 cm from the carotid artery bulb) along a
standard 1 cm length that was automatically determined by a computer-generated ruler. This
method standardizes the timing, location, and distance over which CIMT is measured, ensuring
comparability across participants [331, 332]. Duplicate scans were conducted 2.5 days apart on
average for CIMT (n=44) and the intra-class correlation between replicate scans was 0.84.
111
Immediately after the IMT scan, blood pressure and heart rate were measured by standard
techniques after the subject was recumbent for at least 10 minutes. The subject’s standing height
and weight were also measured as described previously and body mass index (BMI) was
calculated using weight (kg) divided by square of height (m) [252].
Statistical methods
Descriptive analyses were performed to examine the distribution of subject characteristics.
Pearson correlation coefficients were calculated to evaluate the correlation structure of PM10. We
tested the association between methylation at birth and PM10 of the preceding 12 months (lag 0
to 11 months) using distributed lag models (DLM) and explored how the association changed
across lags. To explore the lag structure, the unconstrained distributed lag model can be written
as
E(𝑌 𝑡 ) = 𝛼 + covariates + 𝛽 0
𝑍 0
+ 𝛽 1
𝑍 1
+ … + 𝛽 𝑗 𝑍 𝑗 + 𝜀 𝑡 (1)
where 𝑌 𝑡 is the methylation level at birth at a specific CpG locus, j is the lag number and 𝑍 𝑗 is the
PM10 level at j+1 months prior to birth. Despite giving an unbiased estimate of the overall effect,
this model is too noisy to provide accurate estimates of individual lag effects due to collinearity.
Therefore, placing a constraint on the coefficients to allow smooth variation with lag number has
been applied to estimate the shape of curve more efficiently [325]. The polynomial distributed
lag model, which constrains the 𝛽 𝑗 to follow a polynomial pattern with the lag number, is the
most common approach [325]; that is,
𝛽 𝑗 = ∑ 𝛾 𝑘 𝑗 𝑘 𝑑 𝑘 =0
, for j=0…11 (2)
where d is the degree of the polynomial.
In this analysis, we investigated the lag structure of PM10 effects by combining two
functions that define the conventional exposure-response relationship and the additional lag-
112
response relationship, respectively. Specifically, for the PM10-methylation relationship we fitted
both beta regression models of methylation beta values, and general linear regression (GLR) of
M values, adjusted for sex, methylation plate, cord blood cell types, admixture, parental smoking
during pregnancy and parental education. These two methods were used to address the non-
normal distribution of DNA methylation values, which are bounded by 0 and 1 and in many
cases heavily skewed toward one end or the other [333]. 248 subjects with complete information
on PM10 and all adjusted covariates were included in the analyses. The confounding effects from
ethnicity, city of residence at study entry, season of birth and gestational age were evaluated with
additional sensitivity analyses and were found to be minimal (results not shown). For the lag-
response relationship, a third-degree polynomial function was used, which imposes constraints
and gives enough flexibility to estimate a biologically plausible lag structure, and controlling for
multi-collinearity. These two functions for exposure-response and lag-response relationships
were combined through a cross-basis parameterization, defining 4 cross-basis variables with
associated parameters (𝛾 𝑘 ’s) that represent the whole exposure-lag-response surface.
We took a two-step modeling approach to assess the association between PM10 exposure
during the 12 months prior to birth and methylation patterns at birth. In the first step, Equation 1
and Equation 2 were combined and a separate distributed lag model was fitted at each of the
384,310 CpG sites to generate the estimators of 𝛾 𝑘 ’s associated with the 4 cross-basis variables.
A likelihood-ratio test for testing the null hypothesis that 𝛾 1
= 𝛾 2
= 𝛾 3
= 𝛾 4
= 0 was used to
determine the association between PM10 exposure and methylation at each CpG site by
comparing the model with and without the 4 cross-basis variables. All regression analyses were
adjusted for multiple testing at a false discovery rate (FDR) of 0.1, using the method of
Benjamini and Hochberg [267]. In the second step, for the overlapping CpGs at FDR of 0.1 from
113
both beta regression of methylation beta values and GLR of M values, defined in this study as
PM10-associated loci, the coefficients 𝛽 ̂
𝑗 by lag were obtained from Equation 2 using the
estimated 𝛾 𝑘 ’s. We then plotted the 𝛽 ̂
𝑗 ’s by each month to visualize the lag structure of PM10
effects at each PM10-associated locus.
For the PM10-associated loci, we also compared results from DLM with the individual lag
models, which fit a regression model of methylation values with PM10 at each month separately,
adjusted for the same covariates as described above.
We then annotated the nearest gene for each PM10-associated CpG using the UCSC Genome
Browser in the FDb.InfiniumMethylation.hg19 R package. We also performed Gene Ontology
(GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses using the
missMethyl R package that corrects for the number of probes per gene on the HM450 array.
Finally, we used linear or logistic regression models to evaluate the association between
methylation of the PM10-associated loci and childhood cardiovascular and respiratory outcomes
including CIMT, blood pressure, BMI, asthma and related symptoms, adjusting for age, sex,
ethnicity, parental smoking during pregnancy, methylation plate and parental education level.
BMI was additionally adjusted for outcomes of CIMT, blood pressure, asthma and related
symptoms.
All testes assumed a two-sided alternative hypothesis. All analyses were conducted using the
R programming language, version 3.3.2. The dlnm R package was used to fit the polynomial
distributed lag model and the betareg R package was used to fit beta regression models.

4.4 RESULTS
Characteristics of the study participants
114
Demographic characteristics of the 248 study subjects are shown in Table 4.1. There were
fewer males than females and more than half of the participants were of Hispanic ethnicity. 18%
of the subjects were exposed to tobacco smoking during pregnancy from either mother or father.
Subjects averaged 11 years of age at the time of health outcomes assessment. Distribution of
PM10 at each month during the 1-year study period was shown in Table 4.1 and Figure 4.1. The
mean 24-hour average PM10 ranged from 41 to 46 𝜇 g/m
3
during the study period. Pearson
correlations between PM10 at each month were calculated (Table 4.2). PM10 at months close
together were moderately to highly correlated. PM10 during preconception and the last trimester
of pregnancy were also relatively highly correlated because they were in similar season.
DNA methylation at birth in association with PM10 exposure
We used third-degree polynomial distributed lag models to evaluate the association between
prenatal PM10 exposure and DNA methylation at birth of the 384,310 CpG loci across the
genome. We identified 13 overlapping CpG associated with PM10 with FDR-adjusted P-value
less than 0.1, using both beta regression of methylation beta values and GLR of M values for the
exposure-response relationship (Table 4.3). We then calculated the lag-specific coefficients
associated with a 2SD change in PM10 at each month using only beta regression of methylation
beta values, and plotted the coefficients against time (Figure 4.2-left panel). The lag structures
were highly consistent using beta regression of beta values and GLR of M values (results not
shown). The results indicated temporal variation across lags in the association between PM10
exposure and DNA methylation (Figure 4.2-left panel). For example, after taking account of the
lag structure, higher preconception PM10 exposure was significantly associated with lower
cg00589488 methylation at birth. However, for cg01303372, higher preconception PM10
exposure was significantly associated with increased methylation, while exposure during the first
115
trimester was associated with decreased methylation. The complete list of lag-specific estimates
associated a 2SD change in PM10 from beta regression for each of the 13 CpG loci was shown in
Table 4.4.
For each of the 12 months of PM10 exposure under investigation, we then evaluated how
many of the 13 PM10-associated loci had significant lag-specific effects (P-value < 0.05) (Table
4.5). Preconception month 3 and 2 showed the most significant lag-specific results (11 CpG loci
for both). The second trimester (pregnancy month 4-6) also showed a fair number of significant
lag-specific results.
We also compared results from DLM (Figure 4.2-left panel) of the 13 PM10-associated loci
with individual lag models (Figure 4.2-right panel) which fit a separate model for each lag using
beta regression with the same adjusted covariates. The two models produced generally consistent
results, with more similar patterns for the preconception window, but not for the last trimester of
pregnancy. The magnitudes of lag-specific estimates from the two models were also close.
Functional enrichment
We performed functional enrichment analyses for the PM10-associated CpG sites. Using the
missMethyl R package that corrects for the number of probes per gene on the HM450 array, there
were no significant enrichment for GO terms or KEGG terms at FDR 0.05. However, the top GO
and KEGG terms with nominally significant P-values suggested enrichment for pathways
associated with fructose and mannose metabolism, amino sugar and nucleotide sugar
metabolism, and apoptosis signaling pathway (Table 4.6).
Association with health outcomes
Finally, we related methylation level at birth of the 13 PM10-associated CpG sites with
cardiovascular and respiratory outcomes at age 11 (Table 4.7 and 4.8). Only very few of the 13
116
loci was associated with health outcomes. A 1% increase in methylation level at cg00589488,
which was negatively associated with prenatal PM10 exposure for most of the time periods under
investigation, was associated with a 0.25 kg/m
2
decrease in BMI at 11 years of age. On the other
hand, methylation of cg15972506 was positively associated with PM10 exposure during the
preconception period and negatively associated in the second trimester, was positively associated
with SBP. A 1% higher methylation level at cg25586305 at birth was significantly associated
with a 23% increase in the odds of ever wheezing and a 42% increase in the odds of recent
wheezing episodes at age 11 years. However, none of these relationships remained significant
after adjusting for multiple testing.

4.5 DISCUSSION
In this study, we applied the polynomial distributed lag model to evaluate the association
between ambient PM10 exposure during the 1-year time prior to birth and newborn DNA
methylation patterns. We identified 13 CpG loci significantly associated with prenatal PM10
exposure at FDR 0.1. The lag structure of PM10 effects on methylation was further investigated
for these loci and we found many of the 13 CpG loci were showing significant associations with
PM10 exposure during the 3-month preconception window.
The preconception and in utero period is considered a particularly sensitive window to
epigenetic alterations because this is the time when DNA synthesis rate is high and epigenetic
marks undergo critical modifications. Initial epigenetic reprogramming occurs during
gametogenesis, followed by genome-wide demethylation after fertilization, and DNA
methylation levels are then restored by de novo genome-wide methylation that triggers cell
lineage differentiation, which can be largely maintained but can also be altered by environmental
117
exposures before and/or after birth [217, 218]. There has been a growing body of evidence
supporting the role of environmental exposures during preconception, a previously understudied
critical exposure window, on both maternal and neonatal outcomes [317, 334], yet few studies
have evaluated the mechanism behind these effects.
To our knowledge, this is the first study exploring preconception and in utero air pollution
exposure simultaneously to identify CpG loci associated with PM10 across the epigenome. In
fact, based on our exploration for the lag structure of PM10 effects on the methylation level of the
associated loci, many of the 13 CpG loci showed significant associations with PM10 exposure
during the 3-month preconception window. Although we cannot draw this conclusion for the
whole methylome, these results suggest the potential epigenetic modification role of
preconception environmental exposure [335]. Possible mechanisms include the regulation of
PM10 exposure through histone alterations for which a small amount are retained in the fetus and
may carry information onto subsequent generations to affect DNA methylation [336]. Future
research is warranted to further investigate the influences of preconception exposures on the
offspring through epigenetic changes, and how this is related to disease phenotypes.
Given the developmental plasticity of the prenatal period, studies have investigated the
effects of air pollution exposure and newborn methylation patterns, both genome-wide and on
target genes. In a genome-wide meta-analysis of newborn methylation in relation to prenatal
nitrogen dioxide (NO2) air pollution exposure averaged during the whole pregnancy, the authors
identified 3 epigenome-wide significant CpG sites in mitochondria-related genes [227]. We
previously evaluated the effects of trimester-specific average PM10 and PM2.5 exposure on DNA
methylation of 178,309 promoter regions in the same population of this study, and identified 31
associated loci using FDR-corrected P-values of less than 0.15 [226]. However, none of the top
118
13 PM10-associated CpG loci in the current results overlapped with the previously-reported 31
loci, and possible reasons include different preprocessing methods for methylation data,
improved methods for estimating PM10 and different modeling approaches. Of the 13 PM10-
associated loci, cg26253663 methylation was previously reported to associate with mid-term (28
days) PM2.5 exposure in a genome-wide analysis in an elderly population [337], while
cg01303372 methylation in newborns was identified as responsive to maternal smoking during
pregnancy in a meta-analysis [224].
We used the distributed lag model to investigate the contribution of exposure at each month
to the methylation pattern at birth, while controlling for multiple collinearity. Although this
modeling approach has been utilized to investigate time-varying environmental exposure on
DNA methylation at targeted genes [329, 330], it has not been applied to genome-wide analysis
to identify responsive CpG loci. In this study, we applied the third-degree polynomial DLM on
all 384,310 CpG sites and used likelihood-ratio test to determine global significance of the 4
cross-basis variables, which greatly reduced the number of parameters tested compared to the
non-restricted DLM. After identifying the PM10-associated loci, we obtained the lag-specific
estimates and also compared them with those obtained from the individual lag models, which did
not adjust for exposure at other months. The consistent patterns for preconception window
further adds credibility to the importance of this period. However, we observed effect estimates
of opposite directions for the last trimester using DLM and individual lag models, suggesting the
confounding effects of exposure during previous months in estimating this association. At some
CpG sites (e.g., cg01303372 and cg01472961), the lag structure from DLM showed dramatic
change from preconception to first trimester, possibly reflecting the dynamic effects of exposure
on the time near conception, when the whole epigenome undergoes demethylation and de novo
119
methylation [218]. Results from this paper suggests that applying DLM helps to reduce the
number of parameters tested while accounting for the collinearity between time-varying
exposures. Future research is needed for optimal application of this method to genome-wide
analyses.
The CpG loci associated with prenatal PM10 exposure were also evaluated for association
with cardiovascular and respiratory outcomes at age 11 years. However, we only found a small
number of CpG loci associated with health outcomes and none of the associated genes have been
previously reported in cardiovascular or asthma-related phenotypes. Since enrichment analyses
of the PM10-associated loci suggested involvement in metabolism pathways, future research may
evaluate the associations with more metabolism-relevant outcomes with larger sample size and
possibly at an earlier age.
The strengths of this study include the temporal separation of PM10 exposure assessment and
measurement of DNA methylation, and for the first time studying the effects of preconception air
pollution exposure on the alterations in newborn methylome. Several limitations should be noted.
First of all, we evaluated newborn DNA methylation patterns in blood spot, which is a mixture of
cell types. Although we adjusted for this by adding the estimated cord blood cell types into the
model, the small changes in methylation observed may still be related to shifts in cell populations
of smaller subtypes [338], and we could be missing important associations in subtypes of
immune cells, or in other tissues of interest. Secondly, we used a relatively loose FDR threshold
(0.1) to determine epigenome-wide significant results due to a small sample size and exploratory
analysis of applying DLM to the HM450 array. Although we made some effort to correct for this
by only investigating the overlapping loci from both beta regression and GLR, it is still possible
that we obtained false positive results. Thirdly, there could be measurement error of PM10. We
120
only measured ambient PM10 using the major residential address during pregnancy. Since we
don’t have information on how much time the mother spent outdoors near home and the
exposure level on secondary addresses if subjects moved, it is likely that the measured PM10
level may not reflect the true exposed level. Future studies with personal air pollution monitors
are needed to address this issue.
In conclusion, we identified several CpG loci associated with preconception and pregnancy
PM10 exposure using the polynomial distributed lag model to account for correlation of exposure
across each month. Results suggested the importance of preconception exposure on methylation
at associated CpG sites.

121
4.6 TABLES AND FIGURES
Table 4.1. Demographic characteristics of study participants (N=248)
N %
Male Sex 103 41.5%
Ethnicity

Hispanic 138 55.6%
Non-Hispanic White 81 32.7%
Asian/Black/Other 29 11.7%
Parental education

High school or less 83 33.5%
Some college 87 35.1%
Finshed college or some graduate training 78 31.5%
In-utero tobacco smoke from either parent 44 17.7%
Age at health outcomes assessment, years (mean, sd) 11.2 0.6
CIMT, µm (mean, sd) 562.8 41.3
SBP, mmHg (mean, sd) 105.2 8.7
DBP, mmHg (mean, sd) 57.1 5.6
BMI, kg/m
2
(mean, sd) 19.9 3.9
PM10 at preconception month 3, µg/m
3
(mean, sd) 41.4 15.6
PM10 at preconception month 2, µg/m
3
(mean, sd) 41.5 15.1
PM10 at preconception month 1, µg/m
3
(mean, sd) 41.5 15.6
PM10 at pregnancy month 1, µg/m
3
(mean, sd) 42.2 14.8
PM10 at pregnancy month 2, µg/m
3
(mean, sd) 43.6 16.6
PM10 at pregnancy month 3, µg/m
3
(mean, sd) 45.2 17.7
PM10 at pregnancy month 4, µg/m
3
(mean, sd) 46.6 17.0
PM10 at pregnancy month 5, µg/m
3
(mean, sd) 46.0 17.5
PM10 at pregnancy month 6, µg/m
3
(mean, sd) 45.3 16.9
PM10 at pregnancy month 7, µg/m
3
(mean, sd) 44.4 16.8
PM10 at pregnancy month 8, µg/m
3
(mean, sd) 43.8 16.6
PM10 at pregnancy month 9, µg/m
3
(mean, sd) 43.7 16.5
CIMT: carotid intima-media thickness; BMI: body mass index; SBP: systolic
blood pressure; DBP: diastolic blood pressure; BMI: body mass index.
Numbers do not always add up to 100% due to missing data
122

Table 4.2. Pearson correlation between PM10 levels measured at each month

Preconcepti
on month 3
Preconcepti
on month 2
Preconcepti
on month 1
Pregnan
cy
month 1
Pregnan
cy
month 2
Pregnan
cy
month 3
Pregnan
cy
month 4
Pregnan
cy
month 5
Pregnan
cy
month 6
Pregnan
cy
month 7
Pregnan
cy
month 8
Pregnan
cy
month 9
Preconcepti
on month 3
1.00 0.72* 0.49* 0.18* 0.06 -0.07 0.02 0.11 0.22* 0.49* 0.65* 0.67*
Preconcepti
on month 2
1.00 0.77* 0.53* 0.32* 0.13* 0.02 0.04 0.03 0.19* 0.42* 0.48*
Preconcepti
on month 1
1.00 0.77* 0.59* 0.36* 0.19* 0.12 0.02 0.06 0.16* 0.34*
Pregnancy
month 1
1.00 0.82* 0.62* 0.40* 0.23* 0.03 -0.01 -0.04 0.08
Pregnancy
month 2
1.00 0.85* 0.64* 0.45* 0.15* 0.003 -0.08 0.02
Pregnancy
month 3
1.00 0.81* 0.60* 0.32* 0.08 -0.09 -0.03
Pregnancy
month 4
1.00 0.78* 0.57* 0.34* 0.10 0.11
Pregnancy
month 5
1.00 0.75* 0.53* 0.28* 0.15*
Pregnancy
month 6
1.00 0.78* 0.54* 0.33*
Pregnancy
month 7
1.00 0.76* 0.59*
Pregnancy
month 8
1.00 0.75*
Pregnancy
month 9
1.00
* p-value < 0.05
123
Table 4.3. Association between prenatal PM10 exposure and DNA methylation at birth
Probe Chr Position
Nearest
gene
Genomic location
Mean
methylation
Beta regression of beta
values*
General linear
regression of M
values*
P-value
FDR-
adjusted
P-value
P-value
FDR-
adjusted
P-value
cg00589488 3 49760249 GMPPB Gene body 0.82 4.31E-07 0.08 4.61E-07 0.06
cg01303372 10 121415189 BAG3 Gene body 0.13 1.85E-06 0.09 2.05E-06 0.07
cg01472961 9 136215657 RPL7A Gene body 0.06 1.88E-06 0.09 2.77E-06 0.07
cg04760604 16 22436179 RRN3P3 Gene body 0.80 1.21E-06 0.09 1.37E-06 0.07
cg15972506 3 11079298 SLC6A1 3' UTR 0.25 3.48E-07 0.08 5.02E-07 0.06
cg16311339 3 180397336 CCDC39 Upstream of TSS 0.07 3.41E-06 0.09 3.70E-06 0.08
cg18569695 8 143433909 TSNARE1 Gene body 0.85 2.66E-06 0.09 2.90E-06 0.07
cg21244580 8 79934476 IL7 Upstream of TSS 0.48 1.41E-06 0.09 1.48E-06 0.07
cg24851906 6 168613133 DACT2 3' UTR 0.85 1.80E-06 0.09 1.13E-06 0.07
cg25586305 2 234281690 DGKD Gene body 0.76 2.28E-06 0.09 2.23E-06 0.07
cg26253663 22 25799504 LRP5L Upstream of TSS 0.33 2.75E-06 0.09 4.13E-07 0.06
cg26374686 11 78357700 TENM4 3' UTR 0.17 7.95E-07 0.09 1.18E-06 0.07
cg26907544 1 6284915 ICMT 3' UTR 0.95 3.02E-06 0.09 2.90E-06 0.07
* Assessed by third-degree polynomial distributed lag models adjusted for sex, plate, cord blood cell types, admixture, parental
smoking during pregnancy and highest parental education level. P values were obtained from likelihood-ratio test testing the fit of the 4
cross-basis variables (see statistical methods for details).
FDR was used to adjust for tests performed at all 384,310 CpG loci.
124
Table 4.4. Association between a 2SD increase in prenatal PM10
exposure at each month and DNA methylation at birth assessed
by third-degree polynomial distributed lag models*
Probe
β** P-value Time
cg00589488
-0.039 0.04 Preconception month 3
cg00589488
-0.029 8.29E-05 Preconception month 2
cg00589488
-0.019 8.59E-05 Preconception month 1
cg00589488
-0.010 0.09 Pregnancy month 1
cg00589488
-0.001 0.78 Pregnancy month 2
cg00589488
0.005 0.21 Pregnancy month 3
cg00589488
0.009 0.03 Pregnancy month 4
cg00589488
0.011 0.07 Pregnancy month 5
cg00589488
0.008 0.18 Pregnancy month 6
cg00589488
0.002 0.71 Pregnancy month 7
cg00589488
-0.009 0.21 Pregnancy month 8
cg00589488
-0.026 0.17 Pregnancy month 9
cg01303372
0.103 9.64E-05 Preconception month 3
cg01303372
0.042 7.05E-05 Preconception month 2
cg01303372
0.003 0.64 Preconception month 1
cg01303372
-0.017 0.04 Pregnancy month 1
cg01303372
-0.023 0.002 Pregnancy month 2
cg01303372
-0.020 7.34E-04 Pregnancy month 3
cg01303372
-0.010 0.10 Pregnancy month 4
cg01303372
0.001 0.93 Pregnancy month 5
cg01303372
0.010 0.28 Pregnancy month 6
cg01303372
0.012 0.08 Pregnancy month 7
cg01303372
0.004 0.68 Pregnancy month 8
cg01303372
-0.018 0.51 Pregnancy month 9
cg01472961
0.107 2.13E-05 Preconception month 3
cg01472961
0.046 4.64E-06 Preconception month 2
cg01472961
0.009 0.16 Preconception month 1
cg01472961
-0.008 0.30 Pregnancy month 1
cg01472961
-0.012 0.11 Pregnancy month 2
cg01472961
-0.006 0.32 Pregnancy month 3
cg01472961
0.005 0.38 Pregnancy month 4
cg01472961
0.016 0.04 Pregnancy month 5
cg01472961
0.022 0.01 Pregnancy month 6
cg01472961
0.018 0.009 Pregnancy month 7
cg01472961
-0.001 0.92 Pregnancy month 8
cg01472961
-0.039 0.13 Pregnancy month 9
125
cg04760604
-0.067 2.47E-05 Preconception month 3
cg04760604
-0.029 3.09E-06 Preconception month 2
cg04760604
-0.006 0.15 Preconception month 1
cg04760604
0.006 0.21 Pregnancy month 1
cg04760604
0.010 0.03 Pregnancy month 2
cg04760604
0.008 0.02 Pregnancy month 3
cg04760604
0.003 0.50 Pregnancy month 4
cg04760604
-0.004 0.48 Pregnancy month 5
cg04760604
-0.008 0.15 Pregnancy month 6
cg04760604
-0.007 0.09 Pregnancy month 7
cg04760604
0.001 0.92 Pregnancy month 8
cg04760604
0.018 0.26 Pregnancy month 9
cg15972506
0.063 2.90E-04 Preconception month 3
cg15972506
0.035 4.43E-07 Preconception month 2
cg15972506
0.015 0.001 Preconception month 1
cg15972506
0.001 0.87 Pregnancy month 1
cg15972506
-0.008 0.13 Pregnancy month 2
cg15972506
-0.012 0.002 Pregnancy month 3
cg15972506
-0.014 9.22E-04 Pregnancy month 4
cg15972506
-0.014 0.01 Pregnancy month 5
cg15972506
-0.013 0.02 Pregnancy month 6
cg15972506
-0.014 0.003 Pregnancy month 7
cg15972506
-0.016 0.02 Pregnancy month 8
cg15972506
-0.022 0.22 Pregnancy month 9
cg16311339
0.036 0.30 Preconception month 3
cg16311339
0.000 0.998 Preconception month 2
cg16311339
-0.023 0.01 Preconception month 1
cg16311339
-0.036 0.001 Pregnancy month 1
cg16311339
-0.040 5.85E-05 Pregnancy month 2
cg16311339
-0.039 5.97E-07 Pregnancy month 3
cg16311339
-0.034 6.17E-05 Pregnancy month 4
cg16311339
-0.027 0.01 Pregnancy month 5
cg16311339
-0.021 0.08 Pregnancy month 6
cg16311339
-0.018 0.05 Pregnancy month 7
cg16311339
-0.020 0.15 Pregnancy month 8
cg16311339
-0.029 0.41 Pregnancy month 9
cg18569695
-0.047 0.02 Preconception month 3
cg18569695
-0.009 0.26 Preconception month 2
cg18569695
0.011 0.03 Preconception month 1
cg18569695
0.018 0.005 Pregnancy month 1
126
cg18569695
0.014 0.01 Pregnancy month 2
cg18569695
0.004 0.37 Pregnancy month 3
cg18569695
-0.009 0.05 Pregnancy month 4
cg18569695
-0.021 6.21E-04 Pregnancy month 5
cg18569695
-0.029 1.53E-05 Pregnancy month 6
cg18569695
-0.028 9.25E-08 Pregnancy month 7
cg18569695
-0.015 0.05 Pregnancy month 8
cg18569695
0.013 0.51 Pregnancy month 9
cg21244580
0.127 6.31E-06 Preconception month 3
cg21244580
0.041 2.22E-04 Preconception month 2
cg21244580
-0.008 0.30 Preconception month 1
cg21244580
-0.028 0.002 Pregnancy month 1
cg21244580
-0.027 7.44E-04 Pregnancy month 2
cg21244580
-0.013 0.04 Pregnancy month 3
cg21244580
0.008 0.23 Pregnancy month 4
cg21244580
0.027 0.002 Pregnancy month 5
cg21244580
0.037 1.09E-04 Pregnancy month 6
cg21244580
0.030 7.76E-05 Pregnancy month 7
cg21244580
-0.002 0.87 Pregnancy month 8
cg21244580
-0.065 0.02 Pregnancy month 9
cg24851906
-0.115 2.23E-06 Preconception month 3
cg24851906
-0.045 2.85E-06 Preconception month 2
cg24851906
-0.006 0.39 Preconception month 1
cg24851906
0.011 0.15 Pregnancy month 1
cg24851906
0.011 0.11 Pregnancy month 2
cg24851906
0.001 0.92 Pregnancy month 3
cg24851906
-0.014 0.01 Pregnancy month 4
cg24851906
-0.026 5.09E-04 Pregnancy month 5
cg24851906
-0.030 2.21E-04 Pregnancy month 6
cg24851906
-0.019 0.004 Pregnancy month 7
cg24851906
0.014 0.15 Pregnancy month 8
cg24851906
0.074 0.003 Pregnancy month 9
cg25586305
-0.063 3.72E-06 Preconception month 3
cg25586305
-0.024 1.45E-05 Preconception month 2
cg25586305
-0.002 0.64 Preconception month 1
cg25586305
0.007 0.13 Pregnancy month 1
cg25586305
0.005 0.19 Pregnancy month 2
cg25586305
-0.002 0.48 Pregnancy month 3
cg25586305
-0.011 4.18E-04 Pregnancy month 4
cg25586305
-0.018 1.56E-05 Pregnancy month 5
127
cg25586305
-0.020 1.81E-05 Pregnancy month 6
cg25586305
-0.011 0.002 Pregnancy month 7
cg25586305
0.011 0.04 Pregnancy month 8
cg25586305
0.051 2.37E-04 Pregnancy month 9
cg26253663
-0.061 0.01 Preconception month 3
cg26253663
-0.038 4.76E-05 Preconception month 2
cg26253663
-0.020 0.001 Preconception month 1
cg26253663
-0.006 0.39 Pregnancy month 1
cg26253663
0.003 0.65 Pregnancy month 2
cg26253663
0.009 0.08 Pregnancy month 3
cg26253663
0.011 0.04 Pregnancy month 4
cg26253663
0.011 0.14 Pregnancy month 5
cg26253663
0.008 0.33 Pregnancy month 6
cg26253663
0.002 0.72 Pregnancy month 7
cg26253663
-0.005 0.58 Pregnancy month 8
cg26253663
-0.014 0.56 Pregnancy month 9
cg26374686
0.034 0.23 Preconception month 3
cg26374686
0.026 0.02 Preconception month 2
cg26374686
0.016 0.03 Preconception month 1
cg26374686
0.006 0.53 Pregnancy month 1
cg26374686
-0.004 0.62 Pregnancy month 2
cg26374686
-0.012 0.07 Pregnancy month 3
cg26374686
-0.016 0.02 Pregnancy month 4
cg26374686
-0.014 0.12 Pregnancy month 5
cg26374686
-0.007 0.50 Pregnancy month 6
cg26374686
0.009 0.23 Pregnancy month 7
cg26374686
0.035 0.002 Pregnancy month 8
cg26374686
0.071 0.02 Pregnancy month 9
cg26907544
-0.111 2.82E-06 Preconception month 3
cg26907544
-0.043 6.42E-06 Preconception month 2
cg26907544
-0.004 0.54 Preconception month 1
cg26907544
0.012 0.10 Pregnancy month 1
cg26907544
0.012 0.08 Pregnancy month 2
cg26907544
0.001 0.83 Pregnancy month 3
cg26907544
-0.014 0.01 Pregnancy month 4
cg26907544
-0.026 4.11E-04 Pregnancy month 5
cg26907544
-0.030 1.46E-04 Pregnancy month 6
cg26907544
-0.019 0.002 Pregnancy month 7
cg26907544
0.012 0.19 Pregnancy month 8
cg26907544
0.072 0.003 Pregnancy month 9
128
* Adjusted for sex, plate, cord blood cell types, admixture, parental
smoking during pregnancy, and highest parental education level (plots
are shown in Figure 4.1). Exposure-response relationship was modeled
using beta regression of methylation beta values. SD for PM 10 at each
month was shown in Table 4.1.
** Coefficients from beta regression modeling on the logit scale

129
Table 4.5. Number of PM10-associated CpGs having
significant lag-specific results (P-value < 0.05) for each
month*
Time
Number of CpGs with
significant lag-specific results
Preconception month 3 11
Preconception month 2 11
Preconception month 1 6
Pregnancy month 1 4
Pregnancy month 2 5
Pregnancy month 3 5
Pregnancy month 4 8
Pregnancy month 5 8
Pregnancy month 6 7
Pregnancy month 7 7
Pregnancy month 8 3
Pregnancy month 9 5
* Assessed by third-degree polynomial distributed lag
models adjusted for sex, plate, cord blood cell types,
admixture, parental smoking during pregnancy and
highest parental education level (plots are shown in
Figure 4.2). Exposure-response relationship was modeled
using beta regression of methylation beta values.

130

Table 4.6. Top 10 enriched Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) terms for
PM10-associated CpGs.
GO ID GO term
Occurences
in genome
Occurences
in sample
P value for
overrepresentation
FDR
value
GO:0004671
protein C-terminal S-isoprenylcysteine
carboxyl O-methyltransferase activity
1 1 3.56E-04 1.00
GO:0004475
mannose-1-phosphate guanylyltransferase
activity
1 1 6.74E-04 1.00
GO:0008905
mannose-phosphate guanylyltransferase
activity
1 1 6.74E-04 1.00
GO:0051939 gamma-aminobutyric acid import 1 1 6.75E-04 1.00
GO:0006481 C-terminal protein methylation 2 1 0.001 1.00
GO:0003880
protein C-terminal carboxyl O-
methyltransferase activity
2 1 0.001 1.00
GO:0005332
gamma-aminobutyric acid:sodium symporter
activity
4 1 0.002 1.00
GO:0035469
determination of pancreatic left/right
asymmetry
4 1 0.002 1.00
GO:0000774 adenyl-nucleotide exchange factor activity 5 1 0.002 1.00
GO:0003356 regulation of cilium beat frequency 5 1 0.002 1.00
KEGG ID KEGG Pathway
Occurences
in genome
Occurences
in sample
P value for
overrepresentation
FDR
value
hsa00900 Terpenoid backbone biosynthesis 22 1 0.003 0.65
hsa00051 Fructose and mannose metabolism 32 1 0.006 0.65
hsa00520 Amino sugar and nucleotide sugar metabolism 47 1 0.008 0.65
hsa00561 Glycerolipid metabolism 57 1 0.01 0.65
131

hsa04727 GABAergic synapse 83 1 0.02 0.65
hsa00564 Glycerophospholipid metabolism 92 1 0.02 0.65
hsa01100 Metabolic pathways 1196 2 0.02 0.65
hsa05231 Choline metabolism in cancer 95 1 0.02 0.65
hsa04070 Phosphatidylinositol signaling system 94 1 0.02 0.65
hsa03010 Ribosome 125 1 0.02 0.70
132
Table 4.7. Association between DNA methylation at birth and cardiovascular health outcomes in childhood*
Probe Gene
CIMT SBP DBP BMI
β P-value β P-value β P-value β P-value
cg00589488
GMPPB
-0.03 0.98

0.14 0.56

-0.17 0.31

-0.25 0.03
cg01303372
BAG3
0.02 0.99

0.20 0.35

0.22 0.14

0.03 0.76
cg01472961
RPL7A
-2.58 0.23

0.06 0.88

-0.19 0.51

0.04 0.83
cg04760604
RRN3P3
1.39 0.28

-0.25 0.31

-0.17 0.32

-0.16 0.16
cg15972506
SLC6A1
-0.24 0.82

0.48 0.02

0.23 0.11

0.02 0.84
cg16311339
CCDC39
-0.47 0.76

-0.37 0.22

0.10 0.62

0.17 0.24
cg18569695
TSNARE1
-0.56 0.68

-0.21 0.43

-0.09 0.63

0.01 0.93
cg21244580
IL7
0.16 0.75

0.04 0.69

0.04 0.55

0.07 0.14
cg24851906
DACT2
0.01 0.99

-0.20 0.37

0.03 0.82

0.03 0.76
cg25586305
DGKD
0.24 0.88

-0.08 0.79

-0.20 0.33

0.07 0.59
cg26253663
LRP5L
0.07 0.87

0.14 0.09

0.04 0.50

0.01 0.77
cg26374686
TENM4
-0.10 0.90

-0.02 0.88

0.06 0.60

0.03 0.69
cg26907544
ICMT
-2.75 0.32 0.04 0.93 0.74 0.04 0.11 0.65
* Linear regression model evaluating the change in CIMT (µm), SBP (mmHg), DBP (mmHg) and BMI (kg/m
2
)
per 1% increase in DNA methylation. Models were adjusted for age, sex, ethnicity, parental smoking during
pregnancy, plate, highest parental education and additionally adjusted for BMI for outcomes of CIMT, SBP and
DBP.
CIMT: carotid intima-media thickness; BMI: body mass index; SBP: systolic blood pressure; DBP: diastolic
blood pressure; BMI: body mass index
Bold values are statistically significant (P-value < 0.05).

133

Table 4.8. Association between DNA methylation at birth and asthma and related symptoms in childhood*
Probe Gene
Ever MD-
diagnosed asthma
Ever wheezing
Wheezing in the
previous 12
months

Bronchitic
symptoms in the
previous 12 months
OR P-value OR P-value OR P-value OR P-value
cg00589488 GMPPB 1.10 0.36 1.07 0.41 1.04 0.76 1.14 0.23
cg01303372 BAG3 0.96 0.59 0.98 0.80 1.01 0.95 0.92 0.34
cg01472961 RPL7A 0.69 0.07 0.91 0.49 1.03 0.86 1.14 0.35
cg04760604 RRN3P3 1.28 0.02 1.17 0.06 0.96 0.74 1.03 0.74
cg15972506 SLC6A1 1.10 0.22 1.03 0.62 1.15 0.11 1.06 0.44
cg16311339 CCDC39 0.86 0.23 1.09 0.35 0.99 0.95 1.42 0.003
cg18569695 TSNARE1 1.07 0.47 0.96 0.59 0.93 0.49 1.17 0.14
cg21244580 IL7 0.95 0.20 0.95 0.11 0.98 0.70 0.91 0.01
cg24851906 DACT2 1.06 0.47 1.02 0.75 1.00 0.99 1.03 0.71
cg25586305 DGKD 1.23 0.09 1.23 0.03 1.42 0.02 1.08 0.50
cg26253663 LRP5L 0.97 0.28 1.00 0.88 0.97 0.48 0.99 0.81
cg26374686 TENM4 0.89 0.10 0.87 0.01 0.94 0.37 0.92 0.17
cg26907544 ICMT 1.03 0.89 0.96 0.80 1.46 0.18 1.58 0.06
* Logistic regression model evaluating the odds ratio for risk of asthma and related symptoms per 1% increase in DNA
methylation. Models were adjusted for age, sex, ethnicity, parental smoking during pregnancy, plate, highest parental
education and BMI.
Bold values are statistically significant (P-value < 0.05).
134

Figure 4.1. Distribution of PM10 exposure for each month under investigation.

-3 to -1: preconception month 3 to 1; 1 to 9: pregnancy month 1 to 9.

135
Figure 4.2. Association between a 2SD increase in prenatal PM10 exposure at each month and
DNA methylation

136
Figure 4.2. Continued. Association between a 2SD increase in prenatal PM10 exposure at each
month and DNA methylation.

137
Figure 4.2. Continued. Association between a 2SD increase in prenatal PM10 exposure at each
month and DNA methylation.

138

Figure 4.2. Continued. Association between a 2SD increase in prenatal PM10 exposure at each
month and DNA methylation.

Left panel: lag structures assessed by third-degree polynomial distributed lag models.
Right panel: assessed by individual lag models for PM10 at each month (12 models).
-3 to -1: preconception month 3 to 1; 1 to 9: pregnancy month 1 to 9.
Dots show the coefficients obtained from beta regression of methylation beta values, and the
vertical lines and whiskers are the 95% confidence intervals.

139

CHAPTER 5: SUMMARY AND FUTURE RESEARCH

5.1 SUMMARY AND CONCLUSION
The goal of this dissertation was to first explore the epigenetic regulation of a candidate
gene, AXL, in childhood asthma pathogenesis and lung function development. Epigenetic
modifications have been associated with the risk of asthma and related phenotypes [110, 119]
and lung function development [16, 17], yet studies are needed to address the effects of DNA
methylation at birth to identify early origins of diseases. Therefore, I investigated the epigenetic
regulation of AXL – a gene playing an essential role in regulating effective phagocytosis and
innate immune responses – in childhood asthma and lung development. Given the dynamic
nature of DNA methylation, this dissertation also aimed to study how prenatal air pollution
exposure affects DNA methylation patterns at birth. The distributed lag modeling approach was
utilized to perform the time series analyses of PM10 exposure during the 1-year time prior to
birth.
In Chapter 2, we reported that higher average methylation of CpGs in AXL at birth was
associated with higher risk for parent-reported wheezing at age 6 years. The association was
stronger in girls than in boys and reflected the methylation status of the gene-body region near
the 5’ end. The association with asthma and related symptoms was replicated using
Pyrosequencing for one CpG locus, cg00360107. Furthermore, AXL DNA methylation was
strongly linked to underlying genetic polymorphisms.
In Chapter 3, we observed higher methylation in AXL promoter region at birth measured by
Pyrosequencing was associated with lower lung function growth from 10 to 18 years of age. We
also reported similar association for the growth from 11 to 15 years of age using methylation
140
data measured by HM450 array. One CpG locus in the promoter region, cg10564498, was
significantly associated with decreased lung function growth from 10 to 18 years of age and the
negative associations were observed in a similar age range in the replication population.
Taken together, these findings suggest the role of AXL methylation in childhood asthma and
related symptoms, and lung function growth during adolescence. The presented findings are
consistent with AXL’s importance in inhibiting uncontrolled inflammatory responses [164], and
regulating effective phagocytosis in both human and mouse airway and maintaining lung
immune homeostasis [203, 204]. AXL has been extensively studied in various cancers [198] and
was recently reported as responsive to prenatal tobacco smoke exposure [161, 163]. To date, the
presented findings are the first human studies reporting the epigenetic regulation of AXL in
childhood respiratory health. Interestingly, we observed sex-specific effects in both studies, the
mechanism behind which is unclear but warrants further investigation. To functionally translate
the observed methylation differences, we also related AXL methylation to mRNA expression in
lung using paired data from the TCGA database. Furthermore, we performed a comprehensive
evaluation of the genetic polymorphisms in AXL in relation to methylation and provided
evidence for the contribution of local genomic environment to establishing epigenetic marks of
this gene. The findings from this dissertation support the previously-unrecognized role of AXL in
regulating airway inflammatory responses and pulmonary function development, and also
provide evidence for accounting for genetic variation when investigating the epigenetic
regulation of this gene. These results also contribute to unraveling the role of epigenetic
mechanism in programming susceptibility to asthma and related phenotypes, and setting up early
origins for lung function development.
141
In Chapter 4, we studied the effects of preconception and pregnancy PM10 exposure
simultaneously to identify associated loci across the whole epigenome for the first time. To fully
utilize the time series nature of air pollution data while accounting for collinearity, we used the
polynomial distributed lag model and identified 13 PM10-associated loci using FDR-corrected P-
values of less than 0.1. Further examination of the lagged structure of PM10 effects suggested the
importance of preconception exposure, a previously-understudied critical window. This is the
first study applying distributed lag model genome-widely to study the effects of prenatal air
pollution. Our results illustrated this modeling approach reduced the number of parameters tested
compared to the unrestricted model, while taking into account the correlation between time-
varying exposures. There is also some evidence showing the enrichment of metabolism-related
pathways for the PM10-associated loci. These findings shed lights on the necessity to include
preconception window when studying the effects of air pollution.

5.2 IMPLICATIONS AND FUTURE DIRECTIONS
Given the novelty of our findings, more studies are needed to see if our results are replicable
in other population settings, such as populations with different demographic distributions and of
other age groups. Additionally, there are several suggestions for future research. First of all, our
investigation of AXL methylation was only conducted in bloodspot, which is a mixture of cell
types. To better understand the development of asthma and lung function, airway or lung tissue
samples will be more pathologically relevant, but are much more difficult to obtain for children.
Nasal epithelial cells are a particularly attractive source of airway epithelial cells because of the
greater ease of access compared to the bronchial epithelium and the potential for repeated
isolation from the same individual [339]. Studies have evaluated the use of nasal epithelial cells
142
as surrogates for lower airway cells and suggested that physiologically, nasal epithelial cells
constitute an accessible surrogate for bronchial epithelial cells to facilitate the study of lower
airway inflammation in hitherto largely inaccessible populations including children [339, 340].
Thus, future studies may explore the epigenetic profiles of AXL in nasal epithelial cells from
children and the regulation of childhood respiratory health in a pathologically more relevant
tissue.
Besides, due to the low expression level of AXL in blood, we studied the correlation between
methylation and mRNA using TCGA samples obtained from normal lung tissues of lung cancer
subjects, a majority of whom were moderate to heavy smokers. Therefore, future studies are
needed to evaluate the methylation pattern of AXL across somatic tissues to demonstrate if these
epigenetic marks carry over from blood to other tissues. Several studies exploring the DNA
methylation profiles across various human tissues and conditions are limited by the use of human
autopsy specimens from diseased subjects, or small sample size of peripheral tissue samples
from normal individuals [42, 341, 342]. To overcome these issues, one may characterize the
persistence of AXL DNA methylation pattern across cell types in primary cell lines using
publicly-available databases such as the ENCODE Project Consortium [343]. Studies in samples
from healthier and younger subjects are also needed to assess in which tissues the methylation
correlates with expression. The ongoing eGTEx project seeks to complete the gene expression
phenotypes determined in the GTEx project with diverse molecular assays across the same
tissues and individuals, including DNA methylation, ChIP-seq, protein expression, somatic
mutation, etc. Such effort will enable future investigators to evaluate epigenetic regulation of
AXL in multiple tissues at high resolution [344].
143
In Chapters 2 and 3, we observed association between AXL methylation with both childhood
asthma and lung development. Although we hypothesized that AXL may participate in asthma
pathogenesis through inhibiting inflammatory responses in the airway, and contribute to lung
homeostasis via regulating effective phagocytosis, it is unclear whether these processes involve
different pathways or both. To elucidate the underlying mechanisms, researchers may start from
evaluating the activation of AXL and related genes in each pathway in mouse asthma models, and
mouse models with inflamed airway and/or lung [203]. For example, we may measure the
expression of downstream genes induced by AXL in anti-inflammation pathways including
SOCS1, SOCS3 and/or the nuclear factor- B repressor, Twist [164, 199]. Similarly,
investigators have explored the coregulation of AXL with factors that mediate the recognition and
engulfment of apoptotic cells to determine whether the AXL-mediated apoptosis pathway is
activated [345]. Therefore, future animal- or human- based designs are needed to clarify the
molecular mechanism underlying AXL’s epigenetic regulation of childhood asthma and lung
development, and the mechanism behind the sex-specific effects.
Since the three TAM receptors are often co-expressed in multiple tissues and cell types
[164], and AXL and its major ligand GAS6 are functionally connected, studying the role of all
TAM family genes in respiratory health outcomes comprehensively will benefit our
understanding of this pathway, and the identification of new drug targets for treating respiratory
disease. Future studies may evaluate the comprehensive epigenetic profile of all three TAM
receptors and their ligands in relation to respiratory diseases, including DNA methylation,
genetic polymorphisms, mRNA and miRNA expression, and soluble protein forms [346].
Furthermore, since the TAM family genes are involved in multiple pathways including inhibition
of uncontrolled inflammation signaling by TLR and cytokine receptors, apoptosis pathways and
144
the development of natural killer cell, investigators may look more closely at the relationship
between TAM genes and these pathways. For example, we may study the epigenetic regulation
of TAM genes and Toll-like receptors in the pathogenesis of childhood asthma at the same time,
such as TLR3, TLR4, TLR9 and multiple points in TLR signal transduction cascades, including
MAPK, ERK1/ERK2, NF-𝜅 B, TNF, IL6 and type I interferons [164]. Findings from such studies
will benefit our understanding for the role of the TAM genes in autoimmune diseases and
inflammation pathways. Lastly, methylation of AXL was only measured at birth in this
dissertation. Thus, longitudinal studies evaluating methylation at multiple time points during
childhood and adolescence will help us understand whether the epigenetic marks observed at
birth persist into later life.
The findings in Chapter 4 lends preliminary support for the importance of preconception
exposure in environmental epigenetics studies. We also explored the application of distributed
lag model to genome-wide analysis. More research is needed to further evaluate the biological
role of the identified PM10-associated loci, and whether epigenetic changes in these genes are
mediating the adverse health effects of air pollution. Future studies are warranted to incorporate
preconception exposure when investigating the effects of prenatal environmental factors on
health outcomes, and elucidate the mechanism behind which preconception exposure alters
epigenetic patterns in the offspring. Furthermore, it will be interesting to expand the distributed
lag model to incorporate multiple pollutants and study their lagged effects simultaneously.

145
REFERENCES
1. Moorman JE, Akinbami LJ, Bailey CM, Zahran HS, King ME, Johnson CA, et al:
National surveillance of asthma: United States, 2001-2010. Vital Health Stat 3 2012:1-
58.
2. Landrigan PJ, Schechter CB, Lipton JM, Fahs MC, Schwartz J: Environmental pollutants
and disease in American children: estimates of morbidity, mortality, and costs for lead
poisoning, asthma, cancer, and developmental disabilities. Environ Health Perspect 2002,
110:721-728.
3. Canonica GW: Treating asthma as an inflammatory disease. Chest 2006, 130:21s-28s.
4. Lim RH, Kobzik L, Dahl M: Risk for asthma in offspring of asthmatic mothers versus
fathers: a meta-analysis. PLoS One 2010, 5:e10134.
5. Duijts L, Jaddoe VW, van der Valk RJ, Henderson JA, Hofman A, Raat H, et al: Fetal
exposure to maternal and paternal smoking and the risks of wheezing in preschool
children: the Generation R Study. Chest 2012, 141:876-885.
6. McConnell R, Berhane K, Yao L, Jerrett M, Lurmann F, Gilliland F, et al: Traffic,
susceptibility, and childhood asthma. Environ Health Perspect 2006, 114:766-772.
7. Islam T, Berhane K, McConnell R, Gauderman WJ, Avol E, Peters JM, et al:
Glutathione-S-transferase (GST) P1, GSTM1, exercise, ozone and asthma incidence in
school children. Thorax 2009, 64:197-202.
8. Martino D, Prescott S: Epigenetics and prenatal influences on asthma and allergic
airways disease. Chest 2011, 139:640-647.
9. Ozanne SE, Constancia M: Mechanisms of disease: the developmental origins of disease
and the role of the epigenotype. Nat Clin Pract Endocrinol Metab 2007, 3:539-546.
10. Knuiman MW, James AL, Divitini ML, Ryan G, Bartholomew HC, Musk AW: Lung
function, respiratory symptoms, and mortality: results from the Busselton Health Study.
Ann Epidemiol 1999, 9:297-306.
11. Sobol BJ, Herbert WH, Emirgil C: The high incidence of pulmonary functional
abnormalities in patients with coronary artery disease. Chest 1974, 65:148-151.
12. Islam T, Gauderman WJ, Berhane K, McConnell R, Avol E, Peters JM, et al:
Relationship between air pollution, lung function and asthma in adolescents. Thorax
2007, 62:957-963.
13. Redline S, Tishler PV, Rosner B, Lewitter FI, Vandenburgh M, Weiss ST, et al:
Genotypic and phenotypic similarities in pulmonary function among family members of
adult monozygotic and dizygotic twins. Am J Epidemiol 1989, 129:827-836.
14. in't Veen JC, Smits HH, Hiemstra PS, Zwinderman AE, Sterk PJ, Bel EH: Lung function
and sputum characteristics of patients with severe asthma during an induced exacerbation
by double-blind steroid withdrawal. Am J Respir Crit Care Med 1999, 160:93-99.
15. Sandberg KL, Pinkerton KE, Poole SD, Minton PA, Sundell HW: Fetal nicotine exposure
increases airway responsiveness and alters airway wall composition in young lambs.
Respir Physiol Neurobiol 2011, 176:57-67.
16. Lepeule J, Baccarelli A, Motta V, Cantone L, Litonjua AA, Sparrow D, et al: Gene
promoter methylation is associated with lung function in the elderly: the Normative
Aging Study. Epigenetics 2012, 7:261-269.
146
17. Patil VK, Holloway JW, Zhang H, Soto-Ramirez N, Ewart S, Arshad SH, et al:
Interaction of prenatal maternal smoking, interleukin 13 genetic variants and DNA
methylation influencing airflow and airway reactivity. Clin Epigenetics 2013, 5:22.
18. Bind MA, Baccarelli A, Zanobetti A, Tarantini L, Suh H, Vokonas P, et al: Air pollution
and markers of coagulation, inflammation, and endothelial function: associations and
epigene-environment interactions in an elderly cohort. Epidemiology 2012, 23:332-340.
19. Lemke G: Biology of the TAM receptors. Cold Spring Harb Perspect Biol 2013,
5:a009076.
20. Gao L, Millstein J, Siegmund KD, Dubeau L, Maguire R, Gilliland FD, et al: Epigenetic
regulation of AXL and risk of childhood asthma symptoms. Clinical Epigenetics 2017,
9:121.
21. Robertson KD: DNA methylation and human disease. Nat Rev Genet 2005, 6:597-610.
22. Esteller M: CpG island hypermethylation and tumor suppressor genes: a booming
present, a brighter future. Oncogene 2002, 21:5427-5440.
23. Rodenhiser D, Mann M: Epigenetics and human disease: translating basic biology into
clinical applications. Cmaj 2006, 174:341-348.
24. Feinberg AP, Tycko B: The history of cancer epigenetics. Nat Rev Cancer 2004, 4:143-
153.
25. Jurgens B, Schmitz-Drager BJ, Schulz WA: Hypomethylation of L1 LINE sequences
prevailing in human urothelial carcinoma. Cancer Res 1996, 56:5698-5703.
26. Santourlidis S, Florl A, Ackermann R, Wirtz HC, Schulz WA: High frequency of
alterations in DNA methylation in adenocarcinoma of the prostate. Prostate 1999,
39:166-174.
27. Suter CM, Martin DI, Ward RL: Hypomethylation of L1 retrotransposons in colorectal
cancer and adjacent normal tissue. Int J Colorectal Dis 2004, 19:95-101.
28. Widschwendter M, Jiang G, Woods C, Muller HM, Fiegl H, Goebel G, et al: DNA
hypomethylation and ovarian cancer biology. Cancer Res 2004, 64:4472-4480.
29. De Smet C, Lurquin C, Lethe B, Martelange V, Boon T: DNA methylation is the primary
silencing mechanism for a set of germ line- and tumor-specific genes with a CpG-rich
promoter. Mol Cell Biol 1999, 19:7327-7335.
30. Gopalakrishnan S, Van Emburgh BO, Robertson KD: DNA methylation in development
and human disease. Mutat Res 2008, 647:30-38.
31. Esteller M: Epigenetic gene silencing in cancer: the DNA hypermethylome. Hum Mol
Genet 2007, 16 Spec No 1:R50-59.
32. Li M, Chen WD, Papadopoulos N, Goodman SN, Bjerregaard NC, Laurberg S, et al:
Sensitive digital quantification of DNA methylation in clinical samples. Nat Biotechnol
2009, 27:858-863.
33. Egger G, Liang G, Aparicio A, Jones PA: Epigenetics in human disease and prospects for
epigenetic therapy. Nature 2004, 429:457-463.
34. Issa JP, Kantarjian HM, Kirkpatrick P: Azacitidine. Nat Rev Drug Discov 2005, 4:275-
276.
35. Shen L, Kondo Y, Ahmed S, Boumber Y, Konishi K, Guo Y, et al: Drug sensitivity
prediction by CpG island methylation profile in the NCI-60 cancer cell line panel. Cancer
Res 2007, 67:11335-11343.
36. Tang Q, Cheng J, Cao X, Surowy H, Burwinkel B: Blood-based DNA methylation as
biomarker for breast cancer: a systematic review. Clin Epigenetics 2016, 8:115.
147
37. Kato T, Iwamoto K, Kakiuchi C, Kuratomi G, Okazaki Y: Genetic or epigenetic
difference causing discordance between monozygotic twins as a clue to molecular basis
of mental disorders. Mol Psychiatry 2005, 10:622-630.
38. Skuse DH, James RS, Bishop DV, Coppin B, Dalton P, Aamodt-Leeper G, et al:
Evidence from Turner's syndrome of an imprinted X-linked locus affecting cognitive
function. Nature 1997, 387:705-708.
39. Richardson B: DNA methylation and autoimmune disease. Clin Immunol 2003, 109:72-
79.
40. Ziller MJ, Gu H, Muller F, Donaghey J, Tsai LT, Kohlbacher O, et al: Charting a
dynamic DNA methylation landscape of the human genome. Nature 2013, 500:477-481.
41. Deaton AM, Bird A: CpG islands and the regulation of transcription. Genes Dev 2011,
25:1010-1022.
42. Lokk K, Modhukur V, Rajashekar B, Martens K, Magi R, Kolde R, et al: DNA
methylome profiling of human tissues identifies global and tissue-specific methylation
patterns. Genome Biol 2014, 15:r54.
43. Attwood JT, Yung RL, Richardson BC: DNA methylation and the regulation of gene
transcription. Cell Mol Life Sci 2002, 59:241-257.
44. Jones PA: Functions of DNA methylation: islands, start sites, gene bodies and beyond.
Nat Rev Genet 2012, 13:484-492.
45. Ball MP, Li JB, Gao Y, Lee JH, LeProust EM, Park IH, et al: Targeted and genome-scale
strategies reveal gene-body methylation signatures in human cells. Nat Biotechnol 2009,
27:361-368.
46. Leonard SM, Wei W, Collins SI, Pereira M, Diyaf A, Constandinou-Williams C, et al:
Oncogenic human papillomavirus imposes an instructive pattern of DNA methylation
changes which parallel the natural history of cervical HPV infection in young women.
Carcinogenesis 2012, 33:1286-1293.
47. Liu L, Wylie RC, Andrews LG, Tollefsbol TO: Aging, cancer and nutrition: the DNA
methylation connection. Mech Ageing Dev 2003, 124:989-998.
48. Kerkel K, Spadola A, Yuan E, Kosek J, Jiang L, Hod E, et al: Genomic surveys by
methylation-sensitive SNP analysis identify sequence-dependent allele-specific DNA
methylation. Nat Genet 2008, 40:904-908.
49. Shoemaker R, Deng J, Wang W, Zhang K: Allele-specific methylation is prevalent and is
contributed by CpG-SNPs in the human genome. Genome Res 2010, 20:883-889.
50. Shi J, Marconett CN, Duan J, Hyland PL, Li P, Wang Z, et al: Characterizing the genetic
basis of methylome diversity in histologically normal human lung tissue. Nat Commun
2014, 5:3365.
51. Voisin S, Almen MS, Zheleznyakova GY, Lundberg L, Zarei S, Castillo S, et al: Many
obesity-associated SNPs strongly associate with DNA methylation changes at proximal
promoters and enhancers. Genome Med 2015, 7:103.
52. Gutierrez-Arcelus M, Lappalainen T, Montgomery SB, Buil A, Ongen H, Yurovsky A, et
al: Passive and active DNA methylation and the interplay with genetic variation in gene
regulation. Elife 2013, 2:e00523.
53. Schubeler D: Function and information content of DNA methylation. Nature 2015,
517:321-326.
148
54. Bell JT, Pai AA, Pickrell JK, Gaffney DJ, Pique-Regi R, Degner JF, et al: DNA
methylation patterns associate with genetic and gene expression variation in HapMap cell
lines. Genome Biol 2011, 12:R10.
55. Gibbs JR, van der Brug MP, Hernandez DG, Traynor BJ, Nalls MA, Lai SL, et al:
Abundant quantitative trait loci exist for DNA methylation and gene expression in human
brain. PLoS Genet 2010, 6:e1000952.
56. Zhang D, Cheng L, Badner JA, Chen C, Chen Q, Luo W, et al: Genetic control of
individual differences in gene-specific methylation in human brain. Am J Hum Genet
2010, 86:411-419.
57. Yunginger JW, Reed CE, O'Connell EJ, Melton LJ, 3rd, O'Fallon WM, Silverstein MD:
A community-based study of the epidemiology of asthma. Incidence rates, 1964-1983.
Am Rev Respir Dis 1992, 146:888-894.
58. Expert Panel Report 3 (EPR-3): Guidelines for the Diagnosis and Management of
Asthma-Summary Report 2007. J Allergy Clin Immunol 2007, 120:S94-138.
59. Burr ML: Diagnosing asthma by questionnaire in epidemiological surveys. Clin Exp
Allergy 1992, 22:509-510.
60. Busse WW, Gern JE: Viruses in asthma. J Allergy Clin Immunol 1997, 100:147-150.
61. Platts-Mills TA, Wheatley LM: The role of allergy and atopy in asthma. Curr Opin Pulm
Med 1996, 2:29-34.
62. Wardlaw AJ: The role of air pollution in asthma. Clin Exp Allergy 1993, 23:81-96.
63. Vinay Kumar V, Cotran RS, Robbins SL: Acute and Chronic Inflammation. In Robbins’
Basic Pathology. Sixth edition: Philadelphia: W.B. Saunders Company; 1997: 25-46
64. van Aalderen WM: Childhood asthma: diagnosis and treatment. Scientifica (Cairo) 2012,
2012:674204.
65. Delacourt C: [Bronchial changes in untreated asthma]. Arch Pediatr 2004, 11 Suppl
2:71s-73s.
66. Sont JK, Willems LN, Bel EH, van Krieken JH, Vandenbroucke JP, Sterk PJ: Clinical
control and histopathologic outcome of asthma when using airway hyperresponsiveness
as an additional guide to long-term treatment. The AMPUL Study Group. Am J Respir
Crit Care Med 1999, 159:1043-1051.
67. Postma DS, Kerstjens HA: Characteristics of airway hyperresponsiveness in asthma and
chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1998, 158:S187-192.
68. Cockcroft DW, Davis BE: Mechanisms of airway hyperresponsiveness. J Allergy Clin
Immunol 2006, 118:551-559; quiz 560-551.
69. Cockcroft DW, Ruffin RE, Dolovich J, Hargreave FE: Allergen-induced increase in non-
allergic bronchial reactivity. Clin Allergy 1977, 7:503-513.
70. Homer RJ, Elias JA: Airway remodeling in asthma: therapeutic implications of
mechanisms. Physiology (Bethesda) 2005, 20:28-35.
71. Martinez FD, Wright AL, Taussig LM, Holberg CJ, Halonen M, Morgan WJ: Asthma
and wheezing in the first six years of life. The Group Health Medical Associates. N Engl
J Med 1995, 332:133-138.
72. Azevedo I, de Blic J, Scheinmann P, Vargaftig BB, Bachelet M: Enhanced arachidonic
acid metabolism in alveolar macrophages from wheezy infants. Modulation by
dexamethasone. Am J Respir Crit Care Med 1995, 152:1208-1214.
149
73. Krawiec ME, Westcott JY, Chu HW, Balzar S, Trudeau JB, Schwartz LB, et al:
Persistent wheezing in very young children is associated with lower respiratory
inflammation. Am J Respir Crit Care Med 2001, 163:1338-1343.
74. Goksor E, Amark M, Alm B, Gustafsson PM, Wennergren G: Asthma symptoms in early
childhood--what happens then? Acta Paediatr 2006, 95:471-478.
75. Guilbert T, Krawiec M: Natural history of asthma. Pediatr Clin North Am 2003, 50:523-
538.
76. Edwards CA, Osman LM, Godden DJ, Douglas JG: Wheezy bronchitis in childhood: a
distinct clinical entity with lifelong significance? Chest 2003, 124:18-24.
77. Taussig LM, Wright AL, Holberg CJ, Halonen M, Morgan WJ, Martinez FD: Tucson
Children's Respiratory Study: 1980 to present. J Allergy Clin Immunol 2003, 111:661-
675; quiz 676.
78. McConnell R, Berhane K, Gilliland F, Molitor J, Thomas D, Lurmann F, et al:
Prospective study of air pollution and bronchitic symptoms in children with asthma. Am J
Respir Crit Care Med 2003, 168:790-797.
79. Clarke JR, Jenkins MA, Hopper JL, Carlin JB, Mayne C, Clayton DG, et al: Evidence for
genetic associations between asthma, atopy, and bronchial hyperresponsiveness: a study
of 8- to 18-yr-old twins. Am J Respir Crit Care Med 2000, 162:2188-2193.
80. Malerba G, Pignatti PF: A review of asthma genetics: gene expression studies and recent
candidates. J Appl Genet 2005, 46:93-104.
81. Gao PS, Kawada H, Kasamatsu T, Mao XQ, Roberts MH, Miyamoto Y, et al: Variants of
NOS1, NOS2, and NOS3 genes in asthmatics. Biochem Biophys Res Commun 2000,
267:761-763.
82. Fryer AA, Bianco A, Hepple M, Jones PW, Strange RC, Spiteri MA: Polymorphism at
the glutathione S-transferase GSTP1 locus. A new marker for bronchial
hyperresponsiveness and asthma. Am J Respir Crit Care Med 2000, 161:1437-1442.
83. Heinzmann A, Mao XQ, Akaiwa M, Kreomer RT, Gao PS, Ohshima K, et al: Genetic
variants of IL-13 signalling and human asthma and atopy. Hum Mol Genet 2000, 9:549-
559.
84. Meyers DA, Postma DS, Stine OC, Koppelman GH, Ampleford EJ, Jongepier H, et al:
Genome screen for asthma and bronchial hyperresponsiveness: interactions with passive
smoke exposure. J Allergy Clin Immunol 2005, 115:1169-1175.
85. Romieu I, Ramirez-Aguilar M, Sienra-Monge JJ, Moreno-Macias H, del Rio-Navarro
BE, David G, et al: GSTM1 and GSTP1 and respiratory health in asthmatic children
exposed to ozone. Eur Respir J 2006, 28:953-959.
86. Gold DR, Wright R: Population disparities in asthma. Annu Rev Public Health 2005,
26:89-113.
87. Bjornson CL, Mitchell I: Gender differences in asthma in childhood and adolescence. J
Gend Specif Med 2000, 3:57-61.
88. Hancox RJ, Milne BJ, Poulton R, Taylor DR, Greene JM, McLachlan CR, et al: Sex
differences in the relation between body mass index and asthma and atopy in a birth
cohort. Am J Respir Crit Care Med 2005, 171:440-445.
89. Claudio L, Tulton L, Doucette J, Landrigan PJ: Socioeconomic factors and asthma
hospitalization rates in New York City. J Asthma 1999, 36:343-350.
90. Litonjua AA, Carey VJ, Weiss ST, Gold DR: Race, socioeconomic factors, and area of
residence are associated with asthma prevalence. Pediatr Pulmonol 1999, 28:394-401.
150
91. Devereux G, Turner SW, Craig LC, McNeill G, Martindale S, Harbour PJ, et al: Low
maternal vitamin E intake during pregnancy is associated with asthma in 5-year-old
children. Am J Respir Crit Care Med 2006, 174:499-507.
92. Romieu I, Torrent M, Garcia-Esteban R, Ferrer C, Ribas-Fito N, Anto JM, et al: Maternal
fish intake during pregnancy and atopy and asthma in infancy. Clin Exp Allergy 2007,
37:518-525.
93. Sherriff A, Farrow A, Golding J, Henderson J: Frequent use of chemical household
products is associated with persistent wheezing in pre-school age children. Thorax 2005,
60:45-49.
94. Sunyer J, Torrent M, Munoz-Ortiz L, Ribas-Fito N, Carrizo D, Grimalt J, et al: Prenatal
dichlorodiphenyldichloroethylene (DDE) and asthma in children. Environ Health
Perspect 2005, 113:1787-1790.
95. Goksor E, Amark M, Alm B, Gustafsson PM, Wennergren G: The impact of pre- and
post-natal smoke exposure on future asthma and bronchial hyper-responsiveness. Acta
Paediatr 2007, 96:1030-1035.
96. Gauderman WJ, Avol E, Lurmann F, Kuenzli N, Gilliland F, Peters J, et al: Childhood
asthma and exposure to traffic and nitrogen dioxide. Epidemiology 2005, 16:737-743.
97. Illi S, von Mutius E, Lau S, Niggemann B, Gruber C, Wahn U: Perennial allergen
sensitisation early in life and chronic asthma in children: a birth cohort study. Lancet
2006, 368:763-770.
98. Lau S, Illi S, Sommerfeld C, Niggemann B, Bergmann R, von Mutius E, et al: Early
exposure to house-dust mite and cat allergens and development of childhood asthma: a
cohort study. Multicentre Allergy Study Group. Lancet 2000, 356:1392-1397.
99. Burke H, Leonardi-Bee J, Hashim A, Pine-Abata H, Chen Y, Cook DG, et al: Prenatal
and passive smoke exposure and incidence of asthma and wheeze: systematic review and
meta-analysis. Pediatrics 2012, 129:735-744.
100. Jaakkola JJ, Gissler M: Maternal smoking in pregnancy, fetal development, and
childhood asthma. Am J Public Health 2004, 94:136-140.
101. Gilliland FD, Li YF, Dubeau L, Berhane K, Avol E, McConnell R, et al: Effects of
glutathione S-transferase M1, maternal smoking during pregnancy, and environmental
tobacco smoke on asthma and wheezing in children. Am J Respir Crit Care Med 2002,
166:457-463.
102. Nicolai T, Carr D, Weiland SK, Duhme H, von Ehrenstein O, Wagner C, et al: Urban
traffic and pollutant exposure related to respiratory outcomes and atopy in a large sample
of children. Eur Respir J 2003, 21:956-963.
103. Hulin M, Caillaud D, Annesi-Maesano I: Indoor air pollution and childhood asthma:
variations between urban and rural areas. Indoor Air 2010, 20:502-514.
104. Brauer M, Hoek G, Smit HA, de Jongste JC, Gerritsen J, Postma DS, et al: Air pollution
and development of asthma, allergy and infections in a birth cohort. Eur Respir J 2007,
29:879-888.
105. Gehring U, Cyrys J, Sedlmeir G, Brunekreef B, Bellander T, Fischer P, et al: Traffic-
related air pollution and respiratory health during the first 2 yrs of life. Eur Respir J 2002,
19:690-698.
106. Fedulov AV, Leme A, Yang Z, Dahl M, Lim R, Mariani TJ, et al: Pulmonary exposure to
particles during pregnancy causes increased neonatal asthma susceptibility. Am J Respir
Cell Mol Biol 2008, 38:57-67.
151
107. Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindorff LA, Hunter DJ, et al: Finding
the missing heritability of complex diseases. Nature 2009, 461:747-753.
108. Liu Y, Li H, Xiao T, Lu Q: Epigenetics in immune-mediated pulmonary diseases. Clin
Rev Allergy Immunol 2013, 45:314-330.
109. Cortessis VK, Thomas DC, Levine AJ, Breton CV, Mack TM, Siegmund KD, et al:
Environmental epigenetics: prospects for studying epigenetic mediation of exposure-
response relationships. Hum Genet 2012, 131:1565-1589.
110. Morales E, Bustamante M, Vilahur N, Escaramis G, Montfort M, de Cid R, et al: DNA
hypomethylation at ALOX12 is associated with persistent wheezing in childhood. Am J
Respir Crit Care Med 2012, 185:937-943.
111. Breton CV, Byun HM, Wang X, Salam MT, Siegmund K, Gilliland FD: DNA
methylation in the arginase-nitric oxide synthase pathway is associated with exhaled
nitric oxide in children with asthma. Am J Respir Crit Care Med 2011, 184:191-197.
112. Baccarelli A, Rusconi F, Bollati V, Catelan D, Accetta G, Hou L, et al: Nasal cell DNA
methylation, inflammation, lung function and wheezing in children with asthma.
Epigenomics 2012, 4:91-100.
113. Lee GR, Kim ST, Spilianakis CG, Fields PE, Flavell RA: T helper cell differentiation:
regulation by cis elements and epigenetics. Immunity 2006, 24:369-379.
114. Santangelo S, Cousins DJ, Winkelmann NE, Staynov DZ: DNA methylation changes at
human Th2 cytokine genes coincide with DNase I hypersensitive site formation during
CD4(+) T cell differentiation. J Immunol 2002, 169:1893-1903.
115. Yang Y, Haitchi HM, Cakebread J, Sammut D, Harvey A, Powell RM, et al: Epigenetic
mechanisms silence a disintegrin and metalloprotease 33 expression in bronchial
epithelial cells. J Allergy Clin Immunol 2008, 121:1393-1399, 1399.e1391-1314.
116. Kwon NH, Kim JS, Lee JY, Oh MJ, Choi DC: DNA methylation and the expression of
IL-4 and IFN-gamma promoter genes in patients with bronchial asthma. J Clin Immunol
2008, 28:139-146.
117. Murphy TM, Wong CC, Arseneault L, Burrage J, Macdonald R, Hannon E, et al:
Methylomic markers of persistent childhood asthma: a longitudinal study of asthma-
discordant monozygotic twins. Clin Epigenetics 2015, 7:130.
118. Rastogi D, Suzuki M, Greally JM: Differential epigenome-wide DNA methylation
patterns in childhood obesity-associated asthma. Sci Rep 2013, 3:2164.
119. Yang IV, Pedersen BS, Liu A, O'Connor GT, Teach SJ, Kattan M, et al: DNA
methylation and childhood asthma in the inner city. J Allergy Clin Immunol 2015,
136:69-80.
120. Wang G, Walker SO, Hong X, Bartell TR, Wang X: Epigenetics and early life origins of
chronic noncommunicable diseases. J Adolesc Health 2013, 52:S14-21.
121. Polgar G, Weng TR: The functional development of the respiratory system from the
period of gestation to adulthood. Am Rev Respir Dis 1979, 120:625-695.
122. Quanjer PH, Stanojevic S, Cole TJ, Baur X, Hall GL, Culver BH, et al: Multi-ethnic
reference values for spirometry for the 3-95-yr age range: the global lung function 2012
equations. Eur Respir J 2012, 40:1324-1343.
123. Wang X, Dockery DW, Wypij D, Fay ME, Ferris BG, Jr.: Pulmonary function between 6
and 18 years of age. Pediatr Pulmonol 1993, 15:75-88.
152
124. Tager IB, Segal MR, Speizer FE, Weiss ST: The natural history of forced expiratory
volumes. Effect of cigarette smoking and respiratory symptoms. Am Rev Respir Dis
1988, 138:837-849.
125. Berry CE, Billheimer D, Jenkins IC, Lu ZJ, Stern DA, Gerald LB, et al: A Distinct Low
Lung Function Trajectory from Childhood to the Fourth Decade of Life. Am J Respir Crit
Care Med 2016, 194:607-612.
126. Breton CV, Vora H, Salam MT, Islam T, Wenten M, Gauderman WJ, et al: Variation in
the GST mu locus and tobacco smoke exposure as determinants of childhood lung
function. Am J Respir Crit Care Med 2009, 179:601-607.
127. Gilliland FD, Gauderman WJ, Vora H, Rappaport E, Dubeau L: Effects of glutathione-S-
transferase M1, T1, and P1 on childhood lung function growth. Am J Respir Crit Care
Med 2002, 166:710-716.
128. Islam T, Breton C, Salam MT, McConnell R, Wenten M, Gauderman WJ, et al: Role of
inducible nitric oxide synthase in asthma risk and lung function growth during
adolescence. Thorax 2010, 65:139-145.
129. Kabesch M, Hoefler C, Carr D, Leupold W, Weiland SK, von Mutius E: Glutathione S
transferase deficiency and passive smoking increase childhood asthma. Thorax 2004,
59:569-573.
130. Wilk JB, Walter RE, Laramie JM, Gottlieb DJ, O'Connor GT: Framingham Heart Study
genome-wide association: results for pulmonary function measures. BMC Med Genet
2007, 8 Suppl 1:S8.
131. Hancock DB, Eijgelsheim M, Wilk JB, Gharib SA, Loehr LR, Marciante KD, et al: Meta-
analyses of genome-wide association studies identify multiple loci associated with
pulmonary function. Nat Genet 2010, 42:45-52.
132. Zeiger RS, Dawson C, Weiss S: Relationships between duration of asthma and asthma
severity among children in the Childhood Asthma Management Program (CAMP). J
Allergy Clin Immunol 1999, 103:376-387.
133. Berhane K, McConnell R, Gilliland F, Islam T, Gauderman WJ, Avol E, et al: Sex-
specific effects of asthma on pulmonary function in children. Am J Respir Crit Care Med
2000, 162:1723-1730.
134. Gilliland FD, Berhane K, Li YF, Rappaport EB, Peters JM: Effects of early onset asthma
and in utero exposure to maternal smoking on childhood lung function. Am J Respir Crit
Care Med 2003, 167:917-924.
135. Merkus PJ, Borsboom GJ, Van Pelt W, Schrader PC, Van Houwelingen HC, Kerrebijn
KF, et al: Growth of airways and air spaces in teenagers is related to sex but not to
symptoms. J Appl Physiol (1985) 1993, 75:2045-2053.
136. Xuan W, Peat JK, Toelle BG, Marks GB, Berry G, Woolcock AJ: Lung function growth
and its relation to airway hyperresponsiveness and recent wheeze. Results from a
longitudinal population study. Am J Respir Crit Care Med 2000, 161:1820-1824.
137. Wang X, Mensinga TT, Schouten JP, Rijcken B, Weiss ST: Determinants of maximally
attained level of pulmonary function. Am J Respir Crit Care Med 2004, 169:941-949.
138. Harmsen L, Ulrik CS, Porsbjerg C, Thomsen SF, Holst C, Backer V: Airway
hyperresponsiveness and development of lung function in adolescence and adulthood.
Respir Med 2014, 108:752-757.
153
139. Kotecha SJ, Watkins WJ, Heron J, Henderson J, Dunstan FD, Kotecha S: Spirometric
lung function in school-age children: effect of intrauterine growth retardation and catch-
up growth. Am J Respir Crit Care Med 2010, 181:969-974.
140. Lawlor DA, Ebrahim S, Davey Smith G: Association of birth weight with adult lung
function: findings from the British Women's Heart and Health Study and a meta-analysis.
Thorax 2005, 60:851-858.
141. Giliberti D, Mohan SS, Brown LA, Gauthier TW: Perinatal exposure to alcohol:
implications for lung development and disease. Paediatr Respir Rev 2013, 14:17-21.
142. Erkkola M, Kaila M, Nwaru BI, Kronberg-Kippila C, Ahonen S, Nevalainen J, et al:
Maternal vitamin D intake during pregnancy is inversely associated with asthma and
allergic rhinitis in 5-year-old children. Clin Exp Allergy 2009, 39:875-882.
143. Hayatbakhsh MR, Sadasivam S, Mamun AA, Najman JM, Williams GM, O'Callaghan
MJ: Maternal smoking during and after pregnancy and lung function in early adulthood: a
prospective study. Thorax 2009, 64:810-814.
144. Gilliland FD, Berhane K, McConnell R, Gauderman WJ, Vora H, Rappaport EB, et al:
Maternal smoking during pregnancy, environmental tobacco smoke exposure and
childhood lung function. Thorax 2000, 55:271-276.
145. Li YF, Gilliland FD, Berhane K, McConnell R, Gauderman WJ, Rappaport EB, et al:
Effects of in utero and environmental tobacco smoke exposure on lung function in boys
and girls with and without asthma. Am J Respir Crit Care Med 2000, 162:2097-2104.
146. Gauderman WJ, Gilliland GF, Vora H, Avol E, Stram D, McConnell R, et al: Association
between air pollution and lung function growth in southern California children: results
from a second cohort. Am J Respir Crit Care Med 2002, 166:76-84.
147. Peters JM, Avol E, Gauderman WJ, Linn WS, Navidi W, London SJ, et al: A study of
twelve Southern California communities with differing levels and types of air pollution.
II. Effects on pulmonary function. Am J Respir Crit Care Med 1999, 159:768-775.
148. Schwartz J: Lung function and chronic exposure to air pollution: a cross-sectional
analysis of NHANES II. Environ Res 1989, 50:309-321.
149. Brunekreef B: Particulate matter and lung function growth in children: a 3-yr follow-up
study in Austrian schoolchildren. Eur Respir J 2002, 20:1354-1355; author reply 1355.
150. Gauderman WJ, McConnell R, Gilliland F, London S, Thomas D, Avol E, et al:
Association between air pollution and lung function growth in southern California
children. Am J Respir Crit Care Med 2000, 162:1383-1390.
151. Gauderman WJ, Avol E, Gilliland F, Vora H, Thomas D, Berhane K, et al: The effect of
air pollution on lung development from 10 to 18 years of age. N Engl J Med 2004,
351:1057-1067.
152. Gauderman WJ, Urman R, Avol E, Berhane K, McConnell R, Rappaport E, et al:
Association of improved air quality with lung development in children. N Engl J Med
2015, 372:905-913.
153. Brunekreef B, Janssen NA, de Hartog J, Harssema H, Knape M, van Vliet P: Air
pollution from truck traffic and lung function in children living near motorways.
Epidemiology 1997, 8:298-303.
154. Gauderman WJ, Vora H, McConnell R, Berhane K, Gilliland F, Thomas D, et al: Effect
of exposure to traffic on lung development from 10 to 18 years of age: a cohort study.
Lancet 2007, 369:571-577.
154
155. Lange NE, Sordillo J, Tarantini L, Bollati V, Sparrow D, Vokonas P, et al: Alu and
LINE-1 methylation and lung function in the normative ageing study. BMJ Open 2012, 2.
156. Yang AS, Estecio MR, Doshi K, Kondo Y, Tajara EH, Issa JP: A simple method for
estimating global DNA methylation using bisulfite PCR of repetitive DNA elements.
Nucleic Acids Res 2004, 32:e38.
157. Lepeule J, Bind MA, Baccarelli AA, Koutrakis P, Tarantini L, Litonjua A, et al:
Epigenetic influences on associations between air pollutants and lung function in elderly
men: the normative aging study. Environ Health Perspect 2014, 122:566-572.
158. Qiu W, Baccarelli A, Carey VJ, Boutaoui N, Bacherman H, Klanderman B, et al:
Variable DNA methylation is associated with chronic obstructive pulmonary disease and
lung function. Am J Respir Crit Care Med 2012, 185:373-381.
159. Fu Q, Yu X, Callaway CW, Lane RH, McKnight RA: Epigenetics: intrauterine growth
retardation (IUGR) modifies the histone code along the rat hepatic IGF-1 gene. Faseb j
2009, 23:2438-2449.
160. Joss-Moore LA, Wang Y, Ogata EM, Sainz AJ, Yu X, Callaway CW, et al: IUGR
differentially alters MeCP2 expression and H3K9Me3 of the PPARgamma gene in male
and female rat lungs during alveolarization. Birth Defects Res A Clin Mol Teratol 2011,
91:672-681.
161. Breton CV, Byun HM, Wenten M, Pan F, Yang A, Gilliland FD: Prenatal tobacco smoke
exposure affects global and gene-specific DNA methylation. Am J Respir Crit Care Med
2009, 180:462-467.
162. Mudduluru G, Allgayer H: The human receptor tyrosine kinase Axl gene--promoter
characterization and regulation of constitutive expression by Sp1, Sp3 and CpG
methylation. Biosci Rep 2008, 28:161-176.
163. Breton CV, Salam MT, Gilliland FD: Heritability and role for the environment in DNA
methylation in AXL receptor tyrosine kinase. Epigenetics 2011, 6:895-898.
164. Lemke G, Rothlin CV: Immunobiology of the TAM receptors. Nat Rev Immunol 2008,
8:327-336.
165. Nagata K, Ohashi K, Nakano T, Arita H, Zong C, Hanafusa H, et al: Identification of the
product of growth arrest-specific gene 6 as a common ligand for Axl, Sky, and Mer
receptor tyrosine kinases. J Biol Chem 1996, 271:30022-30027.
166. Prasad D, Rothlin CV, Burrola P, Burstyn-Cohen T, Lu Q, Garcia de Frutos P, et al:
TAM receptor function in the retinal pigment epithelium. Mol Cell Neurosci 2006, 33:96-
108.
167. Lu Q, Gore M, Zhang Q, Camenisch T, Boast S, Casagranda F, et al: Tyro-3 family
receptors are essential regulators of mammalian spermatogenesis. Nature 1999, 398:723-
728.
168. Lu Q, Lemke G: Homeostatic regulation of the immune system by receptor tyrosine
kinases of the Tyro 3 family. Science 2001, 293:306-311.
169. Lemke G, Burstyn-Cohen T: TAM receptors and the clearance of apoptotic cells. Ann N
Y Acad Sci 2010, 1209:23-29.
170. Wu Y, Tibrewal N, Birge RB: Phosphatidylserine recognition by phagocytes: a view to a
kill. Trends Cell Biol 2006, 16:189-197.
171. Xiong W, Chen Y, Wang H, Wang H, Wu H, Lu Q, et al: Gas6 and the Tyro 3 receptor
tyrosine kinase subfamily regulate the phagocytic function of Sertoli cells. Reproduction
2008, 135:77-87.
155
172. Seitz HM, Camenisch TD, Lemke G, Earp HS, Matsushima GK: Macrophages and
dendritic cells use different Axl/Mertk/Tyro3 receptors in clearance of apoptotic cells. J
Immunol 2007, 178:5635-5642.
173. Mahoney JA, Rosen A: Apoptosis and autoimmunity. Curr Opin Immunol 2005, 17:583-
588.
174. Rothlin CV, Ghosh S, Zuniga EI, Oldstone MB, Lemke G: TAM receptors are pleiotropic
inhibitors of the innate immune response. Cell 2007, 131:1124-1136.
175. Iwasaki A, Medzhitov R: Toll-like receptor control of the adaptive immune responses.
Nat Immunol 2004, 5:987-995.
176. Yoshimura A, Naka T, Kubo M: SOCS proteins, cytokine signalling and immune
regulation. Nat Rev Immunol 2007, 7:454-465.
177. Scutera S, Fraone T, Musso T, Cappello P, Rossi S, Pierobon D, et al: Survival and
migration of human dendritic cells are regulated by an IFN-alpha-inducible Axl/Gas6
pathway. J Immunol 2009, 183:3004-3013.
178. Caraux A, Lu Q, Fernandez N, Riou S, Di Santo JP, Raulet DH, et al: Natural killer cell
differentiation driven by Tyro3 receptor tyrosine kinases. Nat Immunol 2006, 7:747-754.
179. Behrens EM, Gadue P, Gong SY, Garrett S, Stein PL, Cohen PL: The mer receptor
tyrosine kinase: expression and function suggest a role in innate immunity. Eur J
Immunol 2003, 33:2160-2167.
180. Roth C, Rothlin C, Riou S, Raulet DH, Lemke G: Stromal-cell regulation of natural killer
cell differentiation. J Mol Med (Berl) 2007, 85:1047-1056.
181. Rothlin CV, Lemke G: TAM receptor signaling and autoimmune disease. Curr Opin
Immunol 2010, 22:740-746.
182. Radic MZ, Shah K, Zhang W, Lu Q, Lemke G, Hilliard GM: Heterogeneous nuclear
ribonucleoprotein P2 is an autoantibody target in mice deficient for Mer, Axl, and Tyro3
receptor tyrosine kinases. J Immunol 2006, 176:68-74.
183. Suh CH, Hilliard B, Li S, Merrill JT, Cohen PL: TAM receptor ligands in lupus: protein
S but not Gas6 levels reflect disease activity in systemic lupus erythematosus. Arthritis
Res Ther 2010, 12:R146.
184. Cheong HS, Lee SO, Choi CB, Sung YK, Shin HD, Bae SC: MERTK polymorphisms
associated with risk of haematological disorders among Korean SLE patients.
Rheumatology (Oxford) 2007, 46:209-214.
185. Ma GZ, Stankovich J, Kilpatrick TJ, Binder MD, Field J: Polymorphisms in the receptor
tyrosine kinase MERTK gene are associated with multiple sclerosis susceptibility. PLoS
One 2011, 6:e16964.
186. McColl A, Bournazos S, Franz S, Perretti M, Morgan BP, Haslett C, et al:
Glucocorticoids induce protein S-dependent phagocytosis of apoptotic neutrophils by
human macrophages. J Immunol 2009, 183:2167-2175.
187. Melaragno MG, Fridell YW, Berk BC: The Gas6/Axl system: a novel regulator of
vascular cell function. Trends Cardiovasc Med 1999, 9:250-253.
188. Collett GD, Sage AP, Kirton JP, Alexander MY, Gilmore AP, Canfield AE:
Axl/phosphatidylinositol 3-kinase signaling inhibits mineral deposition by vascular
smooth muscle cells. Circ Res 2007, 100:502-509.
189. Hurtado B, Munoz X, Recarte-Pelz P, Garcia N, Luque A, Krupinski J, et al: Expression
of the vitamin K-dependent proteins GAS6 and protein S and the TAM receptor tyrosine
kinases in human atherosclerotic carotid plaques. Thromb Haemost 2011, 105:873-882.
156
190. Angelillo-Scherrer A, Burnier L, Flores N, Savi P, DeMol M, Schaeffer P, et al: Role of
Gas6 receptors in platelet signaling during thrombus stabilization and implications for
antithrombotic therapy. J Clin Invest 2005, 115:237-246.
191. Janssen JW, Schulz AS, Steenvoorden AC, Schmidberger M, Strehl S, Ambros PF, et al:
A novel putative tyrosine kinase receptor with oncogenic potential. Oncogene 1991,
6:2113-2120.
192. O'Bryan JP, Frye RA, Cogswell PC, Neubauer A, Kitch B, Prokop C, et al: axl, a
transforming gene isolated from primary human myeloid leukemia cells, encodes a novel
receptor tyrosine kinase. Mol Cell Biol 1991, 11:5016-5031.
193. Schulz AS, Schleithoff L, Faust M, Bartram CR, Janssen JW: The genomic structure of
the human UFO receptor. Oncogene 1993, 8:509-513.
194. Mudduluru G, Vajkoczy P, Allgayer H: Myeloid zinc finger 1 induces migration,
invasion, and in vivo metastasis through Axl gene expression in solid cancer. Mol Cancer
Res 2010, 8:159-169.
195. Mudduluru G, Ceppi P, Kumarswamy R, Scagliotti GV, Papotti M, Allgayer H:
Regulation of Axl receptor tyrosine kinase expression by miR-34a and miR-199a/b in
solid cancer. Oncogene 2011, 30:2888-2899.
196. Linger RM, Keating AK, Earp HS, Graham DK: TAM receptor tyrosine kinases: biologic
functions, signaling, and potential therapeutic targeting in human cancer. Adv Cancer Res
2008, 100:35-83.
197. Ou WB, Corson JM, Flynn DL, Lu WP, Wise SC, Bueno R, et al: AXL regulates
mesothelioma proliferation and invasiveness. Oncogene 2011, 30:1643-1652.
198. Korshunov VA: Axl-dependent signalling: a clinical update. Clin Sci (Lond) 2012,
122:361-368.
199. Sharif MN, Sosic D, Rothlin CV, Kelly E, Lemke G, Olson EN, et al: Twist mediates
suppression of inflammation by type I IFNs and Axl. J Exp Med 2006, 203:1891-1901.
200. Aoki T, Matsumoto Y, Hirata K, Ochiai K, Okada M, Ichikawa K, et al: Expression
profiling of genes related to asthma exacerbations. Clin Exp Allergy 2009, 39:213-221.
201. Hsiao FC, Lin YF, Hsieh PS, Chu NF, Chen YD, Shieh YS, et al: Effect of GAS6 and
AXL Gene Polymorphisms on Adiposity, Systemic Inflammation, and Insulin Resistance
in Adolescents. Int J Endocrinol 2014, 2014:674069.
202. Recarte-Pelz P, Tassies D, Espinosa G, Hurtado B, Sala N, Cervera R, et al: Vitamin K-
dependent proteins GAS6 and Protein S and TAM receptors in patients of systemic lupus
erythematosus: correlation with common genetic variants and disease activity. Arthritis
Res Ther 2013, 15:R41.
203. Fujimori T, Grabiec AM, Kaur M, Bell TJ, Fujino N, Cook PC, et al: The Axl receptor
tyrosine kinase is a discriminator of macrophage function in the inflamed lung. Mucosal
Immunol 2015, 8:1021-1030.
204. Grabiec AM, Denny N, Doherty JA, Happonen KE, Hankinson J, Connolly E, et al:
Diminished airway macrophage expression of the Axl receptor tyrosine kinase is
associated with defective efferocytosis in asthma. J Allergy Clin Immunol 2017.
205. Hou L, Zhang X, Wang D, Baccarelli A: Environmental chemical exposures and human
epigenetics. Int J Epidemiol 2012, 41:79-105.
206. Bollati V, Baccarelli A, Hou L, Bonzini M, Fustinoni S, Cavallo D, et al: Changes in
DNA methylation patterns in subjects exposed to low-dose benzene. Cancer Res 2007,
67:876-880.
157
207. Sullivan J, Sheppard L, Schreuder A, Ishikawa N, Siscovick D, Kaufman J: Relation
between short-term fine-particulate matter exposure and onset of myocardial infarction.
Epidemiology 2005, 16:41-48.
208. Peng RD, Chang HH, Bell ML, McDermott A, Zeger SL, Samet JM, et al: Coarse
particulate matter air pollution and hospital admissions for cardiovascular and respiratory
diseases among Medicare patients. Jama 2008, 299:2172-2179.
209. Kelly FJ, Fussell JC: Size, source and chemical composition as determinants of toxicity
attributable to ambient particulate matter. Atmospheric Environment 2012, 60:504-526.
210. Kreyling WG, Semmler M, Erbe F, Mayer P, Takenaka S, Schulz H, et al: Translocation
of ultrafine insoluble iridium particles from lung epithelium to extrapulmonary organs is
size dependent but very low. J Toxicol Environ Health A 2002, 65:1513-1530.
211. Moller P, Jacobsen NR, Folkmann JK, Danielsen PH, Mikkelsen L, Hemmingsen JG, et
al: Role of oxidative damage in toxicity of particulates. Free Radic Res 2010, 44:1-46.
212. Tarantini L, Bonzini M, Apostoli P, Pegoraro V, Bollati V, Marinelli B, et al: Effects of
particulate matter on genomic DNA methylation content and iNOS promoter
methylation. Environ Health Perspect 2009, 117:217-222.
213. Baccarelli A, Wright RO, Bollati V, Tarantini L, Litonjua AA, Suh HH, et al: Rapid
DNA methylation changes after exposure to traffic particles. Am J Respir Crit Care Med
2009, 179:572-578.
214. Salam MT, Byun HM, Lurmann F, Breton CV, Wang X, Eckel SP, et al: Genetic and
epigenetic variations in inducible nitric oxide synthase promoter, particulate pollution,
and exhaled nitric oxide levels in children. J Allergy Clin Immunol 2012, 129:232-
239.e231-237.
215. Yauk C, Polyzos A, Rowan-Carroll A, Somers CM, Godschalk RW, Van Schooten FJ, et
al: Germ-line mutations, DNA damage, and global hypermethylation in mice exposed to
particulate air pollution in an urban/industrial location. Proc Natl Acad Sci U S A 2008,
105:605-610.
216. Foley DL, Craig JM, Morley R, Olsson CA, Dwyer T, Smith K, et al: Prospects for
epigenetic epidemiology. Am J Epidemiol 2009, 169:389-400.
217. Sassone-Corsi P: Unique chromatin remodeling and transcriptional regulation in
spermatogenesis. Science 2002, 296:2176-2178.
218. Hussain N: Epigenetic influences that modulate infant growth, development, and disease.
Antioxid Redox Signal 2012, 17:224-236.
219. Ho SM, Tang WY, Belmonte de Frausto J, Prins GS: Developmental exposure to
estradiol and bisphenol A increases susceptibility to prostate carcinogenesis and
epigenetically regulates phosphodiesterase type 4 variant 4. Cancer Res 2006, 66:5624-
5632.
220. Goodrich JM, Sanchez BN, Dolinoy DC, Zhang Z, Hernandez-Avila M, Hu H, et al:
Quality control and statistical modeling for environmental epigenetics: a study on in utero
lead exposure and DNA methylation at birth. Epigenetics 2015, 10:19-30.
221. Flom JD, Ferris JS, Liao Y, Tehranifar P, Richards CB, Cho YH, et al: Prenatal smoke
exposure and genomic DNA methylation in a multiethnic birth cohort. Cancer Epidemiol
Biomarkers Prev 2011, 20:2518-2523.
222. Markunas CA, Xu Z, Harlid S, Wade PA, Lie RT, Taylor JA, et al: Identification of DNA
methylation changes in newborns related to maternal smoking during pregnancy. Environ
Health Perspect 2014, 122:1147-1153.
158
223. Joubert BR, Haberg SE, Nilsen RM, Wang X, Vollset SE, Murphy SK, et al: 450K
epigenome-wide scan identifies differential DNA methylation in newborns related to
maternal smoking during pregnancy. Environ Health Perspect 2012, 120:1425-1431.
224. Joubert BR, Felix JF, Yousefi P, Bakulski KM, Just AC, Breton C, et al: DNA
Methylation in Newborns and Maternal Smoking in Pregnancy: Genome-wide
Consortium Meta-analysis. Am J Hum Genet 2016, 98:680-696.
225. Curley JP, Mashoodh R, Champagne FA: Epigenetics and the origins of paternal effects.
Horm Behav 2011, 59:306-314.
226. Breton CV, Gao L, Yao J, Siegmund KD, Lurmann F, Gilliland F: Particulate matter, the
newborn methylome, and cardio-respiratory health outcomes in childhood.
Environmental Epigenetics 2016, 2:1-9.
227. Gruzieva O, Xu CJ, Breton CV, Annesi-Maesano I, Anto JM, Auffray C, et al:
Epigenome-Wide Meta-Analysis of Methylation in Children Related to Prenatal NO2 Air
Pollution Exposure. Environ Health Perspect 2017, 125:104-110.
228. Hollingsworth JW, Maruoka S, Boon K, Garantziotis S, Li Z, Tomfohr J, et al: In utero
supplementation with methyl donors enhances allergic airway disease in mice. J Clin
Invest 2008, 118:3462-3469.
229. Waterland RA, Travisano M, Tahiliani KG, Rached MT, Mirza S: Methyl donor
supplementation prevents transgenerational amplification of obesity. Int J Obes (Lond)
2008, 32:1373-1379.
230. Moorman JE, Rudd RA, Johnson CA, King M, Minor P, Bailey C, et al: National
surveillance for asthma--United States, 1980-2004. MMWR Surveill Summ 2007, 56:1-
54.
231. Mannino DM, Homa DM, Akinbami LJ, Moorman JE, Gwynn C, Redd SC: Surveillance
for asthma--United States, 1980-1999. MMWR Surveill Summ 2002, 51:1-13.
232. Duijts L: Fetal and infant origins of asthma. Eur J Epidemiol 2012, 27:5-14.
233. Pijnenburg MW, De Jongste JC: Exhaled nitric oxide in childhood asthma: a review. Clin
Exp Allergy 2008, 38:246-259.
234. Holgate ST: Epithelium dysfunction in asthma. J Allergy Clin Immunol 2007, 120:1233-
1244; quiz 1245-1236.
235. Lambrecht BN, Hammad H: The airway epithelium in asthma. Nat Med 2012, 18:684-
692.
236. Xiao C, Puddicombe SM, Field S, Haywood J, Broughton-Head V, Puxeddu I, et al:
Defective epithelial barrier function in asthma. J Allergy Clin Immunol 2011, 128:549-
556.e541-512.
237. Li YF, Langholz B, Salam MT, Gilliland FD: Maternal and grandmaternal smoking
patterns are associated with early childhood asthma. Chest 2005, 127:1232-1241.
238. Jerrett M, Shankardass K, Berhane K, Gauderman WJ, Kunzli N, Avol E, et al: Traffic-
related air pollution and asthma onset in children: a prospective cohort study with
individual exposure measurement. Environ Health Perspect 2008, 116:1433-1438.
239. Berhane K, Chang CC, McConnell R, Gauderman WJ, Avol E, Rapapport E, et al:
Association of Changes in Air Quality With Bronchitic Symptoms in Children in
California, 1993-2012. Jama 2016, 315:1491-1501.
240. Cook DG, Strachan DP: Health effects of passive smoking. 3. Parental smoking and
prevalence of respiratory symptoms and asthma in school age children. Thorax 1997,
52:1081-1094.
159
241. Jung KH, Lovinsky-Desir S, Yan B, Torrone D, Lawrence J, Jezioro JR, et al: Effect of
personal exposure to black carbon on changes in allergic asthma gene methylation
measured 5 days later in urban children: importance of allergic sensitization. Clin
Epigenetics 2017, 9:61.
242. Perera F, Tang WY, Herbstman J, Tang D, Levin L, Miller R, et al: Relation of DNA
methylation of 5'-CpG island of ACSL3 to transplacental exposure to airborne polycyclic
aromatic hydrocarbons and childhood asthma. PLoS One 2009, 4:e4488.
243. Wang IJ, Karmaus WJ, Chen SL, Holloway JW, Ewart S: Effects of phthalate exposure
on asthma may be mediated through alterations in DNA methylation. Clin Epigenetics
2015, 7:27.
244. Reik W: Stability and flexibility of epigenetic gene regulation in mammalian
development. Nature 2007, 447:425-432.
245. Stitt TN, Conn G, Gore M, Lai C, Bruno J, Radziejewski C, et al: The anticoagulation
factor protein S and its relative, Gas6, are ligands for the Tyro 3/Axl family of receptor
tyrosine kinases. Cell 1995, 80:661-670.
246. Kasowski M, Grubert F, Heffelfinger C, Hariharan M, Asabere A, Waszak SM, et al:
Variation in transcription factor binding among humans. Science 2010, 328:232-235.
247. McDaniell R, Lee BK, Song L, Liu Z, Boyle AP, Erdos MR, et al: Heritable individual-
specific and allele-specific chromatin signatures in humans. Science 2010, 328:235-239.
248. Cancer Genome Atlas Research N: Comprehensive genomic characterization defines
human glioblastoma genes and core pathways. Nature 2008, 455:1061-1068.
249. Grossman RL, Heath AP, Ferretti V, Varmus HE, Lowy DR, Kibbe WA, et al: Toward a
Shared Vision for Cancer Genomic Data. N Engl J Med 2016, 375:1109-1112.
250. The Cancer Genome Atlas [https://cancergenome.nih.gov/]
251. Peters JM, Avol E, Navidi W, London SJ, Gauderman WJ, Lurmann F, et al: A study of
twelve Southern California communities with differing levels and types of air pollution. I.
Prevalence of respiratory morbidity. Am J Respir Crit Care Med 1999, 159:760-767.
252. Dratva J, Breton CV, Hodis HN, Mack WJ, Salam MT, Zemp E, et al: Birth weight and
carotid artery intima-media thickness. J Pediatr 2013, 162:906-911.e901-902.
253. Hoyo C, Murtha AP, Schildkraut JM, Forman MR, Calingaert B, Demark-Wahnefried W,
et al: Folic acid supplementation before and during pregnancy in the Newborn
Epigenetics STudy (NEST). BMC Public Health 2011, 11:46.
254. Aryee MJ, Jaffe AE, Corrada-Bravo H, Ladd-Acosta C, Feinberg AP, Hansen KD, et al:
Minfi: a flexible and comprehensive Bioconductor package for the analysis of Infinium
DNA methylation microarrays. Bioinformatics 2014, 30:1363-1369.
255. Triche TJ, Weisenberger DJ, Van Den Berg D, Laird PW, Siegmund KD: Low-level
Processing of Illumina Infinium DNA Methylation BeadArrays. Nucleic Acids Res 2013,
(in press).
256. Bolstad BM, Irizarry RA, Astrand M, Speed TP: A comparison of normalization methods
for high density oligonucleotide array data based on variance and bias. Bioinformatics
2003, 19:185-193.
257. Jaffe AE, Irizarry RA: Accounting for cellular heterogeneity is critical in epigenome-
wide association studies. Genome Biol 2014, 15:R31.
258. Bakulski KM, Feinberg JI, Andrews SV, Yang J, Brown S, S LM, et al: DNA
methylation of cord blood cell types: Applications for mixed cell birth studies.
Epigenetics 2016, 11:354-362.
160
259. Tost J, Dunker J, Gut IG: Analysis and quantification of multiple methylation variable
positions in CpG islands by Pyrosequencing. Biotechniques 2003, 35:152-156.
260. Brakensiek K, Wingen LU, Langer F, Kreipe H, Lehmann U: Quantitative high-
resolution CpG island mapping with Pyrosequencing reveals disease-specific methylation
patterns of the CDKN2B gene in myelodysplastic syndrome and myeloid leukemia. Clin
Chem 2007, 53:17-23.
261. Zhou X, Maricque B, Xie M, Li D, Sundaram V, Martin EA, et al: The Human
Epigenome Browser at Washington University. Nat Methods 2011, 8:989-990.
262. Human Epigenome Browser at Washington University in St. Louis
[http://epigenomegateway.wustl.edu/]
263. Li YF, Gauderman WJ, Conti DV, Lin PC, Avol E, Gilliland FD: Glutathione S-
transferase P1, maternal smoking, and asthma in children: a haplotype-based analysis.
Environ Health Perspect 2008, 116:409-415.
264. Torgerson DG, Ampleford EJ, Chiu GY, Gauderman WJ, Gignoux CR, Graves PE, et al:
Meta-analysis of genome-wide association studies of asthma in ethnically diverse North
American populations. Nat Genet 2011, 43:887-892.
265. Barrett JC, Fry B, Maller J, Daly MJ: Haploview: analysis and visualization of LD and
haplotype maps. Bioinformatics 2005, 21:263-265.
266. Pritchard JK, Stephens M, Donnelly P: Inference of population structure using multilocus
genotype data. Genetics 2000, 155:945-959.
267. Hochberg Y, Benjamini Y: More powerful procedures for multiple significance testing.
Stat Med 1990, 9:811-818.
268. Zentner GE, Tesar PJ, Scacheri PC: Epigenetic signatures distinguish multiple classes of
enhancers with distinct cellular functions. Genome Res 2011, 21:1273-1283.
269. Creyghton MP, Cheng AW, Welstead GG, Kooistra T, Carey BW, Steine EJ, et al:
Histone H3K27ac separates active from poised enhancers and predicts developmental
state. Proc Natl Acad Sci U S A 2010, 107:21931-21936.
270. Kent WJ, Sugnet CW, Furey TS, Roskin KM, Pringle TH, Zahler AM, et al: The human
genome browser at UCSC. Genome Res 2002, 12:996-1006.
271. The UCSC Genome Browser [http://genome.ucsc.edu/cite.html]
272. Matsumoto M, Tanaka N, Harada H, Kimura T, Yokochi T, Kitagawa M, et al:
Activation of the transcription factor ISGF3 by interferon-gamma. Biol Chem 1999,
380:699-703.
273. Hess J, Angel P, Schorpp-Kistner M: AP-1 subunits: quarrel and harmony among
siblings. J Cell Sci 2004, 117:5965-5973.
274. Ning S, Pagano JS, Barber GN: IRF7: activation, regulation, modification and function.
Genes Immun 2011, 12:399-414.
275. Holland SJ, Pan A, Franci C, Hu Y, Chang B, Li W, et al: R428, a selective small
molecule inhibitor of Axl kinase, blocks tumor spread and prolongs survival in models of
metastatic breast cancer. Cancer Res 2010, 70:1544-1554.
276. Hong CC, Lay JD, Huang JS, Cheng AL, Tang JL, Lin MT, et al: Receptor tyrosine
kinase AXL is induced by chemotherapy drugs and overexpression of AXL confers drug
resistance in acute myeloid leukemia. Cancer Lett 2008, 268:314-324.
277. Melaragno MG, Wuthrich DA, Poppa V, Gill D, Lindner V, Berk BC, et al: Increased
expression of Axl tyrosine kinase after vascular injury and regulation by G protein-
coupled receptor agonists in rats. Circ Res 1998, 83:697-704.
161
278. Healy AM, Schwartz JJ, Zhu X, Herrick BE, Varnum B, Farber HW: Gas 6 promotes
Axl-mediated survival in pulmonary endothelial cells. Am J Physiol Lung Cell Mol
Physiol 2001, 280:L1273-1281.
279. Melaragno MG, Cavet ME, Yan C, Tai LK, Jin ZG, Haendeler J, et al: Gas6 inhibits
apoptosis in vascular smooth muscle: role of Axl kinase and Akt. J Mol Cell Cardiol
2004, 37:881-887.
280. Ray A, Khare A, Krishnamoorthy N, Qi Z, Ray P: Regulatory T cells in many flavors
control asthma. Mucosal Immunol 2010, 3:216-229.
281. Lesseur C, Armstrong DA, Murphy MA, Appleton AA, Koestler DC, Paquette AG, et al:
Sex-specific associations between placental leptin promoter DNA methylation and infant
neurobehavior. Psychoneuroendocrinology 2014, 40:1-9.
282. Sawalha AH, Wang L, Nadig A, Somers EC, McCune WJ, Hughes T, et al: Sex-specific
differences in the relationship between genetic susceptibility, T cell DNA demethylation
and lupus flare severity. J Autoimmun 2012, 38:J216-222.
283. Shimabukuro M, Sasaki T, Imamura A, Tsujita T, Fuke C, Umekage T, et al: Global
hypomethylation of peripheral leukocyte DNA in male patients with schizophrenia: a
potential link between epigenetics and schizophrenia. J Psychiatr Res 2007, 41:1042-
1046.
284. Davies MN, Volta M, Pidsley R, Lunnon K, Dixit A, Lovestone S, et al: Functional
annotation of the human brain methylome identifies tissue-specific epigenetic variation
across brain and blood. Genome Biol 2012, 13:R43.
285. Grundberg E, Meduri E, Sandling JK, Hedman AK, Keildson S, Buil A, et al: Global
analysis of DNA methylation variation in adipose tissue from twins reveals links to
disease-associated variants in distal regulatory elements. Am J Hum Genet 2013, 93:876-
890.
286. van Eijk KR, de Jong S, Boks MP, Langeveld T, Colas F, Veldink JH, et al: Genetic
analysis of DNA methylation and gene expression levels in whole blood of healthy
human subjects. BMC Genomics 2012, 13:636.
287. Wagner JR, Busche S, Ge B, Kwan T, Pastinen T, Blanchette M: The relationship
between DNA methylation, genetic and expression inter-individual variation in
untransformed human fibroblasts. Genome Biol 2014, 15:R37.
288. Carithers LJ, Moore HM: The Genotype-Tissue Expression (GTEx) Project. Biopreserv
Biobank 2015, 13:307-308.
289. The GTEx Portal [https://gtexportal.org/home/]
290. Fogarty AW, Jones S, Britton JR, Lewis SA, McKeever TM: Systemic inflammation and
decline in lung function in a general population: a prospective study. Thorax 2007,
62:515-520.
291. Gold DR, Wang X, Wypij D, Speizer FE, Ware JH, Dockery DW: Effects of cigarette
smoking on lung function in adolescent boys and girls. N Engl J Med 1996, 335:931-937.
292. McClearn GE, Svartengren M, Pedersen NL, Heller DA, Plomin R: Genetic and
environmental influences on pulmonary function in aging Swedish twins. J Gerontol
1994, 49:264-268.
293. Urman R, McConnell R, Islam T, Avol EL, Lurmann FW, Vora H, et al: Associations of
children's lung function with ambient air pollution: joint effects of regional and near-
roadway pollutants. Thorax 2014, 69:540-547.
162
294. Golbabapour S, Abdulla MA, Hajrezaei M: A concise review on epigenetic regulation:
insight into molecular mechanisms. Int J Mol Sci 2011, 12:8661-8694.
295. Park IK, Giovenzana C, Hughes TL, Yu J, Trotta R, Caligiuri MA: The Axl/Gas6
pathway is required for optimal cytokine signaling during human natural killer cell
development. Blood 2009, 113:2470-2477.
296. Demedts IK, Demoor T, Bracke KR, Joos GF, Brusselle GG: Role of apoptosis in the
pathogenesis of COPD and pulmonary emphysema. Respir Res 2006, 7:53.
297. Gao L, Millstein J, Siegmund KD, Dubeau L, Maguire R, Gilliland FD, et al: Epigenetic
regulation of AXL and risk of childhood asthma symptoms. Clin Epigenetics 2017,
9:121.
298. Breton CV, Yao J, Millstein J, Gao L, Siegmund KD, Mack W, et al: Prenatal Air
Pollution Exposures, DNA Methyl Transferase Genotypes, and Associations with
Newborn LINE1 and Alu Methylation and Childhood Blood Pressure and Carotid Intima-
Media Thickness in the Children's Health Study. Environ Health Perspect 2016,
124:1905-1912.
299. Luna A, Nicodemus KK: snp.plotter: an R-based SNP/haplotype association and linkage
disequilibrium plotting package. Bioinformatics 2007, 23:774-776.
300. Martin TC, Yet I, Tsai PC, Bell JT: coMET: visualisation of regional epigenome-wide
association scan results and DNA co-methylation patterns. BMC Bioinformatics 2015,
16:131.
301. Bernstein BE, Stamatoyannopoulos JA, Costello JF, Ren B, Milosavljevic A, Meissner
A, et al: The NIH Roadmap Epigenomics Mapping Consortium. Nat Biotechnol 2010,
28:1045-1048.
302. Gilliland F, Avol E, McConnell R, Berhane K, Gauderman WJ, Lurmann FW, et al: The
Effects of Policy-Driven Air Quality Improvements on Children’s Respiratory Health.
Research Report I90. Boston, MA: Health Effects Institute; 2017.
303. Hussell T, Bell TJ: Alveolar macrophages: plasticity in a tissue-specific context. Nat Rev
Immunol 2014, 14:81-93.
304. Grabiec AM, Hussell T: The role of airway macrophages in apoptotic cell clearance
following acute and chronic lung inflammation. Semin Immunopathol 2016, 38:409-423.
305. Sin DD, Wu L, Man SF: The relationship between reduced lung function and
cardiovascular mortality: a population-based study and a systematic review of the
literature. Chest 2005, 127:1952-1959.
306. Alexeeff SE, Coull BA, Gryparis A, Suh H, Sparrow D, Vokonas PS, et al: Medium-term
exposure to traffic-related air pollution and markers of inflammation and endothelial
function. Environ Health Perspect 2011, 119:481-486.
307. McConnell R, Berhane K, Gilliland F, London SJ, Islam T, Gauderman WJ, et al:
Asthma in exercising children exposed to ozone: a cohort study. Lancet 2002, 359:386-
391.
308. Clark NA, Demers PA, Karr CJ, Koehoorn M, Lencar C, Tamburic L, et al: Effect of
early life exposure to air pollution on development of childhood asthma. Environ Health
Perspect 2010, 118:284-290.
309. Weldy CS, Liu Y, Liggitt HD, Chin MT: In utero exposure to diesel exhaust air pollution
promotes adverse intrauterine conditions, resulting in weight gain, altered blood pressure,
and increased susceptibility to heart failure in adult mice. PLoS One 2014, 9:e88582.
163
310. Weldy CS, Liu Y, Chang YC, Medvedev IO, Fox JR, Larson TV, et al: In utero and early
life exposure to diesel exhaust air pollution increases adult susceptibility to heart failure
in mice. Part Fibre Toxicol 2013, 10:59.
311. Chen Z, Salam MT, Eckel SP, Breton CV, Gilliland FD: Chronic effects of air pollution
on respiratory health in Southern California children: findings from the Southern
California Children's Health Study. J Thorac Dis 2015, 7:46-58.
312. Bollati V, Baccarelli A: Environmental epigenetics. Heredity (Edinb) 2010, 105:105-112.
313. Breton CV, Salam MT, Wang X, Byun HM, Siegmund KD, Gilliland FD: Particulate
matter, DNA methylation in nitric oxide synthase, and childhood respiratory disease.
Environ Health Perspect 2012, 120:1320-1326.
314. Liu X, Chen Q, Tsai HJ, Wang G, Hong X, Zhou Y, et al: Maternal preconception body
mass index and offspring cord blood DNA methylation: exploration of early life origins
of disease. Environ Mol Mutagen 2014, 55:223-230.
315. Janssen BG, Godderis L, Pieters N, Poels K, Kicinski M, Cuypers A, et al: Placental
DNA hypomethylation in association with particulate air pollution in early life. Part
Fibre Toxicol 2013, 10:22.
316. Zhu Y, Zhang C, Liu D, Grantz KL, Wallace M, Mendola P: Maternal ambient air
pollution exposure preconception and during early gestation and offspring congenital
orofacial defects. Environ Res 2015, 140:714-720.
317. Robledo CA, Mendola P, Yeung E, Mannisto T, Sundaram R, Liu D, et al: Preconception
and early pregnancy air pollution exposures and risk of gestational diabetes mellitus.
Environ Res 2015, 137:316-322.
318. Park HY, Bae S, Hong YC: PM10 exposure and non-accidental mortality in Asian
populations: a meta-analysis of time-series and case-crossover studies. J Prev Med Public
Health 2013, 46:10-18.
319. Bell ML, Samet JM, Dominici F: Time-series studies of particulate matter. Annu Rev
Public Health 2004, 25:247-280.
320. Goodman PG, Dockery DW, Clancy L: Cause-specific mortality and the extended effects
of particulate pollution and temperature exposure. Environ Health Perspect 2004,
112:179-185.
321. Elliott P, Shaddick G, Wakefield JC, de Hoogh C, Briggs DJ: Long-term associations of
outdoor air pollution with mortality in Great Britain. Thorax 2007, 62:1088-1094.
322. Checkoway H, Pearce N, Hickey JL, Dement JM: Latency analysis in occupational
epidemiology. Arch Environ Health 1990, 45:95-100.
323. Braga AL, Zanobetti A, Schwartz J: The time course of weather-related deaths.
Epidemiology 2001, 12:662-667.
324. Samoli E, Zanobetti A, Schwartz J, Atkinson R, LeTertre A, Schindler C, et al: The
temporal pattern of mortality responses to ambient ozone in the APHEA project. J
Epidemiol Community Health 2009, 63:960-966.
325. Almon S: The distributed lag between capital appropriations and expenditures.
Econometrica 1965, 33:178-196.
326. Gasparrini A: Distributed Lag Linear and Non-Linear Models in R: The Package dlnm. J
Stat Softw 2011, 43:1-20.
327. Martins LC, Pereira LA, Lin CA, Santos UP, Prioli G, Luiz Odo C, et al: The effects of
air pollution on cardiovascular diseases: lag structures. Rev Saude Publica 2006, 40:677-
683.
164
328. Berhane K, Zhang Y, Linn WS, Rappaport EB, Bastain TM, Salam MT, et al: The effect
of ambient air pollution on exhaled nitric oxide in the Children's Health Study. Eur
Respir J 2011, 37:1029-1036.
329. Bind MA, Lepeule J, Zanobetti A, Gasparrini A, Baccarelli A, Coull BA, et al: Air
pollution and gene-specific methylation in the Normative Aging Study: association,
effect modification, and mediation analysis. Epigenetics 2014, 9:448-458.
330. Bind MA, Zanobetti A, Gasparrini A, Peters A, Coull B, Baccarelli A, et al: Effects of
temperature and relative humidity on DNA methylation. Epidemiology 2014, 25:561-569.
331. Selzer RH, Hodis HN, Kwong-Fu H, Mack WJ, Lee PL, Liu CR, et al: Evaluation of
computerized edge tracking for quantifying intima-media thickness of the common
carotid artery from B-mode ultrasound images. Atherosclerosis 1994, 111:1-11.
332. Selzer RH, Mack WJ, Lee PL, Kwong-Fu H, Hodis HN: Improved common carotid
elasticity and intima-media thickness measurements from computer analysis of sequential
ultrasound frames. Atherosclerosis 2001, 154:185-193.
333. Du P, Zhang X, Huang CC, Jafari N, Kibbe WA, Hou L, et al: Comparison of Beta-value
and M-value methods for quantifying methylation levels by microarray analysis. BMC
Bioinformatics 2010, 11:587.
334. Murphy LE, Gollenberg AL, Buck Louis GM, Kostyniak PJ, Sundaram R: Maternal
serum preconception polychlorinated biphenyl concentrations and infant birth weight.
Environ Health Perspect 2010, 118:297-302.
335. Day J, Savani S, Krempley BD, Nguyen M, Kitlinska JB: Influence of paternal
preconception exposures on their offspring: through epigenetics to phenotype. Am J Stem
Cells 2016, 5:11-18.
336. Greer EL, Maures TJ, Ucar D, Hauswirth AG, Mancini E, Lim JP, et al:
Transgenerational epigenetic inheritance of longevity in Caenorhabditis elegans. Nature
2011, 479:365-371.
337. Panni T, Mehta AJ, Schwartz JD, Baccarelli AA, Just AC, Wolf K, et al: Genome-Wide
Analysis of DNA Methylation and Fine Particulate Matter Air Pollution in Three Study
Populations: KORA F3, KORA F4, and the Normative Aging Study. Environ Health
Perspect 2016, 124:983-990.
338. Bauer M, Linsel G, Fink B, Offenberg K, Hahn AM, Sack U, et al: A varying T cell
subtype explains apparent tobacco smoking induced single CpG hypomethylation in
whole blood. Clin Epigenetics 2015, 7:81.
339. McDougall CM, Blaylock MG, Douglas JG, Brooker RJ, Helms PJ, Walsh GM: Nasal
epithelial cells as surrogates for bronchial epithelial cells in airway inflammation studies.
Am J Respir Cell Mol Biol 2008, 39:560-568.
340. Thavagnanam S, Parker JC, McBrien ME, Skibinski G, Shields MD, Heaney LG: Nasal
epithelial cells can act as a physiological surrogate for paediatric asthma studies. PLoS
One 2014, 9:e85802.
341. Byun HM, Siegmund KD, Pan F, Weisenberger DJ, Kanel G, Laird PW, et al: Epigenetic
profiling of somatic tissues from human autopsy specimens identifies tissue- and
individual-specific DNA methylation patterns. Hum Mol Genet 2009, 18:4808-4817.
342. Slieker RC, Bos SD, Goeman JJ, Bovee JV, Talens RP, van der Breggen R, et al:
Identification and systematic annotation of tissue-specific differentially methylated
regions using the Illumina 450k array. Epigenetics Chromatin 2013, 6:26.
165
343. Varley KE, Gertz J, Bowling KM, Parker SL, Reddy TE, Pauli-Behn F, et al: Dynamic
DNA methylation across diverse human cell lines and tissues. Genome Res 2013, 23:555-
567.
344. eGTEx Project: Enhancing GTEx by bridging the gaps between genotype, gene
expression, and disease. Nat Genet 2017, 49:1664-1670.
345. Zagorska A, Traves PG, Lew ED, Dransfield I, Lemke G: Diversification of TAM
receptor tyrosine kinase function. Nat Immunol 2014, 15:920-928.
346. Ekman C, Stenhoff J, Dahlback B: Gas6 is complexed to the soluble tyrosine kinase
receptor Axl in human blood. J Thromb Haemost 2010, 8:838-844.

Abstract (if available)

Abstract The burden of childhood respiratory disease remains an important public health problem. Asthma is the most common chronic disease in childhood, affecting approximately 1 in 11 children in the US. The consequences associated with childhood asthma include missed school days, poor sleep and fatigue, and symptoms that interfere with play, sports or other activities. Severe asthma attacks even require emergency treatment or hospital care. According to the Centers for Disease Control and Prevention (CDC), current asthma prevalence increased at a rate of 1.4% per year among children aged 0-17 years, imposing an economic burden of two billion dollars annually on healthcare costs for managing childhood asthma. Airway inflammation plays a central role in the pathogenesis of childhood asthma. Although the etiologies are complex, growing evidence has suggested that childhood asthma is determined by the interplay between genetic, demographic and both in-utero and early-life environmental factors. In recent years, epigenetic modifications, including DNA methylation, have emerged as one mechanism underlying the development of childhood asthma by altering regulation of genes involved in airways development or immune-mediated inflammatory pathways. ❧ In addition to the burden of childhood asthma and other respiratory diseases, normal lung function development during childhood and adolescence is of great importance since it is a prerequisite for optimal respiratory health across the life course. Deficits in pulmonary function have been associated with multiple adverse health outcomes including cardiovascular diseases and chronic obstructive pulmonary disease (COPD) in adults, and increased risk of developing asthma in adolescents. Healthy lung development is a complex process that can be modified by genetic, pathological and environmental factors. Although DNA methylation has been suggested to play a role in the development of lung function, studies were mostly conducted in elder subjects or under disease conditions such as COPD. ❧ The aim of this dissertation is to explore epigenetic marks associated with childhood asthma symptoms and lung function development to identify the early origins of chronic diseases and better understand the underlying mechanisms. In addition, since DNA methylation a dynamic form of modification and alterations can be induced by environmental factors, I also aim to investigate the effects of prenatal air pollution exposure on DNA methylation patterns at birth. The candidate gene we selected to study association with childhood respiratory health is AXL, a member of the TAM family receptor tyrosine kinases. AXL and other TAM family genes are important mediators for effective phagocytosis and play a crucial role in innate immune responses. However, the epigenetic regulation of this gene has rarely been studied in outcomes related to childhood respiratory health. ❧ In Chapter 1, I briefly review the role of DNA methylation in the development of human diseases, the molecular basis of DNA methylation and how it regulates gene expression. I then provide a summary for the pathophysiology and known risk factors (demographic, genetic and environmental factors and epigenetic regulation) of childhood asthma and normal pulmonary function development. Lastly, I discuss how environmental exposures may affect DNA methylation. ❧ In Chapter 2, I examine the association between AXL methylation at birth and the risk of childhood asthma symptoms at age 6 years in subjects from the Children’s Health Study (CHS), taking into consideration the underlying genetic variation in AXL. Findings are evaluated for replication in a separate population of CHS subjects using Pyrosequencing. This manuscript was published in Clinical Epigenetics. ❧ In Chapter 3, I further explore the association between AXL DNA methylation at birth and lung function growth during adolescence. We assess the association between AXL methylation measured by Pyrosequencing and 8-year lung function growth (10 to 18 years of age) in CHS subjects. Findings are evaluated for replication in a separate population of 237 CHS subjects using methylation data from the Illumina HumanMethylation450 (HM450) array when possible. ❧ In Chapter 4, I examine the effects of prenatal PM10 exposure on DNA methylation patterns at birth measured by the HM450 array. I use polynomial distributed lag models to identify CpG loci associated with PM10 exposure during the 3-month preconception and 9-month pregnancy window. ❧ Lastly, in Chapter 5 I provide a summary of main findings and suggestions for future research.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Air pollution, mitochondrial function, and growth in children

PDF

Preprocessing and analysis of DNA methylation microarrays

PDF

Air pollution, smoking, and multigenerational DNA methylation Signatures: a study of two southern California cohorts

PDF

Airway inflammation and respiratory health in the Southern California children's health study

PDF

DNA methylation changes in the development of lung adenocarcinoma

PDF

Genetic epidemiological approaches in the study of risk factors for hematologic malignancies

PDF

Early life risk factors for childhood asthma

PDF

Ambient air pollution and lung function in children

PDF

Early childhood health experience & adult phenotype in twins

PDF

Sensor-based mobile health approaches for personal air pollution and pediatric asthma studies

PDF

Genetic and environmental risk factors for childhood cancer

PDF

DNA methylation of NOS genes and carotid intima-media thickness in children

PDF

Gene expression and angiogenesis pathway across DNA methylation subtypes in colon adenocarcinoma

PDF

Prenatal exposure to organophosphate esters and early neurodevelopment

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

PDF

Long term effects of air pollution on lung function in the European community respiratory health survey

PDF

Lifestyle-related exposures and diseases in twins

PDF

Air pollution and childhood obesity

PDF

Functional DNA methylation changes in normal and cancer cells

PDF

An analysis of conservation of methylation

Asset Metadata

Creator Gao, Lu (author)

Core Title Prenatal air pollution exposure, newborn DNA methylation, and childhood respiratory health

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Epidemiology

Publication Date 07/16/2018

Defense Date 12/12/2017

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

Tag Air pollution,asthma,DNA methylation,epigenetics,lung function,OAI-PMH Harvest

Format application/pdf (imt)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Breton, Carrie (committee chair), Dubeau, Louis (committee member), Millstein, Joshua (committee member), Siegmund, Kimberly (committee member)

Creator Email gaol@usc.edu,sdugaolu0803@gmail.com

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-17383

Unique identifier UC11671958

Identifier etd-GaoLu-6403.pdf (filename),usctheses-c89-17383 (legacy record id)

Legacy Identifier etd-GaoLu-6403.pdf

Dmrecord 17383

Document Type Dissertation

Format application/pdf (imt)

Rights Gao, Lu

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