Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Genetic association studies of age-related macular degeneration from candidate gene to whole genome
(USC Thesis Other)
Genetic association studies of age-related macular degeneration from candidate gene to whole genome
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
GENETIC ASSOCIATION STUDIES OF AGE-RELATED MACULAR
DEGENERATION FROM CANDIDATE GENE TO WHOLE GENOME
by
Nicole Tedeschi-Blok
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
(MOLECULAR EPIDEMIOLOGY)
May 2007
Copyright 2007 Nicole Tedeschi-Blok
ii
Dedication
To my family, and especially
Cameo, Chester, Izzy, Tia, and Tim.
iii
Table of Contents
Dedication ii
List of Tables, v
List of Figures, vii
Abstract, viii
Preface, x
Chapter 1: Introduction, 1
Dissertation Overview, 1
Retinal Biology & Aging, 3
Phenotypes of AMD, 6
Dry, 9
Wet, 9
Descriptive Epidemiology, 10
Overall Rates & Risk Factors, 10
Los Angeles Latino Eye Study, 13
Concepts in Pathogenesis, 14
Oxidative Stress, 14
Inflammatory Mechanisms, 16
Previous Genetic Studies, 17
Chapter 2: Complement Factor H Tyr402His and Bilateral Early AMD in
Latinos, 22
Abstract, 22
Introduction, 23
Methods, 26
Results, 28
Discussion, 31
Chapter 3: Manganese Superoxide Dismutase Ala16Val and Complement
Factor H Tyr402His in Early AMD, 35
Background, 35
Methods, 37
Results, 40
Discussion, 45
iv
Chapter 4: Whole Genome Association Study of Early AMD in Latinos
using GeneChip 500K and 100K SNP Microarrays, 51
Introduction, 51
Methods, 54
Results, 57
Discussion, 66
Chapter 5: Proposal to Evaluate the Complex Etiology of AMD from 600K
SNP Microarrays, Smoking, Sunlight, and Diet, 70
Hypothesis & Specific Aims, 70
Background & Significance, 71
Preliminary Studies, 75
Los Angeles Latino Eye Study, 75
Northern European Cohort, 76
Study Design, 77
Methods, 78
Subjects and Samples, 78
Genotyping, 78
Environmental Assessments, 79
Statistical Analyses, 82
Multiple Testing Considerations, Power and Sample Size, 84
Chapter 6: Summary and Conclusions, 87
Bibliography, 91
Appendices:
Appendix A: Smoking History Questionnaire, 101
Appendix B: Sun Exposure Questionnaire, 105
Appendix C: Diet History Questionnaire, 110
v
List of Tables
Table 1: The Clinical Age-Related Maculopathy System (CARMS) 7
Table 2: Summary of candidate genes previously tested for association with
AMD with negative, one positive or mixed findings 21
Table 3: Early AMD case-control comparisons for the CFH Tyr402His
polymorphism: The Los Angeles Latino Eye Study 29
Table 4: Age, birthplace, smoking status and phenotype distribution for
bilateral and unilateral groups based on drusen type and size among
early AMD cases: The Los Angeles Latino Eye Study 30
Table 5: Bilateral and unilateral intermediate to large soft macular drusen
early AMD case-control comparisons for the CFH Tyr402His
polymorphism 31
Table 6: Genotype and allele frequency distributions for MnSOD A16V
polymorphism for early AMD phenotype-specific case-control
comparisons 42
Table 7: Odds ratios for MnSOD A16V and early AMD phenotypes-specific
case-control comparisons controlling for age 42
Table 8: Odds ratio calculations for single genotype exposures of MnSOD
A16V or CFH Y402H and test for interaction with early AMD
phenotypes-specific case-control comparisons controlling for age 43
Table 9: Odds ratio calculations for all combinations of MnSOD A16V and
CFH Y402H genotypes for phenotypes with pigmentary change in
early AMD controlling for age 44
Table 10: Distribution of demographic and clinical features for cases and
controls 57
Table 11: SNP list of most significant associations for genomewide
case-control comparisons (total SNPs = 616,772) 58
Table 12: SNP list of most significant associations for candidate region,
10q23-10qter, case-control comparisons (total SNPs = 9,509) 59
vi
Table 13: SNP list of most significant associations for candidate region,
1q23-q32, case-control comparisons (total SNPs = 13,710) 60
Table 14: Genotype distribution among cases and controls for highly
significant SNP associations in Astrotactin 63
Table 15: Odds ratios for recessive, dominant and log-additive genetic models
of Astrotactin SNPs comparing bilateral early AMD cases with age,
birthplace, smoking status matched controls 64
Table 16: Available participants by ethnically different populations 86
Table 17: Detectable relative risks for gene-environment interactions using
log-additive genetic models with 80% power 86
vii
List of Figures
Figure 1: The anatomy of the eye 5
Figure 2: Examples of age-related maculopathy grades according to the
Clinical Age-Related Maculopathy Staging system 8
Figure 3: Generalization of phenotype progression of AMD with age, genetic
and/or environmental triggers 10
Figure 4: Gender- and race-/ethnicity-specific AMD prevalence rates with
increasing age in the U.S. 14
Figure 5: Simulation of the effect of peroxynitrite amplification on MnSOD
Ala16Val allele 50
Figure 6: Linkage disequilibrium plot in region of highly associated SNPs
(rs6425398, rs4652199, rs4652201, and rs6685449) 65
viii
Abstract
Recent evidence, in Caucasian populations, implicates several complement
genes and tightly linked genes at chromosomal region 10q26 as major genetic
contributors of AMD. The complex nature of AMD etiology may likely explain
inconsistencies in findings across studies. The Los Angeles Latino Eye Study
(LALES), the largest population-based study of eye disease in any racial/ethnic
group in the U.S., was used to study genetic factors with AMD in Latinos in the
present association studies. Using a candidate gene association study, the
Tyr402His polymorphism of complement factor H (CFH) was associated with
bilateral, but not unilateral, early AMD phenotype. In a separate association study
using the LALES case-control population, both CFH Tyr402His and manganese
superoxide dismutase (MnSOD) Ala16Val polymorphisms were simultaneously
considered with early AMD. Present results suggest that CFH and MnSOD may
play dependent roles in early AMD, especially in susceptibility to AMD with
pigmentary abnormalities. Using a whole genome association study with over
600,000 single nucleotide polymorphisms (SNPs), four tightly linked SNPs located
in a previously implicated chromosomal region, 1q25, were significantly associated
with bilateral early AMD (p ≤ 0.005, Bonferoni adjusted). These associated SNPs
reside in the gene for Astrotactin (ASTN), which translates an important neuronal
cell-adhesion molecule responsible for glial-guided migration of neurons. A
biological role for ASTN in AMD is a novel hypothesis that warrants further study.
ix
In addition, association studies that account for interactions between genes and
environment are ultimately important to fully understand the complex etiology of
AMD. This dissertation also proposes a large whole genome gene-environment
association study using high SNP-density and assessment of smoking, sunlight, and
diet in two independent and ethnically distinct populations.
x
Preface
Age-related macular degeneration (AMD) is a progressive disease and the
late-stage is the leading cause of blindness among the elderly. The public health
impact for the prevention of development and progression of AMD is tremendous.
Due to an aging population, the predicted number of persons with AMD in the U.S.
for year 2020 is 3 million [Friedman, O'Colmain et al. 2004]. AMD-related visual
impairment is prevalent in all races and ethnicities, however in general, the greatest
prevalence rates have been observed for Whites. Over 50 percent of blindness and
20 percent of low vision among Whites is due to AMD [Congdon, O'Colmain et al.
2004]. AMD is a complex disease. Linkage and candidate gene association studies
are typically employed to identify genetic contributors of complex disease.
Candidate genes for AMD are numerous - ranging from a handful of causal genes
of monogenic Mendelian types of retinal degenerations to plausible genes specific
to AMD pathobiology. Linkage studies have implicated several genomic regions.
In general findings from previous genetic studies for AMD susceptibility have been
inconsistent. However recently, complement genes and two tightly linked genes at
chromosome 10q26 are reported to be major genetic contributors to AMD based on
studies in Caucasians. The large contribution of these few genetic factors may
suggest that AMD is an oligogenic rather than multigenic disease. However,
continuing research supports that AMD is a multigenic disease. The most powerful
study to assess genetic components of multigenic disease is a high-density whole
xi
genome association study, which combines the strengths of both linkage and
candidate gene association studies. Advances in both high-throughput genotyping
and knowledge of the genome have contributed to the recent feasibility of whole
genome association studies. Genetic study methods used in this dissertation
include both the candidate gene and whole genome approaches using a Latino
population. The final project included in this dissertation is a proposal to evaluate
complex associations with AMD using a whole genome association study with a
high SNP density and important environmental assessments in two independent and
ethnically distinct populations.
1
Chapter 1: Introduction
Dissertation Overview
Chapter one provides background information pertinent to age-related
macular degeneration (AMD) such as retinal biology with aging, phenotype
definitions, descriptive epidemiology, concepts of pathogenesis, and previous
genetic studies. Recently, genes involved in inflammation and immune response
have become most promising candidates for AMD susceptibility. To date, major
genetic contributors in Caucasians are likely to be complement genes and genes
located at chromosome 10q26 [Fisher, Abecasis et al. 2005; Marx 2006].
Present genetic association studies were conducted using subjects from
within the Los Angeles Latino Eye Study (LALES), the largest population-based
study of eye disease in any racial/ethnic group in the U.S. [Varma, Paz et al. 2004].
Chapter two consists of an association study considering the complement factor H
(CFH) Tyr402His polymorphism. Recent literature supports that the Tyr402His
polymorphism of CFH may play a substantial role in AMD etiology involving
inflammatory mechanisms [Marx 2006]. Presently, this CFH polymorphism was
genotyped among early AMD cases and controls from LALES. Evidence for an
association of CFH Tyr402His with bilateral, but not unilateral, early AMD
phenotype was shown in this population-based study.
Chapter three consists of another association study simultaneously
considering both CFH Tyr402His and manganese superoxide dismutase (MnSOD)
2
Ala16Val polymorphisms with early AMD. Oxidative stress is an important
concept in AMD pathogenesis [Zarbin 2004]. Although it has not been reproduced,
the Ala16Val polymorphism of MnSOD has previously been associated with AMD
[Kimura, Isashiki et al. 2000]. Present results suggest a role for gene-gene
interaction between CFH and MnSOD, especially for cases with pigmentary
abnormalities compared to controls. More specifically, carrying both the
homozygous Ala16 MnSOD genotype and at least one His402 CFH allele is more
frequently observed among early AMD cases as compared to controls. Possible
mechanisms for the suggested gene-gene interaction are discussed.
Chapter four is a whole genome association study using bilateral early
AMD cases and controls from LALES. Genotyping was performed using
Affymetrix GeneChip 500K and 100K SNP set microarrays (GeneChip 500K and
100K). For over 600,000 single nucleotide polymorphisms (SNPs), univariate
logistic regression models were used to identify SNPs associated with bilateral
early AMD using Genetrix software. Bonferroni adjustments of p-values were
used to control for multiple comparisons. For highly significant case-control
differences, conditional logistic regression models were used to further explore
case-control differences and account for matching variables (age, birthplace, and
smoking status) using SAS software. Results suggest a role for the important
neronal cell-adhesion molecule, Astrotactin, which is located in a previously
implicated chromosomal region, 1q25 [Fisher, Abecasis et al. 2005].
3
Chapter five is a summary of the present projects. This chapter compares
present findings in Latinos to previously reported genetic contributors of AMD in
Caucasians. The idea that susceptibility to AMD in Latinos may include similar as
well as unique pathways as that in Caucasians is discussed. Finally, the finding
that Astrotactin may play a role in AMD development in Latinos warrants
immediate further research in other races/ethnicities.
A whole genome association study that also explores interactions with
important environment factors is likely to be the logical next advanced
epidemiologic tool that may further elucidate the complex etiology of AMD.
Chapter six proposes a whole genome association study using two ethnically
different case-control populations: 1) Latinos residing in the U.S. and 2)
Caucasians residing in northern Europe. Environmental factors that will also be
assessed include smoking, sunlight and diet.
Retinal Biology and Aging
Basic anatomy of the eye is shown in figure 1. Detailed retinal biology
pertinent to macular degeneration is found in the book edited by Purves et al., 2001
[Purves, Augustine et al. 2001]. Briefly, the site of primary pathology in AMD is
the retinal pigmented epithelium (RPE), a monolayer of cells situated between the
retinal layers and the choroid. The choroid, a vascular layer that separates the
retina from the sclera and provides nourishment to the back of the eye, is separated
4
from the RPE by Bruch’s membrane. RPE provide nutrients and maintenance for
adjacent photoreceptors, which may be categorized as either rods or cones. The
highest concentration of cone photoreceptors is in the area of the macula, which is
the focal point for light in the retina and is responsible for vision necessary to focus
on objects or text. Photoreceptor outer segments, which are surrounded by apical
processes of the RPE, are composed of densely stacked membrane discs embedded
with photosensitive complexes producing electrical signals carried to the brain via
the optic nerve. As photoreceptor outer segments are renewed by adding new discs
at the inner segment-outer segment interface, old discs at the outer segment-RPE
interface are phagocytosed by surrounding RPE apical processes. These old discs
are broken down by lysis in a diurnal cycle where there is a spike in old disc
phagocytosis with light exposure in the morning.
Fundus photography (see Figure 2) is a diagnostic tool that captures
macroscopic details of retinal anatomy (see Figure 1). In AMD pathology, it is
generally accepted that first RPE dysfunction and undergo apoptosis followed by
degeneration of photoreceptors (initiating in the macula area) and loss of vision
beginning with the ability to focus on objects or text. During senescence or aging,
postmitotic cells, such as RPE, accumulate lipofuscin, sacs of auto-fluorescent
lipid-protein aggregates that are photoinducible free radical generators [Boulton
and Dayhaw-Barker 2001]. A component of lipfuscin that is directly implicated in
AMD due to its phototoxic potential is N-retinyl-N-retinylidene ethanolamine
5
(A2E) [Sparrow, Fishkin et al. 2003]. A2E is a by-product of the visual cycle
formed from components of photoreceptor outer segments, 11-cis-retinal and
phosphatidylethanolamine [Sparrow, Fishkin et al. 2003; Lamb and Simon 2004].
Also with aging, drusen may accumulate between the RPE basal lamina and the
inner collagenous layer of Bruch’s membrane. Small hard druse accumulation of
less than 63 um in diameter are common, and are not considered pathologic [Sarks,
Arnold et al. 1999]. Another aspect of aging biology of the eye that may also incur
pathologic effects is decreased blood flow in the choroid [Zarbin 2004].
Figure 1: The anatomy of the eye. The macula, or fovea, is the area of the retina
at which light is focused. In AMD, photoreceptor degeneration initiates in the macula.
(Drawing freely available at National Eye Institute).
6
Phenotypes of AMD
Classification of AMD is complex and differential methods of classification
have hindered comparisons across epidemiological studies. AMD is diagnosed
using both fundus photography and clinical assessments. However, the
International Age-Related Maculopathy (ARM) Epidemiological Study Group has
implemented an international classification and grading system [Bird, Bressler et
al. 1995]. Standard grading of fundus photographs for AMD is accomplished using
the Wisconsin Age-Related Maculopathy Grading Scheme (WARMGS) [Klein,
Davis et al. 1991]. Drusen is considered the "hallmark" of AMD. Complex
assessments of drusen and other phenotypes are considered in AMD diagnoses.
Recently, the clinical age-related maculopathy staging system (CARMS) was
designed to provide a simple and reliable system for both clinical and research
purposes [Seddon, Sharma et al. 2006]. Table 1 and figure 2 summarize and show
phenotype examples as classified by CARMS. The following sections discuss
phenotypes of AMD which are generally categorized by dry and wet forms.
7
Table 1: The Clinical Age-Related Maculopathy System (CARMS).
Grade of
Maculopathy
Clinical Features
1 No drusen or < 10 small drusen without pigment abnormalities
2 Approximately ≥10 small drusen or <15 intermediate drusen, or pigment abnormalities
associated with ARM
a. Drusen
b. RPE changes (hyperpigmentation and hypopigmentation)
c. Both drusen and RPE changes
3 Approximately ≥15 intermediate drusen or any large drusen
a. No drusenoid RPED
b. Drusenoid RPED
4 Geographic atrophy with involvement of the macular center, or noncentral geographic atrophy
at least 350 µm in size
5 Exudative AMD, including nondrusenoid pigment epithelial detachments, serous or
hemorrhagic retinal detachments, CNVM with subretinal or sub-RPE hemorrhages or fibrosis,
or scars consistent with treatment of AMD
a. Serous RPED, without CNVM
b. CNVM or disciform scar
AMD = age-related macular degeneration; ARM = age-related maculopathy; CNVM = choroidal neovascular
membrane: RPE = retinal pigment epithelium; RPED = retinal pigment epithelial detachment. Small = drusen <
63 mm in diameter located within 2 disc diameters (DDs) of the center of the macula; intermediate = drusen ≥
63 mm but < 125 mm, located within 2 DDs of the center of the macula; large = drusen ≥ 125 mm in diameter
located within 2 DDs of the center of the macula; drusenoid RPED = confluent soft drusen ≥ 500 mm in size.
Reprinted [Seddon, Sharma et al. 2006] with permission of Ophthalmology. See accompanying Fig. 2.
8
Figure 2: Examples of age-related maculopathy grades according to the Clinical Age-Related Maculopathy
Staging system. 1: No drusen. 2a: Several small drusen and no retinal pigment epithelial changes. 2b: Retinal
pigment epithelial alteration, but no drusen. 2c: Both small drusen and retinal pigment epithelial changes. 3a:
Several intermediate-size and large drusen. 3b: Drusenoid retinal pigment epithelial detachment. 4: Geographic
atrophy. 5b: Choroidal neovascular membrane with disciform scar. (5a is a serous retinal pigment epithelial
detachment.) Reprinted [Seddon, Sharma et al. 2006] with permission of Ophthalmology.
9
Dry
The early stage of AMD is a dry form. In general, the early stage of AMD,
also known as age-related maculopathy (ARM), is defined by the absence of signs
of late AMD and the prescence of soft indistinct or reticular drusen or by the
presence of any druse type except hard indistinct, with RPE depigmentation or
increased retinal pigment in the macular area. Druse size is commonly referred to
as intermediate (between 63 and 125 microns in diameter) or large (125 microns or
greater in diameter). Drusen associated with AMD is somewhat similar to that
associated with aging but also have distinct differences, such as the presence of
several complement components [Zarbin 2004]. Early AMD may progress to a dry
form of late AMD, also known as geographic atrophy (GA). In GA, greater
amounts of drusen accumulate and RPE atrophy in a pattern that takes on a
geographic two-dimensional appearance as observed in fundus photographs (see
Figure 2).
Wet
Early AMD may alternately progress to the wet form, also known as
exudative AMD and characterized by choroidal neovasculariztion (CNV) (see
Figure 2). Retinal pigment epithelial detachment or serous detachment of the
sensory retina, subretinal or sub-RPE hemorrhages, and subretinal fibrous scars are
all signs of exudative AMD. Figure 3 summarizes some important AMD
10
nomenclature in a generalization of progression of AMD with genetic and/or
environmental triggers in aging.
Descriptive Epidemiology
Overall Rates & Risk Factors
Figure 4 shows the National Eye Institute reported prevalence rates of AMD
with increasing age for Whites, Blacks, and Hispanics in the U.S. Rates of early
and late AMD phenotypes vary by ethnicity and geographic location [Klein, Peto et
al. 2004]. From the Beaver Dam Study, 10-year incidence rates in an older
population with 99% Whites and 43 to 86 years of age at baseline, were 14.0%,
8.8%, 8.5%, 6.3%, 0.6%, and 0.9% for intermediate drusen, large drusen,
hyperpigmentation, hypopigmentation, GA, and CNV; respectively [Klein, Klein et
al. 2002]. In a Japanese population of participants 50 years and older (65 years,
65+ 45+
Figure 3: Generalization of phenotype progression of AMD with age, and genetic and/or environmental
triggers. While AMD pathogenesis of AMD is uncertain, Early AMD may progress from non-pathologic
drusen accumulation. Early AMD may or may not progress to Late AMD. De novo Late AMD may also
occur (not represented here). GE triggers = Genetic and/or environmental triggers. GA = Geographic
Atrophy. CNV = Choroidal Neovascularization. Drusen may occur in non-pathologic aging that may
progress to pathologic aging lesions of early AMD. GA is a direct progression from early AMD. CNV is an
indirect progression from early AMD.
Aging Early AMD (Dry) Late AMD
None or some Drusen Drusen GA (Dry)
Pigmentary Abnormalities CNV (Wet)
GE triggers GE triggers
Age (years)
11
mean age), rates were 0.5%, 3.2%, 0.2%, and 0.7% for intermediate drusen, any
pigmentary abnormalities, GA, and CNV; respectively [Oshima, Ishibashi et al.
2001]. In an Iceland population of participants 50 years and older, rates were
15.1%, 8.2%, 0.3%, and 0.4% for early AMD, any pigmentary abnormalities, GA,
and CNV; respectively [Jonasson, Arnarsson et al. 2003]. Overall, in a combined
Eurpoean population of participants 65 years and older, rates were 15.4%, 1.2%,
and 2.3% for large drusen, GA, and CNV; respectively [Augood, Vingerling et al.
2006].
In general, progression from early to late AMD is more likely among
Whites than Blacks or Hispanics. For some studies [Klein, Klein et al. 1999; Klein,
Peto et al. 2004], rates of early AMD phenotypes are similar in Whites, Blacks and
Hispanics, while Whites have higher rates of late AMD than Blacks [Klein, Peto et
al. 2004]. The Multi-Ethnic Study of Atherosclerosis (MESA) is a population-
based study of participants 45 to 85 years of age, in which the prevalence of AMD
was compared for Whites (63.0 years, mean age), Blacks (62.4 years, mean age),
Hispanics (61.6 years, mean age), and Chinese (62.4 years, mean age) living within
the U.S. [Klein, Klein et al. 2006]. The prevalence of late AMD reported for
MESA was 0.6%, 0.3%, 0.2%, and 1.0% for Whites, Blacks, Hispanics, and
Chinese; respectively [Klein, Klein et al. 2006]. Klein et al., 2006 [Klein, Klein et
al. 2006], report that for Chinese compared to Whites, CNV was 4.3-times more
prevalent (95% CI = 1.3 to 14.3), although this needs further investigation.
12
After adjusting for age, the most consistent environmental risk factor for
AMD is tobacco smoking. Hypertension is semi-consistently found to increase risk
where uncontrolled hypertension compared to normal pressure increases risk of wet
AMD by 3-fold while uncontrolled hypertension contributes to over 30% of all wet
AMD [Evans 2001; Klein, Peto et al. 2004]. With less consistency, sunlight and
diet may influence risk of AMD [Hawkins, Bird et al. 1999]. Previous studies
suggest antioxidant intake provides protection against AMD [EDCCSG 1992;
EDCCSG 1993; Seddon, Ajani et al. 1994]. The Beaver Dam Eye Study has
previously found both a weak protective effect of zinc intake with early AMD
[Mares-Perlman, Klein et al. 1996] although, in a subsequent study, they found no
overall association of zinc or antioxidants with AMD [VandenLangenberg, Mares-
Perlman et al. 1998]. The Blue Mountain Eye Study found no overall association
of antioxidants and AMD with Autstralian participants [Smith, Mitchell et al. 1999;
Flood, Smith et al. 2002; Kuzniarz, Mitchell et al. 2002]. However, several recent
studies suggest a protective effect of diet and serum levels of antioxidants against
AMD [Delcourt, Cristol et al. 1999; Mares-Perlman, Fisher et al. 2001; Simonelli,
Zarrilli et al. 2002; Snellen, Verbeek et al. 2002; Falsini, Piccardi et al. 2003].
Dietary fat and cholesterol intake may also play role in the development of AMD
[Haddad, Chen et al. 2006]. Furthermore, serum cholesterol as well as
atherosclerosis [McCarty, Mukesh et al. 2001] and cardiovascular disease [Smith,
Assink et al. 2001] have all been positively associated with AMD.
13
Los Angeles Latino Eye Study
The Los Angeles Latino Eye Study (LALES) is a cross-sectional study that
was initially designed to survey health care and eye disease among Latinos in Los
Angeles. LALES includes over 6357 eligible volunteers age 40 years or older
(54.6, mean age) from the city of La Puente [Varma, Paz et al. 2004]. The
participation rate in LALES was 82 percent. The prevalence rates for intermediate
drusen, large drusen, RPE abnormalities, and overall early AMD were 23.0%,
14.5%, 6.0%, and 9.4%; respectively [Varma, Fraser-Bell et al. 2004]. For late
AMD, prevalence rates were 0.15% and 0.29% for GA and CNV; respectively
[Varma, Fraser-Bell et al. 2004]. Lifestyle factors associated with AMD in LALES
include smoking, frequent alcohol drinking, wine consumption, and estrogen use
[Fraser-Bell, Wu et al. 2006]. Smoking slightly increased risk of early AMD (OR
= 1.2, 95% CI = 1.01 to 1.4 for ever versus never smoked). Early AMD cases with
RPE abnormalities were 1.8-times more likely to report heavy alcohol drinking
(95% CI = 1.1 to 3.2). Interestingly, Fraser-Bell and colleagues show that when
alcohol type was considered, wine consumption was protective of RPE
abnormalities (OR = 0.7, 95% CI = 0.4 to 0.98). Also, Fraser-Bell et al., 2006
[Fraser-Bell, Wu et al. 2006], found that among LALES female participants,
estrogen use was protective of soft indistinct drusen in early AMD (OR = 0.5, 95%
CI = 0.3 to 0.8).
14
Rates of AMD
0.00
0.02
0.04
0.06
0.08
0.10
0.12
0.14
0.16
0.18
0.20
Female
55-59
65-69
75-79
85+
Male
55-59
65-69
75-79
85+
Age Groups
Prevalence (%)
White
Black
Hispanic
Figure 4: Gender- and race-/ethnicity-specific AMD prevalence with increasing
age in the U.S. as reported by the National Eye Institute.
Concepts in Pathogenesis
Oxidative Stress
It is well established that oxidative injury is associated with general aging.
Although not necessarily pathological, many age-associated changes to cells may
be a direct product of cumulative oxidative stress [Zarbin 2004]. The retina is
particularly susceptible to oxidative stress. Especially in the area of the macula, the
retina is regularly under photo-oxidative insult and the high content of
15
polyunsaturated fatty acids (PUFA) in retinal membranes provides an excellent
source for oxidative modifications [Beatty, Koh et al. 2000]. Also, the normal
function of phagocytosis of spent photoreceptor outer segments by RPE is a
process that generates reactive oxygen species, primarily hydrogen peroxide. In
addition, lipofuscin, which accumulates in the RPE during senescence from
undegradable photoreceptor outer segments, produces reactive oxygen species
[Terrasa, Guajardo et al. 2003]. Specifically for these reasons, the retina constantly
depends heavily on antioxidants, both endogenous enzymes (superoxide dismutase,
catalase, and glutathione) and exogenous supplements (vitamins C & E and
carotenoids), to minimize oxidative damage. Some degree of oxidative insult is
unavoidable in the retina and therefore, oxidative stress is a plausible hypothesis in
the etiology of AMD. This hypothesis is further supported by the evidence of
oxidized lipids found in drusen, especially for AMD patients [Crabb, Miyagi et al.
2002]. Specifically, Crabb and colleagues [Crabb, Miyagi et al. 2002] showed that
drusen from AMD patients was immunopositive for products of oxidatively
modified docosahexaenoic acid, a highly unsaturated fatty acid making up 50
percent of rod photoreceptor phospholipids, for 82 percent (9 of 11 samples)
compared to 18 percent (2 of 11 samples) of age-matched controls [Crabb, Miyagi
et al. 2002].
16
Inflammatory Mechanisms
Increasing evidence suggests that pathological changes may be a
consequence of inflammation and immune response to retinal injury. In addition to
oxidative modified lipids, several inflammatory and immune-related components
are found in drusen [Donoso, Kim et al. 2006]. Initial RPE injury, which may
result from oxidative stress, may trigger an immune cascade ultimately leading to a
state of chronic inflammation within Bruch's membrane and the choroid [Hageman,
Luthert et al. 2001; Zarbin 2004]. Chronic inflammation is likely to contribute to
abnormalities in the extracellular matrix (ECM), which in conjunction with the
choroid is responsible for supplying the RPE with nutrients for proper function
[Zarbin 2004]. Additional injury to the RPE is likely to result from alterations to
the ECM and choroid vasculature. Further, recent evidence from several genetic
studies show an association of complement factor H (CFH) with AMD in
Caucasian populations [Conley, Thalamuthu et al. 2005; Edwards, Ritter et al.
2005; Hageman, Anderson et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss et
al. 2005; Zareparsi, Branham et al. 2005; Schaumberg, Christen et al. 2006]. CFH
is an important regulator of immune response and is also a component found in
druse [Donoso, Kim et al. 2006]. For these reasons, the inflammation/immune
response is a compelling hypothesis of AMD etiology.
17
Previous Genetic Studies
Genetic studies of AMD have been thoroughly reviewed in Haddad et al.,
2006 [Haddad, Chen et al. 2006]. Briefly, considerable evidence reveals that AMD
is a heritable disease such that first-degree case relatives versus controls have a 2.4-
fold greater risk of overall AMD and 3.1-fold greater risk of exudative AMD (95%
CI = 1.2 to 4.7 and 1.5 to 6.7 for overall and exudative AMD, respectively)
[Haddad, Chen et al. 2006]. From the population-based study by Seddon et al.,
2005 [Seddon, Cote et al. 2005], using U.S. male twins, heritability estimates are
0.46 for any AMD and 0.71 for late AMD. Previously, candidate genes for AMD
have not shown consistent associations across studies. Numerous previous
suspected candidate genes for AMD range from a handful of causal genes of
monogenic Mendelian types of retinal degenerations to plausible genes specific to
AMD pathobiology (See table 2).
Several family-based genome-wide scans implicate different chromosomal
regions, which may be explained by the multifactorial nature of AMD. A recent
genome-scan meta-analysis (GSMA) of six genome-wide scans was performed
using a rank summed analysis weighted by the square root of the number of
affected individuals in each study to give smaller, lower-powered studies less
influence [Fisher, Abecasis et al. 2005]. Fisher and colleagues [Fisher, Abecasis et
al. 2005] report that the two most important chromosomal regions in AMD
susceptibility are 10q and 1q. More specifically, 10q26-10qter (p < 0.0004) is
18
statistically significantly associated and 10q23-10q26 (p < 0.008) is borderline
statistically significantly associated with AMD after adjusting for multiple
comparisons [Fisher, Abecasis et al. 2005]. The third most significantly associated
region by GSMA is 1q23-q31 (p = 0.01) [Fisher, Abecasis et al. 2005]. The
adjacent region, 1q31-1q32 (p = 0.02), was also among other significantly
associated chromosomal regions, which include 2p, 3p, 3q, 4q, 12q, and 16q
[Fisher, Abecasis et al. 2005].
The CFH gene, located at 1q32, is a candidate gene that, recently, has most
consistently been associated with risk of AMD [Conley, Thalamuthu et al. 2005;
Edwards, Ritter et al. 2005; Hageman, Anderson et al. 2005; Haines, Hauser et al.
2005; Klein, Zeiss et al. 2005; Zareparsi, Branham et al. 2005; Schaumberg,
Christen et al. 2006]. Recent clinic-based studies provide evidence suggesting that
43 to 70% of AMD in the population may be attributed to a single genetic
component, the CFH Tyr402His polymorphism [Conley, Thalamuthu et al. 2005;
Edwards, Ritter et al. 2005; Hageman, Anderson et al. 2005; Haines, Hauser et al.
2005; Klein, Zeiss et al. 2005; Zareparsi, Branham et al. 2005]. Specifically, these
studies [Conley, Thalamuthu et al. 2005; Edwards, Ritter et al. 2005; Hageman,
Anderson et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss et al. 2005;
Zareparsi, Branham et al. 2005] observed AMD cases to be 2.1 to 7.4-times more
likely to carry one or more copies of the His402 allele compared to controls. Using
a case-control study nested within the prospective cohort, the Physician's Health
19
Study, Schaumberg and colleagues, confirmed an association between CFH
Tyr402His in White males [Schaumberg, Christen et al. 2006]. In contrast, Gotoh
and colleagues showed no association between CFH polymorphisms and exudative
AMD in Japanese and suggest that association of CFH with AMD may be specific
to Caucasian populations [Gotoh, Yamada et al. 2006]. CFH, an important
regulator of innate immunity, plays a critical role in protecting intact host cells
while allowing destruction of foreign or damaged host cells [Donoso, Kim et al.
2006].
Other complement genes that have recently been implicated in AMD
include complement factor B (CFB) and complement component 2 (C2). CFB and
C2, located 500 base pairs apart from each other at chromosome 6p, are two
proteins involved in the same pathway as CFH. Interestingly, two protective
haplotypes were observed for CFB and C2 [Haddad, Chen et al. 2006]. Hageman
and colleagues found that cases were 55 percent less likely than controls to carry
polymorphisms CFB Lys9His and C2 Glu318Asp (OR = 0.45, 95% CI = 0.33 to
0.61) [Gold, Merriam et al. 2006]. Cases were also 64 percent less likely to carry
CFB Arg32Gln and an intron 10 of C2 polymorphisms (OR = 0.36, 95% CI = 0.23-
0.56) [Gold, Merriam et al. 2006]. Other recently identified genetic candidates
include pleckstrin homology domain containing family A member 1 (PLEKHA1)
or a tightly linked gene, LOC387715, located at 10q26 [Haddad, Chen et al. 2006;
Marx 2006]. Evidence for PLEKHA1/ LOC387715, which possibly functions to
20
activate lymphocytes, suggests that one or more variations may explain
approximately 40% of AMD in the population [Marx 2006].
Table 2: Summary of candidate genes previously tested for association with AMD with negative,
one positive or mixed findings*
No Association Association in One Study Inconsistent Association
Gene Name Symbol Gene Name Symbol Gene Name Symbol
fibulin 4 EFEMP2
cystatin 3 (amyloid
angiopathy and cerebral
hemorrhage) CST3
ATP-binding cassette, sub-
family A member 4 ABCA4
fibulin 1 FBLN1
chemokine (C-X3-C)
receptor 1 CX3CR1
angiotensin I converting
enzyme ACE
fibulin 2 FBLN2 fibulin 5 FBLN5 apolipoprotein E APOE
G protein-coupled receptor 75 GPR75
major histocompatibility
complex HLA genes
elongation of very long chain
fatty acids (stargardt disease
3, autosomal dominant) ELOVL4
interphotoreceptor matrix
proteoglycan 2 IMPG2
low density lipoprotein-
related protein 6 LRP6 fibulin 6 HEMICENTIN-1
laminin, beta 3 LAMB3 matrix metallopeptidase 9 MMP9 paraoxonase PON1
laminin, gamma 1 LAMC1 toll-like receptor 4 TLR4
manganese superoxide
dismutase SOD2
laminin, gamma 2 LAMC2
vascular endothelial growth
factor VEGF
very low density lipoprotein
receptor VLDLR
retinal degeneration, slow
(retinitis pigmentosa 7) RDS
tissue inhibitor of
metalloproteinase 3 (Sorsby
fundus dystrophy,
pseudoinflammatory) TIMP3
vitelliform macular dystrophy
2 (Best disease, bestrophin) VMD2
* Summarized [Haddad, Chen et al. 2006]
21
22
Chapter 2: Complement Factor H Tyr402His and
Bilateral Early AMD in Latinos
ABSTRACT
Recently, a strong association was observed for the complement factor H (CFH)
Tyr402His polymorphism with early and advanced age-related macular
degeneration (AMD) in independent non-Hispanic White clinic-based case-control
studies. These studies suggest the CFH His402 allele is a major risk allele in early
and advanced AMD, explaining 43% to 70% of all AMD in older adults. To
evaluate the CFH Tyr402His polymorphism for an association with early AMD
phenotypes, a population-based case-control study design of early AMD among
Latinos/Hispanics was utilized. This study cohort consisted of 285 early AMD
cases and 570 controls matched on age, birthplace and smoking status. Genotype
determination was performed by allele-specific digestion of polymerase chain
reaction products. No overall statistically significant association with early AMD
was observed among Latinos. However, a subset of early AMD cases that have
bilateral, but not unilateral, intermediate to large soft macular drusen were 1.8
times more likely to carry either the homozygous or heterozygous His402
genotype. The present data suggest that the CFH Tyr402His is not a major risk
23
factor for overall early AMD in this Latino population, but may play a role in
susceptibility to phenotypes of early AMD likely to progress to late AMD.
INTRODUCTION
Age-related macular degeneration (AMD) is a complex multigenic,
multifactorial disease that is typically bilateral. While the accumulation of drusen
is known as the "hallmark" of AMD, certain drusen types of limited size, location,
and extent occur in normal, non-pathologic aging processes. Early stage AMD is
characterized by the absence of signs of late stage AMD and the presence of (1)
soft, indistinct, or reticular drusen or (2) any drusen (except hard, indistinct) with
pigmentary changes in the macular region. Vision loss is minimal in early AMD,
but may advance to significant vision loss or blindness in late AMD. Late AMD,
characterized by geographic atrophy (GA), or choroidal neovascularization (CNV),
is the leading cause of blindness in the U.S among the elderly. Alternatively, some
patients diagnosed with early AMD may never develop signs of late AMD and
consequently, will never develop significant vision loss. Thus, in order to identify
those patients at greatest risk for significant vision loss caused by AMD, the
challenge is to correlate genotypes with early AMD phenotypes most likely to
progress to late AMD.
Bilaterality has been shown to be an important characteristic of those early
AMD patients most likely to progress to late AMD in various population-based
24
studies. Mukesh and colleagues, found that the risk of progression is 5-fold greater
for those with bilateral early AMD compared to those with unilateral early AMD
[Mukesh, Dimitrov et al. 2004]. In another population-based incidence study,
Wang and colleagues, concluded that the risk of blindness associated with AMD
was likely due to bilateral age-related maculopathy lesions [Wang, Mitchell et al.
1998]. They also suggested, based on self-reported AMD family history, that
persons with bilateral early AMD are likely to have a stronger genetic basis than
those with unilateral disease [Wang, Mitchell et al. 1998]. Furthermore, based on
long term follow-up of Age-Related Eye Disease Study (AREDS) participants, a
simplified clinical scale has recently been developed to assess the 5-year risk of
progression to late AMD, in which risk of progression markedly increases with
increasing bilateral lesions [Ferris, Davis et al. 2005].
Familial and linkage studies provide strong evidence of a genetic basis for
AMD [De Jong, Klaver et al. 1997; Seddon, Ajani et al. 1997; Klaver, Wolfs et al.
1998; Seddon, Santangelo et al. 2003; Seddon, Cote et al. 2005]. Seddon and
colleagues, reported that genetic factors may explain as much as 71% of the
variation in overall disease severity in AMD [Seddon, Cote et al. 2005]. It is
generally thought that the genetic component of AMD results from multiple factors
of variable effect, as is typical for any complex disease. Therefore, AMD candidate
genes are numerous, ranging from those involved in monogenic retinal dystrophies
to those involved in inflammatory, apoptotic, or oxidative stress-related pathways
25
[Tuo, Bojanowski et al. 2004; Zarbin 2004]. Recent clinic-based studies provide
evidence suggesting that 43 to 70% of AMD in the population may be attributed to
a single genetic component, the complement factor H (CFH) Tyr402His
polymorphism [Conley, Thalamuthu et al. 2005; Edwards, Ritter et al. 2005;
Hageman, Anderson et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss et al.
2005; Zareparsi, Branham et al. 2005]. Specifically, these studies observed AMD
cases to be 2.1 to 7.4-times more likely to carry one or more copies of the His402
allele compared to controls [Conley, Thalamuthu et al. 2005; Edwards, Ritter et al.
2005; Hageman, Anderson et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss et
al. 2005; Zareparsi, Branham et al. 2005]. Using a case-control study nested within
the prospective cohort, the Physician's Health Study, Schaumberg and colleagues,
confirmed an association between CFH Tyr402His in White males [Schaumberg,
Christen et al. 2006]. In contrast, Gotoh and colleagues showed no association
between CFH polymorphisms and exudative AMD in Japanese [Gotoh, Yamada et
al. 2006]. The present study tests the association of the CFH Tyr402His
polymorphism with early AMD using cases and controls nested within the Los
Angeles Latino Eye Study (LALES), the largest population-based study of eye
disease in any racial/ethnic group in the U.S. [Varma, Fraser-Bell et al. 2004;
Varma, Paz et al. 2004].
26
METHODS
Subjects were ascertained from within LALES, a population-based cross-
sectional study originally designed to assess the prevalence of visual impairment,
ocular disease, and visual functioning in Latinos. Details of LALES study design is
described elsewhere [Varma, Paz et al. 2004]. Briefly, all procedures conformed to
the Declaration of Helsinki for research involving human subjects. The Institutional
Review Board of the University of Southern California approved the project, and
informed consent was obtained from all participants. From February 2000 to May
2003, 6357 Latino participants age 40 years and older from 6 census tracts in Los
Angeles, California, were recruited. The participation rate in LALES was 82
percent. All LALES participants underwent complete eye exams including fundus
photographs. Diagnosis and grading of AMD was performed in a masked manner
as defined according to the Wisconsin Age-Related Maculopathy Grading System
(WARMGS) [Klein, Davis et al. 1991; Klein, Klein et al. 1992]. Whole genome
amplification of DNA from stored blood samples, taken at the time of the eye exam
for all participants entering the study from March, 2001 and beyond, was
performed by multiple displacement amplification technology (Repli-g Service,
Qiagen Sciences, Germantown, MD) and resulted in validated, usable quality for
genotyping assays among 98% of all LALES samples. Early AMD cases and
controls without advanced diabetic retinopathy among LALES participants were
eligible for this study if validated, usable whole genome amplification material was
27
available. Only 8 early AMD cases without advanced diabetic retinopathy were
excluded from the study due to inferior quality of whole genome amplification
material. All eligible early AMD cases (N = 285) from which DNA was collected
were genotyped for the CFH Tyr402His polymorphism. Our study did not include
late AMD cases, since the number of affected individuals in this population was too
small [Varma, Fraser-Bell et al. 2004]. From the remaining LALES participants
(entering the study as of May, 2001, without any AMD or advanced diabetic
retinopathy, 570 randomly matched controls, matched by age (exact or nearest
available year), smoking status (smoker or nonsmoker), and birthplace (U.S.,
Mexico or other) were similarly genotyped for the CFH Tyr402His polymorphism.
CFH Tyr402His genotyping was performed without knowledge of case
status by PCR amplification and restriction enzyme digestion. A 179-bp PCR
product containing the Tyr402His region was amplified in a 50uL reaction with
200 nM of forward primer, 5' TCATTGTTATGGTCCTTAGGAAA3', and reverse
primer, 5' GGAGTAGGAGACCAGCCATT3', 500uM MgCl2, 1X PCR buffer,
and 1 Unit Taq DNA polymerase (Accuprime Taq DNA Polymerase System,
Invitrogen, Carlsbad, CA). PCR amplification was performed with 40 cycles of
30s at 94C, 30s at 58C, and 30s at 68C (GeneAmp 9700 Thermacycler, Applied
Biosystems, Foster City, CA). Purified PCR product was digested with 10 Units of
Tsp509I (New England Biolabs, Ipswich, MA) for 3 hrs at 65C. The digested PCR
products were resolved with 4% NuSieve Agarose (Latitude HT precast gels,
28
Cambrex, East Rutherford, New Jersey) and stained with ethidium bromide.
Automated direct sequencing (3100 Genetic Analyzer, Applied Biosystems, Foster
City, CA) was performed on the reverse strand of the undigested 179-bp PCR
product for selected samples to validate the restriction enzyme assay or to clarify
ambiguous genotypes.
SAS version 9.1 (SAS Institute, Cary, NC), was used to perform all
statistical analyses. The χ
2
test for Hardy-Weinberg equilibrium were carried out
for frequency distributions of genotypes among cases and controls separately.
Conditional logistic regression models were used to compare genotypes or alleles
between cases and their matched controls. Odds ratios and 95% confidence
intervals were calculated for multiplicative, additive, dominant and recessive
models of the minor allele. All statistical tests were performed using a 5%
significance level.
RESULTS
Genotyping data for the 285 cases and 570 controls were in agreement with
the Hardy-Weinberg equilibrium (p = 1.00 for cases and p = 0.49 for controls; table
3). In contrast to previously published clinic-based case-control studies [Conley,
Thalamuthu et al. 2005; Edwards, Ritter et al. 2005; Hageman, Anderson et al.
2005; Haines, Hauser et al. 2005; Klein, Zeiss et al. 2005; Zareparsi, Branham et al.
2005], overall, no apparent statistically significant association of the CFH
29
Tyr402His polymorphism with AMD was observed (table 3). Although, borderline
statistical significance was observed where cases were 1.3-times more likely to
have an additional His402 allele compared to controls (95% C.I. = 0.99 - 1.67, p-
value for apparent dose effect = 0.06; table 3).
Table 3: Early AMD case-control comparisons for the CFH Tyr402His
polymorphism: The Los Angeles Latino Eye Study
Frequency (%)
Model Controls
N = 570
Cases
N = 285
OR
a
95% CI
b
YY 391 (68.6) 180 (63.2) 1 --
HY 165 (29.0) 93 (32.6) 1.2 0.90 to 1.69
HH 14 (2.5) 12 (4.2) 1.9 0.84 to 4.17
H-W p
c
= 0.49
H-W
p
c
= 1.00 p
d
= 0.06
H allele 0.170 0.205 1.3
e
0.99 to 1.67
1.3
f
0.95 to 1.74
1.8
g
0.80 to 3.87
a
Age-, smoking-, birthplace-adjusted odds ratio.
b
CI = Confidence interval.
c
Hardy-Weinberg p-
value.
d
p-Value for apparent dose effect.
e
Additive.
f
Dominant.
g
Recessive.
Since bilaterality in early AMD is an important feature that increases risk of
progression to advanced AMD [Wang, Mitchell et al. 1998; 2001], the cases were
re-categorized based on the presence of soft distinct, soft indistinct or reticular
drusen of sizes greater than 63 microns in diameter. Cases were stratified by (1)
bilateral or (2) unilateral early AMD (table 4). Table 5 shows results for separate
case-control comparisons using bilateral (N = 101) or unilateral (N = 159) cases
and age-, smoking status-, and birthplace-matched controls (N = 360 and 476
matched controls for bilateral and unilateral cases, respectively). Results show that
30
the bilateral early AMD cases are 1.8-times more likely to carry at least one copy of
the His402 allele (p = 0.03) compared to controls (table 5). However, unilateral
early AMD cases are equally likely to carry the His402 allele compared to controls
(table 5).
Table 4: Age, birthplace, smoking status and phenotype distribution for bilateral
and unilateral groups based on drusen type and size among early AMD cases: The
Los Angeles Latino Eye Study
Bilateral Unilateral
Controls
N = 202
Cases
N = 101 p
1
Controls
N = 318
Cases
N = 159 p
1
Age (mean ± SE
2
) 60.7 ± 0.8 61.1 ± 1.2 0.79 56.9 ± 0.6 57.1 ± 0.9 0.89
Birthplace/Smoking Status
U.S. born
1.00
1.00
Nonsmoker 18 (8.9) 9 (8.9) 26 (8.2) 13 (8.20
Smoker 20 (9.9) 10 (9.9) 34 (10.7) 17 (10.7)
Mexico born
Nonsmoker 76 (37.6) 38 (37.6) 132 (41.5) 66 (41.5)
Smoker 60 (29.7) 30 (29.7) 84 (26.4) 42 (26.4)
Other born
Nonsmoker 18 (8.9) 9 (8.9) 22 (6.9) 11 (6.9)
Smoker 10 (5.0) 5 (5.0) 20 (6.3) 10 (6.3)
Drusen Type
Soft Distinct
17 (16.8) 44 (27.7)
Soft Indistinct/Reticular 84 (83.2) 115 (72.3)
Drusen Size
63 to 125 um
16 (15.8) 41 (25.8)
125 to <250 um 52 (51.5) 79 (49.7)
250+ um 28 (27.7) 33 (20.8)
Reticular 5 (5.0) 6 (3.8)
1
p-Value for t-test (continuous variable) or chi-square test (categorical variable) for case-control
comparison.
2
SE = standard error.
31
Table 5: Bilateral and unilateral intermediate to large soft macular drusen early
AMD case-control comparisons for the CFH Tyr402His polymorphism
Bilateral Unilateral
Frequency (%) Frequency (%)
Model
Controls
N = 202
Case
N = 101 OR
a
(95% CI
b
)
Controls
N = 318
Case
N = 159 OR
a
(95% CI
b
)
YY
138 (68.3) 56 (55.5) 1 220 (69.2) 111 (69.8) 1
HY
61 (30.2) 40 (39.6) 1.7 (1.0 – 2.8) 91 (28.6) 44 (27.7) 1.0 (0.6 – 1.5)
HH
3 (1.5) 5 (5.0) 4.0 (0.9 – 17.0) 7 (2.2) 4 (2.5) 1.1 (0.3 – 3.9)
p
c
= 0.19 p
c
= 0.53 p
d
= 0.01 p
c
= 0.50 p
c
= 0.88 p
d
= 0.95
H-Allele 0.166 0.248 1.8
e
(1.1 – 2.8) 0.165 0.164 1.0
e
(0.7 – 1.4)
1.8
f
(1.1– 3.0) 1.0
f
(0.6 – 1.5)
3.3
g
(0.8 – 13.9) 1.1
g
(0.3 – 3.9)
a
Age-, smoking-, birthplace-adjusted odds ratio.
b
CI = Confidence interval.
c
Hardy-Weinberg p-value.
d
p-
Value for apparent dose effect.
e
Additive.
f
Dominant.
g
Recessive.
DISCUSSION
The recent evidence supporting substantial population attributable risks,
ranging from 43% to 70%, for the CFH Tyr402His polymorphism in overall AMD
was previously determined from clinic-based studies [Conley, Thalamuthu et al.
2005; Edwards, Ritter et al. 2005; Hageman, Anderson et al. 2005; Haines, Hauser
et al. 2005; Klein, Zeiss et al. 2005; Zareparsi, Branham et al. 2005]. Schaumberg
and colleagues confirmed an association for this polymorphism with overall AMD
in Whites (OR = 1.5, 95% CI = 1.1 to 2.0 and OR = 2.1, 95% CI = 1.1 to 4.2 for
His402 heterozygous and homozygous genotypes, respectively) and calculated an
attributable fraction of 25% (95% CI = 1% to 44%) [Schaumberg, Christen et al.
2006]. Using a population-based study among Latinos, the present data shows no
32
overall statistically significant association for CFH with early AMD while there is
sufficient power to detect odds ratios of 1.6 and greater for the dominant and log
additive models (for allele frequency 0.17, case-control design with 285 cases and
570 controls; using the program QUANTO [Gauderman and Morrison 2006]).
However, the present results suggests that, in Latinos, CFH is likely to be involved
in the pathogenesis of bilateral, but not unilateral, early AMD with intermediate to
large soft macular drusen.
Since bilaterality is known to increase with age [Wang, Mitchell et al.
1998], bilateral cases are expected to be older than the unilateral cases. On
average, the present bilateral cases are 4 years older than our unilateral cases (61
±12 years and 57 ± 11 years, mean ± standard deviation respectively for bilateral
and unilateral; p < 0.01). Although a statistically significant age difference exists
between the bilateral and unilateral cases, it is not likely that a large enough
number of unilateral cases would convert to bilateral status to change the present
findings within four years. Wang and colleagues showed that bilateral lesions in
early AMD increased by 20 percent over 10 years between the ages of 60 to 69
years [Wang, Mitchell et al. 1998], averaging only 2 percent per year.
Importantly, genotype and allele frequencies for the CFH Tyr402His
polymorphism in this Latino population differ from those recently reported for non-
Hispanic White clinic-based or cohort-based persons. The homozygous His402
genotype is present in only 3% of the Latino population compared to 9-21% in
33
clinic-based non-Hispanic Whites or 14% in cohort-based Whites [Conley,
Thalamuthu et al. 2005; Edwards, Ritter et al. 2005; Haines, Hauser et al. 2005;
Zareparsi, Branham et al. 2005; Schaumberg, Christen et al. 2006]. For Latinos,
the His402 allele is present in 17% of the population compared to 34 to 46% in
clinic-based non-Hispanic Whites or 34% in cohort-based Whites [Conley,
Thalamuthu et al. 2005; Edwards, Ritter et al. 2005; Haines, Hauser et al. 2005;
Zareparsi, Branham et al. 2005; Schaumberg, Christen et al. 2006]. Also, in a
Japanese population that found no association of CFH Tyr402His with exudative
AMD, the His402 allele is present in only 4% of controls [Gotoh, Yamada et al.
2006]. There is evidence from population-based studies that prevalence and
incidence rates of AMD also differ by ethnicity [Klein, Klein et al. 1999]. In
general, non-Hispanic Whites have a higher prevalence and age-specific rates of
CNV than Latinos [Klein, Klein et al. 1999; Varma, Fraser-Bell et al. 2004].
Similarly, Whites have a much higher prevalence of late AMD than Japanese
[Gotoh, Yamada et al. 2006]. In addition to ethnic differences in allele frequencies
of CFH polymorphism and rates of late AMD, Gotoh and colleagues note ethnic
differences in phenotypes, such that soft drusen is less common in Japanese cases
compared to Whites [Gotoh, Yamada et al. 2006]. The present results suggest that
the His402 allele is likely to play a role in susceptibility to early AMD soft drusen
lesions in Latinos that may increase the risk of progression to late AMD.
Therefore, ethnic differences in phenotypes and the frequency of this
34
polymorphism could help to explain the observed differences in rates of late AMD
among ethnic groups - an interesting hypothesis also recently proposed by others
[Grassi, Fingert et al. 2006]. Since all the current data on non-Hispanic Whites is
from clinic-based persons, with the exception of one prospective cohort, data from
other population-based samples of non-Hispanic Whites would further elucidate
this issue. Finally, ethnic differences in genetic susceptibility and disease
phenotypes may also exist for other diseases such as diabetes, inflammatory bowel
disease, autoimmune-disease, and Crohn's disease [Gotoh, Yamada et al. 2006].
The observations reported here demonstrate a statistically significant
association between a genotype and distinct macular phenotype within AMD.
Recently, Magnusson and colleuges have reported that theY402H variant confers
similar risk of soft drusen and both forms of advanced AMD in AMD patients from
Iceland and Utah; however, the role of bilaterality in the soft drusen patients was
not analyzed [Magnusson, Duan et al. 2006]. Although the present data suggests
that the CFH Tyr402His polymorphism is not a major susceptibility factor for
overall early AMD in the Latino population, the CFH His402 allele may play an
important role in bilateral phenotype-specific susceptibility, and especially
phenotypes likely to progress to late AMD. Finally, these results highlight the
importance of utilizing population-based studies to generalize study results and
considering ethnic differences in genotype distributions and disease.
35
Chapter 3: Manganese Superoxide Dismutase Ala16Val
and Complement Factor H Tyr402His in Early AMD
Background
Numerous key pathways are likely to be involved in the development of
AMD. Recently, considerable evidence has established a role for inflammation and
immune response pathways [Marx 2006]. Variations in the complement factor H
(CFH) gene are considered to be major contributors to genetic susceptibility in
Caucasians [Donoso, Kim et al. 2006]. In addition, after reviewing several current
concepts of AMD pathology, Zarbin presents a model for the overlap between non-
pathologic aging and AMD in which both oxidation and inflammation play key
roles [Zarbin 2004]. Zarbin explains that cumulative oxidation is likely to result in
primary insults to RPE, and possibly choriocapillaries, in aging that is followed by
an inflammatory response that eventually leads to RPE cell death and pathogenic
extracellular matrix changes [Zarbin 2004]. Although a role for oxidative stress in
AMD pathology is well established [Beatty, Koh et al. 2000], candidate genes
involved in oxidant or antioxidant processes have not been consistently observed in
association with AMD [Haddad, Chen et al. 2006].
Mitochondrial respiration, a major cellular source of reactive oxygen
species (ROS), plays a key role in cumulative oxidation in aging. Manganese
36
superoxide dismutase (MnSOD) is an important endogenous antioxidant enzyme
localized to the mitochondrial matrix to catalyze the dismutation of superoxide to
form hydrogen peroxide - or dismutation of a higher ROS to form a lesser ROS.
Previously, the Ala16Val polymorphism of MnSOD has been associated with
exudative AMD in Japanese [Kimura, Isashiki et al. 2000], although no association
was found for MnSOD haplotypes with exudative AMD in Caucasians of Northern
Ireland [Esfandiary, Chakravarthy et al. 2005]. There is a large degree of
inconsistencies in candidate gene studies of AMD in general, as expected with
complex diseases. Both the multifactorial etiology of AMD and possible ethnic-
specific as well as phenotype-specific susceptibility are likely to contribute to
inconsistencies across studies.
In a recent study, Ferris and colleagues report that intermediate and large
drusen as well as pigmentary abnormalities are important lesions considered to
increase risk of progression to late AMD [Ferris, Davis et al. 2005]. In addition,
bilateral lesions in early AMD increase risk of progression to advanced AMD
[Wang, Mitchell et al. 1998; 2001; Ferris, Davis et al. 2005]. The CFH Tyr402His
polymorphism is associated with early AMD with bilateral, but not unilateral,
intermediate to large soft drusen in Latinos (See Chapter 2). In contrast, no
differences across phenotypes were reported in several previous studies in
Caucasians, including one prospective study, which found significant association of
this polymorphism with AMD [Conley, Thalamuthu et al. 2005; Edwards, Ritter et
37
al. 2005; Hageman, Anderson et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss
et al. 2005; Zareparsi, Branham et al. 2005; Schaumberg, Christen et al. 2006].
The only previously reported finding of no association with the CFH Tyr402His
polymorphism and AMD is for Japanese, which have different general phenotype
distributions as compared to Caucasians [Gotoh, Yamada et al. 2006]. Especially
because of the multifactiorial etiology of AMD, it is important to explore gene-
gene interactions in AMD. Because both cumulative oxidation and inflammation
are considered critical concepts in AMD pathogenesis [Zarbin 2004],
simultaneously consideration of CFH Tyr402His and MnSOD Ala16Val candidate
polymorphisms with distinct AMD phenotypes likely to progress were explored in
the present study. It is possible that an indirect effect of carrying both CFH and
MnSOD high-risk genotypes may be to increase susceptibility to cumulative
oxidation and inflammation to further elevate the risk of developing AMD with age
over just carrying either one high-risk genotype alone.
Methods
Subjects were ascertained from within the Los Angeles Latino Eye Study
(LALES), a population-based cross-sectional study originally designed to assess
the prevalence of visual impairment, ocular disease, and visual functioning in
Latinos. Details of LALES study design is described elsewhere [Varma, Paz et al.
2004]. Briefly, all procedures conformed to the Declaration of Helsinki for
38
research involving human subjects. The Institutional Review Board of the
University of Southern California approved the project, and informed consent was
obtained from all participants. From February 2000 to May 2003, 6357 Latino
participants age 40 years and older from 6 census tracts in Los Angeles, California,
were recruited. The participation rate in LALES was 82 percent. All LALES
participants underwent complete eye exams including fundus photographs.
Diagnosis and grading of AMD was performed in a masked manner as defined
according to the Wisconsin Age-Related Maculopathy Grading System
(WARMGS) [Klein, Davis et al. 1991; Klein, Klein et al. 1992]. Whole genome
amplification of DNA from stored blood samples, taken at the time of the eye exam
for all participants entering the study from March, 2001 and beyond, was
performed by multiple displacement amplification technology (Repli-g Service,
Qiagen Sciences, Germantown, MD) and resulted in validated, usable quality for
genotyping assays among 98% of all LALES samples. Early AMD cases and
controls without advanced diabetic retinopathy among LALES participants were
eligible for this study if validated, usable whole genome amplification material was
available. Only 8 early AMD cases without advanced diabetic retinopathy were
excluded from the study due to inferior quality of whole genome amplification
material.
CFH Tyr402His genotyping was performed without knowledge of case
status by PCR amplification and restriction enzyme digestion. A 179-bp PCR
39
product containing the Tyr402His region was amplified in a 50uL reaction with
200 nM of forward primer, 5' TCATTGTTATGGTCCTTAGGAAA3', and reverse
primer, 5' GGAGTAGGAGACCAGCCATT3', 500uM MgCl2, 1X PCR buffer,
and 1 Unit Taq DNA polymerase (Accuprime Taq DNA Polymerase System,
Invitrogen, Carlsbad, CA). PCR amplification was performed with 40 cycles of
30s at 94C, 30s at 58C, and 30s at 68C (GeneAmp 9700 Thermacycler, Applied
Biosystems, Foster City, CA). Purified PCR product was digested with 10 Units of
Tsp509I (New England Biolabs, Ipswich, MA) for 3 hrs at 65C. The digested PCR
products were resolved with 4% NuSieve Agarose (Latitude HT precast gels,
Cambrex, East Rutherford, New Jersey) and stained with ethidium bromide.
Automated direct sequencing (3100 Genetic Analyzer, Applied Biosystems, Foster
City, CA) was performed on the reverse strand of the undigested 179-bp PCR
product for selected samples to validate the restriction enzyme assay or to clarify
ambiguous genotypes.
MnSOD Ala16Val genotyping was performed without knowledge of case
status by PCR amplification and restriction enzyme digestion. A 194-bp PCR
product containing the Ala16Val region was amplified in a 50uL reaction with 1uM
of forward primer, 5' GGCTGTGCTTTCTCGTCTTC 3', and reverse primer, 5'
GGTGACGTTCAGGTTGTTCA 3', 1mM MgCl2, 1X PCR buffer, and 0.5 Unit
Taq DNA polymerase (Accuprime Taq DNA Polymerase System, Invitrogen,
Carlsbad, CA). PCR amplification was performed with 30 cycles of 30s at 94C,
40
30s at 56C, and 30s at 68C (GeneAmp 9700 Thermacycler, Applied Biosystems,
Foster City, CA). Purified PCR product was digested with 5 Units of BsaWI (New
England Biolabs, Ipswich, MA) for 3 hrs at 60C. The digested PCR products were
resolved with 4% NuSieve Agarose (Latitude HT precast gels, Cambrex, East
Rutherford, New Jersey) and stained with ethidium bromide. Automated direct
sequencing (3100 Genetic Analyzer, Applied Biosystems, Foster City, CA) was
performed on the forward strand of the undigested 194-bp PCR product for selected
samples to validate the restriction enzyme assay or to clarify ambiguous genotypes.
SAS version 9.1 (SAS Institute, Cary, NC), was used to perform all
statistical analyses. The χ
2
test for Hardy-Weinberg equilibrium was used to
evaluate frequency distributions of genotypes among cases and controls separately.
Conditional logistic regression models were used to compare genotypes or alleles
for MnSOD between cases and their matched controls. Test for statistical
interaction and combined effects of MnSOD and CFH genotypes were explored
using conditional logistic regression. All statistical tests were performed using a
5% significance level.
Results
Case-control comparisons were performed for categories of cases including
intermediate to large macular soft drusen (MSD) or pigmentary abnormalities. Of
the 285 early AMD cases available, there were 101 cases with bilateral MSD and
41
159 with unilateral MSD. When considering any with pigmentary abnormalities
(defined as either hyper- or hypo-pigmentation of the retinal pigmented
epithelium), there were a total of 177 cases of the 285. Table 6 shows the
frequency distribution of the MnSOD A16V polymorphism for case-control
samples based on case phenotype. Table 7 shows no overall statistically
significant differences in MnSOD alleles or genotypes for cases compared to
controls among any phenotype considered. However, the odds ratio for the
recessive Ala16 model for cases with pigmentary abnormalities compared to
controls is the most significant among all case-control comparisons (OR = 1.2, 95%
CI = 0.8 - 1.8; table 7). For phenotypes considered, main effects as well as
simultaneous effects of the recessive Ala16 MnSOD (AA) and the dominant
His402 CFH (Hx) genotypes are summarized in table 8. As observed previously
(see Chapter 2), table 8 also shows the association of CFH Tyr402His with bilateral
intermediate to large MSD (OR = 1.8, 95% CI = 1.1 to 3.0). Interestingly, a
statistically significant interaction was observed for CFH and MnSOD genotypes
among early AMD cases with pigmentary abnormalities compared to controls (p =
0.039, table 8).
42
Table 6: Genotype and allele frequency distributions for MnSOD A16V
polymorphism for early AMD phenotype-specific case-control comparisons
Macular Soft Drusen (63um or greater)
Bilateral Unilateral
Pigmentary
Abnormalities
Controls
N = 202
Cases
N = 101
Controls
N = 318
Cases
N = 159
Controls
N = 354
Cases
N = 177
Genotype or
Allele
n (%) n (%)
n (%) n (%)
n (%) n (%)
VV 34 (16.8) 16 (15.8) 40 (12.6) 30 (18.9) 56 (15.8) 31 (17.4)
AV 95 (47.0) 46 (45.5) 156 (49.1) 73 (45.9) 169 (47.7) 73 (41.2)
AA 73 (36.1) 39 (38.6) 122 (38.4) 56 (35.2) 129 (36.4) 73 (41.2)
A-allele 0.60 0.61 0.63 0.58 0.61 0.62
H-W
1
p-value 0.74 0.69 0.36 0.48 0.96 0.09
1
Hardy-Weinberg p-value.
Table 7: Odds ratios for MnSOD A16V and early AMD phenotypes-specific
case-control comparisons controlling for age.
Macular Soft Drusen (63um or greater)
Bilateral Unilateral
Pigmentary
Abnormalities
Model
OR
1
(95% CI
2
) OR
1
(95% CI
2
) OR
1
(95% CI
2
)
VV 1 1 1
AV 1.0 (0.5 - 2.0) 0.6 (0.4 - 1.1) 0.8 (0.5 - 1.3)
AA 1.1 (0.6 - 2.3) 0.6 (0.3 - 1.1) 1.0 (0.6 - 1.7)
A-allele
Dominant 1.1 (0.6 - 2.0) 0.6 (0.4 - 1.0) 0.9 (0.6 - 1.4)
Recessive 1.1 (0.7 - 1.9) 0.9 (0.6 - 1.3) 1.2 (0.8 - 1.8)
1
OR = Odds ratio.
2
CI = Confidence interval.
43
Table 8: Odds ratio calculations for single genotype exposures of MnSOD A16V
or CFH Y402H and test for interaction with early AMD phenotypes-specific case-
control comparisons controlling for age.
Macular Soft Drusen (63um or greater)
Model
Bilateral Unilateral
Pigmentary
Abnormalities
MnSOD Control Case Control Case Control Case
no AA
1
129 62 196 103 225 104
AA
2
73 39 122 56 129 73
OR
3
(95% CI
4
) 1.1 (0.7 - 1.9) 0.9 (0.6 - 1.3) 1.2 (0.8 - 1.8)
CFH Control Case Control Case Control Case
no Hx
5
138 56 220 111 243 115
Hx
6
64 45 98 48 111 62
OR
3
(95% CI
4
) 1.8 (1.1 - 3.0) 1.0 (0.6 - 1.5) 1.2 (0.8 - 1.8)
Test for interaction
p-value 0.23 0.27 0.039
OR
MnSOD
7
(95% CI
4
) 1.0 (0.5 - 1.9) 0.8 (0.5 - 1.2) 0.9 (0.6 - 1.5)
OR
CFH
8
(95% CI
4
) 1.5 (0.8 - 2.8) 0.8 (0.5 - 1.4) 0.8 (0.5 - 1.4)
OR
INT
9
(95% CI
4
) 1.9 (0.7 - 5.7) 1.6 (0.7 - 3.7) 2.3 (1.0 - 4.9)
1
No homozygous MnSOD Ala16 genotype.
2
Homozygous MnSOD Ala16 genotype.
3
OR = odds ratio.
4
CI =
confidence interval.
5
No heterozygous CFH His402 genotype.
6
Heterozygous CFH His402 genotype.
7
Odds
ratio for homozygous MnSOD Ala16 genotype controlling for heterozygous CFH His402 genotype and
interaction.
8
Odds ratio for heterozygous CFH His402 genotype controlling for homozygous MnSOD Ala16
genotype and interaction.
9
Odds ratio for interaction.
The combined effects of MnSOD and CFH genotypes were further explored
using genotype combinations for cases with bilateral MSD or pigmentary
abnormalities. The highest risk group for combined genotype is assumed to be
carriers of at least one His402 CFH and two Ala16 MnSOD alleles. Alternatively,
the lowest risk group is assumed to be carriers of no His402 CFH and one or no
44
Ala16 MnSOD alleles. Cell counts and odds ratios for the combined genotype
models are shown in table 9. When considering combined genotypes, statistically
significant odds ratios are only observed when both CFH and MnSOD high-risk
genotypes are present compared to low-risk genotypes. From table 9, bilateral
MSD cases are 2.9-times more likely to carry the combined high-risk versus low-
risk CFH and MnSOD genotypes as compared to controls (95% CI = 1.2 to 7.1).
For cases with pigmentary abnormalities compared to controls, the odds ratio for
carrying the combined high-risk versus low-risk CFH and MnSOD genotypes is 1.8
(95% CI = 1.01 to 3.2; table 9).
Table 9: Odds ratio calculations for all combinations of MnSOD A16V and CFH
Y402H genotypes for phenotypes with pigmentary change in early AMD
controlling for age.
Combined
Genotypes
(CFH, MnSOD)
All Pigmentary
Abnormalities
Pigmentary
Abnormalities +
Bilateral MSD
5
Bilateral MSD
5
Control Case Control Case Control Case
noHx, noAA
1
151 74 45 21 80 33
noHx, AA
2
92 41 36 12 58 23
Hx, noAA
3
74 30 28 16 49 29
Hx, AA
4
37 32 13 12 15 16
OR
8
(95% CI
9
) OR
8
(95% CI
9
) OR
8
(95% CI
9
)
noHx, noAA
1
1 1 1
noHx, AA
2
0.9 (0.6 - 1.5) 0.8 (0.3 - 1.8) 1.0 (0.5 - 1.9)
Hx, noAA
3
0.8 (0.5 - 1.4) 1.2 (0.5 - 2.8) 1.5 (0.8 - 2.8)
Hx, AA
4
1.8 (1.01 - 3.2) 2.2 (0.8 - 6.1) 2.9 (1.2 - 7.1)
1
noHx, noAA = Reference genotypes. No heterozygous CFH His402 and no homozygous MnSOD Ala16
genotypes.
2
Intermediate risk genotypes. No heterozygous CFH His 402 and homozygous MnSOD Ala16
genotypes.
3
Intermediate risk genotypes. Heterozygous CFH His 402 and no homozygous MnSOD Ala16
genotypes.
4
High risk genotypes. Heterozygous CFH His 402 and homozygous MnSOD Ala16 genotypes.
5
MSD = Macular soft drusen of 63 microns or greater in size.
6
b = parameter estimate.
7
se = standard error of
parameter estimate.
8
OR = odds ratio.
9
CI = confidence interval.
45
Discussion
In studying gene-gene interactions, two important issues are sample size
necessary to observe a statistically significant difference and biological plausibility.
In the present study a statistically significant interaction was observed between the
MnSOD Ala16Val and CFH Tyr402His genotypes for cases with pigmentary
abnormalities (N = 177) compared to controls (N = 354). In addition, when
modeling combined genotypes, carrying high-risk versus low-risk genotypes for
both CFH and MnSOD were more frequently observed for cases with either
pigmentary abnormalities or bilateral soft drusen compared to controls (OR = 1.8,
95% CI = 1.01 to 3.2 or OR = 2.9, 95% CI = 1.2 to 7.1; respectively for pigmentary
abnormalities or bilateral soft drusen). Relevance of this interaction observed
especially among those with pigmentary abnormalities compared to controls may
be due to a phenotype that is especially susceptible to oxidative stress. The
pigmentary abnormalities observed for the present population are hyper- or hypo-
melanin content in the RPE. Melanin pigment is generally thought to serve as an
antioxidant in the RPE, although some have found it may promote light damage
depending on the cellular environment [Dontsov, Glickman et al. 1999].
Accumulation of lipofuscin, an autofluorescent pigment of aging, in RPE also
contributes to a pro-oxidative environment [Winkler, Boulton et al. 1999]. Also,
complex granules of melanin and lipofuscin may influence pro-oxidant properties
of the RPE in AMD [Sarangarajan and Apte 2005].
46
There may be evidence for biological plausibility for a statistical interaction
between important proteins of the mitochondrial and innate immunity, such as that
observed between MnSOD and CFH. Recent evidence from studies of hepatitis C
suggests a central role for mitochondria in innate immunity with the discovery of
mitochondrial antiviral signaling (MAVS) protein [Piccoli, Scrima et al. 2006]. In
response to viral infection, MAVS, which is localized to the outer mitochondrial
membrane, activates signaling pathways to enable downstream transcription and
maximize cytokine production [McWhirter, Tenoever et al. 2005]. Furthermore,
evidence suggests that the ability of some pathogens to evade host innate immunity
may be due to binding and cleavage of MAVS, such is the case for the hepatitis C
virus [Piccoli, Scrima et al. 2006]. While the extent and details of the
mitochondrial involvement in innate immunity are yet to be determined, it is
interesting that mitochondrial dysfunction, which is inherently tied to oxidative
stress, could have influential effects on innate immunity through nuclear encoded,
mitochondrial localized proteins.
MnSOD is an important nuclear encoded protein that localizes to the
mitochondrial matrix by a signaling sequence that contains either a valine or
alanine at codon 16. There is evidence that the Ala16 MnSOD allele is targeted
through the mitochondrial membrane more efficiently than the Val16 allele [Hiroi,
Harada et al. 1999; Sutton, Khoury et al. 2003; Sutton, Imbert et al. 2005]. This
polymorphism has been previously associated with several diseases although
47
differentially with respect to which allele confers risk [Taufer, Peres et al. 2005].
MnSOD functions to catalyze the dismutation of a higher ROS (superoxide) into a
lesser ROS (hydrogen peroxide), which may help to explain inconsistencies for
which is the high-risk allele for the Ala16Val polymorphism given a particular
cellular environment. Taufer and colleagues suggest that the differential
association of MnSOD alleles may be explained by oxidative imbalance and sought
out to determine if the Ala16 allele was specifically associated with age-related
disease (breast and prostate cancer) or age-related mortality [Taufer, Peres et al.
2005]. Interestingly, Taufer and colleagues found an increased risk of breast or
prostate cancer as well as presence of immunosenscence markers and DNA damage
among those with homozygous Ala16 MnSOD genotype. However, the
homozygous Ala16 MnSOD genotype was not associated with age-related
mortality [Taufer, Peres et al. 2005]. Taufer and colleagues suggest an explanation
for this paradox may be due to the relationship between oxidative imbalance and
immune response. Furthermore, Taufer and colleagues suggest that homozygous
Ala16 MnSOD genotype may be associated with stronger innate immunity due to
increased generation of hydrogen peroxide levels, especially because activated
immune cells, such as macrophages, produce hydrogen peroxide during
inflammation. This is an interesting hypothesis with relevance to a possible
interaction between MnSOD and CFH, an important regulator of innate immunity.
48
Underlying mechanisms for a possible interaction between MnSOD and
CFH are difficult to explore and remain unclear. However, differential
peroxynitrite amplification may be an interesting hypothesis. Superoxide,
primarily produced from electrons leaking from the respiratory chain and
combining with molecular oxygen in the mitochondrial matrix, reacts with either
MnSOD, a mitochondrial antioxidant enzyme, to form hydrogen peroxide or with
nitric oxide to form peroxynitrite. Nitric oxide readily diffuses through the
mitochondrial membranes and may be produced both outside and inside the
mitochondria. The reaction rate of superoxide with nitric oxide far exceeds that of
superoxide and MnSOD, therefore when mitochondrial nitric oxide concentrations
are high, formation of peroxynitrite is favored over hydrogen peroxide. In aging
and injured tissues, levels of the inducible form of nitric oxide synthase may be
elevated and influence peroxynitrite generation. In addition, there is evidence for
mitochondrial NOS (mtNOS) [Lacza, Snipes et al. 2003], which may be the most
likely source for elevated mitochondrial nitric oxide levels and peroxynitrite
generation. Peroxynitrite is a powerful oxidant and readily nitrates tyrosine
residues. Peroxynitrite-induced tyrosine nitration of MnSOD is catalyzed by the
manganese cations residing in close proximity to critical tyrosine residues and lead
to inactivation of MnSOD [MacMillan-Crow, Crow et al. 1998]. Inactivation
MnSOD then favors further peroxynitrite formation through elevated superoxide
levels. Previously, it has been hypothesized that increased generation of
49
peroxynitrite due to increased levels of nitric oxide may initiate a pathological loop
of peroxynitrite amplification and MnSOD inactivation [MacMillan-Crow, Crow et
al. 1996]. To determine if differential peroxynitrite amplification is biologically
plausible for different MnSOD alleles, a mathematical model was used to simulate
the inactivation of MnSOD (or decreased MnSOD concentration) with increased
levels of nitric oxide (or substrate for peroxynitrite) using real or perceived reaction
rates or constants based on review of the literature [Antunes, Salvador et al. 1996;
Hsu, Hsieh et al. 1996; Quijano, Hernandez-Saavedra et al. 2001; Thomas, Liu et
al. 2001; Radi, Cassina et al. 2002] [Hare and Hodges 1982; Hass and Massaro
1987] using the biochemical simulation software, Gepasi 3.3 [Mendes 1993]
(www.ncgr.org/software/gepasi/) (see Figure 5). For the simplest model, not
considering repair of nitration, a greater affinity for nitration of the Ala16 allele
would yield greater amounts of peroxynitrite causing levels of active MnSOD to
drop off dramatically as compared to the Val16 allele (represented by Ala16_b
compared to Val16, Figure 5). However, there is no direct evidence for a greater
affinity for nitration for the Ala16 allele. Furthermore, since this polymorphism is
located in the targeting sequence, it may be cleaved off before processing of the
peptides to form the mature protein. Further biological experiments would be
necessary to explore the possibility that this polymorphism influences the way the
molecular chaperones recognize the newly translocated peptides, in which case
altered folding could contribute to differential nitration.
50
Effect of Peroxynitrite Amplification on MnSOD Concentration
1.00E-08
1.00E-07
1.00E-06
1.00E-05
1.00E-04
1.00E-05 1.00E-04 1.00E-03
Nitric Oxide Concentration [M]
MnSOD Concentration [M]
Val16
Ala16_a
Ala16_b
Figure 5: Simulation of the effect of peroxynitrite amplification on MnSOD Ala16Val allele.
(Ala16_a and Ala16_b designate the Ala16 allele for the case of no preferential nitration and
preferential nitration, respectively.) Reactions, rates, and rate constants used to produce simulation
data are based on literature or perceived values. The following rates (v) or rate constants (k) were
used: v = 1x10
-5
Ms
-1
for production of superoxide, k = 4.7x10
9
s
-1
for consumption of superoxide, k
= 1x10
3
s
-1
for the production of nitric oxide, k = 2.3x10
-5
s
-1
for the consumption of nitric oxide, k =
6.7x10
9
M
-1
s
-1
for the production of peroxynitrite, k = 100 s
-1
for the consumption of peroxynitrite, k
= 2.3x10
-5
s
-1
equally for the consumption of MnSOD and the consumption of nitrated MnSOD, v =
4x10
-10
Ms
-1
or 1x10
-10
Ms
-1
for the production of MnSOD for Ala16_a (no preferential nitration) or
Val16, respectively; and k = 1x10
4
M
-1
s
-1
equally for the production of nitrated MnSOD for Ala16_b
and Val16; or 1x10
5
M
-1
s
-1
for the production of nitrated MnSOD for Ala16_b (preferential
nitration). Gepasi 3.3 software [Mendes 1993] (www.ncgr.org/software/gepasi/) was used to produce
simulated data.
51
Chapter 4: Whole Genome Association Study of Early
AMD in Latinos using GeneChip 500K and 100K SNP
Microarrays
INTRODUCTION
A genetic basis for AMD is well established, however, few of the several
implicated candidate genes and chromosomal regions (reviewed in [Haddad, Chen
et al. 2006]) identified to date have been reproduced across studies. Several studies
have reported a major role for complement genes; especially for the Tyr402His
polymorphism of complement factor H (CFH) using Caucasian populations. CFH
is located at 1q32, a previously implicated chromosomal region [Weeks, Conley et
al. 2001; Fisher, Abecasis et al. 2005]. Also implicated as major contributors of
AMD are tightly linked genes located at 10q26, pleckstrin homology domain
containing family A member 1 and LOC387715 (PLEKHA1/ LOC387715)
[Haddad, Chen et al. 2006; Marx 2006].
Evidence from a recent gene scan meta-analysis (GSMA) [Fisher, Abecasis
et al. 2005] shows that the two most important chromosomal regions in AMD
susceptibility may be 10q and 1q. More specifically, Fisher and colleagues [Fisher,
Abecasis et al. 2005] found statistically significant association for chromosomal
region 10q26-10qter (p < 0.0004) and borderline significant association with the
52
adjacent region 10q23-10q26 (p < 0.008) after adjusting for multiple comparisons
[Fisher, Abecasis et al. 2005]. The third most significant region implicated by
GSMA is 1q23-q31 (p = 0.01) [Fisher, Abecasis et al. 2005], although not
statistically significant after adjusting for multiple comparisons. Fisher and
colleagues also report that the adjacent region between 1q31-1q32, which contains
CFH, is implicated at the 5% significance level (p = 0.02) [Fisher, Abecasis et al.
2005].
Susceptibility to AMD in Latinos may include similar or unique genetic
factors as that in Caucasians. For CFH, genotype and allele frequencies are very
different for Latinos compared to Caucasians (3% and 14% compared to 17% and
34% for His402 homozygous genotype and His402 allele frequencies for Latino
compared to Caucasian controls, respectively). An interesting hypothesis also
recently proposed by others [Grassi, Fingert et al. 2006], may be that ethnic
differences in the frequency of the CFH Tyr402His polymorphism may help to
explain the observed differences in rates of late AMD among ethnic groups such as
Latinos and Caucasions. There is evidence exists to suggest ethnic- and phenotype-
specific roles for genetic susceptibility [Klein, Rowland et al. 1995; Klein, Peto et
al. 2004; Klein, Klein et al. 2006], which may likely account for inconsistent results
across studies.
Ideally, whole genome association studies have the greatest advantage in
detecting consistent genetic contributors of complex disease [Dekker and van Duijn
53
2003]. Because genotyping technology as well as technical and statistical
haplotype resources have vastly improved in recent years, whole genome
association studies of complex disease are now feasible. Affymetrix GeneChip
Mapping 500K and 100K (GeneChip 500K and 100K) single nucleotide
polymorphism (SNP) set microarrays allow the simultaneous genotyping of over
600,000 genomewide SNPs. Furthermore, evidence suggests that genotyping
approximately 500,000 carefully selected genome-wide SNPs is sufficient to
identify haplotype tagging (tagSNPs) and capture nearly all common variation with
80 percent power [Barrett and Cardon 2006].
The present study is a whole genome association study using bilateral early
AMD cases and controls nested within the Los Angeles Latino Eye Study
(LALES), the largest population-based study of eye disease in any racial/ethnic
group in the U.S. [Varma, Fraser-Bell et al. 2004; Varma, Paz et al. 2004]. In the
present study, genotyped SNPs from GeneChip 500K and 100K SNP set
microarrays were used for case-control comparisons for the whole genome as well
as candidate regions. Candidate regions for the present study were based on
important results from the recent GSMA study [Fisher, Abecasis et al. 2005],
including 10q23-10qter and 1q23-q32.
54
METHODS
Subjects were ascertained from within LALES, a population-based cross-
sectional study originally designed to assess the prevalence of visual impairment,
ocular disease, and visual functioning in Latinos. Details of LALES study design is
described elsewhere [Varma, Paz et al. 2004]. Briefly, all procedures conformed to
the Declaration of Helsinki for research involving human subjects. The Institutional
Review Board of the University of Southern California approved the project, and
informed consent was obtained from all participants. From February 2000 to May
2003, 6357 Latino participants age 40 years and older from 6 census tracts in Los
Angeles, California, were recruited. The participation rate in LALES was 82
percent. All LALES participants underwent complete eye exams including fundus
photographs. Diagnosis and grading of AMD was performed in a masked manner
as defined according to the Wisconsin Age-Related Maculopathy Grading System
(WARMGS) [Klein, Davis et al. 1991; Klein, Klein et al. 1992]. Whole genome
amplification of DNA from stored blood samples, taken at the time of the eye exam
for all participants entering the study from March, 2001 and beyond, was
performed by multiple displacement amplification technology (Repli-g Service,
Qiagen Sciences, Germantown, MD) and resulted in validated, usable quality for
genotyping assays among 98% of all LALES samples. Early AMD cases and
controls without advanced diabetic retinopathy among LALES participants were
eligible for this study if validated, usable whole genome amplification material was
55
available. Only 8 early AMD cases without advanced diabetic retinopathy were
excluded from the study due to inferior quality of whole genome amplification
material. Of 285 available cases, bilateral intermediate to large soft drusen or
reticular drusen were present for 101 casess. Controls were matched to cases (2:1
controls:case) by exact or nearest age, birthplace (U.S., Mexico, or other), and
smoking status (smoker or nonsmoker). When more than 2 controls were available
per case, controls were selected randomly.
Genotyping was performed according to manufacturer protocol
(Affymetrix, Santa Clara, CA) using whole genome amplification product as
template for GeneChip Mapping 500K and 100K SNP set microarrays for a total of
616,772 SNPs. Briefly, the assays were performed by digesting template with
restriction enzymes, such as XbaI or Hind III, and ligating restriction fragments to
adapters. The adapter-template fragments (250 to 2000 bp in size) were amplified
using a generic primer that recognizes the adapter sequence. The amplified
segments were fragmented, labeled and hybridized to the GeneChip 500K and
100K microarrays. GeneChip DNA Analysis Software (GDAS; Affymetrix, Santa
Clara, CA) was used to determine genotype calls.
Genetrix (Epicenter, Pasadena, CA) software was used to determine
differences between cases and controls for genotyped SNPs. Candidate regions
included (i) 10q23-10qter designated by 9,509 SNPs between rs7084418 to
rs9629975 and (ii) 1q23-1q32 designated by 13,710 SNPs between rs1804478 to
56
rs11120004. Bonferroni adjusted p-values were used to correct for multiple
comparisons. Because case-control comparisons using Genetrix software did not
account for matching variables, genotypes for selected SNPs were imported into
SAS version 9.1 (SAS Institute, Cary, NC) for further analysis. The χ
2
test for
Hardy-Weinberg equilibrium was used to evaluate frequency distributions of
genotypes among cases and controls separately for selected SNPs using SAS
version 9.1. Conditional logistic regression models were used to compare
recessive, dominant, and log-additive genetic models for cases versus age-,
birthplace-, smoking status-matched controls using SAS version 9.1. Fisher's exact
probabilities were computed when cell counts were low (below 5).
Using algorithms implemented within Genetrix similar to the software
Haploview (www.broad.mit.edu/personal/jcbarret/haplo/), SNP genotype was used
to characterize the linkage disequilibrium (LD) and define haplotype blocks for
selected regions. LD block structure was examined using criteria described by
Gabriel et al., 2002, [Gabriel, Schaffner et al. 2002] with the proportion of
informative SNP pairs in strong LD limited to greater than 90%, the lower bound of
greater than 70% and upper bound greater than 98% on the D’ statistic for strong
LD. The recombination upper bound of D-prime used was less than 90%, and only
SNPs with minor allele frequencies equal to 20% or greater were used.
57
RESULTS
AMD cases were selected for further analysis from LALES based on
bilateral status of soft drusen. Bilateral cases included 35.5 percent (= 101/285) of
eligible AMD cases for which DNA was collected between years 2001 and 2003.
Table 10: Distribution of demographic and clinical features for cases and controls.
Controls
N = 202
Bilateral Cases
N = 101 p-value
Age (mean ± SE
1
) 60.7 ± 0.8 61.1 ± 1.2 0.79
Birthplace/Smoking Status
U.S. born
Nonsmoker 18 (8.9) 9 (8.9)
Smoker 20 (9.9) 10 (9.9)
Mexico born
Nonsmoker 76 (37.6) 38 (37.6)
Smoker 60 (29.7) 30 (29.7)
Other born
Nonsmoker 18 (8.9) 9 (8.9)
Smoker 10 (5.0) 5 (5.0)
1.00
Drusen Type
Soft Distinct
17 (16.8)
Soft Indistinct/Reticular 84 (83.2)
Drusen Size
63 to 125 um
16 (15.8)
125 to <250 um 52 (51.5)
250+ um 28 (27.7)
Reticular 5 (5.0)
1
SE = standard error
Table 11: SNP list of most significant associations for genomewide case-control comparisons (total SNPs = 616,772).
Minor Allele
Frequency
Chromosome
Location SNP Nearest Gene
Control
(%)
Case
(%)
Genetic
Model p-value Adj
1
p-value
1q25.2 rs4652201 Astrotactin (Intron 18) 18.9 33.7 Recessive 1.5E-07 0.0925
1q25.2 rs6425398 Astrotactin (Exon23) 20.8 33.5 Recessive 2.5E-07 0.1542
1q25.2 rs4652199 Astrotactin (Intron 20) 20.3 33.8 Recessive 2.1E-07 0.1295
1q25.2 rs6685449 Astrotactin (Intron 17) 19.4 36.7 Recessive 2.7E-07 0.1665
8p12-p11 rs10503852 Dual specificity phosphatase 4 (262Kb 5') 20.1 34.7 Recessive 3.0E-06 1.0000
5q21 rs430123 Ephrin-A5 (634Kb 3' ) 1.1 9.4 Dominant 3.6E-06 1.0000
9p24.2 rs10511437 KIAA0020 (70Kb 5') 1 8.8 Dominant 3.6E-06 1.0000
10q23.1 rs7919358 SH2 domain containing 4B (342Kb 3') 39.3 57.7 Recessive 4.3E-06 1.0000
10q22.3 rs11002634 Retinoic acid induced 17 (464Kb 5') 0 5.5 Dominant 4.4E-06 1.0000
1
Bonferroni adjusted for 616,772 tests.
58
Table 12: SNP list of most significant associations for candidate region, 10q23-10qter, case-control comparisons
(total SNPs = 9,509).
Minor Allele
Frequency
SNP
Nearest Gene
Control
(%)
Case
(%)
Genetic
Model
p-value
Adj
1
p-value
rs1900500 c10orf82 (18,6Kb 5') 51.7 37.8 Recessive 0.00014 1.0000
rs902474 Slit homolog 1 [Drosophilia] (Intron 4) 43.2 54.7 Dominant 0.00028 1.0000
1
Bonferroni adjusted for 9,509 tests.
59
Table 13: SNP list of most significant associations for candidate region, 1q23-q32, case-control comparisons
(total SNPs = 13,710).
Minor Allele
Frequency
SNP
Nearest Gene
Control
(%)
Case
(%)
Genetic
Model
p-value
Adj
1
p-value
rs4652201 Astrotactin (Intron 18) 18.9 37.1 Recessive 1.50E-07 0.0021
rs6425398 Astrotactin (Exon23) 20.8 36.9 Recessive 2.10E-07 0.0029
rs6685449 Astrotactin (Intron 17) 19.4 36.7 Recessive 2.70E-07 0.0037
rs4652199 Astrotactin (Intron 20) 20.3 36.9 Recessive 2.70E-07 0.0037
rs7555335 Astrotactin (Intron 16) 25.9 40 Recessive 5.30E-05 0.7266
rs5024503
Sodium- and chloride-activated ATP-sensitive
potassium channel (279Kb 3') 42.8 54.3 Recessive 6.20E-05 0.8500
rs6679037
Pleckstrin homology domain containing, family A
member 6 (Intron 17) 40 52.8 Recessive 1.20E-04 1.0000
rs2427837 Fc fragment of IgE, high affinity I receptor (958b 5') 8.9 20.7 Dominant 1.30E-04 1.0000
rs4652206 Astrotactin (Intron 16) 19.6 28.4 Recessive 1.60E-04 1.0000
rs1041236 Glycoprotein A33 (transmembrane) [Intron 1] 34.7 21.4 Dominant 1.70E-04 1.0000
rs16850453 Astrotactin (Intron 11) 18.1 26.6 Recessive 2.20E-04 1.0000
rs352293 Ring findger and WD repeat domain 2 (94Kb 5') 53.3 38.6 Dominant 2.30E-04 1.0000
rs1779304
Plakophilin 1 (ectodermal dysplasia/skin fragility
syndrome) (Intron 2) 38.2 48.2 Recessive 4.00E-04
1.0000
1
Bonferroni adjusted for 13,710 tests 60
61
Table 10 summarizes the distribution of demographic and clinical features for cases
(N = 101) and matched controls (N = 202). Tables 10, 11, and 12 show the most
significant case-control differences for the all SNPs and two candidate regions
using available genotyped SNPs from GeneChip 500K and 100K microarrays (total
SNPs = 616,772; 9,509; and 13,710; respectively). After Bonferroni correction for
multiple comparison, p-values below 1.0 for genomewide case-control differences
were observed for only 4 SNPs, (0.09 < p < 0.17, table 11), all from within
Astrotactin (ASTN). For candidate region, 10q23-10qter, no Bonferroni adjusted
p-values below 1.0 for case-control differences were observed (table 12). Table 13
shows that after Bonferoni correction when limited to candidate region 1q23-q32,
statistically significant case-control differences were observed for the same 4 SNPs
from within ASTN (0.001 < p-value < 0.02).
Table 14 shows genotype frequency distributions for the 4 ASTN SNPs
with statistically significant differences for cases compared to controls. Because
low cell counts exist for homozygous genotypes of the minor allele for controls,
Fisher's exact p-values were computed. Fisher's exact p-values for ASTN
genotype differences between cases and controls were statistically significant after
Bonferoni correction ( 0.005 ≤ p-value ≤ 0.008; table 14). Estimates of relative
risk for ASTN SNPs are shown in table 15. For the recessive model with respect to
the minor allele, cases were 29.6-times more likely to have the high-risk versus
low-risk genotype as compared to controls for the most influential SNP, rs6685449
62
(95% CI = 3.9 to 223.4, table 15). However, the causal SNP may only be linked to
the genotyped SNPs. The significantly associated SNPs and the, yet to be
determined causal SNP, may all be contained within one haplotype block. Linkage
Table 14: Genotype distribution among cases and controls for highly significant SNP associations in Astrotactin.
Frequency (%)
Astrotactin SNP (Location)
BB
1
AB
2
AA
3
Hardy-Weinberg
p-value
Fisher's Exact p-value (Adj p-
value
4
)
Controls: 119 (60.0) 79 (39.4) 2 (1.0) 0.0045
rs6425398 (Exon 23)
Cases: 49 (49.0) 35 (35.0) 16 (16.0) 0.0320
8.7E-07 (0.007)
Controls: 121 (60.6) 77 (38.4) 2 (1.0) 0.0066
rs4652199 (Intron 20)
Cases: 48 (48.5) 35 (35.4) 16 (16.2) 0.0363
7.8E-07 (0.006)
Controls: 124 (63.4) 70 (35.6) 2 (1.0) 0.0201
rs4652201 (Intron 18)
Cases: 47 (49.5) 32 (33.7) 16 (16.8) 0.0881
6.0E-07 (0.005)
Controls: 115 (63.0) 65 (35.4) 3 (1.7) 0.0661
rs6685449 (Intron 17)
Cases: 41 (45.6) 32 (35.6) 17 (18.9) 0.0261
1.1E-06 (0.008)
1
Homozygous genotype for major allele (major allele = B).
2
Heterozygous genotype.
3
Homozygous genotype for minor allele (minor allele = A).
4
Bonferroni
adjusted for 13,710 tests.
63
Table 15: Odds ratios for recessive, dominant and log-additive genetic models of Astrotactin SNPs comparing bilateral
early AMD cases with age, birthplace, smoking status matched controls
Odds Ratio (95% Confidence Interval) Astrotactin
SNP (Location) Recessive Dominant Log-Additive
BB
1
or AB
2
1 BB
1
1 BB
1
1
AB
2
1.1 (0.7 - 1.9)
rs6425398 (Exon 23)
AA
3
16.0 (3.7 - 69.6) AB
2
or AA
3
1.6 (0.9 - 2.6)
AA
3
17.0 (3.8 - 75.5)
BB
1
or AB
2
1 BB
1
1 BB
1
1
AB
2
1.2 (0.7 - 2.1)
rs4652199 (Intron 20)
AA
3
16.0 (3.7 - 69.6) AB
2
or AA
3
1.6 (1.0 - 2.7)
AA
3
17.4 (3.9 - 77.5)
BB
1
or AB
2
1 BB
1
1 BB
1
1
AB
2
1.2 (0.7 - 2.1)
rs4652201 (Intron 18)
AA
3
14.6 (3.3 - 63.6) AB
2
or AA
3
1.7 (1.0 - 2.8)
AA
3
15.9 (3.6 - 70.5)
BB
1
or AB
2
1 BB
1
1 BB
1
1
AB
2
1.4 (0.8 - 2.5)
rs6685449 (Intron 17)
AA
3
29.6 (3.9 - 223.4) AB
2
or AA
3
2.0 (1.1 - 3.3)
AA
3
34.7 (4.5 - 268.2)
1
Homozygous genotype for major allele (major allele = B).
2
Heterozygous genotype.
3
Homozygous genotype for minor allele (minor allele = A).
Note: N =101 cases compared to N = 202 matched controls.
64
65
Exon 23 Intron 20 Intron 18 Intron 17 Intron 16
Pappalysin 2 Astrotactin
Figure 6: Linkage disequilibrium (LD) plot* in region of highly associated
SNPs (rs6425398, rs4652199, rs4652201, and rs6685449), indicated above
by red arrows. Intron 16 SNPs listed among most significantly associated
SNPs (rs7555335 and rs4652206, see table 12) are indicated above by gray
arrows. The suggested haplotype block of interest is 52.6Kb in length
extending from Exon 23 to mid-way through Intron 16 of Astrotactin as
indicated above. *Note: SNPs included in above LD plot were limited to those with minor allele
frequencies greater than 19%. SNP rs10913271 was assumed to be a new mutation and was excluded
from LD plot. Proportion of informative SNP pairs in strong LD was limited to greater than 90%.
Strong LD lower and upper bound of D-prime was limited to greater than 70% and 98%, respectively.
Recombination upper bound of D-prime was limited to less than 90%
66
disequilibrium in the region of the significantly associated ASTN SNPs implicates
a haplotype block that is 52.6Kb in length and extends from mid-Intron 16 to Exon
23 of ASTN (See Figure 6).
DISCUSSION
The present study used high-density/-throughput genotyping of individual
cases and controls to test over 600,000 genome-wide SNPs for association with
AMD. By genotyping individuals rather than case/control pools, it was possible to
adjust for important factors such as age, birthplace and smoking status. Overall
results of the present study provide evidence for association of SNPs at
chromosome 1q25 and implicate the ASTN gene in development of bilateral early
AMD.
In general, two major concerns exist that may increase false positive results
in whole genome association studies: (1) multiple comparisons and (2) population
stratification. In the present study multiple comparisons were conservatively dealt
with by using Bonferroni corrections for p-values. While Bonferroni adjustments
are likely to be overly conservative, observed differences for ASTN withstood p-
value adjustments in a candidate region analysis. Also for overall genomewide
results, the ASTN SNP case-control differences were considerably more significant
than for any other genotyped SNPs.
67
Regarding population stratification, genetic association studies are
especially susceptible to population stratification as a result of admixture between
subpopulations where disease prevalence varies by subpopulation and cases are not
adequately matched based on ethnicity. Latinos are known to be an admixed
population and therefore, genetic association studies in Latinos should be
cautiously interpreted. In addition to ancestry and acculturation, socioeconomic
status is regarded as an important adjustment in association studies to control for
confounding by racial/ethnic groups. Latino ethnicity is typically associated with a
lower socioeconomic status as compared to non-Hispanic Whites in the U.S.
[Gonzalez Burchard, Borrell et al. 2005]. Interestingly, Hispanics have lower rates
than non-Hispanic Whites for some chronic diseases including AMD despite lower
seocioeconomic measures, a phenomena known as the "Hispanic Paradox"
[Gonzalez Burchard, Borrell et al. 2005]. Because cases and controls for the
present Latino population were from the same neighborhood in the Los Angeles
area, socioeconomic status is likely to be similar. In addition, the present study
takes into account birthplace (U.S., Mexico, or other) in an attempt to control for
ancestry. For these reasons, the mixture of ethnic/genetic subgroups is likely to be
similar among cases and controls. With equivalent heterogeneity of ethnic/genetic
subgroups among cases as compared to controls, stratification bias will not occur
[Ardlie, Lunetta et al. 2002].
68
For a candidate region at 1q, several SNPs in ASTN, an important neuronal
cell-adhesion molecule, remained significantly associated with AMD after
Bonferroni adjustment for 13,710 genotyped SNPs at 1q. When considering
genomewide associations (N = 616,772 SNPs), p-values for ASTN SNP
associations were the most significant, although no p-values remained significant
after Bonferroni adjustment. Chromosomal region 1q contains two previously
implicated genes for AMD, CFH at 1q32 and fibulin-6 (FBLN6) at 1q25.3. No
evidence exists for linkage of ASTN with either CFH or FBLN6. CFH, located
approximately 19.5Mb from ASTN, is a major contributor of AMD based on
studies in Caucasians [Marx 2006]. FBLN6, located approximately 8.5Mb from
ASTN, is an extracellular matrix protein of the fibulin family. The Gln5345Arg
mutation of FBLN6 was found to segregate with AMD in one large family in
Oregon. FBLN6 may be a causal gene for AMD although the attributable fraction
(AF) is nearly negligible due to low allele frequencies, which have previously been
reported to be 0.4% or less depending on race/ethnicity [McKay, Clarke et al. 2004;
Schultz, Weleber et al. 2005]. For ASTN, while the minor allele frequency (MAF)
for the causal SNP in the controls is likely to be small, the magnitudes of the main
effect may be quite large, which may imply a sizable AF (e.g., MAF = 1.7%, OR =
29.6, 95% CI = 3.9 to 223.4, AF = 33% for rs6685449 in Intron 17).
In summary, Bonferroni adjusted p-values for case-control difference using
616,772 genomewide SNPs and 13,710 regional SNPs at 1q23-1q32 especially
69
implicates the ASTN gene located at 1q25.2. A causal SNP in ASTN may reside in
a haplotype block of 52.6Kb in size at the 3' end of the gene. ASTN may be a
major contributor to the development of a phenotype in early AMD likely to
progress. Based on present observations, cases may be approximately 15- to 30-
times more likely to carry high-risk versus low-risk genotype for ASTN compared
to controls. This is the first study to implicate ASTN in AMD susceptibility. In
light of the present findings, immediate further directions currently underway
include (1) verifying an association of ASTN with AMD in an independent
population, (2) identifying the causal variant, and (3) determining a biological role
for ASTN in the retina. The current hypothesis for a biological role for ASTN in
AMD focuses on photoreceptor remodeling in response to initial photoreceptor
degeneration following sufficient insult(s). In vivo studies indicate that without
ASTN, there is a reduction in neuronal migration and abnormal development of
synaptic partner systems occurs in the brain resulting in poorer balance and
coordination of the organism [Adams, Tomoda et al. 2002]. Plausibility for a role
for ASTN in the neural retina may be supported initially by detection of its
expression at the photoreceptor-glial interface during photoreceptor remodeling.
70
Chapter 5: Proposal to Evaluate the Complex Etiology of
AMD from 600K SNP Microarrays, Smoking, Sunlight,
and Diet
Hypothesis and Specific Aims
Genetic studies support that AMD is a complex disease likely to involve
multiple genetic factors with environmental and genetic interactions [Haddad, Chen
et al. 2006]. To date, major genetic contributors are likely to be complement genes
and one or more genes of yet unknown function located at chromosome 10q26
[Fisher, Abecasis et al. 2005; Marx 2006]. AMD rates and phenotype distributions
vary by ethnicity [Klein, Rowland et al. 1995; Klein, Klein et al. 1999; Klein, Peto
et al. 2004]. In light of the complex etiology of AMD, ethnic- and phenotype-
specific risk factors may help explain inconsistencies across studies. The overall
goal of this proposed study is to identify both genetic and environmental risk
factors as well as important gene-environment interactions for AMD that are either
common or specific to different ethnicities (Latinos and Caucasians). Specific aims
include the following:
1) Perform genome-wide genotyping for ethnically different populations
using Affymetrix GeneChip Mapping 500K and 100K single nucleotide
polymorphism (SNP) set microarrays (GeneChip 500K and 100K).
71
2) Survey environmental exposures of smoking, sunlight, and diet for
ethnically different populations.
3) Assess case-control comparisons for genotypes, and environmental
exposures considering gene-environment interactions for candidate
genes or regions. Compare results between ethnicities and phenotypes.
Background and Significance
Familial aggregation, heritability, linkage analysis and candidate gene
association studies have all confirmed a role for genetics in AMD [Haddad, Chen et
al. 2006]. Haddad and colleagues [Haddad, Chen et al. 2006] recently reviewed
genetic epidemiology studies of AMD. To date, several chromosomal regions and
gene candidates have been identified although few have been reproduced across
studies. Evidence suggests ethnic- and phenotype-specific roles for genetic
susceptibility [Klein, Rowland et al. 1995; Klein, Peto et al. 2004; Klein, Klein et
al. 2006], which may help account for inconsistent results across studies.
Variations in complement genes and pleckstrin homology domain
containing family A member 1 and LOC387715 (PLEKHA1/ LOC387715) are
considered major contributors of AMD in Caucasians [Marx 2006]. More
specifically, several studies support that approximately 50% of all AMD in the
population may be explained by the Tyr402His polymorphism of complement
factor H (CFH) using Caucasian populations [Marx 2006]. CFH is located at 1q32,
72
a previously implicated chromosomal region [Weeks, Conley et al. 2001; Fisher,
Abecasis et al. 2005]. PLEKHA1/ LOC387715 is located at 10q26, the strongest
linked chromosal location to AMD based on a recent genome-scan meta-analysis
(GSMA) of six genome-wide scans [Fisher, Abecasis et al. 2005]. GSMA results
also implicate other linked chromosomal regions, which include 1q, 2p, 3p, 3q, 4q,
12q, and 16q [Fisher, Abecasis et al. 2005].
Simultaneous evaluation of multiple genetic and environmental factors as
well as comparison between different ethnicities and phenotypes is particularly
important in understanding the etiology of complex diseases such as AMD.
Ideally, whole genome association studies have the greatest advantage in detecting
consistent genetic contributors of complex disease [Dekker and van Duijn 2003].
Because genotyping technology as well as technical and statistical haplotype
resources have vastly improved in recent years, whole genome association studies
of complex disease are now feasible. Affymetrix GeneChip Mapping 500K and
100K SNP (GeneChip 500K and 100K) sets allow the simultaneous genotyping of
over 600,000 genomewide SNPs. Evidence suggests that genotyping
approximately 500,000 carefully selected genome-wide SNPs is sufficient to
identify haplotype tagging (tagSNPs) and capture nearly all common variation with
80 percent power [Barrett and Cardon 2006]. Although high-throughput
genotyping technology to date may not include the optimum SNP selection for
tagSNPs, genome-wide coverage is comparable among various platforms and
73
valuable information can be extracted, especially with follow-up work in functional
studies [Barrett and Cardon 2006].
After adjusting for age, the most consistent environmental risk factor for
AMD is tobacco smoking. With less consistency, sunlight and diet may influence
risk of AMD [Hawkins, Bird et al. 1999]. Previous studies suggest antioxidant
intake provides protection against AMD [EDCCSG 1992; EDCCSG 1993; Seddon,
Ajani et al. 1994]. The Beaver Dam Eye Study has previously found both a weak
protective effect of zinc intake with early AMD [Mares-Perlman, Klein et al. 1996]
although, in a subsequent study, they found no overall association of zinc or
antioxidants with AMD [VandenLangenberg, Mares-Perlman et al. 1998]. The
Blue Mountain Eye Study found no overall association of antioxidants and AMD
with Autstralian participants [Smith, Mitchell et al. 1999; Flood, Smith et al. 2002;
Kuzniarz, Mitchell et al. 2002]. However, several recent studies suggest a
protective effect of diet and serum levels of antioxidants against AMD [Delcourt,
Cristol et al. 1999; Mares-Perlman, Fisher et al. 2001; Simonelli, Zarrilli et al.
2002; Snellen, Verbeek et al. 2002; Falsini, Piccardi et al. 2003]. Dietary fat and
cholesterol intake may also play role in the development of AMD [Haddad, Chen et
al. 2006].
Despite the importance to clarifying inconsistencies in candidate gene
studies of AMD, few studies have successfully explored gene-environment or gene-
gene interactions in AMD. Schmidt and colleagues recently found evidence that
74
smoking modifies the effects of PLEKHA1/LOC387715 genotype in AMD
[Schmidt, Hauser et al. 2006]. Using logistic regression in a case-control data set
and ordered-subset analysis in an independent family-based data set, Schmidt and
colleagues report that history of cigarette smoking and the Ala29Ser polymorphism
of PLEKHA1/LOC387715 confers statistically significant greater risk when
considered together than each considered alone (p = 0.05 and 0.007, for case-
control and family-based data sets, respectively) [Schmidt, Hauser et al. 2006].
In the present proposal, two different ethnicities (Latinos residing in Los
Angeles, U.S.A. and Caucasians residing in Northern Europe) will be used to
conduct separate population-based whole genome association studies from over
600,000 genotyped SNPs using GeneChip 500K and 100K SNP sets.
Environmental factors, smoking, sunlight and diet will be similarly characterized
for both populations. Gentoyping for both populations will be conducted with
identical technology. Because late AMD is very rare in Latinos, in this population
only early AMD case phenotypes will be explored, which include combinations of
soft drusen and pigmentary abnormalities. Similar early AMD phenotypes will be
used to assess genetic and environmental contributors among the Caucasian
population. Both late AMD forms, geographic atrophy (GA) and choroidal
neovascularization (CNV), will be represented in the Northern European
population. Common findings between each population will be especially
75
interesting to overall AMD susceptibility. Population-specific findings will be
important to understandings inconsistencies previously reported across studies.
Preliminary Studies
Los Angeles Latino Eye Study
Los Angeles Latino Eye Study (LALES) was initially designed to survey
health care and eye disease among Latinos in Los Angeles, and includes over 6357
eligible volunteers age 40 years or older from the city of La Puente with clinical
examination [Varma, Paz et al. 2004]. Briefly, all procedures conformed to the
Declaration of Helsinki for research involving human subjects. The Institutional
Review Board of the University of Southern California approved the project, and
informed consent was obtained from all participants. The participation rate in
LALES was 82 percent. Each LALES participant receives a thorough eye exam
including fundus photography and grading of AMD was performed in a masked
manner. AMD diagnoses were defined according to the Wisconsin Age-Related
Maculopathy Grading System (WARMGS) [Klein, Davis et al. 1991]. Of LALES
participants with a clinical exam and gradable photographs, 5875 subjects, there are
551 prevalent early and 25 late AMD cases. Whole genome amplification of DNA
from stored blood samples, taken at the time of the eye exam for all participants
entering the study from March, 2001 and beyond, was performed by multiple
displacement amplification technology (Repli-g Service, Qiagen Sciences,
76
Germantown, MD) and resulted in validated, usable quality for genotyping assays
among 98% of all LALES samples. Early AMD cases and controls without
advanced diabetic retinopathy among LALES participants will be eligible for this
study. Currently, there are 285 early AMD cases and 2,687 controls with DNA
collected and without advanced diabetic retinopathy.
Previously, a whole genome association study using GeneChip 500K and
100K microarrays was conducted for 101 early AMD cases with soft drusen and/or
pigmentary abnormality phenotypes and 202 age-, smoking status-, and birthplace-
matched controls from LALES. Results especially implicates the astrotactin
(ASTN) gene located at 1q25 using Bonferroni adjusted p-values for case-control
difference using 616,772 genomewide SNPs and 13,710 regional SNPs at 1q23-
1q32 (See Chapter 3). A causal SNP in ASTN may reside in a haplotype block of
52.6Kb in size at the 3' end of the gene.
Northern European Cohort
The Northern European Cohort (NEC) is a population-based Caucasian
cross-sectional study model. While NEC data is fictional, distributions of
demographics and clinical variables are based on the European Eye Study
(EUREYE) [Augood, Vingerling et al. 2006]. The NEC model includes Caucasian
volunteers aged 65 years or older throughout Northern Europe. Each NEC
participant receives a thorough eye exam including fundus photography and
77
grading of AMD was performed in a masked manner. AMD diagnoses were
defined according to the International Classification System [Bird, Bressler et al.
1995], which is based on the Wisconsin Age-Related Maculopathy Grading System
(WARMGS) [Klein, Davis et al. 1991]. Based on a participation rate of 71
percent, NEC includes 6340 (8930/0.71) volunteers with a clinical exam. For 100
percent of those with clinical exams, fundus photographs were gradable and
eligible for the study. Of the eligible NEC participants (N = 6340), there are 2,219
prevalent early and 254 late AMD, 190 with CNV and 64 with GA cases (based on
prevalence rates of 35.0, 4.0, 3.0, and 1.0 % for early, late, CNV, and GA;
respectively). The average age of eligible participants in NEC was 73.4 years.
Prevalence of advanced diabetic retinopathy was 1 (634/6340) percent in NEC.
Study Design
The study design will include two independent population-based
retrospective case-control studies. For participants of both populations,
environmental exposures will be assessed by identical questionnaires. For the
LALES population, samples include whole genome DNA amplification material
from finger-prick blood spots or whole blood donations. For the NEC population,
samples include DNA extract from whole blood donations.
78
Methods
Subjects and samples
LALES
Eligible LALES AMD cases will include those without advanced diabetic
retinopathy, without late AMD (because of the small availability), and without any
evidence of early onset AMD (age at diagnosis less than 45 years). Controls will
be randomly selected from the remaining eligible participants without any AMD
and without advanced diabetic retinopathy. Two controls for each case will be
randomly selected and matched on age and birthplace.
NEC
Eligible NEC AMD cases will include those without advanced diabetic
retinopathy and without any evidence of early onset AMD (age at diagnosis less
than 45 years). Two controls for each case will be randomly selected and matched
on age and birthplace.
Genotyping
Genotyping will be performed according to manufacturer protocol
(Affymetrix, Santa Clara, CA) using whole genome amplification product as
template for GeneChip Mapping 500K and 100K SNP set microarrays for a total of
616,772 SNPs. Briefly, the assays will be performed by digesting template with
79
restriction enzymes, such as XbaI or Hind III, and ligating restriction fragments to
adapters. The adapter-template fragments (250 to 2000 bp in size) will be
amplified using a generic primer that recognizes the adapter sequence. The
amplified segments will then be fragmented, labeled and hybridized to the
GeneChip 500K and 100K microarrays. GeneChip DNA Analysis Software
(GDAS; Affymetrix, Santa Clara, CA) will be used to determine genotype calls.
Environmental Assessments
Identical questionnaires will be administered to participants from both
populations. For LALES participants, all environmental exposure questionnaires
will be translated to Spanish. For NEC participants, the questionnaires will be
translated into the national language of their respective European country of
residence. Eligible cases and controls will be sent a letter and telephoned to
schedule an in-person interview at the subjects’ home or other location of their
choice. Interviews will be conducted with consenting subjects. Smoking history,
sunlight, and diet history questionnaires are shown in Appendices 1, 2, and 3;
respectively. All environmental data will be collected using a computerized
database system with built-in quality control checks at the time of the in-home
interview based on pre-coded values in the questionnaires. Computerized data will
be further screened and processed using SAS version 9.1 (SAS Institute, Cary, NC)
to generate appropriate covariate variables. Appropriate covariate variables will be
80
imported into Genetrix (Epicenter Software, Pasadena) software for further
analyses comparing cases and controls.
Smoking History Questionnaire
The smoking history questionnaire consists of 34 questions from the
National Cancer Institute, Division of Cancer Epidemiology and Genetics (DCEG)
[http://dceg/cancer.gov/modules/tobacco.pdf] (Appendix A). Regular cigarette
smoking will be assessed by 7 questions beginning with the yes or no questions:
(1) "Have you ever smoked a total of 100 cigarettes or more over your lifetime?"
and (2) Did you ever smoke cigarettes regularly, that is, at least one per day for six
months or longer?" For individuals answering "yes" to these first two questions,
using the average number of cigarettes smoked per day, pack-years of smoking will
be calculated as (cigarettes per day x years smoked)/20 cigarettes per pack. The
most general measurement of smoking history will be generated as "ever or never".
Other tobacco related information includes questions on cigarette filters, cigars,
pipes, and chewing tobacco.
Sun Exposure Questionnaire
The sun exposure questionnaire consists of 23 questions from the National
Cancer Institute, Division of Cancer Epidemiology and Genetics (DCEG)
http://dceg/cancer.gov/modules/SunExposure2.pdf] (Appendix B). Sun exposure
81
assessment will begin with questions on eye color, natural hair color at age 20, and
having light, medium or dark skin complexion. Questions will also include ages at
which sun protective clothing (hat or long-sleeved shirt) or sunscreen lotions were
used. If sunscreen lotions were used, sunscreen protective factor number will be
asked. Skin reaction to strong sunlight will also be ask by a series of questions
regarding how badly sunburns, frequency of sunburns and ages at which blistering
sunburns occurred. Ages and frequency at which time was spent in the strong
sunlight (between 9am to 3pm) will also be included. Use of sunlamps and tanning
booths will also be asked. Sun exposure variables that will be created based on
answers to questions include cumulative sunlight exposure, an index of skin
sensitivity to sunlight (strong-, medium-, light- sensitivity), and cumulative sun
protection use.
Diet Questionnaire
The Diet questionnaire will consist of 144 food-frequency questions from
the US National Cancer Institutes’ (Block’s) Health Habits and History
Questionnaire [Block, Hartman et al. 1986] and will include questions on portion
sizes (Appendix C). All diet questions will be asked for the year(s) prior to AMD
diagnosis for cases and for the previous year for controls. In general, the
frequency choice ranges from "never'" to "6+ per day" or "never" to 2+ per day".
Portion sizes will also be asked according to the number of pieces, or either small,
82
medium, or large; using visual aids and 3-dimensional models of appropriate
volumes. Multivitamins and supplements will also be included on the FFQ. Years
of use (“less than 1”, “1 to 4”, “5 to 9”, or “10 or more”) will be asked for multi-
vitamins, beta-carotene, vitamin A, vitamin C, vitamin E, and calcium. Intake on
other supplements will also be included as "more than 1 time per week". Only one
value will be assigned for each food frequency choice by converting the unit of
year, month or day to the number of times per day. For ranges of frequencies, the
mid-value will be assigned before conversion, e.g. 2.5 times per month will be
assigned if the ‘2-3 per month’ category is given. Using a standard food item-
specific nutrient content table, average daily nutrient values will be computed for
each subject based on reported food-frequency and portions size values. For
example, intake of each food item will be converted to grams consumed per day by
multiplying the frequency and portion reported against the table of item-specific
nutrient content.
Statistical Analyses
Genetrix (Epicenter, Pasadena, CA) software will be used to determine
differences between cases and controls for individual genotyped SNPs and
haplotypes. Bonferroni adjusted p-values will be used to correct for multiple
comparisons. The χ
2
test for Hardy-Weinberg equilibrium will be used to evaluate
frequency distributions of genotypes among cases and controls separately.
83
Conditional logistic regression models will be used to compare recessive,
dominant, and log-additive genetic models for cases versus age-, birthplace-, -
matched controls. Fisher's exact probabilities will be computed when cell counts
were low (below 5). Using algorithms implemented within Genetrix similar to the
software Haploview (www.broad.mit.edu/personal/jcbarret/haplo/), SNP genotype
will be used to characterize the linkage disequilibrium (LD) and define haplotype
blocks for selected regions. LD block structure will be examined using criteria
described by Gabriel et al., 2002, [Gabriel, Schaffner et al. 2002] with 90%
confidence bounds on the D’ statistic and only SNPs with minor allele frequencies
equal to 10% or greater will be used. All LD block boundaries will be tested for
suitability by evaluating the effects of including surrounding SNPs, with minor
allele frequencies as low as 5%, on the total number and identities of common
haplotypes within each block. For each LD block defined, haplotypes will be
reconstructed and haplotype tagging SNPs (htSNPs) will be selected as the
minimum set of SNPs that can be used to define haplotypes within each block
using algorithms implemented within Genetrix similar to the software, tagSNPs
(www-rcf.usc.edu/~stram/), which estimates the squared correlation between the
true haplotype, R
2
h
, and the estimate from the Excoffier-Slatkin expectation-
maximization (E-M) algorithm [Excoffier and Slatkin 1995]. After forcing
nonsynonomous SNPs to be included as htSNPs, all SNPs of each block that
provide an R
2
h
of 0.7 or greater for haplotypes with an estimated frequency of 5%
84
or greater will be designated as htSNPs, similar to that described in Stram, DO et
al., [Stram, Haiman et al. 2003]. Haplotype dosage estimates, number of copies (0,
1, or 2) of a given haplotype for each individual, will be computed from the E-M
algorithm using individual htSNP genotyping data and haplotype frequency
estimates from the combined case and control data set (justified by the null
hypothesis that there is no difference in haplotype frequencies between cases and
controls).
Gene-environment interaction covariates will be constructed for genes with
the greatest significant p-values for candidate genomic regions. For each candidate
genomic regions previously identified by GSMA [Fisher, Abecasis et al. 2005]
(1q23.3-1q32, 2p12-2p25.1, 3p14.1-3p25.3, 3q12.3-3q22.1, 4q13.3-4q24, 4q28.3-
4q32.1, 10q23.3-10qter, 12q23.2-12q24.3, 16p13-16q23.1), candidate genes will be
chosen based on Bonferroni adjusted p-values or the most significant p-values.
Highly significant SNPs or haplotypes from these candidate genes will then be used
to test for interactions with smoking, sunlight, and diet.
Multiple Testing Considerations, Power & Sample Size
Initially, Bonferroni adjusted p-values will be used to control for multiple
comparisons for candidate genomic regions. However, because it is likely that
several htSNPs may actually be associated with each other, traditional methods of
power and sample size determination are likely to be highly conservative in their
85
adjustments of multiple comparisons. Therefore, to determine the most appropriate
working significance threshold, permutation analysis will be performed with all
htSNPs. (Alternatively, false-discovery-rate analysis will also be considered by
automatic calibration of p-values using the software, QVALUE
(http://genomine.org/qvalue), to describe statistically significant findings. [Storey
and Tibshirani 2003]) Therefore, the power to determine statistically significant
findings in light of simultaneous multiple testing of genome-wide genotypes will be
most appropriately addressed empirically. Table 16 below shows the overall
distribution of available participants from each population. In table 17 below,
using traditional calculations assuming 80% power, and a significance level set at
either 0.05 or 0.000001 (a primary adjustment for multiple comparisons),
detectable relative risks are described for gene-environment interactions based on
log-additive genetic models for a sample size of 200 cases (with 600 matched
controls) and 700 cases (with 2100 matched controls). From table 17, assuming
cases are 1.6-times and 1.4-times more likely than controls to have a particular
genetic and environmental exposures, respectively, detectable gene-environmental
relative risks are 8.1 or 2.7 for a sample size of 200 or 700 cases (with 3 to 1 age-
matched controls), respectively, with 80% power at a significance level of
0.000001.
86
Table 16: Available participants by ethnically different populations.
LALES
1
NEC
2
Demographic
N = 5874 N = 6340
Age (mean ± SE
3
) 54.9 ± 0.14 73.4 ± 0.12
Diabetic
Retinopathy (n %)
890 (15%) 634 (1%)
AMD
None (n %) 5298 (88 %) 3867 (61%)
Early AMD (n %) 551 (10 %) 2219 (35 %)
Late AMD (n %) 25 (0.5 %) 254 (4 %)
CNV
4
(n %) 17 (0.3 %) 190 (3 %)
GA
5
(n %) 8 (0.2 %) 64 (1 %)
1
LALES = Los Angeles Latino Eye Study.
2
NEC = Northern European Cohort.
3
SE = standard error.
4
CNV = choroidal neovascularization.
5
GA = geographic atrophy.
Table 17: Detectable relative risks for gene-environment interactions using log-
additive genetic models with 80% power.
Prevalence 200 Cases + 600 Controls 700 Cases + 2100 Controls
Significance Level
&
Relative Risks
(unadjusted)
MAF
1
E
2
R_g
3
R_e
4
R_ge
5
R_g
3
R_e
4
R_ge
5
0.1 0.1 2 2 4.7 1.6 1.6 2.5
0.3 2.5 1.7 3.4 1.8 1.5 2
0.25 0.1 1.6 2.7 2.5 1.4 1.9 1.7
0.3 2.1 1.9 2.6 1.6 1.5 1.7
0.5 0.1 1.7 4.2 2.4 1.4 2.4 1.6
p -value = 0.05
&
R_g
3
= 1.3
R_e
4
= 1.3
0.3 1.9 3.2 2 1.5 2 1.4
p - value = 1E-6
&
R_g
3
= 1.6
R_e
4
= 1.4
0.1 0.3 7.8 6.0 8.1 2.7 2.2 2.7
1
MAF = minor allele frequency.
2
E = environmental factor frequency.
3
R_g = relative risk for genetic factor.
4
R_e = relative risk for environmental factor.
5
R_ge = relative risk for gene-environment interaction.
87
Chapter 6: Summary and Conclusions
Present results from a candidate gene association study indicate that a
subset of early AMD cases with bilateral intermediate to large soft drusen are 1.8
times more likely to carry either the homozygous or heterozygous His402 genotype
of the complement factor H (CFH) Tyr402His polymorphism (See Chapter 2).
Among Latino controls, the His402 allele frequency is 17 percent, while the
reported allele frequency is at least 34 percent among Caucasians. Several
previous studies in Caucasians suggest that the CFH Tyr402His polymorphism
plays a major role in AMD explaining approximately 50 percent of AMD overall
[Conley, Thalamuthu et al. 2005; Edwards, Ritter et al. 2005; Hageman, Anderson
et al. 2005; Haines, Hauser et al. 2005; Klein, Zeiss et al. 2005; Zareparsi, Branham
et al. 2005]. However, the population attributable fraction for Latinos is only 18
percent based on carrying either the His402 heterozygous or homozygous
genotypes. Comparison of the impact of CFH Tyr402His for overall AMD risk
implies that a major role for CFH may be more specific to Caucasians than Latinos.
Further candidate gene association analysis simultaneously considering CFH and
manganese superoxide dismutase (MnSOD) genotypes, suggest these genes may
play a dependent role in developing pigmentary abnormalities and bilateral soft
drusen phenotypes in early AMD (OR = 1.8, 95% CI = 1.01 to 3.2 or OR = 2.9,
95% CI = 1.2 to 7.1 for carrying both high-risk versus low-risk genotypes
88
respectively for early AMD cases with any pigmentary abnormalities or bilateral
soft drusen compared to controls; See Chapter 3). Relevance of this interaction
observed especially among those with pigmentary abnormalities compared to
controls may be due to a phenotype that is especially susceptible to oxidative stress.
The mechanism of interaction for MnSOD and CFH in AMD pathogenesis remains
unclear.
Roles for both oxidative stress and inflammatory mechanisms have been
established in AMD pathogenesis [Zarbin 2004; Donoso, Kim et al. 2006]. An
interesting hypothesis recently proposed by Taufer and colleagues [Taufer, Peres et
al. 2005] suggests that homozygous Ala16 MnSOD genotype may be associated
with stronger innate immunity due to increased generation of hydrogen peroxide
levels, especially because activated immune cells, such as macrophages, produce
hydrogen peroxide during inflammation. This hypothesis is relevant to the present
interaction observed between MnSOD and CFH, an important regulator of innate
immunity. Further, recent evidence from studies of hepatitis C suggests a central
role for mitochondria in innate immunity with the discovery of a nuclear encoded
protein, which localizes to the outer mitochondrial membrane and activates
signaling pathways to enable downstream transcription and maximize cytokine
production [McWhirter, Tenoever et al. 2005; Piccoli, Scrima et al. 2006]. This
recent observation shows that mitochondrial dysfunction, which is inherently tied
89
to oxidative stress, could have influential effects on innate immunity through
nuclear encoded, mitochondrial localized proteins.
Present results from a whole genome association study using GeneChip
500K and 100K SNP set microarrays strongly implicate the Astrotactin (ASTN)
gene in AMD development in Latinos. This is the first study to implicate ASTN in
AMD susceptibility. ASTN is located at 1q25, an important genomic region based
on previous linkage studies [Fisher, Abecasis et al. 2005]. ASTN may be a major
contributor to the development of a phenotype in early AMD likely to progress
such as bilateral intermediate to large soft drusen. ASTN is an important neural
cell adhesion molecule [Adams, Tomoda et al. 2002]. The current hypothesis for a
biological role for ASTN in AMD focuses on photoreceptor remodeling in response
to initial photoreceptor degeneration following sufficient insult(s). The present
findings of a statistical association between ASTN and early AMD warrant
immediate further directions which include (1) verifying an association of ASTN
with AMD in an independent population, (2) identifying the causal variant, and (3)
determining a biological role for ASTN in the retina.
A whole genome association study that also explores interactions with
important environment factors is likely to be the logical next advanced
epidemiologic tool that may further elucidate the complex etiology of AMD. The
final project in this dissertation is a proposal for a large whole genome gene-
environment association study that explores gene interactions with smoking,
90
sunlight, and diet in two independent and ethnically distinct populations. Both
common and ethnic-specific associations will be helpful for fully understanding
AMD susceptibility.
91
Bibliography
A.R.E.D.S. (2001). "A randomized, placebo-controlled, clinical trial of high-dose
supplementation with vitamins C and E, beta carotene, and zinc for age-
related macular degeneration and vision loss: AREDS report no. 8." Arch
Ophthalmol 119(10): 1417-36.
Adams, N. C., T. Tomoda, et al. (2002). "Mice that lack astrotactin have slowed
neuronal migration." Development 129(4): 965-72.
Antunes, F., A. Salvador, et al. (1996). "Lipid peroxidation in mitochondrial inner
membranes. I. An integrative kinetic model." Free Radic Biol Med 21(7):
917-43.
Ardlie, K. G., K. L. Lunetta, et al. (2002). "Testing for population subdivision and
association in four case-control studies." Am J Hum Genet 71(2): 304-11.
Augood, C. A., J. R. Vingerling, et al. (2006). "Prevalence of age-related
maculopathy in older Europeans: the European Eye Study (EUREYE)."
Arch Ophthalmol 124(4): 529-35.
Barrett, J. C. and L. R. Cardon (2006). "Evaluating coverage of genome-wide
association studies." Nat Genet 38(6): 659-62.
Beatty, S., H. Koh, et al. (2000). "The role of oxidative stress in the pathogenesis of
age-related macular degeneration." Surv Ophthalmol 45(2): 115-34.
Bird, A. C., N. M. Bressler, et al. (1995). "An international classification and
grading system for age-related maculopathy and age-related macular
degeneration. The International ARM Epidemiological Study Group." Surv
Ophthalmol 39(5): 367-74.
Block, G., A. M. Hartman, et al. (1986). "A data-based approach to diet
questionnaire design and testing." Am J Epidemiol 124(3): 453-69.
Boulton, M. and P. Dayhaw-Barker (2001). "The role of the retinal pigment
epithelium: topographical variation and ageing changes." Eye 15(Pt 3): 384-
9.
Congdon, N., B. O'Colmain, et al. (2004). "Causes and prevalence of visual
impairment among adults in the United States." Arch Ophthalmol 122(4):
477-85.
92
Conley, Y. P., A. Thalamuthu, et al. (2005). "Candidate gene analysis suggests a
role for fatty acid biosynthesis and regulation of the complement system in
the etiology of age-related maculopathy." Hum Mol Genet 14(14): 1991-
2002.
Crabb, J. W., M. Miyagi, et al. (2002). "Drusen proteome analysis: an approach to
the etiology of age-related macular degeneration." Proc Natl Acad Sci U S
A 99(23): 14682-7.
De Jong, P. T., C. C. Klaver, et al. (1997). "Familial aggregation of age-related
maculopathy." Am J Ophthalmol 124(6): 862-3.
Dekker, M. C. and C. M. van Duijn (2003). "Prospects of genetic epidemiology in
the 21st century." Eur J Epidemiol 18(7): 607-16.
Delcourt, C., J. P. Cristol, et al. (1999). "Age-related macular degeneration and
antioxidant status in the POLA study. POLA Study Group. Pathologies
Oculaires Liees a l'Age." Arch Ophthalmol 117(10): 1384-90.
Donoso, L. A., D. Kim, et al. (2006). "The role of inflammation in the pathogenesis
of age-related macular degeneration." Surv Ophthalmol 51(2): 137-52.
Dontsov, A. E., R. D. Glickman, et al. (1999). "Retinal pigment epithelium pigment
granules stimulate the photo-oxidation of unsaturated fatty acids." Free
Radic Biol Med 26(11-12): 1436-46.
EDCCSG (1992). "Risk factors for neovascular age-related macular degeneration.
The Eye Disease Case-Control Study Group." Arch Ophthalmol 110(12):
1701-8.
EDCCSG (1993). "Antioxidant status and neovascular age-related macular
degeneration. Eye Disease Case-Control Study Group." Arch Ophthalmol
111(1): 104-9.
Edwards, A. O., R. Ritter, 3rd, et al. (2005). "Complement factor H polymorphism
and age-related macular degeneration." Science 308(5720): 421-4.
Esfandiary, H., U. Chakravarthy, et al. (2005). "Association study of detoxification
genes in age related macular degeneration." Br J Ophthalmol 89(4): 470-4.
Evans, J. R. (2001). "Risk factors for age-related macular degeneration." Prog Retin
Eye Res 20(2): 227-53.
93
Excoffier, L. and M. Slatkin (1995). "Maximum-likelihood estimation of molecular
haplotype frequencies in a diploid population." Mol Biol Evol 12(5): 921-7.
Falsini, B., M. Piccardi, et al. (2003). "Influence of short-term antioxidant
supplementation on macular function in age-related maculopathy: a pilot
study including electrophysiologic assessment." Ophthalmology 110(1): 51-
60; discussion 61.
Ferris, F. L., M. D. Davis, et al. (2005). "A simplified severity scale for age-related
macular degeneration: AREDS Report No. 18." Arch Ophthalmol 123(11):
1570-4.
Fisher, S. A., G. R. Abecasis, et al. (2005). "Meta-analysis of genome scans of age-
related macular degeneration." Hum Mol Genet 14(15): 2257-64.
Flood, V., W. Smith, et al. (2002). "Dietary antioxidant intake and incidence of
early age-related maculopathy: the Blue Mountains Eye Study."
Ophthalmology 109(12): 2272-8.
Fraser-Bell, S., J. Wu, et al. (2006). "Smoking, alcohol intake, estrogen use, and
age-related macular degeneration in Latinos: the Los Angeles Latino Eye
Study." Am J Ophthalmol 141(1): 79-87.
Friedman, D. S., B. J. O'Colmain, et al. (2004). "Prevalence of age-related macular
degeneration in the United States." Arch Ophthalmol 122(4): 564-72.
Gabriel, S. B., S. F. Schaffner, et al. (2002). "The structure of haplotype blocks in
the human genome." Science 296(5576): 2225-9.
Gauderman, W. and J. Morrison (2006). "QAUANTO 1.1: A computer program for
power and sample size calculations for genetic-epidemiology studies,
http://hyra.usc.edu/gxe."
Gold, B., J. E. Merriam, et al. (2006). "Variation in factor B (BF) and complement
component 2 (C2) genes is associated with age-related macular
degeneration." Nat Genet 38(4): 458-62.
Gonzalez Burchard, E., L. N. Borrell, et al. (2005). "Latino populations: a unique
opportunity for the study of race, genetics, and social environment in
epidemiological research." Am J Public Health 95(12): 2161-8.
94
Gotoh, N., R. Yamada, et al. (2006). "No association between complement factor H
gene polymorphism and exudative age-related macular degeneration in
Japanese." Hum Genet 19: 19.
Grassi, M. A., J. H. Fingert, et al. (2006). "Ethnic variation in AMD-associated
complement factor H polymorphism p.Tyr402His." Hum Mutat 27(9): 921-
5.
Haddad, S., C. A. Chen, et al. (2006). "The genetics of age-related macular
degeneration: a review of progress to date." Surv Ophthalmol 51(4): 316-
63.
Hageman, G. S., D. H. Anderson, et al. (2005). "A common haplotype in the
complement regulatory gene factor H (HF1/CFH) predisposes individuals to
age-related macular degeneration." Proc Natl Acad Sci U S A 102(20):
7227-32.
Hageman, G. S., P. J. Luthert, et al. (2001). "An integrated hypothesis that
considers drusen as biomarkers of immune-mediated processes at the RPE-
Bruch's membrane interface in aging and age-related macular
degeneration." Prog Retin Eye Res 20(6): 705-32.
Haines, J. L., M. A. Hauser, et al. (2005). "Complement factor H variant increases
the risk of age-related macular degeneration." Science 308(5720): 419-21.
Hare, J. F. and R. Hodges (1982). "Turnover of mitochondrial matrix polypeptides
in hepatoma monolayer cultures." J Biol Chem 257(21): 12950-3.
Hass, M. A. and D. Massaro (1987). "Differences in CuZn superoxide dismutase
induction in lungs of neonatal and adult rats." Am J Physiol 253(1 Pt 1):
C66-70.
Hawkins, B. S., A. Bird, et al. (1999). "Epidemiology of age-related macular
degeneration." Mol Vis 5: 26.
Hiroi, S., H. Harada, et al. (1999). "Polymorphisms in the SOD2 and HLA-DRB1
genes are associated with nonfamilial idiopathic dilated cardiomyopathy in
Japanese." Biochem Biophys Res Commun 261(2): 332-9.
Hsu, J. L., Y. Hsieh, et al. (1996). "Catalytic properties of human manganese
superoxide dismutase." J Biol Chem 271(30): 17687-91.
95
Jonasson, F., A. Arnarsson, et al. (2003). "The prevalence of age-related
maculopathy in iceland: Reykjavik eye study." Arch Ophthalmol 121(3):
379-85.
Kimura, K., Y. Isashiki, et al. (2000). "Genetic association of manganese
superoxide dismutase with exudative age-related macular degeneration."
Am J Ophthalmol 130(6): 769-73.
Klaver, C. C., R. C. Wolfs, et al. (1998). "Genetic risk of age-related maculopathy.
Population-based familial aggregation study." Arch Ophthalmol 116(12):
1646-51.
Klein, R., M. D. Davis, et al. (1991). "The Wisconsin age-related maculopathy
grading system." Ophthalmology 98(7): 1128-34.
Klein, R., B. E. Klein, et al. (1999). "The prevalence of age-related maculopathy by
geographic region and ethnicity." Prog Retin Eye Res 18(3): 371-89.
Klein, R., B. E. Klein, et al. (1999). "Age-related maculopathy in a multiracial
United States population: the National Health and Nutrition Examination
Survey III." Ophthalmology 106(6): 1056-65.
Klein, R., B. E. Klein, et al. (2006). "Prevalence of age-related macular
degeneration in 4 racial/ethnic groups in the multi-ethnic study of
atherosclerosis." Ophthalmology 113(3): 373-80.
Klein, R., B. E. Klein, et al. (1992). "Prevalence of age-related maculopathy. The
Beaver Dam Eye Study." Ophthalmology 99(6): 933-43.
Klein, R., B. E. Klein, et al. (2002). "Ten-year incidence and progression of age-
related maculopathy: The Beaver Dam eye study." Ophthalmology 109(10):
1767-79.
Klein, R., T. Peto, et al. (2004). "The epidemiology of age-related macular
degeneration." Am J Ophthalmol 137(3): 486-95.
Klein, R., M. L. Rowland, et al. (1995). "Racial/ethnic differences in age-related
maculopathy. Third National Health and Nutrition Examination Survey."
Ophthalmology 102(3): 371-81.
Klein, R. J., C. Zeiss, et al. (2005). "Complement factor H polymorphism in age-
related macular degeneration." Science 308(5720): 385-9.
96
Kuzniarz, M., P. Mitchell, et al. (2002). "Use of vitamin and zinc supplements and
age-related maculopathy: the Blue Mountains Eye Study." Ophthalmic
Epidemiol 9(4): 283-95.
Lacza, Z., J. A. Snipes, et al. (2003). "Mitochondrial nitric oxide synthase is not
eNOS, nNOS or iNOS." Free Radic Biol Med 35(10): 1217-28.
Lamb, L. E. and J. D. Simon (2004). "A2E: a component of ocular lipofuscin."
Photochem Photobiol 79(2): 127-36.
MacMillan-Crow, L. A., J. P. Crow, et al. (1996). "Nitration and inactivation of
manganese superoxide dismutase in chronic rejection of human renal
allografts." Proc Natl Acad Sci U S A 93(21): 11853-8.
MacMillan-Crow, L. A., J. P. Crow, et al. (1998). "Peroxynitrite-mediated
inactivation of manganese superoxide dismutase involves nitration and
oxidation of critical tyrosine residues." Biochemistry 37(6): 1613-22.
Magnusson, K. P., S. Duan, et al. (2006). "CFH Y402H confers similar risk of soft
drusen and both forms of advanced AMD." PLoS Med 3(1): e5.
Mares-Perlman, J. A., A. I. Fisher, et al. (2001). "Lutein and zeaxanthin in the diet
and serum and their relation to age-related maculopathy in the third national
health and nutrition examination survey." Am J Epidemiol 153(5): 424-32.
Mares-Perlman, J. A., R. Klein, et al. (1996). "Association of zinc and antioxidant
nutrients with age-related maculopathy." Arch Ophthalmol 114(8): 991-7.
Marx, J. (2006). "Genetics. A clearer view of macular degeneration." Science
311(5768): 1704-5.
McCarty, C. A., B. N. Mukesh, et al. (2001). "Risk factors for age-related
maculopathy: the Visual Impairment Project." Arch Ophthalmol 119(10):
1455-62.
McKay, G. J., S. Clarke, et al. (2004). "A novel diagnostic test detects a low
frequency of the hemicentin Gln5345Arg variant among Northern Irish age
related macular degeneration patients." Mol Vis 10: 682-7.
McWhirter, S. M., B. R. Tenoever, et al. (2005). "Connecting mitochondria and
innate immunity." Cell 122(5): 645-7.
97
Mendes, P. (1993). "GEPASI: A software package for modelling the dynamics,
steady states and control of biochemical and other systems." Comp Appl
Biosci 9: 563-571.
Mukesh, B. N., P. N. Dimitrov, et al. (2004). "Five-year incidence of age-related
maculopathy: the Visual Impairment Project." Ophthalmology 111(6):
1176-82.
Oshima, Y., T. Ishibashi, et al. (2001). "Prevalence of age related maculopathy in a
representative Japanese population: the Hisayama study." Br J Ophthalmol
85(10): 1153-7.
Piccoli, C., R. Scrima, et al. (2006). "Mitochondrial dysfunction in hepatitis C virus
infection." Biochim Biophys Acta 19: 19.
Purves, D., G. J. Augustine, et al. (2001). Neuroscience. Sunderland, MA, Sinauer
Associates, Inc.
Quijano, C., D. Hernandez-Saavedra, et al. (2001). "Reaction of peroxynitrite with
Mn-superoxide dismutase. Role of the metal center in decomposition
kinetics and nitration." J Biol Chem 276(15): 11631-8.
Radi, R., A. Cassina, et al. (2002). "Peroxynitrite reactions and formation in
mitochondria." Free Radic Biol Med 33(11): 1451-64.
Sarangarajan, R. and S. P. Apte (2005). "Melanin aggregation and polymerization:
possible implications in age-related macular degeneration." Ophthalmic Res
37(3): 136-41.
Sarks, S. H., J. J. Arnold, et al. (1999). "Early drusen formation in the normal and
aging eye and their relation to age related maculopathy: a
clinicopathological study." Br J Ophthalmol 83(3): 358-68.
Schaumberg, D. A., W. G. Christen, et al. (2006). "A Prospective Assessment of
the Y402H Variant in Complement Factor H, Genetic Variants in C-
Reactive Protein, and Risk of Age-Related Macular Degeneration." Invest
Ophthalmol Vis Sci 47(6): 2336-40.
Schmidt, S., M. A. Hauser, et al. (2006). "Cigarette smoking strongly modifies the
association of LOC387715 and age-related macular degeneration." Am J
Hum Genet 78(5): 852-64.
98
Schultz, D. W., R. G. Weleber, et al. (2005). "HEMICENTIN-1 (FIBULIN-6) and
the 1q31 AMD locus in the context of complex disease: review and
perspective." Ophthalmic Genet 26(2): 101-5.
Seddon, J. M., U. A. Ajani, et al. (1997). "Familial aggregation of age-related
maculopathy." Am J Ophthalmol 123(2): 199-206.
Seddon, J. M., U. A. Ajani, et al. (1994). "Dietary carotenoids, vitamins A, C, and
E, and advanced age-related macular degeneration. Eye Disease Case-
Control Study Group." Jama 272(18): 1413-20.
Seddon, J. M., J. Cote, et al. (2005). "The US twin study of age-related macular
degeneration: relative roles of genetic and environmental influences." Arch
Ophthalmol 123(3): 321-7.
Seddon, J. M., S. L. Santangelo, et al. (2003). "A genomewide scan for age-related
macular degeneration provides evidence for linkage to several chromosomal
regions." Am J Hum Genet 73(4): 780-90.
Seddon, J. M., S. Sharma, et al. (2006). "Evaluation of the clinical age-related
maculopathy staging system." Ophthalmology 113(2): 260-6.
Simonelli, F., F. Zarrilli, et al. (2002). "Serum oxidative and antioxidant parameters
in a group of Italian patients with age-related maculopathy." Clin Chim
Acta 320(1-2): 111-5.
Smith, W., J. Assink, et al. (2001). "Risk factors for age-related macular
degeneration: Pooled findings from three continents." Ophthalmology
108(4): 697-704.
Smith, W., P. Mitchell, et al. (1999). "Dietary antioxidants and age-related
maculopathy: the Blue Mountains Eye Study." Ophthalmology 106(4): 761-
7.
Snellen, E. L., A. L. Verbeek, et al. (2002). "Neovascular age-related macular
degeneration and its relationship to antioxidant intake." Acta Ophthalmol
Scand 80(4): 368-71.
Sparrow, J. R., N. Fishkin, et al. (2003). "A2E, a byproduct of the visual cycle."
Vision Res 43(28): 2983-90.
Storey, J. D. and R. Tibshirani (2003). "Statistical significance for genomewide
studies." Proc Natl Acad Sci U S A 100(16): 9440-5.
99
Stram, D. O., C. A. Haiman, et al. (2003). "Choosing haplotype-tagging SNPS
based on unphased genotype data using a preliminary sample of unrelated
subjects with an example from the Multiethnic Cohort Study." Hum Hered
55(1): 27-36.
Sutton, A., A. Imbert, et al. (2005). "The manganese superoxide dismutase
Ala16Val dimorphism modulates both mitochondrial import and mRNA
stability." Pharmacogenet Genomics 15(5): 311-9.
Sutton, A., H. Khoury, et al. (2003). "The Ala16Val genetic dimorphism modulates
the import of human manganese superoxide dismutase into rat liver
mitochondria." Pharmacogenetics 13(3): 145-57.
Taufer, M., A. Peres, et al. (2005). "Is the Val16Ala manganese superoxide
dismutase polymorphism associated with the aging process?" J Gerontol A
Biol Sci Med Sci 60(4): 432-8.
Terrasa, A. M., M. H. Guajardo, et al. (2003). "Peroxidation stimulated by lipid
hydroperoxides on bovine retinal pigment epithelium mitochondria: effect
of cellular retinol-binding protein." Int J Biochem Cell Biol 35(7): 1071-84.
Thomas, D. D., X. Liu, et al. (2001). "The biological lifetime of nitric oxide:
implications for the perivascular dynamics of NO and O2." Proc Natl Acad
Sci U S A 98(1): 355-60.
Tuo, J., C. M. Bojanowski, et al. (2004). "Genetic factors of age-related macular
degeneration." Prog Retin Eye Res 23(2): 229-49.
VandenLangenberg, G. M., J. A. Mares-Perlman, et al. (1998). "Associations
between antioxidant and zinc intake and the 5-year incidence of early age-
related maculopathy in the Beaver Dam Eye Study." Am J Epidemiol
148(2): 204-14.
Varma, R., S. Fraser-Bell, et al. (2004). "Prevalence of age-related macular
degeneration in Latinos: the Los Angeles Latino eye study." Ophthalmology
111(7): 1288-97.
Varma, R., S. H. Paz, et al. (2004). "The Los Angeles Latino Eye Study: design,
methods, and baseline data." Ophthalmology 111(6): 1121-31.
Wang, J. J., P. Mitchell, et al. (1998). "Bilateral involvement by age related
maculopathy lesions in a population." Br J Ophthalmol 82(7): 743-7.
100
Weeks, D. E., Y. P. Conley, et al. (2001). "Age-related maculopathy: an expanded
genome-wide scan with evidence of susceptibility loci within the 1q31 and
17q25 regions." Am J Ophthalmol 132(5): 682-92.
Winkler, B. S., M. E. Boulton, et al. (1999). "Oxidative damage and age-related
macular degeneration." Mol Vis 5: 32.
Zarbin, M. A. (2004). "Current concepts in the pathogenesis of age-related macular
degeneration." Arch Ophthalmol 122(4): 598-614.
Zareparsi, S., K. E. Branham, et al. (2005). "Strong Association of the Y402H
Variant in Complement Factor H at 1q32 with Susceptibility to Age-Related
Macular Degeneration." Am J Hum Genet 77(1): 149-53.
101
Appendix A: Smoking History Questionnaire
102
103
104
105
Appendix B: Sun Exposure Questionnaire
106
107
108
109
110
Appendix C: Diet History Questionnaire
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
Abstract (if available)
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Genomic risk factors associated with Ewing Sarcoma susceptibility
PDF
Association of traffic-related air pollution and age-related macular degeneration in the Los Angeles Latino Eye Study
Asset Metadata
Creator
Tedeschi-Blok, Nicole
(author)
Core Title
Genetic association studies of age-related macular degeneration from candidate gene to whole genome
School
Keck School of Medicine
Degree
Doctor of Philosophy
Degree Program
Molecular Epidemiology
Publication Date
01/18/2009
Defense Date
08/28/2006
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
Astrotactin,Complement Factor H,manganese superoxide dismutase,OAI-PMH Harvest,whole genome association study
Language
English
Advisor
Hinton, David R. (
committee member
), Ingles, Sue A. (
committee member
), Triche, Timothy J. (
committee member
)
Creator Email
tedeschi@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-m226
Unique identifier
UC1487138
Identifier
etd-Tedeschi-20070118 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-161358 (legacy record id),usctheses-m226 (legacy record id)
Legacy Identifier
etd-Tedeschi-20070118.pdf
Dmrecord
161358
Document Type
Dissertation
Rights
Tedeschi-Blok, Nicole
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Repository Name
Libraries, University of Southern California
Repository Location
Los Angeles, California
Repository Email
cisadmin@lib.usc.edu
Tags
Astrotactin
Complement Factor H
manganese superoxide dismutase
whole genome association study