EXPLORING MITOCHONDRIAL DNA ABNORMALITIES IN HIV-EXPOSED

UNINFECTED CHILDREN DIAGNOSED WITH AUTISM SPECTRUM DISORDER: A

CASE CONTROL STUDY

by

Matthew Budd

B.A., Mount Royal University, 2012

A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF

THE REQUIREMENTS FOR THE DEGREE OF

MASTER OF SCIENCE

in

THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES

(Pathology and Laboratory Medicine)

THE UNIVERSITY OF BRITISH COLUMBIA

(Vancouver)

April 2016

© Matthew Budd, 2016 Abstract

Background: Antiretroviral therapy has reduced mother-to-child HIV transmission from 25-

40% to less than 2%. Thus, increasing numbers of HIV-exposed uninfected (HEU) children are being born with perinatal exposure to antiretrovirals. Recently, a Canadian HIV clinic noticed a high prevalence of autism spectrum disorder (ASD) in HEU children. This prompted our analysis of HEU children enrolled in the pan-Canadian Children & Women AntiRetrovirals & Markers of

Aging (CARMA) cohort study. Significant differences in mitochondrial DNA to nuclear DNA ratio (mtDNA content) have been observed in ASD children and HEUs as a potential marker for mitochondrial dysfunction, which has been theorized as a possible mechanism underlying abnormal neurodevelopment. We hypothesized that HEU children with ASD would have significantly different leukocyte mtDNA content than HEU children without ASD and/or HUU children with and without ASD.

Methods: CARMA HEU children with ASD (n=14) were matched 1:3 on age, sex, and ethnicity to HEU children without ASD (n=42), HUU anonymous controls (n=42), and HUU children with ASD in the BC Autism Spectrum Interdisciplinary Research (ASPIRE) program (n=42).

Non-ASD HUU siblings of ASD children (n=9) were also studied and grouped with the HUU controls for the purposes of analyses (n=51 total). MtDNA content was assessed using qPCR.

Results: Among 299 HEU children in CARMA, 14 (4.7%) were diagnosed with ASD, substantially (>3-fold) above North American prevalence estimates (1.5%).

ii

HEU children with ASD had higher mtDNA content (median[interquartile range]: 163[150–

179]) than non-ASD HEUs (115[91–153], p=0.02), HUUs with ASD (110[99–132], p=0.0001), and HUU controls (100[73–121], p<0.0001). Non-autistic HEU children and ASD children with no HIV/cART exposure had higher mtDNA content than controls (p=0.004 and p=0.03, respectively), but did not significantly differ from each other (p=0.2).

Conclusions: Our results suggest a possible cumulative association between elevated leukocyte mtDNA content and both HEU and ASD status. This may implicate mitochondrial dysfunction as a contributor to the high ASD prevalence in our cohort. It is unclear if this effect is modulated by exposure to antiretrovirals or maternal HIV but it is consistent with studies suggesting increased mtDNA content as an adaptive mechanism to mitochondrial dysfunction.

iii

Preface

The contents of this thesis are my own original work. The experiments were designed by me in conjunction with my supervisor, Dr. Hélène Côté, and Dr. Jason Brophy of the Children’s

Hospital of Eastern Ontario. All conducted research was previously approved by the University of British Columbia Research Ethics Board (H08-02018). The collaborative use of clinical samples from the BC Autism Spectrum Interdisciplinary Program was approved August 5, 2015 by the University of British Columbia Research Ethics Board as an amendment to the original protocol (H08-02018-A024). Written parental consent was received for all enrolled participants, and informed assent was sought from those participants with the capacity to provide it. A copy of the consent form is attached in the appendix.

A version of this study is currently being prepared in manuscript format to be published before the completion of my graduate program.

A manuscript describing the monochrome, multiplex method of mtDNA quantitation is currently in preparation.

I conducted most of the measurements described in Chapters 3 and 4, organized the data and matching, and performed all statistical analyses.

iv

Table of Contents

Abstract ...... ii

Preface ...... iv

Table of Contents ...... v

List of Tables ...... x

List of Figures ...... xii

List of Abbreviations ...... xiv

Acknowledgements ...... xviii

Dedication ...... xx

Chapter 1: Introduction ...... 1

1.1 Introduction to thesis...... 1

1.2 Human immunodeficiency virus (HIV) ...... 3

1.2.1 Origin and transmission ...... 3

1.2.2 Epidemiology ...... 3

1.2.3 Morphology and genome structure ...... 4

1.2.4 Replication cycle ...... 7

1.2.5 Immunopathogenesis and stages of infection ...... 9

1.2.6 Combination antiretroviral therapy (cART) ...... 11

1.2.6.1 Classes of cART ...... 12

1.2.6.1.1 Nucleoside reverse transcriptase inhibitors ...... 12

1.2.6.1.2 Non-nucleoside reverse transcriptase inhibitors ...... 15

1.2.6.1.3 Protease inhibitors ...... 15

v

1.2.6.1.4 Integrase inhibitors ...... 16

1.2.6.1.5 Fusion and entry inhibitors ...... 17

1.2.7 Guidelines for HIV treatment and prevention ...... 18

1.2.7.1 In Canada ...... 18

1.2.7.2 In the developing world ...... 19

1.2.7.3 HIV and cART in pregnancy ...... 21

1.3 Mitochondria ...... 24

1.3.1 Structure and function ...... 24

1.3.1.1 The electron transport chain and oxidative phosphorylation ...... 25

1.3.2 Mitochondrial DNA (mtDNA) ...... 27

1.3.2.1 Replication ...... 28

1.3.2.2 ROS-induced oxidative damage ...... 29

1.3.2.3 Maintenance ...... 30

1.4 Autism spectrum disorder (ASD) ...... 30

1.4.1 Diagnostic criteria and instrumentation ...... 31

1.4.1.1 DSM-V ...... 31

1.4.1.2 Screening and assessment ...... 33

1.4.1.3 ADOS-G and ADI-R...... 34

1.4.2 Demographics ...... 35

1.4.3 Other factors associated with ASD ...... 35

1.5 ASD, HIV, cART, and mitochondrial dysfunction ...... 37

Chapter 2: Study design and participant demographics ...... 39

2.1 Recruitment ...... 39 vi

2.2 Eligibility and selection criteria ...... 41

2.3 Participant demographics ...... 43

2.4 Sample collection and preparation ...... 44

Chapter 3: MtDNA content measurements ...... 46

3.1 Measurement technique ...... 46

3.2 Quality control ...... 48

3.3 Statistical analyses of data ...... 50

3.3.1 Univariate between-group comparisons ...... 50

3.3.2 Multivariable linear regression ...... 51

3.3.3 Sensitivity analyses ...... 52

Chapter 4: MtDNA apparent oxidative damage measurements ...... 54

4.1 Measurement technique ...... 54

4.1.1 Sample preparation ...... 55

4.1.2 Long PCR of mtDNA ...... 55

4.1.3 Quantitative PCR and AOD calculations ...... 57

4.2 Quality control ...... 59

4.3 Statistical analyses of data ...... 61

Chapter 5: Results and analyses ...... 62

5.1 MtDNA content ...... 64

5.1.1 Quality control ...... 64

5.1.2 Reproducibility ...... 65

5.1.3 Statistical analyses ...... 66

5.1.3.1 Between-group comparisons ...... 67 vii

5.1.3.2 Multivariable linear regression ...... 73

5.2 MtDNA AOD...... 78

5.2.1 Quality control ...... 79

5.2.2 Statistical analyses ...... 82

5.3 Sensitivity analyses ...... 85

5.3.1 Ethnicity ...... 85

5.3.2 Albumin ...... 87

5.3.3 Platelet count ...... 90

Chapter 6: Discussion and conclusions ...... 92

6.1 Interpretation of results ...... 92

6.1.1 MtDNA content measurements...... 92

6.1.1.1 Quality control ...... 92

6.1.1.2 Primary outcomes and predictors...... 93

6.1.1.2.1 HEU status...... 93

6.1.1.2.2 ASD diagnosis ...... 96

6.1.1.2.3 Other variables of interest ...... 97

6.1.2 MtDNA AOD measurements ...... 99

6.1.2.1 Quality control ...... 99

6.1.2.2 Primary outcomes ...... 101

6.2 Limitations and biases...... 102

6.3 Future directions ...... 106

6.4 Conclusions ...... 107

References ...... 109 viii

Appendix: CARMA consent form ...... 127

ix

List of Tables

Table 1. Fetal-to-maternal antiretroviral drug concentrations in blood after in utero cART exposure, and the ability of cART drugs to cross the blood-brain barrier ...... 23

Table 2. Diagnostic criteria for autism spectrum disorder, as outlined in the Diagnostic and

Statistical Manual of Mental Disorders, Fifth Edition ...... 32

Table 3. CARMA recruitment sites, follow-up parameters, and HEU-relevant demographics for participants in our study ...... 40

Table 4. Forward and reverse primer sequences used to measure mtDNA content via monochrome, multiplex qPCR...... 46

Table 5. Thermal cycler settings for monochrome, multiplex qPCR of nuclear (albumin) and mitochondrial (D-loop) sequences ...... 47

Table 6. Forward and reverse primer sequences used for long PCR of mtDNA ...... 56

Table 7. Thermal cycler settings for long PCR of mtDNA ...... 57

Table 8. Forward and reverse primer sequences used for qPCR of the mtDNA D-loop ...... 58

Table 9. Thermal cycler settings for mtDNA D-loop quantification using qPCR...... 58

Table 10. Characteristics of the 5 ICs used in long PCR and qPCR for the mtDNA AOD assay 59

Table 11. Demographic characteristics of participants in each experimental group ...... 63

Table 12. Variability in mtDNA content measurements of the two ICs over six runs, and within- run measurements of 12 IC High replicates ...... 65

Table 13. Multivariable logistic regression analysis of potential explanatory variables for ASD in

106 study participants ...... 67

x

Table 14. MtDNA content, maternal and paternal age, and prevalence of ASD-relevant factors in

HEU and HUU children with and without ASD ...... 70

Table 15. Maternal cART exposure and mtDNA content for HEUs with ASD (n=14) and non-

ASD HEU matches (n=42) ...... 72

Table 16. Multivariable linear regression analysis between mtDNA content and potential explanatory variables for 106 study participants ...... 74

Table 17. Multivariable linear regression analysis between mtDNA content and potential predictor variables for participants with ASD (n=52) ...... 75

Table 18. Multivariable linear regression analysis between mtDNA content and potential predictor variables for HEU participants with known perinatal cART exposure (n=49) ...... 77

Table 19. Quality control parameters for the three Happy IC candidates used across 12 D-loop runs, corresponding to six long PCR runs ...... 81

Table 20. Rate of amplification of three ICs used in seven long PCR runs ...... 81

Table 21. Absolute amplification, amplification relative to an undamaged IC, and AOD of mtDNA for HEU and HUU children with and without ASD ...... 82

Table 22. Distributions of age, sex, and mtDNA content for three matched groups used in a sensitivity analysis of possible relationship between mtDNA content and ethnicity ...... 85

xi

List of Figures

Figure 1. The HIV genome ...... 5

Figure 2. Structure of a mature HIV virion ...... 6

Figure 3. Life cycle of HIV...... 8

Figure 4. Chemical compound structures of the four DNA nucleosides and their respective NRTI analogues...... 14

Figure 5. Ultrastructure of a mitochondrion ...... 25

Figure 6. The mitochondrial electron transport chain, oxidative phosphorylation, and production of ROS ...... 26

Figure 7. The circular mtDNA genome ...... 28

Figure 8. Selection process, exclusions, and design of our cross-sectional, case control study ... 43

Figure 9. MtDNA AOD assay and calculations...... 54

Figure 10. Correlation between the original and repeat measurements of 28 semi-randomly selected samples. R2 and Pearson’s r are shown ...... 66

Figure 11. Univariate between-group comparisons between mtDNA content of HEU and HUU children with and without ASD. P-values from Student’s t-test or Mann-Whitney U test shown, as appropriate ...... 69

Figure 12. Correlation between mtDNA content and duration of in utero maternal cART exposure in HEU children (n=52) ...... 78

Figure 13. Agarose gel electrophoresis (0.8%) of n=16 samples with quantitatively high (groups

A and C) and quantitatively low (groups B and D) mtDNA AOD ...... 80

xii

Figure 14. Univariate between-group comparisons between mtDNA AOD of HEU and HUU children with and without ASD. P-values from Student’s t-test are shown ...... 83

Figure 15. Correlations between mtDNA content and relative mtDNA AOD in HEU children with ASD (A), HEU children without ASD (B), HUU children with ASD (C), HUU children without ASD (D), and all participants (E). R2 and either Spearman’s ρ or Pearson’s r are shown, as appropriate ...... 84

Figure 16. Univariate between-group comparisons between three sex- and age-matched (1:2) groups of HEU children without ASD. P-values calculated via Student’s t-test ...... 86

Figure 17. Correlation between albumin copy number and mtDNA content for all samples in the study. R2 and Spearman’s ρ are shown ...... 87

Figure 18. MtDNA content measurements after diluting all samples to approximately 5000 copies of albumin ...... 88

Figure 19. Correlation between mtDNA content measurements before and after dilution to a common concentration. R2 and Pearson’s r are shown ...... 89

Figure 20. Correlation between leukocyte mtDNA content and platelet count in HEU children with ASD (A), HEU children without ASD (B), and all HEU children in the study (C). R2 and either Spearman’s ρ or Pearson’s r are shown, as appropriate ...... 91

xiii

List of Abbreviations

3TC Lamivudine

ABC Abacavir

ADI-R The Autism Diagnostic Interview – Revised

ADOS-G The Autism Diagnostic Observation Schedule – Generic

AIDS Acquired immune deficiency syndrome

AOD Apparent oxidative damage

ASD Autism spectrum disorder

ATP Adenosine triphosphate

ATV Atazanavir

AZT Zidovudine cART Combination antiretroviral therapy

CCR5 C-C chemokine receptor type 5

CD4 Cluster of differentiation 4

CD8 Cluster of differentiation 8 cDNA Complementary DNA

CI Confidence interval

CTL Cytotoxic T-lymphocyte

CV Coefficient of variation

CXCR4 C-X-C chemokine receptor type 4 d4T Stavudine

xiv

ddC Zalcitabine ddI Didanosine

DLV Delavirdine

DNA Deoxyribonucleic acid

DRV Darunavir

DSM-V Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition

DTG Dolutegravir

EDTA Ethylenediaminetetraacetic acid

EFV Efavirenz

ENF Enfuvirtide

ETC Electron transport chain

ETR Etravirine

EVG Elvitegravir

FPV Fosamprenavir

FTC Emtricitabine

HEU HIV-exposed uninfected

HIV Human immunodeficiency virus (Type 1)

HUU HIV-unexposed uninfected

IC Internal control

ID Intellectual disabilities

IDV Indinavir

IQR Interquartile range

xv

kb Kilobase

LPV/r Ritonavir-boosted lopinavir

LTR Long terminal repeat mRNA Messenger RNA mtDNA Mitochondrial DNA

MVC Maraviroc nDNA Nuclear DNA

NFV Nelfinavir

NNRTI Non-nucleoside reverse transcriptase inhibitor

NRTI Nucleoside reverse transcriptase inhibitor

NVP Nevirapine

PCR Polymerase chain reaction

PI Protease inhibitor pVL Plasma viral load qPCR Quantitative polymerase chain reaction

RAL Raltegravir

RNA Ribonucleic acid

ROS Reactive oxygen species

RPV Rilpivirine

RT Reverse transcriptase

RTV Ritonavir

SD Standard deviation

xvi

SQV Saquinavir

TDF Tenofovir disoproxil fumarate

TPV Tipranavir

WHO World Health Organization

xvii

Acknowledgements

I would first like to thank my supervisor, Dr. Hélène Côté, for all of her guidance, feedback, and understanding throughout this project and others. The knowledge I have gained through her mentorship has been instrumental in my development as a student, professional, and scientist.

My gratitude also goes out to my supervisory committee, Drs. Haydn Pritchard, Andre Mattman,

Chris Shaw, Ariane Alimenti, and Anthony Bailey for their hard work, expert advice, and commitment to this project.

I would also like to acknowledge the many members of the Côté lab for their friendship, feedback, and aid during my time here. I could not have asked for a better group of people with which to share a workspace. To my peers, Anthony, Abhinav, Sara, and Marta: your assistance and input regarding this project were invaluable, and I wish each of you the very best in your future endeavors. To Izabelle and Beheroze: many thanks for the countless hours you have spent training, teaching, and troubleshooting with me, and for all of the tireless work you put in on my behalf when timelines were bearing down. To our lab’s undergraduate researchers, Adam,

Omair, Mayanne, and Shamim: it has been a pleasure to know each you, and may the knowledge you gain here serve you well going forward.

Finally, I would like to give my sincerest and most heartfelt thanks to my friends and family for their continued support, encouragement, and words of wisdom. I am incredibly fortunate to have made many great friends within the department, and I hope for those friendships to be maintained for years to come. Words cannot express my gratitude toward my parents, without xviii

whom none of this would have been possible. The unconditional love and support they have shown me throughout the years has shaped me into the person I am today. Their kindness, generosity, and unwavering willingness to sacrifice for their family have been nothing short of inspirational.

xix

Dedication

To my parents, Gordon and Cheryl

xx

Chapter 1: Introduction

1.1 Introduction to thesis

The use of combination antiretroviral therapy (cART) has been successful in reducing perinatal mother-to-child transmission rates of human immunodeficiency virus (HIV) from 25-40% to less than 2% in the developed world [1,2]. This has resulted in an increasing number of HIV- exposed, uninfected (HEU) children born each year with months of perinatal exposure to cART.

Although the benefits of reducing rates of vertical HIV transmission are evident, there have recently been clinical concerns raised with regards to long-term outcomes of perinatal cART exposure in HEU infants [2–4]. The nucleoside reverse transcriptase inhibitor class of cART drugs have previously been associated with mitochondrial toxicity in adults [4], and some are able to cross the placental barrier [5], prompting a growing body of research exploring perinatal cART and HIV exposure as potential risk factors for mitochondrial disease and dysfunction [6–

8]. In addition to increased risk of developing various metabolic abnormalities and immune irregularities [9], one potential outcome of mitochondrial dysfunction in young children is an increased risk of developing autism spectrum disorder (ASD) [10].

As of 2010, the global prevalence of ASD in children above diagnostic age has been estimated at

1.47% [11]. However, much higher rates of ASD (9/158, 5.7%) were recently observed in a study of HEU children being followed by a Canadian paediatric HIV clinic [12]. Many of these children are enrolled in the Children & Women AntiRetrovirals & Markers of Aging (CARMA) cohort study. Of the HEU children enrolled in CARMA, 14/299 (4.7%) received a diagnosis of

ASD, and 2 additional cases are strongly suspected and currently undergoing formal assessment.

Of the 14 children diagnosed with ASD, 11 (79%) received cART exposure in utero, and 13/14 1

(93%) were given between 3 and 7 weeks of postnatal prophylaxis with zidovudine, in accordance with Canadian guidelines. In contrast, only 2/144 (1.39%) among the HIV+ children enrolled in CARMA received an ASD diagnosis, consistent with general population estimates.

The objectives of my MSc research are to:

1) Explore and compare potential quantitative and qualitative differences in mitochondrial

DNA (mtDNA) among HEU and HIV-unexposed uninfected (HUU) children with and

without a diagnosis of ASD

2) Explore potential associations between ASD and previously described ASD-relevant

clinical and demographic risk factors, where this information is available

The hypotheses of this research are:

Primary hypothesis:

Compared to age-, sex-, and ethnicity-matched HEU children without ASD, HIV-unexposed uninfected (HUU) controls, or ASD children with no HIV/cART exposure, HEU children diagnosed with ASD will differ in terms of leukocyte mtDNA

a) content

b) apparent oxidative damage

2

1.2 Human immunodeficiency virus (HIV)

1.2.1 Origin and transmission

HIV is classified into two major strains: HIV-1 and HIV-2. The latter is thought to have originated from a non-pathogenic strain of simian immunodeficiency virus observed in Old

World monkeys, and in humans its spread is primarily contained to countries in West Africa

[13]. Due to its geographical restriction, reduced pathogenicity, and lower plasma viral load

(pVL) levels in infected patients compared to HIV-1, the prevalence of HIV-2 appears to be declining in high-risk countries [14–16]. Therefore, any further references to HIV within this thesis shall refer to HIV-1, currently a major global health concern and a growing epidemic dating back to the 1980’s.

HIV is transmitted through three modes of primary contact: sexual, through contact of infected seminal, rectal, or vaginal fluid with mucosal tissue; percutaneous inoculation, as through needle sharing; and vertically, from mother to child in utero, during childbirth, or through breastfeeding

[17]. The risk of transmission can vary greatly and depends on a number of environmental and genetic factors [18], although due to repeated exposures there are several high-risk groups including men who have sex with men, injection drug users, prison inmates, transgender individuals, and sex workers [17,19].

1.2.2 Epidemiology

Globally, as of 2014 there were an estimated 37 million people living with HIV [20], of which

17.4 million (47%) were women of childbearing age, and 2.6 million (7%) were children aged 15 or younger [21]. Many people living with HIV and AIDS, both children and adults, live in sub- 3

Saharan Africa (25.8 million, 70% of global cases) or Southeast Asia (5.0 million, 14% of global cases) [21].

According to figures presented by the Public Health Agency of Canada in 2014 [22], an estimated 71,300 people in Canada were living with HIV and AIDS by the end of 2011, up from

64,000 in 2008 (an 11.4% increase). Approximately 16,600 (23.3%) of these reported cases were women, consistent with 2008 estimates in which women accounted for 14,740 (23.0%) of the

64,000 reported cases [22]. The incidence of new HIV cases within Canada has remained relatively stable in recent years, with approximately 1000 to 1100 new cases reported annually between 2002 and 2011 [22]. The increasing prevalence of HIV in the developed world can largely be attributed to reduced rates of mortality among those infected, a result of advances in available care and treatment options as well as increased access to treatment [21–23]. Currently there are an estimated 300 children receiving care for HIV infection in Canada, the vast majority of whom inherited the virus vertically [1]. Globally, almost 3.3 million children are born each year to HIV+ mothers [24], of which 200-300 are born in Canada [25].

1.2.3 Morphology and genome structure

HIV is a lentivirus containing two copies of a positive-sense RNA genome. The HIV genome is approximately 9.8 kilobases (kb) in length, including a pair of flanking regions of long terminal repeat (LTR) sequences (Figure 1). These LTR regions facilitate the viral genome’s integration into its host’s chromosomal DNA [26], as well as serving as sites of transcriptional regulation

[27–29]. The genome itself contains three regions (gag, pol, and env) encoding for structural

4

proteins, two sites (tat and rev) encoding for regulatory factors, and four accessory proteins which aid replication (vif, vpr, vpu, and nef).

Figure 1. The HIV genome

During viral maturation, the protein encoded by gag is cleaved by viral proteases into four subunits [30]. These subunits form the matrix, capsid, and nucleocapsid proteins, as well a membrane-tethered p6 protein; the function of this p6 unit is not well-understood but appears to be integral in the release of virions from the host cell [30,31]. Pol codes for the viral integrase, reverse transcriptase (RT), and protease enzymes necessary for integration, replication and maturation [32]. The env region codes for two viral envelope glycoproteins, gp120 and gp41, which mediate the fusion and entry of the viral particle into the host cell [33]. Figure 2 illustrates the structure of a mature HIV virion.

5

Figure 2. Structure of a mature HIV virion

The viral envelope is composed of a lipid bilayer derived from the host cell, and embedded within the envelope are numerous transmembrane gp41 subunits non-covalently bonded to surface gp120 [33]. Within the envelope, a p17 protein matrix is formed which preserves the structural integrity of the virion, and upon binding to host cell receptors stimulates the production of proinflammatory cytokines favourable to viral replication [34]. Viral integrase, protease, and RT are contained within a conical capsid, along with viral RNA sheathed by nucleocapsids which protect against the activity of nucleases [35].

6

1.2.4 Replication cycle

HIV primarily infects macrophages and T-lymphocytes expressing the glycoprotein CD4 on their surface (CD4+), although some reports have also documented lesser, likely inefficient levels of viral replication in various other classes of cells such as astrocytes, dendritic cells, endothelial cells, monocytes, CD8+ T-lymphocytes, and hematopoietic stem cells [36–42]. The first step in the replication cycle of HIV is binding of the virion’s gp120 surface subunit to host cell CD4

[28,43]. This binding induces a conformational change in gp120 which allows the virion to bind to a cell surface co-receptor, typically either of the chemokine receptors CCR5 or CXCR4 [28].

This in turn induces a second conformational change which exposes the N-terminal fusion peptide of gp41, allowing for fusion of the viral envelope with host cell membrane [28,44].

Subsequently, viral RNA and enzymes are deposited into the host cell and RT directs the synthesis of double-stranded viral cDNA, which is actively transported within a preintegration complex into the nucleus [44]. HIV integrase then catalyses the insertion of the provirus into chromosomal DNA by recognizing specific sequences of the LTR and ligating these sequences to nonspecific cleavage sites within the host genome [28,44]. Once inserted, transcription of the provirus is dependent on host cell mechanisms and transcription factors, although expression is mediated by promoter signals within the LTR [29]. Viral polyproteins and env gp120 and gp41 are synthesized using host ribosomes, and are assembled in the cytoplasm along with transcribed viral messenger RNA (mRNA) for release in a process directed by gag precursors [44].

Throughout this process and shortly after virion release, polyproteins are cleaved into their constitutive protein forms by both viral and cellular proteases, forming a mature and infectious virion once structural proteins have sufficiently condensed [44]. A visual summary of this process can be seen in Figure 3. 7

Figure 3. Life cycle of HIV 8

1.2.5 Immunopathogenesis and stages of infection

HIV infection is broadly categorized into three clinical stages: the initial or acute phase, the chronic or latent phase, and transition to AIDS. For approximately 7-12 days post-exposure, HIV

RNA and antigens are undetectable in plasma, a period referred to as the eclipse phase [45,46]. It is during the eclipse phase that HIV first begins to replicate within tissues localized at the site of primary exposure, and thus the risk of further transmission during this period is relatively low

[17,47,48]. However, due to the lack of viremia biomarkers and clinical symptoms during this time there is a risk that those infected may not be diagnosed until a later stage, when the probability of transmission is much greater [49]. After infecting localized cells, HIV is shuttled via dendritic cells to the body’s draining lymph nodes, where it is able to spread and infect other tissues throughout the body [45].

Following the eclipse phase is the acute infection stage, characterized in general by rapid proliferation of HIV and a sudden, dramatic decrease in CD4+ T-cell count [50]. The most substantial loss of CD4+ cells occurs in the gut-associated lymphoid tissue, which houses the majority of the body’s activated CCR5-expressing T-lymphocytes [51]. The accompanying activation of CD8+ cytotoxic T-lymphocytes (CTLs) suggests that HIV-induced cell lysis is mediated at least in part by the adaptive immune response [52]. However, HIV-infected cells also display increased sensitivity to ligand-mediated apoptosis, potentially implicating this pathway as a secondary mechanism of depletion [53].

9

One to two weeks after acquisition, circulating levels of HIV RNA increase exponentially and breach the threshold of detection (40-100 copies/mL of plasma) [49]. This is followed by the appearance of other HIV biomarkers such as viral p24 antigen and antibody, detectable by current fourth-generation assays between two and three weeks post-infection [48]. Of note is that peak levels of viremia are reached during the acute phase, with as many as 22 million copies of viral RNA per millilitre of plasma having previously been reported [54]. This peak typically occurs just prior to seroconversion, the point at which HIV antibodies become detectable in the blood. This is followed by a gradual decrease in pVL as infected cells are suppressed by CD8+

CTLs [55,56]. Throughout the acute phase, HIV-infected patients may present with mild to severe clinical symptoms mimicking influenza, known as acute retroviral syndrome [57].

Common symptoms include fever, rash, swollen lymph nodes, sore throat, and muscle pain, usually manifesting two to six weeks after primary infection [57].

Following the acute infection stage is a period of clinical latency known as the chronic phase.

The chronic phase is associated with persistent immune cell activation, elevated levels of circulating proinflammatory cytokines, and translocation of microbial products likely originating from extensive damage to gut-associated lymphoid tissue [58]. During the chronic phase, circulating HIV viral load also experiences a decline from its peak value, establishing a homeostatic viral load set point. Factors that appear to influence the set point include viral load at acquisition, the specificity and effectiveness of the CD8+ CTL response to HIV infection during the acute phase, and the presence of polymorphisms in the HIV genome [45,59–61].

Sequencing assays as well as expansions in the virus’s tropism during the chronic phase suggest

10

that most HIV infections stem from a single progenitor genotype and mutate in vivo, significantly complicating treatment and vaccination efforts [45,62].

The third and final stage of HIV infection is transition to AIDS. If left untreated, chronic HIV infection typically progresses to AIDS within an average of 8 to 10 years after primary infection

[63]. A formal diagnosis of AIDS can be made based on a CD4+ T-cell count of <200 cells/mm3 in blood, or the presence of one or more severe opportunistic illnesses resulting from immune deficiency [64,65]. A full list of conditions which must accompany established HIV infection to diagnose AIDS in children and adults is provided by the World Health Organization (WHO)

[65]. Prognosis and survival time after an AIDS diagnosis is highly variable and dependent upon a number of factors including the severity and number of acquired opportunistic illnesses, but estimates range from between one and two years median survival time in the absence of treatment [66–68].

1.2.6 Combination antiretroviral therapy (cART)

Although antiretroviral monotherapy has been available since 1987, the mid-nineties marked a substantial shift in the HIV treatment landscape. Drugs of different classes were made available, which when combined could inhibit multiple stages of the HIV replication cycle [69]. The administration of antiretroviral triple therapy spanning more than one class of drugs became known as combination antiretroviral therapy (cART), and featured much greater levels of tolerability, viral suppression, and safety than available monotherapies [69,70].

11

1.2.6.1 Classes of cART

There are six classes of cART drugs licensed by the United States Food and Drug Administration

(FDA) for clinical use. The nucleoside reverse transcriptase inhibitor (NRTI) class was the first to receive FDA approval in 1987 [69]. This was followed by protease inhibitors (PIs) in 1995 and non-nucleoside reverse transcriptase inhibitors (NNRTIs) in 1996 [70]. There are also three newer classes of drugs available: fusion inhibitors, licensed in 2003, as well as entry inhibitors and integrase inhibitors, both approved in 2007 [70]. Therapy in the developed world is usually administered as a combination of two NRTIs and one drug from the remaining classes, typically a PI [69]. Before initiating any form of cART treatment, physicians are recommended to administer drug resistance testing through genotypic and phenotypic assays in order to optimize treatment outcomes [71].

1.2.6.1.1 Nucleoside reverse transcriptase inhibitors

NRTIs are a class of drugs with chemical structures mimicking those of naturally occurring nucleosides. As they lack a 3’-hydroxyl group on the deoxyribose moiety, they are unable to form 5’-3’ phosphodiester bonds. Thus, they act as competitive inhibitors of HIV RT by inducing chain termination of proviral DNA during polymerization. There are seven NRTIs currently approved by the FDA for treatment of HIV. Abacavir (ABC), emtricitabine (FTC), lamivudine (3TC), tenofovir disoproxil fumarate (TDF), and zidovudine (AZT) represent the most widely prescribed NRTIs, and have comparatively mild side effects [72]. Two NRTIs developed at a relatively early stage in the history of HIV treatment, didanosine (ddI) and stavudine (d4T), are still approved for use but are no longer recommended, especially in combination treatments, due to their harsher side effects and toxicity [72]. One NRTI, zalcitabine 12

(ddC), was removed from the market in 2006 due to its toxicity. The structures of these drugs and the naturally occurring nucleosides that they mimic are presented in Figure 4.

13

Figure 4. Chemical compound structures of the four DNA nucleosides and their respective

NRTI analogues 14

1.2.6.1.2 Non-nucleoside reverse transcriptase inhibitors

There are currently five FDA approved NNRTIs in clinical use: nevirapine (NVP), efavirenz

(EFV), delavirdine (DLV), etravirine (ETR), and rilpivirine (RPV). In contrast with NRTIs,

NNRTIs inhibit viral RT via non-competitive inhibition. These drugs bind to a hydrophobic pocket adjacent to the RT enzyme’s active site and induce conformational changes which compromise its catalytic function [73]. After binding, RT is unable to coordinate and position divalent metal ions necessary for phosphodiester bond formation, halting DNA polymerization through allosteric regulation [73]. NNRTIs have much longer plasma half-lives compared to PIs, and thus require administration only once daily as opposed to twice or more [74]. This yields

NNRTI-based regimens more appealing in resource-limited environments where affordability and access to treatment are limited, as well as to those experiencing difficulty in maintaining adherence. However, these drugs are particularly susceptible to cross-resistance, as despite their unique chemical structures they all bind to the same hydrophobic pocket on RT. Thus, a single point mutation in pol can confer universal resistance to NNRTI treatment, whereas mutations conferring resistance to NRTIs and PIs are more complex and can be obviated by altering drug combinations [75]. There is also evidence that resistance-conferring mutations for NNRTIs may be more common among newly-infected patients naïve to cART, by as much as three times

(7.5% contra 2.5% for PI resistance) [75].

1.2.6.1.3 Protease inhibitors

A critical step in the replication and maturation of HIV is the proteolytic cleavage of gag and pol polypeptide precursors into their active protein forms. PIs are compounds reminiscent in structure of peptides, but feature hydroxyethylene bonds in place of the hydrolysable peptide 15

bonds of the protease’s natural substrates [76]. These compounds have a high affinity for the active site of aspartyl proteases, and once bound will block any further proteolytic activity via competitive inhibition [76]. Protease mutations conferring PI resistance are referred to as either primary or secondary, depending on the step in resistance development in which the mutation is conferred [77]. Primary mutations are a result of naturally-occurring mutations in the viral genome which result from poor proofreading of viral RT, and typically confer weak levels of resistance in isolation [77]. However, secondary mutations arise as a result of selective but incomplete pressure on PI treatment, and confer much higher-level resistance when present in combination with a primary mutation [77]. Like NNRTIs, the structures of different PIs are highly conserved and vulnerable to cross-resistance, so treatment is usually administered by

‘boosting’ plasma PI levels with low doses of ritonavir, which inhibits PI metabolism [77]. As a result, multiple mutations in HIV protease are required for the virus to escape treatment when administered in this fashion [77]. Eight PIs are licensed by the FDA for use: atazanavir (ATV), darunavir (DRV), fosamprenavir (FPV), indinavir (IDV), nelfinavir (NFV), ritonavir (RTV), saquinavir (SQV), and tipranavir (TPV). In addition, a combination drug of ritonavir-boosted lopinavir (LPV/r) is available.

1.2.6.1.4 Integrase inhibitors

Once deposited into the host’s cytoplasm, viral integrase primes cDNA for integration via endonucleolytic cleavage of several base pairs at the 3’ end, in a process known as 3’-processing

[78]. This integrase-cDNA complex is then shuttled into the nucleus along with other viral and cellular proteins, whereupon integrase catalyses the insertion of its cDNA substrate into host chromosomes, known as strand transfer [78]. The three integrase inhibitors currently being 16

marketed, dolutegravir (DTG); elvitegravir (EVG); and raltegravir (RAL), all inhibit integrase at the 3’-processing stage [78,79]. These drugs bind to Mg2+ ions at the enzyme’s active site and act as competitive inhibitors against viral cDNA [78]. However, other drugs are currently being developed and patented which inhibit integrase at the strand transfer step in a process that appears to be mediated by integrase binding to lens epithelial-derived growth factor [79].

Currently, integrase inhibitors are used primarily to treat patients who have developed resistance to other classes of drugs [78]. Integrase inhibitors are also highly effective at rapidly reducing pVL, and thus are an attractive option for treating pregnant women who present with HIV late in pregnancy [80,81].

1.2.6.1.5 Fusion and entry inhibitors

Only a single fusion inhibitor, enfuvirtide (ENF), and a single entry inhibitor, maraviroc (MVC), are licensed for distribution at the current time, although they share similar mechanisms.

Enfuvirtide mimics a short region of the viral gp41 transmembrane protein and selectively binds it following gp120-CD4 complex formation [82]. In doing so, the drug inhibits further conformational changes in gp41 which are necessary for viral envelope fusion with the host membrane. Maraviroc is a small molecule which blocks fusion of gp120 with the host membrane by serving as a functional antagonist to CCR5 [82,83]. By inhibiting the binding of the receptor to its naturally occurring chemokine ligands, the conformation of CCR5 can no longer facilitate binding to gp120, halting entry of the virus into its host [82,83]. However, MVC is used infrequently to treat HIV due to the fact that its efficacy is limited to CCR5-tropic strains, and will not treat CXCR4-tropic variants. ENF is also used infrequently as the daily injections are known to cause local reactions, and are less convenient than orally-administered drugs. 17

1.2.7 Guidelines for HIV treatment and prevention

1.2.7.1 In Canada

Management guidelines for antiretroviral administration and HIV treatment in pregnancy are kept by the Society of Obstetricians and Gynaecologists of Canada [84], supplemented by guidelines from the U.S. Department of Health and Human Services [85].

Voluntary HIV screening is recommended for all pregnant women in the U.S. and Canada, as early on in pregnancy as possible. Many provinces in Canada perform these tests on an opt-out basis. Fourth generation combined immunoassay testing for p24 antibody and antigen is recommended for diagnosing and screening for HIV infection in pregnancy, as the assay is highly sensitive and specific for HIV, is able to detect infection in the early acute phase (see section 1.2.5), and results can be made available within 24 hours or sooner [85]. Women at high risk for contracting HIV (section 1.2.1) are recommended additional screening during the second and third trimesters [84].

For any woman testing positive for HIV during pregnancy screening, or who is known to be living with HIV and becomes pregnant or plans to become pregnant, treatment with cART is recommended regardless of their CD4 count and pVL [84]. Canadian guidelines recommend viral genotyping as a screen for drug-resistant HIV phenotypes, allowing for the optimization of drug combinations [84]. Typically a drug combination of two NRTIs and one PI or NNRTI is administered during pregnancy, with drugs known to cross the placenta used preferentially [84].

18

Intravenous AZT is administered during the onset of labour until delivery, in tandem with orally administered maternal cART [84].

For children born to HIV+ mothers, diagnostic screening for infection using HIV RNA nucleic acid testing is recommended at birth, 2-4 weeks after birth, 1-2 months after birth, and finally at

4-6 months after birth, although this can differ by region (in British Columbia infants are tested at birth, four weeks, and three months of age) [84,85]. Prophylactic AZT treatment for the first six weeks of life is recommended for all infants born to mothers with lower pVL (<1000 copeis/mL), Recommendations for infants born to mothers who either had a higher pVL or took no cART during pregnancy include a three-drug AZT-containing cART combination for the first six weeks of life, in addition to three doses of NVP in the first week of life and two weeks of

3TC treatments [84].

1.2.7.2 In the developing world

Recommendations to physicians working in resource-limited settings for HIV treatment in pregnancy have been maintained by the WHO since 2000 [86,87]. Like the Canadian guidelines, recommendations in resource-limited settings have undergone several changes over time to accommodate evolutions in preferred regimens, timing of cART initiation, and prophylaxis.

The WHO’s initial recommendations, aimed at women with known HIV+ status, included maternal treatment with AZT and/or 3TC starting between 14 and 36 weeks of pregnancy, followed by 1-6 weeks of postpartum AZT prophylaxis for the infant [87]. The 2004 and 2006 guideline updates recommended all pregnant women be screened for HIV early on in pregnancy, 19

with treatment options stratified according to stage of infection and CD4 count [86,88]. Those mothers requiring treatment for their own health were recommended a combination regimen to start as soon as possible and be maintained throughout life (typically AZT or d4T+3TC+NVP), while those without indications for initiating treatment were recommended AZT prophylaxis treatment starting at week 28 of pregnancy and a single dose of NVP during labour [86,88].

Irrespective of the mother’s treatment needs, infants were given one week of AZT prophylaxis

(or up to 6 weeks if the mother initiated treatment late in pregnancy) and/or a single dose of NVP at birth.

A major revision to these guidelines was published in 2010, expanding and streamlining recommendations for treatment eligibility, initiation time, suggested regimens, and prophylaxis options [89]. First-line NRTI treatment for pregnant women was expanded from AZT+3TC to include TDF+FTC-based regimens, and were recommended to any pregnant woman with a CD4 count <350 cells/mm3, irrespective of clinical staging [89]. Prophylaxis was recommended starting at 14 weeks of pregnancy rather than 28, and options were stratified into two different classes; Option A, based on maternal antepartum AZT prophylaxis with extended infant NVP prophylaxis (administered until one week after cessation of breastfeeding), and Option B based on maternal antepartum triple drug therapy and 4-6 weeks of infant prophylaxis with either NVP or AZT [89]. A third prophylaxis plan, Option B+, was added in 2012, recommending triple drug therapy for the mother starting at diagnosis and continuing throughout life, with 4-6 weeks infant

AZT/NVP prophylaxis [90].

20

The 2013 update to these guidelines offered a more standardized approach, eliminating Options

A, B, and B+ while recommending in their place a single, simplified regiment for all HIV+ pregnant women of any CD4 count [91]. A first-line regimen of TDF+FTC (or 3TC) +EFV was recommended, to begin as soon as HIV was diagnosed and lasting at least until after the cessation or breastfeeding; those meeting eligibility criteria could be kept on this triple drug regimen for life [91]. Recommended infant treatment was also standardized to six weeks of daily

NVP prophylaxis, or twice-daily AZT [91]. At the time of writing, the most recent update to these guidelines came in September of 2015, recommending all HIV+ pregnant women be maintained on cART for life [19].

1.2.7.3 HIV and cART in pregnancy

Although the risk of vertical HIV transmission appears to be greatest during childbirth when the newborn is directly exposed to infected maternal body fluids [92,93], it can also occur both in utero [94] and postpartum through breastfeeding [95]. The mechanism through which HIV is able to cross the placenta has not been fully elucidated, but there is some evidence of placental trophoblasts, Hofbauer cells, and dendritic cells supporting HIV infection and adsorption, presenting a potential pathway linking infected maternal blood to the fetus during pregnancy

[96–98]. While most in utero infections occur late in pregnancy, viral antigens and RT activity have been observed in various fetal tissues, including the developing brain, from as early as 15 weeks [94]. The barriers protecting newborn and fetal brains from drugs, toxins, and pathogens are not as well-developed as those of adults [99], and various animal models have shown a causal relationship between prenatal immune activation and adverse neurodevelopmental

21

outcomes [99]. Such findings have led to questions regarding the potential neurodevelopmental consequences of HIV and cART exposure during pregnancy.

Despite gaps in knowledge regarding the potential adverse effects of cART administration during pregnancy, it is well-established that the prevention of mother-to-child HIV transmission far outweighs any potential risks. However, evidence of adverse mitochondrial function and mitochondrial toxicity in placental tissue, maternal peripheral blood, and cord blood has been observed in HIV-infected pregnant women receiving cART treatment [5,101]. Some studies also suggest differences in mtDNA content and mitochondrial gene expression in infants with perinatal ART exposure compared to HIV-unexposed infants [6,102]. Table 1 displays a summary of commonly used antiretrovirals along with their respective maternal to cord plasma drug concentration ratios (a marker of transplacental permeation) and cerebrospinal fluid to plasma drug concentration ratios (a marker of the drug’s ability to cross the adult blood-brain barrier). Note that for the latter, averages and measurements of variation were heterogeneously reported as means, medians, 95% confidence intervals (CI), interquartile ranges (IQR), standard deviations (SD), and ranges [103], so values were binned and reported as high, medium, or low.

As there is very little information regarding the ability of these drugs to cross the fetal blood- brain barrier during pregnancy, it is plausible that some or all of these drugs could more effectively permeate the developing central nervous systems compared to the adult data presented in Table 1.

22

Table 1. Fetal-to-maternal antiretroviral drug concentrations in blood after in utero cART exposure, and the ability of cART drugs to cross the blood-brain barrier

Range of average C:M drug CSF:P drug concentration Drug Class concentration ratiosa ratiob 3TC 0.93 – 1.22 Low ABC 1.03 – 1.06 Moderate AZT 0.81 – 1.6 High NRTIs d4T 1.0 – 1.3 Moderate ddI 0.38 NDA FTC 1.2 – 1.7 NDA TDF 0.82 – 1.1 Low EFV 0.37 – 0.74 High

NNRTIsc ETV 0.33 – 0.51 High NVP 0.59 – 1.0 High RPV 0.74 NDA ATV 0.13 – 0.24 Low DRV 0.15 – 0.32 Low FPV 0.27 NDA IDV 0 – 0.12 Low PIs LPV/r 0.16 – 0.22 Low NFV 0 – 0.49 Not stated (low detection) SQV 0 – 0.04 Low TPV 0.41 NDA RTV 0 – 0.55 NDA

Fusion, entry, & ENF Undetectable NDA integrase MVC 0.37 NDA inhibitors RAL 1.00 Low a Averages usually reported as medians, but occasionally as means. Values retrieved from Table 1, McCormack and Best [104] b Data retrieved from Tables 1, 3, 5, and 7 of Decloedt et al [103]. “Low” refers to an average ≤ 0.33, “Moderate” to an average > 0.33 but ≤ 0.67, and “High” to an average > 0.67 c Blood-brain barrier penetration was high when the drug was unbound from protein, low when bound C:M, cord plasma to maternal plasma; CSF:P, cerebrospinal fluid to plasma; NDA, no data available 23

1.3 Mitochondria

Mitochondria are found in nearly all eukaryotic cells, including simplified germ cells such as spermatozoa [105]. A critical exception is the mammalian red blood cell, which when fully differentiated contains neither a nucleus nor any mitochondria [106]. The number of mitochondria in other eukaryotic cells can vary considerably, from a single mitochondrion to several million depending on the organism, cell type, and physiological conditions [107]. The number of mitochondria is also fluid within any given cell, as mitochondria are highly dynamic and constantly undergo fusion and fission [107].

1.3.1 Structure and function

Mitochondria are approximately 0.5 to 1 μm in diameter, and contain both an inner and outer phospholipid membrane. The outer membrane is highly permeable to ions, nutrients, and small molecules such as adenosine triphosphate (ATP), while the selectively permeable inner membrane serves as the primary site of ATP generation via the electron transport chain.

Furthermore, each mitochondrion houses an average of 2 to 10 copies of mitochondrial DNA

(mtDNA) [108]. A simple schematic of the mitochondrion is shown in Figure 5.

24

Figure 5. Ultrastructure of a mitochondrion

1.3.1.1 The electron transport chain and oxidative phosphorylation

The electron transport chain (ETC) is a series of protein complexes (I through IV) lining the inner mitochondrial membrane, which serves as the primary source of intracellular energy production via oxidative phosphorylation. Electrons enter the ETC via oxidation of NADH and

FADH2, and are shuttled to complex IV by ubiquinone (also known as coenzyme Q10) and cytochrome c [109]. Complex IV then catalyses the reduction of oxygen to water, and in the process creates a proton gradient across the inner mitochondrial membrane. This proton gradient is used by ATP synthase (also known as complex V) to generate ATP from adenosine 25

diphosphate (ADP) and inorganic phosphate [109]. During this process, reactive oxygen species

- (ROS) such as hydrogen peroxide (H2O2) and superoxide (O2 ) are generated at complexes I and

III [110–112]. An overview of this process is presented in Figure 6. Increased production of ROS has been observed in cells with impaired ETC activity and mitochondrial DNA (mtDNA) damage [113], and ETC abnormalities have been associated with neurodegenerative diseases such as idiopathic Parkinson’s [114,115], Alzheimer’s [116,117], and Huntington’s [118,119].

Figure 6. The mitochondrial electron transport chain, oxidative phosphorylation, and production of ROS

26

1.3.2 Mitochondrial DNA (mtDNA)

The human mtDNA genome is double-stranded, circular and 16.5kb long. Unlike nuclear DNA

(nDNA), nearly every base pair in mtDNA is part of a coding sequence, apart from a highly polymorphic non-coding stretch of approximately 1100 base pairs known as the D-loop (or control region) [107]. MtDNA encodes 2 ribosomal RNAs (rRNA), 22 transfer RNAs (tRNA), and 13 proteins in total (Figure 7), comprising subunits of four out of five ETC complexes necessary for oxidative phosphorylation [108]. Human mtDNA is maternally inherited, owing to a ubiquitination/proteolysis mechanism in mammalian oocyte cytoplasm which selects against sperm mitochondria [120]. In addition, oocytes contain on average 200,000 copies of mtDNA compared to an average of 5 in sperm [121].

27

Figure 7. The circular mtDNA genome

1.3.2.1 Replication

MtDNA replicates independently of both cellular division and mitochondrial proliferation, although the number of mtDNA copies per cell is typically maintained within a homeostatic range [122]. The highly conserved sequences comprising the origins of replication and transcription for mtDNA both reside within the D-loop [123]. Initiation of mtDNA replication begins with the synthesis of a small mRNA molecule which primes the RNA-dependent DNA polymerase γ for transcription [124,125]. Polymerase γ, in addition to its replicative function, also features 3’-5’ exonuclease and 5’-deoxyribose phosphate lyase activities [125]. The copy 28

number of mtDNA per cell can be modulated based on the cell’s energetic needs as well as the presence of the nuclear-encoded mitochondrial transcription factor A, which acts on the promoter regions of mtDNA to regulate replication and transcription [122,126]. Elevated levels of mtDNA per cell in both brain and blood tissue have been associated with advanced age and/or markers of cellular senescence, suggestive of a physiological response to oxidative stress

[6,127].

1.3.2.2 ROS-induced oxidative damage

Excess endogenous ROS production and its interplay with mitochondria and mtDNA has been postulated as one of a number of theories which may explain the concept of cellular aging [123].

MtDNA is particularly susceptible to oxidative lesions and mutations compared to nDNA, due to its proximity to the primary site of endogenous ROS production, its lack of a protective histone core, and its relatively simple repair mechanisms [128].

ROS may induce mitochondrial dysfunction directly through a number of mechanisms, such as damage to mitochondrial proteins, enzymes, lipids, mtDNA, and nDNA [123]. ROS-mediated oxidative damage to mtDNA may produce a number of different modified nitrogenous bases, sugar radicals, strand breaks, and DNA-protein crosslinks [129,130]. The most well- characterized and oft-studied of these derivatives is 8-hydroxy-2’-deoxyguanosine, a commonly used marker of oxidative stress and DNA damage [129,130]. This lesion can result in loss of base-pair specificity, inducing G→T substitutions at the site of damage [131].

29

1.3.2.3 Maintenance

Although a small concentration of ROS is required by some tissues for cell signalling and

- pathogen defense, under normal physiological conditions excess ROS are reduced from O2 to the less toxic form H2O2 by the superoxide dismutase family of enzymes [132]. Three enzymes from this class (SOD1, SOD2, and SOD3) have been isolated from mammalian systems; the manganese-based SOD2 appears to be the only isoform localized to mitochondria [132]. While several repair mechanisms for mtDNA have been described, base-excision repair appears to be the preferential mode of repair for ROS-induced oxidative lesions [129]. In this process the damaged nitrogenous base is removed by a DNA glycosylase, leaving an abasic intermediate in its place [129,133]. These abasic sites can also arise as a byproduct of oxidative damage, and can block normal DNA replication leading to potential downstream cytotoxic events [133]. The abasic site is excised by a combination of lyases and endonucleases, which create a single-strand incision at the site of damage [129,133]. Finally, a nucleotide is re-inserted by polymerase γ and patched into the sequence by DNA ligase [129,133].

1.4 Autism spectrum disorder (ASD)

ASD encompasses a heterogeneous group of developmental disorders sharing a common set of behavioural phenotypes. It is characterized primarily by deficits in social interaction and a highly restricted set of behaviours, and can be accompanied by intellectual and language impairments

[134]. As is the case with several neurodevelopmental disorders, ASD etiology is largely idiopathic; however, unlike disorders with similar symptoms such as Rett syndrome and Fragile

X syndrome, there is no single, unified genetic risk factor for ASD. The clinical heterogeneity of symptoms, severity, and concomitant developmental disorders has complicated efforts to 30

uncover the genetic basis of ASD [135]. However, twin studies [136–138] as well as neuroimaging and neuroanatomical analyses [139,140] have provided strong evidence implicating a heritable, multifactorial genetic basis which may be exacerbated by environmental factors during prenatal development.

1.4.1 Diagnostic criteria and instrumentation

1.4.1.1 DSM-V

The American Psychiatric Association’s Diagnostic and Statistical Manual of Mental Disorders

[141] provides a standardized list of criteria for the diagnosis and categorization of mental disorders, including ASD (Table 2). The manual is currently in its fifth edition (referred to as the

DSM-V) which was published in 2013, replacing the DSM-IV published in 1994. The most relevant change to ASD between the DSM-IV and DSM-V is the re-classification of several previously distinct diagnoses of Asperger Syndrome, Autistic Disorder, Childhood

Disintegrative Disorder, and Pervasive Development Disorder – Not Otherwise Specified into the unified diagnosis of ASD [142]. Because the diagnostic criteria for these disorders were not significantly altered between editions, children given a diagnosis under DSM-IV criteria would typically not be required to be re-assessed under the DSM-V [142,143].

31

Table 2. Diagnostic criteria for autism spectrum disorder, as outlined in the Diagnostic and

Statistical Manual of Mental Disorders, Fifth Edition

Diagnostic Criteria for ASD Example Items (A, B, C, and D must all be met)

A. Deficits in social and communicative behaviour, evident in all three of the following categories:

Failure to respond to initiations of social 1. Absence of reciprocal social behaviour interaction; inability to maintain conversations

2. Poorly adapted nonverbal communication Abnormal eye contact; lack of facial patterns expression

3. Difficulty building and maintaining social Lack of interest in making friends; poor relationships adherence to social norms

B. Repetitive or restrictive behaviour, evident in at least two of the following categories:

Repeating certain words or phrases; 1. Repetitive speech or motor movements compulsion to line up or stack objects

Eating the same meal every day; extreme 2. Unusually ritualized or routine behaviour distress over small changes

3. Abnormal fixation on a restricted set of Strong attachment to certain items; lack of interests explorative behaviour

4. Inappropriate sensitivity to or Indifference to hot/cold; excessive preoccupation with sensory information smelling or touching

C. Symptoms are present in early childhood (before

36 months of age)

D. Symptoms must impair normal functioning

32

1.4.1.2 Screening and assessment

There is presently no formal screening tool available with sufficiently-demonstrated sensitivity and specificity to identify ASD. The American Academy of Pediatrics recommends all children be assessed by their family physicians for developmental delays and ASD at 18 and 24 months of age, but data suggest that a minority of primary care physicians actually conduct such surveillance [144,145]. Many cases are identified based on parental concerns at regular well- child visits, which have been shown to approach the standards of sensitivity and specificity to which other formal screening tests are held [146]. Some screening instruments recommended

(with appropriate caution) by both the American Academy of Pediatrics and the British

Columbia Ministry of Health include the Screening Tool for Autism in Two-Year-Olds (STAT)

[147], the Autism Screening Questionnaire (ASQ) [148], and the Checklist for Autism in

Toddlers (CHAT) [149].

Once a child is recognized by their primary care physician as being either at high risk for ASD or having suspected ASD symptoms, he/she is typically referred to a team of specialists for multidisciplinary assessment. Formal diagnoses can only be made by a paediatrician, child psychiatrist, or clinical psychologist with sufficient experience and training to recognize ASD, but depending on clinical presentation the child may also receive assessment by other professionals such as speech-language pathologists, occupational therapists, and neurologists

[144]. A typical diagnosis can be considered sufficiently reliable by approximately 2 years of age

[150], although Canadian and U.S. population studies suggest the average age of diagnosis is actually between 4 and 6 years [151]. Age of ASD identification and diagnosis has been shown to vary based on severity, access to care and screening, socioeconomic status, and whether the 33

child has any siblings with ASD [152–154]. Recent trends have shown that the age of diagnosis has steadily decreased over the last 25 years, likely owing to greater awareness, more accurate diagnostic tools, and increased efforts among care providers to promote early detection [154].

1.4.1.3 ADOS-G and ADI-R

There are currently two major instruments accepted for use in the standardized diagnosis of

ASD: The Autism Diagnostic Observation Schedule – Generic (ADOS-G) [155] and the Autism

Diagnostic Interview – Revised (ADI-R) [156]. The ADOS-G is a 30 minute physician- administered observational test which consists of four modules of 10 to 15 activities. These activities are designed to elicit skills and behaviours relevant to ASD, such as stereotyped use of words, reciprocal communication, and facial expressions. The module used depends on the subject’s level of expressive language skills, meaning the test can be administered across a broad age range, including adulthood. Behaviours are scored on a 3-point ordinal scale (0=normal,

1=infrequent/possible abnormality, 2=definite abnormality), and an algorithm is used to generate aggregate scores on three domains: social, communication, and social-communication total.

Classifications are made based on the number of domains for which the subject meets or exceeds predetermined cut-off values.

The ADI-R is a set of 93 standardized interview questions, administered with the child’s primary caregiver in approximately 90-120 minutes. Questions are designed to evaluate caregiver descriptions of the child’s current and historical behaviour across three sections: communication, social development, and repetitive/restrictive behaviour. Evaluators administer a 4-point ordinal score (0=normal, 1=infrequent or mild abnormality, 2=definite abnormality, 3=severe 34

abnormality) for each question, and an algorithm is used to calculate total scores for each of the three domains. ASD is signified when the child meets or exceeds cut-off scores in all three domains. The ADI-R and ADOS-G are typically administered in tandem, providing a holistic picture of the child’s observed current behaviour and developmental history.

1.4.2 Demographics

Based on 2010 data from the U.S. Centers for Disease Control and Prevention, ASD prevalence among children aged 8 years is approximately 1 in 68 or 1.47% [11]. Accurate estimations of

Canadian prevalence rates are complicated by the lack of a national surveillance system, a limitation acknowledged by the Public Health Agency of Canada [157]. Figures based on 2008 survey data from Newfoundland & Labrador, Prince Edward Island, and Southeastern Ontario estimate the Canadian prevalence at 98 in 10,000 children (0.98%) [158]. A global estimate of 62 cases per 10,000 (0.62%) was calculated in 2012, although data from Africa, the Middle East, and Southeast Asia are sparse due to inaccessibility of diagnostic services, lack of experienced diagnosticians, and relative dearth of population-level studies [159]. Several studies have reported increases in ASD prevalence over time, although it remains to be determined if this can be attributed to a true rate of incidence rather than systematic improvements in diagnostic availability, broadened diagnostic criteria, and increased awareness of ASD leading to earlier and more frequent identification [158,159].

1.4.3 Other factors associated with ASD

Maternal viral infections in pregnancy, particularly those acquired during the first trimester coinciding with a critical period in prenatal neurodevelopment [160], have previously been 35

associated with an increased risk of the child developing ASD. Animal models have shown that offspring of mice infected with influenza during pregnancy display abnormal behaviour reminiscent of ASD, and that at least part of this effect is directly attributable to activation of the maternal immune response [161]. Advanced maternal age (≥35yrs) and paternal age (≥40yrs) have also been associated with ASD, possibly due to mutations in germ cell DNA which accumulate with age [162].

ASD has a strong overlap with intellectual disabilities (ID), previously referred to as mental retardation, a related but distinct set of neurodevelopmental disorders characterized by IQ levels below 70-75 and compromised everyday life skills [163,164]. As ASD symptom severity can vary heavily from one end of the spectrum to the other, so too does the extent of compromise to intellectual capacity in affected individuals. It has been postulated that IQ level in ASD individuals can serve as a reliable indicator for the severity of ASD symptoms [163,164].

Genetic overlap has also been observed between ASD and ID, in terms of both heritable [165] and somatic [166] mutations, suggesting the two disorders may share certain characteristics in etiology at the cellular level as well as the phenotypic level.

At the population level, elevated incidence rates of certain gastrointestinal symptoms have been described in ASD children compared to those of non-ASD children [167]. It has been suggested that certain characteristic behavioural traits such as irritability and sudden mood changes in non- verbal ASD children may be attributable to very high observed rates of reflux esophagitis and gastrointestinal inflammation [168]. Distinct qualitative and quantitative changes in gastrointestinal flora have also been shown in ASD children, which may be permissive to the 36

colonization of microbial species which produce powerful neurotoxins [169]. However, more research is required before these differences can be fully understood.

Evidence has also suggested that ASD children are at an increased risk of experiencing seizures and epilepsy, particularly during adolescence [170–172]. This relationship appears to be more pronounced in those ASD children with significant ID [171,173]. While clinical research has yet to elucidate a pathophysiological mechanism underlying this relationship, various neuroimaging studies have demonstrated that ASD children are far more likely than the general population to harbor structural brain abnormalities, and on average have reduced activity in the frontal and temporal lobes [174,175]. This evidence has led to some postulation that the relationship may be underpinned by disruptions and damage at both the organic and molecular levels [172].

1.5 ASD, HIV, cART, and mitochondrial dysfunction

Both HIV and cART have been shown to put selective pressure on mtDNA through a variety of mechanisms. NRTIs can inhibit mtDNA replication directly through inhibition of polymerase γ, which incorporates the nucleoside analogs during mtDNA synthesis [176]. Additionally, patients treated with NRTIs have exhibited significant increases in somatic mtDNA mutation burden, which upon clonal expansion can lead to systemic mitochondrial dysfunction [177]. HIV infection, alone or in combination with cART, has been associated with implied oxidative stress

[178,179], which may cause oxidative lesions and large-scale deletions in the mtDNA genome.

Animal models have suggested that certain cART drugs may induce neuronal damage and cell death [180], and damage or disease to the central nervous system has been shown to result from chronic HIV-associated inflammation and immunological activation [181]. Findings with regards 37

to mtDNA content in peripheral blood of HEU children have been mixed, with some studies reporting decreases [7,182] and others reporting increases [6,183].

Several clinical observations in leukocyte mtDNA are shared between children with perinatal

HIV/cART exposure and children with ASD. These include significant differences (increases or decreases) in mtDNA quantity compared to controls [6,184], the presence of inferred markers of oxidative stress [185,186], and an increased number of mtDNA mutations [187,188]. Thus, quantitative and qualitative changes to mtDNA, which may lead to mitochondrial dysfunction, could present a mechanism through which HIV and cART exposure in pregnancy may be associated with an increased risk of developing ASD.

38

Chapter 2: Study design and participant demographics

Participants for this study were enrolled in either the Children & Women AntiRetrovirals &

Markers of Aging (CARMA) cohort study, or the BC Autism Spectrum Interdisciplinary

Research (ASPIRE) program.

2.1 Recruitment

CARMA participants for this study were prospectively enrolled from June, 2009 to February,

2015 as part of a larger cohort study examining the effects of HIV and cART on markers of cellular aging and oxidative stress (e.g. leukocyte telomere length, mtDNA quantity/quality), as well as HIV comorbidities affecting bone health, reproduction, and endocrine function.

Participants in the CARMA cohort primarily include HIV+ and HIV- women (including pregnant women), as well as HIV+ and HIV-exposed uninfected (HEU) children and adolescents. Most HEU children have perinatal cART exposure. Pediatric participants were recruited among four sites across Canada and had a CARMA visit approximately once a year between 2008 and 2012, while this was reduced to once every 2 to 2.5 years from 2013 onward.

Each site/province follows HEU children for a different length of time as per standard of care

(Table 3). Eligible individuals were approached for enrolment by CARMA staff during a routine visit to their HIV clinic, and received a $20 honorarium per visit. Attached in the appendix is a copy of the consent form used at the Oak Tree Clinic in Vancouver; other sites used similar versions.

ASPIRE participants were recruited in Vancouver, BC via word of mouth for the purposes of investigating genetic factors which may underlie the development of ASD. Children with ASD 39

as well as their siblings and parents were invited to participate in the study, yielding a mixed cohort of ASD and non-ASD participants. ASPIRE visits occurred at the Child & Family

Research Institute at the BC Children’s Hospital in Vancouver. For both CARMA and ASPIRE participants, whole blood samples were collected at each study visit.

Table 3. CARMA recruitment sites, follow-up parameters, and HEU-relevant demographics for participants in our study

Centre Children’s Oak Tree The Hospital for Hospitalier Hospital of CARMA visit site Clinic Sick Children Universitaire Eastern Ontario (Vancouver) (Toronto) Sainte-Justine (Ottawa) (Montréal) Length of time participants are Until age 18 Until age 19 Until age ~5 Until age 19 followed by the clinic months with blood collection HEU participants within this study 25 (45) 0 2 (4) 29 (52) n (%)

HEU participants with ASD 11 (79) 0 2 (14) 1 (7) n (%)

Maternal cART regimen taken in pregnancya AZT+3TC+PI/NNRTI 9 - 1 16 TDF+FTC+PI/NNRTI 2 - 0 2 ABC+3TC+PI/NNRTI 4 - 0 4 None 5 - 0 0 Other 4 - 0 7 Unknown 1 - 1 0 a If the mother switched regimens during pregnancy, the regimen taken for the majority of pregnancy was reported

40

2.2 Eligibility and selection criteria

CARMA enrolment occurred in two phases: the first phase occurred from 2008 to 2012, in which all persons living with HIV and/or exposed to cART in utero were eligible to participate. The second phase has been ongoing since 2013, in which visits for previous enrolees were continued, but enrolment of new participants was limited to HEUs aged 10-19 years, new HIV+ cases (any age), HEU children with diagnosed or suspected ASD (any age), and any child with no pre- or postnatal exposure to cART who was born to an HIV+ mother. Exclusion criteria for CARMA are few, with only those participants unable to provide informed consent and/or adequately speak and understand English/French not considered for inclusion.

In an earlier iteration of our study, HEU children with ASD (n=16) were matched 1:2 on sex, ethnicity, and age with HEU children without ASD (n=32) and 1:2 on sex and age with HIV- unexposed, uninfected (HUU) anonymous controls (n=32). Through a collaboration with

ASPIRE, we were able to add a fourth group to our study, namely HUU children with ASD. This study design would allow us to explore how ASD and HEU status might differentially affect mtDNA. Therefore, we re-matched our study groups under blindness to all of our previous matching in order to obviate any potential bias related to our previous partial analyses. We also expanded our matching from 2:1 to 3:1 in order to increase the statistical power of our study, given the relatively low sample size of our group of interest. It was also discovered upon follow- up that two of the 16 HEU children who had previously been identified as having ASD by their physician were found not to meet diagnostic criteria, and therefore they, along with their matches, were removed from all analyses. 41

Thereafter, CARMA HEU children with a formal diagnosis of ASD (n=14) were matched 1:3 on sex, ethnicity, and age with CARMA HEU children without ASD (n=42), and on sex and age to

HUU anonymous controls (n=42) for whom only age and sex data were available (see section

2.4). They were also matched 1:3 on sex, age, and whenever possible ethnicity with ASPIRE

HUU children with ASD (n=42), and any available HUU non-ASD siblings of ASPIRE ASD children (n=9). We combined the ASPIRE HUU siblings and HUU anonymous controls into a single control group for the purposes of statistical analyses (n=51).

For those HEU participants with ASD for whom multiple whole blood samples were available, the sample collected closest after the date of ASD diagnosis was used. If the date of diagnosis was unavailable at the time, the most recent available sample was used. The full number of selections and exclusions are provided in the study design diagram below (Figure 8).

42

Figure 8. Selection process, exclusions, and design of our cross-sectional, case control study

2.3 Participant demographics

At each CARMA visit, clinical and demographic information was collected. A history of the mother’s pregnancy was also collected at enrolment. Some demographic and anthropometric factors collected by CARMA included the child’s date and country of birth, estimated gestational age at birth, sex, ethnicity (as self-reported), and paternal and maternal dates of birth. Clinical data collected included, among other factors, platelet count, blood and lipid panel results, type and duration of any perinatal cART exposure, and any maternal substance use the child was exposed to in pregnancy (e.g. tobacco, alcohol, illicit drugs) if available. In addition, any 43

reported health issues in the child related to such factors as gastrointestinal health, psychological conditions, haematology, or other conditions requiring hospitalization were recorded.

2.4 Sample collection and preparation

For CARMA HEU participants, two aliquots of whole blood (4mL each) were drawn at each visit via arm venipuncture, and collected in lavender K2 ethylenediaminetetraacetic acid (EDTA) tubes (BD Vacutainer®). All blood samples were shipped at room temperature to the BC

Women’s Hospital in Vancouver, British Columbia and transferred to 2mL microcentrifuge tubes and frozen at -80˚C within 48h of collection. ASPIRE blood draws were collected in the same manner, but were frozen at -20˚C instead of -80˚C before being transferred to UBC

Hospital for this study, where they were frozen at -80˚C. The collection dates of ASPIRE samples used in this study ranged from approximately 2002 to 2014; CARMA HEU samples used were collected between January 2010 and April 2015. HUU anonymous controls were comprised of leftover whole blood samples from routine blood work at the emergency ward of the BC Children’s Hospital in Vancouver. These samples were anonymized and only age and sex information was provided to CARMA.

Before extraction, CARMA whole blood was diluted 1:1 with 1x phosphate-buffered saline.

Genomic DNA was extracted using the QIAcube and QIAamp DNA Mini Kit (Qiagen) using the

Blood & Body Fluid protocol (https://www.qiagen.com/us/resources/download.aspx?id=

67893a91-946f-49b5-8033-394fa5d752ea&lang=en, accessed 21-Feb-16). Samples were eluted in 100μL of Buffer AE (Qiagen) containing 10mM Tris-Cl and 0.5mM EDTA (pH 9.0). The average concentration of sample DNA was determined via spectrophotometry as approximately 44

40ng/μL. ASPIRE samples were extracted using the Puregene DNA Isolation Kit (Gentra) with

Qiagen’s Whole Blood protocol (https://www.qiagen.com/us/resources/download.aspx?id= a9e6a609-4600-4b03-afbd-974318590ce5&lang=en, accessed 21-Feb-16) and eluted into

Puregene DNA hydration solution. DNA concentration of ASPIRE samples was determined by spectrophotometry to be on average 35ng/μL.

All whole blood DNA samples were diluted 1:10 with Buffer AE and randomly assigned a blinding tag by an independent lab member before any assays were performed. The translation key for unblinding samples was kept in an encrypted spreadsheet until statistical analyses of data were required.

45

Chapter 3: MtDNA content measurements

3.1 Measurement technique

MtDNA content was expressed as the ratio between mtDNA copy number and the copy number of a single-copy nuclear gene, as determined via real-time quantitative polymerase chain reaction (qPCR). The non-coding mitochondrial D-loop region and albumin were chosen as the mitochondrial and nuclear genes, respectively. The primer sequences used are shown in Table 4, and primers were purified via high-performance liquid chromatography (Integrated DNA Technologies). Samples were assayed using a monochrome, multiplex technique adapted from Cawthon et al [189], allowing for the quantification of mitochondrial and nuclear genes in the same well.

Table 4. Forward and reverse primer sequences used to measure mtDNA content via monochrome, multiplex qPCR

Primer identifier Sequence

Albumin Albu F 5’-CGGCGGCGGGCGGCGCGGGCTGGGCGGAAATGCTGCACAGAATCCTTG-3’ Albd R 5’-GCCCGGCCCGCCGCGCCCGTCCCGCCGGAAAAGCATGGTCGCCTGTT-3’

D-loop D-loop_MPLX(2) F 5’-ACGCTCGACACACAGCACTTAAACACATCTCTGC-3’ D-loop_MPLX(2) R 5’-GCTCAGGTCATACAGTATGGGAGTGRGAGGGRAAAA-3’

46

For each reaction, 2μL of sample DNA was added to 8μL of master mix in each well of a white,

96-well LightCycler 480® Multiwell plate (Roche). All samples were assayed in duplicate, and each plate included a negative control, a 7-sample standard curve, and two internal controls (IC), allowing for quantification of 40 samples per plate. Children who were matched with one another were assayed on the same plate in order to reduce inter-assay bias when performing between-group comparisons, and the positions of samples within plates were randomly assigned to control for plate effects. Master mix reagents included 1X FastStart SYBR Green Master

(Roche), 1.2mM EDTA, and the two primer pairs noted in Table 5, each at 0.9μM. Between delivery into plate wells and qPCR, plates were centrifuged at 1500*g for two minutes at room temperature. The thermal cycling profile used is shown in Table 7. Samples were assayed on the

LightCycler® 480 platform (Roche), Software Version 1.5.1.62 SP2.

Table 5. Thermal cycler settings for monochrome, multiplex qPCR of nuclear (albumin) and mitochondrial (D-loop) sequences

Target temp Acquisition Hold time Temp. ramp Program No. of cycles (°C) mode (mm:ss) rate (°C/s) Pre-incubation 1 95 None 15:00 4.4 94 None 00:15 2.2 62 None 00:10 2.2 Amplification 40 74 Single 00:15 4.4 84 None 00:10 4.4 88 Single 00:15 4.4 95 None 01:00 4.4 Melting Curves 1 45 None 00:01 2.0 95 Continuous ------Cooling 1 40 None 00:01 1.5

47

The standard curve on each plate was generated by serial dilution (1:5) of two cloned plasmid

DNA sequences containing the albumin and D-loop regions of interest mixed in a 1:50 ratio.

Standards ranged from 5,075,625 to 325 copies of albumin, and from 237,952,683 to 15,229 copies of D-loop, resulting in a 15,625-fold linear range (R2>0.99). The two ICs were derived from a 1:4 dilution of pooled volunteer whole blood DNA (low mtDNA content), and an extract from cultured SKBR3 cells (high mtDNA content). The negative control contained Buffer AE in place of template DNA.

Dual-signal information was processed by importing fluorescence data in text format from the

LightCycler 480 software into a spreadsheet and sorting results into 74°C (D-loop) acquisitions and 88°C (albumin) acquisitions. Data were converted to a grid format using the

LC480Conversion Version 2.0 software and threshold cycles for each sample calculated using

LinRegPCR Version 2012.1 software (both available through the Heart Failure Research Centre,

Amsterdam). Copy numbers for mitochondrial and nuclear DNA were calculated by comparing the crossing points of individual samples to a plot of the standard curve generated by Microsoft®

Excel 2010. PCR efficiency was calculated for each gene from the standard curve.

3.2 Quality control

Runs were rejected if the negative control was not a true negative, or >3 cycles below the lowest point on the standard curve. Individual samples were rejected and re-assayed if the copy number of either of the two genes lay outside the standard curve, or if the absolute % difference between duplicate ratios was >15%. Samples outside the standard curve were re-assayed undiluted. If the absolute % difference between duplicates was >15% in two different runs then an average of all 48

four replicates was taken as the accepted value unless the coefficient of variation (CV), defined as SD*100%/mean, between all four values was >10%. Additionally, for each run the mtDNA content ratios for standards and ICs were logged, along with the PCR efficiencies of each gene and the average absolute % difference between duplicates. Runs were rejected and repeated if they failed to meet at least four of the following five conditions:

1) MtDNA content ratios for the first IC must be within ±2 standard deviations from the

mean of all IC ratios measured in our lab using the same standards and controls

2) MtDNA content ratios for the second IC must be within ±2 standard deviations from

the mean of all IC ratios measured in our lab using the same standards and controls

3) PCR efficiency for each gene must be >1.90, and <2.05

4) The difference between the PCR efficiencies of the two genes must be <0.05

5) The average of all absolute % differences between duplicates for the run must be

≤10%

The reproducibility of our results were tested by semi-randomly selecting n=28 samples from the main study arm and repeating the assay on these samples, keeping the position of standards and controls on the plate fixed. Measurements from the two independent runs were plotted on a linear regression curve and the goodness-of-fit coefficient (R2), slope of the best-fit line, and Pearson product-moment correlation were calculated in XLSTAT 2015 v2.02 (Addinsoft).

49

3.3 Statistical analyses of data

3.3.1 Univariate between-group comparisons

The two-tailed Student’s t-test and Mann-Whitney U test were used to compare mtDNA content between groups, after ascertaining the normality of data within each group via the Shapiro-Wilk test. For between-group comparisons, a p-value <0.05 was considered statistically significant.

Various univariate association analyses between mtDNA content and type/duration of perinatal cART exposure in HEU children were also explored. All analyses were conducted in XLSTAT.

Unpaired t-tests were employed for between-group comparisons for several reasons. Firstly, matching between study groups was inexact (see section 3.3.3) and the potential effects of age, sex, and ethnicity on mtDNA content are not currently well-described. Any differences which could be accounted for, or at least influenced by, mismatches between groups would thus render pairwise comparisons inappropriate. Another important consideration is that matching was performed 3:1, or more in the case of the HUU control group. Accordingly, it would be necessary to compare means rather than individual observations in certain cases, in order for the number of observations between groups to match for pairwise evaluation. However, as an a priori decision was made to omit data from the study if certain quality control standards were not met (see previous section), in addition to the aforementioned >3:1 matching for the HUU control group, the number of observations used in the calculation of these means would fluctuate on a case-by-case basis. This presents two matters of complication: firstly, the degrees of freedom assumed in these pairwise comparisons would not be representative, thus influencing p-values in an unpredictable fashion, and secondly, in the event that all 3:1 matches within a group failed to meet quality control standards, paired comparisons for this case would be impossible. We thus 50

opted to perform unpaired tests both to present the most conservative an impartial estimates possible, and to prevent the unnecessary removal of data from an already small sample size in our group of interest (n=14). For interest, if we re-perform all univariate analyses using paired

Student’s t-tests and Wilcoxon signed-rank tests, we note that p-values are similar to those of their unpaired counterparts (p=0.003 to p=0.3).

3.3.2 Multivariable linear regression

We designed three multivariable linear regression models to further analyse and describe our results. The first model included all children for whom maternal age at birth, mtDNA content, age, sex, and ethnicity data were available (n=106 total). The model also included HEU status and ASD diagnosis as variables. The second multivariable regression model included factors thought to be associated with ASD for which we had data available from both CARMA and

ASPIRE children with ASD (n=52 children in total). The model included demographic (age, maternal age at birth) and clinical factors (chronic gastrointestinal conditions, history of seizures/epilepsy, hypotonia, HEU status). Coding for categorical variables in this model was limited to binary inputs (e.g. ‘yes/no’ answers) due to our sample size and the lack of more consistently available detailed information between the two cohorts. Whenever more details were available we report it in a qualitative manner. The third model included explanatory variables related to perinatal maternal cART exposure (trimester of cART initiation, type of NRTIs taken, duration of exposure) suggested in literature as potential risk factors for mitochondrial toxicity and adverse fetal neurodevelopmental outcomes. ASD and non-ASD HEU children with known cART drug exposure were included (n=49). All three of these models were constructed and analysed using the ANCOVA algorithm in XLSTAT. 51

3.3.3 Sensitivity analyses

Due to an ethnicity mismatch between CARMA and ASPIRE children, a sensitivity analysis was performed to examine whether mtDNA content differed significantly between well-matched

Asian/South Asian, Caucasian, and African-Canadian ethnic groups. CARMA HEU children without ASD of Asian or South Asian ethnicity (n=10) were matched 1:2 on sex and age with non-ASD HEU children of Caucasian (n=20) and African-Canadian (n=20) ethnicity. All samples were blinded and randomized in the same manner as in the main arm of the study, with matched samples assayed on the same plate. Between-group comparisons were done using two- tailed Student’s t-test. We also conducted power/sample size calculations using a web-based module provided by the UBC Department of Statistics (http://www.stat.ubc.ca/~rollin/stats/ ssize/n2.html, accessed 19-Feb-16).

Additionally, after all data were collected we examined the correlation between mtDNA content and albumin gene count. This would allow us to determine whether samples with high content could be attributed to low nuclear gene count, possibly indicating sample degradation, leukopenia, or polymorphisms at the albumin primer binding site. Spearman’s rank correlation test was used to analyse these data.

Finally, for HEU children with and without ASD we examined the correlation between mtDNA content and platelet count at the time of the child’s CARMA visit. Since platelets contain a small amount of mtDNA but no nDNA, mtDNA content may have been confounded by an abnormally

52

high platelet count. We used either Spearman’s rank correlation or Pearson product-moment correlation tests to analyse data, as appropriate.

53

Chapter 4: MtDNA apparent oxidative damage measurements

4.1 Measurement technique

Figure 9 summarizes the measurement of mtDNA apparent oxidative damage (AOD) using a long-range polymerase chain reaction (PCR) technique as described below.

Figure 9. MtDNA AOD assay and calculations

54

4.1.1 Sample preparation

The same 1:10 dilutions of whole blood DNA used to calculate mtDNA content were used for the mtDNA AOD assay, except for one sample that contained fewer than 20,000 D-loop copies per μL. In this case the original, undiluted whole blood DNA was assayed, and the D-loop copy number of this sample was extrapolated by multiplying the observed copy number from the mtDNA content assay by 10. Full details of sample preparation are contained in section 2.4.

Aliquots of each sample (ranging between 2 and 8μL) were delivered into 1.7mL microtubes and diluted using Buffer AE to a common concentration of 20,000 D-loop copies/μL. Each tube contained at least 14μL of diluted sample in order to be assayed twice if necessary. The 20,000 copy number was chosen based on prior assays showing that at 27 cycles when the long PCR is stopped, samples are still in the exponential growth phase.

4.1.2 Long PCR of mtDNA

Following dilution, samples were prepared for long-range PCR of the mtDNA genome using reagents from the Expand Long Range dNTPack (Roche) and long PCR primers (Table 6). The primers used cover half the mtDNA genome (approximately 8.5kb), and have been shown to produce results that are sufficiently reliable and reproducible to extrapolate to the entire mtDNA genome. The reagent concentrations used in the long PCR master mix were as follows: 1X Long

Range Buffer 3 (without MgCl2), 2.75mM MgCl2, 0.5mM dNTP, 0.3mM primers MT7988F and

MT708R, 3% dimethyl sulfoxide (DMSO), and 1.25U Expand Long Range Enzyme Mix. In each reaction 3μL of sample DNA was added to 22μL of master mix in a 0.2mL PCR tube and 55

kept on ice to minimize the pre-reaction activity of the heat-sensitive PCR polymerase. A negative control containing PCR-grade water (Roche) in place of DNA was prepared to monitor for contaminants present at the long PCR preparation step.

Table 6. Forward and reverse primer sequences used for long PCR of mtDNA

Primer identifier Sequence

MT7988F 5’-CTCCTTGACGTTGACAATCGAGTAGT-3’ Long PCR fragment MT708R 5’-GGGGATGCTTGCATGTGTAATCTTAC-3’

Reactions were carried out in duplicate on the 96-well MyCycler™ platform (Bio-Rad), with the thermal cycling profile shown in Table 7. The holding cycle was maintained for no longer than

18 hours after completion of the reaction. The outer edges of the 96-well platform were used only for the negative and severe-damage controls (described in section 4.2) due to previous assays showing highly variable results in these positions. To ensure that samples would lie within the standard curve during the following quantification step, 2μL of long PCR products, including the negative control, were diluted 1:500 in 998μL of DNAse/RNAse-free UltraPure™ distilled water (Invitrogen) and stored in 1.5mL elution tubes at -20˚C until use in qPCR.

56

Table 7. Thermal cycler settings for long PCR of mtDNA

Program No. of cycles Target temp (°C) Hold time (mm:ss)

Denaturation 1 93 02:00

Denaturation 93 00:10

Annealing 27 58 00:30

Extension 68 08:00

Extension 1 68 07:00

Holding 1 4 ∞

4.1.3 Quantitative PCR and AOD calculations

The 1:500 dilutions of long PCR products and the 20,000 copy pre-PCR samples were quantified using qPCR on the same LightCycler 480 platform and software used for the mtDNA content assay. However, as nDNA quantification was not required in this assay the only primers included in the master mix were shortened D-loop primers used for monoplex quantification (Table 8). As in the multiplex assay, 2μL of sample template was added to 8μL of master mix in each plate well. The master mixed used in this monoplex assay consisted of 1X FastStart SYBR Green

Master (Roche) and 1μM MT325F and MT474R primers, diluted to concentration with PCR- grade water (Roche). The thermal cycling profile used for the monoplex D-loop assay is shown in Table 9. The D-loop standard curve used in this assay was prepared using the same cloning conditions as for the multiplex qPCR standards, but was produced via 1:10 serial dilution rather than 1:5. Individual standards ranged from 35,400,000 to 354 D-loop copies, a 100,000-fold linear range (R2>0.99).

57

Table 8. Forward and reverse primer sequences used for qPCR of the mtDNA D-loop

Primer identifier Sequence

MT325F 5’-CACAGCACTTAAACACATCTCTGC-3’ qPCR: D-loop region MT474R 5’-AGTATGGGAGTGRGAGGGRAAAA-3’

Each individual sample had its pre-long PCR and post-long PCR counterparts assayed on the same qPCR plate, in order to eliminate inter-assay bias during calculation of mtDNA AOD. As outlined in Figure 9, the amplification efficiency of each sample was expressed as the ratio of its post-long PCR D-loop copy number to its pre-long PCR copy number. These values were then normalized to the amplification efficiency of an undamaged internal control and expressed as mtDNA AOD. As some samples were found to amplify more efficiently than the theoretically undamaged control, we multiplied all AOD values in our study by a common factor to ensure they would fall between 0 and 1, rendering results more intuitive without affecting analyses.

Table 9. Thermal cycler settings for mtDNA D-loop quantification using qPCR

Target temp Acquisition Hold time Temp. ramp Program No. of cycles (°C) mode (mm:ss) rate (°C/s)

Pre-incubation 1 95 None 10:00 4.4 95 None 00:05 4.4 Amplification 45 60 None 00:10 2.2 72 Single 00:05 4.4 95 None 01:00 4.4 Melting Curves 1 45 None 00:01 2.0 95 Continuous ------Cooling 1 40 None 00:01 1.5 58

4.2 Quality control

Each long PCR included between 3 and 5 ICs (each assayed in duplicate) intended to model mtDNA that was undamaged (Happy control), moderately damaged (Medium control), or severely damaged (Unhappy control). For the first six long PCR runs three separate Happy ICs were included on each plate: two were derived from volunteer whole blood DNA and one was extracted from an immortalized cell line. The two whole blood samples were assayed at different pre-long PCR copy numbers in order to ascertain how the assay accommodates large-scale variations in DNA input concentration, and the cell culture control was included to determine whether DNA collected from two different media would amplify in the same manner. It was determined a priori that we would normalize our results to whichever of these three controls behaved most consistently across six runs. Descriptions of the controls used are provided in

Table 10.

Table 10. Characteristics of the 5 ICs used in long PCR and qPCR for the mtDNA AOD assay

Control Description

Happy 1 Volunteer whole blood DNA diluted to 30,000 copies

Happy 2 Volunteer whole blood DNA diluted to 50,000 copies

Happy 3 DNA extracted from JEG-3 cell culture (human placenta choriocarcinoma) DNA extracted from CEM cell culture (human T-cell leukemia), treated with Medium 100mM H2O2

Unhappy DNA extracted from JEG-3 cell culture treated with 100mM H2O2

59

Quality control parameters for the qPCR stage of the assay were similar to those for the mtDNA content multiplex assay (section 5.1.1), except for a relaxed set of criteria related to variation between duplicates. This is owed to a global increase in intra-assay variability using monoplex qPCR compared to the multiplex technique. Individual samples were re-assayed if the absolute

% difference between duplicates was >20% rather than 15%, and runs were rejected and repeated only if the average of absolute % differences for all samples was >15% rather than 10%. Long

PCR runs were repeated if the average absolute % difference between amplification efficiencies of the long PCR duplicates was >20%. If a sample was re-assayed at the qPCR stage, then both long PCR duplicates and the pre-long PCR counterpart were each re-assayed. If a sample’s amplification was not within ±2 SD of the average on two separate long PCR runs, then it was excluded from analyses.

Gel electrophoresis on 0.8% agarose was used to visualize samples that appeared to fail to amplify during long PCR, signified by an amplification value >2 SDs below the mean of all values. Gel electrophoresis was also used to compare samples that amplified at high efficiency and those that amplified at low efficiency. Samples were compared qualitatively in terms of product size and band intensity. This would allow us to determine if there were any apparent large insertions or deletions in the mtDNA genomes of these samples, which may account for differences in their amplification efficiencies. A 1kb Plus DNA ladder (Life Technologies) diluted 1:10 was used for fragment size comparisons.

60

4.3 Statistical analyses of data

Between-group comparisons of mean mtDNA AOD were done using the same methodology as for the univariate analyses of mtDNA content data (section 3.3.1).

To quantify variability due to the 1:500 dilution step, a random long PCR product was diluted 10 times in the manner described in section 4.1.2. The D-loop copy numbers of these individual dilutions were quantified in duplicate via monoplex qPCR and the CV between average copy numbers of each dilution was calculated.

61

Chapter 5: Results and analyses

Because the methods used to measure both mtDNA content and mtDNA AOD were adapted from other techniques and represent currently unpublished work, this thesis will cover both the performance of these assays as well as the case control study of ASD in HEU children.

A novel finding of this study is the increased prevalence of ASD in HEU children within our cohort (14/299, 4.7%) compared to population estimates of 1.5% (section 1.1). We hypothesized that HEU children with ASD would differ significantly in terms of leukocyte mtDNA content and leukocyte mtDNA AOD compared to well-matched groups of non-ASD HEU children and/or HUU children with and without ASD. Differences observed may be indicative of mitochondrial dysfunction, which has been associated with perinatal exposure to HIV and cART, and with an increased risk of ASD.

Within our study’s group of interest (HEU with ASD), participants were predominantly male and of African-Canadian ethnicity, with ages ranging from 2 to 16 years at time of first CARMA visit. Table 11 provides a more detailed overview of the demographics of children within our four study groups. As much as possible, efforts were made to collect information that might be relevant to ASD such as ASD symptom severity, date of diagnosis, intellectual capacity, and any available ADOS-G or ADI-R scores. This information was gathered by petitioning CARMA physicians to reference patient charts within the confines of CARMA’s ethics agreement, or to provide their own qualitative observations.

62

Table 11. Demographic characteristics of participants in each experimental group

HEU with ASD HEU without ASD HUU with ASD HUU without ASD

n=14 n=42 n=42 n=51 Male sex 10 (71) 30 (71) 30 (71) 36 (71) Age at sample collection (years) 6 [4 – 8] (2 – 16) 6 [4 – 8] (2 – 16) 6 [4 – 8] (2 – 16) 6 [4 – 9] (2 – 16) Self-reported ethnicity Black/African Canadian 11 (79) 33 (79) 4 (10) 0 White/Caucasian 3 (21) 9 (21) 20 (48) 5(10) Asian/South Asian 0 0 6 (14) 4 (8) Other/Mixed Ethnicitya 0 0 12 (29) 0 Unknown (anonymous) 0 0 0 42 (82) Maternal cART regimen taken in pregnancyb AZT+3TC+PI/NNRTI 6 (43) 19 (45) - - TDF+FTC+PI/NNRTI 2 (14) 2 (5) - - ABC+3TC+PI/NNRTI 0 8 (19) - - None 3 (21) 2 (5) - - Other 1 (7) 10 (24) - - Unknown 1 (7) 1 (2) - - PI-based cART 8 (57) 35 (83) - - NNRTI-based cART 1 (7) 2 (12) - - Both PI and NNRTI 0 1 (2) - - NRTIs only 1 (7) 1 (2) - - Infants who received AZT prophylaxis 13 (93) 42 (100) - - Length of maternal cART exposure (weeks)c 11 [6 – 35] (0 – 38) 33 [20 – 39] (0 – 41) - - Length of AZT prophylaxis (weeks)d 6 [4 – 6] (0 – 7) 6 [6 – 6] (4 – 7) - - Median [IQR] (range) or N (% of total) a Includes Aboriginal, Latin American, and children of mixed background (e.g. half Aboriginal & half Caucasian) b If the mother switched regimens during pregnancy, the regimen taken for the majority of pregnancy was reported c Maternal cART exposure time was available for 12/14 and 40/42 HEU children with and without ASD respectively d Duration of AZT prophylaxis was available for 13/14 HEU children with ASD 63

5.1 MtDNA content

Using a monochrome, multiplex qPCR technique, we measured leukocyte mtDNA content

(mtDNA:nDNA copy number ratio) in genomic DNA extracts taken from whole blood. This measurement would provide an estimate of the number of mtDNA copies per cell, indicative of the relative synthetic activity of mitochondrial polymerase γ. Inhibition of this enzyme may be associated with significant increases or decreases in mtDNA content.

5.1.1 Quality control

We quantified inter-and intra-run variability of the assay by monitoring the performance of an IC derived from the SKBR3 adenocarcinoma cell line (IC High). MtDNA content of our study samples was measured across a total of six LightCycler runs, each containing 40 samples. Inter- assay CV (5.3%) was calculated based on the mtDNA content of IC High across each of the six runs, while the same IC High control was delivered into n=12 wells on a single plate to calculate the intra-assay CV (4.9%). The inter-assay CV of an IC derived from whole blood (IC Low,

4.8%) did not differ from that of IC High (Table 12). Based on our quality control parameters, two samples fell below the lower limit of the assay and were repeated undiluted but were ultimately excluded from analyses as the CV of the four replicates was >10%.

64

Table 12. Variability in mtDNA content measurements of the two ICs over six runs, and within-run measurements of 12 IC High replicates

IC Low IC High IC High (between runs) (between runs) (within run) n=6 n=6 n=12

MtDNA content 99 ± 5 1303 ± 70 1225 ± 60 Mean ± SD (range) (91 – 104) (1205 – 1400) (1126 – 1319)

CV (%) 4.8 5.3 4.9

5.1.2 Reproducibility

We next sought to examine the reproducibility of the multiplex assay’s measurements by re- quantifying mtDNA content in a selection of study samples. To examine the assay’s performance across the entire range of our observations, samples included in the reproducibility analysis were intended to span a representatively broad range of mtDNA content measurements (ranging from

40 to 215), but within these constraints they were otherwise selected randomly. Original measurements spanned all four experimental groups and five out of six runs. The results of this analysis are shown in Figure 10. The correlation between the two datasets was high (R2=0.93), however, the slope of the regression line indicates that on average there was a slight increase in measurements upon re-assaying, indicative of a minor systematic shift in the standard curve between the two runs.

65

Figure 10. Correlation between the original and repeat measurements of 28 semi-randomly selected samples. R2 and Pearson’s r are shown

5.1.3 Statistical analyses

Between-group comparisons of mtDNA content were performed using univariate and multivariable analyses, both to control for the effects of demographic and clinical factors as well as to explore other potential variables which may be associated with significant changes in mtDNA content in HEU and/or ASD children. Potential relationships between maternal cART exposure (type and duration) and mtDNA content were explored qualitatively, as sample sizes were generally not permissive to more formal analytical techniques. 66

5.1.3.1 Between-group comparisons

Figure 11 shows box plots of mtDNA content in each group as well as between-group comparisons. HEU children with ASD had the highest mtDNA content of any of the four study groups (p<0.0001 to p=0.02 via Mann-Whitney U test). HEU children without ASD and HUU

ASD children also had higher mtDNA content than the HUU control group (p=0.004 and p=0.03, respectively, via Student’s t-test). However, the HEU non-ASD group and HUU ASD group did not significantly differ from one another (p=0.2 via Student’s t-test). The independent predictive ability of mtDNA content for the presence of ASD across HEU and HUU participants was verified using a nominal logistic regression model (McFadden R2=0.307, Table 13).

Table 13. Multivariable logistic regression analysis of potential explanatory variables for

ASD in 106 study participants

Odds Ratio Explanatory variables (for presence of ASD) P-value (95% CI) MtDNA content 1.02 (1.00 – 1.03) 0.016 HEU status (vs. HUU) 0.03 (0.01 – 0.20) 0.0002 Male sex (vs. female) 0.97 (0.33 – 2.89) 0.96 Ethnicity (vs. Caucasian) African-Canadian 1.15 (0.22 – 5.89) 0.87 Asian/Other 0.73 (0.16 – 3.32) 0.68 Child’s age at sample collection, years 0.97 (0.82 – 1.14) 0.68 Maternal age at childbirth, years 0.94 (0.86 – 1.03) 0.16

Table 14 provides summary statistics of the mtDNA content for the four groups along with

ASD-relevant demographic and clinical data that were available. Student’s t-tests showed that

67

study groups did not differ in terms of maternal age (p=0.08 to p=0.6) or paternal age (p=0.1 to p=0.7). Other data were sparse, but qualitative observations suggest that gastrointestinal abnormalities were far more common in HEU children with ASD (71%) compared to other groups (17% and 21% in HEU without ASD and HUU with ASD, respectively). The severity of

ASD symptoms appeared to be greater in HEUs than HUUs: 54% of the HEU children with ASD were described as being on the more severe end of the spectrum, compared to 31% in the HUU

ASD group. Low muscle tone was more common in ASD children (21-33%) compared to non-

ASD children (5-11%), along with epilepsy and seizures, which occurred exclusively in ASD children. Diagnoses of ID and developmental delays were far more common in ASD children

(71% in both HEU and HUU with ASD) than non-ASD children (19% and 0% for HEU and

HUU children, respectively). Notably, no specific diagnoses of ID were found in HEU children with ASD. However, we did note that of those children for which CARMA staff were able to retrieve cognitive and developmental test scores (n=10), nine had scores in the first or second percentile for cognitive skills (Bayley Scale of Infant Development, Mullen Scale of Early

Learning, or Stanford-Binet Intelligence Scale) and/or adaptive intelligence (Vineland Adaptive

Behaviour Scale, Adaptive Behaviour Assessment System). The one child who did not have scores in this range (age 16) was noted as being of normal cognitive ability, but had specific areas of weakness in verbal skills (16th percentile) and short term memory (9th percentile).

68

Figure 11. Univariate between-group comparisons between mtDNA content of HEU and

HUU children with and without ASD. P-values from Student’s t-test or Mann-Whitney U test shown, as appropriate

69

Table 14. MtDNA content, maternal and paternal age, and prevalence of ASD-relevant factors in HEU and HUU children with and without ASD

HEU HEU HUU HUU with ASD without ASD with ASD without ASD n=14 n=42 n=42 n=49 163 [148 – 177] 118 [93 – 159] 111 [98 – 131] 100 [79 – 124] MtDNA content (48 – 215) (20 – 220) (30 – 179) (35 – 167) 30.4 [23.6 – 32.3] 33.1 [30.7 – 36.1] 31.0 [28.0 – 34.8] 27.0 [26.0 – 36.0] Maternal age at birth (years)a (18.2 – 41.7) (24.7 – 41.7) (18.0 – 42.0) (21.0 – 39.0) 35.0 [29.4 – 41.8] 37.0 [33.8 – 40.5] 35.0 [30.0 – 38.0] 30.0 [28.0 – 38.0] Paternal age at birth (years)b (25.0 – 43.7) (23.6 – 62.8) (20.0 – 51.0) (26.0 – 44.0) Developmental disorders/delaysc ID 0 0 17 (40) n/a Motor delay 2 (14) 5 (12) 0 0 Language delay 3 (21) 2 (5) 0 0 Global developmental delay 5 (36) 0 5 (12) 0 Other/Unable to assess 0 1 (2) 8 (19) 0 Severity of ASD symptomsd Mild or mild/moderate 5 (38) - 7 (44) - Moderate 1 (8) - 4 (25) - Moderate/severe or severe 7 (54) - 5 (31) - History of seizures/epilepsyc 2 (14) 0 5 (12) 0 Low muscle tonec 3 (21) 2 (5) 14 (33) 1 (11) Chronic gastrointestinal disordersc 10 (71) 9 (21) 7 (17) 0 Median [IQR] (range) or N (%) of total a Maternal age at birth was available for 41/42 HEU children without ASD and 9/49 HUU without ASD b Paternal age at birth was available for 11/14 HEU with ASD, 38/42 HEU without ASD, 41/42 HUU with ASD, and 9/49 HUU without ASD c Developmental history, muscle tone, seizure, and gastrointestinal data were available for 9/49 HUU children without ASD d Severity of ASD symptoms was available for 13/14 and 16/42 HEU and HUU children with ASD, respectively 70

We also explored potential relationships between mtDNA content and type/duration of maternal cART exposure (Table 15). 3TC was the most commonly administered drug within our study group, with 44/54 (81%) children with known maternal cART status receiving prenatal 3TC exposure at some point in pregnancy. The next most common exposure in pregnancy was AZT at

33/54 (61%). Other commonly prescribed drugs were NFV (21/54, 39%), LPV/r (17/54, 31%), and ABC (14/54, 26%). Other cART drugs were prescribed approximately 10% of the time or less.

Of the HEU children for whom maternal cART information was available (n=54/56, 96%), a total of 49 (90%) received perinatal exposure. Nearly all (44/49, 90%) of these children received exposure to PI-containing cART at some point in pregnancy, while 6/49 (12%) received exposure to NNRTI-based cART (NVP used in all six cases), three of whom also received exposure to PIs. As the use of PIs has been associated with premature cellular senescence and oxidative stress, which may be exacerbated by RTV-boosting [190], we examined the association between mtDNA content and exposure to RTV-boosted regimens. There were no significant differences between children on RTV-boosted PIs (n=23, mean mtDNA content ± SD: 123 ± 47) and children on non-boosted PI regimens (n=21, mean ± SD: 131 ± 50, p=0.6 via Student’s t- test).

Further analyses of these parameters in a multivariable model are described in the following section (5.3.1.2, Table 18).

71

Table 15. Maternal cART exposure and mtDNA content for HEUs with ASD (n=14) and non-ASD HEU matches (n=42)

HEU with ASD HEU without ASD, match 1 HEU without ASD, match 2 HEU without ASD, match 3 Maternal cART MtDNA Maternal cART MtDNA Maternal cART MtDNA Maternal cART MtDNA (weeks of exposure) content (weeks of exposure) content (weeks of exposure) content (weeks of exposure) content AZT, 3TC, LPV/r (5) AZT, 3TC, NFV (11) 164 ABC, AZT, 3TC, LPV/r (38) 118 107 AZT, 3TC, LPV/r (22) 119 FTC, TDF, LPV/r (11) AZT, 3TC (7) 175 AZT, 3TC, NFV (35) 40 ABC, 3TC, ATV (39) 193 ABC, 3TC, LPV/r (40) 90 AZT, 3TC, LPV/r (33) TDF, FTC, LPV/r (38) 171 AZT, 3TC, LPV/r (38) 136 AZT, 3TC, NFV (13) 91 84 ABC, 3TC, LPV/r (6) AZT, 3TC (16) D4T, 3TC, IDV, RTV (n/a) AZT, 3TC, NFV (37) 70 154 151 AZT, 3TC, LPV/r (39) 81 AZT, 3TC, NFV (24) AZT, 3TC, NFV (9) None taken 162 Unknown 165 DDI, NVP, NFV (34) 72 None taken 93 None taken 195 AZT, 3TC, NFV (15) 75 AZT, 3TC, NFV (15) 110 AZT, 3TC, NFV (19) 115 AZT, 3TC, NFV (n/a) 147 AZT, 3TC, NFV (12) 196 D4T, 3TC, NFV (39) 105 AZT, 3TC, NFV (20) 133 ABC, 3TC, LPV/r (16) LPV/r only (13) 154 AZT, 3TC, NVP (41) 154 ABC, 3TC, NFV (39) 115 116 AZT, 3TC, NVP (8) D4T, 3TC, NVP, LPV/r AZT, 3TC, LPV/r (10) 148 ABC, 3TC, ATV (40) 202 AZT, 3TC, NFV (21) 175 148 (31) FTC, TDF, ATV, RTV TDF, FTC, ATV (10) 160 AZT, 3TC, LPV/r (22) 42 AZT, 3TC, LPV/r (33) 118 93 (33) ABC, AZT, 3TC, LPV/r None taken 48 D4T, 3TC, NVP (40) 126 SQV, IDV, RTV (38) 170 85 (33) ABC, 3TC, LPV/r (30) AZT, 3TC, NFV (12) AZT, 3TC, NVP (34) 182 AZT, 3TC, NFV (21) 220 188 135 TDF, 3TC, NFV (2) ABC, TDF, NFV (3) ABC, 3TC, LPV/r (12) 181 ABC, AZT, 3TC (10) 111 TDF, 3TC, TPV, RTV (40) 217 AZT, 3TC, NFV (16) 89 AZT, 3TC, LPV/r (26) AZT, 3TC, NFV (4) Unknown 215 ABC, 3TC, ATV, RTV (37) 20 96 None taken 109 ABC, 3TC, ATV, RTV (35) Bold font: Child appeared to be conceived on cART (date of cART initiation was on or before date used to calculate gestational age)

72

5.1.3.2 Multivariable linear regression

In a multivariable analysis of all participants for whom demographic information was available

(n=106; HEU with ASD=14, HEU without ASD=41, HUU with ASD=42, HUU without ASD=

9), the relationships between mtDNA content and both ASD diagnosis and HEU status were examined after controlling for demographic factors (age, sex, and ethnicity), as well as maternal age. As paternal age was found to be significantly collinear with maternal age (Pearson’s r=0.6, p<0.0001), it was not included as a variable. The R2 value of this model was relatively low

(R2=0.132), indicating that these factors only explain a small portion of the variance.

Nevertheless, both ASD diagnosis and HEU status were significantly and independently associated with elevated mtDNA content (p=0.01 and p=0.03, respectively). Removing non- significant variables and leaving only HEU status and ASD diagnosis in the model did not improve the R2 value (R2=0.117), and did not alter the p-values of ASD or HEU status substantially (p=0.013 and p=0.0004, respectively).

The full list of variables in this model as well as their respective effect sizes on mtDNA content and p-values are shown in Table 16.

73

Table 16. Multivariable linear regression analysis between mtDNA content and potential explanatory variables for 106 study participants

Effect size Explanatory variables P-value (95% CI)

ASD diagnosis (vs. non-ASD) 24.5 (5.7 – 43.3) 0.01

HEU status (vs. HUU) 30.4 (3.8 – 57.0) 0.03

Male sex (vs. female) 6.4 (-10.6 – 23.4) 0.46

Ethnicity (vs. Caucasian)

African-Canadian 5.8 (-18.9 – 30.5) 0.64

Asian/Other 3.3 (-19.5 – 26.2) 0.77

Child’s age at sample collection, years 0.03 (-2.4 – 2.5) 0.98

Maternal age at childbirth, years 0.7 (-0.8 – 2.1) 0.37

A second multivariable model was developed to examine the association between mtDNA content and various pre-identified maternal and non-maternal factors in ASD children. The variables included were selected on the basis of their relevance to ASD (section 1.4.3) as well as the availability of these data within the CARMA and ASPIRE databases. For data that were not available as part of the CARMA database, physicians and research staff at the relevant research sites were asked to provide qualitative or anecdotal information whenever possible. A total of 52 children with ASD (n=14 HEU and n=38 HUU) were included in the model. Table 17 provides the full list of variables and analyses.

74

Table 17. Multivariable linear regression analysis between mtDNA content and potential predictor variables for participants with ASD (n=52)

Effect size Explanatory variables P-value (95% CI)

HEU status (vs. HUU) 53.7 (25.5 – 81.8) <0.0001

Age at sample collection, years -1.3 (-1.9 – 4.5) 0.42

Developmental disorder/delay (vs. no delay) 3.5 (-18.0 – 24.9) 0.75 (ID, language, motor, global) Chronic gastrointestinal abnormalities (vs. no recorded abnormalities) -17.8 (-46.0 – 10.5) 0.21 (constipation, diarrhoea, reflux, etc.) Maternal age at childbirth, years -0.09 (-1.9 – 1.8) 0.93 History of seizures and/or epilepsy (vs. no -5.0 (-38.1 – 28.2) 0.77 history)

Hypotonia (vs. normal muscle tone) 1.1 (-22.9 – 25.1) 0.93

The R2 value of this model was moderate (R2=0.312), but in agreement with other univariate and multivariable analyses, HEU status was independently and significantly associated with elevated mtDNA content (p<0.0001). None of the other factors showed an association. Additionally, there were no significant mtDNA content differences between groups when stratified based on these categorical variables and univariately examined via Student’s t-test (p=0.38 to p=0.85). There were also no significant correlations between mtDNA content and maternal or child age (p=0.28 via Spearman’s and p=0.98 via Pearson’s, respectively).

75

Finally, to further examine the effects of maternal cART type, duration, and initiation time on mtDNA content, a third multivariable model was developed that included all HEU participants with known exposure to maternal cART in pregnancy (n=49). The R2 of this model was 0.195.

As work from our lab and others has demonstrated that different NRTI regimens affect mtDNA in different ways, we explored any association with particular NRTI backbones. We also examined whether exposure time and the trimester at cART initiation showed any association with mtDNA content. PI exposure was included due to its association with markers of oxidative stress, as previously outlined. ASD diagnosis was also included to adjust results for any independent effects of this condition.

The full list of explanatory variables included is shown in Table 18, along with effect sizes and p-values.

76

Table 18. Multivariable linear regression analysis between mtDNA content and potential predictor variables for HEU participants with known perinatal cART exposure (n=49)

Unadjusted effect Unadjusted Adjusted effect size Adjusted Explanatory variables size (95% CI) p-valuea (95% CI) p-value ASD diagnosis (vs. no 32.4 (0.1 – 64.7) 0.03 16.2 (-36.1 – 68.4) 0.53 ASD) Duration of in utero -0.1 (-0.4 – 0.2) 0.89 1.1 (-2.8 – 4.9) 0.33 cART exposure, weeks NRTI taken in pregnancyb AZT (vs. no AZT) -9.0 (-36.5 – 18.5) 0.51 -21.8 (-81.9 – 38.3) 0.47 ABC (vs. no ABC) -10.4 (-42.8 – 22.0) 0.52 -32.9 (-99.9 – 34.2) 0.33 TDF (vs. no TDF) 22.3 (-22.0 – 66.7) 0.31 -0.35 (-88.2 – 87.5) 0.99 ddI or d4T (vs. no ddI -9.9 (-54.6 – 34.9) 0.66 -37.6 (-127.4– 52.2) 0.40 nor d4T) Trimester of cART initiation (vs. those who conceived on cART)c First trimester 20.2 (-30.0 – 70.4) 0.41 21.4 (-37.5 – 80.4) 0.46 Second trimester 13.7 (-47.9 – 75.4) 0.64 37.4 (-47.8 – 122.5) 0.38 Third trimester 40.1 (-38.0 – 118.2) 0.28 59.7 (-52.7 – 172.1) 0.29 Exposure to boosted PIs and NNRTIs (vs. unboosted PIs) Boosted PIs -6.7 (-36.8 – 23.4) 0.65 0.62 (-39.8 – 41.1) 0.98 NNRTIs 25.4 (-28.7 – 79.5) 0.34 29.8 (-36.4 – 95.9) 0.37 a P-values obtained via Student’s t-test, Mann-Whitney U test, or Pearson product-moment correlation test b Children with exposure to two regimens (n=9) were categorized according to the regimen of longer duration c First trimester defined as <14 weeks of gestation, second as >14 but <28, and third as 28 weeks to birth, as per standards of the American Congress of Obstetricians and Gynecologists

No significant mtDNA content differences were observed with regards to type, duration, or initiation time of maternal cART, which was recapitulated in univariate analyses. There was no correlation between mtDNA content and in utero exposure time (Figure 12). 77

Figure 12. Correlation between mtDNA content and duration of in utero maternal cART exposure in HEU children (n=52)

5.2 MtDNA AOD

In order to compare potential differences in mtDNA quality (apparent oxidative damage) between groups, we measured the amplification efficiency of a long-range PCR polymerase in each sample. Theoretically, the polymerase’s progression is slowed by the presence of oxidative lesions, so by comparing each sample’s efficiency of amplification to that of an undamaged IC, it is possible to quantify the extent of ROS-induced damage for each group. Knowing that the final product should encompass half the mtDNA genome (approximately 8.5 to 9kb), by comparing

78

the size of long PCR products on an agarose gel we could also visualize any large insertions or deletions which may be present in the original template.

5.2.1 Quality control

This assay was relatively new in the lab and a stock of “undamaged” or Happy DNA had to be identified for normalization purposes across runs. Three candidate extracts were evaluated, as described in section 4.2. The performance of each candidate was monitored for consistency over the first six long PCR runs before we decided which of these candidates would be used to normalize results. In total, we evaluated n=12 D-loop quantifications per Happy IC, representing n=6 long PCR runs. As seen in Table 19, IC 1 was the most consistent in terms of inter-run CV, both for pre- and post-long PCR copy number. Therefore, IC 1 was the only Happy IC included in the seventh long PCR. Accordingly, the amplifications of all samples were normalized to IC 1 in order to calculate mtDNA AOD (n=20 D-loop runs in total). Based on our quality control parameters (outlined in section 4.2), two samples from the HUU with ASD group were excluded from analyses, as the amplifications of these samples were not within ±2 SD of the mean across two independent long PCR runs. The average absolute % difference between D-loop copy numbers of long PCR replicates was 19.2%

A selection of samples with the highest mtDNA AOD (n=2) and the lowest (n=2) from each study group were qualitatively examined on a 0.8% agarose gel. This experiment revealed no significant insertions or deletions, as all bands were of similar size (Figure 13). There were no strong differences in band intensity between groups, apart from a single HUU control (lane 8).

These findings are discussed further in section 6.1.2.1. 79

Figure 13. Agarose gel electrophoresis (0.8%) of n=16 samples with quantitatively high (groups A and C) and quantitatively low (groups B and D) mtDNA AOD 80

Table 19. Quality control parameters for the three Happy IC candidates used across 12 D- loop runs, corresponding to six long PCR runs

Happy IC 1 Happy IC 2 Happy IC 3 Pre-long PCR D-loop copies (x103) 30.0 ± 2.5 49.7 ± 15.0 11.2 ± 5.7 (copy number) (25.2 – 35.0) (16.5 – 72.7) (2.9 – 21.7) CV (%) 8.3 30.2 51.5 Post-long PCR D-loop copies (x106) 3.7 ± 1.1 5.3 ± 1.9 0.4 ± 0.2 (copy number) (2.0 – 5.0) (2.2 – 8.5) (0.1 – 0.9) CV (%) 28.7 36.5 67.3 Ratio of post-long PCR copies to pre- 41.1 ± 12.6 37.5 ± 12.6 12.4 ± 7.5 long PCR copies (x103) (21.3 – 57.2) (13.4 – 57.4) (3.8 – 28.3) CV (%) 30.7 33.6 60.5 Mean ± SD (range)

One of the seven long PCR runs was repeated since most samples on the initial assay did not amplify, which was confirmed by gel electrophoresis. One D-loop plate failed to meet our quality control parameters and was repeated. Of the D-loop runs that passed quality control, the inter-assay CV of IC High was 21.0% (n=20); the intra-run CV was 8.9% (n=10). The CV between a series of equivalent dilutions (n=10) of the same random long PCR product assayed in duplicate was 18.1%. Happy, Medium, and Unhappy controls appropriately reflected a stepwise decrease in post-long/pre-long PCR copy number ratio, as shown in Table 20.

Table 20. Rate of amplification of three ICs used in seven long PCR runs

Happy IC 1 Medium IC Unhappy IC

n=20 n=7 n=7 Rate of amplification (ratio of post-long PCR copies 39.0 ± 12.6 21.7 ± 9.8 1.3 ± 0.7 to pre-long PCR copies, x103) (Mean ± SD) CV (%) 29.3 45.2 52.0

81

5.2.2 Statistical analyses

Absolute and IC-relative measurements of amplification ratios, as well as the resulting mtDNA

AOD averages for each group are shown in Table 21. The coefficient used to normalize all AOD measurements in the study to values between 0 and 1 was 0.56. There was a total of n=44 samples in the study which amplified more efficiently than the Happy IC: of these, n=1 was an

HEU child with ASD, n=9 HEU children without ASD, n=22 HUU children with ASD, and n=12 HUU non-ASD controls.

Table 21. Absolute amplification, amplification relative to an undamaged IC, and AOD of mtDNA for HEU and HUU children with and without ASD

HEU HEU HUU HUU with ASD without ASD with ASD without ASD n=14 n=42 n=40 n=51 Rate of amplification (ratio of post-long PCR 31.9 ± 10.4 28.8 ± 8.8 35.4 ± 12.4 32.3 ± 9.8 copies to pre-long PCR (16.5 – 47.5) (11.1 – 55.6) (10.3 – 62.2) (9.1 – 48.1) copies, x103)

Rate of amplification 0.80 ± 0.15 0.82 ± 0.35 0.95 ± 0.33 0.80 ± 0.28 relative to Happy IC 1 (0.57 – 1.19) (0.23 – 1.77) (0.18 – 1.63) (0.25 – 1.41)

0.55 ± 0.09 0.53 ± 0.20 0.47 ± 0.18 0.55 ± 0.16 MtDNA AODa (0.33 – 0.68) (0.00 – 0.87) (0.08 – 0.90) (0.20 - 0.86)

Mean ± SD (range) a AOD values expressed after adjustments were applied to ensure values ranged between 0 and 1

There were no observed differences in mtDNA AOD between groups (p=0.1 to p=0.9), as shown by the boxplots in Figure 14.

82

Figure 14. Univariate between-group comparisons between mtDNA AOD of HEU and

HUU children with and without ASD. P-values from Student’s t-test are shown

There was no evidence of a relationship between mtDNA content and mtDNA AOD within any of our four study groups, or among all study participants as a whole. As seen in Figure 15 A-E, correlation coefficients ranged from 0.02 (Spearman’s) to 0.3 (Pearson), with none of the relationships reaching statistical significance (p-values ranging from p=0.07 to p=0.9).

83

Figure 15. Correlations between mtDNA content and relative mtDNA AOD in HEU children with ASD (A), HEU children without ASD (B), HUU children with ASD (C), HUU children without ASD (D), and all participants (E). R2 and either Spearman’s ρ or

Pearson’s r are shown, as appropriate

84

5.3 Sensitivity analyses

5.3.1 Ethnicity

To inform us whether the unknown ethnicity of anonymous controls, or ethnicity mismatches between CARMA and ASPIRE samples were likely to contribute to our main findings, we performed a sensitivity analysis and compared differences in mtDNA content between ethnic groups.

Matched groups were well with respect to sex. Although the age range was similar between ethnicities (paired Student’s t-tests p-values ranged from p=0.98 to p=1.0), overall, the average age of children included in this sensitivity analysis was significantly younger than that of the main study (Table 22).

Table 22. Distributions of age, sex, and mtDNA content for three matched groups used in a sensitivity analysis of possible relationship between mtDNA content and ethnicity

Asian & South African- Caucasian Asian Canadian n=20 n=10 n=20

Male sex 5 (50) 10 (50) 10 (50) n (%)

Median age, years 1.3 [0.6 – 2.5] 1.3 [0.5 – 2.6] 1.3 [0.5 – 2.6] [Q1 – Q3] (range) (1.1 – 15.1) (0.2 – 15.7) (0.3 – 15.6)

MtDNA content 147 ± 45 153 ± 60 168 ± 54 mean ± SD (range) (90 – 203) (54 – 259) (62 – 268)

Absolute % difference between ages of 10.3 ± 10.7 8.2 ± 7.7 matched participants (vs. Asian) - (0.3 – 38.6) (0.3 – 22.8) mean ± SD (range)

85

There were no significant mtDNA content differences observed between any of the three groups

(p=0.3 to 0.8, Figure 16). We also performed sample size calculations using these measurements

(power=0.80 p<0.05), and found that between Caucasian and Asian/South Asian children (effect size=6, common SD=55), a sample size of n=1233 would provide 80% power to detect a difference between these two ethnicities. Differences between African-Canadian children and

Caucasians (effect size=15, common SD=57) as well as Asian/South Asian children (effect size=21, common SD=51) would require group sizes of n=215 and n=89 to reach 80% power, respectively.

Figure 16. Univariate between-group comparisons between three sex- and age-matched

(1:2) groups of HEU children without ASD. P-values calculated via Student’s t-test 86

5.3.2 Albumin

We correlated the mtDNA content of all samples used in the study with their respective albumin copy numbers to ascertain whether between-group differences could be confounded by systematic differences in DNA extract concentration. Across all samples that passed quality control (n=147), a relatively weak but highly significant (p=0.001) correlation was observed between mtDNA content and albumin copy number which reflects the DNA concentration of the extract (Figure 17).

Figure 17. Correlation between albumin copy number and mtDNA content for all samples in the study. R2 and Spearman’s ρ are shown 87

Between-group comparisons revealed that there were significant (p<0.0001) differences between groups in terms of albumin count, implying that our findings could be biased by differences in

DNA extract concentration. To account for this, we re-diluted all samples in our study that passed initial mtDNA content quality checks (n=147) to a common albumin concentration, between 5000 and 7000 copies per 2uL on average (mean ± SD: 5906 ± 806, range 3872 to

9551). We then re-assayed all samples using the multiplex assay, following the exact same procedures described in Chapter 3. The results are shown in Figure 18.

Figure 18. MtDNA content measurements after diluting all samples to approximately 5000 copies of albumin

88

MtDNA content was compared between the original runs and the runs performed after diluting samples to a common concentration. A poor coefficient of correlation or a large systematic shift in mtDNA content (i.e. a slope significantly different from 1 when plotted via linear regression) would be indicative of the measurement’s dependence on sample concentration. Original and repeat measurements were highly conserved, as shown in Figure 19. The slope of the regression line (1.03) also indicates that there was a minimal amount of systematic shift in the standard curve between runs. Therefore, the original values of the initial dataset were kept.

Figure 19. Correlation between mtDNA content measurements before and after dilution to a common concentration. R2 and Pearson’s r are shown 89

5.3.3 Platelet count

Correlations between platelet count and mtDNA were measured in the two HEU groups to assess whether higher platelets count (platelets contain mtDNA but not nDNA) could predict higher mtDNA content. Platelet counts from clinical laboratory tests were obtained from the CARMA database. Values were available for all HEU participants in the study except for two HEU children with ASD, but were not available for children from ASPIRE or HUU anonymous controls. No significant relationships were observed between platelet count and mtDNA content among HEU children, as a whole, with ASD, or without ASD (Figure 20 A-C). These findings suggest that the mtDNA content differences observed between groups were unlikely to be confounded by excess mtDNA derived from platelets, but were likely related to differences in leukocyte mtDNA.

90

Figure 20. Correlation between leukocyte mtDNA content and platelet count in HEU children with ASD (A), HEU children without ASD (B), and all HEU children in the study

(C). R2 and either Spearman’s ρ or Pearson’s r are shown, as appropriate

91

Chapter 6: Discussion and conclusions

6.1 Interpretation of results

We have shown that HEU children with ASD have significantly elevated leukocyte mtDNA content compared to non-ASD children (HEU and HUU), as well as ASD children with no

HIV/cART exposure. These results could be indicative of mitochondrial dysfunction, and may help explain the increased prevalence of ASD observed in HEU children within CARMA. The performance of the mtDNA content and mtDNA AOD assays are discussed below, as well as the relevant findings from our case-control study.

6.1.1 MtDNA content measurements

6.1.1.1 Quality control

Inter- and intra-assay CVs using the multiplex method (both ~5%) were significantly lower than values reported by other groups using a monoplex method (10-20%) [191,192]. Results were highly reproducible (R2=0.93) and melting curve analyses showed minimal production of nonspecific products, demonstrating the robustness of this technique when used to quantify products with a high degree of Tm separation (approximately 14ºC in this case, see Table 5).

Samples with extreme values (low and high) had the largest residuals when plotting the correlation between original and repeat measurements, as these samples were on more extreme ends of the standard curve where quantification is less stable between runs. This finding highlights the need for rigorous quality control metrics, especially when applied toward samples with very low nDNA or mtDNA concentrations.

92

Overall, the multiplex technique features several advantages over traditional monoplex methods, including higher throughput, a reduced level of complexity, and more reproducible results. This method also demands fewer reagents and thus is significantly more cost-effective than monoplex techniques, owed to the necessity of only a single master mix as opposed to two. We are in the process of validating this assay for use in other tissue types, with preliminary results appearing promising.

6.1.1.2 Primary outcomes and predictors

6.1.1.2.1 HEU status

We observed that HEU status, irrespective of ASD diagnosis, was associated with elevated leukocyte mtDNA content compared to controls. This relationship was significant in both univariate and multivariable analyses after controlling for age, sex, ethnicity, and maternal age.

As mentioned in section 1.5, these results are consistent with some previous literature, published both by our lab [6] and other groups (McComsey et al., Ross et al.) [102,183], but they are inconsistent with other reports of decreased mtDNA content in HEU children (Aldrovandi et al.,

Poirier et al.) [7,182]. However, several caveats must be made with regards to the discrepant findings reported by the Aldrovandi and Poirier groups. Firstly, these are both older studies

(published in 2009 and 2003 respectively), and as such the landscape of HIV treatment and prevention was much different than that for both our cohort as well as those of the McComsey and Ross groups (2015 and 2012 respectively). For example, under the 2002 American perinatal treatment guidelines, pregnant women living with HIV who were not already receiving cART were recommended treatment with five daily doses of AZT only (100mg), whereas by current standards the treatment most commonly used would be two daily doses of AZT/3TC 93

(300mg/150mg respectively) in combination with a boosted PI. This is a critical deliberation, as people on monotherapy are significantly more vulnerable to developing drug-resistant HIV than those on dual or triple therapy regimens [193,194], meaning that children exposed only to AZT may have had increased exposure to resistant strains of maternal HIV. Resistance is especially important when considering the shift from five doses to only two, since incomplete adherence to antiretrovirals, even in the range of 10-30% of doses missed, is a significant risk factor for developing resistant HIV strains [195]. As the virus has been associated with mtDNA depletion independently of cART treatment [196], this could be a plausible explanation for why the Poirier group observed decreased mtDNA content in children exposed to AZT monotherapy. This interpretation is also corroborated by the Aldrovandi group’s observations that exposure to a greater number of antiretrovirals in HEUs was associated with progressive increases in mtDNA content [182].

Additionally, Aldrovandi et al. report that their control group was comprised of children from a different cohort than the HEUs, with different blood collection protocols. EDTA was used as the anticoagulant agent for control samples while heparin or acid-citrate-dextrose solution B was used for HEU participants, and no evidence was provided that results were not affected by this discrepancy. Other potential differences between the control group and treatment group which could be attributed to the fact that they come from two different cohorts are not discussed. These two factors could bias the study results. Moreover, they found that by 2 to 5 years after birth, mtDNA content in HEUs increased to levels close to those of controls. This suggests that after toxic strains are removed there may be a physiological rebound effect in mtDNA, which would

94

present at least a partial agreement with our interpretations that elevated mtDNA content may be a physiological response to mitochondrial pressure exerted by maternal HIV and/or cART.

Another possible explanation for elevated mtDNA content in HEUs is that certain drugs may have differential effects on mtDNA content, which we were not powered to explore. For example, there is evidence that the dideoxy NRTIs d4T and ddI can significantly decrease mtDNA content in HIV patients and in cell culture models [197–199], and although our multivariable analyses showed no significant differences, our numbers were small (Table 18).

Only 5/56 (9%) HEU children in the study were exposed to d4T or ddI in pregnancy, compared to 26/56 (46%) with AZT exposure, which has considerably milder side effects (see Table 15 for more detail). Additionally, some of these drugs are more effective at crossing the placenta and blood-brain barrier than others (Table 3), such as AZT, ABC, and d4T, which are able to cross both with moderate to high efficiency. Perinatal exposure to certain drug combinations may therefore exert differential effects on mtDNA content compared to those containing drugs that are not as effective at crossing these barriers. Reproduction of these results in another cohort with more power to examine differences in maternal cART exposure will be key to understanding the exploratory results presented here.

The timing of cART initiation did not associate with differences in mtDNA content, although sample sizes were small for certain groups (n=5 for those conceived on cART, n=6 for those exposed starting in the 3rd trimester). Intuitively, we expect that children whose mothers started treatment early or conceived on these drugs may be more at risk to develop ASD, as the effects of the drugs themselves and/or maternal inflammation and oxidative stress may play a role in 95

adverse outcomes during this critical stage of fetal neurodevelopment. But, in a study by Brogly et al [200] examining over 1000 HEU children, it was demonstrated that the risk of mitochondrial dysfunction was three to four times higher in children who received first exposure to 3TC or AZT/3TC therapy during the third trimester compared to HEU children whose mothers initiated therapy earlier in pregnancy. Other studies in HEUs have demonstrated an association between later maternal cART initiation and lactic acidemia [201] and higher venous lactate levels [202]. Taken together, these findings suggest that exposure to cART during some critical period of fetal development may be an important consideration in studies of this nature.

6.1.1.2.2 ASD diagnosis

Another important finding in this study was that ASD status was independently associated with elevated leukocyte mtDNA content, consistent with literature published by other groups

[184,203]. The fact that HEU children with ASD had the highest mtDNA content of all four groups suggests the possibility of a cumulative relationship between HEU status and ASD in terms of their effects on mtDNA content. This may signify that children showing signs of mitochondrial dysfunction, including but not limited to significant changes in mtDNA content, may be preferentially predisposed toward a diagnosis of ASD. These findings may also indicate that ASD, which in the literature has been associated with signs of chronic oxidative stress [185] and mitochondrial dysfunction [10,204], exerts (or is exerted by) a persistent deleterious effect on mitochondria, whereas non-ASD HEU children are presumably relieved of this mitochondrial pressure after the perinatal exposure period (i.e. shortly after cessation of AZT prophylaxis). This would support our working explanation that elevated mtDNA content may be a physiological response mechanism to mitochondrial dysfunction. 96

6.1.1.2.3 Other variables of interest

There was no association observed between mtDNA content and any other variable included in our multivariable analyses. However, the relatively low R2 values of these models (0.1 to 0.3) suggest that there are likely other factors that are at play which could not be considered here.

Animal and human studies have shown that offspring of mothers who smoked during pregnancy have altered mtDNA content compared to those born to mothers who did not smoke, which this study was not powered to address [6,205]. Information pertaining to other maternal factors which may influence mtDNA content as well as rates of ASD in HEU children, such as infections contracted during pregnancy, medications taken during the perinatal period, and environmental factors such as smoke exposure were either not available or were heterogeneously reported.

Qualitatively, there appeared to be no differences in the prevalence of maternal smoking and drug use between the CARMA and ASPIRE groups (where this information was available), but the potential effects of these factors on mtDNA content could not be assessed in this study.

A sensitivity analysis for ethnicity (section 5.3.1) did not demonstrate significant mtDNA content differences between children of different ethnic origins (p=0.3 to p=0.8). Sample size calculations using these measurements revealed that group sizes between n=89 and n=1233, well beyond numbers used in our study, would be required in order to reach 80% power to detect differences between groups. These results suggest that any underlying differences, if any, which could be accounted for by ethnic mismatch are not highly pronounced. Therefore, it is highly unlikely that either the ethnic mismatch between CARMA and ASPIRE children, or the

97

unknown ethnicities of the HUU anonymous controls, could contribute meaningfully to the observed differences in the main arm of the study.

Factors that were ASD-related (Table 17) did not appear to be significant predictors of mtDNA content. However, any effects could be masked by the small sample size of our group of interest, compounded with the fact that some of this information (e.g. developmental delay) may not have been easily accessible to the CARMA team. As the cohort was not designed to examine factors relevant to ASD, it is also possible that information such as chronic GI conditions may not have been assiduously recorded for non-ASD children. This also applies diagnoses of ID, of which none were noted within either of the HEU groups (Table 14) despite its strong association with

ASD. Obtaining this information retroactively would not only be difficult from a logistical point of view, but may also introduce bias. We note however, that qualitative assessments of functional ability made by CARMA physicians as well as cognitive test scores we were able to find (section 5.1.3.1) suggest that there are very likely some cases of ID among the HEU with

ASD group.

As was mentioned in section 6.1.1.2.1, type and duration of maternal cART exposure may have an effect on mtDNA content, as the literature has demonstrated that HEU infants with cART exposure have elevated mtDNA content compared to HEU children with no exposure [182], and that different individual drugs exert more toxic effects on mtDNA than others [197–199]. While an attempt was made to examine these factors (Table 15, Table 18), the experimental design and sample sizes used in this exploratory study were not suitable to accommodate well-powered, formal examinations of these parameters. Qualitatively, no obvious differences were seen 98

between HEU children with and without ASD in terms of maternal regimens taken, although

ASD children on average tended to have less maternal exposure to cART than non-ASD children

(mean 11 weeks of in utero exposure for ASD children contra 33 for non-ASD). While this could implicate increased perinatalHIV “milieu” exposure as an explanatory factor for the increased rates of ASD in HEU children, this would not explain why only 1.4% of CARMA’s HIV+ children were diagnosed with ASD. Alternatively, this could suggest that initiating cART later in pregnancy may increase the risk of ASD, something that should be further explored in another cohort. More work is needed to elucidate any potential specific associations with different maternal drug combinations, initiation times, and adherence parameters.

6.1.2 MtDNA AOD measurements

6.1.2.1 Quality control

Potential optimizations and quality control parameters for the mtDNA AOD assay were explored, as the performance of this assay using biological samples has not been well established to date. We examined the robustness of results produced by this assay at the long PCR, qPCR, and 1:500 dilution steps (Figure 9) by monitoring the behaviour of various ICs and quantifying inter- and intra-run variability of measurements at these various steps.

While the intra-run and inter-run CVs of the High ICs used in the monoplex D-loop step were higher (9% and 21%, respectively) than those reported for the mtDNA content multiplex assay

(5% each, Table 12), these differences are not dissimilar to numbers reported by other groups.

There was a moderate amount of variation (19%) inherent to the 1:500 dilution step, likely owing to a combination of the variability of the p1000 pipette itself, intra-run variability during D-loop 99

quantification, and delivery technique. The variability of pre-long PCR copy numbers in these controls is of concern, as only IC 1 displayed a CV similar to that of the High IC (8%). The variability observed for ICs 2 and 3 (30 and 50%, respectively) could be suggestive of freeze- thaw effects, as controls were stored at -20 °C but thawed to room temperature before assaying.

As the assays took place over the course of several weeks, the effects of long-term freezing and thawing of samples on long PCR mtDNA amplification should be examined.

Post-long PCR copy numbers for ICs were also highly variable between runs, which could be a result of the long-range PCR enzyme used in this assay. As the enzyme is suspended in a heavy concentration of glycerol, this component of the master mix is highly viscous. It is possible that within the master mix tube there was a density gradient formed, and the enzyme could have been more concentrated near the bottom of the tube. Although efforts were made to ensure the master mix was well vortexed and centrifuged prior to assaying, the time to deliver this master mix to sample tubes may have allowed for the formation of such a gradient. And, although the enzyme and enzyme-containing master mix were kept on ice for as long as possible during reaction setup, some of the observed variability may be due to inactivation or over-activation of the heat- sensitive enzyme during delivery into sample tubes.

In order to confirm long PCR product specificity and visualize the relative amplification of templates with high AOD and low AOD in each group (Figure 13), we used agarose gel electrophoresis. There was no observable evidence of primer dimer or non-specific product formation in any of the 16 templates used. Also, we did note any systematic differences in amplicon sizes or band intensity, apart from one low-damage HUU non-ASD sample that 100

amplified more efficiently than the undamaged control. This finding was expected, as this sample did not pass quality controls checks during the mtDNA content assay, and thus we did not have an accurate measure of the true D-loop copy number in this sample. We found that during monoplex D-loop quantification, this sample had a higher pre-long PCR copy number than other samples, resulting in an exponentially higher post-long PCR D-loop copy number and thus the overloading observed on the gel. However, as the sample passed all quality control checks for the mtDNA AOD assay, we had no grounds to reject it from this arm of the study.

Other samples had roughly equivalent band intensities, suggesting that D-loop copy numbers

(and correspondingly, long PCR efficiency) in low-damage and high-damage samples were too similar to visualize differences in this manner, especially before normalization to initial template copy number.

6.1.2.2 Primary outcomes

Although we did not observe any significant differences in leukocyte mtDNA AOD between groups, any differences could be masked by high leukocyte turnover and physiological repair and destruction mechanisms selecting against highly damaged mtDNA. However, the variability of this assay between and within runs suggests that it is not suitable at this time to examine potentially subtle differences in blood. Nonetheless, this work has highlighted important considerations in terms of IC selection, quality control parameters, and interpretations, which may inform future studies wishing to optimize this assay for use with biological specimens.

No correlation was observed between mtDNA content and AOD within any of the groups, or among samples as a whole. While this may be attributed at least in part to a high degree of 101

variability inherent to the assay, there are other possible explanations for these results. The differences in blood may be too subtle to detect with an assay designed to measure differences spanning a much broader range of values, resulting from a greater amount of oxidative insult than is typically found in biological tissue [206,207]. Furthermore, mtDNA oxidative damage in tissue is mitigated by antioxidants such as SOD2 and catalase (section 1.3.2.3). Since drug pressure in HEU children had long since been removed by the time of blood collection, it is likely that any residual oxidative damage had been repaired, or that highly damaged mitochondria had undergone mitophagy. Thus, any long-standing changes to mtDNA resulting from oxidative stress would likely appear in the form of somatic mutations and/or mtDNA heteroplasmy arising from the clonal expansion of previously damaged mtDNA copies. This is especially true in the context of blood, which is a high-turnover tissue: non-lymphocytic leukocytes circulate in the blood for a matter of hours [208] to an upward of 9 or 10 days [209] depending on subset. Therefore, we would not expect to observe major oxidative lesions in leukocyte mtDNA, except perhaps in the case of chronic ASD-related oxidative stress.

6.2 Limitations and biases

There are several limitations and biases inherently imposed upon this observational study by its cross-sectional, case control nature. Firstly, we are unable to comment on any longitudinal changes that may better elucidate how changes in mtDNA may relate to the progressive development of ASD symptoms, although this was controlled for as best as possible by selecting samples close to the date of ASD diagnosis. This study was also not designed to examine mechanistic pathways of causation that may underlie this apparent relationship, which future studies may be better equipped to address. 102

A possible confounding element of this study stems from the fact that the HUU control group is largely composed of anonymized samples, for which only sex and age data were available. The assumption was made for the purposes of this study that these children do not have ASD and are not HEU. However, given our observations that on average ASD and HEU children had elevated mtDNA content compared to controls, any ASD or HEU cases among this anonymous group would likely present a conservative bias. The sensitivity analyses presented suggest that potential differences in ethnicity, platelet count or leukocyte count are not meaningful confounders of our results.

Several potential recruitment biases inherent to the CARMA cohort must also be considered. As outlined in section 2.2, there was a recruitment bias toward selectively enrolling HEU children with ASD during the second CARMA enrolment phase, which at face value would artificially inflate the prevalence of ASD in HEU children we reported (14/299, 4.7%). However, based on discussions with CARMA staff and physicians, there was also a bias against recruiting ASD children during the first enrolment phase, as the children could become distressed during blood draw. It should be noted that exactly half (7/14) of the HEU children with ASD included in this study were recruited during each enrolment phase. Thus, the selective bias toward ASD in the second enrolment phase was likely counterbalanced by a selective bias against ASD enrolment in the first phase. Notably, in addition to the HEU children with ASD included in this study, we are also aware through personal communications with CARMA staff and investigators of other cases of interest at CARMA sites. Upon retrospective review, at least four cases of ASD in HEU children (born between 1998 and 2006) have been noted at the Oak Tree Clinic in Vancouver, 103

but because HEU children are not routinely followed past 18 months, could not be enrolled in

CARMA. One child was described as relatively high-functioning with a mild case of Asperger’s

Syndrome, and three were noted as more severe cases with a number of other psychiatric conditions such as epilepsy, fetal alcohol syndrome, and ID. Two additional cases of ASD are strongly suspected in CARMA’s HEU children, but are currently undergoing assessment and have not been diagnosed at the time of writing. One of these children is enrolled in Vancouver

(approximately 18 months of age at the time of visit) and the other in Toronto (aged 3.5 years at the time of visit). Finally, one case of ASD was noted in a child born to a mother thought to be at high risk for HIV infection. The mother was given cART treatment during pregnancy as a precaution, but did not test positive for HIV infection.

While successful recruitment rates for CARMA were very high (>90%), there may be a number of inherent selection biases among those who were approached. For example, as the study was not funded to provide translation services only persons who could speak and understand English or French were invited to participate, as speakers of other languages could not be expected to provide proper informed consent. This may present a bias against recent immigrants, which is particularly relevant in the context HIV. A significant proportion of those living with the disease in Canada are immigrants from countries where HIV is endemic [22], many of whom may not speak fluent English or French. This may therefore affect the generalisability of these results to the HIV-infected community at large.

Another limitation of this study is that we were only able to examine differences in whole blood, although intuitively we would expect the most relevant effects to occur in tissues of the central 104

nervous system. One previous study showed altered mitochondrial dynamics and signs of oxidative stress in the temporal lobes of ASD individuals, with no changes observed in mtDNA content or the expression of mitochondrial gene transcription factors [210]. Another study examining patients affected with major depressive disorder, bipolar disorder, and schizophrenia found no significant differences in mtDNA content of post-mortem brain tissue compared to controls, although there was evidence of mitochondrial gene deletions and increased mitochondrial gene expression in these groups [211]. These findings highlight not only the important role of mtDNA in psychiatric conditions, but the importance of examining mitochondrial dynamics and expression patterns outside the realm of mtDNA content.

While we have demonstrated an association between HEU status and elevated mtDNA content, it is unclear whether these results are modulated by the effects of HIV or cART. Limited information was available for the mothers of these children, and thus we have no measures of maternal HIV pVL or CD4 count during the perinatal period. Consequently, we have no way of examining whether the increased prevalence of ASD in HEU children is associated with higher levels of circulating virus and/or maternal inflammatory responses during pregnancy. The lack of information regarding maternal medication is also a confounding factor that we could not account for. It is well-established that affective disorders such as clinical depression are common among those living with HIV, and that depression is a significant risk factor for HIV progression and mortality owed to poor adherence to HIV therapy [212]. Furthermore, the use of certain antidepressants in pregnancy have been associated with teratogenicity [213], which we did not have the data to examine. An exploration of a potential association between ASD and perinatal

105

exposure to antidepressants presents an intriguing next step for subsequent studies in other HEU cohorts.

6.3 Future directions

Long-term effects of HIV and cART exposure in pregnancy on HEUs are a topic of great interest within the HIV caregiver and research community. But, since cART administration in pregnancy is relatively recent (since the mid-1990’s) there is much more research required in this area before we fully understand the range of potential consequences these drugs may have on fetal development. This study was designed to be exploratory and hypothesis generating, given that to our knowledge there has been no prior work examining potential associations between ASD and perinatal HIV/cART exposure.

Future work should focus on identifying other potential factors that may underlie the mtDNA elevation, such as the effects of different maternal cART regimens, cART adherence during pregnancy, the effects of smoking and alcohol, and the timing of cART initiation. While our limited data suggest in a qualitative manner that there were no overt differences between ASD and non-ASD children on these parameters, the data are heterogeneous and incomplete, rendering more formal analyses highly impractical, if not impossible.

Further work is also required in order to assess whether the differences in leukocyte mtDNA content translate to differences in the mitochondria of other tissues such as muscle and brain, an inherent element of uncertainty in our study. We have ultimately theorized that:

106

1) mtDNA content elevation in blood reflects mitochondrial dysfunction in blood cells,

which in turn may reflect mitochondrial dysfunction in neurons at the time of central

nervous system development that is relevant to development of ASD

2) the neuronal mitochondrial dysfunction mediates the abnormal neuronal

development associated with ASD, and finally

3) the neuronal mitochondrial dysfunction is attributable, at least in part, to perinatal

HIV and cART exposure in tandem with other clinical, environmental and/or genetic

risk factors

This theory, which requires confirmation in a prospective evaluation, was developed after mtDNA content was measured. Our initial hypothesis was tested by examining blood mitochondrial DNA content of HIV-exposed ASD patients relative to controls, as this was the only tissue type available for children in this study. The intervening steps in our proposed theory remain uncertain. Examinations of blood and brain tissue in animal models of ASD, which have been previously described [214], may be useful to answer the question of whether changes in leukocyte mtDNA content predict mtDNA damage in brain tissue and neurodevelopmental outcomes.

6.4 Conclusions

Antiretroviral prophylaxis and cART treatment in pregnancy are highly successful at preventing vertical HIV transmission. However, there are concerns regarding long-term health outcomes in

HEU children with perinatal exposure to these drugs, which have well-documented effects on mitochondria. A primary objective of this thesis was to explore the relationship between 107

perinatal HIV/cART exposure with quantitative and qualitative differences in mtDNA, which may in part explain recent and novel findings of increased rates of ASD among these children.

Leukocyte mtDNA content was measured via monochrome, multiplex qPCR between HEU and

HUU children with and without ASD, and clinical and demographic predictors of mtDNA content were examined. Leukocyte mtDNA AOD was also explored in order to investigate for mtDNA damage induced by buildup of endogenous ROS.

HEU status was independently associated with elevated leukocyte mtDNA content in children and adolescents between the ages of two and 16 years. ASD diagnosis was also independently associated with elevated mtDNA content. In children with both of these conditions mtDNA content was further elevated, suggesting a potential additive or cumulative relationship. We propose that elevated mtDNA content could represent a physiological response to systemic mitochondrial dysfunction, which is more pronounced in children with two conditions known to associate with mitochondrial damage and oxidative stress. However, a more comprehensive interpretation of these findings, including a mechanistic and/or causative evaluation of the interplay of oxidative stress, mitochondrial dysfunction, maternal risk factors, and progressive changes in mtDNA at the molecular level, would require prospective analyses which were outside the scope of our study. These results clearly warrant further investigation to examine the effects of other maternal, clinical, demographic, and environmental factors, as well as to validate whether these findings are reproducible in other cohorts. Nevertheless, the elevated prevalence of

ASD in HEU children is a novel finding, and one that indicates more detailed evaluations may be in order regarding neurodevelopmental outcomes in the rapidly growing worldwide population of HEU children and adolescents. 108

References

1. Forbes JC, Alimenti AM, Singer J, Brophy JC, Bitnun A, Samson LM, et al (2012). A national review of vertical HIV transmission. AIDS 26: 757–63.

2. Thorne C, Newell M-L (2007). Safety of Agents Used to Prevent Mother-to-Child Transmission of HIV. Drug Saf 30: 203–13.

3. Thorne C, Newell M-L (2005). The safety of antiretroviral drugs in pregnancy. Expert Opin Drug Saf 4: 323–35.

4. Lewis W, Day BJ, Copeland WC (2003). Mitochondrial toxicity of NRTI antiviral drugs: an integrated cellular perspective. Nat Rev Drug Discov 2: 812–22.

5. Shiramizu B, Shikuma KM, Kamemoto L, Gerschenson M, Erdem G, Pinti M, et al (2003). Placenta and cord blood mitochondrial DNA toxicity in HIV-infected women receiving nucleoside reverse transcriptase inhibitors during pregnancy. J Acquir Immune Defic Syndr 32: 370–4.

6. Côté HCF, Raboud J, Bitnun A, Alimenti A, Money DM, Maan E, et al (2008). Perinatal exposure to antiretroviral therapy is associated with increased blood mitochondrial DNA levels and decreased mitochondrial gene expression in infants. J Infect Dis 198: 851–9.

7. Poirier MC, Divi RL, Al-Harthi L, Olivero OA, Nguyen V, Walker B, et al (2003). Long- Term Mitochondrial Toxicity in HIV-Uninfected Infants Born to HIV-Infected Mothers. JAIDS J Acquir Immune Defic Syndr 33: 175–83.

8. Divi RL, Walker VE, Wade NA, Nagashima K, Seilkop SK, Adams ME, et al (2004). Mitochondrial damage and DNA depletion in cord blood and umbilical cord from infants exposed in utero to Combivir. AIDS 18: 1013–21.

9. Morris G, Berk M, Walder K, Maes M (2015). Central pathways causing fatigue in neuro- inflammatory and autoimmune illnesses. BMC Med 13: 28.

10. Frye RE, Rossignol DA (2011). Mitochondrial dysfunction can connect the diverse medical symptoms associated with autism spectrum disorders. Pediatr Res 69: 41R – 7R.

11. (CDC) DDMNSY 2010 PIC for DC and P (2014). Prevalence of Autism Spectrum Disorder Among Children Aged 8 Years — Autism and Developmental Disabilities Monitoring Network, 11 Sites, United States, 2010. MMWR Surveill Summ 63: 1–21.

12. Bowes J, Brophy JC, Samson L, Gentile C (2011). Pervasive Developmental Disorder in Antiretroviral- and HIV-Exposed, Uninfected Children [Abstract]. Can J Infect Dis Med Microbiol 22: 12B. 109

13. Sharp PM, Hahn BH (2011). Origins of HIV and the AIDS pandemic. Cold Spring Harb Perspect Med 1: a006841.

14. Berry N, Jaffar S, Schim van der Loeff M, Ariyoshi K, Harding E, N’Gom PT, et al (2002). Low level viremia and high CD4% predict normal survival in a cohort of HIV type-2-infected villagers. AIDS Res Hum Retroviruses 18: 1167–73.

15. Van der Loeff MFS, Awasana AA, Sarge-Njie R, van der Sande M, Jaye A, Sabally S, et al (2006). Sixteen years of HIV surveillance in a West African research clinic reveals divergent epidemic trends of HIV-1 and HIV-2. Int J Epidemiol 35: 1322–8.

16. Da Silva ZJ, Oliveira I, Andersen A, Dias F, Rodrigues A, Holmgren B, et al (2008). Changes in prevalence and incidence of HIV-1, HIV-2 and dual infections in urban areas of Bissau, Guinea-Bissau: is HIV-2 disappearing? AIDS 22: 1195–202.

17. Shaw GM, Hunter E (2012). HIV transmission. Cold Spring Harb Perspect Med 2.

18. Shattock RJ, Moore JP (2003). Inhibiting sexual transmission of HIV-1 infection. Nat Rev Microbiol Nature Publishing Group; 1: 25–34.

19. World Health Organization (2015). Guideline on When to Start Antiretroviral Therapy and on Pre-Exposure Prophylaxis for HIV. World Health Organization; 2015.

20. The Joint United Nations Programme on HIV/AIDS (UNAIDS) (2014). Fact Sheet: 2014 Global Statistics. 2014.

21. World Health Organization (2014). Global summary of the AIDS epidemic. 2014.

22. Government of Canada PHA of C (2011). HIV/AIDS Epi Updates - July 2014 - Public Health Agency of Canada.

23. Hall HI, Green TA, Wolitski RJ, Holtgrave DR, Rhodes P, Lehman JS, et al (2010). Estimated future HIV prevalence, incidence, and potential infections averted in the United States: a multiple scenario analysis. J Acquir Immune Defic Syndr 55: 271–6.

24. Abu-Raddad LJ, Barnabas R V, Janes H, Weiss HA, Kublin JG, Longini IM, et al (2013). Have the explosive HIV epidemics in sub-Saharan Africa been driven by higher community viral load? AIDS 27: 981–9.

25. Government of Canada PHA of C (2014). HIV and AIDS in Canada: Surveillance Report to December 31st, 2013 - Public Health Agency of Canada.

26. Gray GE, McIntyre JA (2007). HIV and pregnancy. BMJ 334: 950–3.

110

27. Muesing MA, Smith DH, Cabradilla CD, Benton C V, Lasky LA, Capon DJ (1985). Nucleic acid structure and expression of the human AIDS/lymphadenopathy retrovirus. Nature 313: 450–8.

28. Nielsen MH, Pedersen FS, Kjems J (2005). Molecular strategies to inhibit HIV-1 replication. Retrovirology 2: 10.

29. Klaver B, Berkhout B (1994). Comparison of 5’ and 3' long terminal repeat promoter function in human immunodeficiency virus. J Virol 68: 3830–40.

30. Freed EO (1998). HIV-1 gag proteins: diverse functions in the virus life cycle. Virology 251: 1–15.

31. Muller B, Patschinsky T, Krausslich H-G (2002). The Late-Domain-Containing Protein p6 Is the Predominant Phosphoprotein of Human Immunodeficiency Virus Type 1 Particles. J Virol 76: 1015–24.

32. Hill M, Tachedjian G, Mak J (2005). The packaging and maturation of the HIV-1 Pol proteins. Curr HIV Res 3: 73–85.

33. Gallo SA, Finnegan CM, Viard M, Raviv Y, Dimitrov A, Rawat SS, et al (2003). The HIV Env-mediated fusion reaction. Biochim Biophys Acta - Biomembr 1614: 36–50.

34. De Francesco MA, Baronio M, Fiorentini S, Signorini C, Bonfanti C, Poiesi C, et al (2002). HIV-1 matrix protein p17 increases the production of proinflammatory cytokines and counteracts IL-4 activity by binding to a cellular receptor. Proc Natl Acad Sci U S A 99: 9972–7.

35. Lapadat-Tapolsky M, De Rocquigny H, Van Gent D, Roques B, Plasterk R, Darlix JL (1993). Interactions between HIV-1 nucleocapsid protein and viral DNA may have important functions in the viral life cycle. Nucleic Acids Res 21: 831–9.

36. Lunardi-Iskandar Y, Nugeyre MT, Georgoulias V, Barré-Sinoussi F, Jasmin C, Chermann JC (1989). Replication of the human immunodeficiency virus 1 and impaired differentiation of T cells after in vitro infection of bone marrow immature T cells. J Clin Invest 83: 610–5.

37. Tsubota H, Ringler DJ, Kannagi M, King NW, Solomon KR, MacKey JJ, et al (1989). CD8+CD4- lymphocyte lines can harbor the AIDS virus in vitro. J Immunol 143: 858–63.

38. Langhoff E, Terwilliger EF, Bos HJ, Kalland KH, Poznansky MC, Bacon OM, et al (1991). Replication of human immunodeficiency virus type 1 in primary dendritic cell cultures. Proc Natl Acad Sci U S A 88: 7998–8002.

111

39. Steffan AM, Lafon ME, Gendrault JL, Schweitzer C, Royer C, Jaeck D, et al (1992). Primary cultures of endothelial cells from the human liver sinusoid are permissive for human immunodeficiency virus type 1. Proc Natl Acad Sci U S A 89: 1582–6.

40. Monté D, Groux H, Raharinivo B, Plouvier B, Dewulf J, Clavel T, et al (1992). Productive human immunodeficiency virus-1 infection of megakaryocytic cells is enhanced by tumor necrosis factor-alpha. Blood 79: 2670–9.

41. Carter CC, Onafuwa-Nuga A, McNamara LA, Riddell J, Bixby D, Savona MR, et al (2010). HIV-1 infects multipotent progenitor cells causing cell death and establishing latent cellular reservoirs. Nat Med 16: 446–51.

42. Churchill MJ, Wesselingh SL, Cowley D, Pardo CA, McArthur JC, Brew BJ, et al (2009). Extensive astrocyte infection is prominent in human immunodeficiency virus-associated dementia. Ann Neurol 66: 253–8.

43. Nakai M, Goto T (1996). Ultrastructure and morphogenesis of human immunodeficiency virus. J Electron Microsc (Tokyo) 45: 247–57.

44. Ferguson MR, Rojo DR, von Lindern JJ, O’Brien WA (2002). HIV-1 replication cycle. Clin Lab Med 22: 611–35.

45. McMichael AJ, Borrow P, Tomaras GD, Goonetilleke N, Haynes BF (2010). The immune response during acute HIV-1 infection: clues for vaccine development. Nat Rev Immunol 10: 11–23.

46. Hecht FM, Wellman R, Busch MP, Pilcher CD, Norris PJ, Margolick JB, et al (2011). Identifying the early post-HIV antibody seroconversion period. J Infect Dis 204: 526–33.

47. Royce RA, Seña A, Cates W, Cohen MS (1997). Sexual transmission of HIV. N Engl J Med 336: 1072–8.

48. Cohen MS, Gay CL, Busch MP, Hecht FM (2010). The detection of acute HIV infection. J Infect Dis 202 Suppl : S270–7.

49. Fiebig EW, Wright DJ, Rawal BD, Garrett PE, Schumacher RT, Peddada L, et al (2003). Dynamics of HIV viremia and antibody seroconversion in plasma donors: implications for diagnosis and staging of primary HIV infection. AIDS 17: 1871–9.

50. Schuetz A, Deleage C, Sereti I, Rerknimitr R, Phanuphak N, Phuang-Ngern Y, et al (2014). Initiation of ART during early acute HIV infection preserves mucosal Th17 function and reverses HIV-related immune activation. PLoS Pathog 10: e1004543.

51. Guadalupe M, Reay E, Sankaran S, Prindiville T, Flamm J, McNeil A, et al (2003). Severe CD4+ T-cell depletion in gut lymphoid tissue during primary human immunodeficiency 112

virus type 1 infection and substantial delay in restoration following highly active antiretroviral therapy. J Virol 77: 11708–17.

52. Brenchley JM, Schacker TW, Ruff LE, Price DA, Taylor JH, Beilman GJ, et al (2004). CD4+ T cell depletion during all stages of HIV disease occurs predominantly in the gastrointestinal tract. J Exp Med 200: 749–59.

53. Kaplan D, Sieg S (1998). Role of the Fas/Fas ligand apoptotic pathway in human immunodeficiency virus type 1 disease. J Virol 72: 6279–82.

54. Piatak M, Saag MS, Yang LC, Clark SJ, Kappes JC, Luk KC, et al (1993). High levels of HIV-1 in plasma during all stages of infection determined by competitive PCR. Science 259: 1749–54.

55. Koup RA, Safrit JT, Cao Y, Andrews CA, McLeod G, Borkowsky W, et al (1994). Temporal association of cellular immune responses with the initial control of viremia in primary human immunodeficiency virus type 1 syndrome. J Virol 68: 4650–5.

56. Schmitz JE, Kuroda MJ, Santra S, Sasseville VG, Simon MA, Lifton MA, et al (1999). Control of viremia in simian immunodeficiency virus infection by CD8+ lymphocytes. Science 283: 857–60.

57. Clark SJ, Shaw GM (1993). The acute retroviral syndrome and the pathogenesis of HIV-1 infection. Semin Immunol 5: 149–55.

58. Brenchley JM, Price DA, Schacker TW, Asher TE, Silvestri G, Rao S, et al (2006). Microbial translocation is a cause of systemic immune activation in chronic HIV infection. Nat Med Nature Publishing Group; 12: 1365–71.

59. Kelley CF, Barbour JD, Hecht FM (2007). The Relation Between Symptoms, Viral Load, and Viral Load Set Point in Primary HIV Infection. JAIDS J Acquir Immune Defic Syndr 45: 445–8.

60. Fellay J, Shianna K V, Ge D, Colombo S, Ledergerber B, Weale M, et al (2007). A whole- genome association study of major determinants for host control of HIV-1. Science 317: 944–7.

61. Streeck H, Jolin JS, Qi Y, Yassine-Diab B, Johnson RC, Kwon DS, et al (2009). Human Immunodeficiency Virus Type 1-Specific CD8+ T-Cell Responses during Primary Infection Are Major Determinants of the Viral Set Point and Loss of CD4+ T Cells. J Virol 83: 7641–8.

62. Connor RI, Sheridan KE, Ceradini D, Choe S, Landau NR (1997). Change in coreceptor use correlates with disease progression in HIV-1--infected individuals. J Exp Med 185: 621–8. 113

63. Muñoz A, Wang MC, Bass S, Taylor JM, Kingsley LA, Chmiel JS, et al (1989). Acquired immunodeficiency syndrome (AIDS)-free time after human immunodeficiency virus type 1 (HIV-1) seroconversion in homosexual men. Multicenter AIDS Cohort Study Group. Am J Epidemiol 130: 530–9.

64. (2014). Revised surveillance case definition for HIV infection--United States, 2014. MMWR Recomm Rep 63: 1–10.

65. World Health Organization (2007). WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children. 2007.

66. Mocroft A, Youle M, Morcinek J, Sabin CA, Gazzard B, Johnson MA, et al (1997). Survival after diagnosis of AIDS: a prospective observational study of 2625 patients. Royal Free/Chelsea and Westminster Hospitals Collaborative Group. BMJ 314: 409–13.

67. Lee LM, Karon JM, Selik R, Neal JJ, Fleming PL (2001). Survival after AIDS diagnosis in adolescents and adults during the treatment era, United States, 1984-1997. JAMA 285: 1308–15.

68. Jacobson LP, Kirby AJ, Polk S, Phair JP, Besley DR, Saah AJ, et al (1993). Changes in survival after acquired immunodeficiency syndrome (AIDS): 1984-1991. Am J Epidemiol 138: 952–64.

69. Vella S, Schwartländer B, Sow SP, Eholie SP, Murphy RL (2012). The history of antiretroviral therapy and of its implementation in resource-limited areas of the world. AIDS 26: 1231–41.

70. Palmisano L, Vella S (2011). A brief history of antiretroviral therapy of HIV infection: success and challenges. Ann Ist Super Sanita 47: 44–8.

71. Hirsch MS, Günthard HF, Schapiro JM, Brun-Vézinet F, Clotet B, Hammer SM, et al (2008). Antiretroviral drug resistance testing in adult HIV-1 infection: 2008 recommendations of an International AIDS Society-USA panel. Clin Infect Dis 47: 266– 85.

72. Carr A, Cooper DA (2000). Adverse effects of antiretroviral therapy. Lancet (London, England) 356: 1423–30.

73. Schauer GD, Huber KD, Leuba SH, Sluis-Cremer N (2014). Mechanism of allosteric inhibition of HIV-1 reverse transcriptase revealed by single-molecule and ensemble fluorescence. Nucleic Acids Res 42: 11687–96.

74. Gazzard B (2000). What are the perceived advantages and disadvantages of non- nucleoside reverse transcriptase inhibitors as first-line therapy? HIV Med 1: 11–4. 114

75. Boden D (1999). HIV-1 Drug Resistance in Newly Infected Individuals. JAMA American Medical Association; 282: 1135.

76. De Clercq E (2007). The design of drugs for HIV and HCV. Nat Rev Drug Discov 6: 1001–18.

77. Maarseveen N van, Boucher C (2006). Resistance to protease inhibitors. Mediscript; 2006.

78. Pommier Y, Johnson AA, Marchand C (2005). Integrase inhibitors to treat HIV/AIDS. Nat Rev Drug Discov 4: 236–48.

79. Pendri A, Meanwell NA, Peese KM, Walker MA (2011). New first and second generation inhibitors of human immunodeficiency virus-1 integrase. Expert Opin Ther Pat 21: 1173– 89.

80. Westling K, Pettersson K, Kaldma A, Navér L (2012). Rapid decline in HIV viral load when introducing raltegravir-containing antiretroviral treatment late in pregnancy. AIDS Patient Care STDS Mary Ann Liebert, Inc. 140 Huguenot Street, 3rd Floor New Rochelle, NY 10801 USA; 26: 714–7.

81. Boucoiran I, Tulloch K, Pick N, Kakkar F, van Schalkwyk J, Money D, et al A case series of third-trimester raltegravir initiation: Impact on maternal HIV-1 viral load and obstetrical outcomes. Can J Infect Dis Med Microbiol = J Can des Mal Infect la Microbiol médicale / AMMI Canada 26: 145–50.

82. Briz V, Poveda E, Soriano V (2006). HIV entry inhibitors: mechanisms of action and resistance pathways. J Antimicrob Chemother 57: 619–27.

83. Esté JA, Telenti A (2007). HIV entry inhibitors. Lancet (London, England) 370: 81–8.

84. Money D, Tulloch K, Boucoiran I, Caddy S, Yudin MH, Allen V, et al (2014). Guidelines for the Care of Pregnant Women Living With HIV and Interventions to Reduce Perinatal Transmission: Executive Summary. J Obstet Gynaecol Canada 36: 721–34.

85. Panel on Antiretroviral Therapy and Medical Management of HIV-Infected Children (2016). Guidelines for the Use of Antiretroviral Agents in Pediatric HIV Infection. 2016.

86. World Health Organization (2006). WHO | Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants. World Health Organization;

87. Research WHOD of RH and, HIV UI-ATT on M-CT of (2001). New data on the prevention of mother-to-child transmission of HIV and their policy implications : conclusions and recommendations : WHO Technical consultation on behalf of the UNFPA/UNICEF/WHO/UNAIDS Inter-Agency Task Team on Mother-to-Child Transmission . Geneva : World Health Organization; 115

88. World Health Organization (2004). Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants [Internet]. 2004.

89. World Health Organization (2010). WHO | Antiretroviral drugs for treating pregnant women and preventing HIV infection in infants. World Health Organization;

90. World Health Organization (2012). WHO | Use of antiretroviral drugs for treating pregnant women and preventing HIV infection in infants. World Health Organization;

91. World Health Organization (2013). WHO | Consolidated guidelines on the use of antiretroviral drugs for treating and preventing HIV infection. World Health Organization;

92. Magder LS, Mofenson L, Paul ME, Zorrilla CD, Blattner WA, Tuomala RE, et al (2005). Risk factors for in utero and intrapartum transmission of HIV. J Acquir Immune Defic Syndr 38: 87–95.

93. EHRNST A (1991). HIV in pregnant women and their offspring: evidence for late transmission. Lancet 338: 203–7.

94. Sprecher S, Soumenkoff G, Puissant F, Degueldre M (1986). VERTICAL TRANSMISSION OF HIV IN 15-WEEK FETUS. Lancet 328: 288–9.

95. Dunn D, Newell M, Ades A, Peckham C (1992). Risk of human immunodeficiency virus type 1 transmission through breastfeeding. Lancet 340: 585–8.

96. Mognetti B, Moussa M, Croitoru J, Menu E, Dormont D, Roques P, et al (2000). HIV-1 co-receptor expression on trophoblastic cells from early placentas and permissivity to infection by several HIV-1 primary isolates. Clin Exp Immunol 119: 486–92.

97. Soilleux EJ, Morris LS, Lee B, Pöhlmann S, Trowsdale J, Doms RW, et al (2001). Placental expression of DC-SIGN may mediate intrauterine vertical transmission of HIV. J Pathol 195: 586–92.

98. Soilleux EJ, Coleman N (2003). Transplacental transmission of HIV: a potential role for HIV binding lectins. Int J Biochem Cell Biol 35: 283–7.

99. Saunders NR, Liddelow SA, Dziegielewska KM (2012). Barrier mechanisms in the developing brain. Front Pharmacol 3: 46.

100. Meyer U, Feldon J, Fatemi SH (2009). In-vivo rodent models for the experimental investigation of prenatal immune activation effects in neurodevelopmental brain disorders. Neurosci Biobehav Rev 33: 1061–79.

116

101. Hernàndez S, Morén C, López M, Coll O, Cardellach F, Gratacós E, et al (2012). Perinatal outcomes, mitochondrial toxicity and apoptosis in HIV-treated pregnant women and in- utero-exposed newborn. AIDS 26: 419–28.

102. Ross AC, Leong T, Avery A, Castillo-Duran M, Bonilla H, Lebrecht D, et al (2012). Effects of in utero antiretroviral exposure on mitochondrial DNA levels, mitochondrial function and oxidative stress. HIV Med 13: 98–106.

103. Decloedt EH, Rosenkranz B, Maartens G, Joska J (2015). Central nervous system penetration of antiretroviral drugs: pharmacokinetic, pharmacodynamic and pharmacogenomic considerations. Clin Pharmacokinet 54: 581–98.

104. McCormack SA, Best BM (2014). Protecting the fetus against HIV infection: a systematic review of placental transfer of antiretrovirals. Clin Pharmacokinet 53: 989–1004.

105. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P (2002). Sperm. Garland Science; 2002.

106. Cimen MYB (2008). Free radical metabolism in human erythrocytes. Clin Chim Acta 390: 1–11.

107. Alberts B, Johnson A, Lewis J, Raff M, Roberts K, Walter P (2002). The Genetic Systems of Mitochondria and Plastids. Garland Science; 2002.

108. Muftuoglu M, Mori MP, de Souza-Pinto NC (2014). Formation and repair of oxidative damage in the mitochondrial DNA. Mitochondrion 17: 164–81.

109. Cooper GM (2000). Metabolic Energy. Sinauer Associates; 2000.

110. St-Pierre J, Buckingham JA, Roebuck SJ, Brand MD (2002). Topology of superoxide production from different sites in the mitochondrial electron transport chain. J Biol Chem 277: 44784–90.

111. Adam-Vizi V (2005). Production of reactive oxygen species in brain mitochondria: contribution by electron transport chain and non-electron transport chain sources. Antioxid Redox Signal Mary Ann Liebert, Inc. 2 Madison Avenue Larchmont, NY 10538 USA; 7: 1140–9.

112. Liu Y, Fiskum G, Schubert D (2002). Generation of reactive oxygen species by the mitochondrial electron transport chain. J Neurochem 80: 780–7.

113. Indo HP, Davidson M, Yen H-C, Suenaga S, Tomita K, Nishii T, et al (2007). Evidence of ROS generation by mitochondria in cells with impaired electron transport chain and mitochondrial DNA damage. Mitochondrion 7: 106–18.

117

114. Parker WD, Boyson SJ, Parks JK (1989). Abnormalities of the electron transport chain in idiopathic Parkinson’s disease. Ann Neurol 26: 719–23.

115. Benecke R, Strümper P, Weiss H (1993). Electron transfer complexes I and IV of platelets are abnormal in Parkinson’s disease but normal in Parkinson-plus syndromes. Brain 116 ( Pt 6: 1451–63.

116. Parker WD, Filley CM, Parks JK (1990). Cytochrome oxidase deficiency in Alzheimer’s disease. Neurology 40: 1302–1302.

117. Parker WD, BA JP, Filley CM, Kleinschmidt-DeMasters BK (1994). Electron transport chain defects in Alzheimer’s disease brain. Neurology 44: 1090–1090.

118. Parker WD, Boyson SJ, Luder AS, Parks JK (1990). Evidence for a defect in NADH: ubiquinone oxidoreductase (complex I) in Huntington’s disease. Neurology 40: 1231– 1231.

119. Koroshetz WJ, Jenkins BG, Rosen BR, Beal MF (1997). Energy metabolism defects in Huntington’s disease and effects of coenzyme Q10. Ann Neurol 41: 160–5.

120. Sutovsky P, Moreno RD, Ramalho-Santos J, Dominko T, Simerly C, Schatten G (2000). Ubiquitinated sperm mitochondria, selective proteolysis, and the regulation of mitochondrial inheritance in mammalian embryos. Biol Reprod 63: 582–90.

121. Gabriel MS, Chan SW, Alhathal N, Chen JZ, Zini A (2012). Influence of microsurgical varicocelectomy on human sperm mitochondrial DNA copy number: a pilot study. J Assist Reprod Genet 29: 759–64.

122. Lee H-C, Wei Y-H (2005). Mitochondrial biogenesis and mitochondrial DNA maintenance of mammalian cells under oxidative stress. Int J Biochem Cell Biol 37: 822– 34.

123. Lagouge M, Larsson N-G (2013). The role of mitochondrial DNA mutations and free radicals in disease and ageing. J Intern Med 273: 529–43.

124. Falkenberg M, Larsson N-G, Gustafsson CM (2007). DNA replication and transcription in mammalian mitochondria. Annu Rev Biochem Annual Reviews; 76: 679–99.

125. Kaguni LS (2004). DNA P OLYMERASE γ, T HE M ITOCHONDRIAL R EPLICASE 1. Annu Rev Biochem Annual Reviews 4139 El Camino Way, P.O. Box 10139, Palo Alto, CA 94303-0139, USA; 73: 293–320.

126. Virbasius J V, Scarpulla RC (1994). Activation of the human mitochondrial transcription factor A gene by nuclear respiratory factors: a potential regulatory link between nuclear

118

and mitochondrial gene expression in organelle biogenesis. Proc Natl Acad Sci 91: 1309– 13.

127. Barrientos A, Casademont J, Cardellach F, Estivill X, Urbano-Marquez A, Nunes V (1997). Reduced steady-state levels of mitochondrial RNA and increased mitochondrial DNA amount in human brain with aging. Brain Res Mol Brain Res 52: 284–9.

128. Shigenaga MK, Hagen TM, Ames BN (1994). Oxidative damage and mitochondrial decay in aging. Proc Natl Acad Sci 91: 10771–8.

129. Boesch P, Weber-Lotfi F, Ibrahim N, Tarasenko V, Cosset A, Paulus F, et al (2011). DNA repair in organelles: Pathways, organization, regulation, relevance in disease and aging. Biochim Biophys Acta 1813: 186–200.

130. Dizdaroglu M, Jaruga P, Birincioglu M, Rodriguez H (2002). Free radical-induced damage to DNA: mechanisms and measurement. Free Radic Biol Med 32: 1102–15.

131. Cheng KC, Cahill DS, Kasai H, Nishimura S, Loeb LA (1992). 8-Hydroxyguanine, an abundant form of oxidative DNA damage, causes G----T and A----C substitutions. J Biol Chem 267: 166–72.

132. Zelko IN, Mariani TJ, Folz RJ (2002). Superoxide dismutase multigene family: a comparison of the CuZn-SOD (SOD1), Mn-SOD (SOD2), and EC-SOD (SOD3) gene structures, evolution, and expression. Free Radic Biol Med 33: 337–49.

133. Sung J-S, Demple B (2006). Roles of base excision repair subpathways in correcting oxidized abasic sites in DNA. FEBS J 273: 1620–9.

134. McPartland JC, Reichow B, Volkmar FR (2012). Sensitivity and specificity of proposed DSM-5 diagnostic criteria for autism spectrum disorder. J Am Acad Child Adolesc Psychiatry 51: 368–83.

135. Schaefer GB, Mendelsohn NJ (2008). Clinical genetics evaluation in identifying the etiology of autism spectrum disorders. Genet Med Nature Publishing Group; 10: 301–5.

136. Bailey A, Le Couteur A, Gottesman I, Bolton P, Simonoff E, Yuzda E, et al (1995). Autism as a strongly genetic disorder: evidence from a British twin study. Psychol Med Cambridge University Press; 25: 63.

137. Folstein S, Rutter M (1977). INFANTILE AUTISM: A GENETIC STUDY OF 21 TWIN PAIRS. J Child Psychol Psychiatry 18: 297–321.

138. Hallmayer J, Cleveland S, Torres A, Phillips J, Cohen B, Torigoe T, et al (2011). Genetic heritability and shared environmental factors among twin pairs with autism. Arch Gen Psychiatry American Medical Association; 68: 1095–102. 119

139. Bauman ML, Kemper TL (2005). Neuroanatomic observations of the brain in autism: a review and future directions. Int J Dev Neurosci 23: 183–7.

140. Miller MT, Strömland K, Ventura L, Johansson M, Bandim JM, Gillberg C (2005). Autism associated with conditions characterized by developmental errors in early embryogenesis: a mini review. Int J Dev Neurosci 23: 201–19.

141. Association AP (2013). Diagnostic and statistical manual of mental disorders. 5th ed. American Psychiatric Publishing; 2013.

142. Hyman SL (2013). New DSM-5 includes changes to autism criteria. AAP News American Academy of Pediatrics;

143. Huerta M, Bishop SL, Duncan A, Hus V, Lord C (2012). Application of DSM-5 criteria for autism spectrum disorder to three samples of children with DSM-IV diagnoses of pervasive developmental disorders. Am J Psychiatry 169: 1056–64.

144. Dua V (2003). Standards and Guidelines for the Assessment and Diagnosis of Young Children with Autism Spectrum Disorder in British Columbia. 2003.

145. Dosreis S, Weiner CL, Johnson L, Newschaffer CJ (2006). Autism spectrum disorder screening and management practices among general pediatric providers. J Dev Behav Pediatr 27: S88–94.

146. Glascoe FP (1997). Parents’ Concerns About Children's Development: Prescreening Technique or Screening Test? Pediatrics 99: 522–8.

147. Stone WL, Coonrod EE, Ousley OY Brief Report: Screening Tool for Autism in Two- Year-Olds (STAT): Development and Preliminary Data. J Autism Dev Disord Kluwer Academic Publishers-Plenum Publishers; 30: 607–12.

148. Berument SK, Rutter M, Lord C, Pickles A, Bailey A (1999). Autism screening questionnaire: diagnostic validity. Br J Psychiatry 175: 444–51.

149. Baird G, Charman T, Baron-Cohen S, Cox A, Swettenham J, Wheelwright S, et al (2000). A screening instrument for autism at 18 months of age: a 6-year follow-up study. J Am Acad Child Adolesc Psychiatry 39: 694–702.

150. Lord C, Risi S, DiLavore PS, Shulman C, Thurm A, Pickles A (2006). Autism from 2 to 9 years of age. Arch Gen Psychiatry 63: 694–701.

151. Shattuck PT, Durkin M, Maenner M, Newschaffer C, Mandell DS, Wiggins L, et al (2009). Timing of identification among children with an autism spectrum disorder: findings from a population-based surveillance study. J Am Acad Child Adolesc Psychiatry 48: 474–83. 120

152. Coo H, Ouellette-Kuntz H, Lam M, Yu CT, Dewey D, Bernier FP, et al (2012). Correlates of age at diagnosis of autism spectrum disorders in six Canadian regions. Chronic Dis Inj Can 32: 90–100.

153. Ouellette-Kuntz HMJ, Coo H, Lam M, Yu CT, Breitenbach MM, Hennessey PE, et al Age at diagnosis of autism spectrum disorders in four regions of Canada. Can J public Heal = Rev Can santé publique 100: 268–73.

154. Daniels AM, Mandell DS (2014). Explaining differences in age at autism spectrum disorder diagnosis: a critical review. Autism 18: 583–97.

155. Lord C, Risi S, Lambrecht L, Jr EHC, Leventhal BL, DiLavore PC, et al (2000). The Autism Diagnostic Observation Schedule—Generic: A Standard Measure of Social and Communication Deficits Associated with the Spectrum of Autism. J Autism Dev Disord Kluwer Academic Publishers-Plenum Publishers; 30: 205–23.

156. Lord C, Rutter M, Le Couteur A (1994). Autism Diagnostic Interview-Revised: a revised version of a diagnostic interview for caregivers of individuals with possible pervasive developmental disorders. J Autism Dev Disord 24: 659–85.

157. Government of Canada HCPAC and CB (2008). Building an Autism Spectrum Disorders Surveillance System for Canada - Diseases and Conditions - Health Canada. 2008.

158. Ouellette-Kuntz H, Coo H, Lam M, Breitenbach MM, Hennessey PE, Jackman PD, et al (2014). The changing prevalence of autism in three regions of Canada. J Autism Dev Disord 44: 120–36.

159. Elsabbagh M, Divan G, Koh Y-J, Kim YS, Kauchali S, Marcín C, et al (2012). Global prevalence of autism and other pervasive developmental disorders. Autism Res 5: 160–79.

160. Atladóttir HO, Thorsen P, Østergaard L, Schendel DE, Lemcke S, Abdallah M, et al (2010). Maternal infection requiring hospitalization during pregnancy and autism spectrum disorders. J Autism Dev Disord 40: 1423–30.

161. Shi L, Fatemi SH, Sidwell RW, Patterson PH (2003). Maternal influenza infection causes marked behavioral and pharmacological changes in the offspring. J Neurosci 23: 297–302.

162. Durkin MS, Maenner MJ, Newschaffer CJ, Lee L-C, Cunniff CM, Daniels JL, et al (2008). Advanced parental age and the risk of autism spectrum disorder. Am J Epidemiol 168: 1268–76.

163. Ben Itzchak E, Lahat E, Burgin R, Zachor AD (2008). Cognitive, behavior and intervention outcome in young children with autism. Res Dev Disabil 29: 447–58.

121

164. Matson JL, Dempsey T, Lovullo S V, Wilkins J (2008). The effects of intellectual functioning on the range of core symptoms of autism spectrum disorders. Res Dev Disabil 29: 341–50.

165. Laumonnier F, Bonnet-Brilhault F, Gomot M, Blanc R, David A, Moizard M-P, et al (2004). X-linked mental retardation and autism are associated with a mutation in the NLGN4 gene, a member of the neuroligin family. Am J Hum Genet 74: 552–7.

166. McCarthy SE, Gillis J, Kramer M, Lihm J, Yoon S, Berstein Y, et al (2014). De novo mutations in schizophrenia implicate chromatin remodeling and support a genetic overlap with autism and intellectual disability. Mol Psychiatry 19: 652–8.

167. Ibrahim SH, Voigt RG, Katusic SK, Weaver AL, Barbaresi WJ (2009). Incidence of gastrointestinal symptoms in children with autism: a population-based study. Pediatrics 124: 680–6.

168. Horvath K, Papadimitriou JC, Rabsztyn A, Drachenberg C, Tildon JT (1999). Gastrointestinal abnormalities in children with autistic disorder. J Pediatr 135: 559–63.

169. Finegold SM, Molitoris D, Song Y, Liu C, Vaisanen M-L, Bolte E, et al (2002). Gastrointestinal microflora studies in late-onset autism. Clin Infect Dis 35: S6–16.

170. Deykin EY, MacMahon B (1979). The incidence of seizures among children with autistic symptoms. Am J Psychiatry American Psychiatric Assn; 136: 1310–2.

171. Volkmar FR, Nelson DS (1990). Seizure disorders in autism. J Am Acad Child Adolesc Psychiatry 29: 127–9.

172. Hughes JR, Melyn M (2005). EEG and Seizures in Autistic Children and Adolescents: Further Findings with Therapeutic Implications. Clin EEG Neurosci 36: 15–20.

173. Parmeggiani A, Barcia G, Posar A, Raimondi E, Santucci M, Scaduto MC (2010). Epilepsy and EEG paroxysmal abnormalities in autism spectrum disorders. Brain Dev 32: 783–9.

174. Miles JH, Hillman RE (2000). Value of a clinical morphology examination in autism. Am J Med Genet 91: 245–53.

175. Muñoz-Yunta JA, Freixas A, Valls-Santasusana A, Maldonado A (1999). [Stereotypes, developmental disorders and neuroimaging studies]. Rev Neurol 28 Suppl 2: S124–30.

176. Brinkman K, ter Hofstede HJ, Burger DM, Smeitink JA, Koopmans PP (1998). Adverse effects of reverse transcriptase inhibitors: mitochondrial toxicity as common pathway. AIDS 12: 1735–44.

122

177. Payne BAI, Wilson IJ, Hateley CA, Horvath R, Santibanez-Koref M, Samuels DC, et al (2011). Mitochondrial aging is accelerated by anti-retroviral therapy through the clonal expansion of mtDNA mutations. Nat Genet Nature Publishing Group, a division of Macmillan Publishers Limited. All Rights Reserved.; 43: 806–10.

178. Pace GW, Leaf CD (1995). The role of oxidative stress in HIV disease. Free Radic Biol Med 19: 523–8.

179. Lagathu C, Eustace B, Prot M, Frantz D, Gu Y, Bastard J-P, et al (2007). Some HIV antiretrovirals increase oxidative stress and alter chemokine, cytokine or adiponectin production in human adipocytes and macrophages. Antivir Ther 12: 489–500.

180. Akay C, Cooper M, Odeleye A, Jensen BK, White MG, Vassoler F, et al (2014). Antiretroviral drugs induce oxidative stress and neuronal damage in the central nervous system. J Neurovirol 20: 39–53.

181. Deeks SG (2011). HIV infection, inflammation, immunosenescence, and aging. Annu Rev Med 62: 141–55.

182. Aldrovandi GM, Chu C, Shearer WT, Li D, Walter J, Thompson B, et al (2009). Antiretroviral exposure and lymphocyte mtDNA content among uninfected infants of HIV-1-infected women. Pediatrics 124: e1189–97.

183. McComsey GA, Kang M, Ross AC, Lebrecht D, Livingston E, Melvin A, et al (2015). Increased mtDNA Levels Without Change in Mitochondrial Enzymes in Peripheral Blood Mononuclear Cells of Infants Born to HIV-Infected Mothers on Antiretroviral Therapy. HIV Clin Trials Taylor & Francis;

184. Chen S, Li Z, He Y, Zhang F, Li H, Liao Y, et al (2015). Elevated mitochondrial DNA copy number in peripheral blood cells is associated with childhood autism. BMC Psychiatry BioMed Central; 15: 50.

185. James SJ, Cutler P, Melnyk S, Jernigan S, Janak L, Gaylor DW, et al (2004). Metabolic biomarkers of increased oxidative stress and impaired methylation capacity in children with autism. Am J Clin Nutr 80: 1611–7.

186. Selvam A, Buhimschi IA, Makin JD, Pattinson RC, Anderson R, Forsyth BW (2015). Hyperferritinemia and markers of inflammation and oxidative stress in the cord blood of HIV-exposed, uninfected (HEU) infants. HIV Med 16: 375–80.

187. Jitratkosol MHJ, Sattha B, Maan EJ, Gadawski I, Harrigan PR, Forbes JC, et al (2012). Blood mitochondrial DNA mutations in HIV-infected women and their infants exposed to HAART during pregnancy. AIDS 26: 675–83.

123

188. Graf WD, Marin-Garcia J, Gao HG, Pizzo S, Naviaux RK, Markusic D, et al (2000). Autism Associated With the Mitochondrial DNA G8363A Transfer RNALys Mutation. J Child Neurol 15: 357–61.

189. Cawthon RM (2009). Telomere length measurement by a novel monochrome multiplex quantitative PCR method. Nucleic Acids Res 37: e21–e21.

190. Lefèvre C, Auclair M, Boccara F, Bastard J-P, Capeau J, Vigouroux C, et al (2010). Premature senescence of vascular cells is induced by HIV protease inhibitors: implication of prelamin A and reversion by statin. Arterioscler Thromb Vasc Biol 30: 2611–20.

191. Biesecker G, Karimi S, Desjardins J, Meyer D, Abbott B, Bendele R, et al (2003). Evaluation of mitochondrial DNA content and enzyme levels in tenofovir DF-treated rats, rhesus monkeys and woodchucks. Antiviral Res 58: 217–25.

192. Dimmock D, Tang L-Y, Schmitt ES, Wong L-JC (2010). Quantitative evaluation of the mitochondrial DNA depletion syndrome. Clin Chem 56: 1119–27.

193. Larder B, Kemp S, Harrigan P (1995). Potential mechanism for sustained antiretroviral efficacy of AZT-3TC combination therapy. Science (80- ) American Association for the Advancement of Science; 269: 696–9.

194. Deeks SG (1997). HIV-1 Protease Inhibitors. JAMA American Medical Association; 277: 145.

195. Sethi AK, Celentano DD, Gange SJ, Moore RD, Gallant JE (2003). Association between adherence to antiretroviral therapy and human immunodeficiency virus drug resistance. Clin Infect Dis 37: 1112–8.

196. Garrabou G, López S, Morén C, Martínez E, Fontdevila J, Cardellach F, et al (2011). Mitochondrial damage in adipose tissue of untreated HIV-infected patients. AIDS 25: 165–70.

197. Chen C, Vazquez-Padua M, Cheng Y (1991). Effect of anti-human immunodeficiency virus nucleoside analogs on mitochondrial DNA and its implication for delayed toxicity. Mol Pharmacol 39: 625–8.

198. Medina DJ, Tsai CH, Hsiung GD, Cheng YC (1994). Comparison of mitochondrial morphology, mitochondrial DNA content, and cell viability in cultured cells treated with three anti-human immunodeficiency virus dideoxynucleosides. Antimicrob Agents Chemother 38: 1824–8.

199. Côté HCF, Yip B, Asselin JJ, Chan JW, Hogg RS, Harrigan PR, et al (2003). Mitochondrial:nuclear DNA ratios in peripheral blood cells from human

124

immunodeficiency virus (HIV)-infected patients who received selected HIV antiretroviral drug regimens. J Infect Dis 187: 1972–6.

200. Brogly SB, Ylitalo N, Mofenson LM, Oleske J, Van Dyke R, Crain MJ, et al (2007). In utero nucleoside reverse transcriptase inhibitor exposure and signs of possible mitochondrial dysfunction in HIV-uninfected children. AIDS 21: 929–38.

201. Alimenti A, Burdge DR, Ogilvie GS, Money DM, Forbes JC (2003). Lactic acidemia in human immunodeficiency virus-uninfected infants exposed to perinatal antiretroviral therapy. Pediatr Infect Dis J 22: 782–9.

202. Gingelmaier A, Grubert TA, Kost BP, Setzer B, Lebrecht D, Mylonas I, et al (2009). Mitochondrial toxicity in HIV type-1-exposed pregnancies in the era of highly active antiretroviral therapy. Antivir Ther 14: 331–8.

203. Giulivi C, Zhang Y-F, Omanska-Klusek A, Ross-Inta C, Wong S, Hertz-Picciotto I, et al (2010). Mitochondrial dysfunction in autism. JAMA American Medical Association; 304: 2389–96.

204. Rossignol DA, Frye RE (2012). Mitochondrial dysfunction in autism spectrum disorders: a systematic review and meta-analysis. Mol Psychiatry Macmillan Publishers Limited; 17: 290–314.

205. Westbrook DG, Anderson PG, Pinkerton KE, Ballinger SW (2010). Perinatal tobacco smoke exposure increases vascular oxidative stress and mitochondrial damage in non- human primates. Cardiovasc Toxicol 10: 216–26.

206. Santos JH, Mandavilli BS, Van Houten B (2002). Measuring oxidative mtDNA damage and repair using quantitative PCR. Methods Mol Biol 197: 159–76.

207. Kovalenko OA, Santos JH (2009). Analysis of oxidative damage by gene-specific quantitative PCR. Curr Protoc Hum Genet Chapter 19: Unit 19.1.

208. Cartwright GE, Athens JW, Wintrobe MM (1964). The kinetics of granulopoiesis in a normal man. Blood American Society of Hematology; 24: 780–803.

209. Ottesen J (1954). On the age of human white cells in peripheral blood. Acta Physiol Scand 32: 75–93.

210. Tang G, Gutierrez Rios P, Kuo S-H, Akman HO, Rosoklija G, Tanji K, et al (2013). Mitochondrial abnormalities in temporal lobe of autistic brain. Neurobiol Dis 54: 349–61.

211. Torrell H, Montaña E, Abasolo N, Roig B, Gaviria AM, Vilella E, et al (2013). Mitochondrial DNA (mtDNA) in brain samples from patients with major psychiatric

125

disorders: gene expression profiles, mtDNA content and presence of the mtDNA common deletion. Am J Med Genet B Neuropsychiatr Genet 162B: 213–23.

212. Starace F, Ammassari A, Trotta MP, Murri R, De Longis P, Izzo C, et al (2002). Depression is a risk factor for suboptimal adherence to highly active antiretroviral therapy. J Acquir Immune Defic Syndr 31 Suppl 3: S136–9.

213. Gentile S (2005). The Safety of Newer Antidepressants in Pregnancy and Breastfeeding. Drug Saf 28: 137–52.

214. Shaw CA, Li Y, Tomljenovic L (2013). Administration of aluminium to neonatal mice in vaccine-relevant amounts is associated with adverse long term neurological outcomes. J Inorg Biochem 128: 237–44.

126

Appendix: CARMA consent form

127

Mechanism of Aging Following Exposure to HIV Antiretroviral Drugs CIHR Team Grant in HIV Therapy and Aging CARMA-2: MITOCHONDRIAL & TELOMERE STUDIES IN A PROSPECTIVE COHORT and MEASURING MITOCHONDRIAL AGING, APPLICATION TO HIV INFECTION AND THERAPY and CELLULAR AGING AND HIV COMORBIDITIES IN WOMEN AND CHILDREN

carma-2 ~ Informed Consent ~ Study Participants ~

 Site Principal Investigator: Dr Deborah Money

Executive Director – WHRI, OBGYN - Oak Tree Clinic 604 875 3459

Local CIHR Team Co-Investigators

Dr Helene Cote Dr Neora Pick Dr Melanie Murray Dr Ariane Alimenti Associate Professor Medical Director and Adult Physician Pediatrician Dept of Path & Lab Med, Adult Physician, Oak Tree Oak Tree Clinic Oak Tree Clinic UBC Clinic 604 875 2212 604 875 2212 604 822 9777 604 875 2212

Emergency contact 24 hrs/7 days/week ~ Evelyn Maan RN at 604 767 5044

INTRODUCTION

Throughout this consent form, when we say “you” or “your”, we mean you or your child. You are being asked to participate in the research study, named above, because you are living with HIV, or your mother has HIV and she took anti-HIV drugs when she was pregnant with you. The study team, listed above, is trying to better understand the effects of the HIV virus and the drugs that you are taking, or have taken, or were exposed to, on the cells of your body. Also we are trying to better

128

understand the bone and endocrine health (a system of glands in the body that secrete hormones which help control the chemical reactions in the body needed to maintain life) of people living with HIV.

This consent form will provide you with information on the options available and purpose of the study, how it may help you, any risks to you, and what is expected of you during the study. Once you understand the study, if you agree to take part in the study, you will be asked to sign this consent form. You will be given a copy of this form to keep for your records.

This project is funded by a grant from the Canadian Institutes of Health Research (CIHR).

PURPOSE

Anti-HIV drugs are used to treat people living with HIV and are also used to reduce the chance of HIV transmission from mother to child from 25% to less than 1%. It has been shown that some anti-HIV drugs, as an unwanted side effect, may have a toxic effect on the cells of the body. Some of the drugs can have an effect on different body systems that leads to mitochondrial (energy-producing part of body cells) dysfunction. When the mitochondria are not working properly (mitochondrial toxicity) the body can start to build up high levels of lactate (a byproduct of cell function). Also, when mitochondria are affected, they make molecules (small particles called free radicals) that can cause damage to DNA.

As people living with HIV are living much longer, there is increasing evidence that HIV may cause ‘early aging’ with several possible complications including endocrine system dysfunction and disturbance of normal bone metabolism which can cause low bone mass and increased fractures.

The purpose of this study is to investigate the effect of taking anti-HIV drugs on adults and children with HIV and on anti-HIV medication-exposed children who themselves do not have HIV (born to women with HIV), using two experimental laboratory tests. One is for mitochondrial DNA (mtDNA) and will test the level of function of the mitochondria and the other will look at the possible damage drugs may do to the length of DNA at the end of chromosomes. Additionally, we would like to better understand the bone health of adult women living with HIV using specialized x-ray scans in combination with several other measures. One of the x- rays is called a DXA scan (dual energy x-ray absorptiometry or bone scan) and is

129

used to look at the mass (size and structure) of your bones, one is called a pQCT scan (peripheral quantitative computed tomography) and is used to assess the strength of your bones and the other is called a high-resolution (shows finer detail) pQCT scan, which can only be used on your lower arm or leg but can show the inside of your bones with much better detail than the other scans. The pQCT and HRpQCT will be offered to women who participated in the Bone Health Study conducted at the Oak Tree Clinic from 2001-2003. Each of these tests together offer very low levels of radiation; less than a dental x-ray and approximately equal to the amount of radiation you would get from taking a plane from Vancouver to Calgary and back. Lastly, we would also like to understand the endocrine system health of both women and girls (age 12 and up) living with HIV by doing some extra blood tests and asking some detailed endocrine health questions.

YOU SHOULD NOT HAVE X-RAYS DONE IF YOU ARE PREGNANT

This study has up to FOUR possible options for you to consider: 1. OPTION A – Core/Basic Aging Study 2. OPTION B – Bone Health Study 3. OPTION C – Endocrine Health Study 4. OPTION D – Hepatitis C Treatment Study

STUDY ELIGIBILITY/SCREENING FOR OPTION A – Core Aging Study In order to be eligible to participate, you must: 1. Be either: • living with HIV (any age) and taking or have taken anti-HIV drugs • not living with HIV (any age) and exposed to anti-HIV drugs during your mother’s/your pregnancy 2. Agree to have a medical examination and detailed medication history done 3. Agree to have blood drawn from a vein at the same time as routine laboratory monitoring, if living with HIV – OR - blood drawn from a vein/finger poke for the study if not living with HIV and exposed to anti-HIV drugs during your mother’s pregnancy 4. Agree to have a mouthswab taken

You are not eligible to participate if you are: 1. Not living with HIV or have not been exposed to anti-HIV drugs during your mother’s/your pregnancy

130

STUDY ELIGIBILITY/SCREENING FOR OPTION B – Bone Health Study In order to be eligible to participate, you must: 1. Be an adult female living with HIV and age 19 or older 2. Agree to answer detailed questions about diet, exercise and any broken bones you may have had 3. Agree to have blood drawn from a vein for bone-specific lab tests at the same time as routine laboratory monitoring 4. Agree to have a bone density scan 5. Agree to have a pQCT scan and an HRpQCT scan if you were in the 2001- 2003 Bone Health Study

You are not eligible to participate if you are: 1. Pregnant 2. Unable to communicate/read in English where the presence of an interpreter is not available

STUDY ELIGIBILITY/SCREENING FOR OPTION C – Endocrine Health Study In order to be eligible to participate, you must: 1. Be a youth or adult female living with HIV, age 12 or older and having had your first menstrual period 2. Agree to answer detailed questions about your endocrine system health 3. Agree to have blood drawn from a vein for endocrine-specific lab tests at the same time as routine laboratory monitoring

You are not eligible to participate if you are: 1. Pregnant 2. Unable to communicate/read in English where the presence of an interpreter is not available

STUDY ELIGIBILITY/SCREENING FOR OPTION D – Hepatitis C Treatment Study In order to be eligible to participate, you must: 1. Be an adult female living with HIV and Hepatitis C (HCV), age 19 or older 2. Be planning to start HCV treatment with interferon-free medications 3. Agree to have blood drawn from a vein at the same time as routine laboratory monitoring, once within 3 months of starting treatment and a second time at 3 months after treatment completion 4. Agree to have a mouthswab taken 131

You are not eligible to participate if you are: 1. Planning to start HCV treatment with medications that include interferon 2. Unable to communicate/read in English where the presence of an interpreter is not available

STUDY ENTRY

If you decide to take part in this study, and you sign this consent form, the following outline describes the study schedule. You can choose to do any or all of the following:

Schedule of Visits for OPTION A – Core Aging Study

If you are living with HIV: You will have one study visit every 1 ½ to 2 years for up to 5 years. Each study visit will be linked to a routine clinic visit for your regular health care. At the same time as your regular blood work, a study blood sample and a mouthswab will be collected. At each of these visits the following will be done:  General health questions will be asked  A brief examination (as routine for your visit) will be conducted  A mouthswab will be collected Blood will be drawn from an arm vein for routine laboratory tests including HIV viral load, CD4 cell count, routine chemistry, routine hematology, and lactate. At the same time as your routine tests, blood will be drawn to test for mtDNA quantity, quality, mtRNA, DNA length, mitochondrial proteins, nutritionally relevant biomarkers (elements in the blood such as vitamin B12, vitamin D, folate and omega-3 fatty acids), inflammation biomarkers (elements in the blood that show inflammation is present) and if you are age 14 years and above, a few endocrine- health tests will also be done. We will also test for a series of viral infections that are very common in humans and can be in the body for a long time with no symptoms if the immune system is healthy. We will test for viruses such as those that cause chickenpox (varicella zoster virus or VZV), herpes (herpes simplex virus or HSV), mononucleosis (Epstein Barr virus or EBV), as well as cytomegalovirus (CMV), and the virus formerly known as Hepatitis G (GB virus C or GBVC). This testing may include antibody testing (a protein in the blood made in response to a foreign substance or a toxin - like an infection) and viral DNA (molecules in the blood that carry the virus’ genetic information) and viral RNA (molecules that 132

carry the virus’ instructions from the DNA into proteins) testing. Because we are using non-diagnostic methods of testing (for research use only), we will not be giving these results to you or your doctor. Twenty ml of blood will be collected for adults and children 6 years or older, 5-10 ml for young children 2 to 5 years old and 0.5-2 ml for infants less than 2 years. The study visits do require a small amount of additional time over a usual clinic appointment. About 15-20 minutes at each visit will be needed for study related activities. All of the scheduled blood work is routine except for the mtDNA quantity, quality, mtRNA, DNA length, mitochondrial proteins, nutritionally relevant biomarker, inflammation biomarker and viral infection tests; these tests require 20 ml (4 teaspoons of blood) for adults and children (6 years or older), 5-10ml (1 -2 teaspoons) for young children (2 to 5 years) and 0.5-2 ml for infants (less than 2 years), which will be drawn at the same time as the routine blood tests. The results of all the blood tests and mouthswabs will be charted in the study paperwork.

Baseline information will be extracted from the clinical record and, if available, will include information such as: health history, anti-HIV drug history, other drugs, any toxic exposures, etc. If you participated in the Bone Health Study done at the Oak Tree Clinic between 2001-2003 and consent to participate in OPTION B (below), we will also access the data from that study to provide a 10-year reflection of your bone health.

If you have been exposed to anti-HIV drugs in your mother’s/your pregnancy but are not living with HIV: You will have one study visit every 1 ½ to 2 years for up to 5 years. Each study visit will not necessarily be linked to a routine clinic visit. A study blood sample and a mouthswab will be collected.

At each of these visits the following will be done:  General health questions will be asked  A mouthswab will be collected  Blood will be drawn from a finger poke or an arm vein (20 ml of blood for adults and children 6 years or older, 5-10 ml for young children 2 to 5 years and 0.5-2 ml for infants less than 2 years).

The study visits do require some of your time. About 15-20 minutes will be needed for study related activities. Your blood will be tested for mtDNA quantity, quality, mtRNA, DNA length, mitochondrial proteins, nutritionally relevant biomarker,

133

inflammation biomarker and viral infection tests; these tests require 20 ml or 4 teaspoons of blood for adults and children (6 years or older), 5-10 ml (1- 2teaspoons) for young children (2 to 5 years) and 0.5-2 ml for infants (less than 2 years). The results of all the blood tests and mouthswabs will be charted in the study paperwork.

Baseline information will be extracted from the existing clinical record if available, and will include information such as: health history, anti-HIV drug exposure history, other drugs, any toxic exposures, etc.

Schedule of Visits for OPTION B – Bone Health Study

In addition to the visit details outlined for OPTION A above, you will have one study visit for the Bone Health Study. If we are successful with additional funding we may offer you a second visit approximately 1.5-3 years after the first. At each of these visits the following will be done:  A few additional bone-health specific blood tests will be done such as calcium, phosphate, and vitamin D (about 15 ml or 3 teaspoons of blood will be needed for these tests)  A detailed history of your diet, exercise and any broken bones will be asked  One x-ray is called a DXA scan (dual energy x-ray absorptiometry or bone scan) and is used to look at the mass (size and structure) of your bones. You will have had this scan does as part of your routine clinical care and we will collect the results from your chart.  If you were in the 2001-2003 Bone Health Study, two more x-rays will do done: one is called a pQCT scan (peripheral quantitative computed tomography) and is used to assess the strength of your bones and the other is called a high- resolution (shows finer detail) pQCT scan, which can only be used on your lower arm or leg but can show the inside of your bones with much better detail than the other scans. These 2 additional scans will be done at the Centre for Hip Health and Mobility (CHHM) at Laurel and 10th Avenue in Vancouver. Each of these x-rays offer very low levels of radiation; less than a dental x-ray and if you have all three of them the radiation you will be exposed to is approximately equal to the amount of radiation exposure you would get from taking a plane from Vancouver to Calgary and back.

YOU SHOULD NOT HAVE X-RAYS DONE IF YOU ARE PREGNANT

This study visit does require some of your time. Participating in OPTION B will

134

require about 90-120 minutes (120 if having scans at the CHHM) for study related activities.

Schedule of Visits for OPTION C – Endocrine Health Study

In addition to the visit details outlined for OPTION A above, you will have one study visit for the Endocrine Health Study. If we are successful with additional funding we may offer you a second visit approximately 1.5-3 years after the first. At each of these visits the following will be done:  A few additional endocrine-health specific blood tests will be done (about 15 ml or 3 teaspoons of blood will be needed for these tests). Whenever possible this blood sample will be collected in the morning in the fasting state (no food or drink after midnight the night before (water is ok))  You will be sent home with a kit to collect saliva samples two times. Once is at 10pm and then at 8am. We will also provide you with a stamped and addressed envelope to return the samples to us.  A detailed endocrine health history will be asked The study visits do require some of your time. Participating in OPTION C will require about 20-30 minutes for study related activities.

Participating in all of the options listed above (A, B and C including the scans at the CHHM) would require approximately 4-5 hours of your time and approximately 35 ml or 7 teaspoons of your blood for research. All of the study procedures can be completed over the course of 2-3 routine clinical visits and 2 blood collections and within a total of 6 months.

Schedule of Visits for OPTION D – Hepatitis C Treatment Study

You will have two study visits – the first visit will be within 3 months of starting your HCV treatment and the second visit will be 3 months after you finish your HCV treatment. Each study visit will be linked to a routine clinic visit for your regular health care. At the same time as your regular blood work, a study blood sample and a mouthswab will be collected. At each of these visits the following will be done (the same as described for Option A):  General health questions will be asked  A brief examination (as routine for your visit) will be conducted  A mouthswab will be collected 135

Blood will be drawn from an arm vein for the same laboratory tests and using the same blood volumes as described for Option A.

The study visits do require a small amount of additional time over a usual clinic appointment. About 15-20 minutes at each visit will be needed for study related activities. The results of all the blood tests and mouthswabs will be charted in the study paperwork.

Baseline information will be extracted from the clinical record and, if available, will include information such as: health history, anti-HIV drug history, other drugs, any toxic exposures, etc.

RISKS AND/OR DISCOMFORTS

Risks from Blood Drawing for OPTIONS A, B, C and D Blood drawing may cause some discomfort, bleeding or bruising where the needle enters the body. A small blood clot may form where the needle enters the body or there may be swelling in the area. Rarely, fainting or a local infection at the puncture site may occur.

Risks from the DXA and pQCT scans for OPTION B The DXA scan (done as part of routine care) and the pQCT scans are types of x- rays. All three x-rays will expose you to less radiation than a dental x-ray and approximately equal the amount of radiation you would be exposed to from taking a plane from Vancouver to Calgary and back (this is a natural form of radiation called ‘cosmic-radiation’, mainly from the sun, that is always around us and when you are in a plane flying high above the earth where the atmosphere is thinner, the radiation exposure is somewhat higher than when you are on the ground). Participants will be exposed to:  7.4 μSv (microsievert, a measure of radiation) from the DXA scan  0.72 μSV from the pQCT scan  <3 μSV from the HRpQCT scan This equals a maximum total of 11.12 μSv of radiation exposure from all three scans.

Radiation is not safe during pregnancy and you should not have these tests done if you are or think you may be pregnant.

We will make every effort to protect your privacy and confidentiality during the

136

study; however, it is possible that people may learn that you are participating in the study, and this may make you uncomfortable.

There are no other known risks associated with participation in this study.

BENEFITS for OPTION A: We will not be able to use any of the results from this study to tell you whether or not you are at an increased risk of these side effects and you will not receive any benefit from these results. However, knowledge gained from this study may, in the future, help others who are living with HIV.

BENEFITS for OPTION B: We will not be able to use any of the results from the PQCT and HRpQCT scans to tell you whether or not you are at an increased risk of bone problems because these tests are for research only and you will therefore not receive any benefit from these results. Your DXA however, done as part of routine care, and the bone-specific extra blood tests may be used by your doctor as part of your care and treatment (the extra blood tests are done in batches for research and there may be a delay of a few months before results are available). Knowledge gained from this study may, in the future, provide information about proper dosing of vitamin D and calcium for women living with HIV and may recommend changes to current guidelines for the screening and prevention of problems with bone health.

BENEFITS for OPTION C: The endocrine-specific extra blood tests may be used by your doctor as part of your care and treatment (the extra blood tests are done in batches for research and there may be a delay of a few months before results are available), yet you may not receive any benefit from this study. However, knowledge gained from this study may, in the future, help others who are living with HIV.

BENEFITS for OPTION D: You will not receive any benefit from participating in this parto of the study. However, knowledge gained from this study may, in the future, help others who are living with HIV and/or HCV.

NEW FINDINGS

You will be told of any new information learned during the course of the study that might cause you to change your mind about staying in the study. At the end of the study, you will be told when study results may be available and how to learn about

137

them.

VOLUNTARY PARTICIPATION

Your participation in this research study is strictly voluntary. You may choose not to participate in this study or to withdraw yourself from participation in the study at any time without providing any reasons for your decision. It will not influence the availability or quality of your present or future health care at this facility. Please take time to read this information carefully and to discuss it with your family, friends, and doctor before you decide.

COSTS AND REIMBURSEMENT for OPTION A: You will be paid $20.00 for each study visit to help with the cost of transportation, parking or childcare. No receipts are required for this and you will be paid at the time of each visit.

COSTS AND REIMBURSEMENT for OPTION B: You will be paid $20 in addition to the $20 indicated above for Option A (Option A is built into both Options B and C), for a total of $40, if you are not doing the 2 extra scans at the CHHM; and $30 in addition to the $20 indicated for Option A, for a total of $50 if you are doing the 2 extra scans for each study visit to help with the cost of transportation, parking or childcare. No receipts are required for this and you will be paid at the time of each visit.

COSTS AND REIMBURSEMENT for OPTION C: You will be paid $10 in addition to the $20 indicated above for Option A, for a total of $30, (Option A is built into both Options B and C) for each study visit to help with the cost of transportation, parking or childcare. No receipts are required for this and you will be paid at the time of each visit.

Dr Money and the other doctors involved in the study will not receive any money for your participation in this study. You should know that one of the investigators, Dr. Cote, is an inventor on a patent that has been filed by the University of British Columbia, on the mtDNA test used in this study. Therefore, she and UBC could one day receive a financial benefit from this research. You have the right to request more information about this financial benefit. You will not be eligible to receive any additional financial benefit from participating in this study even if the test should become commercialized.

COSTS AND REIMBURSEMENT for OPTION D: You will be paid $20.00 for 138

each study visit to help with the cost of transportation, parking or childcare. No receipts are required for this and you will be paid at the time of each visit.

IN CASE OF RESEARCH RELATED INJURIES

Signing this consent form in no way limits your legal rights against the sponsor, investigators, or anyone else, and you do not release the study doctors or participating institutions from their legal and professional responsibilities.

CONFIDENTIALITY

Your confidentiality will be respected.

You will be assigned a unique study number as a participant in this study. Only this number will be used on any research-related information collected about you during the course of this study, so that your identity [i.e. your name or any other information that could identify you] as a participant in this study will be kept confidential. Information that contains your identity will remain only with the Principal Investigator and/or designate. The list that matches your name to the unique study number that is used on your research-related information will not be removed or released without your consent unless required by law

Your tissue samples will be stored in a deep-freezer at the Cote laboratory at the UBC Hospital, Department of Pathology. The freezer is located in a locked room which is further located in the Cote laboratory which is locked after hours and on weekends. The custodian of these samples is Dr. Helene Cote. Samples are batched and tests are run in batches for quality assurance. All tissue samples are identified with your study ID only and will be stored for up to 25 years, except in cases where the Optional Tissue Banking Consent has been signed and then tissues may be stored for an indefinite period.

No information or records that disclose your identity will be published without your consent, nor will any information or records that disclose your identity be removed or released without your consent unless required by law.

Your rights to privacy are legally protected by federal and provincial laws that require safeguards to insure that your privacy is respected and also give you the right of access to the information about you that has been provided to the sponsor and, if need be, an opportunity to correct any errors in this information. Further details 139

about these laws are available on request to your study doctor.

ADDITIONAL INFORMATION

If you have any questions or need more information about this study at any time, please contact Dr Deborah Money 604 875 3459, or the study coordinator, Evelyn Maan RN at 604 767 5044.

If you have any concerns or complaints about your rights as a research participant and/or your experiences while participating in this study, contact the Research Subject Information Line in the University of British Columbia Office of Research Services by e-mail at [email protected] or by phone at 604-822-8598 (Toll Free: 1- 877-822-8598).

140

PARENT/GUARDIAN and PARTICIPANT CONSENT

 I have read and understood the participant information and consent form.  I have had sufficient time to consider the information provided and to ask for advice if necessary.  I have had the opportunity to ask questions and have had satisfactory responses to my questions.  I understand that all of the information collected will be kept confidential and that the result will only be used for scientific objectives.  I understand that my or my child’s participation in this study is voluntary and that myself or my child are completely free to refuse to participate or to withdraw from this study at any time without changing in any way the quality of care that myself or my child receive.  I understand that I am not waiving any of my or my child’s legal rights as a result of signing this consent form.  I understand that this study may provide no specific benefit to myself or my child.  I have been told that I will receive a dated and signed copy of this form.

I have read this form and I consent to participate in this study. I would like to participate in OPTION A – Core Aging Study OPTION B – Bone Health Study OPTION C – Endocrine Health Study OPTION D – Hepatitis C Treatment Study OPTION A, B, C and D (if applicable)

______Printed name and signature of participant Date

______Printed name of parent or legal guardian, relationship to child, and signature Date

______Printed name of second parent/legal guardian Date (if applicable), relationship to child, signature

______Printed name and signature of person obtaining consent Date

141

This consent was done in the following language: ______The person signing below acted as an interpreter/translator for the participant, during the consent process. ______Printed name and signature of person assisting in consent discussion Date

142