The Roles of Complement C4A and C4B Genetic Diversity and HLA DRB1 Variants on Disease Associations with Juvenile Dermatomyositis and Systemic Lupus Erythematosus

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By

Katherine E. Lintner, B.S.

Graduate Program in Molecular, Cellular and Developmental Biology

The Ohio State University

2016

Dissertation Committee:

Carlos Alvarez, Ph.D.

Heithem El-Hodiri, Ph.D.

Wael Jarjour, M.D.

Yusen Liu, Ph.D.

Chack-Yung Yu, D.Phil., Advisor

Copyright by

Katherine E. Lintner

2016

Abstract

Complement C4 is an immune protein with a wide range of effector functions, including disposal of apoptotic materials, clearance of immune complexes, and activation of the classical complement pathway resulting in cytolysis of microbes. Homozygous deficiencies of C4 or early complement components (C1q, C1r, or C1s), albeit rare, are strongly associated with the autoimmune disease systemic lupus erythematosus (SLE).

Much more common than a complete genetic deficiency is “low” gene copy number

(GCN) of C4, which varies among human genomes from two to eight copies. Low GCN of C4, specifically of the isotype C4A, is associated with SLE disease risk. However, it is known that C4A-deficient haplotypes in European Americans are in strong linkage disequilibrium (LD) with HLA allele DRB1*0301 on chromosome 6, which has been associated with other autoimmune diseases, including juvenile dermatomyositis (JDM). It remains a puzzle whether C4A deficiency, DRB1*0301, or both are responsible for the primary disease association because of the strong LD exhibited between the two genetic variants. We assessed GCNs for C4A, C4B, and HLA-DRB1 alleles in genetic risk of

JDM.

C4A deficiency was a risk factor for JDM independent of DRB1*0301, but the effect size was stronger when C4A deficiency and DRB1*0301 were present together.

JDM patients with C4A deficiency had higher prevalence of elevated serum muscle ii enzymes at disease diagnosis and elevated erythrocyte-bound C4-derived activation products (E-C4d). We also observed that C4A deficiency is a strong risk factor for pediatric SLE susceptibility as is the case for adult SLE reported previously, but the effect size was greater in pediatric populations. Our regression analyses of Caucasian

SLE patients and controls suggested that DRB1*0301 was likely secondary to C4A deficiency on disease susceptibility.

Given the common observation of C4A deficiency in patients afflicted with JDM or SLE, a logical step forward would be the establishment of C4-based therapy for these patients. In a series of luciferase reporter assays, we confirmed the 3’ long terminal repeat

(3’LTR) that flanks a retroviral insertion in intron 9 of C4 displays promoter activity, which could possibly initiate an antisense C4 transcript from exon 9 to exon 1. Moreover, we demonstrated the 3’LTR, in both orientations, enhanced C4 promoter activity in different human cultured cells. In healthy subjects from three different races, we showed a significant, negative correlation between the retroviral insert and C4 plasma protein levels. Cell culture assays and genotype-phenotype data indicate that the retroviral 3’LTR may interfere with C4 gene expression, possibly through antisense inhibition.

Several observations here require thoughtful and investigative extensions of this work. It is of interest to study the 3’LTR and other cis-acting elements that may affect C4 gene expression. Furthermore, increasing the sample size of the study population and analyzing complement and HLA genetic risk factors in autoimmune diseases in other races will allow for a greater understanding of disease pathways, identify any common or

iii race-specific genetic factors, and impact the field of human health by strengthening the prevention, treatment, maintenance, or cure of JDM, SLE, or other human immune- mediated diseases.

iv

Dedicated to Anthony Miller.

His love and encouragement wholeheartedly helped me to complete this document.

v

Acknowledgments

I wish to thank my advisor Dr. C. Yung Yu for supporting me during my time in his laboratory. He has given me every opportunity to succeed as a researcher, pushing me to limits that I sometimes did not know I could (or want to!) reach. He has taught me not only about complement genetics and immunology, but also about history, politics, and of course sports, including my favorites the Cincinnati Bengals, Cincinnati Reds, and The

Ohio State Buckeyes. Thank you, Dr. Yu, for sharing all of your knowledge and passion with me.

I also wish to thank my graduate studies committee- Carlos Alvarez, Heithem El-

Hodiri, Wael Jarjour, and Yusen Liu. Their thoughtful input, challenging questions, and dedicated time to my committee meetings and presentations have helped me to enhance my research and complete this degree. In addition to my committee members, I thank Dr.

John P. Atkinson (Washington University School of Medicine) for his insights and assistance with writing of my Chapter 1- Introduction.

As for the rest of the 4th floor Wexner crew, I wish to thank the following people:

Bi Zhou, for relentlessly performing DNA and cell isolation from patient blood samples and Southern blots to obtain the precious RCCX genotypes for all of our recruited human subjects; Carolyn Moore, for her administrative support; Jessi Buescher and Jennifer

vi

Bosse, for their friendship and willingness to lend a listening ear, when all I wanted to do was complain; and Dr. Loyal Coshway, for her friendship and moral support.

The bulk of my research and greatest accomplishment of my Ph.D. was the publication of our JDM study. Undoubtedly, this project took a number of years as well as a number of collaborators. I wish to thank Dr. Rabheh Abdul-Aziz and Dr. Charles

Spencer for their role in recruiting JDM patients. I also thank Dr. Abdul-Aziz for her assistance in helping me to obtain important clinical data. This manuscript would not exist without the collaboration of researchers in the Environmental Autoimmunity Group at the National Institute of Environmental Health Sciences. I wish to thank Dr. Lisa

Rider, Dr. Terrance O’Hanlon, and Dr. Frederick Miller for their efforts in recruiting

JDM patients and shipping blood and DNA samples for use in our study and for their essential input regarding data analysis and manuscript production.

I must say that I would not be where I am in life without the support of my family. My parents have always instilled in me the importance of education. In the words of my father, my career has been a “professional student.” I told you I would be finished someday, Pops! No more student. Onward to a career… Although my older sister Emily and twin brother Rodney live multiple states away, I always looked forward to visiting with them and their families. It was the best way to break away from school for a bit. My nieces are the light of my life.

Last but definitely not least, I want to express my sincere gratitude and love for my fiancé, Anthony Miller. We met in seminar class during our first year of graduate school. Although stressful and overwhelming at times, graduate classes were much more

vii enjoyable with him by my side. He has stood by my side, never judgmental or critical, during my candidacy exam, important academic presentations, stages of manuscript writing and, of course, during the process of writing this document. So many times I wanted to give up. So many times I would complain every day about the same thing. He showed me nothing but love, encouragement, and empathy. Someday (very soon, I hope!)

I can show him the same support as he begins to prepare his dissertation. Thank you,

Anthony, for loving me through this.

viii

Vita

2001-2005………………………………..Stephen T. Badin High School, Hamilton, OH

2005-2009………………………………..B.S. Biology, Walsh University, North Canton, OH

2011-2016………………………………..Graduate Research Associate, Molecular, Cellular, &

Developmental Biology, The Ohio State

University, Columbus, OH

Publications

Lintner K, Wu YL, Yang Y, Spencer C, Hauptmann G, Hebert L, Atkinson J, Yu CY .

(2016) Early components of the complement classical activation pathway in human

systemic autoimmune diseases. Frontiers in Immunology 7:36.

Chen JY, Wu YL, Mok MY, Wu YJ, Lintner K, Wang CM, Chung EK, Yang Y, Zhou

B, Wang H, Yu D, Alhomosh A, Jones K, Spencer CH, Nagaraja HN, Lau YL, Lau CS,

Yu CY. (2016) Effects of complement C4 gene copy number variations, size dichotomy

and C4A deficiency on genetic risk and clinical presentation of East-Asian SLE. Arthritis

and Rheumatology doi: 10.1002/art.39589

ix

Lintner K, Patwardhan A, Rider L, Abdul-Aziz R, Wu YL, Lundstrom E, Padyukov L,

Zhou B, Alhomosh A, Newsom D, White P, Jones K, O’Hanlon T, Miller F, Spencer C,

Yu CY. (2015) Gene copy number variations of complement C4 and C4A deficiency as genetic risk factors and in pathogenesis of juvenile dermatomyositis. Annals of the

Rheumatic Diseases doi :10.1136/annrheumdis. 2015.207762

Yu CY, Driest K, Wu YL, Lintner K, Patwardhan A, Spencer C, Rigby W, Hebert L,

Hauptmann G. (2013) Complement in rheumatic diseases. Encyclopedia of Medical

Immunology: Autoimmune Diseases. New York: Springer Science. P. 286-302

Yan J, Meng X, Wancket L, Lintner K, Nelin L, Chen B, Francis K, Smith C, Rogers L,

Liu Y. (2012) Glutathione reductase facilitates host defense by sustaining phagocytic oxidative burst and promoting the development of neutrophil extracellular traps. Journal of Immunology 188(5):2316-27

Field of Study

Major Field: Molecular, Cellular and Developmental Biology

x

Table of Contents

Abstract ...... ii

Acknowledgments...... vi

Vita ...... ix

Table of Contents ...... xi

List of Tables ...... xvi

List of Figures ...... xviii

List of Abbreviations ...... xxi

Chapter 1: Introduction ...... 1

1.1 Overview of the Complement System ...... 1

1.2 Immunity and Self-Tolerance...... 5

1.3 Gene Copy Number Variations (CNVs) and Complement Component C4...... 7

1.4 The Human Leukocyte Antigen (HLA) Region and Autoimmune Diseases ...... 10

1.5 Systemic Lupus Erythematosus (SLE) ...... 12

1.5.1 SLE Background...... 12

1.5.2 SLE and Serum Complement Levels...... 14

1.5.3 Cell-bound C4d as a Biomarker of Complement Activation for Humoral

Immunity, Alloreactivity, and Autoimmunity ...... 15 xi

1.5.4 Hereditary Genetic Deficiencies of Early Complement Components in SLE .. 17

Complement C1q Deficiency ...... 18

Complement C1r and C1s Deficiency ...... 21

Complement C2 Deficiency ...... 24

Complement C4 Gene Copy Number Variations and C4A or C4B Isotype Deficiency

...... 25

1.5.5 Acquired Deficiencies and Autoantibodies to Complement Components in SLE

...... 31

Anti-C1 Autoantibodies ...... 32

Anti-C1 Inhibitor Autoantibodies ...... 33

C3 and C4 Nephritic Factors ...... 34

1.6 Juvenile Dermatomyositis (JDM) ...... 35

1.6.1 JDM Background ...... 35

1.6.2 Complement in JDM ...... 36

Chapter 2: Gene Copy Number Variations (CNVs) of Complement C4 and C4A

Deficiency in Genetic Risk and Pathogenesis of Juvenile Dermatomyositis (JDM)...... 64

Abstract ...... 64

2.1 Introduction ...... 65

2.2 Materials and Methods ...... 67

xii

2.2.1 Study Populations ...... 67

2.2.2 Determination of total C4, C4A, and C4B Genotypes and Phenotypes ...... 68

2.2.3 Flow Cytometric Detection of Erythrocyte-Bound Complement Activation

Fragments ...... 68

2.2.4 HLA-DRB1 Typing ...... 69

2.2.5 Gene Expression Profiling ...... 69

2.2.6 Statistical Analysis ...... 70

2.2.7 Additional Declarations ...... 70

2.3 Results ...... 72

2.3.1 Gene CNVs of Total C4, C4A, and C4B in JDM and Race-matched Healthy

Controls ...... 72

Total C4 genes ...... 72

C4A and C4B Genes...... 72

2.3.2 HLA-DRB1 Alleles, C4A-GCN, and C4A Deficiency on Disease Risks of

JDM ...... 73

2.3.3 Levels of Erythrocyte-bound C4d (E-C4d) or C3d (E-C3d) in JDM and

Controls ...... 74

2.3.4 JDM Patients With Elevated Levels of Multiple Serum Muscle Enzymes had 75

Low GCN of C4A...... 75

2.3.5 Differential Gene Expression Profiling of JDM and Controls ...... 76 xiii

2.4 Discussion ...... 77

Chapter 3: Immunogenetic Studies of C4 Gene Copy Numbers and HLA-DRB1 Alleles in Systemic Lupus Erythematosus (SLE) in Caucasian Populations ...... 110

Abstract ...... 110

3.1 Introduction ...... 111

3.2 Materials and Methods ...... 113

3.2.1 SLE and Control Populations ...... 113

3.2.2 HLA-DRB1 Typing ...... 113

3.2.3 Determination of Total C4, C4A, and C4B Genotypes ...... 114

3.2.4 Statistical Analysis ...... 114

3.2.5 Additional Declarations ...... 115

3.3 Results ...... 116

3.3.1 C4 Gene Copy Number Distribution and HLA-DRB1 Allele Frequencies ... 116

3.3.2 C4A Deficiency in the Presence and Absence of HLA-DRB1*0301 ...... 118

3.3.3 Strong Genetic Association Between DR3 and C4A deficiency in Study

Populations ...... 119

3.3.4 Analysis of C4 and HLA-DRB1 Alleles in Adult SLE vs. Pediatric SLE

Populations ...... 120

3.4 Discussion ...... 120

xiv

Chapter 4: Regulation of Complement C4 Gene Expression ...... 129

Abstract ...... 129

4.1 Introduction ...... 130

4.2 Methods ...... 132

4.2.1 Cells and Reagents...... 132

4.2.2 Construction of Luciferase Plasmids ...... 133

4.2.3 Cell Transfection and Dual Luciferase Reporter Assays ...... 134

4.2.4 C4 Genotype and Phenotype Determination ...... 134

4.2.5 Human Subject Recruitment ...... 135

4.2.6 Additional Declarations ...... 135

4.3 Results ...... 136

4.3.1 Activity of the C4 Promoter Region ...... 136

4.3.2 Analysis of HERV-K(C4) LTR Activity ...... 137

4.3.3 In Vitro Treatment of C4-promoter and HERV-K(C4)-LTR Transfected Cells

...... 138

4.3.4 Variations in Human C4 Protein Expression In Vivo ...... 138

4.4 Discussion ...... 139

Chapter 5: Conclusion...... 156

References ...... 161

xv

List of Tables

Table 1 Complete deficiency of A, B, or C chain genes of C1q ...... 49

Table 2 Molecular defects and clinical presentation of complete deficiency for C1s or

C1r ...... 54

Table 3 Molecular defects and clinical presentations of complete deficiency of complement C4A and C4B ...... 57

Table 4 Demographic features and clinical characteristics of JDM patients ...... 95

Table 5 RCCX haplotypes (H1 and H2), complement C4B and C4A GCN and protein polymorphism, and HLA-DRB1 haplotypes (H1 and H2) in eight JDM patients ...... 96

Table 6 A comparison of plasma protein levels for complement C4 and C3 between JDM and race-matched controls ...... 97

Table 7 HLA-DRB1 Alleles and C4 Gene CNVs in JDM and controls ...... 98

Table 8 Elevation of serum muscle enzymes at disease diagnosis in JDM patients with and without C4A deficiency ...... 100

Table 9 RCCX structures and HLA-DRB1 genotypes of JDM patients for microarray studies ...... 101

Table 10 Patient demographics, C4 CNV, HLA-DRB1 genotypes, serum muscle enzymes and medications of 19 JDM patients (at recruitment) for microarray gene profiling experiments ...... 102

xvi

Table 11 Differentially expressed genes in blood samples of JDM patients...... 105

Table 12 Demographics of SLE study populations ...... 125

Table 13 Comparison of C4 isotype deficiencies and HLA allele frequencies in SLE and controls ...... 126

Table 14 Effect sizes of homozygous vs. heterozygous deficiencies of C4A ...... 128

Table 15 Primers used for C4 promoter and HERV-K(C4) 3’LTR cloning ...... 153

Table 16 Plasmid constructs designed for luciferase reporter assays ...... 154

xvii

List of Figures

Figure 1 The activation pathways of the complement system ...... 38

Figure 2 Gene size dichotomy and gene copy number variation of complement C4 ...... 40

Figure 3 Genetic architecture of the HLA region in the human genome ...... 42

Figure 4 Typical serial serum protein profiles of complement C4 and C3 in human SLE patients ...... 43

Figure 5 SLE patients with a homozygous deficiency of early components for the classical pathway of complement activation ...... 45

Figure 6 Comparisons of frequencies for total C4, C4A, and C4B gene copy number groups in SLE (red) and controls (blue) ...... 47

Figure 7 Race-specific distribution patterns of RCCX modules in human populations .. 48

Figure 8 Variations of C4 haplotypes and gene copy numbers (GCNs) of total C4, C4A and C4B in JDM subjects and race-matched healthy controls ...... 82

Figure 9 Genetic diversity of complement C4 ...... 84

Figure 10 Variations of C4 haplotypes and gene copy numbers (GCNs) of total C4, C4A and C4B, and C4 protein polymorphisms in eight selected JDM subjects ...... 86

Figure 11 Relationship between C4 plasma protein levels plotted versus total C4 GCN in

95 European American JDM subjects ...... 88

Figure 12 Erythrocyte-bound E-C4d and E-C3d in JDM patients and controls ...... 89

xviii

Figure 13 Gene expression profiling of PAXgene blood RNA in JDM and healthy controls ...... 91

Figure 14 Correlations of transcript levels for interferon-stimulated genes IFI17 (left) and IFI44 (right) with polymorphonuclear cells present in leukocytes of JDM blood samples ...... 93

Figure 15 Correlation of CCR5 transcripts with lymphocytes present in leukocytes of

JDM blood samples...... 94

Figure 16 Sample of HLA-DRB1 genotyping gel ...... 122

Figure 17 Comparison of C4 isotype deficiencies and HLA allele frequencies in SLE and controls ...... 123

Figure 18 Overview of multivariate regression analysis ...... 124

Figure 19 Detection of C4 expression in various tissues ...... 142

Figure 20 Dichotomous size variation of human C4 gene ...... 143

Figure 21 pGL4.10[luc2] vector map ...... 144

Figure 22 Promoter sequence of human C4 and strategy for making promoter deletion constructs...... 145

Figure 23 Example of construct cloned for luciferase assays ...... 146

Figure 24 C4 reporter gene activity in 8505C cultured cells ...... 147

Figure 25 C4 reporter gene activity in HepG2 cultured cells ...... 149

Figure 26 C4 reporter gene activity upon IFN-gamma treatment ...... 150

Figure 27 Temporal expression of C4 promoter activity in response to IFN-gamma treatment ...... 151

xix

Figure 28 Variations of human C4 plasma protein and GCN of C4-long genes (C4L) in different races...... 152

xx

List of Abbreviations

ACR American College of Rheumatology

AH50 complement alternative pathway hemolytic activity 50%

ANA anti-nuclear antibodies

AP alternative pathway of complement activation

BMI body mass index

C1-inh complement C1-inhibitor

C4 complement component 4

C4bp C4 binding protein

CBCAPS cell-bound complement activation products cDNA complementary DNA

CH50 total complement hemolytic activity 50%

CI confidence interval

CNV copy number variation

CP classical pathway of complement activation

CR1 complement receptor 1

DAF decay-accelerating factor

DNA deoxyribonucleic acid

E-C3d erythrocyte-bound C3-derived activation fragment

E-C4d erythrocyte-bound C4-derived activation fragment

xxi

FH complement factor H

GCN gene copy number

GWAS genome-wide association study

HERV-K(C4) human endogenous retroviral insertion of lysine (K) family in C4

HLA human leukocyte antigen

IC immune complex

IgG immunoglobulin G

IgM immunoglobulin M

IRB Institutional Review Board

IFN interferon

IQR interquartile range

JDM juvenile dermatomyositis

LD linkage disequilibrium

LTR long terminal repeat

MAC membrane attack complex

MASP1/2 mannan-binding lectin associated serine peptidase 1 or 2

MBL mannan-binding lectin

MFI median fluorescent intensity

MHC major histocompatibility complex

OR odds ratio

PCR polymerase chain reaction qPCR quantitative polymerase chain reaction

xxii

RCCX RP-C4-CYP21-TNX

SD standard deviation

SLE systemic lupus erythematosus

SLEDAI SLE disease activity index

SLICC Systemic Lupus International Collaborating Clinics

SNP single nucleotide polymorphism

TNF-a tumor necrosis factor alpha

xxiii

Chapter 1: Introduction

1.1 Overview of the Complement System

The complement system is a humoral recognition and effector system that facilitates in the elimination of invading pathogens. The components of the complement system coordinate in a sequential series of reactions beginning with C1, followed by activation of C4, C2, C3, C5, C6, C7, C8, and C9, which are numbered in the order of their discovery. The activation pathways of the complement system converge at complement C3 and progress to the formation of the membrane attack complexes (MAC or C5b-C9) on a target membrane, which includes the creation of pores across the target membrane, inducing cell lysis, loss of cytoplasm, and osmotic lysis. The cascades of the complement system are mediated and adjusted according to the type of initiator and the microenvironment in which complement activation is occurring. Proteins of the complement system cooperate and coordinate to differentiate among invading microbes, immune complexes, apoptotic cells, cellular debris, and physiologic host cells (Ricklin et al. 2010; Sturfelt and Truedsson 2012; Yu et al. 2014).

Three distinct pathways trigger cascades of activation through proteolysis of zymogens or precursors present in the circulation (Figure 1). The classical pathway (CP) is predominantly triggered by IgM or IgG immune complexes. The formation of antigen- antibody complexes exposes binding sites for complement C1q on Fc-regions of

1 immunoglobulins IgG or IgM, triggering the assembly and activation of the multi- molecular C1 complex, C1q-C1r2-C1s2 (Pathway 3, Figure 1). Conformational changes in the C1 complex are induced upon binding of C1q to antibody, leading to activation of the serine protease subunits C1r and then C1s. As a result, C1s next activates complement C4 and complement C2, leading to the formation of the classical pathway C3 convertase

(abbreviated C4b2a). Following the early components’ activation, later components of the complement cascade form the membrane attack complex (MAC), which perturbs membranes, including the creation of pores across the target membrane, inducing cell lysis, loss of cytoplasm, and osmotic shock.

In the mid-1950s, Pillemer and colleagues of Case Western Reserve University observed that complement activation could occur in the absence of a specific antibody

(Pillemer et al. 1954). The existence of such an “alternative” pathway (AP) of activation was challenged but was confirmed more than two decades later (Fearon et al. 1974).

Specific proteins involved in the AP are named factors, such as factor B, factor D, factor

H, and factor P (properdin). This pathway (Pathway 1, Figure 1) is initiated by a “tick- over” mechanism, in which a small proportion of complement C3 in the circulation is continuously hydrolyzed at slow rate (~1-2 %/hour) by water to form C3(H2O). C3(H2O) binds to factor B, which is activated by factor D, to form C3(H2O)Bb. C3(H2O)Bb accordingly acts as a relatively labile C3 convertase, constantly initiating C3 cleavage.

Properdin stabilizes the short-lived C3 convertase. Under the appropriate circumstances, a C5 convertase (C3bBbP) is formed, and the cascade progresses to MAC formation on a foreign cell surface, similar to that of the CP. The binding of P to C3bBb on a microbial

2

(or protected) surface will stabilize and protect the convertase from inactivation by regulatory proteins, thereby enhancing the convertase activity. The AP actually represents an ancient mechanism of innate immune host defense. The tick-over mechanism of complement activation enables a continuous surveillance for the host, executing the first line of defense against foreign invaders. With the development of a circulatory system, a system of host defense that both worked in seconds and was pathogen-destructing became mandatory.

A third pathway of complement activation involves the specific pattern recognition of biomolecules. One strategy for organisms to achieve species-specific diversity is by modification of biomolecules such as glycolipids and glycoproteins with different complexities of sugars. Typically, carbohydrate moieties on glycoproteins among vertebrates consist of complex sugars with secondary modifications (biantennary type) and ending with sialic acids. By contrast, the carbohydrate moieties in prokaryotes generally consist of simpler polymers of saccharides such as mannose. Pattern recognition of biomolecules is a universal theme of innate immunity. This pathway of complement activation is initiated by the binding of pattern recognition molecules including mannan-binding lectin (MBL) or ficolins to a bacterial membrane that expresses arrays of simple carbohydrates such as mannose and N-acetylglucosamine

(Jensenius 2005; Degn et al. 2012). Such binding triggers the assembly of MBL/MASP2 and ficolin/MASP2, or MBL/MASP1 and ficolin/MASP1 complexes (Pathway 2, Figure

1). MASP2 and MASP1 are both serine proteases. MASP2, associated with MBL or

3 ficolin, activates both C4 and C2. As a result, a C3 convertase identical to that generated by the CP is formed.

Thus, all three complement activation pathways pass through the focal point on the activation of C3 to C3a and C3b, and then C5a and C5b, leading to the assembly of sublytic or lytic complexes on target membrane. It is noteworthy that all three activation pathways can be amplified by the positive feedback mechanism of the AP. In addition to cell lysis, effects of complement activation include opsonization to enhance phagocytosis of target cells, clearance of apoptotic bodies, solubilization and removal of immune complexes, stimulation of cytokine production, and anaphylatoxin-mediated effects. To summarize, the complement system has been designed in evolution primarily to activate on the membranes of bacteria and certain viruses. Opsonization via C3b and cellular activation via the anaphylatoxins C3a and C5a are its two primary functions.

Because the activated complement components C4b, C3b, and C5b67 can attach to any nearby cell surfaces including host cells, regulatory mechanisms have evolved to restrict complement activation to damaged self and foreign targets. Inherently, all activated complement proteins spontaneously undergo intrinsic decay or inactivation when not stabilized by other pathway components or factors. In addition, several regulatory proteins in plasma or on the cell membrane can dissociate (decay) multi- molecular (activated) complexes and also proteolytically degrade the anchor proteins such as C4b and C3b. Upon initiation of classical or lectin pathways, C1-inhibitor (C1- inh) is a serine protease inhibitor that mimics the substrates for C1s in the C1 complex, and MASP2 or MASP1 in the MBL or ficolin complex. C1-inh forms a complex with

4 activated C1r and C1s, leading to the dissociation of the enzymatic C1r/C1s subunit from the recognition C1q subunit and preventing further activation of C4 and C2.

Importantly, there are also key strategies of regulation that act upon the assembly and stability of the C3 convertases. Fluid phase proteins C4b-binding protein (C4bp) and factor H (FH), and membrane proteins complement receptor 1 (CR1) and decay accelerating factor (DAF) all dissociate the recognition and enzymatic subunits of C3 convertases. Moreover, C4bp, CR1, and membrane cofactor protein (MCP) serve as cofactor proteins for factor I-mediated degradation of C4b, while FH, CR1, and MCP each serves as a cofactor for the factor I-mediated proteolysis of C3b. Notably, C4bp and

FH recognize exposed host glycoproteins with glycosaminoglycans and sialic acids. The presence of such regulatory molecules on self-surfaces, but absence from most foreign particle surfaces, allows the regulators to prevent activation on host tissues while restricting complement activity to designated, foreign targets. Dysfunctional or uncontrolled complement activation can lead to destruction of body cells, overt release of inflammatory mediators, and tissue damage, as evidenced by clinical complications experienced by patients with the autoimmune disease systemic lupus erythematosus

(SLE).

1.2 Immunity and Self-Tolerance

Innate immunity is known as the first line of defense. This response is ingrained and is the initial response to invading microbes. In many instances, the response eliminates the microbe, but upon reinfection, the response is the same (i.e., no memory

5 response; not enhanced). Components of the innate immune response include: physical barrier (skin, mucosal epithelial); blood constituents (complement, inflammatory mediators); phagocytic cells (neutrophils, macrophages, natural killer cells); and cytokines, which regulate and coordinate the activity of immune cells.

Unlike innate immunity, adaptive immunity has memory and is stimulated by previous exposure to infectious agents; in other words, each time one is exposed to a pathogen, the resulting response increases in magnitude and also is initiated quicker with each successive exposure. Adaptive immunity is mediated by B and T cells and the production of specific antibodies.

Jan Klein once defined immunology as the science of self-nonself discrimination

(Klein 1982). A fundamental property of a normally-functioning immune system is self- tolerance, or unresponsiveness to self-antigens. Self-tolerance is induced when immune cells that recognize self-antigens are either deleted or inactivated and therefore unable to stimulate a complete immune response. Failure of self-tolerance results in immune reactions against one’s own tissues, or autoimmunity. The potential for autoimmunity exists in all individuals because many self-antigens are readily accessible to immune cells. However, it has been found that many individuals are more or less susceptible due to environmental and/or genetic factors. The exact mechanisms by which self-tolerance is disrupted remain unclear and would provide understanding of the pathogenesis of autoimmune diseases.

6

1.3 Gene Copy Number Variations (CNVs) and Complement Component C4

Genetic diversities in mammals exist in multiple forms. Heritable genetic variations include small and local changes, such as single nucleotide polymorphisms

(SNPs), mini-insertions and deletions (indels) of genomic DNA, expansions and contractions of variable tandem repeats, and microsatellite DNAs. Larger variations include insertion and deletion of retrotransposable elements, inversion and translocation of sub-chromosomal fragments that change orders of genetic markers, dichotomous and continuous segmental duplications that engage DNA sequences in the range of kilobases to megabases, and varying number of sex chromosomes. Together with epigenetic variations, such as posttranslational modifications of histones (i.e., acetylations, methylations, and phosphorylations) and methylations of DNA, quantitative and qualitative, plus spatial and temporal, variations in the expression of gene products are achieved. Such genetic and epigenetic variations not only contribute to specificities in various growth and development processes, but also provide a large repertoire of variants to facilitate adaptations to environmental fluctuations and serve as the driving force for evolution. Until recently, it was a common belief that SNPs were the major contributor to genetic variability. However, genome-wide studies in the past decade employing array comparative genomic hybridization have uncovered frequent, inter-individual deletions and insertions of genomic DNA sequences >1 kb in size, or copy number variations

(CNVs) (Iafrate et al. 2004; Sebat et al. 2004). CNVs < 1Mb are not usually detectable by conventional cytogenetic techniques. Structural variations involving CNVs can account for nearly 70% of differences in DNA sequence between two human subjects (Girirajan

7 et al. 2011). While many CNVs are located outside of gene-encoding DNA, some CNVs produce extensive genetic variation which may alter quantity as well as function of the resulting protein. Inter-individual variations in common CNVs, particularly amongst genes involved in metabolism and immune effector function, have been associated with susceptibility to several complex diseases and disorders: Low copy numbers of the salivary amylase gene AMY1 is associated with predisposition to obesity (Falchi et al.

2014); deletion of the complement regulatory related genes CFHR3-R1 bears protection against age-related macular degeneration (Hughes et al. 2006) but is a risk factor for atypical hemolytic uremic syndrome (Zipfel et al. 2007); increased copy number of beta- defensin genes (DEFB4, DEFB103, DEFB104) are associated with psoriasis (Hollox et al. 2008); polymorphic variants and low GCN of Fc receptor family genes are associated with chronic inflammatory diseases of autoimmune origin (Bournazos et al. 2009). In particular, our laboratory has demonstrated associations of variations in GCN of C4 with

SLE (Yang et al. 2007), JDM (Lintner et al. 2015), Type 1 Diabetes (Kingery et al.

2012), and Rheumatoid Arthritis (Rigby et al. 2012).

The complement C4 gene is located in the HLA class III region on the short arm of chromosome 6, telomeric of the C2 gene (Carroll et al. 1984). Remarkably, two to eight copies of C4 genes can be present in a diploid genome (Yang et al. 1999; Wu et al.

2007) (Figure 2). Such common CNVs are a rare phenomenon in mammalian genetics, as they deviate from the conventional diploid genome of two (one from each parent). Here, one to four copies of nearly identical genes co-exist on a single chromosome, thereby creating a gene dosage effect for a quantitative phenotype. Segmental duplications for C4

8 always include the RP (STK19) gene upstream of C4, and the downstream genes CYP21 and TNX (RCCX), which unlike C4, are always duplicated as fragments, or pseudogenes

(Yang et al. 1999). Each C4 gene either encodes for an acidic C4A or a basic C4B protein, with only four amino acid changes (PCPVLD 1120-1125 for the C4A isotype and

LSPVIH for the C4B isotype), but this results in substantial differences in chemical reactivity for peptide and carbohydrate antigens (Isenman and Young 1984; Law et al.

1984; Yu et al. 1986; Kidmose et al. 2012). C4A favors binding to amino groups (i.e., immune complexes), while C4B favors binding to hydroxyl or carbohydrate-rich groups.

Approximately 40 protein variants for complement C4 have been documented (Mauff et al. 1998). Technically, it is noteworthy to mention that the differential binding to hydroxyl- or amide- group containing substrates/immune complexes, or hemolytic activities between activated C4A and C4B can be readily demonstrated using purified component proteins or proteins resolved by gel electrophoresis, but not from sera when many regulatory proteins are present (Reilly 1999; Yang et al. 2003). Additionally, C4 exists in a long form and short form. The long gene is 20.6 kb, and the short gene is 14.2 kb. In a long C4 gene, an endogenous retrovirus HERV-K(C4), which is 6.4 kb in size, integrated into C4 intron 9. Among healthy subjects of European ancestry, 76% of C4 genes belong to the long form and 24% belong to the short form (Blanchong et al. 2000).

The primary site for C4 biosynthesis is in the liver. However, multiple tissues also synthesize C4, presumably for local consumption, particularly after stimulation by interferon-gamma (Mitchell et al. 1996). A thioester bond is present but hidden in native

C4. A proteolytic cleavage by activated C1s removes a 74 amino acid C4a peptide and

9 leads to a remarkable change of conformation in C4 (Mortensen et al. 2015).

Consequently, the protected thioester bond becomes exposed to the exterior. In activated

C4B, one of the four isotypic residues (Histidine-1125) serves as a catalyst and facilitates a rapid nucleophilic attack, resulting in formation of a covalent ester linkage between

C4B and the target surface (Carroll et al. 1990; Dodds et al. 1996). In activated C4A,

Aspartic Acid-1125 is present instead of Histidine-1125, and such a catalytic reaction does not occur. Instead, C4A reacts effectively with an amino group on an immune complex or a protein molecule to form a covalent amide bond. Such a difference in chemical reactivity appears to diversify the functional roles of C4A and C4B in the clearance of immune complexes and the propagation of activation pathways, respectively.

Quantitative real-time PCR (qPCR) is one technique to absolutely identify gene copy number (GCN) of total C4. Quantitative real-time PCR (qPCR) assays were developed using TaqMan chemistry and based on sequences specific for C4A and C4B genes, structural characteristics corresponding to the long and short forms of C4 genes, and the breakpoint region of RCCX modular duplication (Wu et al. 2007). Relatively speaking, qPCR is a cost-efficient, accurate method and requires only a small amount of

DNA (~30ng per amplicon, in duplicate).

1.4 The Human Leukocyte Antigen (HLA) Region and Autoimmune Diseases

The HLA region (or MHC, major histocompatibility complex), located on the short arm of chromosome 6, is one of the most gene-dense and polymorphic stretches of

DNA (Figure 3). Contributions of multiple variants at this polymorphic locus have been

10 associated with various autoimmune, infectious, and inflammatory diseases (Gough and

Simmonds 2007; Fernando et al. 2008; Trowsdale and Knight 2013; Lenz et al. 2015) and therefore, the HLA region is one of the most extensively studied regions in the human genome. Recent advances in technology and genomic analysis methods, including SNP genotyping, GWAS, and high-throughput sequencing have been utilized to uncover such disease associations. The first-discovered association of HLA with disease was in 1967 when HLA-B antigens were found at increased frequency in patients with Hodgkin’s lymphoma (Amiel 1967). Serological typing of this sort uncovered many more HLA- disease associations. This seemed fitting because the HLA region encodes for several proteins that play key roles in the immune system. The unique amino acid sequences of

HLA class I and class II alleles encode for molecules responsible for antigen presentation and processing and determine, to a large extent, the nature of an individual’s response to both foreign and self-antigens. The class III genes are heterogeneous in structures and function, and include genes encoding for components of the complement system C4, C2 and factor B, and for cytokines such as tumor necrosis factor-α (TNF-α).

Variation within the HLA has been found to be associated with almost every autoimmune disease. However, several hallmarks of the genetic architecture of the HLA alleles create obstacles for interpreting HLA-disease associations. A) Extensive linkage disequilibrium (LD) that exists among alleles throughout this locus make it difficult to determine whether variant(s) are responsible for the primary disease association or are simply present due to LD; B) the extent of sequence variability exhibited by some HLA genes make routine typing strategies extremely laborious and time-consuming, which is

11 not ideal for large case-control analyses; C) often there is more than one association signal from the HLA region for a specific disease; D) specific HLA haplotypes can be associated with risk of one disease yet be associated with protection for another disease.

For all of these reasons noted above, most published studies are, at best, association studies, and the pathogenic mechanisms underlying HLA-disease associations have yet to be determined.

Of interest to us, it is noteworthy to mention that certain ancestral haplotypes encompassing the HLA-DRB1 region can be in strong LD with specific C4 haplotypes, given that the C4 locus is 350 kb away in the same chromosomal region. Previous studies revealed that class II genes DRB1 allele *0301 and DQA1 allele *0501, class III gene

TNFA-308A allele, and class I gene variants B*08 and A*01 are in strong linkage disequilibrium among human subjects of European descent and this is described as the ancestral haplotype AH8.1 (Friedman et al. 1983; Dawkins et al. 1999; Pachman et al.

2001). Also present in AH8.1 is a single C4B gene but the absence of a C4A gene

(Dawkins et al. 1999; Yu and Whitacre 2004; Horton et al. 2008). That is, AH8.1 is deficient of the C4A gene.

1.5 Systemic Lupus Erythematosus (SLE)

1.5.1 SLE Background

SLE is a systemic autoimmune disease characterized by the presence of multiple autoantibodies against self-nuclear antigens (ANA). SLE affects mainly females, especially those of child-bearing years. The overall reported prevalence of SLE in the

12 population is 20-150 cases per 100,000, although prevalence rates are greater for

Caucasian women (164 per 100,000) and even greater for African American women (406 per 100,000) (Chakravarty et al. 2007; Pons-Estel et al. 2010).

Although the name implies and describes skin lesions, SLE is a systemic disease, affecting every vital organ via autoimmune response. The following 11 criteria, of which

4 are needed for a formal SLE diagnosis, were originally established by the American

College of Rheumatology (ACR) in 1971 and updated in 1982 and again in 1997: malar rash, discoid rash, photosensitivity, oral ulcers, arthritis, serositis, kidney disorder, neurologic disorder, blood disorder, immunologic disorder, and abnormal antinuclear antibodies (Fries and Siegel 1973; Tan et al. 1982; Hochberg 1997). Most recently, the

Systemic Lupus International Collaborating Clinics (SLICC) revised the ACR criteria to include hypocomplementemia, or low complement levels in serum (Petri et al. 2012). In addition, the SLICC classification criteria require at least one clinical and one immunologic criterion to meet SLE diagnosis, which aims to exclude clinical criteria alone or positive serologic tests alone to be considered a diagnosis of SLE.

SLE is a multifactorial and complex disease. As with most complex diseases, combinations of genetics with environmental factors probably contribute to disease.

Environmental influences such as Epstein barr viral infections and DNA hypomethylation attributed to certain medications have been associated with the development of SLE

(Tsokos 2011). Although in rare cases SLE may develop from a single complete genetic deficiency, the disease most commonly results from a complex combination of genetic variants and SNPs (Moser et al. 2009) GWAS have identified approximately 50 gene loci

13 but only accounts for a small percent of heritability and susceptibility to SLE (Graham et al. 2009; Boackle 2013; Rullo and Tsao 2013).

Complement activation products and genetic polymorphisms have been observed and shown to contribute to inflammation and tissue damage in SLE. Next is a brief summary of findings of complement activation and involvement in SLE.

1.5.2 SLE and Serum Complement Levels

SLE patients commonly present with evidence of complement consumption leading to low serum levels of C4 and C3 (Elliott and Mathieson 1953; Lewis et al.

1971). Initially, up to one-half of SLE patients will have low C4 and C3. In most established patients, serum C4 levels are biomarkers for lupus disease activity; low levels correlate with a flare, while normal levels correspond with remission (Birmingham et al.

2010). Longitudinal studies of serum C4 protein levels in SLE patients revealed different expression profiles characterized by three distinct groups (Figure 4) (Wu et al. 2006). The first group exhibited persistently low C4 levels throughout the course of the study, and many of these patients had a low copy number of C4 genes. The second group featured periodic fluctuations of C4 that paralleled disease activity, while the third group had normal C4 levels most of the time. The typical pattern in active SLE patients is that both

C4 and C3 are low simultaneously. However, exceptions occur. C3 levels are usually 3-6 fold higher than C4 levels, therefore consumption of complement by immune complexes could reduce C4 below normal but leave C3 in the normal range. With a positive response to treatment, both C4 and C3 levels will rise. As noted, up to one-half of SLE

14 patients will present with serum C4 and C3 in a normal range, which does not rule out a lupus diagnosis. The CH50 and AP50 measure the lysis of red blood cells by the respective pathway and thus are functional tests. Furthermore, in vivo complement activation can also be assayed by testing for complexes or split products formed during complement activation (Atkinson and Yu 2015).

Copy number variation (CNV) of C4 can affect serum C4 protein concentrations.

In most Caucasian populations, about 60% of individuals have four copies of the C4 gene, while 28.5% have three (or less), and 12.5% have five (or more). In lupus, the number of patients with three or less C4 genes may increase to 42.2% and is therefore considered a risk factor for SLE (Yang et al. 2007). If an individual has low copy of C4 genes, the baseline C4 antigenic level may be 12 to 18 (approximately 6 to 8 mg/dl per copy of a C4 gene). In this situation, it doesn’t take much activation to lower the C4 out of the normal range. Additionally, a subject’s body mass index (BMI) is positively associated with serum C4 or C3 protein concentrations (Yang et al. 2003; Saxena et al.

2009). All things considered, the care of each patient must be individualized. Repeated, longitudinal serum measurements of C3 and C4 are usually clinically utilized.

1.5.3 Cell-bound C4d as a Biomarker of Complement Activation for Humoral Immunity,

Alloreactivity, and Autoimmunity

In the past decade, cell-bound levels of processed complement activation products

(CBCAPS), especially erythrocyte-bound C4 (E-C4d), has been proposed to assist in the diagnosis and clinical monitoring of SLE (Manzi et al. 2004; Putterman et al. 2014). E-

15

C4d is a stable, proteolytic end fragment of C4 which is covalently bound to the surface of erythrocytes following activation of the CP or MBL pathway. On cell surfaces, the activated C4b is processed to the cell-bound C4d with the release of soluble C4c through factor-I mediated proteolysis in the presence of a cofactor (i.e., CR1 on the plasma membrane or C4bp from plasma). In 2004, Manzi et al. found that erythrocytes from

SLE patients had markedly higher levels of E-C4d when compared to healthy controls or patients with other diseases (Manzi et al. 2004). Additional studies explored the biomarker utility of CBCAPS on T and B lymphocytes, platelets, and reticulocytes in

SLE (Liu et al. 2005; Calano et al. 2006; Navratil et al. 2006; Liu et al. 2010). The levels of C4d bound to the membrane of these cells were significantly higher in SLE than healthy controls or patients with other diseases. In a study of 304 SLE patients, 285 patients with other rheumatic diseases and 205 healthy controls, Putterman and colleagues reported that CBCAPS on erythrocytes or B cells had higher diagnostic sensitivity than standard complement levels (serum C3 and C4) and anti-dsDNA measurements when distinguishing between SLE and non-SLE, suggesting that CBCAPS could be more specific and sensitive biomarkers for diagnosis and prognosis of SLE

(Putterman et al. 2014). However, despite the relative simplicity and low cost of serum

C4 and C3 measurements, it remains to be seen how assessment of cell-bound C4d and

C3d will contribute to the clinical care of patients with SLE.

16

1.5.4 Hereditary Genetic Deficiencies of Early Complement Components in SLE

Deficiencies or genetic polymorphisms of early complement components are strongly associated with increased risk of developing SLE or a lupus-like disease (Figure

5). Complement deficiencies are hypothesized to be associated with increased susceptibility of SLE for several reasons. Functionally, the role of complement includes the identification, opsonization, and proper disposal of apoptotic cells and immune complexes formed regularly between antibodies and foreign or self-antigens (Walport

2001; Ricklin et al. 2010). An inability to efficiently clear apoptotic cells could render them a source of autoantigens and thereby drive autoantibody production. Impaired clearance of immune complexes and “self” debris provides a logical explanation for complement deficiency in the induction of SLE (Navratil et al. 2001). While there are multiple complement pathways to assist the host with clearance of these types of materials that accumulate continuously in healthy subjects, the CP is essential through at least C4, and to a lesser degree C2, to properly handle and dispose of immune complexes and apoptotic debris.

Another hypothesis that attempts to explain the association of complement deficiency with SLE suggests that the complement system is involved in immune tolerance (Carroll 2004). In other words, early components of the CP are engaged in the

“cross-talk” to the adaptive immune system to achieve tolerance against self-antigens, or in the discrimination of “self” versus “non-self”. A complement deficiency that results in a breach of self-tolerance provides a reasonable explanation for association with SLE

(Prodeus et al. 1998; Carroll 2000). Normally, complement receptor 1 (CR1/CD35) and

17 complement receptor 2 (CR2/CD21) on follicular dendritic cells of peripheral lymphoid tissues (such as the spleen and lymph nodes) bind to and deliver self-antigens coated with complement fragments to the autoreactive B-cells, which are anergized or kept away from germinal center reactions. Previously, it was demonstrated in mouse models that complement was necessary for the elimination of self-reactive lymphocytes during the maturation of the immune system (Prodeus et al. 1998). Specifically, Prodeus et al. showed that deficiency of either CR1/CR2 or of C4 in a well-defined mouse model of peripheral tolerance resulted in high titers of anti-nuclear antibodies (ANAs) and a severe lupus-like disease.

These two concepts of debris clearance and regulation of self-tolerance are not exclusive and likely overlap, or together explain the development of SLE in humans deficient in early components of the CP. In the following sections we will describe genetic complement deficiency states associated with SLE for each component of the early classical pathway (C1q, C1r, C1s, C4 and C2).

Complement C1q Deficiency

Three different genes (C1qA, C1qB, C1qC) closely linked on the short arm of chromosome 1 encode for the C1q protein, which is composed of 18 polypeptide chains.

Inter-chain disulfide bonds are formed via the cysteine residue in the N-terminal region of each chain. Following the N-terminal region is a collagen-like region of approximately

81 residues. One A chain and one B chain form a heterodimer during biosynthesis, and two C chains form a homodimer, both through disulfide linkages via conserved cysteine

18 residues. Two A-B heterodimers associate with one C-C homodimer to form a hexameric structure, of the composition ABC-CBA. Three of these hexamers, with a total of 18 polypeptide chains together, form the tulip-like structure of C1q with a collagenous tail and six globular regions, each with globular heads ghA, ghB, and ghC.

The CP of activation is initiated by the C1 complex, of which C1q is the first subcomponent. When C1q in the C1 complex binds to IgM or IgG present in an immune complex, a binding site of C1r/C1s is exposed, allowing further activation of the complement pathway (Duncan and Winter 1988; Zlatarova et al. 2006; Gadjeva et al.

2008). C1q is an important opsonin to promote phagocytosis of apoptotic cells or debris, which can be archived directly without complement activation through binding at the collagenous region of C1q to calreticulin in apoptotic cell blebs and to CD91 on phagocytes.

The number of reported cases of homozygous genetic deficiency of C1q has increased to 74 (Pickering et al. 2000; Schejbel et al. 2011; Atkinson and Yu 2015;

Stegert et al. 2015). Among the reported C1q-deficient subjects, the median age of (any) disease onset was 6 years. The clinical presentations among C1q-deficient patients varied considerably, but the two common observations were: (a) SLE or lupus-like disease in

88% and (b) recurrent bacterial infections in 41% (Lipsker and Hauptmann 2010;

Schejbel et al. 2011). Some patients (17%) died at a young age secondary to septicemia.

Among the C1q-deficient patients with SLE or lupus-like disease, cutaneous disorders, especially photosensitivity, were prominent with a frequency of 84%. Glomerulonephritis and neurologic disease affected about 30% and 19% of patients, respectively. Oral

19 ulceration occurred in 22% and arthritis/arthralgia in 16%. Immunologically, most C1q- deficient patients had normal serum levels of complement C4 and C3, high frequency of

ANAs (particularly anti-Ro/SSA) but a low frequency of anti-dsDNA autoantibodies.

Several different causative mutations have been identified in patients with complete C1q deficiency. A variety of mutations (including nonsense, frame-shift indels, and splice site) result in the absence of biosynthesis of one of the three C1q (a, b, or c) chains. Other mechanisms have been identified by which C1q protein is defective in secretion, structure, or function. We have summarized in Table 1 the molecular basis of the genetic mutations leading to complete absence or complete functional deficiency of

C1q protein for which genetic information is reported, as well as the accompanying clinical manifestations.

Most of the causative mutations associated with homozygous C1q genetic deficiency are the result of a consanguineous marriage. The affected patients are likely close descendants of ancestors carrying the specific deleterious mutation. In fact, screening of large SLE cohorts from those countries with reported cases of C1q- deficiency to determine the prevalence of C1q deleterious mutations have yielded negative results (Topaloglu et al. 1996; Chew et al. 2008), suggesting that the mutations are “private” and rare but with very large effect size, as documented in many complex diseases (Manolio et al. 2009; McClellan and King 2010).

20

Complement C1r and C1s Deficiency

The genes for human C1s and C1r are located on the short arm of chromosome 12

(Nguyen et al. 1988). According to bioinformatics studies and previous publications

(Kusumoto et al. 1988), C1r and C1s are configured in a tail-to-tail orientation with their

3’ends separated by approximately 9 kb. The DNA sequence for the genomic region harboring human C1s and C1r coding sequences is still incomplete in the Reference

Genome (Annotation release 106, January 2015) and consists of several gaps.

C1r and C1s are paralogous proteins that share 38% identity and 55% similarity.

In circulation, C1r and C1s are proenzymes that exist as a tetrameric structure, C1s-C1r-

C1r-C1s, which assembles in the presence of Ca2+ with C1q to form the multi-molecular

C1 complex. Upon activation of C1q (e.g., through binding of its globular heads to the

Fc-regions of IgG or IgM in an immune complex), the tetramer interacts with the hinge region of C1q to form the activated C1 complex. Autoactivation of the two C1r by proteolytic cleavages between Arg-463 and Ile-464 is followed by activation of C1s by proteolysis between Arg438 and Ile-439, which release the enzymatic activity of the nascent C1 complex. The C1s in this C1 complex activates C4 and then C2 which together form the classical pathway C3 convertase, C4b2a.

Deficiencies in subcomponents C1r and C1s were among the earliest reports linking complement deficiency with human glomerulonephritis or a lupus-like disease

(Pickering et al. 1970; Day et al. 1972; Moncada et al. 1972). A total of 20 cases of C1r and/or C1s deficiencies have been reported, which include twelve cases of C1r deficiency from eight families and eight cases of C1s deficiency from five families. Among the

21

C1r-deficient patients, there was a consistent reduction in the serum protein levels of C1s to 30% of its normal level, but highly elevated serum protein levels of C4, C2, and C1- inhibitor (200-400% of their corresponding normal ranges). C3 was also elevated by approximately 50%, but C1q levels were normal. A similar phenomenon was observable among C1s-deficient patients; C1s-deficient patients had greatly reduced serum levels of

C1r, markedly elevated levels of C4, C2, C1-inh, and C3, and normal levels of C1q.

Among the C1r/C1s deficient subjects, all but three had recurrent bacterial, viral, or fungal infections (85%), and many patients died at young age because of a severe infection. Thirteen subjects (65%) developed SLE or a lupus-like disease. The prevalence of ANA among these patients was about 60%. Mortality at young age from infections likely explains the slightly lower frequency of lupus disease association compared to C1q deficiency. Most C1r/C1s deficient patients had severe cutaneous lesions (Figure 5).

Eight patients (40%) had renal disease due to lupus nephritis. Such presentations underscore the inter-dependence of C1r and C1s in sustaining a stable tetrameric structure that would otherwise be susceptible to a high turnover rate. A deficiency of C1r or C1s prevents the formation of the C1 complex and diminishes the need for engagement of C1- inh and other regulators of complement activation. When C1 is not functional, the CP is not activated, and consumption of C4, C2, and C3 is greatly reduced, resulting in high levels of these proteins in the circulation (Wu et al. 2011). This and other results strongly indicate a chronic turnover of component proteins for the CP (Manderson et al. 2001).

The molecular defects leading to C1r or C1s deficiency have been determined in one case of C1r deficiency and seven cases of C1s deficiency (Table 2). Relative to C1r

22 deficiency, the defect was a homozygous C to T substitution in exon 10 resulting in the

R380X nonsense mutation in the second CCP domain, resulting in no detectable protein in the serum (Wu et al. 2011). The proband developed SLE at 3 months of age and presented with reduced levels of C1s (similar to other C1r-deficient patients), but highly elevated protein levels of C4, C2, and C1 inhibitor.

For C1s-deficiency, several deleterious mutations have been identified. A C→G mutation in exon 6 (Y204X) resulted in a premature stop codon, and no protein products could be detected (Amano et al. 2008). This particular nonsense mutation was homozygous in four siblings, and all showed no detectable C1s protein, but only two developed SLE at the ages of 7 and 13 years. The other two siblings, at ages 10 and 20, did not have clinical symptoms of SLE. Significantly reduced levels of C1r and elevated serum C4 were detected in all four siblings. In another case report, a 4-bp deletion

(TTTG) in exon 10 that led to a frame-shift and a nonsense mutation in exon 12 (E597X) was detected in a single patient (Inoue et al. 1998; Endo et al. 1999). This patient developed unique symptoms including virus-associated hemophagocytic syndrome and died after a long period of a comatose state. A mutation documented in another patient from the same family was a heterozygous G→T mutation in exon 12 leading to E597X and on the other allele, a novel missense mutation G630Q (Abe et al. 2009). This patient displayed symptoms that were similar to the other, related patient, including fever of unknown origin and short-term disturbances of consciousness. A second C1s genetic variant is a nonsense C→T mutation in exon 12 (R534X) (Dragon-Durey et al. 2001).

This patient had undetectable serum C1s, normal C1r and C1q, and absence of CH50

23 activity. The patient was 2 years of age and presented with several autoimmune diseases, including a lupus-like syndrome, Hashimoto’s thyroiditis, and autoimmune hepatitis.

Each of these clinical observations and the similarities in autoimmune presentation with

C1r/C1s deficiencies highlight the role of a dysfunction in the CP of complement leading to systemic autoimmune disease.

Complement C2 Deficiency

The complement C2 gene is located in the HLA class III region on the short arm of chromosome 6. Serum C2 is a precursor protein that is cleaved by activated C1 into two fragments: C2b and C2a. C2a is a serine protease and forms the C3 convertase along with C4b (denoted C4b2a) (Walport 2001; Walport 2001). C2 also functions as a critical component in the lectin pathway. MBL or ficolins in complex with MASP-1 bind to relevant carbohydrate molecules and activate MASP-2, which then cleaves C2 and C4, forming a C3 convertase identical to that formed in the CP (Wallis et al. 2007). Overall,

C2 functions as a key component in the classical and lectin pathway, thereby providing defense against microbial infection and assisting in removal of immune complexes.

Among individuals of European descent, C2 deficiency occurs with an estimated prevalence of 1/20,000, which probably accounts for less than 1% of SLE patients. There are two types of C2 deficiency (Agnello 1986; Johnson et al. 1992). Type 1 C2 deficiency is caused by nonsense mutations leading to the absence of protein biosynthesis. The predominant form of such type 1 deficiency was a 28-bp deletion that removed 9-bp from the 3 end of exon 6 and 19-bp from the 5 end of intron 6 in the C2

24 gene, leading to a skipping of exon 6 in the C2 mRNA and generation of premature stop codon (Johnson et al. 1992). Such 28-bp deletion is present in the HLA haplotype with

A10 (A25) and B18 in the class I region, BF-S, C2Q0, C4A4 and C4B2 in the class III region, and DRB1*15 (DR2) in the class II region. The second form of Type 1 deficiency is present in HLA A3, B35, DR4, BF-F, C2Q0, C4A3 and C4A2 (Wang et al. 1998). The cause is a 2-bp deletion in exon 2 of C2 gene that leads to a nonsense mutation.

About 10% of C2 deficiency is secondary to the Type II deficiency in which the

C2 protein is synthesized but not secreted. The molecular defects identified as missense mutations are C111Y, S189F, and G444R (Wetsel et al. 1996; Zhu et al. 1998). It is not clear how these mutations block the secretion of C2 protein.

Unlike a deficiency of proteins for the C1 complex or C4 described earlier, the penetrance of C2 deficiency on SLE is about 10%. Similar to other risk factors for SLE, there is a female predominance for patients with C2 deficiency. C2-deficient SLE patients tend to have early childhood onset but a milder disease process with prominent photosensitive dermatologic manifestations, speckled ANAs [the autoantibody specificity is common for the Ro/SSA antigen], and a family history of SLE. Anti-DNA antibody tests are usually negative and severe kidney disease is rare.

Complement C4 Gene Copy Number Variations and C4A or C4B Isotype Deficiency

A complete or homozygous genetic deficiency of both complement C4A and C4B has been reported in 28 individuals (Nordin Fredrikson et al. 1991; Lokki et al. 1999;

Rupert et al. 2002; Yang et al. 2004; Wu et al. 2009). The subjects came from 19 families

25 with different racial backgrounds and were characterized by 16 different HLA haplotypes. The female to male ratio was 1:1. SLE or lupus-like disease was diagnosed in

22 (78.6%) of the C4-deficient subjects, plus four others had renal disease including glomerulonephritis. Early disease onset, severe photosensitive skin rash, the presence of autoantibodies against ribonuclear protein Ro/SSA, and high titers of ANA were common clinical features of the subjects. Many of the C4-deficient patients also had severe proliferative glomerulonephritis. We have summarized the molecular basis of complete

C4 deficiency determined in 15 cases (Table 3).

While a complete deficiency of C4 is rare, a selective deficiency of either C4A or

C4B is much more commonly observed and has been implicated in several autoimmune diseases (Yang et al. 2007; Liu et al. 2011; Rigby et al. 2012; Hou et al. 2014; Lintner et al. 2015). To investigate the C4 genetic diversities in SLE, a study population consisting of 216 female SLE patients, 17 male SLE patients, 362 first degree relatives, 389 unrelated, healthy female controls, and 128 male controls was investigated (Yang et al.

2007). In the study group of European Americans, total gene copy number (GCN) of C4 ranged from 2 to 6 copies, GCN of C4A ranged from 0 to 5 copies, and GCN of C4B ranged from 0 to 4 copies. In comparison to healthy controls, SLE patients had significant reductions of GCN of total C4 (Figure 6). Among the SLE patients, 9.3% had only two copies of C4 genes, compared to 1.5% in healthy controls. The effect size of SLE disease risk (or odds ratio) for subjects who had only two copies of C4 genes was 6.51. Of the two C4 isotypes C4A and C4B, there were no significant differences detected among

GCN of C4B between SLE and controls. However, significant decreases of GCN of C4A

26 were noted in SLE patients. Among SLE patients, 6.5% had a homozygous deficiency

(i.e., 0 copies) of C4A and 26.4% had a heterozygous deficiency (i.e., 1 copy), compared to 1.3% and 18.2%, respectively, in healthy controls. The odds ratio for SLE for a subject with C4A homozygous deficiency was 5.27. In other words, a total C4 GCN=2 or a C4A

GCN=0 (C4A deficiency) are large effect size genetic risk factors for human SLE.

To date, no other common genetic variant has been identified to be so strongly associated with SLE. Remarkably, 32.9% of SLE subjects carried the risk factor of low

GCN of C4A. As with most common genetic variants associated with autoimmune disease, the risk factor is also present in the general population with considerable frequency (19.5% in this study). From another way of statistical analysis, C4 gene copy number variations are continuous variations and therefore the mean of gene copy numbers for total C4, C4A, C4B, long genes and short genes can each be compared by

Student’s t-test between patients and controls. The mean GCN (±standard deviation) for total C4 between female controls and female SLE patients were 3.81±0.75 and

3.56±0.77, respectively (p=0.0001). The mean C4A GCN (±SD) was 2.05±0.79 in controls and 1.81±0.89 in SLE (p=0.0005). The mean long C4 was 2.91±1.03 in controls and 2.66±1.14 in SLE (p=0.005). In other words, when compared with controls, SLE showed reductions of 0.25 copies of total C4, 0.24 copies of C4A, or 0.25 copies of long

C4. For C4B or short C4 genes, no significant differences were observed between the

European American SLE and race-matched controls.

As one of the original publications to document the highly prevalent, multi-allelic gene CNVs of complement C4 in SLE, this work (Yang et al. 2007) went through

27 independent data generation and validation processes to confirm the CNV-calls. Those processes included (a) pulsed-field gel electrophoresis to resolve PmeI-digested genomic

DNA fragments and Southern blot analysis for long-range mapping; (b) TaqI restriction digests and genomic Southern blot analyses (restriction fragment length polymorphisms) to resolve the relative dosages of RP1-C4L (7.0 kb), RP1-C4S (6,4 kb), RP2-C4L (6.0 kb), RP2-C4S (5.4 kb), plus relative dosages of CYP21B to CYP21A, and relative dosages of TNXB to TNXA; (c) PshAI-PvuII digests of genomic DNA and Southern blot analysis to segregate C4A and C4B and yield their relative dosages; (d) immunofixation of EDTA- plasma for polymorphic variants of C4A and C4B proteins resolved by high voltage agarose gel electrophoresis; and (e) corroboration of C4 genotypes and phenotypes of study subjects from data of family members.

Subsequently, a TaqMan-based quantitative PCR for total C4, C4A, C4B, long genes and short genes was developed and applied for replication studies, particularly when quantities of genomic DNA for patients and controls was limited. In the latter case, internal data validation is achieved when GCNs of total C4 = GCNs of C4A+C4B =

GCNs of C4 long + C4 short. Such qPCR strategy is sensitive and highly robust when the quality of genomic DNA is excellent, which can be reflected by internal data validation of the independent amplicons. Our experience suggested that genomic DNA samples at low concentrations (≤15 ng/l) are relatively unstable in storage and tend to yield inconsistent data in analyses of multi-allelic CNVs. This makes individual internal data validation crucial for data accuracy.

28

Association of lower gene copy number of total C4 and C4A deficiency as a risk factor of SLE have also been observed in three independent East-Asian studies (Lv et al.

2012; Kim et al. 2013; Chen et al. 2016). C4A deficiency in subjects of European ancestry is primarily attributed to the presence of a single short C4B gene (mono-S) in the

HLA that is predominantly in linkage disequilibrium with HLA A*01, B*08 and

DRB1*0301 that is dubbed ancestral haplotype 8.1 (AH8.1) (Dawkins et al. 1999).

Intriguingly, such AH8.1 haplotype is basically absent among East-Asian subjects. A different mechanism leading to C4A deficiency is prevalent in Asian SLE (Figure 7)

(Chen et al. 2016). Certainly other genetic or environmental risk factors, combined with the C4A deficiency, contributed to SLE development in genetically predisposed patients.

Still, it is of interest to examine if restoring C4A in such (C4A-deficient) patients would result in positive therapeutic outcomes.

On a recent case/control study of British SLE (cases, N=501; controls, N=719), total C4 gene copy numbers were determined by a paralog ratio test, which employed a set of PCR primers (16-mer) that hybridized to and amplified the duplicated regions of complement C4 plus a unique and non-variable region in chromosome 19 as a reference control (Boteva et al. 2012). C4A and C4B were deduced from the ratios of C4A and C4B determined by NlaIV restriction digest of a different PCR product spanning the C4- isotypic site and were resolved by capillary electrophoresis. While extensive gene copy number variations and associated polymorphisms were observable for complement C4, the difference on the mean copy number of total C4 between British SLE and controls was only marginal (3.79±0.98 in SLE; 3.89±0.98 in controls; p=0.046). On the other

29 hand, the mean copy numbers of C4A (1.82±0.93 in SLE, 2.08±0.93 in controls; p<0.001) and C4B (1.96±0.93 in SLE and 1.81±0.72 in controls; p<0.001) were both significantly reduced in SLE. However, in multiple logistic regression models, deficiencies of C4A or

C4B both became insignificant in the presence of DRB1*0301. This led the investigators to conclude that “…partial complement C4 deficiency states are not independent risk factors for SLE in UK…”. It is notable that there were significant differences in the architecture of GCN group distributions between the British (Boteva et al. 2012) and the

US study populations (Yang et al. 2007).

Going through the Boteva study on C4-CNV determination, two issues could have been improved. The first would have been to employ an internal data validation for GCN calls. The second would have been to add an additional procedure to overcome the artifacts introduced by heteroduplex formation between C4A and C4B alleles during

PCR, which would be resistant to restriction digest (NlaIV) and therefore would skew relative dosages of C4A and C4B (Uejima et al. 2000; Chung et al. 2002). To dissect the relative roles of C4-CNVs or C4A deficiency and HLA-DRB1 alleles in SLE, the issues on the lack of internal validation and method choice for the reliability of C4-CNV calls cannot be ignored. Further investigations are necessary to settle the issue of C4A deficiency and HLA-DRB1*0301 on the genetic risk of SLE of northern European ancestry.

The unprecedented variations of gene copy numbers with high frequencies of homozygous or heterozygous deficiency of C4A or C4B, and continuous variations in copy numbers from one to four copies of C4 genes on each copy of chromosome 6 (or

30 haplotype) among healthy subjects and SLE patients inevitably pose great challenges both technically and conceptually: the former for accurate data acquisition and the latter for accurate data interpretation. Without deliberate design and rigorous and independent validation strategies, unfortunately, many studies were inherently tainted with misinterpretations or partially correct or inappropriate conclusions. On determining the roles of common and multiallelic CNVs in health and disease such as those for complement C4A and C4B, immunoglobulin Fc receptors FCGR3A and FCBR3B, and neutrophil alpha-defensins DEFA3 and DEFA1, it is critically important to have meticulous experimental design and methods to acquire accurate and consistent data.

Realization of the DNA sequence basis that causes a phenotype or functional diversity is essential for specific experimental method. As mentioned above, the heteroduplex issue during the PCR process needs to be considered and resolved. Moreover, genetic studies of a complex disease usually require hundreds to thousands of genomic DNA samples from cases and controls, which are generally obtained from multiple sources. For determination of common and continuous CNVs, the high quality of genomic DNA samples is essential. Heterogeneous quality of DNA samples has a high tendency to yield inconsistent data. Under those conditions, independent replication and rigorous internal validation methods of samples from every subject becomes a necessity.

1.5.5 Acquired Deficiencies and Autoantibodies to Complement Components in SLE

Autoantibodies have been reported that bind with high affinity to complement proteins, particularly in SLE patients (Dragon-Durey et al. 2014). Most of these

31 antibodies are not directed against native proteins, but instead directed against neoepitopes. Such epitopes becomes exposed in active or inactivated proteins, or upon multi-molecular complex assembly in the activation process or following proteolytic cleavage. The binding of these autoantibodies to complement proteins could lead to a state of an acquired deficiency and contribute to disease pathogenesis similar to the way genetic deficiencies do so. We will describe a series of autoantibodies that have been detected in systemic autoimmune disease against early components and regulators of the

CP.

Anti-C1 Autoantibodies

Approximately 30% of SLE patients synthesize autoantibodies against C1q. Their presence correlates with anti-dsDNA, nephritis, and low C4 and C3 in about 75% of such patients (Fremeaux-Bacchi et al. 1996; Walport 2002; Mahler et al. 2013). A relevant question is whether the anti-C1q amplifies the complement activation by immune complexes. The development of anti-C1q antibodies may arise in response to activation of the CP. Following CP activation, C1q remains attached to immune complexes and therefore is located at the site of inflammation. Proteases at the inflammatory site may degrade IgG and C1q (autoantigen), generating multiple proteolytic fragments of IgG and

C1q. This may be an explanation for C1q antibodies and anti -IgG and –IgM (rheumatoid factors) that develop in SLE patients.

A large study was recently reported in which one objective was to assess the specificity of anti-C1q antibodies and their associations with SLE manifestations and

32 diagnostic tests (Orbai et al. 2015). The authors confirmed the association of anti-C1q antibodies, low complement (C4 and C3), and anti-dsDNA antibodies. Further, this combination had the highest serological association with renal disease. Anti-C1q antibodies were detected in 28% of all SLE patients but were observed more frequently in patients with renal disease (approximately 68%). Anti-C1q autoantibodies were observed in 5-10% of patients with related systemic rheumatic diseases. It was suggested that the presence of anti-C1q antibodies contributes to a nephritis flare (Mahler et al. 2013).

The presence of autoantibodies against other components of the C1 complex is less established. A report of fifteen SLE patients demonstrated that seven of them had autoantibodies to C1s (He and Lin 1998). The binding of these antibodies to C1s was shown to enhance its enzymatic activity for C4, providing possible additional explanation for low serum C4 in these individuals.

Anti-C1 Inhibitor Autoantibodies

An IgG autoantibody that inactivates C1-inhibitor (anti-C1-Inh) was initially described in a patient with the acquired angioedema (AAE) syndrome that mimics hereditary angioedema (Jackson et al. 1986). Anti-C1-Inh autoantibodies have also been described in SLE, especially those exhibiting symptoms of angioedema (Nakamura et al.

1991; Ochonisky et al. 1993). Anti-C1-Inh antibodies bind and inactivate C1-Inh so that it is no longer available to participate in the regulation of C1. As a result, the CP is excessively activated, leading to development of angioedema as well as possibly more severe renal diseases. A more recent study of 202 SLE patients and 134 healthy controls

33 detected anti-C1-Inh autoantibodies in 17% of SLE patients and 4% of controls

(Meszaros et al. 2010). In SLE patients, the anti-C1-Inh levels correlated with the duration and activity of SLE but did not correlate with SLE laboratory parameters, including serum levels of C3 and C4. Conversely, 1-2% of hereditary angioedema patients develop SLE, probably related to the chronically very low C4 and C2.

C3 and C4 Nephritic Factors

C3 and C4 nephritic factors are IgG autoantibodies that bind to and stabilize the alternative pathway C3 convertase and the classical pathway C3 convertase, respectively.

By binding the C3 convertases, C3 and C4 nephritic factors prolong the half-life by preventing the regulation of C3 convertases. This results in uncontrolled complement activation and increased consumption and depletion of serum C3. C3 and C4 nephritic factors are associated with membranoproliferative glomerulonephritis, acquired partial lipodystrophy, and post-infectious acute glomerulonephritis (Daha and van Es 1980;

Miller et al. 2012; Jozsi et al. 2014). Both of these autoantibodies have been detected in

SLE patients, and suggested to be associated with renal disease, but their prevalence and role in pathogenesis of systemic autoimmune disease is not well documented (Daha and van Es 1980; Pickering and Cook 2008; Miller et al. 2012).

34

1.6 Juvenile Dermatomyositis (JDM)

1.6.1 JDM Background

JDM is a rare, autoimmune, multi-systemic inflammatory disease affecting mostly muscle and skin. The incidence in the United States is 3.2 per million children per year, with some differences in ethnic groups (Symmons et al. 1995; Ramanan and Feldman

2002; Mendez et al. 2003). Disease onset occurs at or below 16 years of age, with an average age of onset of 7 years, and the ratio of female:male is 2.3:1 (Feldman et al.

2008). Characteristic clinical features and diagnostic criteria include proximal muscle weakness and inflammation, distinct skin rashes (Gottron’s sign or heliotrope eyelid rash), increased levels of serum muscle enzymes, and pathological changes upon muscle biopsy (Bohan and Peter 1975; Bohan and Peter 1975; Dalakas and Hohlfeld 2003;

Feldman et al. 2008). The autoimmune origin of JDM is supported by the presence of autoantibodies and its association with genes located in the MHC. However, there is no one specific target antigen that has been identified, and the agents initiating self- sensitization remain unknown.

Similar to many autoimmune rheumatic diseases, JDM is thought to arise from immunological dysfunction resulting from a complex interplay of genetic susceptibility factors and environmental stimuli, including infections, UVB exposure, and season of birth (Pachman et al. 1997; Tezak et al. 2002; Dalakas and Hohlfeld 2003; Rider et al.

2010; Shah et al. 2013). Within three months prior to disease diagnosis, most JDM patients had symptoms related to upper respiratory track or abdominal infections, suggesting that infections may be playing a role in triggering the disease onset (Pachman

35 et al. 2005). Using RNA isolated from muscles and blood from established pediatric and adult DM patients (years after disease onset and with no signs of infection), gene expression profiling experiments consistently demonstrate in most patients a “signature pattern” that is characteristic of stimulation by type I interferons (Tezak et al. 2002;

Walsh et al. 2007; Greenberg 2010; Rider et al. 2010). Type I interferons are mostly secreted by dendritic cells, mast cells and other myeloid cells in response to viral infections. A significant number of patients with JDM are positive for antinuclear antibody (ANA, 23%) (Pachman et al. 1985). In addition to the presence of ANA that are common in most systemic autoimmune diseases, many JDM patients have myositis- specific autoantibodies against cellular components essential for protein synthesis and secretion, such as autoantibodies against synthetases of tRNA and “signal particles”

(Hengstman et al. 2001; Rider et al. 2013).

JDM is probably a complex autoimmune disease triggered by environmental factors such as infections in genetically predisposed human subjects, leading to a chronic inflammatory state with continuous secretion of type I interferons, dysfunctional innate and adaptive immunity leading to the loss of immune tolerance and production of autoantibodies, dysregulated complement activation and destruction of the vascular system surrounding muscle fibers, skin, and sometimes other vital organs.

1.6.2 Complement in JDM

Of the inflammatory myopathies, DM is the one most reported to be associated with a complement-mediated pathogenesis (Dalakas 2015). Complement-mediated

36 destruction of perivascular endothelium and perifascicular ischemia of muscle fibers in biopsies from DM patients have been demonstrated by multiple investigators (Whitaker and Engel 1972; Kissel et al. 1986; Emslie-Smith and Engel 1990; Kissel et al. 1991;

Mascaro et al. 1995; Burgin et al. 2014). Circulating immune complexes, IgG and IgM, complement C3 and late components of complement activation C5b-C9 MAC were detected in DM muscle and skin biopsies. A single case of an individual with a complete

C2 genetic deficiency was reported to have DM (Leddy et al. 1975). The observation of complement components and complement fragments indicates the possible involvement of complement pathways in tissue damage.

The HLA allele DRB1*0301 (DR3) has been identified as a major immunogenetic risk factor for JDM and many autoimmune diseases (Friedman et al. 1983; Thorsby and

Lie 2005; Mamyrova et al. 2006; O'Hanlon et al. 2006) and was recently confirmed as the strongest genetic risk factor for JDM through genome-wide association studies (GWAS) and dense genotyping of immune-related loci (Miller et al. 2013; Rothwell et al. 2015).

Also present in the HLA region is the locus for complement C4. Specifically, DR3 in

European subjects is in strong linkage disequilibrium with a particular C4 haplotype: a single C4B gene but the absence of a C4A gene (Yu and Whitacre 2004; Horton et al.

2008). Consequently, it might be theorized that lack of C4A contributes an additional risk factor for this DR3 haplotype. Despite evidence for implication of immune complexes and complement activation in pathogenesis of JDM, as well as the extensive knowledge of C4 genetic variation, there are no studies to date that investigate C4 genetics in JDM patients.

37

Figure 1 The activation pathways of the complement system

The three activation pathways of the complement system are shown according to

evolution and physiologic sequences. Pathway 1 is known as the alternative pathway. It

is activated through a tick-over mechanism because of continuous hydrolysis of the

thioester bond in C3, which enables the formation with factor B to form a weak C3

convertase. Pathway 2 is known as the MBL or lectin pathway. It is initiated through the

binding of mannan binding lectin (MBL) or ficolin to arrays of simple sugar molecules in

glycosylated antigens on microbes. This is a pattern recognition mechanism

characteristic of the innate immune system. Pathway 3 is called the classical pathway

and is initiated through the binding of specific antibodies IgM or IgG to antigens. It is an

effector arm of the humoral adaptive immune system. Each activation pathway engages

the formation of a multi-molecular initiation complex, followed by the assembly of a C3

convertase and a C5 convertase for activations of C3 and C5, respectively, and

culminates in the formation of membrane attack complexes (MAC) on the target

38 (continued) membrane. All three pathways can be amplified through a positive feedback mechanism, as C3b generated by any C3 convertase can feed to the alternative pathway through association with factor B to form new C3 convertase after activation by factor D

(pathway 1). Anaphylatoxins C3a and C5a are produced during the activation process.

For brevity, by-products generated during the activation of C4, C2 and factor B are not shown. Red arrows show activation of component proteins through cleavage by serine proteases. A dotted horizontal arrow denotes multiple steps are involved in the formation of the membrane attack complex. Early components of the classical pathway C1q, C1r,

C1s and C4 are engaged in the differentiation of immunity and autoimmunity, as genetic or acquired deficiency in any of these components are linked to pathogenesis of SLE.

Complement C2 is also be involved in the protection against autoimmunity but its effect size is smaller [Modified from Reference (Yu et al. 2014)].

39

Figure 2 Gene size dichotomy and gene copy number variation of complement C4 (continued)

40

A human C4 gene consists of 41 exons coding for a precursor protein of 1744 amino acids including a signal peptide of 19 amino acids. (A) There are two forms of C4 genes.

The long gene is 20.6 kb, and the short gene is 14.2 kb. In a long C4 gene, an endogenous retrovirus HERV-K(C4), which is 6.4 kb in size, integrated into its ninth intron. Among healthy subjects of European ancestry, 76% of C4 genes belong to the long form and 24% belong to the short form. (B) Among European subjects, one to four copies of C4 genes are present in the central region of the major histocompatibility complex (MHC) located on chromosome 6p21.3. Thus, there is a continuous variation in copy number of C4 genes from two to eight copies among different human subjects. (C)

The duplication of a C4 gene occurs in a modular fashion, with a 0.9 kb fragment of RP

(STK19) upstream of complement C4, a full steroid cytochrome P450 21-hydroxylase

(CYP21) and a 4.0 kb fragment of the tenascin (TNX) at the downstream region of C4

(known as a RCCX module). The duplication of CYP21 gene can be a pseudogene

(CYP21A or CYP21A1P) or an intact functional gene (CYP21B or CYP21A2). Each C4 gene in the RCCX module may either code for an acidic C4A or a basic C4B. Each C4 gene may be either long or short [Adapted from References (Yu 1991; Dangel et al. 1994;

Chung et al. 2002)].

41

Figure 3 Genetic architecture of the HLA region in the human genome

The HLA region in humans encompasses 7.6Mp on chromosome 6p21. The region

contains >400 genes and is subdivided into five subregions: extended class I, class I,

class III, class II, and the extended class II. The HLA exhibits the most dense linkage

disequilibrium in the human genome, sometimes extending >500kb. Of interest, the

region harboring the C4 locus in class III displays strong LD with specific ancestral

haplotypes of the DRB1 region in class II, which are 350kb apart. Adapted from reference

(Undlien et al. 2001).

42

Figure 4 Typical serial serum protein profiles of complement C4 and C3 in human SLE (continued) patients 43

Serum C4 (red, solid line) and C3 (green, dashed line) protein levels tend to go up and down together in most SLE patients. The horizontal dotted line indicates the low level of serum C4 (<10 mg/dL), below which usually requires clinical attention. The profiles shown are taken from three individual patients over a 24 month period and represent three common profiles typically observed in SLE patients. In the first profile (A), levels of C4 and C3 were chronically low. In some patients, even if C3 levels rose to normal range, C4 levels remained low. Patients with this profile are often characterized by low copy number of C4 genes. (B) The second profile had frequent and parallel fluctuations of serum C3 and C4. Patients with this profile had active disease, and low C3 and low

C4 roughly correlated with disease activity. In the third profile (C), C4 and C3 protein levels stayed in the normal range most of the time, except at the time of diagnosis and during a disease relapse. Patients with this profile had relatively inactive disease.

Patients with the second and third profiles have normal gene copy number of total C4 but may have a heterozygous deficiency of C4A. [Modified from Reference (Wu et al.

2006)].

44

Figure 5 SLE patients with a homozygous deficiency of early components for the classical pathway of complement activation

Severe cutaneous lesions are common clinical presentations in SLE patients with a

complete complement deficiency. (A) A homozygous C1q-deficient male child with

cutaneous infection (upper panel), and with discoid lupus erythematosus and scarring

lesions on face when he was 22 years old (lower panel). (B) A male child with discoid

lupus at 16-month old with homozygous C1r-deficiency. This patient experienced

generalized seizures, developed a scissoring gait with toe walking, spasticity and

weakness of the legs. At 18 years old, he was diagnosed with class IV lupus nephritis

and progressed to end-stage renal disease. (C) A complete C4-deficient girl at three

years old with butterfly rash and cheilitis (upper panel), and osteomyelitis of the femur at

45 (continued)

10 years old (lower panel). This patient died at age 12 because of pulmonary infection and cardiovascular failure. (D) A homozygous C2-deficient young woman with acute cutaneous lupus erythematosus. The upper panel shows the butterfly rash, and the lower panel shows photosensitive lesions on sun-exposed areas. [Adapted from References (Yu et al. 2007; Wu et al. 2009; Lipsker and Hauptmann 2010; Wu et al. 2011)].

46

Figure 6 Comparisons of frequencies for total C4, C4A, and C4B gene copy number groups in SLE (red) and controls (blue)

SLE patients (N=216) of European ancestry showed significantly higher frequencies for lower copy numbers of total C4 (GCN=2 or 3) and C4A (GCN=0 or 1) compared to healthy, race-matched controls (N=389). Mean GCN for Total C4 in SLE (3.56±0.77) was significantly lower than in controls (3.84±0.69; p=5.3x10-6, t-test). Similarly, mean GCN for C4A in SLE (1.81±0.89) was significantly lower than in controls (2.06±0.76; p=2.0x10-

4, t-test) [Modified from Reference (Yang et al. 2007)].

47

Figure 7 Race-specific distribution patterns of RCCX modules in human populations

The size dichotomy of C4 genes and copy number variation of RCCX modules on an

MHC haplotype together create a repertoire of length variants among different human

subjects, which also exist with race-specific distribution patterns. (A) The most prevalent

haplotypes of RCCX in Whites and Asian-Indians are the bimodular long-long (LL), and

bimodular long-short (LS) in Blacks and East-Asians. (B) Notably, monomodular-short

(mono-S or S) haplotypes with a single short C4B gene and C4A deficiency is relatively

common in White and Black subjects but almost absent in Asians [Modified from

Reference (Yu and Whitacre 2004)].

48

Table 1 Complete deficiency of A, B, or C chain genes of C1q

Age of onset Molecular Location Mutation (years), sex, Clinical Presentations References Defect ethnicity 1 C1qA M1R start codon nk, M, African- SLE (Namjou et al. mutation; no American2 2012) detectable protein 2 C1qA M1R As above nk, M, African- Lupus, premature death (Namjou et al. American2 2012)

49

3 C1qA Q208X Nonsense 10, M, Turkish History of ear and oral infections, recurrent (Berkel et al. mutation skin lesions, premature death at age 10 1977; Berkel from septicemia et al. 1979; Petry et al. 1997) 4 C1qA Q208X As above 4, F, Turkish1 Malar rash, stomatitis, ANA, premature (Petry et al. death at age 6 from sepsis 1997; Berkel et al. 2000) 5 C1qA Q208X As above 6, F, Turkish1 Facial swelling, hematuria (Petry et al. 1997; Berkel et al. 2000) 6 C1qA Q208X As above nk, F, Turkish Asymptomatic at age 22 (Berkel et al. 1997)

7 C1qA Q208X As above 3, F, Turkish3 SLE, glomerulonephritis, arthralgias, (Topaloglu et

49 Continued

Table 1 continued

photosensitivity, anti-Ro autoantibodies al. 1996) 8 C1qA Q208X As above 15, F, Turkish3 Photosensitive rash, microscopic (Topaloglu et hematuria, IgA nephropathy al. 1996) 9 C1qA Q208X As above 4, M, Turkish SLE-like disease, recurrent meningitis, (Hoppenreijs pneumonia, meningococcal sepsis, ANA et al. 2004) 10 C1qA Q208X As above 1, M, Turkish Rash, recurrent upper respiratory tract (Sun-Tan et infections, low ANA al. 2010) 11 C1qA Q208X As above 1, M, Iraqi Erythematous rashes, recurrent, otitis (Schejbel et media, glomerulonephritis, fatigue, al. 2011) photosensitivity, ANA

50 12 C1qA W216X Nonsense 0.5, F, SLE, cutaneous lupus, bacterial meningitis, (Schejbel et 4 mutation Sudanese ANA al. 2011)

13 C1qA W216X As above 3, M, Sudanese4 SLE, cutaneous lupus, bacterial meningitis, (Schejbel et bacterial keratitis, polyarthritis, ANA al. 2011) 14 C1qA 1-bp Frameshift 3, M, Caucasian Photosensitivity, malar rash (Schejbel et deletion; mutation al. 2011) Q64X premature stop codon 15 C1qB point premature 4, M, Pakistani SLE-like disease, history of fever, (Reid and mutation stop codon; glomerulonephritis, discoid facial lesions, Thompson (RFLP functionally ANA, premature death at age 8 1983; analysis deficient McAdam et only) protein al. 1988)

16 C1qB G42D Glycine 16, F, SLE, arthralgia (Chapuis et al. mutation; Moroccan5 1982; Petry et LMW C1q; al. 1997) 50 Continued

Table 1 continued complete functional deficiency 17 C1qB G42D As above 23, M, SCLE, ANA, anti-Sm autoantibodies (Chapuis et al. Moroccan5 1982; Petry et al. 1997) 18 C1qB G42D As above 3, M, SLE, ANA, anti-ds DNA autoantibodies, (Chapuis et al. Moroccan5 thrombocytopenia, growth retardation 1982; Petry et al. 1997) 19 C1qB G42D As above nk, M, Asymptomatic at age 42 (Chapuis et al. Moroccan5 1982; Petry et

51 al. 1997) 6 20 C1qB G244R Glycine 3, F, Inuit DLE, photosensitive malar rash, ANA, (Marquart et mutation; recurrent skin and mucosal lesions al. 2007) no detectable protein 21 C1qB G244R As above 14, F, Inuit6 SLE, ANA, arthritis (Marquart et al. 2007) 22 C1qB G244R As above 2, F, Inuit6 Lupus erythematosus, vasculitis, (Marquart et pneumonia, ANA al. 2007)

23 C1qB G63S Missense 20, M, Arabian SLE, CNS involvement, recurrent (Roumenina et mutation; infections; premature death due to bacteria- al. 2011) C1q unable induced septic shock to associate with C1r and C1s 24 C1qB 6251A>C splice site 2, M, Caucasian Recurrent upper airway infections, history (van

51 Continued

Table 1 continued

mutation; no of fever and seizures Schaarenburg detectable et al. 2015) protein 25 C1qB 187G>T splice site 4, F, Japanese DLE, history of fever, facial erythema, (Higuchi et al. mutation; joint pain, oral ulcerations 2013) complete functional deficiency 26 C1qC G34R Glycine 4, F, Indian7 DLE, photosensitivity, ANA (Slingsby et mutation; al. 1996) LMW C1q; 52 complete functional deficiency 27 C1qC G34R As above 0.8, M, Indian7 Umbilican sepsis, erythematous rash, (Slingsby et ANA, parotitis, anti-Ro autoantibodies al. 1996) 28 C1qC G34R As above 0.5, F, Arabian SLE-like disease with CNS involvement, (Tsuge et al. recurrent bacterial infections, ANA, anti- 2010) dsDNA autoantibodies, hyper-IgM syndrome 29 C1qC G34R As above 21, F, Adult-onset of SLE-like disease, history of (Pickering et Caucasian fever, oral ulcerations, bacterial meningitis, al. 2008) ANA, anti-Ro autoantibodies 30 C1qC R69X Nonsense 9, F, Caucasian Severe SLE, with cutaneous and CNS (Bowness et mutation involvement, ANA, anti-Sm and anti-Ro al. 1994; autoantibodies, cytomegalovirus retinitis, Slingsby et al. premature death at age 28 from CNS 1996) involvement Continued 52

Table 1 continued 31 C1qC R69X As above 10, M, Kosova Malar and discoid rash, ANA, oral (Schejbel et ulcerations al. 2011) 32 C1qC G76R Glycine 8, F, Turkish Recurrent meningitis, pneumonia (Gulez et al. mutation; no 2010) detectable protein 33 C1qC delC43 frameshift 1, M, SLE-like disease, photosensitivity, (Mikuska et fs108X premature Yugoslavian8 butterfly rash, glomerulonephritis, ear al. 1983; stop codon at infections, ANA, anti-Sm and anti-Ro Slingsby et al. 108 autoantibodies 1996) 34 C1qC delC43 As above 3, M, SLE-like disease, cutaneous vasculitis, (Mikuska et 8 53 fs108X Yugoslavian ANA al. 1983; Slingsby et al. 1996) 35 C1qC 1bp frameshift 6, F, Pakistani Erythematosus rash, recurrent urinary and (Stone et al. deletion mutation respiratory infections, ANA, anti-Sm 2000; Mehta  83X premature autoantibodies et al. 2010) stop codon at 83 1-8 individuals marked with matching superscripts are from the same family; nk, age of onset is not known;

DLE, discoid lupus erythematosus; SCLE, subacute cutaneous lupus erythematosus; SLE, systemic lupus erythematosus.

53

Table 2 Molecular defects and clinical presentation of complete deficiency for C1s or C1r

Age of onset

Location Mutation Molecular Defect (years), sex, Clinical Presentations References

ethnicity

1 C1s -Exon 6 Y204X nonsense mutation; 7, F, Brazilian1 SLE, recurrent infections (pneumonia, (Amano et al.

no detectable protein septic arthritis, sinusitis), ANA, anti-Sm 2008)

54

autoantibodies, arthritis, proteinuria,

deposition of IgG and C1q on the

glomeruli

2 C1s -Exon 6 Y204X nonsense mutation; 13, M, SLE, arthritis, ANA, anti-Sm (Amano et al.

no detectable protein Brazilian1 autoantibodies, photosensitivity 2008)

3 C1s -Exon 6 Y204X nonsense mutation; nk, M, Asymptomatic at age 20 (Amano et al.

no detectable protein Brazilian1 2008)

4 C1s -Exon 6 Y204X nonsense mutation; nk, M, Asymptomatic at age 10 (Amano et al.

54 Continued

Table 2 continued

no detectable protein Brazilian1 2008)

5 C1s – Exon 4bp frameshift mutation 4, M, Japanese2 Virus-associated hemophagocytic (Inoue et al.

10, Exon 12 deletion leading to nonsense syndrome; history of fever; seizures with 1998; Endo et

+ E597X mutation; no loss of consciousness leading to premature al. 1999)

detectable protein death at age 7

6 C1s, Exon 12 G630Q missense mutation; 13, F, History of fever and pain, ANA, seizures (Abe et al.

55 +E597X nonsense mutation; Japanese2 and periods of unconsciousness 2009)

truncated protein

(functionally

inactive) detectable

at extremely low

levels in serum

7 C1s, Exon 12 R534X nonsense mutation; 2, F, Caucasian Recurrent malar rash, mild fever, pain and (Dragon-

no detectable protein swelling in joints, lupus-like syndrome, Durey et al.

55 Continued

Table 2 continued

Table 2 continued Hashimoto's thyroiditis, autoimmune 2001)

hepatitis

8 C1r R380X nonsense mutation; 0.3, M, African SLE, discoid lupus rash, diffuse (Wu et al.

no detectable protein American proliferative glomerulonephritis, transverse 2011)

56 myelitis

1-2 individuals marked with matching superscripts are from the same family; nk, age of onset is not known;

SLE, systemic lupus erythematosus.

56

Table 3 Molecular defects and clinical presentations of complete deficiency of complement C4A and C4B

Location Mutation Age of onset Clinical Presentations RCCX HLA References

(years), sex,

race/ethnicity

57 1 Exon 20 Homozygous 1- 2, F, Swedish SLE-like disease, atypical L (C4A) A30 B18 DR3 (Kjellman

bp C-deletion, rash, ANA, rheumatoid et al. 1982;

codon 830; factor, persistent exanthema, Nordin

premature stop glomerulonephritis Fredrikson

et al. 1991;

Fredrikson

et al. 1998)

2 Exon 29 Heterozygous; 30, F, Finnish1 Malar rash, photosensitivity, LS (C4A-C4B) A2 B39 Cw7 (Lokki et

identical 2-bp polyarthritis, leukopenia, / DRB1*1501 / al. 1999)

insertion to ANA (1/320), anti-Sm

57 Continued

Table 3 continued

codon 1232 of (1/1280), weakly positive L (C4A) A2 B40 Cw3

all three C4 rheumatoid factor DRB1*1501

genes

3 Exon 29 As above nk, M, Finnish1 Photosensitivity As above As above (Lokki et

al. 1999) 58

4 Exon 29; 2-bp insertion in 9, M, US- SLE, arthralgia, malar rash; LS (C4A- A2 B12 DR6 (Rupert et

Exon 13 codon 1232 of French descent2 photosensitivity; ANA C4B) al. 2002)

C4A; 1-bp (1/10240), positive for anti-

deletion in Sm, anti-U1 ribonuclear

codon 541 of protein, anti-cardiolipins;

C4B class III nephritis,

neurological disease, brain

vasculitis; Sjogren’s

Continued 58

Table 3 continued

syndrome, recurrent

infections, Raynaud's

phenomenon; died at age 23

5 Exon 29; As above 42, M, US- discoid rash, polyarthralgias, As above As above (Rupert et

2

59 Exon 13 French descent oral ulcers al. 2002)

6 Exon 13 GT-deletion in 10, M, History of fever, L (C4A) A24 Cw7 B38 (Yang et al.

codon 516 Austrian/Italian macrohematuria, mesangial DR13 2004)

GN; infection, nephrotic

syndrome; membranous GN

7 Exon 13 As above 5, M, Renal failure, mesangial GN; L (C4A) A24 Cw7 B38 (Yang et al.

Austrian/Italian3 skin disease with facial rash; DR13 2004)

a brother died at 3 with

cerebral vasculitis and sepsis.

59 Continued

Table 3 continued

8 Exon 13 As above 2, F, SLE, history of fever, skin As above As above (Yang et al.

Austrian/Italian3 rash, oral ulcers, microscopic 2004)

hematuria, mesangial GN;

skin transplant 60

9 Intron g8127a (gtat) 17, M, Henoch-Schoenlein purpura, SS (C4B- A30 B18 DR7 (Yang et al.

28, both C4B genes Austrian/Italian macrohematuria, nephrotic C4B) 2004)

splice syndrome, mesangial GN;

donor hemodialysis at 23; renal

graft at 24; hematuria and

proteinuria recurred at 26;

mesangial GN; chronic

allograft nephropathy;

hemodialysis at 28; second

renal graft at 36. A younger

60 Continued

Table 3 continued

brother with complete C4

deficiency (details

unavailable).

10 Intron g8127a (gtat) 6, F, SLE, hypertension, erythema SS (C4B- A30 B18 DR7 (Yang et al.

28, both C4B genes Austrian/Italian4 at face, hands and arms; C4B) 2004)

splice microhematuria, proteinuria,

donor membranoproliferative GN;

61

chronic renal failure;

hemodialysis at 26; renal

graft at 31.

11 Intron As above 5, M, SLE, skin lesions, As above As above (Yang et al.

28, Austrian/Italian4 microhematuria, proteinuria, 2004)

splice MPGN; hemodialysis at 16,

donor cadaveric renal transplant at

61 Continued

Table 3 continued

18; chronic renal graft

nephropathy at 23;

hemodialysis at 24;

meningitis

62 12 Intron As above 5, F, Hematuria and proteinuria, As above As above (Yang et al.

28, Austrian/Italian4 MPGN; facial maculopapular 2004)

splice rash; biopsy-proven skin

donor vasculitis; mental disorder,

severe cerebral vasculitis.

13 Exon 13 R559X 6, M, SLE, malar rash, L (C4A) A2 B17 DRB1*07 (Wu et al.

Moroccan5 photosensitivity, discoid rash, 2009)

ANA (1/1280), positive anti-

Ro/SSA, proteinuria and

microscopic hematuria, GN.

62 Continued

Table 3 continued

14 Exon 13 As above 17, M, Recurrent infections, L (C4A) A2 B17 DRB1*07 (Wu et al.

Moroccan5 hematuria 2009)

63

15 Exon 36 4-bp (GACT) 12, F, Algerian Malar rash, ANA (1/1024); LS (C4A- A1 B17 DRB1*13 (Wu et al.

insertion at anti-Ro, anti-Sm; renal C4B) 2009)

codon 1555, disease, lung infections,

Y1556X; both bacterial meningitis;

C4A and C4B osteomyelitis; died at age 12

related to cardiopulmonary

complications.

1-5 individuals marked with matching superscripts are from the same family; nk, age of onset is not known;

GN, glomerulonephritis; MPGN, membranoproliferative glomerulonephritis; SLE, systemic lupus erythematosus.

63

Chapter 2: Gene Copy Number Variations (CNVs) of Complement C4 and C4A

Deficiency in Genetic Risk and Pathogenesis of Juvenile Dermatomyositis (JDM)

Abstract

Complement-mediated vasculopathy of muscle and skin are clinical features of

JDM. We assessed gene copy number variations (CNVs) for complement C4 and its isotypes, C4A and C4B, in genetic risks and pathogenesis of JDM. The study population included 105 JDM patients and 500 healthy European Americans. Gene copy numbers

(GCNs) for total C4, C4A, C4B and HLA-DRB1 genotypes were determined by Southern blots and PCRs. Processed activation product C4d bound to erythrocytes (E-C4d) was measured by flow cytometry. Global gene-expression microarrays were performed in 19

JDM and 7 controls using PAXgene-blood RNA. Differential expression levels for selected genes were validated by qPCR.

Significantly lower GCNs and differences in distribution of GCN groups for total

C4 and C4A were observed between JDM and controls. Lower GCN of C4A in JDM remained among HLA DR3-positive subjects (p=0.015). Homozygous or heterozygous

C4A deficiency was present in 40.0% of JDM compared to 18.2% of controls [odds ratio

(OR)=3.00 (1.87-4.79), p=8.2x10-6]. JDM had higher levels of E-C4d than controls

(p=0.004). In JDM, C4A-deficient subjects had higher levels of E-C4d (p=0.0003) and

64 higher frequency of elevated levels of multiple serum muscle enzymes at diagnosis

(p=0.004). Microarray profiling of blood RNA revealed upregulation of type I

Interferon-stimulated genes and lower abundance of transcripts for T-cell and chemokine function genes in JDM, but this was less prominent among C4A-deficient or DR3- positive patients.

2.1 Introduction

Juvenile dermatomyositis (JDM) is a rare, autoimmune, multi-system inflammatory disease affecting primarily muscle and skin in children. Characteristic clinical features and diagnostic criteria include proximal muscle weakness and inflammation, increased levels of serum muscle enzymes, distinct skin rashes such as

Gottron’s papules or heliotrope rash, and pathological changes on muscle biopsy or magnetic resonance imaging (MRI) (Bohan and Peter 1975; Bohan and Peter 1975;

Dalakas and Hohlfeld 2003; Feldman et al. 2008; Rider and Miller 2011; Robinson and

Reed 2011; Dalakas 2015).

The HLA class II gene DRB1 allele *0301 (also known as DR3) has been identified as a major immunogenetic risk factor for JDM and was reaffirmed as the predominant risk locus of juvenile and adult dermatomyositis in a recent GWAS (Dalakas and

Hohlfeld 2003; Feldman et al. 2008; Robinson and Reed 2011; Miller et al. 2013). HLA class I and class II genes are engaged in antigen presentation and processing. The class III genes are heterogeneous in structures and function, and include genes encoding for components of the complement system C4, C2 and factor B, and for cytokines such as

65 tumor necrosis factor-α (TNF-α), and α and β lymphotoxins (Figure 8A) (Horton et al.

2004). Previous studies revealed that class II genes DRB1 allele *0301 and DQA1 allele

*0501, class III gene TNFA-308A allele, and class I gene variants B*08 and A*01 are in strong LD among human subjects of European descent and this is described as the ancestral haplotype AH8.1 (Friedman et al. 1983; Dawkins et al. 1999; Pachman et al.

2001). Also present in AH8.1 is a single C4B gene but the absence of a C4A gene

(Dawkins et al. 1999; Yu and Whitacre 2004; Horton et al. 2008). Remarkably, there are extensive inter-individual gene CNVs and gene-size dichotomy for complement C4

(Figure 9). Briefly, two to eight copies of C4 genes can be present in a diploid genome.

Each C4 gene either encodes for an acidic C4A or a basic C4B protein, with only four amino acid changes (PCPVLD 1120-1125 for C4A and LSPVIH for C4B), but these result in substantial differences in chemical reactivity for peptide and carbohydrate antigens (Isenman and Young 1984; Law et al. 1984; Yu et al. 1986; Kidmose et al.

2012).

Complement-mediated destruction of perivascular endothelium and perifascicular ischemia of muscle fibers in biopsies from dermatomyositis patients have been demonstrated by multiple investigators (Whitaker and Engel 1972; Kissel et al. 1986;

Emslie-Smith and Engel 1990; Kissel et al. 1991; Mascaro et al. 1995; Burgin et al.

2014). Circulating immune complexes, immunoglobulins IgG and IgM, complement C3 and C5b-9 membrane attack complex were shown in DM muscle and skin biopsies.

However, the initiation for complement activation and the potential role of complement in the breakdown of immune tolerance in JDM remain unclear. Continuous CNVs with

66 one to four copies of C4 genes on a haplotype with different combinations of C4A and

C4B genes in human populations have only been established since 1999 (Yang et al.

1999; Blanchong et al. 2000; Chung et al. 2002). Many earlier epidemiologic studies of complement C4A and C4B in rheumatic diseases, including JDM, were based on an incomplete or inaccurate model with two-locus (C4A-C4B) on a haplotype for data interpretation, and thus conclusions drawn became uncertain (O'Neill et al. 1978; Robb et al. 1988; Moulds et al. 1990; Reed et al. 1991). Here we perform a fresh and meticulous investigation of C4 genetic diversity and examine their effects on the risk and pathogenesis of JDM, with further considerations to the presence and absence of HLA-

DRB1 risk and protective alleles.

2.2 Materials and Methods

2.2.1 Study Populations

IRB approval was obtained from Nationwide Children’s Hospital (NCH) and the

National Institutes of Health (NIDDK/NIAMS, NIH). One hundred five JDM patients were enrolled, of which 45 were from NCH and 60 were from the NIH. Each patient met the diagnostic criteria for JDM according to the Bohan and Peter criteria (Bohan and

Peter 1975; Bohan and Peter 1975). Typical characteristic MRI abnormalities of muscle were applied in place of biopsy as a modification of the Bohan and Peter criteria for the

NCH cohort (Davis et al. 2011). The mean age (±SD) at recruitment was 10.8±7.6 years old, and at disease diagnosis was 7.4±4.2 years old. The self-reported racial distribution was 90.5% Caucasian, 6.7% African American, and 2.9% Hispanic. Complete

67 demographics and disease characteristics are displayed in Table 4. Ten non-Caucasian

JDM patients were excluded from genetic analysis. Race-matched healthy control subjects included 500 European Americans residing in Midwest-America.

2.2.2 Determination of total C4, C4A, and C4B Genotypes and Phenotypes

Previously, we described protocols for genotyping and phenotyping of complement C4 by Southern blot analyses and immunofixation, respectively (Sim and

Cross 1986; Chung et al. 2002; Chung et al. 2005). C4 plasma protein levels were assayed using Radial Immunodiffusion kits (The Binding Site, UK) (Table 6). In cases of limited DNA quantities or ambiguous results, quantitative real-time PCR experiments for

GCN of total C4, C4A, and C4B were performed as described (Wu et al. 2007). All C4-

CNV calls were validated by independent technology, or multiple amplicons in qPCR, and matched genotype and phenotype interpretations (Figures 8, 10; Table 5).

2.2.3 Flow Cytometric Detection of Erythrocyte-Bound Complement Activation

Fragments

Erythrocytes from whole blood were used for antibody staining and flow cytometry. Mouse monoclonal antibodies specific for human C4d, for human C3d, or the isotype-matched control MOPC21 (Quidel, San Diego, CA) were used (Manzi et al.

2004; Putterman et al. 2014). PE-conjugated goat anti-mouse IgG F(ab’)2 (Jackson

ImmunoResearch, West Grove, PA) was used as a secondary antibody. FlowJo software

(Tree Star Inc., Version 7.6) was used to electronically gate erythrocytes based on

68 forward and sideward scatter properties. Among the gated cells, E-C4d or E-C3d was reported as median fluorescence intensity (MFI), which was calculated using C4d- specific (or C3d-specific) MFI minus the MOPC-isotype control MFI.

2.2.4 HLA-DRB1 Typing

Genotyping of HLA-DRB1 alleles for JDM and all control samples were determined using genomic DNA for sequence-specific primer PCR (Hui and Bidwell

1993). HLA-DRB1 frequency was calculated by the number of allele-positive subjects divided by the total number of subjects.

2.2.5 Gene Expression Profiling

RNA was extracted from whole blood using PAXgene collection tubes

(PreAnalytiX, Becton, Dickinson and Company). Microarray analysis was performed by the Biomedical Genomics Core facility at the NCH. RNA samples passing quality control were labeled with Agilent’s One-Color microarray-based gene expression analysis labeling protocol and hybridized to the SurePrint G3 Human v2 GE 8x60K Microarray

(AMADID 039494). Images were analyzed with Feature Extraction 10.9 (Agilent

Technologies). Median foreground intensities were obtained for each spot and imported into the mathematical software package R. After pre-processing, the data were quantile normalized using the LIMMA package (Smyth and Speed 2003). Statistical analysis was performed via Significance Analysis of Microarray (SAM) implemented using the

Bioconductor Siggenes package to identify differentially expressed genes between JDM

69 and control groups (Tusher et al. 2001). Changes in expression ≥1.5 fold and a 15% false discovery rate as estimated by SAM were considered significantly different. Selected genes were validated by SYBR-Green qPCR using PAXgene RNA.

2.2.6 Statistical Analysis

Statistical analyses were performed using Prism6 (GraphPad Software Inc., San

Diego, CA) and JMP Genomics 6.0 (SAS Institute Inc., Cary, NC) software. Descriptive statistics are displayed as mean ± standard deviation (SD) for normally distributed data, and simple comparisons were made using Student’s t-test for continuous data, or by chi- square analysis for categorical data. Odds Ratios (OR) with 95% confidence intervals

(CI) are reported. For non-normally distributed data, median with interquartile range

(IQR) is reported, and Mann-Whitney test was used for comparisons. For all analyses, p≤0.05 was considered to be significant.

2.2.7 Additional Declarations

This was a major study with JDM patients recruited at Nationwide Children’s

Hospital (NCH) and National Institutes of Health (National Institute of Diabetes and

Digestive and Kidney Diseases (NIDDK)/National Institute of Arthritis and

Musculoskeletal and Skin Diseases (NIAMS), NIH. The following physicians were involved with referral and recruitment of JDM patients: NCH: Anjali Patwardhan,

Rabheh Abdul-Aziz, Charles Spencer, Stacy Ardoin, Sharon Bout-Tabaku, and Gloria

Higgins; NIH: Bita Arabshahi, Ruy Carrasco, Victoria Cartwright, Anne Eberhard,

70

Barbara Edelheit, Harry Gewanter, Donald Goldsmith, Beth Gottlieb, Thomas Griffin,

Melissa Hawkings-Holt, Roger Hollister, Yukiko Kimura, Patrick Knibbe, Lauren

Pachman, Maria Perez, Abigail Smukler, Sangeeta Sule, Carol Wallace, and Lawrence

Zeme. All healthy controls were recruited in central Ohio with the help of Jeanie Shaw,

RN at Nationwide Children’s Hospital. Gene expression profiling microarray analysis was performed by the Biomedical Genomics Core facility at NCH. Hierarchical clustering analysis and results were interpreted with the expertise of Peter White (NCH).

C4 gene copy numbers were determined by research associate Bi Zhou (NCH) using Southern blot analyses and by Yee Ling Wu (NCH) and Katherine Lintner (NCH) using qPCR assays. HLA genotypes were determined by Katherine Lintner (NCH) and by the NIH HLA laboratory. E-C4d levels were determined by Yee Ling Wu, Bi Zhou, and Katherine Lintner.

A manuscript of this work has been published. The citation is as follows: Lintner

K, Patwardhan A, Rider LG, Abdul-Aziz R, Wu YL, Lundstrom E, Padyukov L, Zhou B,

Alhomosh A, Newsom D, White P, Jones KB, O’Hanlon TP, Miller FW, Spencer CH,

Yu CY: Gene copy-number variations (CNVs) of complement C4 and C4A deficiency in genetic risk and pathogenesis of juvenile dermatomyositis. Ann Rheum Dis doi:

10.1136/annrheumdis.2015.207762

71

2.3 Results

2.3.1 Gene CNVs of Total C4, C4A, and C4B in JDM and Race-matched Healthy

Controls

Total C4 genes

In healthy controls (N=500), the variation of C4 GCN showed a normal distribution pattern, ranging from 2-7 total copies. In Caucasian JDM (N=95), total C4 genes ranged from 2-5 copies, with a shift of distribution to the lower copy number compared to controls [Χ2=20.7; degrees of freedom (df)=4; p=0.0004, Χ2 analysis)

(Figure 8B). JDM had a lower mean GCN than did controls (JDM: 3.49±0.71; controls:

3.83±0.69; p=1.8x10-5, t-test; Table 7). Low copy number of total C4 (C4T, GCN≤3) was present in 50.5% of JDM and 29.2% of controls [OR=2.48 (1.59-3.87); p=7.5x10-5]. The means of normalized C4 protein levels positively correlated with GCN of total C4

(Figure 11). However, it is clear that there are large degrees of variations in protein levels within each GCN group.

C4A and C4B Genes

The reduction in copy number of total C4 in JDM can be the result of a decrease in C4A, C4B, or both. We observed a significant shift to lower GCN of C4A in JDM patients (p=0.0004). The presence of homozygous or heterozygous deficiency of C4A genes (GCN=0 or 1) had a frequency of 40.0% in JDM, compared to 18.2% in controls

72

[OR=3.00 (1.87-4.79); p=8.2x10-6]. The overall mean GCN of C4A observed in JDM was

1.79±0.86 compared to 2.09±0.75 in controls (p=0.0004).

As for C4B, no significant differences in the distribution of C4B-GCN or C4B- deficiency were observed between JDM and controls. Therefore, the basis for decreased

GCN of total C4 in Caucasian JDM patients was attributable to lower GCN of C4A

(Figure 8).

2.3.2 HLA-DRB1 Alleles, C4A-GCN, and C4A Deficiency on Disease Risks of JDM

The frequency of HLA-DRB1*0301 alleles (DR3) was 46.3% in Caucasian JDM patients (N=95), compared with 25.8% in a race-matched healthy population (N=500).

HLA-DR3 was associated with JDM with an OR=2.48 (1.58-3.89) and a p-value of

9.5x10-5. The concurrence of C4A deficiency and DR3 in a subject was present in 36.8% of JDM and 15.4% of controls, with an odds ratio of 3.20 (1.98-5.19) and a p-value of

4.8x10-6. By contrast, the frequency of HLA-DRB1*1501 (DR2) was 11.6% in Caucasian

JDM compared to 27.8% in healthy controls. DRB1*1501 was a protective factor against

JDM with OR=0.34 (0.18-0.66) and p=0.0004 (Figure 8E; Table 7A).

Multiple logistic regression analyses were performed to investigate if C4A deficiency, the presence of DR3 and the presence of DR2 could serve as independent risk factors for JDM, conditional upon presence of other factor(s) in five different combinations of regression models (Table 7B). In models when DR3+ and C4A deficiency were put together as individual factors (model a or b), the relevance of DR3+ as an independent parameter became insignificant. The presence of DR2 plus C4A

73 deficiency, or DR3+, or C4A deficiency with DR3+ all yielded statistical significance to account for JDM genetic risks. The last model yielded the best fit: C4A deficiency plus

DR3+ was a strong risk factor with odds ratio of 2.96 (1.84-4.80) and DR2 was protective factor with odds ratio of 0.34 (0.21-0.55).

To further evaluate the relative roles of C4A deficiency and DR3 on disease risk of JDM, we performed subgroup analyses (Table 7C). Among the DR3+ subjects, JDM patients had a significantly lower mean-GCN of C4A than controls (JDM: 1.18±0.54; controls: 1.47±0.72; p=0.015). Similarly among the DR3+ subjects, C4A deficiency had a greater prevalence in JDM (79.6%) than controls (59.7%) [OR=2.63 (1.17-5.92), p=0.014]. Among DR3-negative subjects, however, there were no apparent differences in mean-GCN of C4A or the prevalence of C4A deficiency between JDM and controls, suggesting the heightened risk of lower C4A-GCN or C4A deficiency on JDM required a

DR3+ background.

Reciprocal subgroup analyses to compare the prevalence of DR3+ between JDM and controls among C4A-deficient subjects (GCN≤1), or among C4A-proficient subjects

(GCN≥2) revealed slight but insignificant increases in the frequency of DR3 in JDM

(Table 7C).

2.3.3 Levels of Erythrocyte-bound C4d (E-C4d) or C3d (E-C3d) in JDM and Controls

Cell-bound complement levels were determined in 40 Columbus JDM patients and 206 healthy subjects of European ancestry by flow cytometry. Comparing between

JDM and healthy controls, significant elevation of E-C4d levels (p=0.004, Mann Whitney

74 test, Figure 12A) but not E-C3d levels (Figure 12B) was observed in JDM. The median fluorescent intensities (MFI) for E-C3d levels were substantially lower than those of E-

C4d levels, which is consistent with the presence of complement regulation mechanisms on self-cell surfaces.

We investigated if there was a correlation of E-C4d levels with C4A or C4B gene dosages in JDM (Figure 12 C&D). The C4A deficiency group (GCN≤1; N=15) had a

MFI of 1426 (IQR: 601-1744), which was significantly higher than that of the C4A- proficient group (GCN≥2; N=25; MFI=454 (234-718); p=0.0003, Mann-Whitney test).

On the other hand, the C4B-deficiency group (GCN≤1; N=11) had a median E-C4d MFI of 308 (226-505), which was significantly lower than that of the C4B-proficient group

(GCN≥2, N=29; MFI=775 (495-1458); p=0.003, Mann-Whitney test). Thus, C4A and

C4B appeared to play opposite roles on the deposition of cell-bound E-C4d: high GCN of

C4A dampened activation, while high GCN of C4B amplified activation.

2.3.4 JDM Patients With Elevated Levels of Multiple Serum Muscle Enzymes had

Low GCN of C4A

JDM patients exhibited elevated levels of a variety of serum muscle enzymes.

We performed intragroup comparisons to investigate if C4A deficiency was related to elevated muscle enzyme levels at the time of disease diagnosis. Indeed, patients with C4A deficiency had higher prevalence of abnormal serum muscle enzymes such as creatine kinase (C4A-deficient: 86.1%, C4A proficient: 58.2%; p=0.0034) and aldolase (C4A- deficient: 94.1%, C4A-proficient: 78.4%; p=0.038) and elevations of multiple serum

75 muscle enzymes (C4A-deficient: 97.1%; C4A- proficient: 74.5%; p=0.0025) than C4A- proficient patients. However, the prevalence of elevated levels of serum aspartate aminotransferase and lactate dehydrogenase was not associated with C4A deficiency

(Table 8).

2.3.5 Differential Gene Expression Profiling of JDM and Controls

Global gene expression profiling was performed using PAXgene blood RNA from

19 consecutive JDM patients and 7 controls (Figure 13; Tables 9-11). Expression profiles revealed differential expression of transcripts in JDM from 56 genes that were significantly different using SAM criteria. Differentially expressed genes included 24 upregulated and 32 downregulated genes (Table 11). Of the upregulated genes, the most remarkable are nine type I interferon (IFN-I) response genes and three genes related to B- cell functions. Of the downregulated genes, the most distinct were genes related to T-cell functions, chemokines and chemokine receptors. Six JDM patients exhibited the most polarized upregulation of IFN-I genes and downregulation of genes for chemokine/chemokine receptor and T-cell functions (Figure 13A). Of interest, five of these six JDM patients were C4A-proficient (GCN≥2), and did not carry the HLA-DR3 allele.

To validate gene expression changes from microarray, we performed SYBR-

Green qPCR using cDNA from PAXgene-blood RNA for IFI44, IFI17, CXCR6, and

CCR5 transcripts. Results revealed in JDM upregulation of IFI44 (2.7-fold; p=0.028) and

IFI17 (3.0-fold; p=0.0054), and downregulation of CXCR6 (1.9 fold; p=0.031) and CCR5

76

(1.5 fold; p=0.048) (Figure 13B). The housekeeping gene GAPDH was used as a normalization standard (Willems et al. 2008).

2.4 Discussion

Genetic markers or polymorphic variants of HLA class II and class I alleles have been associated with juvenile and adult DM through conventional genotyping techniques over the past three decades (Friedman et al. 1983; Reed and Stirling 1995; Mamyrova et al. 2006) and by a recent GWAS (Miller et al. 2013). Among Caucasians, genetic risk factor(s) for JDM have been found to be strongest in the HLA, particularly with haplotypes that consisted of polymorphic markers or alleles of class I gene B*08 and class II genes DR3 or DRB1*0301 and DQA1*0501 (Mamyrova et al. 2006). Among human subjects of north and western European ancestry, these alleles exist together in strong LD as ancestral haplotype AH8.1. Besides JDM, AH8.1 is a common risk haplotype for several other autoimmune diseases including SLE, type 1 diabetes (T1D), and Sjögren’s Syndrome (Dawkins et al. 1999; Lessard et al. 2013).

A great challenge for studying complex diseases associated with the HLA region, including JDM, is to determine which gene(s) or polymorphic variants contribute to disease development under the background of strong allelic associations or LD. There are many unique features of the HLA region that confound traditional genetic association analyses, including: (a) the extreme allelic polymorphisms of class I and II genes, some with hundreds to thousands of alleles (e.g., HLA-DRB1 and HLA-B); (b) the large number of genes present in the HLA locus at 6p21.3, including 57 different structural genes in the

77 class III region; (c) the inter-individual gene copy number variations for DRB in the class

II region and for C4 modules in the class III region; and (d) the long-range LD of variant markers for class I, II, and III genes spanning thousands of Kb of genomic DNA.

Additionally, many HLA haplotypes appear to be ancient but race-specific (Yu and

Whitacre 2004). Therefore, it is difficult to identify definitive genetic susceptibility factors without examining potential candidates individually and systematically.

This study aims to decipher the gene CNVs for complement C4 and its isotypes C4A and

C4B in JDM with definitive techniques, and to dissect the relative roles of HLA-

DRB1*0301 (DR3) and C4A deficiency on JDM disease risk in subjects of European ancestry.

The carriage of DR3 was present in 46.3% of our JDM patients compared to

25.8% of race-matched healthy controls. Therefore, DR3 is a medium effect-size risk factor for JDM (OR=2.48). Homozygous and heterozygous deficiency of complement

C4A had a frequency of 40.0% in Caucasian JDM and 18.2% in healthy controls

(OR=3.00). The co-existence of HLA-DR3 and C4A deficiency confers higher risk than either individual risk factors, and such concurrence in a diploid genome was present in

36.8% of JDM and 15.4% of controls with an OR=3.20. The independent roles of DRB1 variants and C4A deficiency in JDM were further validated by multiple logistic regression analyses. Moreover, among DR3-positive subjects, lower mean GCN of C4A or higher prevalence of C4A deficiency persisted in JDM versus controls. An interpretation to this phenomenon is that DR3-positivity contributes to a permissive background and C4A deficiency significantly elevated the vulnerability to an autoimmune

78 disease including JDM. While HLA-DRB1 and complement C4 each is engaged in specific immunologic functions such as antigen presentation to effector T-cells, and complement-mediated cytolysis and immunoclearance, they are both involved in the recognition of self and non-self, and are key players for the process of archiving memory and tolerance in the immune system.

Destruction of perifascicular capillaries by complement and subsequent ischemia of muscle fibers in dermatomyositis have been demonstrated by multiple investigators over the past three decades (Whitaker and Engel 1972; Kissel et al. 1986; Emslie-Smith and Engel 1990; Kissel et al. 1991; Mascaro et al. 1995; Burgin et al. 2014). Activation of complement can be initiated via the classical pathway that is triggered by immune complexes formed between myositis-associated or myositis-specific autoantibodies and self-antigens abundant in muscles and skin. Physiologically, low GCN or low production of C4A protein systemically may dampen immune complex clearance and therefore promote autoimmunity. Compared with controls, we observed a moderate but significant increase in the deposition of C4d on erythrocytes among JDM patients, which reiterates involvement of complement activation in the pathogenesis of JDM. Almost all JDM patients with two or more elevated muscle enzymes at disease diagnosis had a C4A deficiency but normal mean GCN of C4B. In other words, immune-mediated tissue injuries in JDM might have been resulted through the activation of C4B. Consistent with this phenomenon, we observed increased deposition of processed complement activation product E-C4d in JDM patients than in controls. Interestingly, the levels of E-C4d were directly proportional to the GCN of C4B, and inversely proportional to the GCN of C4A.

79

Physiologically, activated C4B protein is highly reactive and over-activation could lead to complement-mediated injuries. In addition to its role in immunoclearance and protection against autoimmunity, the presence of activated C4A may attenuate the activity of C4B and minimize its potential deleterious effect.

Among the JDM patients, 63.2% were not associated with C4A deficiency on a

DR3+ background, and the underlying genetic risk factors in this group of patients (C4A- proficient and DR3-negative) are yet to be identified. An emerging feature in JDM is the upregulation of type I interferon-stimulated gene expression in many patients (Tezak et al. 2002; Walsh et al. 2007; Baechler et al. 2011). Our microarray studies of peripheral blood samples revealed marked increase in transcripts in JDM for many IFN-I stimulated genes and B-cell specific genes, but diminished transcript levels of many genes related to chemokines and T-cell functions. Such differential levels of transcripts reflect both different gene expression levels and also compositions of leukocytes in the peripheral blood samples (Figures 14&15). The polarized upregulation of IFN-stimulated genes and

B-cell function genes, and downregulation of chemokine receptor and T cell function genes were more marked among C4A-proficient or DR3-negative patients, implying the presence of additional or alternative mechanisms leading to the pathogenesis of JDM.

In conclusion, we report the novel finding of low GCN of complement C4 and

C4A deficiency associated with JDM. Many hypotheses exist as to why a partial deficiency of complement C4 (in particular isotype C4A) drives an autoimmune response in humans. Popular and sustainable theories conclude that C4A deficiency results in: impaired removal of apoptotic cells and immune complexes (“debris clearance”; “waste-

80 disposal” hypothesis) (Walport 2001) and/or predisposition to abnormal regulation of autoreactive B cells (“regulation of self-tolerance”) (Paul, 2002; Prodeus, 1998,

Chatterjee, 2013). Indeed, mouse models deficient in C4 are characterized by autoimmune response, including high titers of spontaneous antinuclear antibodies, glomerular deposition of immune complexes, and glomerulonephritis (Chen et al. 2000).

In the HLA region, class II haplotypes are hypothesized to be associated with autoimmune diseases because of the ability of specific alleles to code for molecules that bind and present different, specific self-peptides. In addition, JDM patients with C4A deficiency were more likely to have elevated levels of multiple serum muscle enzymes at diagnosis and high levels of E-C4d. Further in-depth studies through HLA-DRB1 and

C4A genotypes, cell-bound C4d levels and differentially expressed genes including those engaged in muscle-cell functions and signaling, and characterization of clinical phenotypes (Shah et al. 2013) may help to understand the pathogenic mechanisms, to enable patient stratification, and to facilitate genotype and gene expression guided therapies in patients with JDM.

81

Figure 8 Variations of C4 haplotypes and gene copy numbers (GCNs) of total C4, C4A and

C4B in JDM subjects and race-matched healthy controls

(continued)

82

A. A simplified map of the human major histocompatibility complex (MHC) showing genes of immunologic functions. B, C and D. Gene copy number variations of complement C4, C4A and C4B in JDM and controls. E. A summary of genetic risk factors in the MHC for JDM.

83

Figure 9 Genetic diversity of complement C4

A. Copy number variation (CNV) of RP-C4-CYP21-TNX (RCCX) modules in the class

III region of the HLA. B. Gene size dichotomy of long and short C4 genes. Exon-intron

structures of long and short C4 genes are shown. The long gene is due to endogenous

retrovirus integrated into intron 9 of the C4 gene. C. DNA sequences and derived amino

acid sequences determining C4A and C4B. C4A and C4B isotypic sequences are present

in Exon 26 of the C4 genes. The long/short C4 genes and C4A/C4B genes are (continued) 84 independent structural features of C4 genes that can be distinguished by TaqI RFLP and

PshAI RFLP, respectively.

85

Figure 10 Variations of C4 haplotypes and gene copy numbers (GCNs) of total C4, C4A and C4B, and C4 protein polymorphisms in eight selected JDM subjects

(continued)

86

A. Specific C4 haplotype structures were determined by long range mapping by PFGE following PmeI digestion of genomic DNA and hybridization to a C4d-specific probe.

One to three copies of C4 genes in an RCCX module on each Chromosome 6 were demonstrated. B. Total GCNs of C4 and details on long and short C4 genes together with their neighboring genes RP1 and RP2 were determined by Southern Blot analyses using

TaqI-digested genomic DNA. C. Relative dosages of C4A and C4B copy numbers were determined by PshAI-PvuII RFLP. D. C4A and C4B protein polymorphisms were revealed by immunofixation of EDTA-plasma resolved by high-voltage agarose gel electrophoresis. For example, Patient No. 30 (first lane) had homozygous single short (S)

C4 genes with 107 kb PmeI fragment, which was confirmed by the presence of the 6.4 kb

TaqI fragment (Panel B) and the presence of single C4B gene and absence of C4A gene

(Panel C). Immunofixation of plasma revealed the presence of C4B1 protein and a deficiency of C4A protein (Panel D).

87

Figure 11 Relationship between C4 plasma protein levels plotted versus total C4 GCN in

95 European American JDM subjects

C4 plasma levels are normalized and graphed as fold changes, which were calculated as

the fold change compared to the mean plasma level of either the Columbus plasma

samples or NIH plasma samples GCN, gene copy number. Red dotted line represents the

trend line.

88

Figure 12 Erythrocyte-bound E-C4d and E-C3d in JDM patients and controls

A and B. A comparison of E-C4d and E-C3d, respectively, in JDM and controls. C. A

comparison of E-C4d in C4A-deficient and C4A-proficient JDM patients. D. A

(continued)

89 comparison of E-C4d in C4B-deficient and C4B-proficient JDM patients. The median for each group is indicated by a horizontal bar, while the shorter bars represent interquartile ranges; the p-value for Mann Whitney test is indicated.

90

Figure 13 Gene expression profiling of PAXgene blood RNA in JDM and healthy controls

(continued)

91

A. Hierarchical clustering analysis of PAXgene blood RNA gene expression microarray data in 19 JDM and 7 healthy subjects; red and green represent upregulated and downregulated genes, respectively; vertical columns represent the data for each subject, and rows indicate different genes (see Table 11 for details). B. Four genes from microarray results were selected for SYBR-green qPCR analysis using cDNA from

PAXgene blood RNA to validate the upregulation and downregulation in JDM of the chosen genes. The white column shows normalized value to 1 for controls, while the fold-change in JDM for each respective gene is indicated by colored columns; p-values for Mann-Whitney tests are indicated. Number of subjects (N) used for SYBR-Green qPCR assays are: IFI44 - 28 JDM and 14 controls; IFI17 - 17 JDM and 15 controls;

CXCR6 - 24 JDM and 19 controls; and CCR5 - 24 JDM and 19 controls.

92

Figure 14 Correlations of transcript levels for interferon-stimulated genes IFI17 (left) and

IFI44 (right) with polymorphonuclear cells present in leukocytes of JDM blood samples

93

Figure 15 Correlation of CCR5 transcripts with lymphocytes present in leukocytes of JDM blood samples

94

Table 4 Demographic features and clinical characteristics of JDM patients

Features N (%)

Age at recruitment: mean ± SD, yrs. old 10.8 ± 7.6

Age at diagnosis: mean ± SD, yrs. old 7.4 ± 4.2

Sex: female / male 67 (63.8%) / 38 (36.2%)

Race/Ethnicity: Caucasian / African / Hispanic 95 (90.5%) / 7 (6.7%) / 3 (2.9%)

Calcinosis* 15 / 97 (15.5%)

Ulcerations 17 / 97 (17.5%)

Lipodystrophy 8 / 96 (8.3%)

Disease course†

Monocyclic 16 / 66 (24.2%)

Polycyclic 12 / 66 (18.2%)

Chronic continuous 38 / 66 (57.6%)

Positive ANA 62 / 82 (75.6%)

* Values given indicate number of subjects (percentage) for which data was available.

† Only patients who have been followed ≥2 years were categorized by disease courses.

95

Table 5 RCCX haplotypes (H1 and H2), complement C4B and C4A GCN and protein polymorphism, and HLA-DRB1 haplotypes (H1 and H2) in eight JDM patients

DM RCCX RCCX C4B C4A C4B C4A DRB1* DRB1* ID# H1 H2 protein protein H1 H2 GCN GCN

30 S S 2 0 B1,B1 Q 03 03

18 LS S 2 1 B1,B1 A3 01 03

29 LL S 2 1 B1,B1 A3 15 03

31 LS LS 2 2 B2B2 A4,A2 04 08

03† LS LS 1 3 B3¶ A91¶A12¶A3 03 07

22 LL LL 1 3 B1 A3,A3,A3 04 15

32 LLL LS 0 5 Q A3,A3,A3,A3,A3 04 13

24 LL LS 2 2 B1,B1 A3,A3 03 07

GCN, gene copy number; † DM03 was African American; H1, haplotype 1; H2,

haplotype 2; ¶B3, A91 and A12 had similar electrophoretic mobilities in an agarose

typing gel.

96

Table 6 A comparison of plasma protein levels for complement C4 and C3 between JDM and race-matched controls

JDM Control p-value C4 protein, mg/dL: median (IQR) 28.3 (22.4-35.2) 30.1 (23.8-37.4) <0.0001 N 33 361 C3 protein, mg/dL: median (IQR) 141.4 (129.7-163.5) 170.7 (145.2-208.7) <0.0001 N 33 361

97

Table 7 HLA-DRB1 Alleles and C4 Gene CNVs in JDM and controls

A. Analysis of single genetic risk factors in JDM

JDM (N=95) Control (N=500) p-value OR (95% CI)

a. Distribution, N (%) C4T GCN ≤ 3 48 (50.5%) 146 (29.2%) 7.5x10-5 2.48 (1.59-3.87) C4A deficiency, GCN ≤1 38 (40.0%) 91 (18.2%) 8.2x10 3.00 (1.87-4.79) C4B deficiency, GCN ≤1 28 (29.5%) 153 (30.6%) 0.83† HLA DRB1*1501 11 (11.6%) 139 (27.8%) 0.0004 0.34 (0.18-0.66) HLA DRB1*0301 (DR3+) 44 (46.3%) 129 (25.8%) 9.5x10 2.48 (1.58-3.89) C4A deficiency with DR3+ 35 (36.8%) 77 (15.4%) 4.8x106 3.20 (1.98-5.19)

b. Mean GCN ± SD Total C4 genes 3.49 ± 0.71 3.83 ± 0.69 1.8x10-5 C4A genes 1.79 ± 0.86 2.09 ± 0.75 0.0004 C4B genes 1.71 ± 0.58 1.73 ± 0.63 0.68†

B. Multiple logistic regression models for genetic factors of JDM

Models and parameters R2 2 p-value OR (95% CI) a. C4A deficiency, DRB1*1501 and DR3+ 0.058 30.2 1.2x106 DRB1*1501 8.95 0.0028 0.39 (0.19-0.74) C4A deficiency 6.21 0.013 2.27 (1.19-4.40) DR3+ 0.70 0.40† 1.31 (0.69-2.44) b. C4A deficiency and DR3+ 0.041 21.3 2.4x10 C4A deficiency 6.05 0.014 2.26 (1.18-4.39) DR3+ 1.39 0.24† 1.47 (0.77-2.74) c. C4A deficiency and DRB1*1501 0.057 29.5 3.9x107 DRB1*1501 9.64 0.0045 0.38 (0.19-0.71) C4A deficiency 16.8 2.7x10 2.76 (1.71-4.43) d. DR3+ and DRB1*1501 0.046 24.0 6.1x106 Continued

98

Table 7 continued DRB1*1501 8.80 0.003 0.40 (0.19-0.74) DR3+ 11.3 0.0008 2.21 (1.39-3.49) e. C4A deficiency with DR3+ and DRB1*1501 0.059 30.6 2.2x107 C4A deficiency with DR3+ 17.9 2.4x10 2.96 (1.81-4.80) DRB1*1501 9.7 0.0018 0.34 (0.21-0.55)

C. Subgroup analyses of C4A-GCN, C4A deficiency and HLA-DR3 in JDM JDM Control p-value OR (95% CI)

Mean C4A GCN, DR3+ 1.18±0.54 1.47±0.72 0.015 Mean C4A GCN, DR3  2.31±0.73 2.31±0.63 0.97†

C4A-deficient, DR3+, N (%) 35 (79.6) 77 (59.7) 0.014 2.63 (1.17-5.92) C4A-proficient, DR3+, N (%) 9 (20.5) 52 (40.3)

C4A-deficient, DR3 , N (%) 3 (5.9) 14 (3.8) 0.50† C4A-proficient, DR3 , N (%) 48 (94.1) 357 (96.2)

DR3+, C4A-deficient, N (%) 35 (92.1) 77 (84.6) 0.23† DR3 , C4A-deficient, N (%) 3 (7.9) 14 (15.4)

DR3+, C4A-proficient, N (%) 9 (15.8) 52 (12.7) 0.53† DR3 , C4A-proficient, N (%) 48 (84.2) 357 (87.3)

Abbreviations: C4T, total copies of C4 genes; CI, confidence interval; GCN, gene copy number; N, number; OR, odds ratio. C4A-deficient: C4A GCN =0 or 1; C4A-proficient:

C4A GCN ≥2; †, not statistically significant; categorical data were compared by Χ2 analyses; continuous data were compared by t-tests.

99

Table 8 Elevation of serum muscle enzymes at disease diagnosis in JDM patients with and without C4A deficiency

Muscle Enzyme N (%) with elevation of muscle enzyme levels

C4A-deficient C4A-proficient p-value*

Aldolase 32 (94.1) 40 (78.4) 0.038

Aspartate aminotransferase 27 (81.8) 36 (67.9) NS

Creatine kinase 31 (86.1) 32 (58.2) 0.0034

Lactate dehydrogenase 24 (82.8) 31 (77.5) NS

≥ 2 Muscle enzymes 33 (97.1) 38 (74.5) 0.0025

*by Χ2 analysis; NS, not significant.

100

Table 9 RCCX structures and HLA-DRB1 genotypes of JDM patients for microarray studies

Code-2 Sex RCCX-1 RCCX-2 C4B C4A DRB1_H1 DRB1_H2 m_DM01 F LL LS 0 4 18 13.1 m_DM02 F LLS S 2 2 03 13.1 m_DM03 F LL LS 2 2 09 11 m_DM04 F LL LL 2 2 01 14 m_DM05 M LL LL 1 3 15 04 m_DM06 F LL LS 2 2 01 04 m_DM07 F LS S 2 1 03 07 m_DM08 M LS S 2 1 03 07 m_DM09 F LS LS 1 3 07 18 m_DM10 M LS S 2 1 07 03 m_DM11 M LLL LS 1 4 01 01 m_DM12 F LL LS 2 2 07 15 m_DM13 M LL S 2 1 15 03 m_DM14 F LS S 2 1 01 03 m_DM15 M LL LL 1 3 04 04 m_DM16 F LL S 1 2 01 03 m_DM17 F LL S 2 1 03 11 m_DM18 F LS LS 2 2 04 08 m_DM19 M LSS S 3 1 03 07

The RCCX genetic backgrounds and the HLA-DRB1 genotypes of the 19 JDM PAXgene peripheral blood samples subjected to global gene expression analyses using Agilent microarrays.

101

Table 10 Patient demographics, C4 CNV, HLA-DRB1 genotypes, serum muscle enzymes and medications of 19 JDM patients (at recruitment) for microarray gene profiling experiments

ID Age Age Sex Race Total C4B C4A DRB1 DRB1 Elevated Disease Current

(current) (diagnosis) C4 GCN GCN H1 H2 Muscle Status Treatment

GCN Enzymes

DM04 18.7 17.9 F W 4 2 2 1 14 AST, CK Active MTX

DM12 16.6 13.1 F W 4 2 2 7 15 Aldolase, Active MTX 102

AST, CK

DM18 17.3 17.3 F W 4 2 2 4 8 Aldolase, Active MTX, Pred

AST

DM05 14.6 8 M W 4 1 3 4 15 None Active MTX

DM10 9.5 6.3 M W 3 2 1 3 7 None Active IVIG, MTX,

Pred

DM09 10.9 8.8 F B 4 1 3 7 18 None Active IVIG, Pred

102 Continued

Table 10 continued

DM08 14.7 10.7 M W 3 2 1 3 7 None Flare IVIG, MTX,

Pred

DM06 10.5 4.1 F W 4 2 2 1 4 None Remission None

DM16 15.7 14.1 F W 3 1 2 1 3 n/a Remission MTX

DM17 3.1 1.7 F W 3 2 1 3 11 None Remission MTX

DM03 12.6 9.6 F B 4 2 2 9 11 None Remission MTX

DM01 6.0 2.9 F B 4 0 4 13.1 18 Aldolase Active MTX 103

DM15 9.4 4.4 M W 4 1 3 4 4 None Remission None

DM19 12.0 5.6 M W 4 3 1 3 7 None Remission IVIG, Pred

DM14 8.9 2.1 F W 3 2 1 1 3 None Remission MTX

DM02 13.8 6 F W 4 2 2 3 13.1 None Active IVIG, Dexa-

methasone

DM07 10.6 3.7 F W 3 2 1 3 7 None Active MTX

103 Continued

Table 10 continued

DM13 7.7 4 M W 3 2 1 3 15 Aldolase Flare MTX

DM11 7.2 4 M W 5 1 4 1 1 none Remission MTX, Pred

F, female; M, male; W, White; B, Black; RCCX: RP-C4-CYP21-TNX; GCN, gene copy number; H1, haplotype 1; H2,

haplotype 2; AST, aspartate aminotransferase; CK, creatine kinase; MTX, methotrexate; Pred, prednisone; IVIG, intravenous

immunoglobulin IgG.

104

104

Table 11 Differentially expressed genes in blood samples of JDM patients.

Gene Symbol Gene Name GenBank Mean Fold- putative functions Accession Change* Interferon response - antiviral IFI44 interferon-induced protein 44 NM_006417 2.9 EPSTI1 epithelial stromal interaction 1 (breast) NM_033255 2.6 USP18 ubiquitin specific peptidase 18 NM_017414 2.1 downregulates interferon responses OAS3 2'-5'-oligoadenylate synthetase 3 NM_006187 2.2

105 100kDa; binds and activates RNase L, inhibition of cellular protein synthesis and viral infection resistance EIF2AK2 eukaryotic translation initiation factor 2-alpha kinase 2 NM_001135652 2.1 inhibition; IFN-stimulated IFITM1 interferon induced transmembrane protein 1 (9-27) NM_003641 1.6 ZC3HAV1 zinc finger CCCH-type, antiviral 1 NM_024625 1.5 prevents infection by retroviruses TRIM5 tripartite motif containing 5 NM_033092 1.5 E3 ubiquitin-ligase and ubiqutinates itself, retroviral restriction PYHIN1 pyrin and HIN domain family, member 1; NM_198930 -1.6 HIN-200 family of interferon-inducible proteins

Immune system B cell functions VPREB1 pre-B lymphocyte 1 NM_007128 1.8 Continued 105

Table 11 continued

FCRL1 Fc receptor-like 1 NM_001159397 1.6 IL37 interleukin 37 NM_014439 1.5 IL-1 family, binds IL-1 receptor T cells functions CD2 CD2 molecule NM_001767 -1.5 TIGIT T cell immunoreceptor with Ig and ITIM domains NM_173799 -1.6 ANKRD35 ankyrin repeat domain 35 NM_144698 -1.6 MID2 midline 2 NM_012216 -1.7 localizes to microtubular structures in the cytoplasm, TRIM member; antiviral 106 WNT1 wingless-type MMTV integration site family, member 1 NM_005430 -1.7

mesencephalon and cerebellum, cell signaling, embryogenesis; T cell development CTSW cathepsin W NM_001335 -1.7 regulation of T-cell cytolytic activity; T cells CTSZ cathepsin Z NM_001336 -1.8 lysosomal cysteine proteinase, carboxyl mono and di-peptidase activities EOMES eomesodermin NM_005442 -1.9 differentiation of effector CD8+ T cells ; essential during trophoblast development and gastrulation; T-box DNA binding protein RORC RAR-related orphan receptor C NM_005060 -2.4 nuclear hormone receptors; DNA binding lymphoid organogenesis ; inhibit the expression of Fas ligand and IL2 GZMK granzyme K NM_002104 -2.5 106 Continued

Table 11 continued

granzyme 3; tryptase II Innate immunity /chemokine receptors CFH complement factor H NM_001014975 -1.7 CCR5 chemokine (C-C motif) receptor 5 NM_000579 -2.3 expressed on T cells and macrophages CXCR3 chemokine (C-X-C motif) receptor 3; NM_001504 -2.2 G protein-coupled receptor with selectivity for three chemokines CXCR6 chemokine (C-X-C motif) receptor 6 NM_006564 -2.5 homing of T-cells to skin

107

Blood vessel functions PHACTR1 phosphatase and actin regulator 1 ENST00000379350 1.7 blood vessel, endothelial cells; a key component in the angiogenic process CYP4A11 cytochrome P450, family 4, subfamily A, polypeptide 11 NM_000778 1.6 coronary artery disease ZNF831 zinc finger protein 831 NM_178457 -1.5 blood pressure EMILIN1 elastin microfibril interfacer 1 NM_007046 -1.5 blood vessels KLKB1 kallikrein B, plasma (Fletcher factor) 1 NM_000892 -1.6 surface-dependent activation of blood coagulation, fibrinolysis, kinin generation and inflammation

Signaling pathways, membrane transports XAF1 XIAP associated factor 1 NM_017523 2.1 107 Continued

Table 11 continued

apoptosis signaling GNG11 guanine nucleotide binding protein (G protein), gamma 11 NM_004126 1.9 guanine nucleotide-binding protein CDCA7 cell division cycle associated 7 NM_031942 1.6 c-Myc - phosphorylated by Akt-transformation; apoptosis KIAA1958 KIAA1958 ENST00000374244 1.5 PLEKHG3 pleckstrin homology domain containing family G (with RhoGef NM_015549 -1.5 domain) member 3; rhoGTPase exchange PERP PERP, TP53 apoptosis effector NM_022121 -1.7

108 p53 effector

IGFBP3 insulin-like growth factor binding protein 3 NM_001013398 -1.8

proapoptotic and antiproliferative; TGF- signaling -receptor

Nervous system, synapses, Ca+2, muscle functions SNX22 sorting nexin 22 NM_024798 1.7 BAALC brain and acute leukemia, cytoplasmic ENST00000297574 1.7 GABRA3 gamma-aminobutyric acid (GABA) A receptor, alpha 3 NM_000808 1.7 SAMD9L sterile alpha motif domain containing 9-like NM_152703 1.7 mutated in familial tumoral calcinosis; cell proliferation, lost in mouse ENKUR enkurin, TRPC channel interacting protein NM_145010 1.6 signal transduction to Ca-channels ABCC3 ATP-binding cassette, sub-family C (CFTR/MRP), member 3 NM_001144070 1.6 NGFRAP1 nerve growth factor receptor (TNFRSF16) associated protein 1 NM_014380 1.5 PARD3 par-3 partitioning defective 3 homolog (C. elegans) NM_001184792 1.5

108 Continued

Table 11 continued polarized cell growth, neural tube DLG3 discs, large homolog 3 (Drosophila) NM_021120 -1.5 excitoray synapses CAMK2N1 calcium/calmodulin-dependent protein kinase II inhibitor 1 NM_018584 -1.5 negative regulation of protein kinase CACNA1I calcium channel, voltage-dependent, T type, alpha 1I subunit NM_021096 -1.5 CACNA2D2 calcium channel, voltage-dependent, alpha 2/delta subunit 2 NM_001005505 -1.6 early infantile epileptic encephalopathy TMED9 transmembrane emp24 protein transport domain containing 9 NM_017510 -1.7 Golgi and ER

LRFN3 leucine rich repeat and fibronectin type III domain containing 3 NM_024509 -1.7 109 adhesion, synapses CLIC5 chloride intracellular channel 5 NM_016929 -1.8 associates with actin-based cytoskeletal structures; hair cell stereocilia formation, myoblast proliferation and glomerular podocyte and endothelial cell maintenance ATXN7L2 ataxin 7-like 2 NM_153340 -1.9 SLC4A10 solute carrier family 4, sodium bicarbonate transporter, member NM_022058 -2.3 10 HCO3-, Cl- transport; regulates pH

Other LOC144571 uncharacterized LOC144571 NR_026971 -1.8 * from 19 JDM patients and 7 healthy controls.

109

Chapter 3: Immunogenetic Studies of C4 Gene Copy Numbers and HLA-DRB1 Alleles

in Systemic Lupus Erythematosus (SLE) in Caucasian Populations

Abstract

Homozygous or heterozygous genetic deficiencies of Complement C4A have been reported to be associated with systemic lupus erythematosus (SLE) disease risk. We and others have documented the tight linkage disequilibrium (LD) shared between specific haplotypes of complement C4 and HLA ancestral haplotypes, specifically DRB1*0301

(DR3) in LD with C4A deficiency. Like C4A deficiency, it has previously been reported that DR3 is also associated with risk of SLE. In the present study, we examined the association of HLA-DRB1 alleles and C4 gene copy numbers (GCNs) in Caucasian populations from central Ohio (pediatric SLE, n=55; adult SLE, n=256; controls, n=814) to determine the relative contributions of HLA-DRB1 and Complement C4 loci in determining susceptibility to SLE. Multivariate, backward, stepwise logistic regression analysis identified C4A deficiency and HLA-DRB1*15 (DR2) as factors that were significantly associated with SLE in both pediatric and adult populations. C4A deficiency remained a significant, independent risk factor for SLE in subgroup analyses of subjects stratified based on the presence or absence of DRB1*0301 alleles. Additional analyses revealed a significantly greater effect size for homozygous C4A deficiency (C4A=0)

[Odds ratio: 9.58] compared to a heterozygous C4A deficiency (C4A=1) [Odds ratio:

110

1.89] in adult SLE. We conclude that C4A deficiency and HLA-DRB1*15 are independently associated with SLE, while HLA-DRB1*0301 is not an independent risk factor for SLE and is present due to its high association (i.e., linkage disequilibrium) with

C4A deficiency.

3.1 Introduction

The etiology of SLE is believed to arise from a combination of predisposing genetic and epigenetic factors, sex hormones, cytokines, plus environmental triggers such as infections, stress, and UV exposure (Banchereau and Pascual 2006; Flesher et al. 2010;

Tsokos 2011). Homozygous (GCN=0) or heterozygous (GCN=1) genetic deficiencies of

Complement C4A have been reported to be associated with SLE disease risk (Howard et al. 1986; Yang et al. 2007; Pereira et al. 2016). In particular, the SLE odds ratio for individuals with homozygous C4A deficiency among female subjects was previously reported to be 5.3 (Yang et al. 2007). Furthermore, many lines of evidence suggest the malfunction or dysregulation of the complement system play important roles in the pathogenesis and disease progression of SLE; complete genetic deficiencies of early complement components (i.e., C1q/r/s, C2, or C4) are strongly associated with increased risk of developing SLE or a lupus-like disease (Lintner et al. 2016).

We and others have documented the tight LD shared between specific haplotypes of complement C4 and HLA ancestral haplotypes, specifically DRB1*0301 (DR3) in LD with C4A-deficient haplotypes (Dawkins et al. 1999; Yu and Whitacre 2004; Horton et al.

2008). The ancestral haplotype AH8.1 in the HLA that contains DR3 harbors one C4B

111 gene and zero C4A genes (i.e., C4A deficiency). Like C4A deficiency, it has previously been reported that DR3 has a greater than expected frequency in Caucasian patients with

SLE and is therefore associated with disease risk (Scherak et al. 1978; Tan and Arnett

1998; Graham et al. 2002). HLA associations (including DR2 and DR3) in several other autoimmune diseases besides SLE support the hypothesis of HLA involvement in general autoimmunity (Horton et al. 2004; Parkes et al. 2013). Generally, the effect size of SLE disease risk (or odds ratios) is between 2-3 for given HLA associations in SLE.

For the above reasons and given the complex architecture of the HLA genomic region, it is difficult to resolve whether C4A deficiency is responsible for the primary disease association and DRB1*0301 alleles are present due to LD, or vice versa, or if both are involved with susceptibility to SLE. It seems biologically reasonable to believe that HLA or complement C4 could both contribute to development of SLE. It may be hypothesized that HLA genes predispose to autoantibody formation, while C4A deficiency leads to either abnormal regulation of self-tolerance or impaired clearance of apoptotic cells and immune complexes and may therefore promote autoimmunity

(Prodeus et al. 1998; Paul et al. 2002; Chatterjee et al. 2013).

In efforts to confirm the previous findings of C4A deficiency associations with

SLE and to investigate the relationship of C4 haplotypes with HLA alleles in SLE, we undertook a case-control analysis of Caucasian adult and pediatric patients with SLE and race-matched healthy controls. In the present study, we aim to determine the relative contributions of Complement C4 loci and HLA-DRB1 alleles in determining susceptibility to SLE.

112

3.2 Materials and Methods

3.2.1 SLE and Control Populations

IRB approval was obtained from Nationwide Children’s Hospital and The Ohio

State University Wexner Medical Center, and consent for all enrolled study subjects was obtained according to protocol. Study subjects included 256 Caucasian adult patients with

SLE, 55 Caucasian pediatric patients with SLE, and 814 race-matched healthy controls subjects recruited in central Ohio. Each patient met the ACR diagnostic criteria for SLE

(Tan et al. 1982; Hochberg 1997; Petri et al. 2012). The mean age (±SD) at recruitment for adult SLE patients was 43.5 ± 11.7 years old, for pediatric SLE patients at recruitment was 15.2 ± 3, and for race-matched controls was 37.2 ± 11.6 years old. The female: male ratio in adult patients was ~13:1, in pediatric patients was ~8:1, and in controls was ~3:1.

Complete demographics of the study populations are displayed in Table 12.

3.2.2 HLA-DRB1 Typing

Genotyping of HLA-DRB1 alleles for SLE and all control samples were determined using genomic DNA for sequence-specific primer PCR (Hui and Bidwell

1993). Each DNA sample undergoes 20 different PCR reactions, containing 17 different oligonucleotide primer mixes (PM) designed to amplify a specific DRB1 allele or allele group, plus 3 primer mixes for determining “superspecificities” (PM 51, 52, 53). Thus, allele assignment for each sample is merely based on the presence or absence of amplified product visualized by agarose gel electrophoresis. A sample PCR gel with final

113 allelic calls is shown in Figure 16. HLA-DRB1 frequency was calculated by the number of allele-positive subjects divided by the total number of subjects.

3.2.3 Determination of Total C4, C4A, and C4B Genotypes

Previously, we described protocols for genotyping of complement C4 by Southern blot analyses (Chung et al. 2005). In some cases of limited DNA quantities or ambiguous results, quantitative real-time PCR experiments for GCN of total C4, C4A, and C4B were performed as previously described (Wu et al. 2007). In the latter case, all C4-CNV calls were validated by multiple C4 amplicons (i.e., each sample was assayed for C4 total,

C4A, and C4B, wherein: GCN of Total C4 = C4A + C4B).

3.2.4 Statistical Analysis

Statistical analyses were performed using Prism6 (GraphPad Software Inc., San

Diego, CA) and JMP Genomics 6.0 (SAS Institute Inc., Cary, NC) software. We used multivariate backward stepwise regression for final model building with entry at p≤0.25 based on bivariate analyses. Factors with the highest p-value (least significant) were eliminated sequentially until all remaining factors had p≤0.05. For model variables, C4A deficiency is defined as ≤1 copy of C4A; subjects were labeled as “positive” for HLA genotypes if they carried at least one copy of the given allele.

Descriptive statistics are displayed as mean ± standard deviation (SD) for normally distributed data, and simple comparisons were made using Student’s t-test for continuous data, or by chi-square analysis for categorical data. Odds Ratios (OR) with

114

95% confidence intervals (CI) are reported. For non-normally distributed data, median with interquartile range (IQR) is reported, and Mann-Whitney test was used for comparisons. For all analyses, p≤0.05 was considered to be significant.

3.2.5 Additional Declarations

This is a large, ongoing study in our laboratory involving many other institutions and collaborators. Contributions to this project from the author (Katherine Lintner) are as follows: Since 2011, subject recruitment and blood sample processing for healthy controls (n=186), adult SLE (n=27), and pediatric SLE (n=13). This includes SSP-PCR to determine HLA-DRB1 genotypes. In addition, the author genotyped HLA-DRB1 alleles from DNA samples from previously recruited healthy controls (n=100) and pediatric SLE

(n=13). All C4 genotyping southern blot analyses were performed by Bi Zhou.

Previously in the laboratory, contributions to this project are as follows: Adult

SLE patient recruitment and sample processing: Yan Yang, Erwin Chung, Yee Ling Wu,

Bi Zhou, Stephanie Savelli, Dan J Birmingham, Brad H. Rovin, Lee A. Hebert, Chack-

Yung Yu, Wael Jarjour. Pediatric SLE patient recruitment and sample processing: Robert

Rennebohm, Gloria C. Higgins, Charles Spencer, Hermine Brunner (Cincinnati

Children’s Hospital Medical Center), Stacy Ardoin, Sharon Bout-Tabaku, Anjali

Patwardhan, Rabheh Abdul Aziz, Yee Ling Wu, Yan Yang, Erwin Chung, Bi Zhou, and

Chack-Yung Yu. Ohio white control recruitment and sample processing: Karla Jones,

Jeanie Shaw, Yee Ling Wu, Yan, Yang, Bi Zhou, Lee A Hebert, and Chack-Yung Yu.

HLA-DRB1 genotyping: Emeli Lundstrom and Leonid Padyukov (Karolinska, Sweden),

115

Yee Ling Wu and Alaaedin Alhomosh. C4 genotyping: Yan Yang, Erwin Chung, Bi

Zhou and Yee Ling Wu. Statistical analyses were performed by Katherine Lintner and

Chack-Yung Yu with the guidance of our statistician, Dr. Haikady Nagaraja (The Ohio

State University).

3.3 Results

3.3.1 C4 Gene Copy Number Distribution and HLA-DRB1 Allele Frequencies

We have previously reported distributions of C4 gene copy number for this group of adult SLE, and it was found that low gene copy number of C4A and C4A deficiency were disease risk factors (Yang et al. 2007). We have substantially increased our race- matched control population by ~300 subjects and also increased our adult SLE patient cohort by 27 patients. Risk factors of C4A deficiency (C4A≤1) and low GCN of C4A remained strongly associated with SLE in our adult populations (Table 13). In addition, we have recruited a population of pediatric patients with SLE (n=55). C4A deficiency and low GCN of C4A were also significantly associated with pediatric SLE (Table 13). In pediatric SLE, 38.18% (21/55) were C4A-deficient compared to controls (19.2%), [OR:

2.61 (95% CI: 1.47-4.61), p= 0.0016]. GCN of C4A was significantly lower in pediatric

SLE [C4A: 1.65 ± 0.75] compared to controls [2.09 ± 0.77, p= 4.82x10-5].

The presence of DR3 was much more frequent in adult SLE (35.9%) and pediatric

SLE (38.2%) compared to our control population (24.7%). Because both DR3 and C4A deficiency have been previously reported as risk factors associated with SLE, we considered both variables to be “non-negotiable” and performed a simple, logistic

116 regression bivariate analysis. In our model with adult SLE vs. controls, DR3 became highly insignificant (p=0.938) in the presence of C4A deficiency (p= 0.0003). Similarly in pediatric SLE vs. controls, DR3 became highly insignificant (p= 0.996) in the presence of C4A deficiency (p= 0.021). In models for both adult and pediatric SLE, DR3 no longer independently contributed to SLE disease risk.

3.3.2 Statistical Analyses and Relative Roles of C4 and HLA-DRB1 Alleles in SLE

Figure 17 summarizes C4 genotypes and HLA allele frequencies in SLE and control groups. We used multivariate backward logistic stepwise regression analysis to determine whether the association of C4 and HLA-DRB1 alleles retained significance when controlling for known risk and protective factors (Figure 18). We determined initial entry of variables into the model of p≤0.25 based on bivariate analyses (Table 13a). In adult SLE, the multivariate analysis identified C4A deficiency [OR: 2.33 (95% CI: 1.70-

3.19), p=2.00x10-7] and HLA-DRB1*15 [OR: 1.40 (95% CI: 1.02-1.91), p=0.035] as factors that were significantly associated with disease risk. Similarly, in pediatric SLE, the multivariate analysis identified significant risk factors to be C4A deficiency [OR: 2.80

(95% CI: 1.55-4.96), p= 0.0008] and HLA-DRB1*15 [OR: 2.41 (95% CI: 1.37-4.22), p=

0.0026]. All other variables became insignificant (p≥0.05) and were rejected by the model.

117

3.3.2 C4A Deficiency in the Presence and Absence of HLA-DRB1*0301

To further evaluate the relative roles of C4A deficiency and DR3 on disease risk of SLE, we stratified subjects based on the presence or absence of DR3 and re-evaluated the association of C4A deficiency with SLE. In a subgroup analysis of subjects without the DR3 allele, C4A deficiency was more frequent in adult SLE compared to controls

(9.76% vs. 4.57%) and remained a significant risk factor [OR: 2.26 (95% CI: 1.19-4.28), p= 0.0165]. Likewise, in subjects with the DR3 allele, C4A deficiency was more frequent in adult SLE compared to controls (79.4% vs. 63.7%) and remained a significant risk factor [OR: 2.19 (95% CI: 1.23-3.92), p= 0.0061]. Identical subgroup analyses in our pediatric SLE study population revealed similar trends supporting the independent role of

C4A deficiency in SLE disease risk, but the number of subjects was too low to reach statistical significance. A reciprocal analysis to test the association of DR3 in C4A- deficient and C4A-proficient subgroups revealed that DR3 becomes a very insignificant factor (p~0.90). C4A deficiency was significantly associated with SLE populations with or without the presence of DR3 allele.

Additional analyses of C4A deficiency revealed a significantly greater effect size of a complete, homozygous C4A deficiency (C4A=0) compared to a partial, heterozygous

C4A deficiency (C4A=1) in SLE populations (Table 14). Adult SLE with homozygous

C4A deficiency (17/256, 6.64%) vs. controls (6/814, 0.74%) gave an odds ratio of 9.58

(95% CI: 3.93-26.79, p= 3.07x10-7) for SLE risk. Odds ratio calculated for heterozygous

C4A deficiency in adult SLE (72/239, 30.13%) vs. controls (150/808, 18.56%) had a lower effect size of 1.89 (95% CI: 1.36-2.62, p= 0.0002). Subjects with homozygous C4A

118 deficiency were excluded from the analysis for heterozygous C4A deficiency. Likewise, in pediatric SLE, the effect size of homozygous C4A deficiency [OR: 7.77 (95% CI:

1.60-30.34), p= 0.014; frequency in pediatric SLE: 3/55 (5.46%)] was significantly greater than heterozygous C4A deficiency [OR: 2.32 (95% CI: 1.25-4.18), p= 0.0083, frequency in pediatric SLE: 18/52 (34.62%)].

3.3.3 Strong Genetic Association Between DR3 and C4A deficiency in Study Populations

In Caucasians, the HLA ancestral haplotype AH8.1 includes DRB1*0301 allele, and this haplotype is also strongly associated with the absence of C4A genes, or “C4A deficiency” (Yu and Whitacre 2004; Horton et al. 2008). There was a strong, negative correlation with the number of DR3 alleles and the number of C4A alleles in both control and SLE populations. Although uncommon, DR3 homozygosity associated with C4A deficiency (C4A= 0 or 1) 100% in SLE. That is, SLE subjects homozygous for DR3 also had C4A deficiency [adult SLE: 100% (17/17); pediatric SLE: 100% (2/2)]. For control subjects homozygous for DR3, 83.3% (5/6) had C4A deficiency. None of our study subjects (control or SLE) had GCN of C4A=0 without the presence of at least one DR3 allele. In other words, we never saw a complete, homozygous deficiency of C4A in DR3- negative subjects. Conversely, we rarely saw C4A deficiency unless DR3 was also present. We also measured the strength of the linear relationship between number of DR3 alleles and number of C4A alleles (grouped into [C4A=0], [C4A=1], or [C4A=2-5]) and found very strong correlation for all study populations (controls: R2=0.439, p=3.60x10-

104; adult SLE: R2=0.588, p=8.64x10-51; pediatric SLE: R2=0.495, p=2.08x10-9).

119

3.3.4 Analysis of C4 and HLA-DRB1 Alleles in Adult SLE vs. Pediatric SLE Populations

We used the same method of logistic regression modeling described above to assess for any differences in predictor variables between pediatric SLE vs. adult SLE.

Although some variables had a greater effect size in pediatric vs. adult SLE, multivariate analyses did not reveal any statistical differences in risk or protective factors between pediatric and adult SLE subjects.

3.4 Discussion

A great challenge for studying complex diseases associated with the HLA region, including SLE, is to determine which gene(s) or polymorphic variants contribute to disease development under the background of strong allelic associations or linkage disequilibrium. This study replicates the findings of gene CNVs for complement C4A associated with SLE risk and dissects the relative roles of HLA-DRB1*0301 and C4A deficiency in SLE disease risk in subjects of European ancestry. We confirmed previous observations of increased frequency of C4A deficiency and low gene copy numbers of

C4A in SLE adult and pediatric populations similar to what was described in other studies

(Yang et al. 2007; Pereira et al. 2016), although the effect size appeared to be greater in pediatric SLE vs. adult populations. While the presence of DR3 was significantly higher in SLE vs. controls, DR3 did not hold as an independent predictor of SLE disease in multivariate analyses. In fact, our data shows that DR3 strongly correlates with C4A

120 deficiency and therefore we conclude that DR3 is more frequent in SLE simply due to the high LD shared between DR3 and C4A-deficient alleles.

Because C4A deficiency contributes to the susceptibility of SLE in Caucasians independently of DR3, stratification was used to test whether C4A deficiency was still associated with SLE with or without the presence of DR3. C4A deficiency remained a significant risk factor strongly associated with SLE in groups with DR3 and in groups without DR3. That is, C4A deficiency can be considered a risk factor for SLE completely independent of the presence of DR3.

We did not detect any significant differences in C4 genotypes or HLA alleles in pediatric vs. adult SLE populations. Although most often diagnosed in adulthood, SLE in pediatric populations is frequently more severe. There are no established genetic pathways known to explain the variability in severity of clinical features or age of disease onset, with the exception of complement deficiencies (homozygous C1q,C1r, C1s, and

C4 deficiencies) which predispose to SLE early in life (Ardoin and Schanberg 2012;

Lintner et al. 2016).

In our study, we demonstrated that DRB1*0301 allele only appeared to be associated with SLE because of its linkage with C4A deficiency. It remains to be determined whether DRB1*0301 (AH8.1 in Caucasians) in other autoimmune diseases contributes independently to disease risk or, as shown here, is C4A deficiency the real culprit?

121

122

Figure 16 Sample of HLA-DRB1 genotyping gel

Each sample has 20 PCR reactions with different primer sets, and after thermocycling, each reaction is run on agarose gel.

The upper band is a positive control; the lower band represents the HLA-DRB1 positive alleles in that individual sample.

Primer mixes 51, 52, and 53 provide superspecificity. For example, sample 1 has positive bands for PM 16 and PM17 only,

thus their HLA-DRB1 diploid genotype is *07/*16.

122

Figure 17 Comparison of C4 isotype deficiencies and HLA allele frequencies in SLE and controls

A summary of distributions in C4 and HLA-DRB1 genotypes in the MHC for Caucasian

SLE (adult and pediatric) and control subjects.

123

Figure 18 Overview of multivariate regression analysis

124

Table 12 Demographics of SLE study populations

Cases Cases Controls N=256 N=55 N=814

Adults SLE Pediatric SLE Healthy Controls N=229 N=27 N=48 N=7 N=814 Origin of subjects Central Ohio Ohio State Nationwide Cincinnati Central Ohio University, Children’s Children’s 125 Wael Jarjour Hospital Hospital Medical

Center, Hermine Brunner Age at enrollment (years), mean ± SD 44.3 ± 12.1 38.0 ± 10.8 15.0 ± 2.8 19.4 ± 6.3 37.2 ± 12.0 Male, n (%) 17 (7.4%) 1 (3.7%) 6 (12.5%) 0 214 (26.3%) Female, n (%) 212 (92.6%) 26 (96.3%) 42 (87.5%) 7 (100%) 600 (73.7%) BMI, mean ± SD 29.1 ± 7.5 26.5 ± 8.3 23.7 ± 5.5 21.3 ± 2.5 27.7 ± 6.6

125

Table 13 Comparison of C4 isotype deficiencies and HLA allele frequencies in SLE and controls

a. Distribution,

N (%)

C4A deficiency, C4B deficiency, HLA HLA HLA HLA HLA

GCN ≤1 GCN ≤1 DRB1*15 DRB1*0301 DRB1*07 DRB1*08 DRB1*09

126 (DR2) (DR3)

SLE, adult 89 (34.8%) 61 (23.8%) 83 (32.4%) 92 (35.9%) 60 (23.4%) 7 (2.73%) 8 (3.13%)

*5.2x10-7 *0.029 *0.11 *0.0006 *NS *0.046 *0.15 (N=256)

SLE, pediatric 21 (38.2%) 9 (16.4%) 25 (45.5%) 21 (38.2%) 7 (12.7%) 6 (10.9%) 0

(N=55) *0.0016 *0.17 *0.0052 *0.033 *0.0423 *0.15 *0.19

Controls 156 (19.2%) 251 (30.8%) 221 (27.2%) 201 (24.7%) 195 (24.0%) 46 (5.65%) 13 (1.60%)

(N=814)

Continued 126

Table 13 continued

b. C4 Mean

GCN ± SD

Total C4 C4A GCN C4B GCN C4-long C4-short

genes GCN GCN

SLE, adult 3.55 ± 0.77 1.78 ± 0.88 1.79 ± 0.56 2.57 ± 1.17 1.02 ± 0.84

†1.3x10-7 †4.7x10-8 †0.19 †2.9x10-6 †0.051 127 3.51 ± 0.69 1.65 ± 0.75 1.85 ± 0.62 2.43 ± 1.06 1.09 ± 0.81 SLE, pediatric

†0.0012 †4.8x10-5 †0.15 †0.0004 †0.10

Controls 3.83 ± 0.71 2.09 ± 0.77 1.73 ± 0.64 2.92 ± 0.99 0.91 ± 0.78

*Chi-square p-value, compared to controls; †T-test p-value, compared to controls; GCN, gene copy number; NS, p>0.25

127

Table 14 Effect sizes of homozygous vs. heterozygous deficiencies of C4A

Risk factor Control subjects Adult SLE OR P-value Pediatric SLE OR P-value

N (frequency) N (frequency) (95% CI) N (frequency) (95% CI) 128

Heterozygous C4A 1.89 2.32 150/808 (18.56%) 72/239 (30.13%) 0.0002 18/52 (34.62%) 0.0083 deficiency* (C4A=1) (1.36-2.62) (1.25-4.18)

Homozygous C4A 9.58 7.77 6/814 (0.74%) 17/256 (6.64%) 3.07x10-7 3/55 (5.46%) 0.014 deficiency (C4A=0) (3.93-26.79) (1.60-30.34)

*Subjects with homozygous C4A deficiency (C4A=0) were excluded from the analysis for heterozygous C4A deficiency.

128

Chapter 4: Regulation of Complement C4 Gene Expression

Abstract

A large majority of patients with JDM or SLE have low gene copy number of

C4A or of total C4 genes. Therefore, it is of interest to study cis-acting regulatory elements of the C4 promoter region and also identify any trans-acting factors such as cytokines, hormones, or other inflammatory mediators that may modulate C4 gene expression, thereby providing better insight into the control of C4 during inflammation or tissue injury and possibly creating therapeutics to increase C4 expression in individuals with partial C4 deficiencies. The C4 gene can differ in size dependent on the absence

(“short gene”) or presence (“long gene”) of a 6.4kb retroviral insertion in intron 9

[termed HERV-K(C4)]. The proviral DNA is flanked on each side by long terminal repeats (LTR), which contain sequences for poly(A+) signal, SV40 enhancer core, TATA box, and other motifs for transcriptional regulation. In a series of luciferase reporter assays, we used C4 promoter fragments and HERV-K(C4) LTR plasmid constructs in cultured cell lines from tissues known to express C4. We observed ~2-fold increase in reporter gene expression in HepG2 and TPC-1 cells lines when the 3’LTR was cloned upstream of the C4 promoter. That the 3’LTR appears to enhance C4 promoter activity has never been reported. Our results also confirmed solo promoter activity of 3’LTR in sense orientation, which could possibly initiate an antisense C4 transcript from exon 9 to 129 exon 1. In 8505C cultured cells, we demonstrated that IFN-gamma has a very strong effect on C4 promoter activity. In healthy subjects from three different races, there was a significant, negative correlation between C4-long gene copy number and C4 protein levels. While the 3’LTR appears to enhance C4 promoter activity and could therefore increase C4 gene expression, in a seemingly different mechanism, the 3’LTR exhibits solo promoter activity in an orientation that may initiate antisense C4 transcripts, possibly interfering with C4 translation and overall protein concentrations. Cell culture studies coupled with our C4 genotype-phenotype data in healthy subjects suggest that the C4 retroviral insert is interfering with C4 gene expression, either through antisense

“silencing” of C4 transcripts, or simply by slowing the rate of transcription, given that the retroviral insert is large, 6.4kb in size.

4.1 Introduction

Human C4 is one of the most polymorphic proteins in circulation. The primary site for C4 biosynthesis is in the liver. However, many studies have also demonstrated extra-hepatic sites of expression of C4, including peripheral blood mononuclear cells, kidney, thyroid, and adrenal cortex/medulla (Whaley 1980; Feucht et al. 1986; Witte et al. 1991) (Figure 19). The C4 protein is synthesized in the liver as a ~200 kDa hemolytically inactive precursor. C1s enzyme cleaves C4a, a 7 kDa fragment, from the amino terminus, resulting in functionally active C4 protein. The promoter region of the human C4 gene has been analyzed previously in reporter gene assays (Vaishnaw et al.

1998; Ulgiati et al. 2000). Results showed that the promoter region immediately upstream

130

(<200 bp) of the transcription start site (TSS) is associated with maximal, basal C4 expression in liver cell lines. C4 promoter lacks consensus TATA and CCAAT boxes, although an Sp1 site and E box motif were found to be essential for activity and probably takes the place of the TATA motif (Vaishnaw et al. 1998).

C4 consists of two forms, C4A and C4B, which are highly conserved in coding regions but can differ in size dependent on the absence (“short gene”) or presence (“long gene”) of a 6.4kb retroviral insertion in intron 9 (Dangel et al. 1994). It is estimated that

8% of the human genome is comprised of a variety of retroviral sequences that were integrated into our DNA more than 25 million years ago (Griffiths 2001). The proviral

DNA is often flanked on each side by long terminal repeats (LTR), which contain sequences for enhancers, poly(A+) signals, TATA box, and other motifs for transcriptional regulation (Buzdin et al. 2003). Although many of the retroviral genes themselves are rendered dysfunctional by deletions, mutations, and multiple stop codons, it is known that many LTRs in the human genome have retained either uni- or bi- directional promoter activity and/or may function as transcriptional regulators and affect the expression of nearby genes (Kidwell and Lisch 1997; Sverdlov 2000; Buzdin et al.

2006). The C4 retroviral insertion, termed HERV-K(C4), integrated in an orientation opposite to the C4 coding strand (Figure 20). If the HERV-K(C4) 3’LTR has promoter activity in its sense orientation, an antisense C4 transcript from exons 9 to exon 1 could, in theory, be expressed, thereby possibly interfering with C4 mRNA transcripts. If the 3’

LTR has promoter activity in its antisense orientation, this could initiate a HERV-K(C4) retroviral antisense transcript. The 3’LTR, but not the 5’LTR, has been shown previously

131 to exhibit promoter activity, but conflicting results were reported regarding the orientation of activity (Dangel et al. 1994; Mack et al. 2004).

A complete, or homozygous, genetic deficiency of both complement C4A and

C4B has been reported in 28 individuals (Nordin Fredrikson et al. 1991; Lokki et al.

1999; Rupert et al. 2002; Yang et al. 2004; Wu et al. 2009), and 22 (78.6%) of them were diagnosed with lupus-like disease, plus four others had renal disease including glomerulonephritis. Partial deficiency of C4, specifically of C4A, is also considered a major risk factor for autoimmune diseases, like SLE and JDM (Yang et al. 2007; Lintner et al. 2015; Chen et al. 2016). Therefore, it is of major interest to study the cis-acting regulatory elements essential for the expression of C4 and also identify any cytokines, hormones, or other inflammatory mediators that may modulate C4 gene expression. This would provide better insight into the control of C4 during inflammation or tissue injury and possibly pave the way for therapeutic potentials to increase C4 expression in individuals with partial C4 deficiencies. To date, the only known inducer of C4 is gamma interferon (IFN-gamma) (Miura et al. 1987; Tsukamoto et al. 1992; Mitchell et al. 1996).

4.2 Methods

4.2.1 Cells and Reagents

The human hepatoma cell line HepG2 was purchased from the American Type

Culture Collection (Rockville, MD). Thyroid cell lines TPC-1 and 8505C were kindly provided by Professor Sissy Jhiang (The Ohio State University). All adherent cell lines were cultured in DMEM (supplemented with 10% FBS, 2mM L-glutamine, 1mM sodium

132 pyruvate, 100 U/mL penicillin, and 100ug/mL streptomycin) and grown at 37°C, 5% C02.

The following recombinant human proteins were purchased from Cell Sciences (Canton,

MA): interferon gamma, tumor necrosis factor alpha, interleukin-10, and transforming growth factor beta.

4.2.2 Construction of Luciferase Plasmids

pGL4.10[luc2] vector was obtained from Promega and used for all plasmid constructs in luciferase assays (Figure 21). pGL4.10[luc2] is a basic vector with no promoter and contains a multiple cloning region to allow cloning of a promoter of choice.

The C4 promoter and HERV-K(C4) LTR fragments were amplified from C4 cosmids using primers listed in Table 15. The resulting PCR fragments were digested with the corresponding restriction enzyme, purified, and cloned into the cloning site in pGL4.10[luc2], upstream of the luc2 region. The entire C4 promoter region from 1kb upstream of the RP1/RP2 breakpoint, downstream to the C4 translation start site

(fragment size= 2.6kb) was initially cloned. Then, a series of 5’primers, progressively closer to the C4 translation start site (=ATG), together with a common 3’primer termed

XC4E1R1 were used to clone different sizes of the C4 promoter region into the BglII site in pGL4.10[luc2] (Figure 22). The 3’LTR from HERV-K(C4) retroviral insertion was cloned alone into pGL4.10[luc2] or together with C4 promoters utilizing NheI restriction site (Figure 23). A complete list of plasmid constructs that were generated are described in Table 16. After cloning, plasmids were transformed into 10-beta competent cells (New

England BioLabs) and plated in dilutions onto agar plates with appropriate antibiotics for

133 selection. Colonies were picked for plasmid purification and sent for Sanger sequencing to verify the correct insert and orientation of the insert.

4.2.3 Cell Transfection and Dual Luciferase Reporter Assays

Cells were seeded in appropriate media containing serum and antibiotics into tissue-treated flasks or dishes at suggested seeding density determined by surface area

(e.g., 3.0x105 cells into 12-well plate). Cells were incubated under normal growth conditions (37°C, 5% CO2) until ~60-80% confluent. On the day of transfection, cells were washed with 1x PBS, and new media was added to cells. Transfection with appropriate constructs was performed according to traditional protocol using Attractene transfection reagent (Qiagen). After 24h incubation, cells were either treated for another

24h with chosen cytokine/reagent or if untreated, directly analyzed using Dual Luciferase

Reporter Assay System according to the manufacturer’s passive lysis method protocol

(Promega). All cell culture wells were plated in duplicate, and luciferase samples were performed in triplicates. Promoter activity is expressed as relative luminescence, firefly activity divided by (normalized) renilla luciferase activity. Furthermore, all signals were normalized to the negative control (pGL4.10[luc2], empty backbone) for each respective assay.

4.2.4 C4 Genotype and Phenotype Determination

Previously, we described protocols for genotyping and phenotyping of complement C4 by Southern blot analyses and immunofixation, respectively (Sim and

134

Cross 1986; Chung et al. 2002; Chung et al. 2005). C4 protein concentrations were measured from plasma using radial immunodiffusion commercial kits (The Binding Site,

UK). Quantitative real-time PCR experiments for GCN of total C4, C4A, and C4B were performed as described (Wu et al. 2007). All C4-CNV calls were validated by independent technology, or multiple amplicons in qPCR, and matched genotype and phenotype interpretations.

4.2.5 Human Subject Recruitment

Healthy control subjects residing in Central Ohio were previously recruited by the laboratory with IRB approval from Nationwide Children’s Hospital (Columbus, OH).

4.2.6 Additional Declarations

The results presented in this chapter are unpublished and are part of an ongoing project in the laboratory. The author (Katherine Lintner) is responsible for designing and cloning all of the plasmid constructs and for all luciferase assays conducted. C4 genotype and phenotype data were already available and determined previously in the laboratory by the following people: C4 genotyping: Bi Zhou, Yee Ling Wu, and Katherine Lintner.

C4 phenotyping: Yee Ling Wu, Yan Yang, Nazreen Esack, and Katherine Lintner.

135

4.3 Results

4.3.1 Activity of the C4 Promoter Region

To analyze activity of the C4 promoter region in luciferase reporter assays and confirm previous findings (Vaishnaw et al. 1998; Ulgiati et al. 2000) of promoter regions required for maximal activity of basal expression of human C4, we generated a series of

C4 promoter deletion constructs, ranging from -300bp (relative to translation start site) to

-2600bp (Figure 22). We performed such assays in a variety of cell lines, including

HepG2 (liver), TPC-1 (thyroid), and 8505C (thyroid).

We showed that the -300bp fragment exhibited the strongest activity of all the fragments tested. In the thyroid cell line 8505C, the promoter strength of

0.3kb>1.6kb>2.6kb (Figure 24). The 300bp fragment efficiency was ~2.9x stronger than

1.6kb and 7.6x stronger than 2.6kb. Similarly, in HepG2 liver cell line, the 0.3kb fragment exhibited the greatest signal in our luciferase reporter assay (Figure 25).

We compared the activity of our 2.6kb promoter fragments cloned from two different C4 cosmids. Using cos3A3, we cloned the fragment from the region homologous to locus II in humans. Using KEM1 cosmid, we cloned the 2.6kb fragment, but the fragment from this cosmid resembles that of the promoter for a single, C4B gene in humans, also known as “mono-short” haplotype. Our analysis showed that the mono-S promoter was about 2x stronger than that of the identical sized locus I promoter (Figure

24).

136

4.3.2 Analysis of HERV-K(C4) LTR Activity

We designed luciferase plasmid constructs with LTR alone, to test promoter activity, and with LTR upstream of 1.6kb C4 fragment, to test if LTR has enhancer activity on the C4 promoter.

Many HERV LTR regions in the human genome have retained either uni- or bi- directional promoter activity (Domansky et al. 2000). To test the HERV-K(C4) LTRs and confirm previous findings (Dangel et al. 1994; Mack et al. 2004), we transfected TPC-1,

8505C, and HepG2 cell lines with 3’LTR luciferase plasmid constructs (cloned in both orientations). The 3’LTR exhibited promoter activity in the sense direction (relative to the HERV-K insert) in TPC-1 and HepG2 cells (Figure 25), but not in 8505C (Figure 24).

To clarify, the 3’LTR in its own sense direction is in opposite orientation to the C4 gene and could, in theory, drive expression of an antisense C4 transcript, from exon 9 to exon

1 (Figure 20). We did not detect any promoter activity for 3’LTR in the antisense direction.

In HepG2 cells, the 3’LTR enhanced activity of the C4 promoter, regardless of which direction the 3’LTR was cloned (sense or antisense); the effect was <2-fold

(Figure 25). Likewise, in TPC-1 cells, the 3’LTR, regardless of orientation, enhanced the

C4 promoter activity >2-fold. This observation has never been reported. In 8505C cells, however, the 3’LTR did not have an enhancer effect on the C4 promoter fragment.

137

4.3.3 In Vitro Treatment of C4-promoter and HERV-K(C4)-LTR Transfected Cells

We treated transfected cells with IFN-gamma (IFN-y), the only known inducer of

C4 expression. In 8505C cell line, we demonstrated very clearly that IFN-y has a very strong effect on C4 promoter fragments, as well as 3’LTR+C4 promoter plasmids, taken at 24 hour time points (Figure 26). In addition, we showed a temporal effect in TPC-1 cell line; we treated with IFN-y and harvested cells at 1hour, 4 hours, and 24 hours. We observed a time-dependent increase on C4 promoter activity (Figure 27). In any of our cell lines tested, IFN-y did not have an effect on the solo promoter activity of 3’LTR alone.

Concentrations of plasma C4 are increased during acute phase response to inflammation or tissue injury (Johansson et al. 1972; Moshage 1997). Therefore we examined the effect of several mediators in vitro. We did not detect any changes upon C4 expression for the following reagents at 24 hour time points: IL-10 [10 ng/mL], TGF-beta

[15 ng/mL], or IFN-alpha [2000 U/mL].

4.3.4 Variations in Human C4 Protein Expression In Vivo

Healthy control plasma samples were measured for C4 levels in three different races: Asian (n= 153), African American (n= 221), and European American (n= 234). To assess the role of the HERV-K(C4) retroviral insertion on C4 protein expression, we performed a bivariate linear regression analysis of number of C4-long genes in individuals vs. C4 plasma protein concentration (Figure 28). For all three races, we only included individuals with four total C4 genes, so as to exclude any protein variations due

138 to total gene copy number differences. For all three races, there was a significant, negative correlation between C4-long genes and C4 protein levels; the C4 protein levels in individuals with long GCN=4

Caucasians. For those with four total C4 genes, the average C4 plasma concentration for groups stratified by long genes was: C4L=4 (protein [mg/L]= 285.50 ± 70.59), C4L=3

(327.46 ± 106.65), C4L=2 (367.61 ± 89.19), C4L=1 (441.32 ± 122.99). For our analysis,

R2= 0.112, p<0.0001. A linear fit produced the equation C4 [mg/L]= 457.16 –

43.14*#C4-long. In other words, for every long gene present, the total C4 plasma protein level was decreased by 43.14 mg/L. Average C4 plasma levels in healthy individuals typically range from 120-700 mg/L.

4.4 Discussion

We found that the C4 promoter region just 300bp upstream of the translation start site was associated with maximal reporter gene expression. These findings are similar to and supported by published data (Vaishnaw et al. 1998). Our data demonstrated that although not as strong, there was still significant activity with other C4 promoter fragments in luciferase assays. These results were similar in all three cell lines used for assays (HepG2, 8505C, TPC-1). We also showed, as previously described, that IFN-y has a very strong effect on C4 promoter activity; experiments in liver and thyroid cell lines demonstrated this effect. Variations in concentrations and time point measurements of

139 other cytokines and reagents that we tested need to be further explored before drawing any solid conclusions on their effects, if any, on C4 expression.

Our results from luciferase assays using clones of HERV-K(C4)-3’LTR demonstrated promoter activity in its sense orientation, which could possibly initiate a C4 transcript in antisense orientation. The presence of transcriptional regulatory motifs such as SV40-type enhancer core sequence and TATA box correlates with 3’LTR promoter activity in this observed orientation (Dangel et al. 1994; Chu et al. 1995). We also speculated that the LTRs of HERV-K(C4) may influence activity of the nearby C4 promoter. We observed ~2-fold increase in reporter gene expression in HepG2 and TPC-

1 cells lines when the 3’LTR was cloned upstream of the C4 promoter. However, we did not observe this effect in 8505C cells. In 8505C, reporter gene expression was similar for

C4 promoter construct compared to C4 promoter with 3’LTR. Further work is required to elucidate whether tissue or cell line specificity of 3’LTR enhancement on C4 promoter is due to expression of specific regulatory proteins which act on LTR promoter.

We are now presented with opposing observations concerning the 3’LTR and how it may affect C4 expression: The 3’LTR appears to enhance C4 promoter activity and could therefore increase C4 expression, while in a seemingly different mechanism, the

3’LTR exhibits promoter activity in an orientation that may initiate antisense C4 transcripts, possibly interfering with C4 translation and overall protein concentrations of

C4. To assess the role of HERV-K(C4)-3’LTR on C4 protein levels, we performed various analyses using C4 genotype and phenotype data from healthy controls from three different races. In all three races, there was a significant, negative correlation between the

140 number of long genes and the amount of C4 protein detected in plasma samples. In other words, individuals with more long genes had lower C4 plasma levels compared to race- matched individuals with fewer long genes. According to this data, we hypothesize that the retroviral insert (in the long gene) is interfering with C4 expression, either through antisense “silencing” of C4 transcripts by blocking the translation of C4 mRNA transcripts, or simply by slowing the rate of transcription, given that the retroviral insert is large, 6.4kb in size.

141

Figure 19 Detection of C4 expression in various tissues

A Northern blot analysis shows that human complement C4 transcripts are expressed in

multiple tissues. RNA blots were hybridized with human C4d specific probe (upper

panel) and with a human RD probe corresponding to a subunit of the negative

transcription elongation factor as a positive control (lower panel). Figure reprinted from

reference (Blanchong et al. 2001).

142

Figure 20 Dichotomous size variation of human C4 gene

In the long C4 gene, the HERV-K(C4) is arranged in opposite orientation with respect to

C4. The retroviral insertion is positioned in intron 9 of C4, is ~6.4kb in size, and has

hallmark structures such as long terminal repeats (LTRs), gag, pol, and env genes.

143

Figure 21 pGL4.10[luc2] vector map

The pGL4.10[luc2] vector (Promega) is a basic, promoterless vector with a multiple

cloning region to allow cloning of any promoter of choice. The vector encodes a

luciferase reporter gene luc2 (Photinus pyralis), allowing for high expression and reduced

background.

144

Figure 22 Promoter sequence of human C4 and strategy for making promoter deletion constructs.

Size variations in C4 promoter constructs at the 5’ end are indicated. All fragments had a common 3’end using XC4E1R1 primer and were cloned into BglII site of pGL4.10[luc2] upstream of the luc2 reporter gene.

145

Figure 23 Example of construct cloned for luciferase assays

A 1.6kb C4 promoter region was cloned into BglII site in the multiple cloning region.

Then, the vector was re-opened with NheI, and the 3’LTR fragment of HERV-K(C4) was

cloned in. Constructs were sequenced after cloning to choose correct orientations of

inserts. In this example, both C4 promoter and 3’LTR are oriented in their sense direction.

146

8505C cell line 600

500

400

300

147

200 Relative Relative Luminescence

100

0 0.3kb, C4 1.6kb, C4 2.6kb, C4 2.6kb, C4 3'LTR (S) 3'LTR (AS) promoter (S) promoter (S) promoter, promoter, locus I (S) mono-short (S)

Figure 24 C4 reporter gene activity in 8505C cultured cells

S, sense orientation ; AS, antisense orientation

(continued) 147

A bar graph representing expression levels of cloned luciferase constructs in 8505C, untreated cultured cells. Expression levels are reported as means + standard error in triplicates. Relative luminescence is indicated as firefly signal divided by renilla signal, to control for transfection efficiency, and all values are normalized to the transfected pGL4.10[luc2] promoter-less plasmid for each assay.

148

HepG2 cell line 400

350

300

250

200

150 Relative Relative Luminescence 100

50

0 0.3kb, C4 1.6kb, C4 3'LTR(S)+1.6kb 3'LTR(AS)+1.6kb 3'LTR (S) 3'LTR (AS) promoter (S) promoter (S) C4 (S) C4 (S)

Figure 25 C4 reporter gene activity in HepG2 cultured cells

S, sense orientation ; AS, antisense orientation

A bar graph representing expression levels of cloned luciferase constructs in HepG2,

untreated cultured cells. Expression levels are reported as means + standard error in

triplicates. Relative luminescence is indicated as firefly signal divided by renilla signal, to

control for transfection efficiency, and all values are normalized to the transfected

pGL4.10[luc2] promoter-less plasmid for each assay.

149

8505C cell line

3000.00

2500.00

2000.00 no treatment 1500.00 IFN-y [250 U/mL] 1000.00

500.00 Relative Luminescence Relative 0.00

Figure 26 C4 reporter gene activity upon IFN-gamma treatment

S, sense orientation ; AS, antisense orientation

A bar graph representing expression levels of cloned luciferase constructs in 8505C

cultured cells with and without IFN-gamma treatment. Expression levels are reported in

triplicates, and standard error bars are shown. Relative luminescence is indicated as firefly

signal divided by renilla signal.

150

1.6kb, C4 promoter, TPC-1 cell line 35000

30000

25000

20000 no treatment 15000 IFN-y [5000U/mL]

10000 Relative Relative Luminescence 5000

0 1 hour 4 hours 24 hours

Figure 27 Temporal expression of C4 promoter activity in response to IFN-gamma treatment

Relative expression of 1.6kb C4 promoter fragment in TPC-1 cultured cells upon IFN-

gamma treatment. Cells were harvested at 1 hour, 4 hours, and 24 hours. Relative

luminescence indicated is relative (normalized) to untreated cells at respective time points.

151

152

Figure 28 Variations of human C4 plasma protein and GCN of C4-long genes (C4L) in different races

C4 plasma protein levels were measured from healthy control samples and plotted against the individual’s GCN of C4 long

(C4L) genes. For all three races, we only included individuals with four total C4 genes, so as to exclude any protein variations

due to total gene copy number differences. Linear regression coefficient of determination (R2) and corresponding p-values are

shown.

152

Table 15 Primers used for C4 promoter and HERV-K(C4) 3’LTR cloning

ID Primer (5’-->3’) Forward/ Size template Reverse (bp) XRP1E4F1 GAC AGA TCT GAC CAA ATG ACA CAG ACC TTT GG F 2600 cos3A3 XRP1E4F1 GAC AGA TCT GAC CAA ATG ACA CAG ACC TTT GG F 2600 KEM1 XRP1RP2F2 GAC AGA TCT CAA GTA CTT TGT TAA AGG TAT CC F 1600 cos3A3 XRP1RP2F3 GAC AGA TCT GCT CAT CAT TGC TCA GCT CCT CAG F 1100 cos3A3 XRP1RP2F4 GAC AGA TCT GTA CCT CTG CCT GTG ATA TTT TCT G F 700 cos3A3 153 XRP1RP2F5 GAC AGA TCT GGT GCT CAG GCA CTG GAA TGA GAG F 300 cos3A3 XC4E1R1 (common reverse) GAC AGA TCT GGC TGG AGG ATC CAA GAG AGG TTA G R X-K(C4)-3'LTR-R GAC GCT AGC GCA TGT TGG GAA AAG GAC TTG TGG R 600 cos3A3 XC4I9-3'LTR-F GAC GCT AGC GTG GAC ATG TGT TGT TCA ATG CC F cos3A3 X-K(C4)-5'LTR-F GAC GCT AGC CTT GAG CGT TTC CTC ACC AGA TTC F 600 cos3A3 XC4I9-5'LTR-R GAC GCT AGC CAC AAG ACA GTG AGC TCC CAG R cos3A3

All primers are tagged on 5’ end with GAC (random three nucleotides) plus BglII restriction site (AGATCT) or NheI

restriction site (GCTAGC) to assist with cloning into pGL4.10[luc2] vector

153

Table 16 Plasmid constructs designed for luciferase reporter assays

Description ID name Size of insert C4 locus Direction of Fwd. primer Rev. primer Cosmid

(kb) insert

C4 promoter region A2-1 2.6 Locus I Sense XRP1E4F1 XC4E1R1 cos3A3

C4 promoter region AA9 2.6 Mono-S Sense XRP1E4F1 XC4E1R1 KEM1 154

C4 promoter region E2 1.6 Locus I/II Antisense XRP1RP2F2 XC4E1R1 cos3A3

C4 promoter region E3 1.6 Locus I/II Sense XRP1RP2F2 XC4E1R1 cos3A3

C4 promoter region F3 1.1 Locus I/II Antisense XRP1RP2F3 XC4E1R1 cos3A3

C4 promoter region G5 0.7 Locus I/II Antisense XRP1RP2F4 XC4E1R1 cos3A3

C4 promoter region H1 0.3 Locus I/II Antisense XRP1RP2F5 XC4E1R1 cos3A3

C4 promoter region H2 0.3 Locus I/II Sense XRP1RP2F5 XC4E1R1 cos3A3

3’LTR 1-8 0.6 Sense XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

3’LTR 2-6 0.6 Antisense XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

3’LTR cloned into 3-2 2.2 LTR XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

154 Continued

Table 16 continued

E3, upstream of C4 (sense), C4

promoter (sense)

3’LTR cloned into 3-3 2.2 LTR XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

E3, upstream of C4 (antisense), 155

promoter C4 (sense)

3’LTR cloned into 1-3 2.2 LTR XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

E2, upstream of C4 (antisense),

promoter C4

(antisense)

3’LTR cloned into 2-4 2.2 LTR XC4I9-3'LTR-F X-K(C4)-3'LTR-R cos3A3

E2, upstream of C4 (sense), C4

promoter (antisense)

155

Chapter 5: Conclusion

It is well-established that homozygous deficiencies of C4 and other early complement components (C1q, C1r, C1s) are strongly associated with immune-mediated clinical features in humans. Here, we have established and determined definitively the relative role(s) of the tightly linked alleles C4A deficiency and HLA-DRB1*0301 in JDM and elaborated on previous findings in SLE. From low GCN of C4A to plasma protein levels and CBCAPS phenotype data, we report novel findings that C4A plays a significant role in JDM.

Complement C4 and HLA in JDM. In this dissertation, we observed C4A deficiency association in patients with JDM, although DRB1*0301 in conjunction with the C4A deficiency permitted the highest risk of JDM. That is, C4A deficiency and

DRB1*0301 appear to be additive risk factors, both of which contribute to disease risk. In a study population of 95 Caucasian JDM cases and 500 race-matched healthy controls, homozygous and heterozygous deficiency of complement C4A had a frequency of 40.0% in JDM and 18.2% in controls (odds ratio=3.0). The co-existence of HLA-DR3 and C4A deficiency confers higher risk than either individual risk factors, and such concurrence in a diploid genome was present in 36.8% of JDM and 15.4% of controls with an odds ratio of 3.2. The independent roles of DRB1 variants and C4A deficiency in JDM were further

156 validated by multiple logistic regression analyses. Moreover, among DR3-positive subjects, lower mean GCN of C4A or higher prevalence of C4A deficiency persisted in

JDM versus controls. An interpretation to this phenomenon is that DR3-positivity contributes to a permissive background and C4A deficiency significantly elevated the vulnerability to an autoimmune disease including JDM. While HLA-DRB1 and complement C4 gene products are each engaged in specific immunologic functions such as antigen presentation to effector T-cells, and complement-mediated cytolysis and immunoclearance, they are both involved in the recognition of self and non-self, and are key players for the process of archiving memory and tolerance in the immune system.

Although this was the first study of GCN of C4 in JDM and race-matched replication cohorts are needed to validate this association, our preliminary results are quite convincing.

Complement C4 and HLA in SLE. In this work, we verified the findings of low

GCN of complement C4A associated with SLE risk and dissected the relative roles of

HLA-DRB1*0301 and C4A deficiency in SLE disease risk in subjects of European ancestry in adult and pediatric populations. Because of the strong LD shared between

C4A-deficient haplotypes and DRB1*0301, identifying the primary disease allele(s) can be difficult. While the presence of DR3 was significantly higher in SLE vs. controls, DR3 did not hold as an independent predictor of SLE disease in multivariate statistical analyses. In fact, our data shows that DR3 strongly correlates with C4A deficiency and therefore we conclude that DR3 association with SLE is simply due to the LD shared

157 between DR3 and C4A-deficient haplotypes. In fact, of the individuals who did not carry the DR3 allele, ~90% of them had a “normal” GCN of C4A (≥2 copies), supporting the idea that DR3 and C4A deficiency nearly always exist together.

C4A deficiency and Autoimmune Disease Mechanistic Hypotheses. It can be reasonably hypothesized why a deficiency of C4A protein is associated with increased susceptibility of autoimmune diseases. Functionally, the role of complement includes the identification, opsonization, and proper disposal of apoptotic cells and immune complexes formed regularly between antibodies and foreign or self-antigens. An inability to efficiently clear apoptotic cells could render them a source of autoantigens and thereby drive autoantibody production. Activated C4A protein reacts effectively with amino groups on immune complexes or protein molecules, whereas C4B, because of the nucleotide differences embedded at the genetic level, reacts more effectively with carbohydrate-rich groups. While C4A-deficient patients usually always have a normal gene copy number of C4B, the functional disparities between the C4A and C4B proteins provide reasonable explanation as to why deficiency of C4A, but not deficiency of C4B, confers risk for SLE and JDM.

To the Clinic and Beyond…In our study populations, approximately 40% of JDM patients and 35% of SLE patients were either heterozygous or homozygous C4A- deficient. Given this common observation of C4A deficiency noted more frequently in patients afflicted with aforementioned autoimmune diseases, a logical and practical

158 forward step would be the establishment of C4-based therapy for these patients. We showed that the HERV-K(C4) 3’LTR imposes an enhancer effect on the C4 promoter and also exhibits solo promoter activity, possibly resulting in antisense C4 transcripts from exon 9 to exon 1. It is of interest to study this 3’LTR and trans-acting factors that affect its activity, thereby creating possibilities to harness 3’LTR activity for use in altering C4 gene expression in vivo in patients.

C4-knockout mice in permissive backgrounds express autoimmune (SLE) phenotypes (Chen et al. 2000; Paul et al. 2002) and could serve as excellent animal models needed to test the roles of C4A and C4B in disease pathogenicity. Successful rescue of disease phenotypes in C4-knockout mice with human C4A and/or C4B transgenes would provide proof-of-principle and justifications for C4-based clinical trials for better therapies of SLE and other autoimmune diseases. Alternatively, using gene editing methods like CRISPR/Cas9 (Sternberg et al. 2015) or homologous recombination

TALEN technology (Baker 2012), one could modify the existing mouse C4 gene to human-specific C4A or C4B sequences. Human and mouse C4 genes share 76% sequence identity, and the mouse C4 is a hybrid of human C4A and C4B isotypes (Yu et al. 2000).

Using genome editing to “humanize” the mouse C4 gene would allow for greater, in- depth investigation on the specific roles of C4A and C4B on the pathogenesis of SLE.

Animal studies must logically precede any clinical trials that would implement a trial of treatment for human patients with SLE or JDM.

Several observations and conclusions from this dissertation require thoughtful and investigative extensions of this work. Logically, increasing the sample size of the study

159 population and/or analyzing replication cohorts will confirm and validate our genetic findings of C4A deficiency associated with autoimmune diseases. Additionally, this dissertation focuses solely on subjects of European American descent (Caucasian).

Insights into C4 and HLA genetic risk factors in autoimmune diseases in other races will allow for a greater understanding of disease pathways and identify any common or race- specific associated genetic factors. SLE is notably a complex disease, with many genetic loci implicated in risk and pathogenesis. There are multiple paths and etiologies that can lead to SLE. As noted in this dissertation, a portion of patients with SLE or JDM did not have the risk factor of C4A deficiency; the underlying genetic risk factor(s) in this group of patients needs to be addressed. The advancements in GWAS, DNA sequencing technologies, complex statistical modeling algorithms, and more in-depth research will hopefully begin to address the questions of what multiple, interacting genetic variants associate with SLE disease risk and pathogenicity, along with or in addition to C4A deficiency. The work in this dissertation and future studies should align with one, overarching goal: to advance the understanding of autoimmune disease pathogenesis and impact the field of human health by strengthening the prevention, treatment, maintenance, or cure of JDM, SLE, or other human immune-mediated diseases.

160

References

Abe, K., Y. Endo, N. Nakazawa, K. Kanno, M. Okubo, T. Hoshino and T. Fujita (2009). "Unique phenotypes of C1s deficiency and abnormality caused by two compound heterozygosities in a Japanese family." J Immunol 182(3): 1681-1688.

Agnello, V. (1986). "Lupus diseases associated with hereditary and acquired deficiencies of complement." Springer Semin Immunopathol 9(2-3): 161-178.

Amano, M. T., V. P. Ferriani, M. P. Florido, E. S. Reis, M. I. Delcolli, A. E. Azzolini, A. I. Assis-Pandochi, A. G. Sjoholm, C. S. Farah, J. C. Jensenius and L. Isaac (2008). "Genetic analysis of complement C1s deficiency associated with systemic lupus erythematosus highlights alternative splicing of normal C1s gene." Mol Immunol 45(6): 1693-1702.

Amiel, J. (1967). Study of the Leukocyte Phenotypes in Hodgkin's Disease. Histocompatibility Testing 1967: Report of a Conference and Workshop E. Curtoni. Copenhagen, Denmark: 79-81.

Ardoin, S. P. and L. E. Schanberg (2012). "Paediatric rheumatic disease: lessons from SLE: children are not little adults." Nat Rev Rheumatol 8(8): 444-445.

Atkinson, J. and C. Yu (2015). The Complement System in Systemic Lupus Erythematosus. Systemic Lupus Erythematosus. G. C. Tsokos. New York, NY, Elsevier: In Press.

Baechler, E. C., H. Bilgic and A. M. Reed (2011). "Type I interferon pathway in adult and juvenile dermatomyositis." Arthritis Res Ther 13(6): 249.

Baker, M. (2012). "Gene-editing nucleases." Nat Methods 9(1): 23-26.

Banchereau, J. and V. Pascual (2006). "Type I interferon in systemic lupus erythematosus and other autoimmune diseases." Immunity 25(3): 383-392.

Berkel, A., M. Loos, O. Sanal, G. Mauff, Y. Gungen, U. Ors, F. Ersoy and O. Yegin (1979). "Clinical and immunochemical studies in a case of selective complete C1q deficiency." Clin Exp Immunol 38(52).

Berkel, A., O. T. Sanal, R. and M. Loos (1977). "A case of selective C1q deficiency." Turkish Journal of Pediatrics 19(101).

161

Berkel, A. I., E. Birben, C. Oner, R. Oner, M. Loos and F. Petry (2000). "Molecular, genetic and epidemiologic studies on selective complete C1q deficiency in Turkey." Immunobiology 201(3-4): 347-355.

Berkel, A. I., F. Petry, O. Sanal, K. Tinaztepe, F. Ersoy, A. Bakkaloglu and M. Loos (1997). "Development of systemic lupus erythematosus in a patient with selective complete C1q deficiency." Eur J Pediatr 156(2): 113-115.

Birmingham, D. J., F. Irshaid, H. N. Nagaraja, X. Zou, B. P. Tsao, H. Wu, C. Y. Yu, L. A. Hebert and B. H. Rovin (2010). "The complex nature of serum C3 and C4 as biomarkers of lupus renal flare." Lupus 19(11): 1272-1280.

Blanchong, C. A., E. K. Chung, K. L. Rupert, Y. Yang, Z. Yang, B. Zhou, J. M. Moulds and C. Y. Yu (2001). "Genetic, structural and functional diversities of human complement components C4A and C4B and their mouse homologues, Slp and C4." Int Immunopharmacol 1(3): 365-392.

Blanchong, C. A., B. Zhou, K. L. Rupert, E. K. Chung, K. N. Jones, J. F. Sotos, W. B. Zipf, R. M. Rennebohm and C. Yung Yu (2000). "Deficiencies of human complement component C4A and C4B and heterozygosity in length variants of RP-C4-CYP21-TNX (RCCX) modules in caucasians. The load of RCCX genetic diversity on major histocompatibility complex-associated disease." J Exp Med 191(12): 2183-2196.

Boackle, S. A. (2013). "Advances in lupus genetics." Curr Opin Rheumatol 25(5): 561- 568.

Bohan, A. and J. B. Peter (1975). "Polymyositis and dermatomyositis (first of two parts)." N Engl J Med 292(7): 344-347.

Bohan, A. and J. B. Peter (1975). "Polymyositis and dermatomyositis (second of two parts)." N Engl J Med 292(8): 403-407.

Boteva, L., D. L. Morris, J. Cortes-Hernandez, J. Martin, T. J. Vyse and M. M. Fernando (2012). "Genetically determined partial complement C4 deficiency states are not independent risk factors for SLE in UK and Spanish populations." Am J Hum Genet 90(3): 445-456.

Bournazos, S., J. M. Woof, S. P. Hart and I. Dransfield (2009). "Functional and clinical consequences of Fc receptor polymorphic and copy number variants." Clin Exp Immunol 157(2): 244-254.

Bowness, P., K. A. Davies, P. J. Norsworthy, P. Athanassiou, J. Taylor-Wiedeman, L. K. Borysiewicz, P. A. Meyer and M. J. Walport (1994). "Hereditary C1q deficiency and systemic lupus erythematosus." QJM 87(8): 455-464. 162

Burgin, S., J. H. Stone, A. S. Shenoy-Bhangle and D. McGuone (2014). "Case records of the Massachusetts General Hospital. Case 18-2014. A 32-Year-old man with a rash, myalgia, and weakness." N Engl J Med 370(24): 2327-2337.

Buzdin, A., E. Kovalskaya-Alexandrova, E. Gogvadze and E. Sverdlov (2006). "At least 50% of human-specific HERV-K (HML-2) long terminal repeats serve in vivo as active promoters for host nonrepetitive DNA transcription." J Virol 80(21): 10752-10762.

Buzdin, A., S. Ustyugova, K. Khodosevich, I. Mamedov, Y. Lebedev, G. Hunsmann and E. Sverdlov (2003). "Human-specific subfamilies of HERV-K (HML-2) long terminal repeats: three master genes were active simultaneously during branching of hominoid lineages." Genomics 81(2): 149-156.

Calano, S. J., P. A. Shih, C. C. Liu, A. H. Kao, J. S. Navratil, S. Manzi and J. M. Ahearn (2006). "Cell-bound complement activation products (CB-CAPs) as a source of lupus biomarkers." Adv Exp Med Biol 586: 381-390.

Carroll, M. C. (2000). "The role of complement in B cell activation and tolerance." Adv Immunol 74: 61-88.

Carroll, M. C. (2004). "A protective role for innate immunity in systemic lupus erythematosus." Nat Rev Immunol 4(10): 825-831.

Carroll, M. C., R. D. Campbell, D. R. Bentley and R. R. Porter (1984). "A molecular map of the human major histocompatibility complex class III region linking complement genes C4, C2 and factor B." Nature 307(5948): 237-241.

Carroll, M. C., D. M. Fathallah, L. Bergamaschini, E. M. Alicot and D. E. Isenman (1990). "Substitution of a single amino acid (aspartic acid for histidine) converts the functional activity of human complement C4B to C4A." Proc Natl Acad Sci U S A 87(17): 6868-6872.

Chakravarty, E. F., T. M. Bush, S. Manzi, A. E. Clarke and M. M. Ward (2007). "Prevalence of adult systemic lupus erythematosus in California and Pennsylvania in 2000: estimates obtained using hospitalization data." Arthritis Rheum 56(6): 2092-2094.

Chapuis, R. M., G. Hauptmann, E. Grosshans and H. Isliker (1982). "Structural and functional studies in C1q deficiency." J Immunol 129(4): 1509-1512.

Chatterjee, P., A. F. Agyemang, M. B. Alimzhanov, S. Degn, S. A. Tsiftsoglou, E. Alicot, S. A. Jones, M. Ma and M. C. Carroll (2013). "Complement C4 maintains peripheral B-cell tolerance in a myeloid cell dependent manner." Eur J Immunol 43(9): 2441-2450. 163

Chen, J. Y., Y. L. Wu, M. Y. Mok, Y. J. Wu, K. E. Lintner, C. M. Wang, E. K. Chung, Y. Yang, B. Zhou, H. Wang, D. J. Yu, A. Alhomosh, K. Jones, C. H. Spencer, H. N. Nagaraja, Y. Lung Lau, C. S. Lau and C. Y. Yu (2016). "Effects of Complement C4 Gene Copy-Number Variations, Size Dichotomy and C4A-Deficiency on Genetic Risk and Clinical Presentation of East-Asian SLE." Arthritis Rheumatol 10.1002/art.39589.

Chen, Z., S. B. Koralov and G. Kelsoe (2000). "Complement C4 inhibits systemic autoimmunity through a mechanism independent of complement receptors CR1 and CR2." J Exp Med 192(9): 1339-1352.

Chew, C. H., K. H. Chua, L. H. Lian, S. M. Puah and S. Y. Tan (2008). "PCR-RFLP genotyping of C1q mutations and single nucleotide polymorphisms in Malaysian patients with systemic lupus erythematosus." Hum Biol 80(1): 83-93.

Chu, X., C. Rittner and P. M. Schneider (1995). "Length polymorphism of the human complement component C4 gene is due to an ancient retroviral integration." Exp Clin Immunogenet 12(2): 74-81.

Chung, E. K., Y. L. Wu, Y. Yang, B. Zhou and C. Y. Yu (2005). "Human complement components C4A and C4B genetic diversities: complex genotypes and phenotypes." Curr Protoc Immunol Chapter 13: Unit 13 18.

Chung, E. K., Y. Yang, R. M. Rennebohm, M. L. Lokki, G. C. Higgins, K. N. Jones, B. Zhou, C. A. Blanchong and C. Y. Yu (2002). "Genetic sophistication of human complement components C4A and C4B and RP-C4-CYP21-TNX (RCCX) modules in the major histocompatibility complex." Am J Hum Genet 71(4): 823- 837.

Chung, E. K., Y. Yang, K. L. Rupert, K. N. Jones, R. M. Rennebohm, C. A. Blanchong and C. Y. Yu (2002). "Determining the one, two, three, or four long and short loci of human complement C4 in a major histocompatibility complex haplotype encoding C4A or C4B proteins." Am J Hum Genet 71(4): 810-822.

Daha, M. R. and L. A. van Es (1980). "Relative resistance of the F-42-stabilized classical pathway C3 convertase to inactivation by C4-binding protein." J Immunol 125(5): 2051-2054.

Dalakas, M. C. (2015). "Inflammatory Muscle Diseases." N Engl J Med 373(4): 393-394.

Dalakas, M. C. and R. Hohlfeld (2003). "Polymyositis and dermatomyositis." Lancet 362(9388): 971-982.

Dangel, A. W., A. R. Mendoza, B. J. Baker, C. M. Daniel, M. C. Carroll, L. C. Wu and C. Y. Yu (1994). "The dichotomous size variation of human complement C4 164

genes is mediated by a novel family of endogenous retroviruses, which also establishes species-specific genomic patterns among Old World primates." Immunogenetics 40(6): 425-436.

Davis, W. R., J. E. Halls, A. C. Offiah, C. Pilkington, C. M. Owens and K. Rosendahl (2011). "Assessment of active inflammation in juvenile dermatomyositis: a novel magnetic resonance imaging-based scoring system." Rheumatology (Oxford) 50(12): 2237-2244.

Dawkins, R., C. Leelayuwat, S. Gaudieri, G. Tay, J. Hui, S. Cattley, P. Martinez and J. Kulski (1999). "Genomics of the major histocompatibility complex: haplotypes, duplication, retroviruses and disease." Immunol Rev 167: 275-304.

Day, N. K., H. Geiger, R. Stroud, M. DeBracco, B. Mancaido, D. Windhorst and R. A. Good (1972). "C1r deficiency: an inborn error associated with cutaneous and renal disease." J Clin Invest 51(5): 1102-1108.

Degn, S. E., L. Jensen, A. G. Hansen, D. Duman, M. Tekin, J. C. Jensenius and S. Thiel (2012). "Mannan-binding lectin-associated serine protease (MASP)-1 is crucial for lectin pathway activation in human serum, whereas neither MASP-1 nor MASP-3 is required for alternative pathway function." J Immunol 189(8): 3957- 3969.

Dodds, A. W., X. D. Ren, A. C. Willis and S. K. Law (1996). "The reaction mechanism of the internal thioester in the human complement component C4." Nature 379(6561): 177-179.

Domansky, A. N., E. P. Kopantzev, E. V. Snezhkov, Y. B. Lebedev, C. Leib-Mosch and E. D. Sverdlov (2000). "Solitary HERV-K LTRs possess bi-directional promoter activity and contain a negative regulatory element in the U5 region." FEBS Lett 472(2-3): 191-195.

Dragon-Durey, M. A., C. Blanc, L. T. Roumenina, N. Poulain, S. Ngo, P. Bordereau and V. Fremeaux-Bacchi (2014). "Anti-factor H autoantibodies assay." Methods Mol Biol 1100: 249-256.

Dragon-Durey, M. A., P. Quartier, V. Fremeaux-Bacchi, J. Blouin, C. de Barace, A. M. Prieur, L. Weiss and W. H. Fridman (2001). "Molecular basis of a selective C1s deficiency associated with early onset multiple autoimmune diseases." J Immunol 166(12): 7612-7616.

Duncan, A. R. and G. Winter (1988). "The binding site for C1q on IgG." Nature 332(6166): 738-740.

165

Elliott, J. A., Jr. and D. R. Mathieson (1953). "Complement in disseminated (systemic) lupus erythematosus." AMA Arch Derm Syphilol 68(2): 119-128.

Emslie-Smith, A. M. and A. G. Engel (1990). "Microvascular changes in early and advanced dermatomyositis: a quantitative study." Ann Neurol 27(4): 343-356.

Endo, Y., K. Kanno, M. Takahashi, K. Yamaguchi, Y. Kohno and T. Fujita (1999). "Molecular basis of human complement C1s deficiency." J Immunol 162(4): 2180-2183.

Falchi, M., J. S. El-Sayed Moustafa, P. Takousis, F. Pesce, A. Bonnefond, J. C. Andersson-Assarsson, P. H. Sudmant, R. Dorajoo, M. N. Al-Shafai, L. Bottolo, E. Ozdemir, H. C. So, R. W. Davies, A. Patrice, R. Dent, M. Mangino, P. G. Hysi, A. Dechaume, M. Huyvaert, J. Skinner, M. Pigeyre, R. Caiazzo, V. Raverdy, E. Vaillant, S. Field, B. Balkau, M. Marre, S. Visvikis-Siest, J. Weill, O. Poulain- Godefroy, P. Jacobson, L. Sjostrom, C. J. Hammond, P. Deloukas, P. C. Sham, R. McPherson, J. Lee, E. S. Tai, R. Sladek, L. M. Carlsson, A. Walley, E. E. Eichler, F. Pattou, T. D. Spector and P. Froguel (2014). "Low copy number of the salivary amylase gene predisposes to obesity." Nat Genet 46(5): 492-497.

Fearon, D. T., K. F. Austen and S. Ruddy (1974). "Properdin factor D. II. Activation to D by properdin." J Exp Med 140(2): 426-436.

Feldman, B. M., L. G. Rider, A. M. Reed and L. M. Pachman (2008). "Juvenile dermatomyositis and other idiopathic inflammatory myopathies of childhood." Lancet 371(9631): 2201-2212.

Fernando, M. M., C. R. Stevens, E. C. Walsh, P. L. De Jager, P. Goyette, R. M. Plenge, T. J. Vyse and J. D. Rioux (2008). "Defining the role of the MHC in autoimmunity: a review and pooled analysis." PLoS Genet 4(4): e1000024.

Feucht, H. E., C. M. Jung, M. J. Gokel, G. Riethmuller, J. Zwirner, A. Brase, E. Held and G. J. O'Neill (1986). "Detection of both isotypes of complement C4, C4A and C4B, in normal human glomeruli." Kidney Int 30(6): 932-936.

Flesher, D. L., X. Sun, T. W. Behrens, R. R. Graham and L. A. Criswell (2010). "Recent advances in the genetics of systemic lupus erythematosus." Expert Rev Clin Immunol 6(3): 461-479.

Fredrikson, G. N., B. Gullstrand, P. M. Schneider, K. Witzel-Schlomp, A. G. Sjoholm, C. A. Alper, Z. Awdeh and L. Truedsson (1998). "Characterization of non-expressed C4 genes in a case of complete C4 deficiency: identification of a novel point mutation leading to a premature stop codon." Hum Immunol 59(11): 713-719.

166

Fremeaux-Bacchi, V., L. Weiss, C. Demouchy, J. Blouin and M. D. Kazatchkine (1996). "Autoantibodies to the collagen-like region of C1q are strongly associated with classical pathway-mediated hypocomplementemia in systemic lupus erythematosus." Lupus 5(3): 216-220.

Friedman, J. M., L. M. Pachman, M. L. Maryjowski, R. M. Radvany, W. E. Crowe, V. Hanson, J. E. Levinson and C. H. Spencer (1983). "Immunogenetic studies of juvenile dermatomyositis: HLA-DR antigen frequencies." Arthritis Rheum 26(2): 214-216.

Fries, J. F. and R. C. Siegel (1973). "Testing the 'preliminary criteria for classification of SLE'." Ann Rheum Dis 32(2): 171-177.

Gadjeva, M. G., M. M. Rouseva, A. S. Zlatarova, K. B. Reid, U. Kishore and M. S. Kojouharova (2008). "Interaction of human C1q with IgG and IgM: revisited." Biochemistry 47(49): 13093-13102.

Girirajan, S., C. D. Campbell and E. E. Eichler (2011). "Human copy number variation and complex genetic disease." Annu Rev Genet 45: 203-226.

Gough, S. C. and M. J. Simmonds (2007). "The HLA Region and Autoimmune Disease: Associations and Mechanisms of Action." Curr Genomics 8(7): 453-465.

Graham, R. R., G. Hom, W. Ortmann and T. W. Behrens (2009). "Review of recent genome-wide association scans in lupus." J Intern Med 265(6): 680-688.

Graham, R. R., W. A. Ortmann, C. D. Langefeld, D. Jawaheer, S. A. Selby, P. R. Rodine, E. C. Baechler, K. E. Rohlf, K. B. Shark, K. J. Espe, L. E. Green, R. P. Nair, P. E. Stuart, J. T. Elder, R. A. King, K. L. Moser, P. M. Gaffney, T. L. Bugawan, H. A. Erlich, S. S. Rich, P. K. Gregersen and T. W. Behrens (2002). "Visualizing human leukocyte antigen class II risk haplotypes in human systemic lupus erythematosus." Am J Hum Genet 71(3): 543-553.

Greenberg, S. A. (2010). "Dermatomyositis and type 1 interferons." Curr Rheumatol Rep 12(3): 198-203.

Griffiths, D. J. (2001). "Endogenous retroviruses in the human genome sequence." Genome Biol 2(6): REVIEWS1017.

Gulez, N., F. Genel, F. Atlihan, B. Gullstrand, L. Skattum, L. Schejbel, P. Garred and L. Truedsson (2010). "Homozygosity for a novel mutation in the C1q C chain gene in a Turkish family with hereditary C1q deficiency." J Investig Allergol Clin Immunol 20(3): 255-258.

167

He, S. and Y. L. Lin (1998). "In vitro stimulation of C1s proteolytic activities by C1s- presenting autoantibodies from patients with systemic lupus erythematosus." J Immunol 160(9): 4641-4647.

Hengstman, G. J., B. G. van Engelen, W. T. Vree Egberts and W. J. van Venrooij (2001). "Myositis-specific autoantibodies: overview and recent developments." Curr Opin Rheumatol 13(6): 476-482.

Higuchi, Y., J. Shimizu, M. Hatanaka, E. Kitano, H. Kitamura, H. Takada, M. Ishimura, T. Hara, O. Ohara, K. Asagoe and T. Kubo (2013). "The identification of a novel splicing mutation in C1qB in a Japanese family with C1q deficiency: a case report." Pediatr Rheumatol Online J 11(1): 41.

Hochberg, M. C. (1997). "Updating the American College of Rheumatology revised criteria for the classification of systemic lupus erythematosus." Arthritis Rheum 40(9): 1725.

Hollox, E. J., U. Huffmeier, P. L. Zeeuwen, R. Palla, J. Lascorz, D. Rodijk-Olthuis, P. C. van de Kerkhof, H. Traupe, G. de Jongh, M. den Heijer, A. Reis, J. A. Armour and J. Schalkwijk (2008). "Psoriasis is associated with increased beta-defensin genomic copy number." Nat Genet 40(1): 23-25.

Hoppenreijs, E. P., P. J. van Dijken, P. J. Kabel and J. M. Th Draaisma (2004). "Hereditary C1q deficiency and secondary Sjogren's syndrome." Ann Rheum Dis 63(11): 1524-1525.

Horton, R., R. Gibson, P. Coggill, M. Miretti, R. J. Allcock, J. Almeida, S. Forbes, J. G. Gilbert, K. Halls, J. L. Harrow, E. Hart, K. Howe, D. K. Jackson, S. Palmer, A. N. Roberts, S. Sims, C. A. Stewart, J. A. Traherne, S. Trevanion, L. Wilming, J. Rogers, P. J. de Jong, J. F. Elliott, S. Sawcer, J. A. Todd, J. Trowsdale and S. Beck (2008). "Variation analysis and gene annotation of eight MHC haplotypes: the MHC Haplotype Project." Immunogenetics 60(1): 1-18.

Horton, R., L. Wilming, V. Rand, R. C. Lovering, E. A. Bruford, V. K. Khodiyar, M. J. Lush, S. Povey, C. C. Talbot, Jr., M. W. Wright, H. M. Wain, J. Trowsdale, A. Ziegler and S. Beck (2004). "Gene map of the extended human MHC." Nat Rev Genet 5(12): 889-899.

Hou, S., J. Qi, D. Liao, J. Fang, L. Chen, A. Kijlstra and P. Yang (2014). "High C4 gene copy numbers protects against Vogt-Koyanagi-Harada syndrome in Chinese Han." Br J Ophthalmol 98(12): 1733-1737.

Howard, P. F., M. C. Hochberg, W. B. Bias, F. C. Arnett, Jr. and R. H. McLean (1986). "Relationship between C4 null genes, HLA-D region antigens, and genetic

168

susceptibility to systemic lupus erythematosus in Caucasian and black Americans." Am J Med 81(2): 187-193.

Hughes, A. E., N. Orr, H. Esfandiary, M. Diaz-Torres, T. Goodship and U. Chakravarthy (2006). "A common CFH haplotype, with deletion of CFHR1 and CFHR3, is associated with lower risk of age-related macular degeneration." Nat Genet 38(10): 1173-1177.

Hui, K. and J. Bidwell (1993). Handbook of HLA Typing Techniques. Boca Raton, FL, CRC Press.

Iafrate, A. J., L. Feuk, M. N. Rivera, M. L. Listewnik, P. K. Donahoe, Y. Qi, S. W. Scherer and C. Lee (2004). "Detection of large-scale variation in the human genome." Nat Genet 36(9): 949-951.

Inoue, N., T. Saito, R. Masuda, Y. Suzuki, M. Ohtomi and H. Sakiyama (1998). "Selective complement C1s deficiency caused by homozygous four-base deletion in the C1s gene." Hum Genet 103(4): 415-418.

Isenman, D. E. and J. R. Young (1984). "The molecular basis for the difference in immune hemolysis activity of the Chido and Rodgers isotypes of human complement component C4." J Immunol 132(6): 3019-3027.

Jackson, J., R. B. Sim, A. Whelan and C. Feighery (1986). "An IgG autoantibody which inactivates C1-inhibitor." Nature 323(6090): 722-724.

Jensenius, J. C. (2005). "The mannan-binding lectin (MBL) pathway of complement activation: biochemistry, biology and clinical implications." Adv Exp Med Biol 564: 21-22.

Johansson, B. G., C. O. Kindmark, E. Y. Trell and F. A. Wollheim (1972). "Sequential changes of plasma proteins after myocardial infarction." Scand J Clin Lab Invest Suppl 124: 117-126.

Johnson, C. A., P. Densen, R. K. Hurford, Jr., H. R. Colten and R. A. Wetsel (1992). "Type I human complement C2 deficiency. A 28-base pair gene deletion causes skipping of exon 6 during RNA splicing." J Biol Chem 267(13): 9347-9353.

Jozsi, M., S. Reuter, P. Nozal, M. Lopez-Trascasa, P. Sanchez-Corral, Z. Prohaszka and B. Uzonyi (2014). "Autoantibodies to complement components in C3 glomerulopathy and atypical hemolytic uremic syndrome." Immunol Lett 160(2): 163-171.

Kidmose, R. T., N. S. Laursen, J. Dobo, T. R. Kjaer, S. Sirotkina, L. Yatime, L. Sottrup- Jensen, S. Thiel, P. Gal and G. R. Andersen (2012). "Structural basis for

169

activation of the complement system by component C4 cleavage." Proc Natl Acad Sci U S A 109(38): 15425-15430.

Kidwell, M. G. and D. Lisch (1997). "Transposable elements as sources of variation in animals and plants." Proc Natl Acad Sci U S A 94(15): 7704-7711.

Kim, J. H., S. H. Jung, J. S. Bae, H. S. Lee, S. H. Yim, S. Y. Park, S. Y. Bang, H. J. Hu, H. D. Shin, S. C. Bae and Y. J. Chung (2013). "Deletion variants of RABGAP1L, 10q21.3, and C4 are associated with the risk of systemic lupus erythematosus in Korean women." Arthritis Rheum 65(4): 1055-1063.

Kingery, S. E., Y. L. Wu, B. Zhou, R. P. Hoffman and C. Y. Yu (2012). "Gene CNVs and protein levels of complement C4A and C4B as novel biomarkers for partial disease remissions in new-onset type 1 diabetes patients." Pediatr Diabetes 13(5): 408-418.

Kissel, J. T., R. K. Halterman, K. W. Rammohan and J. R. Mendell (1991). "The relationship of complement-mediated microvasculopathy to the histologic features and clinical duration of disease in dermatomyositis." Arch Neurol 48(1): 26-30.

Kissel, J. T., J. R. Mendell and K. W. Rammohan (1986). "Microvascular deposition of complement membrane attack complex in dermatomyositis." N Engl J Med 314(6): 329-334.

Kjellman, M., A. B. Laurell, B. Low and A. G. Sjoholm (1982). "Homozygous deficiency of C4 in a child with a lupus erythematosus syndrome." Clin Genet 22(6): 331- 339.

Klein, J. (1982). Immunology: The Science of Self-Nonself Discrimination. New York City, NY, John Wiley & Sons Inc.

Kusumoto, H., S. Hirosawa, J. P. Salier, F. S. Hagen and K. Kurachi (1988). "Human genes for complement components C1r and C1s in a close tail-to-tail arrangement." Proc Natl Acad Sci U S A 85(19): 7307-7311.

Law, S. K., A. W. Dodds and R. R. Porter (1984). "A comparison of the properties of two classes, C4A and C4B, of the human complement component C4." EMBO J 3(8): 1819-1823.

Leddy, J. P., R. C. Griggs, M. R. Klemperer and M. M. Frank (1975). "Hereditary complement (C2) deficiency with dermatomyositis." Am J Med 58(1): 83-91.

Lenz, T. L., A. J. Deutsch, B. Han, X. Hu, Y. Okada, S. Eyre, M. Knapp, A. Zhernakova, T. W. Huizinga, G. Abecasis, J. Becker, G. E. Boeckxstaens, W. M. Chen, A. Franke, D. D. Gladman, I. Gockel, J. Gutierrez-Achury, J. Martin, R. P. Nair, M.

170

M. Nothen, S. Onengut-Gumuscu, P. Rahman, S. Rantapaa-Dahlqvist, P. E. Stuart, L. C. Tsoi, D. A. van Heel, J. Worthington, M. M. Wouters, L. Klareskog, J. T. Elder, P. K. Gregersen, J. Schumacher, S. S. Rich, C. Wijmenga, S. R. Sunyaev, P. I. de Bakker and S. Raychaudhuri (2015). "Widespread non-additive and interaction effects within HLA loci modulate the risk of autoimmune diseases." Nat Genet 47(9): 1085-1090.

Lessard, C. J., H. Li, I. Adrianto, J. A. Ice, A. Rasmussen, K. M. Grundahl, J. A. Kelly, M. G. Dozmorov, C. Miceli-Richard, S. Bowman, S. Lester, P. Eriksson, M. L. Eloranta, J. G. Brun, L. G. Goransson, E. Harboe, J. M. Guthridge, K. M. Kaufman, M. Kvarnstrom, H. Jazebi, D. S. Cunninghame Graham, M. E. Grandits, A. N. Nazmul-Hossain, K. Patel, A. J. Adler, J. S. Maier-Moore, A. D. Farris, M. T. Brennan, J. A. Lessard, J. Chodosh, R. Gopalakrishnan, K. S. Hefner, G. D. Houston, A. J. Huang, P. J. Hughes, D. M. Lewis, L. Radfar, M. D. Rohrer, D. U. Stone, J. D. Wren, T. J. Vyse, P. M. Gaffney, J. A. James, R. Omdal, M. Wahren-Herlenius, G. G. Illei, T. Witte, R. Jonsson, M. Rischmueller, L. Ronnblom, G. Nordmark, W. F. Ng, U. K. P. S. s. S. Registry, X. Mariette, J. M. Anaya, N. L. Rhodus, B. M. Segal, R. H. Scofield, C. G. Montgomery, J. B. Harley and K. L. Sivils (2013). "Variants at multiple loci implicated in both innate and adaptive immune responses are associated with Sjogren's syndrome." Nat Genet 45(11): 1284-1292.

Lewis, E. J., C. B. Carpenter and P. H. Schur (1971). "Serum complement component levels in human glomerulonephritis." Ann Intern Med 75(4): 555-560.

Lintner, K. E., A. Patwardhan, L. G. Rider, R. Abdul-Aziz, Y. L. Wu, E. Lundstrom, L. Padyukov, B. Zhou, A. Alhomosh, D. Newsom, P. White, K. B. Jones, T. P. O'Hanlon, F. W. Miller, C. H. Spencer and C. Y. Yu (2015). "Gene copy-number variations (CNVs) of complement C4 and C4A deficiency in genetic risk and pathogenesis of juvenile dermatomyositis." Ann Rheum Dis 10.1136/annrheumdis-2015-207762.

Lintner, K. E., Y. L. Wu, Y. Yang, C. H. Spencer, G. Hauptmann, L. A. Hebert, J. P. Atkinson and C. Y. Yu (2016). "Early Components of the Complement Classical Activation Pathway in Human Systemic Autoimmune Diseases." Front Immunol 7: 36.

Lipsker, D. and G. Hauptmann (2010). "Cutaneous manifestations of complement deficiencies." Lupus 19(9): 1096-1106.

Liu, C. C., S. Manzi, A. H. Kao, J. S. Navratil and J. M. Ahearn (2010). "Cell-bound complement biomarkers for systemic lupus erythematosus: from benchtop to bedside." Rheum Dis Clin North Am 36(1): 161-172, x.

171

Liu, C. C., S. Manzi, A. H. Kao, J. S. Navratil, M. J. Ruffing and J. M. Ahearn (2005). "Reticulocytes bearing C4d as biomarkers of disease activity for systemic lupus erythematosus." Arthritis Rheum 52(10): 3087-3099.

Liu, Y. H., L. Wan, C. T. Chang, W. L. Liao, W. C. Chen, Y. Tsai, C. H. Tsai and F. J. Tsai (2011). "Association between copy number variation of complement component C4 and Graves' disease." J Biomed Sci 18: 71.

Lokki, M. L., A. Circolo, P. Ahokas, K. L. Rupert, C. Y. Yu and H. R. Colten (1999). "Deficiency of human complement protein C4 due to identical frameshift mutations in the C4A and C4B genes." J Immunol 162(6): 3687-3693.

Lv, Y., S. He, Z. Zhang, Y. Li, D. Hu, K. Zhu, H. Cheng, F. Zhou, G. Chen, X. Zheng, P. Li, Y. Ren, X. Yin, Y. Cui, L. Sun, S. Yang and X. Zhang (2012). "Confirmation of C4 gene copy number variation and the association with systemic lupus erythematosus in Chinese Han population." Rheumatol Int 32(10): 3047-3053.

Mack, M., K. Bender and P. M. Schneider (2004). "Detection of retroviral antisense transcripts and promoter activity of the HERV-K(C4) insertion in the MHC class III region." Immunogenetics 56(5): 321-332.

Mahler, M., R. A. van Schaarenburg and L. A. Trouw (2013). "Anti-C1q autoantibodies, novel tests, and clinical consequences." Front Immunol 4: 117.

Mamyrova, G., T. P. O'Hanlon, J. B. Monroe, D. M. Carrick, J. D. Malley, S. Adams, A. M. Reed, E. A. Shamim, L. James-Newton, F. W. Miller and L. G. Rider (2006). "Immunogenetic risk and protective factors for juvenile dermatomyositis in Caucasians." Arthritis Rheum 54(12): 3979-3987.

Mamyrova, G., T. P. O'Hanlon, J. B. Monroe, D. M. Carrick, J. D. Malley, S. Adams, A. M. Reed, E. A. Shamim, L. James-Newton, F. W. Miller, L. G. Rider and G. Childhood Myositis Heterogeneity Collaborative Study (2006). "Immunogenetic risk and protective factors for juvenile dermatomyositis in Caucasians." Arthritis Rheum 54(12): 3979-3987.

Manderson, A. P., M. C. Pickering, M. Botto, M. J. Walport and C. R. Parish (2001). "Continual low-level activation of the classical complement pathway." J Exp Med 194(6): 747-756.

Manolio, T. A., F. S. Collins, N. J. Cox, D. B. Goldstein, L. A. Hindorff, D. J. Hunter, M. I. McCarthy, E. M. Ramos, L. R. Cardon, A. Chakravarti, J. H. Cho, A. E. Guttmacher, A. Kong, L. Kruglyak, E. Mardis, C. N. Rotimi, M. Slatkin, D. Valle, A. S. Whittemore, M. Boehnke, A. G. Clark, E. E. Eichler, G. Gibson, J. L. Haines, T. F. Mackay, S. A. McCarroll and P. M. Visscher (2009). "Finding the missing heritability of complex diseases." Nature 461(7265): 747-753. 172

Manzi, S., J. S. Navratil, M. J. Ruffing, C. C. Liu, N. Danchenko, S. E. Nilson, S. Krishnaswami, D. E. King, A. H. Kao and J. M. Ahearn (2004). "Measurement of erythrocyte C4d and complement receptor 1 in systemic lupus erythematosus." Arthritis Rheum 50(11): 3596-3604.

Marquart, H. V., L. Schejbel, A. Sjoholm, U. Martensson, S. Nielsen, A. Koch, A. Svejgaard and P. Garred (2007). "C1q deficiency in an Inuit family: identification of a new class of C1q disease-causing mutations." Clin Immunol 124(1): 33-40.

Mascaro, J. M., Jr., G. Hausmann, C. Herrero, J. M. Grau, M. C. Cid, J. Palou and J. M. Mascaro (1995). "Membrane attack complex deposits in cutaneous lesions of dermatomyositis." Arch Dermatol 131(12): 1386-1392.

Mauff, G., B. Luther, P. M. Schneider, C. Rittner, B. Stradmann-Bellinghausen, R. Dawkins and J. M. Moulds (1998). "Reference typing report for complement component C4." Exp Clin Immunogenet 15(4): 249-260.

McAdam, R. A., D. Goundis and K. B. Reid (1988). "A homozygous point mutation results in a stop codon in the C1q B-chain of a C1q-deficient individual." Immunogenetics 27(4): 259-264.

McClellan, J. and M. C. King (2010). "Genetic heterogeneity in human disease." Cell 141(2): 210-217.

Mehta, P., P. J. Norsworthy, A. E. Hall, S. J. Kelly, M. J. Walport, M. Botto and M. C. Pickering (2010). "SLE with C1q deficiency treated with fresh frozen plasma: a 10-year experience." Rheumatology (Oxford) 49(4): 823-824.

Mendez, E. P., R. Lipton, R. Ramsey-Goldman, P. Roettcher, S. Bowyer, A. Dyer and L. M. Pachman (2003). "US incidence of juvenile dermatomyositis, 1995-1998: results from the National Institute of Arthritis and Musculoskeletal and Skin Diseases Registry." Arthritis Rheum 49(3): 300-305.

Meszaros, T., G. Fust, H. Farkas, L. Jakab, G. Temesszentandrasi, G. Nagy, E. Kiss, P. Gergely, M. Zeher, Z. Griger, L. Czirjak, R. Hobor, A. Haris, K. Polner and L. Varga (2010). "C1-inhibitor autoantibodies in SLE." Lupus 19(5): 634-638.

Mikuska, A., Z. Stefanovic, V. Miletic, V.-A. Oxelius and A. Sjoholm (1983). "Systemic lupus erythematosus-like disease in two siblings with complete C1q deficiency." Periodicum Biologorum 85: 271-273.

Miller, E. C., N. M. Chase, P. Densen, M. K. Hintermeyer, J. T. Casper and J. P. Atkinson (2012). "Autoantibody stabilization of the classical pathway C3 convertase leading to C3 deficiency and Neisserial sepsis: C4 nephritic factor revisited." Clin Immunol 145(3): 241-250. 173

Miller, F. W., R. G. Cooper, J. Vencovsky, L. G. Rider, K. Danko, L. R. Wedderburn, I. E. Lundberg, L. M. Pachman, A. M. Reed, S. R. Ytterberg, L. Padyukov, A. Selva-O'Callaghan, T. R. Radstake, D. A. Isenberg, H. Chinoy, W. E. Ollier, T. P. O'Hanlon, B. Peng, A. Lee, J. A. Lamb, W. Chen, C. I. Amos and P. K. Gregersen (2013). "Genome-wide association study of dermatomyositis reveals genetic overlap with other autoimmune disorders." Arthritis Rheum 65(12): 3239-3247.

Miller, F. W., R. G. Cooper, J. Vencovsky, L. G. Rider, K. Danko, L. R. Wedderburn, I. E. Lundberg, L. M. Pachman, A. M. Reed, S. R. Ytterberg, L. Padyukov, A. Selva-O'Callaghan, T. R. Radstake, D. A. Isenberg, H. Chinoy, W. E. Ollier, T. P. O'Hanlon, B. Peng, A. Lee, J. A. Lamb, W. Chen, C. I. Amos, P. K. Gregersen and C. Myositis Genetics (2013). "Genome-wide association study of dermatomyositis reveals genetic overlap with other autoimmune disorders." Arthritis Rheum 65(12): 3239-3247.

Mitchell, T. J., M. Naughton, P. Norsworthy, K. A. Davies, M. J. Walport and B. J. Morley (1996). "IFN-gamma up-regulates expression of the complement components C3 and C4 by stabilization of mRNA." J Immunol 156(11): 4429- 4434.

Miura, N., H. L. Prentice, P. M. Schneider and D. H. Perlmutter (1987). "Synthesis and regulation of the two human complement C4 genes in stable transfected mouse fibroblasts." J Biol Chem 262(15): 7298-7305.

Moncada, B., N. K. Day, R. A. Good and D. B. Windhorst (1972). "Lupus- erythematosus-like syndrome with a familial defect of complement." N Engl J Med 286(13): 689-693.

Mortensen, S., R. T. Kidmose, S. V. Petersen, A. Szilagyi, Z. Prohaszka and G. R. Andersen (2015). "Structural Basis for the Function of Complement Component C4 within the Classical and Lectin Pathways of Complement." J Immunol 194(11): 5488-5496.

Moser, K. L., J. A. Kelly, C. J. Lessard and J. B. Harley (2009). "Recent insights into the genetic basis of systemic lupus erythematosus." Genes Immun 10(5): 373-379.

Moshage, H. (1997). "Cytokines and the hepatic acute phase response." J Pathol 181(3): 257-266.

Moulds, J. M., C. Rolih, R. Goldstein, K. F. Whittington, N. B. Warner, I. N. Targoff, M. Reichlin and F. C. Arnett (1990). "C4 null genes in American whites and blacks with myositis." J Rheumatol 17(3): 331-334.

Nakamura, S., M. Yoshinari, Y. Saku, K. Hirakawa, C. Miishima, K. Murai, K. Tokiyama and M. Fujishima (1991). "Acquired C1 inhibitor deficiency associated 174

with systemic lupus erythematosus affecting the central nervous system." Ann Rheum Dis 50(10): 713-716.

Namjou, B., M. Keddache, D. Fletcher, S. Dillon, L. Kottyan, G. Wiley, P. M. Gaffney, B. E. Wakeland, C. Liang, E. K. Wakeland, R. H. Scofield, K. Kaufman and J. B. Harley (2012). "Identification of novel coding mutation in C1qA gene in an African-American pedigree with lupus and C1q deficiency." Lupus 21(10): 1113- 1118.

Navratil, J. S., S. Manzi, A. H. Kao, S. Krishnaswami, C. C. Liu, M. J. Ruffing, P. S. Shaw, A. C. Nilson, E. R. Dryden, J. J. Johnson and J. M. Ahearn (2006). "Platelet C4d is highly specific for systemic lupus erythematosus." Arthritis Rheum 54(2): 670-674.

Navratil, J. S., S. C. Watkins, J. J. Wisnieski and J. M. Ahearn (2001). "The globular heads of C1q specifically recognize surface blebs of apoptotic vascular endothelial cells." J Immunol 166(5): 3231-3239.

Nguyen, V. C., M. Tosi, M. S. Gross, O. Cohen-Haguenauer, C. Jegou-Foubert, M. F. de Tand, T. Meo and J. Frezal (1988). "Assignment of the complement serine protease genes C1r and C1s to chromosome 12 region 12p13." Hum Genet 78(4): 363-368.

Nordin Fredrikson, G., L. Truedsson, A. G. Sjoholm and M. Kjellman (1991). "DNA analysis in a MHC heterozygous patient with complete C4 deficiency-- homozygosity for C4 gene deletion and C4 pseudogene." Exp Clin Immunogenet 8(1): 29-37.

O'Hanlon, T. P., D. M. Carrick, I. N. Targoff, F. C. Arnett, J. D. Reveille, M. Carrington, X. Gao, C. V. Oddis, P. A. Morel, J. D. Malley, K. Malley, E. A. Shamim, L. G. Rider, S. J. Chanock, C. B. Foster, T. Bunch, P. J. Blackshear, P. H. Plotz, L. A. Love and F. W. Miller (2006). "Immunogenetic risk and protective factors for the idiopathic inflammatory myopathies: distinct HLA-A, -B, -Cw, -DRB1, and - DQA1 allelic profiles distinguish European American patients with different myositis autoantibodies." Medicine (Baltimore) 85(2): 111-127.

O'Neill, G. J., S. Y. Yang and B. Dupont (1978). "Two HLA-linked loci controlling the fourth component of human complement." Proc Natl Acad Sci U S A 75(10): 5165-5169.

Ochonisky, S., L. Intrator, J. Wechsler, J. Revuz and M. Bagot (1993). "Acquired C1 inhibitor deficiency revealing systemic lupus erythematosus." Dermatology 186(4): 261-263.

175

Orbai, A. M., L. Truedsson, G. Sturfelt, O. Nived, H. Fang, G. S. Alarcon, C. Gordon, J. Merrill, P. R. Fortin, I. N. Bruce, D. A. Isenberg, D. J. Wallace, R. Ramsey- Goldman, S. C. Bae, J. G. Hanly, J. Sanchez-Guerrero, A. E. Clarke, C. B. Aranow, S. Manzi, M. B. Urowitz, D. D. Gladman, K. C. Kalunian, M. I. Costner, V. P. Werth, A. Zoma, S. Bernatsky, G. Ruiz-Irastorza, M. A. Khamashta, S. Jacobsen, J. P. Buyon, P. Maddison, M. A. Dooley, R. F. Van Vollenhoven, E. Ginzler, T. Stoll, C. Peschken, J. L. Jorizzo, J. P. Callen, S. S. Lim, B. J. Fessler, M. Inanc, D. L. Kamen, A. Rahman, K. Steinsson, A. G. Franks, Jr., L. Sigler, S. Hameed, N. Pham, R. Brey, M. H. Weisman, G. McGwin, Jr., L. S. Magder and M. Petri (2015). "Anti-C1q antibodies in systemic lupus erythematosus." Lupus 24(1): 42-49.

Pachman, L. M., T. O. Fedczyna, T. S. Lechman and J. Lutz (2001). "Juvenile dermatomyositis: the association of the TNF alpha-308A allele and disease chronicity." Curr Rheumatol Rep 3(5): 379-386.

Pachman, L. M., J. M. Friedman, M. L. Maryjowski-Sweeney, O. Jonnason, R. M. Radvany, G. C. Sharp, M. A. Cobb, N. D. Battles, W. E. Crowe, C. W. Fink and et al. (1985). "Immunogenetic studies of juvenile dermatomyositis. III. Study of antibody to organ-specific and nuclear antigens." Arthritis Rheum 28(2): 151-157.

Pachman, L. M., J. R. Hayford, M. C. Hochberg, M. A. Pallansch, A. Chung, C. D. Daugherty, B. H. Athreya, S. L. Bowyer, C. W. Fink, H. L. Gewanter, R. Jerath, B. A. Lang, I. S. Szer, J. Sinacore, M. L. Christensen and A. R. Dyer (1997). "New-onset juvenile dermatomyositis: comparisons with a healthy cohort and children with juvenile rheumatoid arthritis." Arthritis Rheum 40(8): 1526-1533.

Pachman, L. M., R. Lipton, R. Ramsey-Goldman, E. Shamiyeh, K. Abbott, E. P. Mendez, A. Dyer, D. M. Curdy, L. Vogler, A. Reed, G. Cawkwell, L. Zemel, C. Sandborg, R. Rivas-Chacon, C. Hom, N. Ilowite, A. Gedalia, J. Gitlin and M. Borzy (2005). "History of infection before the onset of juvenile dermatomyositis: results from the National Institute of Arthritis and Musculoskeletal and Skin Diseases Research Registry." Arthritis Rheum 53(2): 166-172.

Parkes, M., A. Cortes, D. A. van Heel and M. A. Brown (2013). "Genetic insights into common pathways and complex relationships among immune-mediated diseases." Nat Rev Genet 14(9): 661-673.

Paul, E., O. O. Pozdnyakova, E. Mitchell and M. C. Carroll (2002). "Anti-DNA autoreactivity in C4-deficient mice." Eur J Immunol 32(9): 2672-2679.

Pereira, K. M., A. G. Faria, B. L. Liphaus, A. A. Jesus, C. A. Silva, M. Carneiro-Sampaio and L. E. Andrade (2016). "Low C4, C4A and C4B gene copy numbers are stronger risk factors for juvenile-onset than for adult-onset systemic lupus erythematosus." Rheumatology (Oxford) 10.1093/rheumatology/kev436. 176

Petri, M., A. M. Orbai, G. S. Alarcon, C. Gordon, J. T. Merrill, P. R. Fortin, I. N. Bruce, D. Isenberg, D. J. Wallace, O. Nived, G. Sturfelt, R. Ramsey-Goldman, S. C. Bae, J. G. Hanly, J. Sanchez-Guerrero, A. Clarke, C. Aranow, S. Manzi, M. Urowitz, D. Gladman, K. Kalunian, M. Costner, V. P. Werth, A. Zoma, S. Bernatsky, G. Ruiz-Irastorza, M. A. Khamashta, S. Jacobsen, J. P. Buyon, P. Maddison, M. A. Dooley, R. F. van Vollenhoven, E. Ginzler, T. Stoll, C. Peschken, J. L. Jorizzo, J. P. Callen, S. S. Lim, B. J. Fessler, M. Inanc, D. L. Kamen, A. Rahman, K. Steinsson, A. G. Franks, Jr., L. Sigler, S. Hameed, H. Fang, N. Pham, R. Brey, M. H. Weisman, G. McGwin, Jr. and L. S. Magder (2012). "Derivation and validation of the Systemic Lupus International Collaborating Clinics classification criteria for systemic lupus erythematosus." Arthritis Rheum 64(8): 2677-2686.

Petry, F., A. I. Berkel and M. Loos (1997). "Multiple identification of a particular type of hereditary C1q deficiency in the Turkish population: review of the cases and additional genetic and functional analysis." Hum Genet 100(1): 51-56.

Petry, F., G. Hauptmann, J. Goetz, E. Grosshans and M. Loos (1997). "Molecular basis of a new type of C1q-deficiency associated with a non-functional low molecular weight (LMW) C1q: parallels and differences to other known genetic C1q- defects." Immunopharmacology 38(1-2): 189-201.

Pickering, M. C., M. Botto, P. R. Taylor, P. J. Lachmann and M. J. Walport (2000). "Systemic lupus erythematosus, complement deficiency, and apoptosis." Adv Immunol 76: 227-324.

Pickering, M. C. and H. T. Cook (2008). "Translational mini-review series on complement factor H: renal diseases associated with complement factor H: novel insights from humans and animals." Clin Exp Immunol 151(2): 210-230.

Pickering, M. C., P. Macor, J. Fish, P. Durigutto, F. Bossi, F. Petry, M. Botto and F. Tedesco (2008). "Complement C1q and C8beta deficiency in an individual with recurrent bacterial meningitis and adult-onset systemic lupus erythematosus-like illness." Rheumatology (Oxford) 47(10): 1588-1589.

Pickering, R. J., G. B. Naff, R. M. Stroud, R. A. Good and H. Gewurz (1970). "Deficiency of C1r in human serum. Effects on the structure and function of macromolecular C1." J Exp Med 131(4): 803-815.

Pillemer, L., L. Blum, I. H. Lepow, O. A. Ross, E. W. Todd and A. C. Wardlaw (1954). "The properdin system and immunity. I. Demonstration and isolation of a new serum protein, properdin, and its role in immune phenomena." Science 120(3112): 279-285.

177

Pons-Estel, G. J., G. S. Alarcon, L. Scofield, L. Reinlib and G. S. Cooper (2010). "Understanding the epidemiology and progression of systemic lupus erythematosus." Semin Arthritis Rheum 39(4): 257-268.

Prodeus, A. P., S. Goerg, L. M. Shen, O. O. Pozdnyakova, L. Chu, E. M. Alicot, C. C. Goodnow and M. C. Carroll (1998). "A critical role for complement in maintenance of self-tolerance." Immunity 9(5): 721-731.

Putterman, C., R. Furie, R. Ramsey-Goldman, A. Askanase, J. Buyon, K. Kalunian, W. W. Chatham, E. Massarotti, K. Kirou, N. Jordan, I. Blanco, A. Weinstein, P. Chitkara, S. Manzi, J. Ahearn, T. O'Malley, J. Conklin, C. Ibarra, D. Barken and T. Dervieux (2014). "Cell-bound complement activation products in systemic lupus erythematosus: comparison with anti-double-stranded DNA and standard complement measurements." Lupus Sci Med 1(1): e000056.

Ramanan, A. V. and B. M. Feldman (2002). "Clinical features and outcomes of juvenile dermatomyositis and other childhood onset myositis syndromes." Rheum Dis Clin North Am 28(4): 833-857.

Reed, A. M., L. Pachman and C. Ober (1991). "Molecular genetic studies of major histocompatibility complex genes in children with juvenile dermatomyositis: increased risk associated with HLA-DQA1 *0501." Hum Immunol 32(4): 235- 240.

Reed, A. M. and J. D. Stirling (1995). "Association of the HLA-DQA1*0501 allele in multiple racial groups with juvenile dermatomyositis." Hum Immunol 44(3): 131- 135.

Reid, K. B. and R. A. Thompson (1983). "Characterization of a non-functional form of C1q found in patients with a genetically linked deficiency of C1q activity." Mol Immunol 20(10): 1117-1125.

Reilly, B. D. (1999). "Analysis of human C4A and C4B binding to an immune complex in serum." Clin Exp Immunol 117(1): 12-18.

Ricklin, D., G. Hajishengallis, K. Yang and J. D. Lambris (2010). "Complement: a key system for immune surveillance and homeostasis." Nat Immunol 11(9): 785-797.

Rider, L. G. and F. W. Miller (2011). "Deciphering the clinical presentations, pathogenesis, and treatment of the idiopathic inflammatory myopathies." JAMA 305(2): 183-190.

Rider, L. G., M. Shah, G. Mamyrova, A. M. Huber, M. M. Rice, I. N. Targoff and F. W. Miller (2013). "The myositis autoantibody phenotypes of the juvenile idiopathic inflammatory myopathies." Medicine (Baltimore) 92(4): 223-243. 178

Rider, L. G., L. Wu, G. Mamyrova, I. N. Targoff and F. W. Miller (2010). "Environmental factors preceding illness onset differ in phenotypes of the juvenile idiopathic inflammatory myopathies." Rheumatology (Oxford) 49(12): 2381-2390.

Rigby, W. F., Y. L. Wu, M. Zan, B. Zhou, S. Rosengren, C. Carlson, W. Hilton and C. Y. Yu (2012). "Increased frequency of complement C4B deficiency in rheumatoid arthritis." Arthritis Rheum 64(5): 1338-1344.

Robb, S. A., A. H. Fielder, C. E. Saunders, N. J. Davey, M. W. Burley, D. H. Lord, J. R. Batchelor and V. Dubowitz (1988). "C4 complement allotypes in juvenile dermatomyositis." Hum Immunol 22(1): 31-38.

Robinson, A. B. and A. M. Reed (2011). "Clinical features, pathogenesis and treatment of juvenile and adult dermatomyositis." Nat Rev Rheumatol 7(11): 664-675.

Rothwell, S., R. G. Cooper, I. E. Lundberg, F. W. Miller, P. K. Gregersen, J. Bowes, J. Vencovsky, K. Danko, V. Limaye, A. Selva-O'Callaghan, M. G. Hanna, P. M. Machado, L. M. Pachman, A. M. Reed, L. G. Rider, J. Cobb, H. Platt, O. Molberg, O. Benveniste, P. Mathiesen, T. Radstake, A. Doria, J. De Bleecker, B. De Paepe, B. Maurer, W. E. Ollier, L. Padyukov, T. P. O'Hanlon, A. Lee, C. I. Amos, C. Gieger, T. Meitinger, J. Winkelmann, L. R. Wedderburn, H. Chinoy and J. A. Lamb (2015). "Dense genotyping of immune-related loci in idiopathic inflammatory myopathies confirms HLA alleles as the strongest genetic risk factor and suggests different genetic background for major clinical subgroups." Ann Rheum Dis 10.1136/annrheumdis-2015-208119.

Roumenina, L. T., D. Sene, M. Radanova, J. Blouin, L. Halbwachs-Mecarelli, M. A. Dragon-Durey, W. H. Fridman and V. Fremeaux-Bacchi (2011). "Functional complement C1q abnormality leads to impaired immune complexes and apoptotic cell clearance." J Immunol 187(8): 4369-4373.

Rullo, O. J. and B. P. Tsao (2013). "Recent insights into the genetic basis of systemic lupus erythematosus." Ann Rheum Dis 72 Suppl 2: ii56-61.

Rupert, K. L., J. M. Moulds, Y. Yang, F. C. Arnett, R. W. Warren, J. D. Reveille, B. L. Myones, C. A. Blanchong and C. Y. Yu (2002). "The molecular basis of complete complement C4A and C4B deficiencies in a systemic lupus erythematosus patient with homozygous C4A and C4B mutant genes." J Immunol 169(3): 1570-1578.

Saxena, K., K. J. Kitzmiller, Y. L. Wu, B. Zhou, N. Esack, L. Hiremath, E. K. Chung, Y. Yang and C. Y. Yu (2009). "Great genotypic and phenotypic diversities associated with copy-number variations of complement C4 and RP-C4-CYP21- TNX (RCCX) modules: a comparison of Asian-Indian and European American populations." Mol Immunol 46(7): 1289-1303. 179

Schejbel, L., L. Skattum, S. Hagelberg, A. Ahlin, B. Schiller, S. Berg, F. Genel, L. Truedsson and P. Garred (2011). "Molecular basis of hereditary C1q deficiency-- revisited: identification of several novel disease-causing mutations." Genes Immun 12(8): 626-634.

Scherak, O., G. Kolarz and W. R. Mayr (1978). "HLA-B8 in caucasian patients with systemic lupus erythematosus." Scand J Rheumatol 7(1): 3-6.

Sebat, J., B. Lakshmi, J. Troge, J. Alexander, J. Young, P. Lundin, S. Maner, H. Massa, M. Walker, M. Chi, N. Navin, R. Lucito, J. Healy, J. Hicks, K. Ye, A. Reiner, T. C. Gilliam, B. Trask, N. Patterson, A. Zetterberg and M. Wigler (2004). "Large- scale copy number polymorphism in the human genome." Science 305(5683): 525-528.

Shah, M., G. Mamyrova, I. N. Targoff, A. M. Huber, J. D. Malley, M. M. Rice, F. W. Miller, L. G. Rider and G. Childhood Myositis Heterogeneity Collaborative Study (2013). "The clinical phenotypes of the juvenile idiopathic inflammatory myopathies." Medicine (Baltimore) 92(1): 25-41.

Shah, M., I. N. Targoff, M. M. Rice, F. W. Miller and L. G. Rider (2013). "Brief report: ultraviolet radiation exposure is associated with clinical and autoantibody phenotypes in juvenile myositis." Arthritis Rheum 65(7): 1934-1941.

Sim, E. and S. J. Cross (1986). "Phenotyping of human complement component C4, a class-III HLA antigen." Biochem J 239(3): 763-767.

Slingsby, J. H., P. Norsworthy, G. Pearce, A. K. Vaishnaw, H. Issler, B. J. Morley and M. J. Walport (1996). "Homozygous hereditary C1q deficiency and systemic lupus erythematosus. A new family and the molecular basis of C1q deficiency in three families." Arthritis Rheum 39(4): 663-670.

Smyth, G. K. and T. Speed (2003). "Normalization of cDNA microarray data." Methods 31(4): 265-273.

Stegert, M., M. Bock and M. Trendelenburg (2015). "Clinical presentation of human C1q deficiency: How much of a lupus?" Mol Immunol 67(1): 3-11.

Sternberg, S. H., B. LaFrance, M. Kaplan and J. A. Doudna (2015). "Conformational control of DNA target cleavage by CRISPR-Cas9." Nature 527(7576): 110-113.

Stone, N. M., A. Williams, J. D. Wilkinson and G. Bird (2000). "Systemic lupus erythematosus with C1q deficiency." Br J Dermatol 142(3): 521-524.

Sturfelt, G. and L. Truedsson (2012). "Complement in the immunopathogenesis of rheumatic disease." Nat Rev Rheumatol 8(8): 458-468.

180

Sun-Tan, C., T. T. Ozgur, G. Kilinc, R. Topaloglu, O. Gokoz, S. Ersoy-Evans and O. Sanal (2010). "Hereditary C1q deficiency: a new family with C1qA deficiency." Turk J Pediatr 52(2): 184-186.

Sverdlov, E. D. (2000). "Retroviruses and primate evolution." Bioessays 22(2): 161-171.

Symmons, D. P., J. A. Sills and S. M. Davis (1995). "The incidence of juvenile dermatomyositis: results from a nation-wide study." Br J Rheumatol 34(8): 732- 736.

Tan, E. M., A. S. Cohen, J. F. Fries, A. T. Masi, D. J. McShane, N. F. Rothfield, J. G. Schaller, N. Talal and R. J. Winchester (1982). "The 1982 revised criteria for the classification of systemic lupus erythematosus." Arthritis Rheum 25(11): 1271- 1277.

Tan, F. K. and F. C. Arnett (1998). "The genetics of lupus." Curr Opin Rheumatol 10(5): 399-408.

Tezak, Z., E. P. Hoffman, J. L. Lutz, T. O. Fedczyna, D. Stephan, E. G. Bremer, I. Krasnoselska-Riz, A. Kumar and L. M. Pachman (2002). "Gene expression profiling in DQA1*0501+ children with untreated dermatomyositis: a novel model of pathogenesis." J Immunol 168(8): 4154-4163.

Thorsby, E. and B. A. Lie (2005). "HLA associated genetic predisposition to autoimmune diseases: Genes involved and possible mechanisms." Transpl Immunol 14(3-4): 175-182.

Topaloglu, R., A. Bakkaloglu, J. H. Slingsby, M. J. Mihatsch, M. Pascual, P. Norsworthy, B. J. Morley, U. Saatci, J. A. Schifferli and M. J. Walport (1996). "Molecular basis of hereditary C1q deficiency associated with SLE and IgA nephropathy in a Turkish family." Kidney Int 50(2): 635-642.

Trowsdale, J. and J. C. Knight (2013). "Major histocompatibility complex genomics and human disease." Annu Rev Genomics Hum Genet 14: 301-323.

Tsokos, G. C. (2011). "Systemic lupus erythematosus." N Engl J Med 365(22): 2110- 2121.

Tsuge, I., Y. Kondo, Y. Nakajima, N. Nakagawa, K. Imai, S. Nonoyama, K. Oshima, O. Ohara, M. Hatanaka, E. Kitano, H. Kitamura and A. Urisu (2010). "Hyper IgM syndrome and complement Clq deficiency in an individual with systemic lupus erythematosus-like disease." Clin Exp Rheumatol 28(4): 558-560.

Tsukamoto, H., K. Nagasawa, S. Yoshizawa, Y. Tada, A. Ueda, Y. Ueda and Y. Niho (1992). "Synthesis and regulation of the fourth component of complement (C4) in

181

the human monocytic cell line U937: comparison with that of the third component of complement (C3)." Immunology 75(4): 565-569.

Tusher, V. G., R. Tibshirani and G. Chu (2001). "Significance analysis of microarrays applied to the ionizing radiation response." Proc Natl Acad Sci U S A 98(9): 5116-5121.

Uejima, H., M. P. Lee, H. Cui and A. P. Feinberg (2000). "Hot-stop PCR: a simple and general assay for linear quantitation of allele ratios." Nat Genet 25(4): 375-376.

Ulgiati, D., L. S. Subrata and L. J. Abraham (2000). "The role of Sp family members, basic Kruppel-like factor, and E box factors in the basal and IFN-gamma regulated expression of the human complement C4 promoter." J Immunol 164(1): 300-307.

Undlien, D. E., B. A. Lie and E. Thorsby (2001). "HLA complex genes in type 1 diabetes and other autoimmune diseases. Which genes are involved?" Trends Genet 17(2): 93-100.

Vaishnaw, A. K., T. J. Mitchell, S. J. Rose, M. J. Walport and B. J. Morley (1998). "Regulation of transcription of the TATA-less human complement component C4 gene." J Immunol 160(9): 4353-4360. van Schaarenburg, R. A., N. A. Daha, J. J. Schonkeren, E. W. Nivine Levarht, D. J. van Gijlswijk-Janssen, F. A. Kurreeman, A. Roos, C. van Kooten, C. A. Koelman, M. R. Ernst-Kruis, R. E. Toes, T. W. Huizinga, A. C. Lankester and L. A. Trouw (2015). "Identification of a novel non-coding mutation in C1qB in a Dutch child with C1q deficiency associated with recurrent infections." Immunobiology 220(3): 422-427.

Wallis, R., A. W. Dodds, D. A. Mitchell, R. B. Sim, K. B. Reid and W. J. Schwaeble (2007). "Molecular interactions between MASP-2, C4, and C2 and their activation fragments leading to complement activation via the lectin pathway." J Biol Chem 282(11): 7844-7851.

Walport, M. J. (2001). "Complement. First of two parts." N Engl J Med 344(14): 1058- 1066.

Walport, M. J. (2001). "Complement. Second of two parts." N Engl J Med 344(15): 1140-1144.

Walport, M. J. (2002). "Complement and systemic lupus erythematosus." Arthritis Res 4 Suppl 3: S279-293.

182

Walsh, R. J., S. W. Kong, Y. Yao, B. Jallal, P. A. Kiener, J. L. Pinkus, A. H. Beggs, A. A. Amato and S. A. Greenberg (2007). "Type I interferon-inducible gene expression in blood is present and reflects disease activity in dermatomyositis and polymyositis." Arthritis Rheum 56(11): 3784-3792.

Wang, X., A. Circolo, M. L. Lokki, P. G. Shackelford, R. A. Wetsel and H. R. Colten (1998). "Molecular heterogeneity in deficiency of complement protein C2 type I." Immunology 93(2): 184-191.

Wetsel, R. A., J. Kulics, M. L. Lokki, P. Kiepiela, H. Akama, C. A. Johnson, P. Densen and H. R. Colten (1996). "Type II human complement C2 deficiency. Allele- specific amino acid substitutions (Ser189 --> Phe; Gly444 --> Arg) cause impaired C2 secretion." J Biol Chem 271(10): 5824-5831.

Whaley, K. (1980). "Biosynthesis of the complement components and the regulatory proteins of the alternative complement pathway by human peripheral blood monocytes." J Exp Med 151(3): 501-516.

Whitaker, J. N. and W. K. Engel (1972). "Vascular deposits of immunoglobulin and complement in idiopathic inflammatory myopathy." N Engl J Med 286(7): 333- 338.

Willems, E., L. Leyns and J. Vandesompele (2008). "Standardization of real-time PCR gene expression data from independent biological replicates." Anal Biochem 379(1): 127-129.

Witte, D. P., T. R. Welch and L. S. Beischel (1991). "Detection and cellular localization of human C4 gene expression in the renal tubular epithelial cells and other extrahepatic epithelial sources." Am J Pathol 139(4): 717-724.

Wu, Y. L., B. P. Brookshire, R. R. Verani, F. C. Arnett and C. Y. Yu (2011). "Clinical presentations and molecular basis of complement C1r deficiency in a male African-American patient with systemic lupus erythematosus." Lupus 20(11): 1126-1134.

Wu, Y. L., G. Hauptmann, M. Viguier and C. Y. Yu (2009). "Molecular basis of complete complement C4 deficiency in two North-African families with systemic lupus erythematosus." Genes Immun 10(5): 433-445.

Wu, Y. L., G. C. Higgins, R. M. Rennebohm, E. K. Chung, Y. Yang, B. Zhou, H. N. Nagaraja, D. J. Birmingham, B. H. Rovin, L. A. Hebert and C. Y. Yu (2006). "Three distinct profiles of serum complement C4 proteins in pediatric systemic lupus erythematosus (SLE) patients: tight associations of complement C4 and C3 protein levels in SLE but not in healthy subjects." Adv Exp Med Biol 586: 227- 247. 183

Wu, Y. L., S. L. Savelli, Y. Yang, B. Zhou, B. H. Rovin, D. J. Birmingham, H. N. Nagaraja, L. A. Hebert and C. Y. Yu (2007). "Sensitive and specific real-time polymerase chain reaction assays to accurately determine copy number variations (CNVs) of human complement C4A, C4B, C4-long, C4-short, and RCCX modules: elucidation of C4 CNVs in 50 consanguineous subjects with defined HLA genotypes." J Immunol 179(5): 3012-3025.

Yang, Y., E. K. Chung, Y. L. Wu, S. L. Savelli, H. N. Nagaraja, B. Zhou, M. Hebert, K. N. Jones, Y. Shu, K. Kitzmiller, C. A. Blanchong, K. L. McBride, G. C. Higgins, R. M. Rennebohm, R. R. Rice, K. V. Hackshaw, R. A. Roubey, J. M. Grossman, B. P. Tsao, D. J. Birmingham, B. H. Rovin, L. A. Hebert and C. Y. Yu (2007). "Gene copy-number variation and associated polymorphisms of complement component C4 in human systemic lupus erythematosus (SLE): low copy number is a risk factor for and high copy number is a protective factor against SLE susceptibility in European Americans." Am J Hum Genet 80(6): 1037-1054.

Yang, Y., E. K. Chung, B. Zhou, C. A. Blanchong, C. Y. Yu, G. Fust, M. Kovacs, A. Vatay, C. Szalai, I. Karadi and L. Varga (2003). "Diversity in intrinsic strengths of the human complement system: serum C4 protein concentrations correlate with C4 gene size and polygenic variations, hemolytic activities, and body mass index." J Immunol 171(5): 2734-2745.

Yang, Y., K. Lhotta, E. K. Chung, P. Eder, F. Neumair and C. Y. Yu (2004). "Complete complement components C4A and C4B deficiencies in human kidney diseases and systemic lupus erythematosus." J Immunol 173(4): 2803-2814.

Yang, Z., A. R. Mendoza, T. R. Welch, W. B. Zipf and C. Y. Yu (1999). "Modular variations of the human major histocompatibility complex class III genes for serine/threonine kinase RP, complement component C4, steroid 21-hydroxylase CYP21, and tenascin TNX (the RCCX module). A mechanism for gene deletions and disease associations." J Biol Chem 274(17): 12147-12156.

Yu, C., K. Driest, Y. Wu, K. Lintner, A. Patwardhan, C. H. Spencer, F. C. Arnett, W. F. Rigby, L. Hebert and G. Hauptmann (2014). Complement in Rheumatic Diseases. Encyclopedia of Medical Immunology: Autoimmune Diseases. B. Diamond and A. Davidson. New York, Springer Reference: 286-302.

Yu, C., G. Hauptmann, Y. Yang, Y. Wu, D. J. Birmingham, B. H. Rovin and L. A. Hebert (2007). Complement Deficiencies in Human Systemic Lupus Erythematosus (SLE) and SLE Nephritis: Epidemiology and Pathogenesis. Systemic Lupus Erythematosus: A Companion to Rheumatology. G. C. Tsokos, C. Gordon and J. Smolen. Philadelphia, PA, Mosby Elsevier: 183-193.

184

Yu, C., Z. Yang, C. A. Blanchong and W. Miller (2000). "The human and mouse MHC class III region: a parade of 21 genes at the centromeric segment." Immunol Today 21(7): 320-328.

Yu, C. Y. (1991). "The complete exon-intron structure of a human complement component C4A gene. DNA sequences, polymorphism, and linkage to the 21- hydroxylase gene." J Immunol 146(3): 1057-1066.

Yu, C. Y., K. T. Belt, C. M. Giles, R. D. Campbell and R. R. Porter (1986). "Structural basis of the polymorphism of human complement components C4A and C4B: gene size, reactivity and antigenicity." EMBO J 5(11): 2873-2881.

Yu, C. Y. and C. C. Whitacre (2004). "Sex, MHC and complement C4 in autoimmune diseases." Trends Immunol 25(12): 694-699.

Zhu, Z. B., T. P. Atkinson and J. E. Volanakis (1998). "A novel type II complement C2 deficiency allele in an African-American family." J Immunol 161(2): 578-584.

Zipfel, P. F., M. Edey, S. Heinen, M. Jozsi, H. Richter, J. Misselwitz, B. Hoppe, D. Routledge, L. Strain, A. E. Hughes, J. A. Goodship, C. Licht, T. H. Goodship and C. Skerka (2007). "Deletion of complement factor H-related genes CFHR1 and CFHR3 is associated with atypical hemolytic uremic syndrome." PLoS Genet 3(3): e41.

Zlatarova, A. S., M. Rouseva, L. T. Roumenina, M. Gadjeva, M. Kolev, I. Dobrev, N. Olova, R. Ghai, J. C. Jensenius, K. B. Reid, U. Kishore and M. S. Kojouharova (2006). "Existence of different but overlapping IgG- and IgM-binding sites on the globular domain of human C1q." Biochemistry 45(33): 9979-9988.

185