Genomic Analysis of Human Spinal Deformity and Characterization of a Zebrafish Disease Model Jillian Gwen Buchan Washington University in St

Home , CNOT11, KATNBL1, MTRNR2L6, NAGPA, NBPF10, OR52A5

Washington University in St. Louis Washington University Open Scholarship

All Theses and Dissertations (ETDs)

Spring 4-22-2014 Genomic Analysis of Human Spinal Deformity and Characterization of a Zebrafish Disease Model Jillian Gwen Buchan Washington University in St. Louis

Follow this and additional works at: https://openscholarship.wustl.edu/etd

Recommended Citation Buchan, Jillian Gwen, "Genomic Analysis of Human Spinal Deformity and Characterization of a Zebrafish Disease Model" (2014). All Theses and Dissertations (ETDs). 1223. https://openscholarship.wustl.edu/etd/1223

This Dissertation is brought to you for free and open access by Washington University Open Scholarship. It has been accepted for inclusion in All Theses and Dissertations (ETDs) by an authorized administrator of Washington University Open Scholarship. For more information, please contact [email protected].

WASHINGTON UNIVERSITY IN ST. LOUIS

Division of Biology & Biomedical Sciences Molecular Genetics and Genomics

Dissertation Examination Committee: Christina A. Gurnett, Chair Carlos Cruchaga Alison M. Goate Matthew I. Goldsmith Kelly R. Monk Lilianna Solnica-Krezel

Genomic Analysis of Human Spinal Deformity and Characterization of a Zebrafish Disease Model

Jillian Gwen Buchan

A dissertation presented to the Graduate School of Arts and Sciences of Washington University in partial fulfillment of the requirements for the degree of Doctor of Philosophy

May 2014

St. Louis, Missouri

TABLE OF CONTENTS

List of Figures ….………………………………………………………………………… iv

List of Tables ….…………………………………………………………………………. vi

Acknowledgements ….…………………………………………………………………… viii

Abstract of the Dissertation ….…………………………………………………………… ix

Chapter 1: Introduction and Perspective …...………………………………………… 1

Figures and Tables ……………………………………………………………………….. 17

References ………………………………………………………………………………... 24

Chapter 2: Copy Number Variation in Adolescent Idiopathic Scoliosis ……………. 33

Summary …………………………………………………………………………………. 34

Introduction ………………………………………………………………………………. 35

Methods ….……………………………………………………………………………….. 36

Results ……………………………………………………………………………………. 38

Discussion ………………………………………………………………………………... 41

Figures and Tables ……………………………………………………………………….. 46

References ………………………………………………………………………………... 58

Chapter 3: Exome Sequencing Identifies Rare Variants in FBN1 and FBN2 Associated with Severe Adolescent Idiopathic Scoliosis ……………………………… 63

Summary …………………………………………………………………………………. 64

Introduction ………………………………………………………………………………. 65

Methods ….……………………………………………………………………………….. 66

Results ……………………………………………………………………………………. 71

Discussion ………………………………………………………………………………... 78

Figures and Tables ……………………………………………………………………….. 84

References ………………………………………………………………………………... 92

Chapter 4: Kinesin Family Member 6 (kif6) is Necessary for Normal Vertebral Development in Zebrafish …………………………………………………... 97

Summary …………………………………………………………………………………. 98

Introduction ………………………………………………………………………………. 99

Methods ….……………………………………………………………………………….. 101

Results ……………………………………………………………………………………. 107

Discussion ………………………………………………………………………………... 113

Figures and Tables ……………………………………………………………………….. 117

References ………………………………………………………………………………... 127

Chapter 5: Conclusions and Future Directions ...…………...………………………… 131

References ………………………………………………………………………………... 141

iii

LIST OF FIGURES

Chapter 1: Introduction and Perspective

Figure 1.1: Scoliosis spinal deformity …………………………………………………… 17

Figure 1.2: Schematic demonstrating the Cobb method of measuring spinal curvature in scoliosis ……………...…………………………………………... 18

Figure 1.3: Prevalence of AIS by severity ……………………………………………….. 19

Figure 1.4: Preoperative and postoperative radiograph of an AIS patient ….……………. 20

Figure 1.5: A proposed genetic model for AIS development ………...………………….. 21

Figure 1.6: Disorganized and reduced fibrillin staining in AIS ………………………….. 22

Chapter 2: Copy Number Variation in Adolescent Idiopathic Scoliosis

Figure 2.1: Proximal chromosome 1q21.1 duplications identified in AIS patients ……… 46

Figure 2.2: Segregation of chromosome 1q21.1 duplications …………………………… 47

Chapter 3: Exome Sequencing Identifies Rare Variants in FBN1 and FBN2 Associated with Severe Adolescent Idiopathic Scoliosis

Figure 3.1: Rare FBN1 and FBN2 variants identified in AIS patients …………………… 84

Figure 3.2: Segregation of FBN1 and FBN2 variants ……………………………..…...... 85

Figure 3.3: Rare FBN1 and FBN2 variants are associated with curve severity in AIS …... 86

Figure 3.4: Elevated pSMAD2 in AIS patients with rare FBN1 variants ………………... 87

Chapter 4: Kinesin Family Member 6 (kif6) is Necessary for Normal Vertebral Development in Zebrafish

Figure 4.1: Skolios mutants develop recessively-inherited curvature of the body axis ….. 117

Figure 4.2: Spinal curvature in skolios mutants occurs independent of major vertebral abnormalities and progressive through adult stages in males ………………… 118

Figure 4.3: Vertebral fusions and malformations occur at similar frequencies in skolios and WT zebrafish ……………………………………………………... 119

Figure 4.4: Meitoc mapping and whole genome sequencing identifies nonsense mutation in kif6 as candidate for skolios ……………………………………… 120

Figure 4.5: Transient knockdown of kif6 using morpholinos ……………………………. 121

Figure 4.6: Design and efficiency of kif6 TALENs ……………………………………… 122

Figure 4.7: TALEN-induced mutations in kif6 recapitulate the skolios phenotype ……… 123

Figure 4.8: RT-PCR expression of kif6 in WT embryos and adult tissues ………………. 124

Figure 4.9: Cilia structure and function are normal in skolios mutants ………………….. 125

LIST OF TABLES

Chapter 1: Introduction and Perspective

Table 1.1: Four broad categories of scoliosis ……………………………………………. 23

Chapter 2: Copy Number Variation in Adolescent Idiopathic Scoliosis

Table 2.1: Demographics of 143 AIS probands analyzed for copy number variation …… 48

Table 2.2: Frequency of 45 genomic copy number disorder regions in AIS patients ……. 49

Table 2.3: Clinical information for individuals identified with chromosome 1q21.1 duplications …………………………………………………………………… 50

Table 2.4: Summary of rare CNVs (<1% frequency in DGV) identified in AIS patients …………………………………………………………………… 51

Table 2.5: Rare CNVs identified in AIS cases …………………………………………… 52

Chapter 3: Exome Sequencing Identifies Rare Variants in FBN1 and FBN2 Associated with Severe Adolescent Idiopathic Scoliosis

Table 3.1: Rare variant frequencies for FBN1 and FBN2 in AIS cases and controls of European ancestry …………………………………………………………….. 88

Table 3.2: Rare FBN1 and FBN2 variants identified in AIS cases (curve ≥10°) ………… 89

Table 3.3: Clinical characteristics of AIS cases (curve ≥10°) of European ancestry with rare variants in FBN1 and FBN2 ……………………………………………… 90

Table 3.4: Clinical features of AIS patients evaluated for Marfan syndrome …………… 91

Chapter 4: Kinesin Family Member 6 (kif6) is Necessary for Normal Vertebral Development in Zebrafish

Table 4.1: Identification of candidate gene using low-coverage whole genome sequencing …………………………………………………………………….. 126

vii

ACKNOWLEDGEMENTS

First and foremost, I’d like to thank my thesis advisor, Dr. Christina Gurnett. Chris was a wonderful mentor who provided me with exceptional guidance, sage advice and unwavering support. I have truly appreciated her patience, encouragement and intelligence throughout my graduate studies.

I’d like to thank the current and past members of the Gurnett lab: Dr. David Alvarado,

Dr. Kyungsoo Ha, Dr. Gabe Haller, Kevin McCall, Piny Yang and Hyuliya Aferol. Whether it was scientific debates, pondering new techniques or trying to figure out Kevin’s puns, I learned a lot from all of you and thoroughly enjoyed my time in the lab. I hope I provided enough “data” to show my appreciation. I’d especially like to thank David, who not only developed several experimental and computational protocols that made much of this work possible, but also served as an additional mentor who was always willing to teach me new techniques and share ideas.

I’d also like to acknowledge the members of my thesis committee: Drs. Alison Goate,

Anne Bowcock, Matthew Goldsmith, Kelly Monk and Lila Solnica-Krezel as well as other

faculty at Washington University in St. Louis who have given me advice and feedback. Thank

you especially to Dr. Matthew Goldsmith for allowing me to collaborate on the skolios project.

Thank you to all of my friends back in Seattle, here in St. Louis and abroad who have

been an invaluable source of inspiration, motivation and comic relief.

Finally, I would have never found the courage or aspiration to pursue a PhD without my

family. Thank you to all my extended family: aunts, uncles, cousins and first cousins once-

removed. I am especially appreciative of my grandparents and my brother, Matt. Most importantly, I could not have gotten here without the love and encouragement from my wonderful parents. viii

ABSTRACT OF THE DISSERTATION

Genomic Analysis of Human Spinal Deformity and Characterization of a Zebrafish Disease Model

Jillian Gwen Buchan

Doctor of Philosophy in Biology & Biomedical Sciences Molecular Genetics and Genomics

Washington University in St. Louis, 2014

Associate Professor Christina Gurnett, Chair

Scoliosis is characterized by a lateral curvature of the spine that requires long-term bracing and invasive spinal surgery in cases with progressive deformity. Some individuals develop scoliosis secondary to congenital malformations or syndromic disorders, but most scoliosis is considered idiopathic and has no known cause. Adolescent idiopathic scoliosis (AIS) onsets in late childhood and causes spinal deformity in approximately 3% of the pediatric population. Despite a strong genetic basis, genetic risk factors for AIS are unknown and the pathogenesis remains poorly understood, which has been further hindered by the lack of a relevant animal model. Therefore, we used multiple approaches to better understand the genetic and molecular etiology of AIS using human and animal-based studies.

First, we performed copy number variation analysis on 143 patients with isolated scoliosis using the Affymetrix Genome-wide Human SNP Array 6.0. We identified a duplication of chromosome 1q21.1 in 2.1% (N=3/143) of AIS patients, which was enriched compared to 0.09% (N=1/1079) of controls (P=0.0057) and 0.07% (N=6/8329) of published

controls (P=0.0004). Other notable findings include trisomy X, which was identified in 1.8%

(N=2/114) of female AIS patients, and rearrangements of chromosome 15q11.2 and 16p11.2 that

may be relevant to scoliosis susceptibility. We also report rare CNVs that will be of use to future

studies investigating candidate genes for AIS.

Second, we performed a genome-wide rare variant burden analysis using exome sequence

data and identified FBN1 (fibrillin-1) as the most significantly associated gene with

AIS. Mutations in FBN1 are most frequently association with Marfan syndrome, a syndromic

condition that causes scoliosis in 60% of patients. Based on these results, FBN1 and a related

gene, FBN2 (fibrillin-2), were sequenced in a total of 852 AIS cases and 669 controls. In

individuals of European ancestry, rare variants in FBN1 and FBN2 were enriched in severely

affected AIS cases (7.6%) compared to in-house controls (2.4%) (OR=3.5, P=5.46×10-4) and

Exome Sequencing Project controls (2.3%) (OR=3.5, P=1.48×10-6). Scoliosis severity in AIS cases was associated with FBN1 and FBN2 rare variants (P=0.0012) and replicated in an

independent Han Chinese cohort (P=0.0376), suggesting that rare variants have utility as

predictors of curve progression. Clinical evaluations revealed that the majority of AIS cases

with rare FBN1 variants do not meet diagnostic criteria for Marfan syndrome, though variants

are associated with tall stature (P=0.0035) and upregulation of the TGF-β pathway.

Finally, we characterized a recessive zebrafish mutant, called skolios, which develops a

spinal deformity phenotype that parallels many features of human AIS, including an isolated

lateral curvature of the spine that arises independent of congenital vertebral malformations.

Skolios was identified in a mutagenesis screen and previously mapped to a 2.7 Mb region on

chromosome 17. Because skolios may be an informative model of scoliosis, we sought to

identify its genetic basis. We performed low coverage whole genome sequencing on a skolios

x mutant and identified a single nonsense mutation in the mapped region, which caused a premature stop in kinesin family member 6 (kif6), a poorly characterized kinesin of unknown function. To determine if the loss of kif6 is responsible for the skolios phenotype, we used

TALENs to create additional mutant alleles. We isolated three new TALEN-induced mutations that caused frameshift mutations in kif6. All zebrafish homozygous or compound heterozygous for kif6 frameshift mutations developed body axis curvature that was indistinguishable from skolios mutants, verifying kif6 as the causative gene. Preliminary investigation into the mechanism of spinal deformity revealed no association with vertebral malformations or cilia defects in skolios mutants.

Overall, these results identify important genetic risk factors for AIS, including clinically relevant copy number variants and rare genetic variation in FBN1 and FBN2. Moreover, we have identified an animal model with scoliosis and demonstrated a novel role for kif6 in the developing spine. These findings have significant impact on our understanding of the genetic basis of AIS and reveal new strategies to identify and treat AIS.

Chapter 1:

Introduction and Perspective

Scoliosis

Spinal deformity is a general term that encompasses several conditions where the spine is

abnormally curved or misaligned. Scoliosis is a form of spinal deformity that causes a complex,

three-dimensional curvature of the spine, with the predominate deviation from the midline

occurring in the frontal plane (Figure 1.1). Scoliosis is therefore quantified by the degree of

lateral curvature. The Cobb method is the standard and most reliable method of assessing the magnitude of spinal curvature in scoliosis and is quantified by measuring the angle, also called the Cobb angle, formed between lines drawn parallel to the most tilted vertebrae on a standing radiograph (Figure 1.2) [1,2]. Slight variation in the curvature of the spine is normal in the population; thus, pathological scoliosis is defined as curves measuring ≥10° using the Cobb method [3].

Scoliosis classification

Scoliosis is classified by the associated etiology, if known, and can be divided into four

general categories: congenital, neuromuscular, syndromic and idiopathic (Table 1.1). In

congenital scoliosis, spinal curvature is secondary to malformations of the vertebrae, such as

wedge-shaped vertebrae and hemivertebrae [4]. These vertebral malformations occur early in

development and are caused by defects during somatogenesis, which can result in imbalanced

spinal growth and scoliosis [5].

In neuromuscular and syndromic scoliosis, scoliosis arises secondary to a primary, systemic condition. For example, scoliosis is common in certain neuromuscular conditions, including 20-25% of patients with cerebral palsy [6] and 75-90% of patients with Duchenne’s

muscular dystrophy [7]. Similarly, scoliosis is a common feature of several syndromic

conditions, including 63% of patients with Marfan syndrome [8], which also causes a variety of

other cardiovascular, pulmonary, ocular and skeletal phenotypes.

When congenital, neuromuscular and syndromic origins of scoliosis have been ruled out,

scoliosis is considered idiopathic. Idiopathic scoliosis is therefore a diagnosis of exclusion that

typically occurs as an isolated condition in otherwise healthy individuals. Approximately 80% of all scoliosis is classified as idiopathic. Some patients with idiopathic scoliosis are identified early in life (0-9 years). However, in the majority of cases, scoliosis is detected in late childhood with progression coinciding with pubertal growth in adolescence. Thus, idiopathic scoliosis that is identified in late childhood is further categorized as adolescent idiopathic scoliosis (AIS).

Adolescent idiopathic scoliosis

AIS affects up to 3% of all children and is generally defined as pathological scoliosis of

unknown etiology that onsets after the age of 10. Mild deformity (curves measuring 10-20°)

affects males and females at similar frequencies [9]. Spinal curvature measuring >20° is less common, with the most severe curves being the least frequent (Figure 1.3). There is a female bias in AIS patients with severe curves. In patients with curves measuring >40°, the females-to-

male ratio is 10:1 [9]. Anthropometric studies have attempted to characterize other physical

observations in children with AIS. Such studies have noted reduced body mass index in both

female and male AIS patients [10,11]. AIS children are also typically taller than normal, age-

matched children, which is enhanced when lost height due to curvature is also factored into

measurements [12,13].

In most cases, AIS causes only mild deformity that rarely worsens after skeletal maturity

[14]. Therefore, mild scoliosis is not typically treated. AIS causing moderate (20-40°) and

severe (>40°) scoliosis has been associated with impaired lung function, resulting in reduced forced expiratory volume and forced vital capacity [15-17]. Severe scoliosis can also lead to

spinal stenosis with radiculopathy, functional limitations, back pain and cause a deformed

appearance that can be psychologically damaging [18-22]. Moreover, AIS measuring >30° continues to worsen after skeletal maturity, at an average of 1° per year [14]. Therefore, treatment is often necessary to improve quality of life and reduce the risk of negative physical and psychological outcomes in AIS patients with moderate and severe deformity.

Bracing is a nonoperative form of treatment that is considered once curves reach 20-25°.

There are many types of braces available to treat scoliosis (reviewed in [23]), but all braces are designed to align the spine using external force with the goal of limiting progression and correcting existing scoliosis. Compared to other types of scoliosis (e.g. congenital, neuromuscular), AIS patients are the most likely to benefit from bracing [23], which has been shown to effectively limit curve progression in some AIS patients [24].

When nonoperative treatments fail, surgery is considered, typically once curves exceed

40-45°. Surgical correction of scoliosis is most frequently achieved through the placement of instrumentation and spinal fusion (Figure 1.4) (reviewed in [25,26]). More than 18,000 children undergo major spinal surgery annually because of severe scoliosis [27]. While surgery is an effective way to correct AIS and prevent further progression of the deformity, there is a risk of complications. In a study of 702 patients who had surgical correction of AIS, non-neurologic complications (e.g. infection, excessive bleeding) occurred in 15.4% of patients [28]. In a separate study, 1,301 AIS patients who underwent spinal surgery were assessed for neurological complications (e.g. spinal cord injury), which occurred in 0.69% of AIS patients [29]. Because surgical correction of AIS is invasive and could potentially lead to complications, bracing is

often prescribed in patients at risk for progression to prevent the need for surgery. Overall, about

10% of AIS patients develop a deformity resulting in bracing or surgery [30]. Risk factors for

progressive AIS include skeletal immaturity, female gender and larger spinal curvature at initial

diagnosis [31].

Genetics of AIS

The pathogenesis of AIS is poorly understood. Many diverse models for the etiology of

AIS have been proposed, including defects in melatonin signaling, asymmetric development of

the spine coupled with biomechanical factors, low bone mass, abnormal estrogen levels,

shortened spinal cord, muscle imbalances and many others (reviewed in [32-35]). Among these models, no single concept is preferred as the direct cause of AIS. Instead, AIS is considered a multifactorial disorder that could arise from multiple etiologies.

Though the pathogenesis of AIS is disputed, it is well accepted that AIS has a genetic

basis. It has been recognized since the late 19th century that spinal curvatures tend to run in families [36]. At least 26% of AIS patients have one or more affected relatives [37], with 6-11%

of first-degree relatives affected [37,38], which exceeds what would be predicted based on AIS

frequencies in the general population. Additionally, monozygotic twins demonstrate higher

concordance for AIS (73%) compared to dizygotic twins (36%) [39]. Curve severity is also

significantly correlated in monozygotic twins [39].

Dominant autosomal and dominant X-linked inheritance has been described [32,40-42],

but most AIS families demonstrate complex, non-Mendelian inheritance. Emerging views of

AIS heritably favor a complex genetic model with large genetic heterogeneity [32,43-46]. With substantial variation in curve magnitude, rates of progression, age of onset and morphology of

the deformity among AIS patients, it is also possible that some of these factors arise from different genetic factors. For example, Cheng et al. [47] proposed a two-step model for the

development of progressive deformity. In the model, predisposition genes contribute to the

overall susceptibility to AIS, while different disease modifier genes contributing to the risk of

progressive scoliosis (Figure 1.5). Though additional studies are needed to support this two-step

model, it demonstrates the potential complexity underlying the genetics of AIS.

AIS linkage studies

Numerous family-based linkage associations have been reported in AIS (reviewed in

[48]). Parametric studies of large, single or multi-families have identified several loci that were significantly linked to AIS, including 3q12.1 [49], 5q13.3 [49], 9q31.2-q34.2 [50], 12p [51],

17p11.22 [52], 19p13.3 [53] and Xq22.3-q27.2 [42]. Nonparametric studies have identified

significant linkage regions at 6q15-q21 [54] and 10q23-q25.3 [54]. The genes within these loci

relevant to AIS susceptibility have not been determined.

Miller et al. [55] completed a genome-wide screen with 1,198 individuals in 202 AIS

families that had two or more affected individuals. In the study, 391 microsatellite markers

spaced on average 9 cM apart were genotyped to assess linkage in each family. Analysis of all

families under multiple models generated varying linkage results depending on the assumptions

made (model independent vs. autosomal dominance, AIS as a quantitative vs. qualitative trait,

varying severity thresholds). The overall findings indicated four major candidate regions on

chromosome 6, 9, 16 and 17. It was also noted that a subset of families with more severe curves

showed evidence of linkage to chromosome 19, though this locus was less significant when all

families were included in the analysis [55]. This subgroup was later studied in greater detail and

verified to overlap with the 19p13 region identified by Chan et al. [53], containing several candidate genes such as fibrillin-3 and thromboxane A2 receptor [56]. Similarly, chromosome

9q31-q34 was first identified by Miller et al. [55] and identified independently in a linkage study of a five generational family with eight affected individuals [50].

Wise et al. [57] identified linkage from a single Caucasian family on chromosome 18q with a multipoint nonparametric LOD score of 8.26. Gurnett et al. [58] performed linkage analysis on a five-generational Caucasian family with AIS affecting thirteen females and identified a novel chromosome 18q12.1-q12.2 locus with a parametric LOD score of 3.31, which was encompassed by the region from Wise et al. [56]. Pectus excavatum, a deformity causing abnormal growth of the ribs and sternum, was also present in one additional female and three male family members. Interestingly, scoliosis and pectus excavatum are often associated [59].

When family members with pectus excavatum were included in the analysis by Gurnett et al.

[58] as affected, the LOD score increased to 3.86. Approximately 45 genes were predicted within the linkage region, but no coding mutations were identified in 21 genes that were resequenced.

While some progress has been made using linkage analysis, successful linkage studies rely on large families with highly penetrant forms of disease in which individuals are unambiguously assigned as affected or unaffected. For diseases such as AIS, with variable expressivity and low penetrance, disease status can be difficult to assign in most families. It has been noted in previous AIS studies that reliable diagnosis can only be made if radiographs are obtained for all family members, since many mild forms of scoliosis can occur without the patient’s knowledge [32,57]. Obtaining radiographs from many individuals in large AIS families can be challenging; therefore, the usefulness of linkage analysis is limited in many AIS cohorts.

AIS candidate genes

Further efforts to understand the genetic basis of AIS have focused on candidate gene studies. Most of these studies have yielded mixed results, with both positive and negative findings (reviewed in [48]). Matrilin-1 (MATN1) was selected as a candidate for AIS bases on a scoliosis-like phenotype observed in MATN1 mutant mice [60]. MATN1 is an extracellular matrix protein that is primarily expressed in cartilage [61]. A microsatellite in the 3’ untranslated region (UTR) of MATN1 was over-transmitted from parents to affected probands in

50 parent-child trios [62]. An association with MATN1 was similarly found in a Chinese population [60]; however, no association was identified in a large Japanese population containing 789 AIS cases and 1,239 controls [63].

Genes involved in melatonin signaling have also been a focus of candidate gene studies based on findings in animal models (see section below). Nearly all studies of the melatonin receptors 1A (MTNR1A) and 1B (MTNR1B) yielded no association with AIS [63-68], with only one study reporting a positive association for MTNR1B [69].

Because AIS develops during puberty and has a female bias, both forms of the estrogen receptors (ESR1 and ESR2) have been implicated as candidate genes. Polymorphic sites in ESR1

(encoding ERα) were associated with progression of AIS >5° from initial evaluation and larger curves in 304 Japanese females [70]. These results were similarly found in a study of 202

Chinese AIS patients [71] and 67 Chinese AIS patients with spinal curves measuring ≥30° [72], but no association was identified in a study of 540 Chinese AIS patients with curves ≥20° [73].

ESR2 (encoding ERβ) was associated with AIS susceptibility in 218 patients [74]. However, a replication study involving 798 AIS patients and 637 controls of Japanese ancestry found no association with polymorphic sites in either ESR1 or ESR2 [75].

Gao et al. [76] used linkage analysis from 53 families to identify the chromodomain

helicase DNA-binding protein 7 gene (CDH7) as a candidate for AIS. Nonsense, missense and

splice-site mutations in CDH7 cause CHARGE syndrome (coloboma of the eye, heart defects, atresia of the choanae, retardation of growth and/or development, genital and/or urinary abnormalities, and ear abnormalities and deafness) [77]. CHARGE syndrome is also associated with high rates of scoliosis, which occurs in >60% of CHARGE adolescent and adult patients

[76,78]; thus, the authors hypothesized that milder mutations could contribute to isolated scoliosis in AIS patients. Sequencing CDH7 in 25 affected patients with AIS did not reveal any

rare single nucleotide polymorphisms (SNPs) that segregated with disease, although a common polymorphism (rs4738824) was shown to have significant overtransmission of the minor allele compared to controls [76].

FBN1 and FBN2 as candidate genes for AIS

Like CDH7 and CHARGE syndrome, other genes causing Mendelian disorders with syndromic forms of scoliosis have been considered as candidates for AIS. Hadley-Miller et al.

[79] hypothesized that the elastic fiber system of the extracellular matrix may could be a possible

etiology of AIS, particularly when there are defects in fibrillin. Fibrillin-1 (FBN1) and fibrillin-2

(FBN2) are glycoproteins that are the major components of 10nm microfibrils and provide structural support to the extracellular matrix [80,81]. Mutations in FBN1 cause Marfan syndrome [82], a rare, autosomal dominant disorders that cause scoliosis in 63% of patients [8].

Marfan syndrome is estimated to occur in 1:5,000-1:10,000 individuals. Aortic dilation is the

hallmark feature of Marfan syndrome and predisposes individuals to life-threatening aortic

aneurysms and dissections [83]. Additional features occur in Marfan syndrome with large

9 phenotypic variability and include ectopia lentis, tall stature, pectus excavatum and carinatum, dolichostenomelia, arachnodactyly and joint hypermobility.

Mutations in FBN2 cause congenital contractural arachnodactyly (also called Beals syndrome) [84], an autosomal dominant disorder that causes scoliosis in 50% of patients [85].

Congenital contractural arachnodactyly is also rare (<1:10,000) and has phenotypic overlap with

Marfan syndrome [86]. Many patients with congenital contractural arachnodactyly are described as having a Marfanoid appearance, with features such as pectus excavatum and carinatum, tall stature, dolichostenomelia and arachnodactyly. Distinctive features of congenital contractural arachnodactyly include “crumpled” ears, muscular hypoplasia and contractures of the major joints. Although congenital contractural arachnodactyly does not appear to be associated with the same risk of life-threatening cardiovascular complications as Marfan syndrome, aortic dilation has been described in several patients and echocardiographic evaluations are currently recommended [87].

FBN1 and FBN2 are large, complex genes that each consists of approximately 11kb of coding sequence spread over 65 exons. In Marfan syndrome, over 1,000 mutations have been described throughout FBN1 [88], with about two-thirds being missense mutations [89]. Most pathogenic FBN1 mutations are unique to a family [90] and about 20% are de novo [91].

Overall, there is little genotype-phenotype correlation within or between families, with the exception of the neonatal region of FBN1 (exons 24-32) that is typically associated with a more severe, early-onset form of Marfan syndrome [89,92]. The neonatal region is also conserved in

FBN2 (exons 23-34) and mutations causing congenital contractural arachnodactyly are limited to this region [87]. Overall, mutations causing congenital contractural arachnodactyly are

considerably less common than those causing Marfan syndrome, with only about 40 FBN2

mutations having been described [87,93-95].

Although Marfan syndrome and congenital contractural arachnodactyly are considered to

be highly penetrant, autosomal dominant disorders, FBN1 mutations have been described in families with isolated kyphoscoliosis [96] and isolated skeletal features including scoliosis [97].

This suggests that fibrillin could play a role in isolated scoliosis. Hadley-Miller et al. [79]

investigated FBN1 with histological specimens collected from AIS patients during surgery.

Immunohistochemical staining of FBN1 from ligamentum flavum showed disorganized and/or

reduced staining in 82% (N=18/22) of AIS patients compared to controls, which occurred

independent of the site of biopsy (e.g. concave, convex, apical of the curve) (Figure 1.6).

Similarly, cultured fibroblasts from AIS patients showed that 17% (N=4/23) of specimens

secreted fibrillin that failed to incorporate into the extracellular matrix. Linkage analysis from

eleven pedigrees containing 96 individuals did not support segregation of the FBN1 locus with

AIS, although segregation within a single family could not be ruled out [98]. Because of the

prohibitively high cost of sequencing these large genes, no further genetic studies of FBN1 or

FBN2 have been pursued in patients with AIS.

AIS genome-wide association studies

Genome-wide association studies (GWAS) rely on the ‘common disease, common

variant’ (CDCV) hypothesis, which argues that common variants in the population, typically

with low disease penetrance, are major contributors to common diseases. Several GWAS have

been performed with AIS cohorts to identify common risk loci. Sharma et al. [99] performed a

GWAS on 419 Texas AIS families with 1,122 individuals and identified common genetic

polymorphisms associated with AIS near the close homologue of L1 (CHL1) gene, although this

locus did not reach genome-wide significance. CHL1 is involved in axonal guidance and

neuronal migration [100,101] and Sharma et al. [99] hypothesized CHL1 could be a good

candidate for AIS based on evidence that incorrect axonal targeting can cause the autosomal

recessive disorder horizontal gaze palsy with progressive scoliosis. Horizontal gaze palsy with

progressive scoliosis is caused by homozygous or compound heterozygous mutations in a similar

axonal guidance gene, roundabout homolog 3 (ROBO3) [102]. Interestingly, other genes involved in axonal guidance pathways were also among the top associations, including down syndrome cell adhesion molecule (DSCAM) and contactin-associated protein-like 2 (CNTNAP2)

[99].

A second GWAS was performed in 1,376 Japanese females with AIS and 11,297 female controls [103]. Three common SNPs (rs11190870, rs625039, rs11598564) near the ladybird homeobox 1 (LBX1) gene were highly associated with AIS in this cohort, reaching genome-wide significance. Further analysis of LBX1 by reverse transcription polymerase chain reaction (RT-

PCR) showed LBX1 was expressed in human fetal skeletal muscle and spinal cord. Takahashi et al. [103] hypothesized that LBX1 could lead to scoliosis through abnormal somatosensory function. The most significant SNP near LBX1 (rs11190870) replicated in independent cohorts consisting of 300 AIS patients and 788 controls in a southern Chinese population [104] and 949

AIS patients and 976 controls in a Han Chinese population [105]. The association replicated for all three SNPs in a cohort of 513 AIS patients and 440 healthy controls from a Han Chinese population [106]. However, only one SNP (rs11598564) was among the top 100 SNPs associated with AIS in the Sharma et al. [99] study, which contained mostly individuals of

European ancestry, suggesting possible genetic heterogeneity among ancestral backgrounds.

The Japanese GWAS from Takahashi et al. [103] was extended by Kou et al. [107],

identifying six additional SNPs with suggestive evidence for association with AIS. One of these

SNPs (rs6570507) reached genome-wide significance after association testing was performed in

an independent cohort of AIS cases and controls. This SNP is in the intron of G protein-coupled

receptor 126 (GPR126). GPR126 was expressed in proliferating chondrocytes in mice and the

authors suggest that GPR126 could affect AIS susceptibility through abnormal skeletal growth.

Animal models for AIS

The genetic and molecular basis of AIS has been difficult to identify, which is in part

attributed to the lack of a relevant animal model. Most animal models of AIS have been

primarily limited to experimental forms of scoliosis, where scoliosis is induced by invasive

surgical procedures, immobilization or the use of the systemic agents (reviewed in [108]). One

of the most extensively studied models of experimentally-induced scoliosis is pinealectomy in chickens. First noted by Thillard [109] in 1959, surgical removal of the pineal gland in the first few days after chicks hatch caused chickens to develop scoliosis that progressed until sexual maturity [110]. The cause of scoliosis was hypothesized to be due to altered melatonin signaling, although later studies showed that chickens with suppressed melatonin secretion by constant light did not develop scoliosis [111]. Pinealectomy did not cause scoliosis in rats [112], unless pinealectomized rats were also made bipedal through amputation of the forelimbs [113].

This finding has led to speculations that bipedalism may be an important consideration for modeling AIS pathogenesis. However, pinealectomy in guppy [114] and salmon [115] induces scoliosis, indicating that the pinealectomy model of experimentally-induced scoliosis is not exclusive to bipedalism.

Because spinal deformity develops secondary to harsh interventions, animal models with

experimentally-induced forms of scoliosis are unlikely to recapitulate many features of human

AIS. Few animal models with sporadic, naturally occuring scoliosis have been identified. One

long-term study utilized a stock of chickens that were normal at birth but developed a scoliotic

phenotype as they approached sexual maturity [116]. Over 10 years, McCarrey et al. [116]

selected for scoliosis to create inbred scoliotic stocks and found that long-term selection

increased the incidence of scoliosis to 88.5%, but did not affect curve severity. Though the study

provided evidence for a major gene affect, follow-up studies to identify the genes have not been

performed.

More recently, a possible genetic model for scoliosis was identified in the guppy

(Poecilia reticulata). A single male guppy with spinal curvature was isolated from a laboratory

stock and crossed to generate a lineage called curveback [114]. Of curveback fish born with normal spines, 79% developed curves within 2 weeks. The age of onset, curve magnitude and progression varied. Severe spinal curvature was observed in a higher percentage of females than males. Numerous crossing schemes suggested that the curveback phenotype has complex inheritance, with a major gene and modifiers, some of which may be sex-linked. While curveback may prove to be a useful model for AIS, the genes causing the curveback phenotype and the potential association to human AIS remain to be elucidated.

Current understanding of AIS

In the past decade, discovery of disease-causing variants was most often achieved through linkage, candidate gene sequencing and association studies. These same approaches

have been utilized in the pursuit of genetic variants associated with AIS. Through linkage

studies, several loci associated with AIS have been successfully identified. However, these

regions have been large, containing many genes and the specific gene(s) relevant to AIS within these loci remain unknown. Candidate gene studies have yielded mixed findings, with most gene associations failing to replicate in other AIS cohorts. While this may be the result of large genetic heterogeneity between ancestral populations or endophenotypes unaccounted for in replication cohorts, many studies evaluating AIS candidate genes have used case-control designs with small sample sizes. Thus, additional studies involving larger cohorts of patients are needed to resolve previous findings. GWAS has successfully identified common variants in LBX1 associated with AIS in a large Japanese cohort [103], which replicated in independent AIS cohorts from other Asian ancestries [104-106]. However, it is unclear if LBX1 is associated with

AIS in other ancestral populations. Furthermore, common polymorphisms account for only a small amount of AIS heritability, suggesting that other forms of genetic variation are likely at play and may have a larger role in AIS susceptibility. Though progress has been made, overall, the genetic basis of AIS remains poorly understood. As an added challenge, few animal models have been described and it is currently unclear which species are appropriate to model genetic findings in human.

Dissertation objectives

The objective of the research presented here is to identify additional genetic risk factors for AIS, especially rare genetic variants that have not been adequately evaluated as a potential cause of disease. Our study design will address the ‘common disease, rare variant’ (CDRV)

hypothesis, which states that common diseases are likely due to rare variants of higher

penetrance [117]. First, we will evaluate rare genomic copy number variants in a cohort of AIS

cases. Rare copy number variation is an established cause of human disease, yet genomic

deletions, duplications and chromosomal aneuploidy have never been explored as a potential risk

factor for AIS. Second, we will utilize the recent advances in next generation sequencing

technology and DNA enrichment to evaluate rare disease-causing sequence variants (single base

mutations and small insertion/deletions) on a genome-wide scale.

The second major objective of this dissertation is to evaluate zebrafish as a potential animal model for AIS using an ENU-forward genetic screen. Naturally occurring scoliosis has been demonstrated in the guppy; therefore we hypothesize that the zebrafish, a highly tractable animal model with diverse molecular and genetic resources, is an ideal model to identify new mutant strains that could illuminate new candidate genes and molecular pathways that also play a role in human AIS.

Figure 1.1: Scoliosis spinal deformity Graphic showing the appearance of normal spine (left) and a spine with scoliosis (right). Figure reproduced from [118].

Figure 1.2: Schematic demonstrating the Cobb method of measuring spinal curvature in scoliosis To quantify the magnitude of scoliosis deformity, spinal curvature is measured from frontal plane using a standing radiograph. The curve, or Cobb angle, is the measured angle created when lines are drawn parallel to the most tilted vertebrae. Figure reproduced from [27].

Figure 1.3: Prevalence of AIS by severity AIS affect 2-3% of the population, but the prevalence decreases with increasing severity. Figure reproduced from [27].

Figure 1.4: Preoperative and postoperative radiograph of an AIS patient Preoperative radiograph (top) shows a patient with severe scoliosis from the frontal and lateral view. Postoperatively (bottom), the placement of instrumentation and spinal fusion has corrected the spinal curvature by 60%. Figure reproduced from [119].

Figure 1.5: A proposed genetic model for AIS development A two-step model of genetic and environmental factors contributing to AIS development. Predisposition genes contribute to the AIS susceptibility whereas disease modifier genes contribute to the risk of progression to severe AIS. Figure reproduced from [47].

Figure 1.6: Disorganized and reduced fibrillin staining in AIS Immunohistochemical staining of FBN1 in an individual not affected with scoliosis (top) and a patient with AIS (bottom). In the AIS patient, reduced staining and/or disorganization was observed in 82% of patients. Figure reproduced from [31].

Table 1.1: Four broad categories of scoliosis

Scoliosis classification Example(s)

Congenital Wedge-shaped vertebrae, hemivertebrae Cerebral palsy, Duchenne muscular dystrophy, arthrogryposis, Neuromuscular spinal muscular atrophy Marfan syndrome, Ehlers-Danlos syndrome, congenital contractural Syndromic arachnodactyly Idiopathic Adolescent idiopathic scoliosis

REFERENCES

1. Cobb JR (1948) Outline for the study of scoliosis. American Academy of Orthopaedic Surgeons Instructional Course Lectures 5: 261-265. 2. Keynan O, Fisher CG, Vaccaro A, Fehlings MG, Oner FC, Dietz J, Kwon B, Rampersaud R, Bono C, France J, Dvorak M (2006) Radiographic measurement parameters in thoracolumbar fractures: a systematic review and consensus statement of the spine trauma study group. Spine (Phila Pa 1976) 31: E156-165. 3. Boachie-Adjei O, Lonner B (1996) Spinal deformity. Pediatr Clin North Am 43: 883-897. 4. Hedequist D, Emans J (2007) Congenital scoliosis: a review and update. J Pediatr Orthop 27: 106-116. 5. Chan G, Dormans JP (2009) Update on congenital spinal deformities: preoperative evaluation. Spine (Phila Pa 1976) 34: 1766-1774. 6. Imrie MN, Yaszay B (2010) Management of spinal deformity in cerebral palsy. Orthop Clin North Am 41: 531-547. 7. Mo F, Cunningham ME (2011) Pediatric scoliosis. Curr Rev Musculoskelet Med 4: 175-182. 8. Sponseller PD, Hobbs W, Riley LH, 3rd, Pyeritz RE (1995) The thoracolumbar spine in Marfan syndrome. J Bone Joint Surg Am 77: 867-876. 9. Lenke L (2004) Idiopathic scoliosis. Philadelphia: Lippincot, Williams & Wilkins 3. 10. Wei-Jun W, Xu S, Zhi-Wei W, Xu-Sheng Q, Zhen L, Yong Q (2012) Abnormal anthropometric measurements and growth pattern in male adolescent idiopathic scoliosis. Eur Spine J 21: 77-83. 11. Barrios C, Cortes S, Perez-Encinas C, Escriva MD, Benet I, Burgos J, Hevia E, Piza G, Domenech P (2011) Anthropometry and body composition profile of girls with nonsurgically treated adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 36: 1470- 1477. 12. Siu King Cheung C, Tak Keung Lee W, Kit Tse Y, Ping Tang S, Man Lee K, Guo X, Qin L, Chun Yiu Cheng J (2003) Abnormal peri-pubertal anthropometric measurements and growth pattern in adolescent idiopathic scoliosis: a study of 598 patients. Spine (Phila Pa 1976) 28: 2152-2157. 13. Yim AP, Yeung HY, Hung VW, Lee KM, Lam TP, Ng BK, Qiu Y, Cheng JC (2012) Abnormal skeletal growth patterns in adolescent idiopathic scoliosis--a longitudinal study until skeletal maturity. Spine (Phila Pa 1976) 37: E1148-1154. 14. Weinstein SL, Ponseti IV (1983) Curve progression in idiopathic scoliosis. J Bone Joint Surg Am 65: 447-455. 15. Johnston CE, Richards BS, Sucato DJ, Bridwell KH, Lenke LG, Erickson M, Spinal Deformity Study G (2011) Correlation of preoperative deformity magnitude and pulmonary function tests in adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 36: 1096-1102.

16. Liu Z, Qiu Y, Wang B (2008) [Factors affecting the preoperative pulmonary function test in patients with scoliosis: a retrospective evaluation of 108 cases]. Zhonghua Yi Xue Za Zhi 88: 2457-2460. 17. Koumbourlis AC (2006) Scoliosis and the respiratory system. Paediatr Respir Rev 7: 152- 160. 18. Ramirez N, Johnston CE, Browne RH (1997) The prevalence of back pain in children who have idiopathic scoliosis. J Bone Joint Surg Am 79: 364-368. 19. Freidel K, Petermann F, Reichel D, Steiner A, Warschburger P, Weiss HR (2002) Quality of life in women with idiopathic scoliosis. Spine (Phila Pa 1976) 27: E87-91. 20. Fowles JV, Drummond DS, L'Ecuyer S, Roy L, Kassab MT (1978) Untreated scoliosis in the adult. Clin Orthop Relat Res: 212-217. 21. Berven S, Deviren V, Demir-Deviren S, Hu SS, Bradford DS (2003) Studies in the modified Scoliosis Research Society Outcomes Instrument in adults: validation, reliability, and discriminatory capacity. Spine (Phila Pa 1976) 28: 2164-2169; discussion 2169. 22. Shapiro GS, Taira G, Boachie-Adjei O (2003) Results of surgical treatment of adult idiopathic scoliosis with low back pain and spinal stenosis: a study of long-term clinical radiographic outcomes. Spine (Phila Pa 1976) 28: 358-363. 23. Heary RF, Bono CM, Kumar S (2008) Bracing for scoliosis. Neurosurgery 63: 125-130. 24. Weinstein SL, Dolan LA, Wright JG, Dobbs MB (2013) Effects of bracing in adolescents with idiopathic scoliosis. N Engl J Med 369: 1512-1521. 25. Deniz Olgun Z, Yazici M (2013) Posterior instrumentation and fusion. J Child Orthop 7: 69- 76. 26. Gummerson NW, Millner PA (2010) Spinal fusion for scoliosis, clinical decision-making and choice of approach and devices. Skeletal Radiol 39: 939-942. 27. Jacobs JJ, Andersson GBJ, Bell J, Weinstein SL, Dormans JP, Furman MB, Lane N, Puzas JE, St. Clair EW, Yelin EH (2011) Spinal Deformity and Related Conditions. The Burden of Musculoskeletal Diseases in the United States: American Academy of Orthopaedic Surgeons. 28. Carreon LY, Puno RM, Lenke LG, Richards BS, Sucato DJ, Emans JB, Erickson MA (2007) Non-neurologic complications following surgery for adolescent idiopathic scoliosis. J Bone Joint Surg Am 89: 2427-2432. 29. Diab M, Smith AR, Kuklo TR, Spinal Deformity Study G (2007) Neural complications in the surgical treatment of adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 32: 2759- 2763. 30. Miller NH (1999) Cause and natural history of adolescent idiopathic scoliosis. Orthop Clin North Am 30: 343-352, vii. 31. Reamy BV, Slakey JB (2001) Adolescent idiopathic scoliosis: review and current concepts. Am Fam Physician 64: 111-116. 32. Kouwenhoven JW, Castelein RM (2008) The pathogenesis of adolescent idiopathic scoliosis: review of the literature. Spine (Phila Pa 1976) 33: 2898-2908. 25

33. Dayer R, Haumont T, Belaieff W, Lascombes P (2013) Idiopathic scoliosis: etiological concepts and hypotheses. J Child Orthop 7: 11-16. 34. Wang WJ, Yeung HY, Chu WC, Tang NL, Lee KM, Qiu Y, Burwell RG, Cheng JC (2011) Top theories for the etiopathogenesis of adolescent idiopathic scoliosis. J Pediatr Orthop 31: S14-27. 35. Burwell RG, Dangerfield PH, Freeman BJ (2008) Concepts on the pathogenesis of adolescent idiopathic scoliosis. Bone growth and mass, vertebral column, spinal cord, brain, skull, extra-spinal left-right skeletal length asymmetries, disproportions and molecular pathogenesis. Stud Health Technol Inform 135: 3-52. 36. Keetley CB (1888) On the Etiology and Essential Nature of Scoliosis. Ann Surg 7: 126-136. 37. Riseborough EJ, Wynne-Davies R (1973) A genetic survey of idiopathic scoliosis in Boston, Massachusetts. J Bone Joint Surg Am 55: 974-982. 38. Wynne-Davies R (1968) Familial (idiopathic) scoliosis. A family survey. J Bone Joint Surg Br 50: 24-30. 39. Kesling KL, Reinker KA (1997) Scoliosis in twins. A meta-analysis of the literature and report of six cases. Spine (Phila Pa 1976) 22: 2009-2014; discussion 2015. 40. Robin GC, Cohen T (1975) Familial scoliosis. A clinical report. J Bone Joint Surg Br 57: 146-148. 41. De George FV, Fisher RL (1967) Idiopathic scoliosis: genetic and environmental aspects. J Med Genet 4: 251-257. 42. Justice CM, Miller NH, Marosy B, Zhang J, Wilson AF (2003) Familial idiopathic scoliosis: evidence of an X-linked susceptibility locus. Spine (Phila Pa 1976) 28: 589-594. 43. Miller NH (2007) Genetics of familial idiopathic scoliosis. Clin Orthop Relat Res 462: 6-10. 44. Ward K, Ogilvie J, Argyle V, Nelson L, Meade M, Braun J, Chettier R (2010) Polygenic inheritance of adolescent idiopathic scoliosis: a study of extended families in Utah. Am J Med Genet A 152A: 1178-1188. 45. Kruse LM, Buchan JG, Gurnett CA, Dobbs MB (2012) Polygenic threshold model with sex dimorphism in adolescent idiopathic scoliosis: the Carter effect. J Bone Joint Surg Am 94: 1485-1491. 46. Miller NH (2011) Idiopathic scoliosis: cracking the genetic code and what does it mean? J Pediatr Orthop 31: S49-52. 47. Cheng JC, Tang NL, Yeung HY, Miller N (2007) Genetic association of complex traits: using idiopathic scoliosis as an example. Clin Orthop Relat Res 462: 38-44. 48. Gorman KF, Julien C, Moreau A (2012) The genetic epidemiology of idiopathic scoliosis. Eur Spine J 21: 1905-1919. 49. Edery P, Margaritte-Jeannin P, Biot B, Labalme A, Bernard JC, Chastang J, Kassai B, Plais MH, Moldovan F, Clerget-Darpoux F (2011) New disease gene location and high genetic heterogeneity in idiopathic scoliosis. Eur J Hum Genet 19: 865-869.

50. Ocaka L, Zhao C, Reed JA, Ebenezer ND, Brice G, Morley T, Mehta M, O'Dowd J, Weber JL, Hardcastle AJ, Child AH (2008) Assignment of two loci for autosomal dominant adolescent idiopathic scoliosis to chromosomes 9q31.2-q34.2 and 17q25.3-qtel. J Med Genet 45: 87-92. 51. Raggio CL, Giampietro PF, Dobrin S, Zhao C, Dorshorst D, Ghebranious N, Weber JL, Blank RD (2009) A novel locus for adolescent idiopathic scoliosis on chromosome 12p. J Orthop Res 27: 1366-1372. 52. Salehi LB, Mangino M, De Serio S, De Cicco D, Capon F, Semprini S, Pizzuti A, Novelli G, Dallapiccola B (2002) Assignment of a locus for autosomal dominant idiopathic scoliosis (IS) to human chromosome 17p11. Hum Genet 111: 401-404. 53. Chan V, Fong GC, Luk KD, Yip B, Lee MK, Wong MS, Lu DD, Chan TK (2002) A genetic locus for adolescent idiopathic scoliosis linked to chromosome 19p13.3. Am J Hum Genet 71: 401-406. 54. Marosy B, Justice CM, Vu C, Zorn A, Nzegwu N, Wilson AF, Miller NH (2010) Identification of susceptibility loci for scoliosis in FIS families with triple curves. Am J Med Genet A 152A: 846-855. 55. Miller NH, Justice CM, Marosy B, Doheny KF, Pugh E, Zhang J, Dietz HC, 3rd, Wilson AF (2005) Identification of candidate regions for familial idiopathic scoliosis. Spine (Phila Pa 1976) 30: 1181-1187. 56. Alden KJ, Marosy B, Nzegwu N, Justice CM, Wilson AF, Miller NH (2006) Idiopathic scoliosis: identification of candidate regions on chromosome 19p13. Spine (Phila Pa 1976) 31: 1815-1819. 57. Wise CA, Barnes R, Gillum J, Herring JA, Bowcock AM, Lovett M (2000) Localization of susceptibility to familial idiopathic scoliosis. Spine (Phila Pa 1976) 25: 2372-2380. 58. Gurnett CA, Alaee F, Bowcock A, Kruse L, Lenke LG, Bridwell KH, Kuklo T, Luhmann SJ, Dobbs MB (2009) Genetic linkage localizes an adolescent idiopathic scoliosis and pectus excavatum gene to chromosome 18 q. Spine (Phila Pa 1976) 34: E94-100. 59. Hong JY, Suh SW, Park HJ, Kim YH, Park JH, Park SY (2011) Correlations of adolescent idiopathic scoliosis and pectus excavatum. J Pediatr Orthop 31: 870-874. 60. Chen Z, Tang NL, Cao X, Qiao D, Yi L, Cheng JC, Qiu Y (2009) Promoter polymorphism of matrilin-1 gene predisposes to adolescent idiopathic scoliosis in a Chinese population. Eur J Hum Genet 17: 525-532. 61. Deak F, Wagener R, Kiss I, Paulsson M (1999) The matrilins: a novel family of oligomeric extracellular matrix proteins. Matrix Biol 18: 55-64. 62. Montanaro L, Parisini P, Greggi T, Di Silvestre M, Campoccia D, Rizzi S, Arciola CR (2006) Evidence of a linkage between matrilin-1 gene (MATN1) and idiopathic scoliosis. Scoliosis 1: 21. 63. Takahashi Y, Matsumoto M, Karasugi T, Watanabe K, Chiba K, Kawakami N, Tsuji T, Uno K, Suzuki T, Ito M, Sudo H, Minami S, Kotani T, Kono K, Yanagida H, Taneichi H, Takahashi A, Toyama Y, Ikegawa S (2011) Lack of association between adolescent

idiopathic scoliosis and previously reported single nucleotide polymorphisms in MATN1, MTNR1B, TPH1, and IGF1 in a Japanese population. J Orthop Res 29: 1055-1058. 64. Morcuende JA, Minhas R, Dolan L, Stevens J, Beck J, Wang K, Weinstein SL, Sheffield V (2003) Allelic variants of human melatonin 1A receptor in patients with familial adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 28: 2025-2028; discussion 2029. 65. Qiu XS, Tang NL, Yeung HY, Cheng JC, Qiu Y (2008) Lack of association between the promoter polymorphism of the MTNR1A gene and adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 33: 2204-2207. 66. Nelson LM, Ward K, Ogilvie JW (2011) Genetic variants in melatonin synthesis and signaling pathway are not associated with adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 36: 37-40. 67. Qiu XS, Tang NL, Yeung HY, Qiu Y, Qin L, Lee KM, Cheng JC (2006) The role of melatonin receptor 1B gene (MTNR1B) in adolescent idiopathic scoliosis--a genetic association study. Stud Health Technol Inform 123: 3-8. 68. Shyy W, Wang K, Gurnett CA, Dobbs MB, Miller NH, Wise C, Sheffield VC, Morcuende JA (2010) Evaluation of GPR50, hMel-1B, and ROR-alpha melatonin-related receptors and the etiology of adolescent idiopathic scoliosis. J Pediatr Orthop 30: 539-543. 69. Qiu XS, Tang NL, Yeung HY, Lee KM, Hung VW, Ng BK, Ma SL, Kwok RH, Qin L, Qiu Y, Cheng JC (2007) Melatonin receptor 1B (MTNR1B) gene polymorphism is associated with the occurrence of adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 32: 1748- 1753. 70. Inoue M, Minami S, Nakata Y, Kitahara H, Otsuka Y, Isobe K, Takaso M, Tokunaga M, Nishikawa S, Maruta T, Moriya H (2002) Association between estrogen receptor gene polymorphisms and curve severity of idiopathic scoliosis. Spine (Phila Pa 1976) 27: 2357-2362. 71. Wu J, Qiu Y, Zhang L, Sun Q, Qiu X, He Y (2006) Association of estrogen receptor gene polymorphisms with susceptibility to adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 31: 1131-1136. 72. Zhao D, Qiu GX, Wang YP, Zhang JG, Shen JX, Wu ZH (2009) Association between adolescent idiopathic scoliosis with double curve and polymorphisms of calmodulin1 gene/estrogen receptor-alpha gene. Orthop Surg 1: 222-230. 73. Tang NL, Yeung HY, Lee KM, Hung VW, Cheung CS, Ng BK, Kwok R, Guo X, Qin L, Cheng JC (2006) A relook into the association of the estrogen receptor [alpha] gene (PvuII, XbaI) and adolescent idiopathic scoliosis: a study of 540 Chinese cases. Spine (Phila Pa 1976) 31: 2463-2468. 74. Zhang HQ, Lu SJ, Tang MX, Chen LQ, Liu SH, Guo CF, Wang XY, Chen J, Xie L (2009) Association of estrogen receptor beta gene polymorphisms with susceptibility to adolescent idiopathic scoliosis. Spine (Phila Pa 1976) 34: 760-764. 75. Takahashi Y, Matsumoto M, Karasugi T, Watanabe K, Chiba K, Kawakami N, Tsuji T, Uno K, Suzuki T, Ito M, Sudo H, Minami S, Kotani T, Kono K, Yanagida H, Taneichi H, Takahashi A, Toyama Y, Ikegawa S (2011) Replication study of the association between

adolescent idiopathic scoliosis and two estrogen receptor genes. J Orthop Res 29: 834- 837. 76. Gao X, Gordon D, Zhang D, Browne R, Helms C, Gillum J, Weber S, Devroy S, Swaney S, Dobbs M, Morcuende J, Sheffield V, Lovett M, Bowcock A, Herring J, Wise C (2007) CHD7 gene polymorphisms are associated with susceptibility to idiopathic scoliosis. Am J Hum Genet 80: 957-965. 77. Vissers LE, van Ravenswaaij CM, Admiraal R, Hurst JA, de Vries BB, Janssen IM, van der Vliet WA, Huys EH, de Jong PJ, Hamel BC, Schoenmakers EF, Brunner HG, Veltman JA, van Kessel AG (2004) Mutations in a new member of the chromodomain gene family cause CHARGE syndrome. Nat Genet 36: 955-957. 78. Doyle C, Blake K (2005) Scoliosis in CHARGE: a prospective survey and two case reports. Am J Med Genet A 133A: 340-343. 79. Hadley-Miller N, Mims B, Milewicz DM (1994) The potential role of the elastic fiber system in adolescent idiopathic scoliosis. J Bone Joint Surg Am 76: 1193-1206. 80. Sakai LY, Keene DR, Engvall E (1986) Fibrillin, a new 350-kD glycoprotein, is a component of extracellular microfibrils. J Cell Biol 103: 2499-2509. 81. Zhang H, Apfelroth SD, Hu W, Davis EC, Sanguineti C, Bonadio J, Mecham RP, Ramirez F (1994) Structure and expression of fibrillin-2, a novel microfibrillar component preferentially located in elastic matrices. J Cell Biol 124: 855-863. 82. Dietz HC, Cutting GR, Pyeritz RE, Maslen CL, Sakai LY, Corson GM, Puffenberger EG, Hamosh A, Nanthakumar EJ, Curristin SM, et al. (1991) Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 352: 337-339. 83. Loeys BL, Dietz HC, Braverman AC, Callewaert BL, De Backer J, Devereux RB, Hilhorst- Hofstee Y, Jondeau G, Faivre L, Milewicz DM, Pyeritz RE, Sponseller PD, Wordsworth P, De Paepe AM (2010) The revised Ghent nosology for the Marfan syndrome. J Med Genet 47: 476-485. 84. Putnam EA, Zhang H, Ramirez F, Milewicz DM (1995) Fibrillin-2 (FBN2) mutations result in the Marfan-like disorder, congenital contractural arachnodactyly. Nat Genet 11: 456- 458. 85. Ramos Arroyo MA, Weaver DD, Beals RK (1985) Congenital contractural arachnodactyly. Report of four additional families and review of literature. Clin Genet 27: 570-581. 86. Viljoen D (1994) Congenital contractural arachnodactyly (Beals syndrome). J Med Genet 31: 640-643. 87. Callewaert BL, Loeys BL, Ficcadenti A, Vermeer S, Landgren M, Kroes HY, Yaron Y, Pope M, Foulds N, Boute O, Galan F, Kingston H, Van der Aa N, Salcedo I, Swinkels ME, Wallgren-Pettersson C, Gabrielli O, De Backer J, Coucke PJ, De Paepe AM (2009) Comprehensive clinical and molecular assessment of 32 probands with congenital contractural arachnodactyly: report of 14 novel mutations and review of the literature. Hum Mutat 30: 334-341. 88. Keane MG, Pyeritz RE (2008) Medical management of Marfan syndrome. Circulation 117: 2802-2813.

89. Faivre L, Collod-Beroud G, Loeys BL, Child A, Binquet C, Gautier E, Callewaert B, Arbustini E, Mayer K, Arslan-Kirchner M, Kiotsekoglou A, Comeglio P, Marziliano N, Dietz HC, Halliday D, Beroud C, Bonithon-Kopp C, Claustres M, Muti C, Plauchu H, Robinson PN, Ades LC, Biggin A, Benetts B, Brett M, Holman KJ, De Backer J, Coucke P, Francke U, De Paepe A, Jondeau G, Boileau C (2007) Effect of mutation type and location on clinical outcome in 1,013 probands with Marfan syndrome or related phenotypes and FBN1 mutations: an international study. Am J Hum Genet 81: 454-466. 90. Judge DP, Dietz HC (2005) Marfan's syndrome. Lancet 366: 1965-1976. 91. Dietz HC (1993) Marfan Syndrome. In: Pagon RA, Bird TD, Dolan CR, Stephens K, Adam MP, editors. GeneReviews. Seattle (WA). 92. Putnam EA, Cho M, Zinn AB, Towbin JA, Byers PH, Milewicz DM (1996) Delineation of the Marfan phenotype associated with mutations in exons 23-32 of the FBN1 gene. Am J Med Genet 62: 233-242. 93. Beroud C, Collod-Beroud G, Boileau C, Soussi T, Junien C (2000) UMD (Universal mutation database): a generic software to build and analyze locus-specific databases. Hum Mutat 15: 86-94. 94. Chen Y, Lei YP, Zheng HX, Wang W, Cheng HB, Zhang J, Wang HY, Jin L, Li H (2009) A novel mutation (C1425Y) in the FBN2 gene in a father and son with congenital contractural arachnodactyly. Genet Test Mol Biomarkers 13: 295-300. 95. Davis MR, Summers KM (2012) Structure and function of the mammalian fibrillin gene family: implications for human connective tissue diseases. Mol Genet Metab 107: 635- 647. 96. Ades LC, Sreetharan D, Onikul E, Stockton V, Watson KC, Holman KJ (2002) Segregation of a novel FBN1 gene mutation, G1796E, with kyphoscoliosis and radiographic evidence of vertebral dysplasia in three generations. Am J Med Genet 109: 261-270. 97. Milewicz DM, Grossfield J, Cao SN, Kielty C, Covitz W, Jewett T (1995) A mutation in FBN1 disrupts profibrillin processing and results in isolated skeletal features of the Marfan syndrome. J Clin Invest 95: 2373-2378. 98. Miller NH, Mims B, Child A, Milewicz DM, Sponseller P, Blanton SH (1996) Genetic analysis of structural elastic fiber and collagen genes in familial adolescent idiopathic scoliosis. J Orthop Res 14: 994-999. 99. Sharma S, Gao X, Londono D, Devroy SE, Mauldin KN, Frankel JT, Brandon JM, Zhang D, Li QZ, Dobbs MB, Gurnett CA, Grant SF, Hakonarson H, Dormans JP, Herring JA, Gordon D, Wise CA (2011) Genome-wide association studies of adolescent idiopathic scoliosis suggest candidate susceptibility genes. Hum Mol Genet 20: 1456-1466. 100. Holm J, Hillenbrand R, Steuber V, Bartsch U, Moos M, Lubbert H, Montag D, Schachner M (1996) Structural features of a close homologue of L1 (CHL1) in the mouse: a new member of the L1 family of neural recognition molecules. Eur J Neurosci 8: 1613-1629. 101. Wright AG, Demyanenko GP, Powell A, Schachner M, Enriquez-Barreto L, Tran TS, Polleux F, Maness PF (2007) Close homolog of L1 and neuropilin 1 mediate guidance of thalamocortical axons at the ventral telencephalon. J Neurosci 27: 13667-13679.

102. Jen JC, Chan WM, Bosley TM, Wan J, Carr JR, Rub U, Shattuck D, Salamon G, Kudo LC, Ou J, Lin DD, Salih MA, Kansu T, Al Dhalaan H, Al Zayed Z, MacDonald DB, Stigsby B, Plaitakis A, Dretakis EK, Gottlob I, Pieh C, Traboulsi EI, Wang Q, Wang L, Andrews C, Yamada K, Demer JL, Karim S, Alger JR, Geschwind DH, Deller T, Sicotte NL, Nelson SF, Baloh RW, Engle EC (2004) Mutations in a human ROBO gene disrupt hindbrain axon pathway crossing and morphogenesis. Science 304: 1509-1513. 103. Takahashi Y, Kou I, Takahashi A, Johnson TA, Kono K, Kawakami N, Uno K, Ito M, Minami S, Yanagida H, Taneichi H, Tsuji T, Suzuki T, Sudo H, Kotani T, Watanabe K, Chiba K, Hosono N, Kamatani N, Tsunoda T, Toyama Y, Kubo M, Matsumoto M, Ikegawa S (2011) A genome-wide association study identifies common variants near LBX1 associated with adolescent idiopathic scoliosis. Nat Genet 43: 1237-1240. 104. Fan YH, Song YQ, Chan D, Takahashi Y, Ikegawa S, Matsumoto M, Kou I, Cheah KS, Sham P, Cheung KM, Luk KD (2012) SNP rs11190870 near LBX1 is associated with adolescent idiopathic scoliosis in southern Chinese. J Hum Genet 57: 244-246. 105. Jiang H, Qiu X, Dai J, Yan H, Zhu Z, Qian B, Qiu Y (2013) Association of rs11190870 near LBX1 with adolescent idiopathic scoliosis susceptibility in a Han Chinese population. Eur Spine J 22: 282-286. 106. Gao W, Peng Y, Liang G, Liang A, Ye W, Zhang L, Sharma S, Su P, Huang D (2013) Association between common variants near LBX1 and adolescent idiopathic scoliosis replicated in the Chinese Han population. PLoS One 8: e53234. 107. Kou I, Takahashi Y, Johnson TA, Takahashi A, Guo L, Dai J, Qiu X, Sharma S, Takimoto A, Ogura Y, Jiang H, Yan H, Kono K, Kawakami N, Uno K, Ito M, Minami S, Yanagida H, Taneichi H, Hosono N, Tsuji T, Suzuki T, Sudo H, Kotani T, Yonezawa I, Londono D, Gordon D, Herring JA, Watanabe K, Chiba K, Kamatani N, Jiang Q, Hiraki Y, Kubo M, Toyama Y, Tsunoda T, Wise CA, Qiu Y, Shukunami C, Matsumoto M, Ikegawa S (2013) Genetic variants in GPR126 are associated with adolescent idiopathic scoliosis. Nat Genet 45: 676-679. 108. Janssen MM, de Wilde RF, Kouwenhoven JW, Castelein RM (2011) Experimental animal models in scoliosis research: a review of the literature. Spine J 11: 347-358. 109. Thillard MJ (1959) [Vertebral column deformities following epiphysectomy in the chick]. C R Hebd Seances Acad Sci 248: 1238-1240. 110. Kanemura T, Kawakami N, Deguchi M, Mimatsu K, Iwata H (1997) Natural course of experimental scoliosis in pinealectomized chickens. Spine (Phila Pa 1976) 22: 1563- 1567. 111. Cheung KM, Lu DS, Poon AM, Wang T, Luk KD, Leong JC (2003) Effect of melatonin suppression on scoliosis development in chickens by either constant light or surgical pinealectomy. Spine (Phila Pa 1976) 28: 1941-1944. 112. O'Kelly C, Wang X, Raso J, Moreau M, Mahood J, Zhao J, Bagnall K (1999) The production of scoliosis after pinealectomy in young chickens, rats, and hamsters. Spine (Phila Pa 1976) 24: 35-43.

113. Machida M, Saito M, Dubousset J, Yamada T, Kimura J, Shibasaki K (2005) Pathological mechanism of idiopathic scoliosis: experimental scoliosis in pinealectomized rats. Eur Spine J 14: 843-848. 114. Gorman KF, Tredwell SJ, Breden F (2007) The mutant guppy syndrome curveback as a model for human heritable spinal curvature. Spine (Phila Pa 1976) 32: 735-741. 115. Fjelldal PG, Grotmol S, Kryvi H, Gjerdet NR, Taranger GL, Hansen T, Porter MJ, Totland GK (2004) Pinealectomy induces malformation of the spine and reduces the mechanical strength of the vertebrae in Atlantic salmon, Salmo salar. J Pineal Res 36: 132-139. 116. McCarrey JR, Abbott UK, Benson DR, Riggins RS (1981) Genetics of scoliosis in chickens. J Hered 72: 6-10. 117. Schork NJ, Murray SS, Frazer KA, Topol EJ (2009) Common vs. rare allele hypotheses for complex diseases. Curr Opin Genet Dev 19: 212-219. 118. Anitha H, Prabhu GK (2012) Identification of apical vertebra for grading of idiopathic scoliosis using image processing. J Digit Imaging 25: 155-161. 119. Sale de Gauzy J, Jouve JL, Accadbled F, Blondel B, Bollini G (2011) Use of the Universal Clamp in adolescent idiopathic scoliosis for deformity correction and as an adjunct to fusion: 2-year follow-up. J Child Orthop 5: 273-282.

Chapter 2:

Copy Number Variation in Adolescent Idiopathic Scoliosis

SUMMARY

Adolescent idiopathic scoliosis (AIS) is a complex genetic disorder that causes spinal deformity in approximately 3% of the population. Candidate gene, linkage and genome-wide association studies have sought to identify genetic variation that predisposes individuals to AIS, but the genetic basis remains unclear. Copy number variants (CNVs) are associated with several isolated skeletal phenotypes, but their role in AIS has never been assessed. In this study, we determined the frequency of recurrent copy number rearrangements, chromosome aneuploidy and rare CNVs in AIS patients. We performed genomic copy number analysis on 143 patients with isolated AIS using the Affymetrix Genome-wide Human SNP Array 6.0. We identified a duplication of chromosome 1q21.1 in 2.1% (N=3/143) of AIS patients, which was enriched compared to 0.09% (N=1/1079) of controls (P=0.0057) and 0.07% (N=6/8329) of published controls (P=0.0004). Other notable findings include trisomy X, which was identified in 1.8%

(N=2/114) of female AIS patients, and rearrangements of chromosome 15q11.2 and 16p11.2 that may be relevant to scoliosis susceptibility. Finally, we report rare CNVs that will be of use to future studies investigating candidate genes for AIS. Overall, our results suggest that chromosomal microarray may reveal clinically useful abnormalities in some AIS patients.

Portions of this chapter are adapted from: Buchan JG, Alvarado DM, Haller G, Aferol H, Miller NH, Dobbs MB, Gurnett CA. (2014) “A Role for Copy Number Variation in Adolescent Idiopathic Scoliosis.” Clinical Orthopedics and Related Research. Accepted manuscript.

INTRODUCTION

Over the past decade, copy number variation has increasingly been recognized as a major

cause of disease, in what has come to be known as genomic disorders [1]. Genomic disorders

arise when deletions, duplication or inversions affect gene dosage or proximity to regulatory

elements such as enhancers [2]. The mutation rate for copy number variants (CNVs) is estimated

to be two to four magnitudes higher than for point mutations [3].

CNV analysis has been successfully applied to patients with intellectual disability, neuropsychiatric disorders and multiple congenital abnormalities, revealing multiple genes and

genomic regions contributing to disease susceptibility (reviewed in [4-6]). In comparison,

relatively few CNV studies have assessed patients with isolated skeletal phenotypes, though

CNVs have been associated with idiopathic short stature [7,8], idiopathic clubfoot [9-11], adult-

onset degenerative lumbar scoliosis [12] and bone mineral density [13]. Moreover, a recent

study showed that recurrent rearrangements of chromosome 16p11.2 are risk factors for non- idiopathic scoliosis and vertebral abnormalities in children with additional developmental, neurological and congenital abnormalities [14]. These results demonstrate the potential for recurrent and other rare CNVs to affected skeletal phenotypes, like scoliosis, and warrants evaluation of CNVs in patients with AIS.

In this study, we determine the frequency of copy number rearrangements, chromosome aneuploidy, and rare CNVs from a genome-wide CNV screen of 143 AIS patients and report an important role for CNVs in AIS pathogenesis.

METHODS

Patient samples

AIS patients were recruited from the Pediatric Orthopedic Surgery clinics at St. Louis

Children’s Hospital and St. Louis Shriners Hospital for Children. All patients had spinal curves

≥10° as measured with the Cobb method [15] on a standing radiograph. Patients with known or

suspected scoliosis etiologies (e.g. Marfan syndrome, congenital abnormalities) were excluded.

Growth parameters were calculated based on data from the National Center for Health Statistics

(http://www.cdc.gov/nchs/). Blood or saliva samples were collected for probands and available

relatives after obtaining informed consent. DNA isolations were performed using the DNA

Isolation Kit for Mammalian Blood (Roche, Indianapolis, IN, USA) or the Oragene Purifier

(DNA Genotek, Kanata, ON, Canada) according to the manufacturer’s instructions. Replication

cohorts for chromosome 1q21.1 (N=120) and trisomy X (N=172) included AIS patients recruited

from St. Louis Children’s Hospital, St. Louis Shriners Hospital for Children and the University

of Colorado. Human Subjects Committees at all three institutions approved this study.

CNV analysis

Copy number analysis was performed for 148 AIS probands and 1079 controls (666

controls from a bipolar disorder study [16] and 413 idiopathic clubfoot patients [11]) using the

Genome-wide Human SNP Array 6.0 (Affymetrix, Santa Clara, CA, USA). Copy number calls were generated with the Genotyping Console software (Affymetrix Affymetrix, Santa Clara, CA,

USA) using a reference set of 270 HapMap controls. AIS samples with contrast quality control

(QC) <0.04 and median absolute pairwise difference (MAPD) >0.35 or total CNVs >two

standard deviations above the average were excluded (N=5). Analysis of the remaining 143 AIS

samples was limited to CNVs ≥125kb in size with ≥50 markers and ≤10kb average distance

between markers. CNVs with >50% overlap with assembly gaps were removed. We identified

rare and novel CNVs by limiting our analysis to CNVs that had <50% overlap with CNVs at

>1% frequency or with all CNVs, respectively, in the Database of Genomic Variants (release

2013-07-23, http://projects.tcag.ca/variation) when ≥100 individuals were evaluated. To

evaluate known, clinically significant copy number disorders in AIS patients, we identified

CNVs that overlapped genomic regions associated with recurrent copy number disorders,

including 45 recurrent genomic disorder regions (<10 Mb) recently examined in a large control

population [17]. Quantitative PCR (qPCR) using ≥3 PCR primer pairs validated selected CNVs.

All sequence coordinates are reported using NCBI assembly build 36 (hg18).

RESULTS

To investigate whether CNVs contribute to AIS susceptibility, we evaluated 143 unrelated AIS patients for CNVs with the Affymetrix Genome-wide Human SNP Array 6.0. The

AIS cohort included primarily familial cases with moderate to severe AIS. The average spinal curve (Cobb angle) for patients in this cohort was 49° and 80% were female (Table 2.1). To

enrich our dataset with high confidence CNVs, our analysis was limited to CNVs ≥125 kb in size, identified by ≥50 markers with ≤10 kb average distance between markers.

Recurrent copy number rearrangements

Recurrent rearrangements of chromosome 16p11.2 are associated with scoliosis and other abnormalities [14], but this locus has never been evaluated for CNVs in a patient cohort with isolated scoliosis and the role of 16p11.2 and other recurrent CNVs in AIS remains unknown.

Therefore, we identified AIS patients with CNVs overlapping 45 loci associated with recurrent copy number disorders [17]. Seven patients (4.9%, N=7/143) in our AIS cohort had CNVs overlapping one of the recurrent genomic disorder regions. These CNVs caused rearrangements of chromosome 1q21.1 (N=3), chromosome 2q13 (N=1), chromosome 15q11.2 (N=2) and

chromosome 16p11.2 (N=1) (Table 2.2).

A duplication of chromosome 1q21.1 was the most frequent finding and was present in

2.1% (N=3/143) of AIS patients compared to 0.09% (N=1/1079) of controls (P=0.0057, one-

tailed Fisher’s exact test) and 0.07% (N=6/8329) of published controls (P=0.0004, one-tailed

Fisher’s exact test) [17]. After correcting for multiple testing, the enrichment of chromosome

1q21.1 duplications in AIS remained significant compared to published controls (P<0.001).

Large segmental duplication blocks mediate four recurrent breakpoints in chromosome 1q21.1

CNVs and define distinct proximal and distal regions, but only the proximal region was duplicated in all three cases (Figure 2.1). To test segregation of the proximal 1q21.1 duplication with AIS, available family members were evaluated for the presence of the CNV using qPCR.

The proximal 1q21.1 duplication segregated with reduced penetrance in all three pedigrees

(Figure 2.2). AIS ranged in severity from mild to severe in the five affected individuals with the

duplication (Table 2.3). Two carriers of the duplication were unaffected, although this was not

confirmed radiographically. Clinical information and family histories did not reveal evidence for

intellectual or developmental disability or other significant comorbidities in the individuals with

the duplication.

To screen for additional AIS patients with CNVs in chromosome 1q21.1, we used qPCR

to identify proximal 1q21.1 CNVs in an independent replication cohort (N=120), but no

additional chromosome 1q21.1 duplications were identified. However, even combined with the

Affymetrix Genome-wide Human SNP Array 6.0 results, the frequency of proximal 1q21.1

duplications in AIS patients remained enriched compared to controls from this study (N=3/263

vs. N=1/1079, P=0.0255, one-tailed Fisher’s exact test) and published controls (N=3/263 vs.

N=6/8329, P=0.0021, one-tailed Fisher’s exact test) [17].

Trisomy X

We identified large-scale aneuploidy in 1.4% (N=2/143) of AIS patients. Both patients

were female with trisomy of the X chromosome (47, XXX). Trisomy X was therefore present in

1.75% (N=2/114) of female AIS patients and was confirmed by qPCR. Both AIS patients had tall stature (>99th percentile), but no additional physical or developmental comorbidities, and

neither had previously been diagnosed with this condition. Only one of 529 females (0.19%) in our control cohort was identified with trisomy X, similar the 0.11% frequency observed by

Nielsen and Wohlert [18] in a large study of 17,038 newborn females. We screened an additional 172 female AIS patients for trisomy X by qPCR and no additional patients were identified. Thus, the overall frequency of trisomy X in AIS females was 0.7% (N=2/286).

Rare CNVs in AIS patients

To determine if the overall burden of rare CNVs differs between AIS patients and controls, we identified CNVs that occurred at <1% in the Database of Genomic Variants (<50% overlap with CNVs present at a frequency of >1% in the Database of Genomic Variants). We identified 177 rare CNVs in 92 AIS patients (Table 2.4). Of these CNVs, 118 were duplications,

59 were deletions and the average CNV size was 328 kb. The frequency of large (>750 kb), autosomal CNVs was similar between AIS patients (4.90%, N=7/143) and controls (5.65%,

N=61/1079). A complete list of rare CNVs identified in AIS patients is available in Table 2.5.

DISCUSSION

Here, we report the first study investigating copy number variation in a cohort of AIS

patients. In this study, 6.3% (N=9/143) of AIS patients were identified with chromosomal

aneuploidy (N=2/143) or a CNV affecting a region previously associated with a recurrent genomic disorder (N=7/143). Our findings illustrate how frequently clinically relevant CNVs will be encountered during microarray testing of patients with isolated scoliosis. Because the current study cohort was ascertained in an orthopedic clinic population and because patients with

obvious developmental and intellectual impairment were excluded from the study, the rate would likely have been much higher if patients with additional comorbidities were included. While several of the clinically significant CNVs identified in this study are unlikely to be related to

AIS, others may have a role in AIS pathogenesis or contribute more generally to multiple spinal phenotypes.

Chromosome 1q21.1 duplications were the most frequently observed chromosomal abnormality identified in AIS patients. Duplications of 1q21.1 were present in 2.1% of AIS

probands but were found in only 0.07-0.08% of healthy controls [17,19]. Chromosome 1q21.1 spans a complex 5.4 Mb region of which only 25% contains unique, non-duplicated sequence and approximately 700 kb of sequence is missing due to gaps in the NCBI assembly (build 36)

[20]. Segmental duplications make the region susceptible to recurrent rearrangements by nonallelic homologous recombination [21]. The most frequently observed breakpoints arise from four large segmental duplication blocks that define two stretches of unique sequence within chromosome 1q21.1: the proximal and distal regions located at genomic positions 144 to 144.34

Mb and 145 to 145.9 Mb, respectively. The distal region is associated with many phenotypes,

including autism [20,22-24], intellectual disability [20,22,25-27], schizophrenia [22,28-33],

microcephaly and macrocephaly [20,22], congenital heart defects [34-36] and abnormalities in

the renal and urinary tract [37]. In contrast, there are far fewer disease associations with the

proximal region. The strongest association is with thrombocytopenia with absent radii (TAR)

syndrome [38], which is frequently caused by compound inheritance of proximal 1q21.1 deletion

and a rare single nucleotide variant in RBM8A [39]. Other phenotypes where proximal

chromosome 1q21.1 duplications and deletions of varying sizes have been noted include

congenital heart defects [40-42], Mayer-Rokitansky-Küster-Hauser syndrome [43], autism [44],

fetal urogenital abnormalities [45] and additional diverse phenotypes [19,46]. Although the three

AIS patients in our cohort had different chromosome 1q21.1 duplication breakpoints, only the

proximal region was common to all three individuals and no individuals reported additional phenotypes previously associated with chromosome 1q21.1 rearrangements. Duplications of

1q21.1 segregated with AIS in small families with incomplete penetrance, but incomplete penetrance has also been previously noted in TAR syndrome [38] and other phenotypes [19,47].

Because additional chromosome 1q21.1 duplications were not detected after screening an additional 120 AIS patient samples, proximal duplications are likely rarer in AIS than estimated from our original cohort of 143 patients. Nevertheless, the combined dataset of 263 patients still showed a significant enrichment of proximal 1q21.1 duplications in AIS.

Scoliosis has not been previously described in patients with proximal chromosome

1q21.1 duplications, though the overall number of patients with this genomic abnormality is relatively small. Scoliosis was not reported in seventeen patients with proximal duplications, although two patients had other spine phenotypes (lumbar lordosis and lumbosacral hyperlordosis) [19]. Of thirty-four patients with proximal deletions, three had a spinal phenotype

(scoliosis, kyphosis and C6-C7 vertebral fusion) [19]. Although previous studies of patients with proximal chromosome 1q21.1 duplications suggest that scoliosis is not a commonly associated

phenotype, scoliosis is often not detected until adolescence and is therefore likely to be

underreported in younger patient cohorts. Furthermore, mild scoliosis may be underreported in

children with multiple congenital anomalies whose other major medical problems are of greater

concern. Additional larger studies are needed to determine the importance of proximal chromosome 1q21.1 duplications in the etiology of AIS.

Several recurrent CNVs identified in our AIS patient cohort are primarily associated with

cognitive impairment or other related phenotypes, but may also contribute to scoliosis and other

spinal phenotypes. Hemivertebrae were present in 20% (N=2/10) of patients with chromosome

15q11.2 duplications, although none of 15 patients with the reciprocal deletion had vertebral

abnormalities [48]. Two patients in our AIS cohort had chromosome 15q11.2 CNVs.

Chromosome 15q11.2 deletions are strongly associated with developmental delay [17], but the

reciprocal duplication is not and the significance and associated phenotypes of 15q11.2 duplications, if any, are not well established. Neurocognitive or developmental phenotypes were

not present in either of our AIS patients with the 15q11.2 CNVs.

Likewise, a variety of spinal abnormalities, including hemivertebrae, syringomyelia and scoliosis have been described in patients with rearrangements of chromosome 16p11.2 [14,49-

54] and this locus was hypothesized to be a risk factor for idiopathic scoliosis [14]. We

identified a chromosome 16p11.2 duplication in one AIS patient with spina bifida occulta. This

patient did not have intellectual or developmental disability, which is consistent with the large

phenotypic variability and reduced penetrance observed in individuals with chromosome

16p11.2 duplications [55]. Interestingly, chromosome 16 was previously associated with

idiopathic scoliosis by linkage analysis in 202 families [56] and fine-mapping linkage analysis of

544 individuals from an additional 95 families narrowed the association to two regions on chromosome 16, which included 16p11.2 [57], providing further support for an association of the

16p11.2 locus with AIS.

We identified trisomy of the X chromosome (47, XXX), also called triple X syndrome, in

0.7% (N=2/286) female AIS patients. Triple X syndrome is estimated to occur in 1 of 1,000 female births and generally causes mild phenotypes, such as premature ovarian failure and learning disabilities, which results in many patients being undiagnosed [58]. Scoliosis is well known to occur more frequently in triple X syndrome than in the general population [59-61].

Specifically, Olanders (1977) described scoliosis in 15.2% (N=5/33) of patients with triple X syndrome [61]. Patients with triple X syndrome are also typically taller than average, with the average height being >80th percentile by the age of fourteen [62]. Both AIS patients with

trisomy X in our series were tall (>99th percentile for height) but reported no additional

phenotypes. Testing for triple X syndrome should be considered in tall females with AIS so that other associated phenotypes, such as premature ovarian failure and developmental delay, can be

monitored.

There are several limitations to our study, including the relatively small sample size. For

the rare CNVs discussed here, a larger sample size would have improved frequency estimates

and provided more statistical power. Moreover, in many cases additional family members were

unavailable for testing and limited our ability to provide additional support for causality.

Second, controls used in this study included a large number of patients with idiopathic clubfoot

and CNVs have already been associated with this phenotype [9-11]. However, we also reference

the frequency of recurrent CNVs in an even larger published independent control dataset [17] and these frequencies are very similar to our control dataset.

In summary, this study reports the first analysis of copy number variation in a large cohort of patients with AIS. Our data shows that over 6% of AIS patients harbor a clinically significant copy number abnormality and many of these are likely risk factors for scoliosis and other spinal phenotypes. Our results suggest that microarray analysis for CNVs may be a useful clinical test in this patient population. Finally, we provide a list of all rare CNVs identified in

AIS patients, which will be useful for future studies evaluating candidate genes for AIS.

Figure 2.1: Proximal chromosome 1q21.1 duplications identified in AIS patients Log2 ratios for three AIS probands identified with CNVs at chromosome 1q21.1 show the location of the duplications (shaded). The four recurrent breakpoints (BP1-BP4) and approximate locations of the proximal and distal regions are shown. Image modified from Genome Browser (http://genome.ucsc.edu). All genomic coordinates are shown for NCBI assembly build 36 (hg18).

Figure 2.2: Segregation of chromosome 1q21.1 duplications Duplications segregated in AIS probands (arrows) and family members with reduced penetrance. Dup, duplication; -, wild type.

Table 2.1: Demographics of 143 AIS probands analyzed for copy number variation Familial Non-familial Total AIS cases 134 (94%) 9 (6%) 143 Male 29 (100%) 0 (0%) 29 (20%) Female 105 (92%) 9 (8%) 114 (80%) Cobb angle Mean 49° (n=122) 50° (n=9) 49° (n=131) Range 10-97° (n=122) 26-74° (n=9) 10-97° (n=131)

Table 2.2: Frequency of 45 genomic copy number disorder regions in AIS patients Controls (published) AIS patients (this study) Controls (this study) Genomic [17] Size Abnormality disorder (kb) Size region (Mb) Patient(s) Phenotype(s) CNV (Mb) Frequency Frequency p-value Frequency p-value (kb)

1q21.1 6041-001 AIS chr1:142.72-144.97 2241 chr1: 2.10% 0.09% 0.07% duplication 340 6267-001 AIS chr1:144.08-144.50 420 0.0057 0.0004 144-144.34 (n=3/143) (n=1/1079) (n=6/8329) (proximal) 6035-004 AIS chr1:144.08-144.46 379

2q13 chr2: 0.70% 0.65% 0.38% 160 6126-001 AIS chr2:110.13-110.52 394 0.6316 0.4304 duplication 110.18-110.34 (n=1/143) (n=7/1079) (n=32/8329)

15q11.2 chr15: 0.70% 0.37% 0.23% 290 6053-001 AIS chr15:19.09-21.04 1949 0.4639 0.2888 deletion 20.35-20.64 (n=1/143) (n=4/1079) (n=19/8329) 49 15q11.2 chr15: 0.70% 0.46% 0.43% 290 6036-001 AIS chr15:20.22-20.97 744 0.5269 0.4681 duplication 20.35-20.64 (n=1/143) (n=5/1079) (n=36/8329)

16p11.2 chr16: AIS, spina 0.70% 0.19% 0.02% 550 6032-001 chr16:29.50-30.09 587 0.3118 0.0498 duplication 29.56-30.11 bifida occulta (n=1/143) (n=2/1079) (n=2/8329)

Genomic coordinates are reported using NCBI assembly build 36 (hg18)

Table 2.3: Clinical information for individuals identified with chromosome 1q21.1 duplications 1q21.1 Affected Family Individual Sex Curve type Treatment status (Cobb angle) 6035 004* F Dup/- Yes (15°) NA None 002 F Dup/- Yes, mild NA None 001 F NA Yes (56°) Right thoracic Surgery 6041 001* M Dup/- Yes (31°) Right thoracic Brace 002 F Dup/- Yes NA Surgery 003 F Dup/- No Not affected Not affected 6267 001* M Dup/- Yes (44°) Right thoracic Surgery 002 F -/- No Not affected Not affected 003 M Dup/- No Not affected Not affected Abbreviations: NA, not available; F, female; M, male; Dup, duplication; -, wild type *Proband

Table 2.4: Summary of rare CNVs (<1% frequency in DGV) identified in AIS patients AIS patients 143 AIS patients with rare CNVs 92 Total rare CNVs 177 Rare CNVs containing Refseq genes 130 Average CNV size 328 kb Median CNV size 217 kb Duplications 118 Deletions 59 Large autosomal CNVs (>750 kb) 7 AIS patients with large CNVs 4.90% In-house controls with large CNVs 5.65% Abbreviations: DGV, Database of Genomic Variants; CNV, copy number variant

Table 2.5: Rare CNVs identified in AIS cases # of Distance Patient Position Size CNV marke between Genes (Haploinsufficiency index [63]) ID (hg18) (kb) rs markers chr1:59061626- 6205-001 Dup 134 105 1 LOC100131060 (NA), 59195775 chr1:102428272- 6278-001 Del 197 152 1 MIR548AI (NA), 102625426 chr1:119724173- HAO2 (93.7%), HSD3B2 (NA), HSD3B1 (69.0%), HSD3BP4 6062-001 Dup 232 232 1 119956321 (NA), LINC00622 (NA), LOC100288142 (NA), LINC00623 (NA), LOC728875 (NA), PPIAL4B (NA), PPIAL4C (NA), PPIAL4A (NA), NBPF12 (NA), PFN1P2 (NA), NBPF8 (NA), NBPF9 (NA), LOC653513 (NA), PDE4DIP (NA), SEC22B (NA), NOTCH2NL (NA), HFE2 (8.3%), TXNIP (46.2%), POLR3GL (66.1%), chr1:142724939- 6041-001 Dup 2241 425 5 ANKRD34A (NA), LIX1L (27.4%), RBM8A (33.3%), 144965462 GNRHR2 (NA), PEX11B (35.1%), ITGA10 (21.5%), ANKRD35 (63.3%), PIAS3 (9.2%), NUDT17 (43.5%), POLR3C (32.7%), RNF115 (NA), CD160 (88.1%), PDZK1 (39.5%), GPR89A (NA), GPR89C (NA), PDZK1P1 (NA), LOC728989 (NA), PPIAL4A (NA), NBPF8 (NA), NBPF9 (NA), HFE2 (8.3%), TXNIP (46.2%), POLR3GL (66.1%), ANKRD34A (NA), chr1:144083908- LIX1L (27.4%), RBM8A (33.3%), GNRHR2 (NA), PEX11B 6035-004 Dup 379 162 2 144463196 (35.1%), ITGA10 (21.5%), ANKRD35 (63.3%), PIAS3 (9.2%), NUDT17 (43.5%), POLR3C (32.7%), RNF115 (NA), CD160 (88.1%), PDZK1 (39.5%), PPIAL4A (NA), NBPF8 (NA), NBPF9 (NA), HFE2 (8.3%), TXNIP (46.2%), POLR3GL (66.1%), ANKRD34A (NA), chr1:144083908- LIX1L (27.4%), RBM8A (33.3%), GNRHR2 (NA), PEX11B 6267-001 Dup 420 166 3 144503409 (35.1%), ITGA10 (21.5%), ANKRD35 (63.3%), PIAS3 (9.2%), NUDT17 (43.5%), POLR3C (32.7%), RNF115 (NA), CD160 (88.1%), PDZK1 (39.5%), GPR89A (NA), chr1:144337051- PPIAL4A (NA), NBPF8 (NA), NBPF9 (NA), RNF115 (NA), 6194-001 Dup 126 53 2 144463196 CD160 (88.1%), PDZK1 (39.5%), chr1:144337051- PPIAL4A (NA), NBPF8 (NA), NBPF9 (NA), RNF115 (NA), 6010-001 Dup 166 57 3 144503409 CD160 (88.1%), PDZK1 (39.5%), GPR89A (NA), PPIAL4A (NA), NBPF8 (NA), NBPF9 (NA), GPR89C (NA), PDZK1P1 (NA), ACP6 (82.8%), GJA5 (83.7%), GJA8 (NA), GPR89B (NA), NBPF11 (NA), NBPF24 (NA), NBPF10 (NA), chr1:145574884- 6041-001 Dup 1983 400 5 MIR5087 (NA), LOC100130000 (NA), FLJ39739 (NA), 147558211 PPIAL4D (NA), PPIAL4E (NA), PPIAL4F (NA), NBPF14 (NA), NBPF15 (NA), NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), LOC388692 (NA), chr1:146797048- 6180-001 Dup 552 57 10 NBPF15 (NA), NBPF16 (NA), LOC645166 (NA), 147349249 chr1:146928878- 6211-001 Del 462 68 7 NBPF16 (NA), LOC645166 (NA), 147390822 chr1:146928878- 6221-001 Del 490 73 7 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147419014 chr1:146928878- 6241-001 Del 490 73 7 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147419014 chr1:146928878- 6293-001 Del 490 73 7 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147419014 chr1:146928878- 6193-006 Del 528 87 6 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147456930 chr1:146928878- 6178-001 Del 540 89 6 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147469205 chr1:146929009- 6053-001 Del 528 84 6 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147456930 chr1:147017157- 6322-001 Del 402 66 6 NBPF16 (NA), LOC645166 (NA), NBPF23 (NA), 147419014 chr1:178525369- 6339-001 Del 132 60 2 ACBD6 (25.1%), 178657666 chr1:245244567- 6044-001 Del 132 70 2 ZNF670 245377000 chr2:1395887- 6068-001 Dup 164 110 1 TPO (51.9%), 1559511

chr2:49389260- 6240-001 Del 125 112 1 49514365 chr2:50774202- 6237-001 Del 384 255 2 NRXN1 (NA), 51158243 chr2:52573819- 6136-004 Dup 131 117 1 52705065 chr2:87026226- RGPD2 (NA), RGPD1 (NA), PLGLB1 (NA), PLGLB2 (NA), 6028-001 Dup 276 53 5 87302219 LOC285074 (NA), MIR4771 TBC1D8 (64.4%), CNOT11 (NA), SNORD89 (NA), RNF149 chr2:101027972- 6206-001 Dup 953 518 2 (82.4%), MIR5696 (NA), CREG2 (82.4%), RFX8 (NA), 101981034 MAP4K4 (21.9%), LOC100506328 (NA), IL1R2 (12.8%), chr2:184585902- 6339-001 Del 413 332 1 184999185 chr2:236540644- 6157-001 Dup 155 86 2 AGAP1 (NA), 236695522 chr2:238308001- 6062-001 Del 172 110 2 LRRFIP1 (80.0%), RBM44 (NA), RAMP1 (75.1%), 238479991 chr3:537642- 6314-001 Dup 344 228 2 882120 chr3:21204662- 6157-001 Del 126 89 1 21330495 chr3:35780275- 6029-001 Dup 136 79 2 ARPP21 (NA), 35916762 chr3:35780275- 6304-001 Dup 136 79 2 ARPP21 (NA), 35916762 chr3:50406198- 6306-001 Dup 186 86 2 CACNA2D2 (23.2%), C3orf18 (36.2%), HEMK1 (96.6%), 50592426 chr3:126907150- 6209-001 Dup 196 77 3 MIR548I1 (NA), 127103540 chr3:126907150- 6306-001 Dup 205 78 3 MIR548I1 (NA), 127112518 chr3:126907150- 6303-001 Dup 241 86 3 MIR548I1 (NA), FAM86JP (NA), ALG1L (NA), 127147683 chr3:126929342- 6050-001 Dup 174 72 2 MIR548I1 (NA), 127103540 chr3:143303623- 6213-001 Dup 266 118 2 TFDP2 (2.9%), GK5 (77.4%), XRN1 (11.8%), 143569349 chr3:143307337- 6050-001 Dup 262 116 2 TFDP2 (2.9%), GK5 (77.4%), XRN1 (11.8%), 143569349 chr3:162898796- 6327-001 Del 133 98 1 163031671 chr3:164396601- 6081-001 Del 437 197 2 CT64 (NA), 164833278 chr4:3840848- 6306-001 Dup 416 104 4 FAM86EP (NA), OTOP1 (84.7%), 4256969 chr4:3992159- 6303-001 Dup 244 69 4 FAM86EP (NA), 4236511 chr4:3992159- 6238-001 Dup 252 75 3 FAM86EP (NA), OTOP1 (84.7%), 4243961 chr4:4060245- 6163-001 Dup 195 75 3 OTOP1 (84.7%), 4254791 chr4:4066344- 6206-001 Dup 170 56 3 4236387 chr4:69054586- 6305-001 Dup 421 68 6 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69475447 chr4:69054586- 6233-001 Dup 538 71 8 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69592854 chr4:69054586- 6306-001 Dup 538 71 8 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69592854 chr4:69057524- 6126-001 Dup 418 67 6 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69475447 chr4:69057524- 6194-001 Dup 461 68 7 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69518358 chr4:69057524- 6303-001 Dup 535 70 8 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69592854 chr4:69057524- 6317-001 Dup 535 70 8 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69592854 chr4:69096355- 6206-001 Dup 496 56 9 UGT2B17 (98.0%), UGT2B15 (NA), TMPRSS11E (56.4%), 69592854

chr4:127358793- 6085-001 Del 145 100 1 127503594 chr4:131801982- 6303-001 Dup 210 120 2 132011871 chr4:135143484- 6217-001 Del 262 126 2 PABPC4L (NA), 135405069 chr4:135190819- 6293-001 Del 214 98 2 PABPC4L (NA), 135405069 FRG1 (NA), LOC283788 (NA), FRG2 (NA), LOC100288255 chr4:191028537- (NA), DUX4L2 (NA), DUX4L3 (NA), DUX4L4 (NA), 6306-001 Dup 226 58 4 191254119 DUX4L5 (NA), DUX4L6 (NA), DUX2 (NA), DUX4L7 (NA), LOC100653046 (NA), chr5:97281278- 6219-001 Del 174 91 2 97455710 chr5:115241061- AP3S1 (NA), AQPEP (NA), ARL14EPL (NA), COMMD10 6010-001 Dup 210 174 1 115450833 (46.9%), chr5:137227123- 6094-001 Dup 134 59 2 MYOT (18.2%), PKD2L2 (33.4%), FAM13B (NA), 137361288 chr6:1388098- 6259-001 Dup 246 208 1 FOXC1 (9.7%), GMDS (54.4%), 1633999 chr6:4200744- 6018-001 Del 216 213 1 4416634 chr6:57431738- 6197-001 Dup 350 157 2 PRIM2 (4.1%), 57781441 chr6:67005222- 6187-001 Del 488 387 1 67492828 chr7:4321584- 6201-001 Del 180 171 1 4501323 chr7:9014463- 6303-001 Dup 345 296 1 9359327 chr7:57593474- 6238-001 Dup 143 56 3 57736164 chr7:63726630- 6094-001 Dup 138 70 2 ZNF107 (NA), 63864919 chr7:63777655- 6300-001 Del 135 56 2 ZNF107 (NA), ZNF138 (NA), 63912674 chr7:100753806- 6171-001 Dup 170 74 2 COL26A1 (NA), 100924034 chr7:123003085- 6289-001 Dup 168 100 2 ASB15 (75.4%), LMOD2 (NA), WASL (4.2%), 123170697 CCDC136 (79.2%), FLNC (34.3%), ATP6V1F (NA), chr7:128220854- 6138-001 Dup 179 71 3 LOC100130705 (NA), KCP (NA), IRF5 (73.9%), TNPO3 128399562 (15.8%), chr7:142001612- 6086-001 Del 134 89 2 MTRNR2L6 (NA), 142135250 chr8:11101956- MTMR9 (82.9%), SLC35G5 (NA), TDH (NA), C8orf12 (NA), 6178-001 Dup 259 214 1 11360722 FAM167A (NA), chr8:39740335- ADAM2 (53.0%), IDO1 (NA), IDO2 (NA), C8orf4 (43.8%), 6337-001 Dup 1482 1129 1 41221920 ZMAT4 (22.1%), chr8:53552367- 6125-001 Dup 343 201 2 FAM150A (NA), RB1CC1 (9.0%), 53895567 chr8:53571918- FAM150A (NA), RB1CC1 (9.0%), NPBWR1 (78.3%), OPRK1 6337-001 Dup 1105 745 1 54677369 (69.1%), chr8:53583779- 6138-001 Dup 384 223 2 FAM150A (NA), RB1CC1 (9.0%), 53968156 chr9:28176821- 6040-001 Del 161 135 1 LINGO2 (4.3%), 28337679 chr9:89592787- CTSLP8 (NA), LOC392364 (NA), SPATA31E1 (NA), 6304-001 Dup 151 94 2 89744027 SPATA31C1 (NA), chr9:119155660- 6171-001 Dup 1190 920 1 ASTN2 (NA), TLR4 (49.1%), 120345735 chr10:35210421- 6128-002 Dup 146 85 2 CUL2 (10.0%), 35356001 chr10:67724031- 6223-001 Del 156 95 2 CTNNA3 (62.1%), 67879925 chr10:68037668- 6219-001 Del 174 142 1 CTNNA3 (62.1%), 68211255 chr10:81775965- 6211-001 Dup 187 124 2 TMEM254 81962789

chr10:93658958- 6044-001 Dup 164 76 2 FGFBP3 (NA), BTAF1 (2.3%), CPEB3 (9.7%), 93822581 chr10:10708990 6233-001 Dup 217 125 2 0-107306982 chr10:10709213 6053-001 Dup 215 124 2 2-107306982 chr10:12836147 6341-001 Del 198 170 1 9-128559804 chr11:5023934- 6026-001 Del 147 127 1 OR52J3 (NA), OR52E2 (NA), OR52A5 (NA), OR52A1 (NA), 5170989 chr11:49601437- 6330-001 Dup 207 112 2 LOC440040 (NA), 49808604 chr11:55442375- OR5I1 (95.5%), OR10AG1 (NA), OR7E5P (NA), OR5F1 (NA), 6085-001 Del 140 106 1 55582755 OR5AS1 (NA), chr11:71029775- 6303-001 Dup 171 51 3 FAM86C1 (NA), ALG1L9P (NA), 71201199 chr11:71029775- 6306-001 Dup 176 53 3 FAM86C1 (NA), ALG1L9P (NA), ZNF705E (NA), 71205347 chr11:71029775- 6206-001 Dup 185 56 3 FAM86C1 (NA), ALG1L9P (NA), ZNF705E (NA), 71214508 chr12:71517646- 6275-001 Del 136 89 2 71654142 chr12:89419571- 6038-001 Del 236 129 2 89655818 chr12:98539959- 6157-001 Del 169 97 2 ANKS1B (51.7%), FAM71C (95.3%), 98709344 chr12:11275241 6316-002 Dup 294 290 1 RBM19 (42.2%), 6-113046826 chr12:13053353 6277-001 Dup 177 181 1 5-130710089 chr13:17943628- 6121-001 Dup 274 60 5 LINC00417 (NA), 18217635 chr13:17943628- 6205-001 Del 324 70 5 LINC00417 (NA), 18267580 chr13:18138676- 6303-001 Dup 178 77 2 LINC00417 (NA), ANKRD20A9P (NA), 18316876 chr13:18184856- 6317-001 Dup 125 59 2 LINC00417 (NA), ANKRD20A9P (NA), 18310275 chr13:19054403- 6180-001 Dup 326 152 2 MPHOSPH8 (75.4%), PSPC1 (65.3%), ZMYM5 (93.2%), 19380872 chr13:33887795- 6337-001 Dup 148 107 1 LINC00457 (NA), 34035511 chr13:54016684- 6206-001 Del 357 211 2 54374043 chr13:75241901- 6306-001 Dup 152 157 1 LMO7 (89.7%), C13orf45 (NA), 75394225 chr14:21524127- 6073-001 Del 507 438 1 22030660 chr14:21555352- 6330-001 Del 395 344 1 21950283 chr14:21575148- 6121-001 Del 455 389 1 22029666 chr14:21656002- 6171-001 Del 130 135 1 21786451 chr14:21656002- 6036-001 Del 352 296 1 22008051 chr14:21660717- 6085-001 Del 360 312 1 22020331 chr14:21664286- 6086-001 Del 353 305 1 22017122 chr14:21668999- 6044-001 Del 132 141 1 21801251 chr14:21697326- 6216-001 Del 253 215 1 21950283 chr14:21697326- 6234-001 Del 323 266 1 22020331 chr14:21711505- 6033-001 Del 253 204 1 21964093 chr14:21721254- 6035-004 Del 250 198 1 21970760

chr14:21721254- 6012-001 Del 287 218 1 22008051 chr14:21817588- 6337-001 Del 133 87 2 21950283 chr14:42027983- 6041-001 Dup 383 244 2 42411046 chr14:60411971- 6073-001 Del 178 82 2 MNAT1 (3.1%), TRMT5 (NA), SLC38A6 (95.6%), 60589931 EVL (7.5%), MIR342 (NA), DEGS2 (NA), YY1 (4.4%), chr14:99592280- 6179-001 Dup 521 246 2 SLC25A29 (NA), MIR345 (NA), SLC25A47 (NA), WARS 100113598 (48.9%), WDR25 (58.5%), BEGAIN (NA), chr15:31753942- RYR3 (37.6%), AVEN (90.0%), CHRM5 (44.2%), EMC7 6068-001 Dup 491 389 1 32244695 (NA), PGBD4 (NA), KATNBL1 (NA), chr15:43934913- 6234-001 Del 136 107 1 44070720 chr16:4971083- 6271-003 Del 317 188 2 SEC14L5 (NA), NAGPA (85.5%), 5288139 chr16:6082024- 6218-001 Dup 203 229 1 6284589 chr16:14678121- PLA2G10 (71.6%), NPIPA2 (NA), NPIPA3 (NA), ABCC6P2 6317-001 Dup 356 82 4 15034391 (NA), NOMO1 (NA), MIR3179 chr16:14684882- PLA2G10 (71.6%), NPIPA2 (NA), NPIPA3 (NA), ABCC6P2 6077-001 Dup 317 75 4 15002197 (NA), NOMO1 (NA), MIR3179 chr16:14684882- PLA2G10 (71.6%), NPIPA2 (NA), NPIPA3 (NA), ABCC6P2 6306-001 Dup 317 75 4 15002197 (NA), NOMO1 (NA), MIR3179 chr16:14796084- 6106-001 Dup 189 52 4 ABCC6P2 (NA), NOMO1 (NA), MIR3179 14984969 chr16:14796084- 6298-001 Dup 199 64 3 ABCC6P2 (NA), NOMO1 (NA), MIR3179 14994666 chr16:14803686- 6238-001 Dup 184 55 3 ABCC6P2 (NA), NOMO1 (NA), MIR3179 14987944 chr16:14803686- 6028-001 Dup 191 63 3 ABCC6P2 (NA), NOMO1 (NA), MIR3179 14994666 chr16:14846829- 6322-001 Dup 141 52 3 NOMO1 (NA), MIR3179 14987944 chr16:14846829- 6221-001 Dup 155 66 2 NOMO1 (NA), MIR3179 15002197 chr16:14846829- 6233-001 Dup 155 66 2 NOMO1 (NA), MIR3179 15002197 chr16:14888466- 6126-001 Del 135 64 2 NOMO1 (NA), MIR3179 15023746 chr16:16540862- 6305-001 Dup 472 198 2 17012694 SLC7A5P1 (NA), SPN (NA), QPRT (NA), C16orf54 (NA), chr16:29498829- 6032-001 Dup 587 198 3 ZG16 (NA), KIF22 (NA), MAZ (32.2%), PRRT2 (41.6%), 30085908 PAGR1 (NA), MVP (69.2%), CDIPT (69.0%), CDIPT chr16:34645472- 6306-001 Dup 187 52 4 34832848 chr16:73926386- 6301-001 Del 155 75 2 CFDP1 (13.2%), TMEM170A (NA), CHST6 (70.7%), 74080901 chr17:692289- 6073-001 Dup 146 82 2 NXN (37.4%), 837961 chr17:15811223- ADORA2B (87.8%), ZSWIM7 (NA), TTC19 (50.4%), NCOR1 6271-003 Dup 223 106 2 16033881 (16.5%), chr17:20166343- CCDC144CP (NA), LGALS9B (NA), KRT16P3 (NA), 6306-001 Dup 418 83 5 20584318 CDRT15L2 (NA), LOC100287072 (NA), chr17:21281893- 6161-001 Dup 195 87 2 C17orf51 (NA), 21476920 chr17:21281893- 6269-001 Dup 213 94 2 C17orf51 (NA), 21495308 ASIC2 (NA), CCL2 (26.0%), CCL7 (95.8%), CCL11 (99.6%), chr17:28981264- 6038-001 Dup 981 937 1 CCL8 (96.3%), CCL13 (99.2%), CCL1 (99.9%), C17orf102 29962135 (NA), TMEM132E (33.5%), chr17:51455471- 6123-001 Dup 217 221 1 ANKFN1 (27.1%), 51672518 chr18:14103916- 6306-001 Dup 237 52 5 ZNF519 (NA), ANKRD20A5P (NA), CYP4F35P (NA), 14341363 chr18:14139471- 6300-001 Dup 249 60 4 ANKRD20A5P (NA), CYP4F35P (NA), 14388602

chr18:60260938- 6033-001 Dup 159 92 2 60420239 chr19:12435345- 6306-001 Del 126 54 2 ZNF709 (NA), ZNF564 (NA), ZNF490 (54.9%), 12561674 chr21:34646609- KCNE2 (89.8%), SMIM11 (NA), KCNE1 (76.7%), RCAN1 6341-001 Dup 177 102 2 34823160 (NA), chrX:2965778- 6218-001 Dup 193 181 1 ARSF (76.4%), 3158320 chrX:6465151- HDHD1 (NA), MIR4767 (NA), STS (45.2%), VCX (NA), 6028-001 Dup 1628 1213 1 8093257 PNPLA4 (64.3%), MIR651 (NA), chrX:7227969- 6036-001 Dup 185 135 1 STS (45.2%), 7412657 chrX:7373945- 6233-001 Dup 408 308 1 VCX (NA), 7782233 chrX:7799976- 6316-002 Dup 295 245 1 PNPLA4 (64.3%), MIR651 (NA), 8094650 chrX:7812245- 6233-001 Dup 280 231 1 PNPLA4 (64.3%), MIR651 (NA), 8092678 chrX:9408122- TBL1X (NA), GPR143 (79.8%), SHROOM2 (70.2%), 6202-001 Dup 589 505 1 9997284 LOC100288814 (NA), WWC3 (NA), LOC550643 (NA), UQCRBP1 (NA), SPIN3 (NA), SPIN2B chrX:56748915- 6280-001 Del 1742 806 2 (43.4%), SPIN2A (NA), FAAH2 (62.4%), ZXDB (NA), ZXDA 58490459 (NA), chrX:57386443- 6094-001 Dup 131 68 2 FAAH2 (62.4%), 57517736 chrX:58298891- 6209-001 Dup 198 56 4 58497269 chrX:61645542- 6026-001 Dup 428 68 6 62073324 chrX:61645542- 6278-001 Dup 467 82 6 62112687 chrX:65616302- 6031-001 Dup 488 241 2 EDA2R (NA), 66104307 chrX:65678294- 6274-001 Dup 172 88 2 EDA2R (NA), 65850037 chrX:89609426- 6328-001 Dup 222 60 4 89831490 chrX:92273794- 6077-001 Dup 450 285 2 92723465 chrY:14608287- 6098-001 Dup 199 120 2 VCY (NA), 14807693 chrY:25889736- 6306-001 Dup 177 58 3 GOLGA2P2Y (NA), GOLGA2P3Y (NA), CSPG4P1Y (NA), 26066330 Abbreviations: Dup, duplication; Del, deletion; CNV, copy number variant; ID, identifier

REFERENCES

1. Lupski JR (1998) Genomic disorders: structural features of the genome can lead to DNA rearrangements and human disease traits. Trends Genet 14: 417-422. 2. Stankiewicz P, Lupski JR Structural variation in the human genome and its role in disease. Annu Rev Med 61: 437-455. 3. Lupski JR (2007) Genomic rearrangements and sporadic disease. Nat Genet 39: S43-47. 4. Cook EH, Jr., Scherer SW (2008) Copy-number variations associated with neuropsychiatric conditions. Nature 455: 919-923. 5. Morrow EM (2010) Genomic copy number variation in disorders of cognitive development. J Am Acad Child Adolesc Psychiatry 49: 1091-1104. 6. Stankiewicz P, Lupski JR (2010) Structural variation in the human genome and its role in disease. Annu Rev Med 61: 437-455. 7. Zahnleiter D, Uebe S, Ekici AB, Hoyer J, Wiesener A, et al. (2013) Rare copy number variants are a common cause of short stature. PLoS Genet 9: e1003365. 8. van Duyvenvoorde HA, Lui JC, Kant SG, Oostdijk W, Gijsbers AC, et al. (2013) Copy number variants in patients with short stature. Eur J Hum Genet. 9. Alvarado DM, Aferol H, McCall K, Huang JB, Techy M, et al. (2010) Familial isolated clubfoot is associated with recurrent chromosome 17q23.1q23.2 microduplications containing TBX4. Am J Hum Genet 87: 154-160. 10. Alvarado DM, McCall K, Aferol H, Silva MJ, Garbow JR, et al. (2011) Pitx1 haploinsufficiency causes clubfoot in humans and a clubfoot-like phenotype in mice. Hum Mol Genet 20: 3943-3952. 11. Alvarado DM, Buchan JG, Frick SL, Herzenberg JE, Dobbs MB, et al. (2013) Copy number analysis of 413 isolated talipes equinovarus patients suggests role for transcriptional regulators of early limb development. Eur J Hum Genet 21: 373-380. 12. Shin JH, Ha KY, Jung SH, Chung YJ (2011) Genetic predisposition in degenerative lumbar scoliosis due to the copy number variation. Spine (Phila Pa 1976) 36: 1782-1793. 13. Chew S, Dastani Z, Brown SJ, Lewis JR, Dudbridge F, et al. (2012) Copy number variation of the APC gene is associated with regulation of bone mineral density. Bone 51: 939-943. 14. Al-Kateb H, Khanna G, Filges I, Hauser N, Grange DK, et al. (2014) Scoliosis and vertebral anomalies: Additional abnormal phenotypes associated with chromosome 16p11.2 rearrangement. Am J Med Genet A. 15. Cobb JR (1948) Outline for the study of scoliosis. American Academy of Orthopaedic Surgeons Instructional Course Lectures 5: 261-265. 16. Dick DM, Foroud T, Flury L, Bowman ES, Miller MJ, et al. (2003) Genomewide linkage analyses of bipolar disorder: a new sample of 250 pedigrees from the National Institute of Mental Health Genetics Initiative. Am J Hum Genet 73: 107-114.

17. Cooper GM, Coe BP, Girirajan S, Rosenfeld JA, Vu TH, et al. (2011) A copy number variation morbidity map of developmental delay. Nat Genet 43: 838-846. 18. Nielsen J, Wohlert M (1991) Chromosome abnormalities found among 34,910 newborn children: results from a 13-year incidence study in Arhus, Denmark. Hum Genet 87: 81- 83. 19. Rosenfeld JA, Traylor RN, Schaefer GB, McPherson EW, Ballif BC, et al. (2012) Proximal microdeletions and microduplications of 1q21.1 contribute to variable abnormal phenotypes. Eur J Hum Genet 20: 754-761. 20. Mefford HC, Sharp AJ, Baker C, Itsara A, Jiang Z, et al. (2008) Recurrent rearrangements of chromosome 1q21.1 and variable pediatric phenotypes. N Engl J Med 359: 1685-1699. 21. Sharp AJ, Locke DP, McGrath SD, Cheng Z, Bailey JA, et al. (2005) Segmental duplications and copy-number variation in the human genome. Am J Hum Genet 77: 78-88. 22. Brunetti-Pierri N, Berg JS, Scaglia F, Belmont J, Bacino CA, et al. (2008) Recurrent reciprocal 1q21.1 deletions and duplications associated with microcephaly or macrocephaly and developmental and behavioral abnormalities. Nat Genet 40: 1466- 1471. 23. Autism Genome Project C, Szatmari P, Paterson AD, Zwaigenbaum L, Roberts W, et al. (2007) Mapping autism risk loci using genetic linkage and chromosomal rearrangements. Nat Genet 39: 319-328. 24. Bucan M, Abrahams BS, Wang K, Glessner JT, Herman EI, et al. (2009) Genome-wide analyses of exonic copy number variants in a family-based study point to novel autism susceptibility genes. PLoS Genet 5: e1000536. 25. Sharp AJ, Hansen S, Selzer RR, Cheng Z, Regan R, et al. (2006) Discovery of previously unidentified genomic disorders from the duplication architecture of the human genome. Nat Genet 38: 1038-1042. 26. de Vries BB, Pfundt R, Leisink M, Koolen DA, Vissers LE, et al. (2005) Diagnostic genome profiling in mental retardation. Am J Hum Genet 77: 606-616. 27. Jaillard S, Drunat S, Bendavid C, Aboura A, Etcheverry A, et al. (2010) Identification of gene copy number variations in patients with mental retardation using array-CGH: Novel syndromes in a large French series. Eur J Med Genet 53: 66-75. 28. Ikeda M, Aleksic B, Kirov G, Kinoshita Y, Yamanouchi Y, et al. (2010) Copy number variation in schizophrenia in the Japanese population. Biol Psychiatry 67: 283-286. 29. International Schizophrenia C (2008) Rare chromosomal deletions and duplications increase risk of schizophrenia. Nature 455: 237-241. 30. Stefansson H, Rujescu D, Cichon S, Pietilainen OP, Ingason A, et al. (2008) Large recurrent microdeletions associated with schizophrenia. Nature 455: 232-236. 31. Kirov G, Grozeva D, Norton N, Ivanov D, Mantripragada KK, et al. (2009) Support for the involvement of large copy number variants in the pathogenesis of schizophrenia. Hum Mol Genet 18: 1497-1503.

32. Need AC, Ge D, Weale ME, Maia J, Feng S, et al. (2009) A genome-wide investigation of SNPs and CNVs in schizophrenia. PLoS Genet 5: e1000373. 33. Levinson DF, Duan J, Oh S, Wang K, Sanders AR, et al. (2011) Copy number variants in schizophrenia: confirmation of five previous findings and new evidence for 3q29 microdeletions and VIPR2 duplications. Am J Psychiatry 168: 302-316. 34. Greenway SC, Pereira AC, Lin JC, DePalma SR, Israel SJ, et al. (2009) De novo copy number variants identify new genes and loci in isolated sporadic tetralogy of Fallot. Nat Genet 41: 931-935. 35. Erdogan F, Larsen LA, Zhang L, Tumer Z, Tommerup N, et al. (2008) High frequency of submicroscopic genomic aberrations detected by tiling path array comparative genome hybridisation in patients with isolated congenital heart disease. J Med Genet 45: 704-709. 36. Christiansen J, Dyck JD, Elyas BG, Lilley M, Bamforth JS, et al. (2004) Chromosome 1q21.1 contiguous gene deletion is associated with congenital heart disease. Circ Res 94: 1429-1435. 37. Weber S, Landwehr C, Renkert M, Hoischen A, Wuhl E, et al. (2011) Mapping candidate regions and genes for congenital anomalies of the kidneys and urinary tract (CAKUT) by array-based comparative genomic hybridization. Nephrol Dial Transplant 26: 136-143. 38. Klopocki E, Schulze H, Strauss G, Ott CE, Hall J, et al. (2007) Complex inheritance pattern resembling autosomal recessive inheritance involving a microdeletion in thrombocytopenia-absent radius syndrome. Am J Hum Genet 80: 232-240. 39. Albers CA, Paul DS, Schulze H, Freson K, Stephens JC, et al. (2012) Compound inheritance of a low-frequency regulatory SNP and a rare null mutation in exon-junction complex subunit RBM8A causes TAR syndrome. Nat Genet 44: 435-439, S431-432. 40. Brunet A, Armengol L, Heine D, Rosell J, Garcia-Aragones M, et al. (2009) BAC array CGH in patients with Velocardiofacial syndrome-like features reveals genomic aberrations on chromosome region 1q21.1. BMC Med Genet 10: 144. 41. Warburton D, Ronemus M, Kline J, Jobanputra V, Williams I, et al. (2013) The contribution of de novo and rare inherited copy number changes to congenital heart disease in an unselected sample of children with conotruncal defects or hypoplastic left heart disease. Hum Genet. 42. Hitz MP, Lemieux-Perreault LP, Marshall C, Feroz-Zada Y, Davies R, et al. (2012) Rare copy number variants contribute to congenital left-sided heart disease. PLoS Genet 8: e1002903. 43. Ledig S, Schippert C, Strick R, Beckmann MW, Oppelt PG, et al. (2011) Recurrent aberrations identified by array-CGH in patients with Mayer-Rokitansky-Kuster-Hauser syndrome. Fertil Steril 95: 1589-1594. 44. Shen Y, Dies KA, Holm IA, Bridgemohan C, Sobeih MM, et al. (2010) Clinical genetic testing for patients with autism spectrum disorders. Pediatrics 125: e727-735. 45. Liao C, Fu F, Yi CX, Li R, Yang X, et al. (2012) Prenatal diagnosis of an atypical 1q21.1 microdeletion and duplication associated with foetal urogenital abnormalities. Gene 507: 92-94.

46. Kaminsky EB, Kaul V, Paschall J, Church DM, Bunke B, et al. (2011) An evidence-based approach to establish the functional and clinical significance of copy number variants in intellectual and developmental disabilities. Genet Med 13: 777-784. 47. Rosenfeld JA, Coe BP, Eichler EE, Cuckle H, Shaffer LG (2013) Estimates of penetrance for recurrent pathogenic copy-number variations. Genet Med 15: 478-481. 48. Abdelmoity AT, LePichon JB, Nyp SS, Soden SE, Daniel CA, et al. (2012) 15q11.2 proximal imbalances associated with a diverse array of neuropsychiatric disorders and mild dysmorphic features. J Dev Behav Pediatr 33: 570-576. 49. Shimojima K, Inoue T, Fujii Y, Ohno K, Yamamoto T (2009) A familial 593-kb microdeletion of 16p11.2 associated with mental retardation and hemivertebrae. Eur J Med Genet 52: 433-435. 50. Shen Y, Chen X, Wang L, Guo J, Shen J, et al. (2011) Intra-family phenotypic heterogeneity of 16p11.2 deletion carriers in a three-generation Chinese family. Am J Med Genet B Neuropsychiatr Genet 156: 225-232. 51. Hernando C, Plaja A, Rigola MA, Perez MM, Vendrell T, et al. (2002) Comparative genomic hybridisation shows a partial de novo deletion 16p11.2 in a neonate with multiple congenital malformations. J Med Genet 39: E24. 52. Bijlsma EK, Gijsbers AC, Schuurs-Hoeijmakers JH, van Haeringen A, Fransen van de Putte DE, et al. (2009) Extending the phenotype of recurrent rearrangements of 16p11.2: deletions in mentally retarded patients without autism and in normal individuals. Eur J Med Genet 52: 77-87. 53. Fernandez BA, Roberts W, Chung B, Weksberg R, Meyn S, et al. (2010) Phenotypic spectrum associated with de novo and inherited deletions and duplications at 16p11.2 in individuals ascertained for diagnosis of autism spectrum disorder. J Med Genet 47: 195- 203. 54. Schaaf CP, Goin-Kochel RP, Nowell KP, Hunter JV, Aleck KA, et al. (2011) Expanding the clinical spectrum of the 16p11.2 chromosomal rearrangements: three patients with syringomyelia. Eur J Hum Genet 19: 152-156. 55. Shinawi M, Liu P, Kang SH, Shen J, Belmont JW, et al. (2010) Recurrent reciprocal 16p11.2 rearrangements associated with global developmental delay, behavioural problems, dysmorphism, epilepsy, and abnormal head size. J Med Genet 47: 332-341. 56. Miller NH, Justice CM, Marosy B, Doheny KF, Pugh E, et al. (2005) Identification of candidate regions for familial idiopathic scoliosis. Spine (Phila Pa 1976) 30: 1181-1187. 57. Miller NH, Justice CM, Marosy B, Swindle K, Kim Y, et al. (2012) Intra-familial tests of association between familial idiopathic scoliosis and linked regions on 9q31.3-q34.3 and 16p12.3-q22.2. Hum Hered 74: 36-44. 58. Gustavson KH (1999) [Triple X syndrome deviation with mild symptoms. The majority goes undiagnosed]. Lakartidningen 96: 5646-5647. 59. Barr ML, Sergovich FR, Carr DH, Saver EL (1969) The triplo-X female: an appraisal based on a study of 12 cases and a review of the literature. Can Med Assoc J 101: 247-258.

60. Otter M, Schrander-Stumpel CT, Curfs LM (2010) Triple X syndrome: a review of the literature. Eur J Hum Genet 18: 265-271. 61. Olanders S (1977) Females with Supernumerary X Chromosomes; A Study of 39 Psychiatric Cases: Esselte Studium. 62. Linden MG, Bender BG, Harmon RJ, Mrazek DA, Robinson A (1988) 47,XXX: what is the prognosis? Pediatrics 82: 619-630. 63. Huang N, Lee I, Marcotte EM, Hurles ME (2010) Characterising and predicting haploinsufficiency in the human genome. PLoS Genet 6: e1001154.

Chapter 3:

Exome Sequencing Identifies Rare Variants in FBN1 and FBN2 Associated with Severe Adolescent Idiopathic Scoliosis

SUMMARY

Despite a strong genetic basis, few genes have been associated with AIS and the pathogenesis remains poorly understood. In a genome-wide rare variant burden analysis using exome sequence data, we identified FBN1 (fibrillin-1) as the most significantly associated gene with AIS. Based on these results, FBN1 and a related gene, FBN2 (fibrillin-2), were sequenced in a total of 852 AIS cases and 669 controls. In individuals of European ancestry, rare variants in

FBN1 and FBN2 were enriched in severely affected AIS cases (7.6%) compared to in-house controls (2.4%) (OR=3.5, P=5.46×10-4) and Exome Sequencing Project controls (2.3%)

(OR=3.5, P=1.48×10-6). Scoliosis severity in AIS cases was associated with FBN1 and FBN2

rare variants (P=0.0012) and replicated in an independent Han Chinese cohort (P=0.0376),

suggesting that rare variants have utility as predictors of curve progression. Clinical evaluations

revealed that the majority of AIS cases with rare FBN1 variants do not meet diagnostic criteria

for Marfan syndrome, though variants are associated with tall stature (P=0.0035) and

upregulation of the TGF-β pathway. Overall, these results expand our definition of fibrillin-

related disorders to include AIS and open up new strategies for diagnosing and treating severe

AIS.

INTRODUCTION

The human exome consists of about 180,000 exons spread across ~38 megabases (Mb).

Although this represents only 1% of the genome, disrupted protein-coding regions make up

about 85% of known variation underlying Mendelian disorders [1]. Therefore, exome sequencing provides an unbiased, genome-wide approach to identify variation in the regions that are most likely to have large affects.

Rare variants are thought to have a greater contribution to disease risk because negative

selection restricts variants of large genetic effect from rising to high frequency [2]. We therefore

hypothesized that rare genetic variants are a major contributor to severe AIS susceptibility.

Although large-scale sequencing studies are needed to detect rare variants, such studies have

recently become feasible due to emerging sequencing technologies and reduction in sequencing cost. Using exome sequencing, we first analyzed rare variants genome-wide in a small cohort of severe AIS cases and identified FBN1 and FBN2 as candidate genes for AIS, which we then sequenced in a larger case and control cohort using a newly described, cost-effective targeted

capture method.

METHODS

Patient samples and controls

AIS patients were recruited from St. Louis Children’s Hospital, St. Louis Shriners

Hospital for Children, the University of Iowa, the University of Colorado and the Chinese

University of Hong Kong. All patients had scoliosis of unknown etiology with spinal curves measuring ≥10°. Curve measurements are reported for the largest lateral spinal curve using the

Cobb method [3]. Patients with developmental delay, multiple congenital anomalies or known underlying medical disorders (e.g. Ehlers-Danlos syndrome, Marfan syndrome) were excluded.

We selected 91 unrelated AIS cases of European ancestry with severe deformity (spinal curves measuring ≥40° or requiring surgery) for the exome sequencing screen. Additional AIS cases of

European (N=344), Han Chinese (N=370) and other ancestries (N=47) were included in subsequent analyses. DNA was collected from affected probands after obtaining informed consent. Growth parameters were calculated using the National Center for Health Statistics

(http://www.cdc.gov/nchs). Systemic features of Marfan syndrome and joint hypermobility were quantified according to the revised Ghent scoring system [4], and the Beighton scoring system

[5], respectively, and were obtained in patients without knowledge of genetic sequencing results.

A subset of AIS cases were recontacted after FBN1 variants previously identified in individuals with Marfan syndrome or novel variants of uncertain significance were identified as a part of this research study. These cases were referred to clinical geneticists at St. Louis Children’s Hospital and St. Louis Shriners Hospital for Children for evaluation of possible Marfan syndrome.

Controls for this study consisted of unrelated healthy individuals or patients with neurological (e.g. Alzheimer’s disease) or other musculoskeletal disorders without spine

involvement (e.g. limb deformity). For the exome sequencing screen, we selected 337 controls

of European ancestry. Additional in-house controls of European (N=249) or Han Chinese

(N=83) ancestry were included in subsequent analyses. Individuals of European ancestry

(N=4,300) were included from the Exome Variant Server, NHLBI GO Exome Sequencing

Project (ESP) (http://evs.gs.washington.edu/EVS) public database as an additional control cohort

(ESP6500SI-V2).

Exon enrichment and sequencing

Exon enrichment was performed using the SureSelect Human All Exon 38Mb and 50Mb

kits (Agilent Technologies, Santa Clara, California) or the TruSeq Exome Enrichment kit

(Illumina, San Diego, California). Exome data was extracted from whole genome sequencing in

a subset of in-house controls. Samples were sequenced on a Genome Analyzer IIx or HiSeq

2000 sequencer (Illumina, San Diego, California) by the Washington University Genome

Technology Access Center.

Targeted capture of FBN1 and FBN2

Target enrichment of FBN1 and FBN2 was performed using the Multiplex Direct

Genomic Selection method (manuscript under review), which utilizes multiplexed capture [6] with targeted selection using bacterial artificial chromosomes (BACs) [7] and next-generation sequencing. DNA samples (300ng) were sonicated using a Covaris E210 focus-ultrasonicator

(Covaris Inc., Woburn, Massachusetts). Samples were indexed using previously described adapter sequences [6]. Indexed samples were pooled in batches of 48-96 prior to capture. Four

BACs spanning the entire coding regions of FBN1 (RP11-1144G24 [chr15:48676783-48822143]

67 and RP11-147E14 [chr15:48822351-48985158]) and FBN2 (RP11-909P14 [chr5:127,567,172-

127,756,046] and RP11-351A8 [chr5:127,752,763-127,906,815]) (BACPAC Resources Center,

Oakland, California) were used as baits. BACs were biotinylated using nick translation with biotin-16-dUTP (Roche Applied Science, Penzberg, Germany). Pooled samples were hybridized with biotinylated BACs. Dynabeads M-280 Streptavidin (Life Technologies Co., Carlsbad,

California) were used to isolate BAC-associated DNA. Post-capture, pooled samples were sequenced using a MiSeq Personal Sequencer (Illumina, San Diego, California).

Alignment and variant calling

Next generation sequencing reads were aligned to hg19 human reference sequence

(Genome Reference Consortium Human Build 37) using Novoalign software (Novocraft

Technologies, Selangor, Malaysia). Variant calling and annotation were completed using

SAMtools [8] and SeattleSeq Annotation 131

(http://snp.gs.washington.edu/SeattleSeqAnnotation131), respectively. All AIS cases and in- house controls were analyzed using an identical methodology.

Quality control and gene burden analysis

For all analyses, variants were defined as rare when absent from the dbSNP database

(build 137). Analysis was restricted to variants altering the coding sequence (nonsense, splice- site, missense and insertion/deletion mutations). We used a collapsing approach to measure gene burden [9]. Rare variant burden was quantified as the sum of all variant alleles for each gene divided by the total alleles in the population and is expressed as the collapsed minor allele frequency (cMAF).

For the exome sequencing screen, only AIS cases and in-house controls with ≥90% of the exome covered at ≥8X coverage were included in the analysis. Variant calls were merged at all sites and filtered for sites called in >80% of individuals. Gene burden analysis was performed exome-wide for each gene with at least one variant site.

For gene burden analysis of FBN1 and FBN2 in the larger dataset, more conservative quality thresholds were applied. Only individual samples with ≥95% of FBN1 (NM_000138) or

FBN2 (NM_001999) covered at ≥8X read depth were included. Low quality variants (phred- scaled quality score <30 or genotype quality score <75) were removed. Controls from the ESP were similarly analyzed using build 137 of the dbSNP database as a filter, which does not include individuals from this dataset. In the Han Chinese patient cohort, variants identified in dbSNP or in 83 Han Chinese controls were excluded from analysis. All rare variants in AIS cases were validated by Sequenom MassARRAY (Sequenom, San Diego, California) by the

Washington University Human Genetics Division of Genotyping Core or by Sanger sequencing using an ABI 3730 Sequencer (Life Technologies, Carlsbad, California). All genotyped variants

(N=32) validated.

Statistical analysis

One-tailed P-values <0.05 were considered statistically significant. A Fisher’s exact test was used for all gene burden analyses. The Shapiro-Wilk test demonstrated non-normally distributed clinical characteristics (P<0.05), therefore, the Mann-Whitney-Wilcoxon test was used to compare clinical features, except gender and AIS family history, which were compared using a Fisher’s exact test. Statistical tests were performed with GraphPad Prism software

(http://www.graphpad.com/scientific-software/prism).

Protein immunoblotting

Immunoblotting was performed using paraspinous muscle collected during surgery for

scoliosis (AIS and Marfan syndrome) or for a ruptured spinal disc in a patient without scoliosis

(unaffected control). All samples were collected from females between 9 and 18 years old.

Denatured protein (200µg) was separated on a 7.5% gel by SDS-PAGE and transferred to a nitrocellulose membrane. Rabbit polyclonal antibody to pSmad2 (1:1,000; EMD Millipore Co.) and mouse monoclonal antibody to actin (AC1-20.4.2; 1:2000; Sigma-Aldrich Co.) were used.

Chemiluminescent signals in AIS cases and Marfan syndrome were quantified using ImageJ

[10]. Phospho-SMAD2 signal was normalized to actin and the unaffected control.

RESULTS

Exome sequencing screen identifies FBN1 and FBN2 as a candidate genes for severe AIS

To identify candidate genes for AIS, exome sequencing was performed on 91 AIS cases

with severe scoliosis (spinal curves measuring ≥40° or surgically treated). We restricted our

analysis to AIS patients with severe scoliosis to reduce the likelihood of latent disease in the

control population, thus increasing power of the association [11]. Furthermore, identifying

genetic variants associated with severe AIS has important clinical utility. Cases in this cohort

had an average spinal curve of 60.2° (SD=15.2°) and 84% were female. The female bias is

consistent with previous studies, where female-to-male ratios are estimated to be as high as 10:1

with severe deformity [12]. Exome sequencing data was also generated for 337 controls. All

cases and controls were unrelated and of European ancestry. Only coding variants that caused nonsense, splice-site, missense or insertion/deletion mutations which were rare (defined as being

absent in dbSNP) were included in the analysis. This filtering strategy was selected to enrich in

mutations that are very rare and most likely to be deleterious.

Using a gene-burden analysis, we compared the genome-wide frequency of rare coding

variants in severe AIS cases and controls. Of nearly 13,000 genes identified with at least one

rare variant, FBN1 (encoding fibrillin-1) was the most significantly associated gene with AIS compared to controls (P=3.17×10-4, Fisher’s exact test). The collapsed minor allele frequency

(cMAF) for all rare variants in FBN1 was 0.044 in cases (N=91), compared to 0.005 in controls

(N=337) (OR=10.4, 95% CI=2.7-39.5). This result was surprising, as mutations in FBN1 are

associated with Marfan syndrome (MIM 154700), an autosomal dominant and highly penetrant

disorder that causes skeletal, ocular and cardiovascular phenotypes. Likely a result of the small

sample size, the association did not reach exome-wide significance (P<2.5×10-6). However,

FBN1 was a compelling candidate for AIS because 63% of Marfan syndrome patients develop scoliosis [13]. Congenital contractural arachnodactyly (MIM 121050) is a related disorder that has large phenotypic overlap with Marfan syndrome, including scoliosis which develops in 50% of patients [14], and is caused by autosomal dominant mutations in fibrillin-2 (FBN2). FBN2 showed weak association with AIS in our exome sequencing screen (cMAF=0.022) compared to controls (cMAF=0.004) (P=0.0404, Fisher’s exact test), but was also a compelling candidate for

AIS due to a recent genome-wide association study that identified an FBN2 single nucleotide polymorphism (SNP) among the top 100 SNPs associated with AIS [15].

Rare coding variants in FBN1 and FBN2 are enriched in severe AIS

Based on the results of our exome sequencing screen, FBN1 and FBN2 were selected as

candidate genes to study in a larger cohort. Both genes were sequenced in additional AIS

patients and controls using a newly described targeted capture approach. In total, 323 severely affected AIS cases and 493 controls of European descent were analyzed for rare variants in

FBN1 or FBN2. All cases and controls had ≥95% coverage at ≥8X read depth to be included in the gene analysis. The average spinal curve in the AIS cohort was 57.4° (SD=12.1°). As an independent control cohort, we also identified rare variants in FBN1 and FBN2 for 4,300 individuals of European ancestry from the NHLBI GO Exome Sequencing Project (ESP) controls (N=4,300).

For FBN1, rare variants were enriched in AIS cases (cMAF=0.021) compared to both in- house controls (cMAF=0.005) (P=0.0041, Fisher’s exact test) and ESP controls (cMAF=0.005)

(P=8.14×10-5, Fisher’s exact test) (Table 3.1). Thirteen AIS patients (4.2%, N=13/311) and five

in-house controls (1.0%, N=5/489) were identified with a rare variant in FBN1. An estimated

1.0% (N=44/4300) of ESP controls had a rare FBN1 variant (assuming each individual harbored only one variant). All AIS variants were heterozygous and caused missense amino acid changes throughout the FBN1 protein (Figure 3.1A, Table 3.2). Segregation was evaluated for nine families where DNA was available from one or more additional family members (Figure 3.2). Variants segregated with AIS in some families. Two variants (p.G1217S and p.G1313S) were located in the neonatal

region of FBN1 (exons 24-32), which is typically associated with a more severe, early-onset

form of Marfan syndrome [16,17]. Three variants were previously reported in Marfan syndrome

(p.L1405R [18], p.M1576T [19,20] and p.I2585T [18,21,22]) and one of these variants

(p.M1576T) was recurrent in three AIS patients (Table 3.2). Variants previously reported in

Marfan syndrome were also identified in the ESP cohort (p.D1479E [18], p.E2019K [23],

p.R2554W [24], p.R2694Q [25] and p.S2832G [25]) whereas none of the in-house control

variants were previously reported as pathogenic.

Similarly, rare FBN2 variants were also enriched in AIS cases (cMAF=0.017) compared

to in-house controls (cMAF=0.006) (P=0.0307, Fisher’s exact test) and ESP controls

(cMAF=0.007) (P=0.0054, Fisher’s exact test) (Table 3.1). Eleven AIS patients (3.5%,

N=11/316), five in-house controls (1.2%, N=5/427) and approximately 1.3% (N=56/4300) of

ESP controls were identified with a variant in FBN2. All AIS variants were heterozygous and

caused missense mutations, except for a five base insertion present in one AIS case (Figure 3.1B,

Table 3.2). DNA was available for one or more additional family members in seven families. Variants segregated in some families (Figure 3.2). In congenital contractural arachnodactyly, pathogenic

mutations are limited to the neonatal region (exons 23-34) [26]. In contrast, variants identified in

AIS patients were located throughout the protein, including the neonatal region (p.R1021C,

p.I1116S, p.L1125V and p.G1271A). None of the variants identified in AIS cases or controls were previously reported in congenital contractural arachnodactyly.

Combining rare variants from both FBN1 and FBN2 yielded a cMAF of 0.039 in AIS cases, 0.012 in controls (P=5.46×10-4, Fisher’s exact test) and 0.012 in ESP controls (P=1.48×10-

6, Fisher’s exact test) (Table 3.1). Only one AIS case and none of the in-house controls had a

variant in both FBN1 and FBN2. Thus, variants in either FBN1 or FBN2 were identified in 7.6%

(N=23/304) of AIS cases, 2.4% (N=10/425) of in-house controls and approximately 2.3%

(N=100/4300) of ESP controls.

In addition to AIS cases with severe scoliosis, we also sequenced 112 AIS cases of

European ancestry with mild to moderate scoliosis (curve <40°). The average spinal curve in this cohort was 28.1° (SD=6.9°). Rare variants in FBN1 and FBN2 were identified in 1.9%

(N=2/106) and 0.9% (N=1/107), respectively, of the mild to moderately affected AIS cases

(Figure 3.1A, Figure 3.1B, Table 3.2). FBN1 and FBN2 rare variant frequencies were not significantly different in mild to moderate AIS cases compared to in-house controls (P=0.47,

Fisher’s exact test) or ESP controls (P=0.42, Fisher’s exact test).

Distinct clinical features of AIS patients with rare fibrillin variants

To determine if AIS patients with rare fibrillin variants manifest unique clinical characteristics, we compared patients with variants in FBN1 (N=15) and FBN2 (N=12) to

patients without rare variants in either gene (N=379) (Table 3.3). Analysis was limited to

individuals of European ancestry, but included all curve severities (≥10°). The average spinal

curve was 20° greater in AIS patients with an FBN1 variant compared to other patients

(P=6.91×10-4, Mann-Whitney-Wilcoxon test). All 15 AIS patients with rare FBN1 variants

developed scoliosis with a Cobb angle of ≥35° (Table 3.2), which would typically warrant bracing or surgery. FBN1 variants were also associated with taller stature standardized to age and gender (P=0.0035, Mann-Whitney-Wilcoxon test). Of all AIS patients in this cohort with height ≥90th percentile, 22.5% (N=9/40) were found to have a rare FBN1 variant. Rare FBN1 variants were also associated with Marfan-associated features, which were quantified using the revised Ghent criteria [4] (P=0.0072, Mann-Whitney-Wilcoxon test). Gender, weight, body mass index, family history of treated AIS and joint hypermobility were not associated with the presence of rare FBN1 variants. There were no unique clinical characteristics in AIS patients with rare FBN2 variants (Table 3.3). However, like FBN1, we noted a trend towards greater spinal curves in AIS patients with FBN2 variants (P=0.10).

Curve severity is associated with rare fibrillin variants in AIS cases of European and Han

Chinese ancestry

AIS patients of European ancestry with a rare variant in either FBN1 or FBN2 (N=26) developed an average spinal curve of 64.0°, compared to 49.0° in cases with no rare variants

(N=364) (P=0.0012, Mann-Whitney-Wilcoxon test) (Figure 3.3). Because predicting curve severity has important clinical implications, we performed a replication study in 370 Han

Chinese AIS patients. Our replication cohort included all curve severities ≥10° (average=34.1°,

SD=12.5°). Twenty-eight (7.6%, N=28/370) Han Chinese cases were identified with a rare coding variant in FBN1 or FBN2. The average spinal curve in these cases (37.2°, N=28) was greater than cases with no rare variant in FBN1 or FBN2 (33.9°, N=342) (P=0.0376, Mann-

Whitney-Wilcoxon test) (Figure 3.3). Moreover, 11.9% (N=12/101) of Han Chinese cases with

75 severe spinal curves measuring ≥40° had a rare variant in FBN1 or FBN2, compared to only

5.9% (N=16/269) of cases with spinal curves <40° (P=0.04826, Fisher’s exact test).

AIS cases of other ancestral background (N=47) were also sequenced in this study and an additional five rare variants in FBN1 and FBN2 were identified (Figure 3.1A, Figure 3.1B, Table

3.2). In all ancestries, the average spinal curve in AIS cases with a rare FBN1 or FBN2 variant was 50.5° (N=59), compared to 42.1° in cases with no fibrillin variant (N=742) (P=0.0026,

Mann-Whitney-Wilcoxon test) (Figure 3.3).

Evaluation of Marfan syndrome in AIS patients with FBN1 variants

Pathogenic mutations in Marfan syndrome result in progressive dilation of the aorta, which predisposes individuals to aortic aneurysms and dissections [4]. Because the life- threatening cardiovascular complications of Marfan syndrome are preventable, it is currently recommended that known or expected pathogenic mutations identified incidentally during clinical sequencing are reported [27]. Therefore, AIS patients with rare FBN1 variants were referred for additional evaluations assessing potential Marfan syndrome. Nine patients agreed to additional evaluations and were subsequently assessed by a clinical geneticist (Table 3.4). Three

AIS patients had Marfan syndrome-associated mutations (p.L1405R, p.M1576T and p.I2585T) and six patients had variants of unknown significance (p.N280T, p.Q697R, p.N703H, p.N2178K, p.Y2225F and p.V2868I). The revised Ghent nosology [4] was employed. AIS patients were between 10 and 24 years old at time of evaluation. Seven patients had imaging studies for aortic root dilation and five patients received a dilated eye exam. Two AIS patients received physical examinations but no cardiac imaging studies. Familial DNA was available for genotyping in nine families. The FBN1 variant was inherited in eight families and one family was inconclusive (Figure 3.2).

None of the AIS patients reported significant features suggestive of Marfan syndrome (ectopia lentis,

unexplained death, aortic dissection or aortic aneurysm) in any family members. Nonspecific features of

Marfan syndrome (e.g. pectus deformity, hypermobility) were present in some family members. Only

one of the nine patients that received additional evaluation met the diagnostic criteria for Marfan

syndrome. This patient (6386001) had aortic root dilation (Z-score=2.6) and significant Marfan-

associated systemic features (Ghent systemic features score=10) (Table 3.4). Of all patients of

European descent evaluated for systemic features (N=118), 6386001 was the only patient that

scored ≥7, which is considered clinically significant [4]. The mutation identified in 6386001

(p.I2585T) was previously described in multiple individuals with Marfan syndrome [18,21,22].

Increased pSMAD2 in paraspinous muscle of AIS patients with rare FBN1 variants

Previous studies have shown that pathogenic FBN1 mutations causing Marfan syndrome

result in upregulated transforming growth factor beta (TGF-β) signaling that can be measured in

plasma or indirectly measured through activation of downstream targets, such as phosphorylation

of SMAD2 (pSMAD2) [28-31]. To determine if AIS patients with rare FBN1 and FBN2 variants

also have upregulated TGF-β, we used western blotting to examine pSMAD2 levels in

paraspinous muscle from three AIS cases harboring a variant in FBN1 (p.P1225L, p.M1576T and

p.G2003R) and one AIS case with a rare variant in FBN2 (p.P740H) (Figure 3.4). Paraspinous muscle was also collected for a patient diagnosed with Marfan syndrome (positive control) and an unaffected individual who did not have scoliosis (negative control). Compared to the unaffected control, pSMAD2 was elevated in all AIS patients with rare FBN1 variants and in the

Marfan syndrome patient. We observed similar pSMAD2 levels the AIS patient with a rare

FBN2 variant compared to the unaffected control.

DISCUSSION

Pathogenic mutations in FBN1 are most frequently associated with Marfan syndrome

[32] where over 1,000 mutations have been described [33], but mutations in FBN1 have also been reported in ectopia lentis syndrome [34], isolated aortic aneurysm [35-38], Shprintzen-

Goldberg syndrome [39], Weill-Marchesani syndrome [40], geleophysic and acromicric dysplasia [41], stiff skin syndrome [42] and MASS phenotype (myopia, mitral valve prolapse, borderline aortic root dilation, skeletal and skin findings) [43]. Mutations in FBN2 have previously been associated only with congenital contractural arachnodactyly [44]. In the current study, we demonstrate an important role for FBN1 and FBN2 in the pathogenesis of AIS, broadening the spectrum of fibrillin-related disorders.

Until now, the majority of FBN1 and FBN2 sequencing studies have been limited to select cohorts of patients with suspected Marfan syndrome and congenital contractural arachnodactyly, particularly because the cost of sequencing FBN1 and FBN2, each consisting of

~11kb of coding sequence spread over 65 exons, is prohibitively high in large cohorts. However,

FBN1 mutations were previously described in a family with isolated kyphoscoliosis [45] and a family with isolated skeletal features including scoliosis [46] suggesting that fibrillin might contribute to isolated skeletal phenotypes, like AIS. In addition, earlier investigations revealed fibrillin abnormalities in the fibroblasts of AIS patients [47], but linkage analysis did not support segregation of the FBN1 locus in large AIS pedigrees [48] and further genetic studies in AIS were not pursued.

Using exome sequencing and a cost-effective targeted capture approach to sequence

FBN1 and FBN2, we found that the burden of FBN1 and FBN2 rare variants is relatively high in

severely affected AIS patients, with 7.6% of patients of European ancestry having a rare variant

in either gene, over three-times the frequency of controls (2.4%). Curve severity was the most striking clinical finding that distinguished patients with rare fibrillin variants, leading to spinal curves that were on average 15° larger than observed in other AIS patients. Interestingly, scoliosis is common in patients with Marfan syndrome [13] and congenital contractural arachnodactyly [14] and, like the AIS cases studied here, often progresses to severe curves

[49,50]. Our findings indicate that FBN1 and FBN2 variants could serve as prognostic genetic markers to predict scoliosis progression. This has important clinical utility because of the potential to develop personalized prevention strategies for those at high risk of severe deformity and to eliminate screening and treatment for patients at low risk [51-53].

In addition to spinal curve progression, we found that rare FBN1 variants are also associated with tall stature in the AIS population. Because nearly one-quarter of AIS cases in our study with height ≥90th percentile had a rare FBN1 variant, this easily obtainable clinical

measure can identify patients who are particularly likely to harbor these genetic variants and may

benefit from FBN1 sequencing studies. We also found that AIS patients with FBN1 variants

collectively scored higher using the Ghent systemic features scoring system [4], which evaluates several nonspecific physical traits that are common in Marfan syndrome. However, the differences were small and the systemic features score is unlikely to be useful for identifying

individual AIS patients with rare FBN1 variants. Despite this, our data confirm that the systemic

features scores do have utility as a screen for unrecognized Marfan syndrome in the AIS

population, as the only patient with a clinically significant systemic feature score (≥7) was

identified with an FBN1 mutation and subsequently diagnosed with Marfan syndrome as a part

of this study after additional clinical testing.

While FBN1 mutations causing Marfan syndrome are associated with life threatening

cardiovascular complications, our data suggest that the majority of AIS-associated variants are

associated with a more benign course. First, physical findings were not suggestive of Marfan

syndrome in most patients (N=8/9) and, had these patients been evaluated without knowledge of

FBN1 mutation status, Marfan syndrome would not have been suspected and FBN1 testing would not have been recommended based on clinical genetics evaluations alone. Second, the majority of patients who received imaging studies of the aorta showed no evidence of aortic

dilation (N=6/7), which is the cardinal feature of Marfan syndrome. Third, family histories did not endorse Marfan syndrome. Rare FBN1 variants were inherited from a parent in at least seven families, but no patient reported family members with features suggestive of Marfan syndrome, including aortic dilation, aortic aneurysm or dissection, ectopia lentis, or unexplained death.

However, despite little evidence supporting a diagnosis of Marfan syndrome in most AIS

patients with rare FBN1 variants, one AIS patient not previously suspected of the disorder was

diagnosed as a direct result of this study. Because there are treatments available to prevent the

life-threatening aortic complications of Marfan syndrome (e.g. losartan [28,54-56]), recognition

of the disorder can significantly affect patient outcomes. Unfortunately, for AIS patients with

FBN1 variants who did not fulfill the diagnostic criteria at the time of examination, the absence

of aortic dilation does not conclusively rule out Marfan syndrome and aortic dilation may

develop later in life [4]. Ongoing screening of all AIS patients with rare FBN1 variants is therefore warranted and is consistent with the current recommendations for patients with FBN1

mutations who do not fulfill diagnostic criteria [23,57]. Regardless of the long-term risks associated with rare FBN1 variants, FBN1 sequencing in the AIS population is likely to have

80 clinical value, either as an early indicator of potential Marfan syndrome or as a predictor of progressive AIS.

Congenital contractural arachnodactyly does not appear to be associated with the same risk of life-threatening cardiovascular complications as Marfan syndrome; however, aortic dilation has been described in several patients and echocardiographic evaluations are currently recommended [26]. AIS patients with rare FBN2 variants did not receive additional clinical evaluation, but patient histories did not indicate “crumpled” ears, muscular hypoplasia, contractures of the major joints or other features that would be suggestive of congenital contractural arachnodactyly.

Although our findings suggest that rare variants in FBN1 and FBN2 are associated with

AIS rather than unrecognized Marfan syndrome or congenital contractural arachnodactyly, additional long-term studies of AIS patients with fibrillin abnormalities are needed. Such studies would clarify the potential future complications, if any, associated with rare fibrillin variants and inform the interpretation of incidentally identified FBN1 and FBN2 variants in future diagnostic or research studies.

FBN1 and FBN2 are expressed with significantly overlapping patterns beginning during early development (see review [58]) and the biological mechanism leading to scoliosis when

FBN1 and FBN2 are mutated remains unclear, even in cases of Marfan syndrome and congenital contractural arachnodactyly. Aberrant activation of the TGF-β pathway has been observed in

Marfan syndrome [30] and may be responsible for some of the skeletal features, like tall stature

[59]. We found that AIS patients with FBN1 variants were also tall, which suggests that these protein changes also alter TGF-β signaling. We showed upregulation of the TGF-β pathway in paraspinous muscle of three AIS patients with rare FBN1 variants, confirming that at least some

81 of the variants identified in AIS have functional effects. Recent studies have shown that FBN2 also regulates TGF-β signaling [60], but we did not observe upregulation of TGF-β in muscle from an AIS patient with a rare FBN2 variant, suggesting that the variant may not have an effect on this aspect of FBN2 function. Unfortunately, we did not have access to muscle on a larger cohort of AIS patients and further studies will be required to demonstrate more definitively that these gene variants activate the TGF-β pathway. Interestingly, the TGF-β pathway is also upregulated in patients with Loeys-Dietz syndrome [61,62], a disorder that has many overlapping features with Marfan syndrome and congenital arachnodactyly and is highly associated with scoliosis [63]. Because scoliosis is a common feature of several Mendelian disorders involving the TGF-β signaling pathway, other genes in the TGF-β pathway may be good candidates for

AIS, which is supported by a recent study that identified transforming growth factor beta-1 as a novel susceptibility gene for AIS [64]. Moreover, an association of TGF-β signaling with AIS could explain why other Marfan-associated features are observed more frequently in AIS patients than in the general population, including tall stature [65], joint hypermobility [66], pectus excavatum [67,68] and dural ectasia [69], Recent studies in mouse models and humans with

Marfan syndrome have shown promising developments in therapeutic strategies to reduce the hyper-activated TGF-β pathway, including TGF-β neutralizing antibody and the angiotensin II type 1 receptor blocker, losartan [28,54-56]. The potential link between AIS and the TGF-β signaling pathway suggests the possibility that drugs like losartan may someday be useful to treat

AIS cases in which the TGF-β pathway is activated.

In summary, we demonstrate an important role for FBN1 and FBN2 in AIS pathogenesis.

We show that rare variants are enriched in severely affected AIS patients and are significantly associated with curve severity. Because reliable methods for predicting scoliosis progression are

not currently available, these results have immediate clinical utility. Furthermore, we show altered TGF-β signaling in AIS patients with fibrillin mutations, which opens up the possibility of novel treatments that have not previously been considered for AIS.

Figure 3.1: Rare FBN1 and FBN2 variants identified in AIS patients Protein structure of FBN1 and FBN2 (redrawn [70]). (A) Missense variants were identified throughout FBN1 in AIS patients of European descent (N=417) and other ancestries (N=47). Variants previously associated with Marfan syndrome are in bold. (B) Missense and other coding changes were identified in FBN2 in AIS patients of European (N=423) and other ancestries (N=47). Asterisk indicates variants identified in other ancestries.

Figure 3.2: Segregation of FBN1 and FBN2 variants Variant segregation was evaluated for AIS families with two or more individuals available for genotyping. The presence of AIS resulting in bracing or surgery (treated AIS), AIS that measured <20° or was self-reported and not treated (mild AIS) and other features of Marfan syndrome and congenital contractural arachnodactyly (pectus carinatum, pectus excavatum, hypermobility, spondylolisthesis) are shown.

Figure 3.3: Rare FBN1 and FBN2 variants are associated with curve severity in AIS The average spinal curve was determined for AIS cases with a rare variant in either FBN1 or FBN2 (Variant) and AIS cases without a rare variant in either gene (No variant). AIS cases of European ancestry (N=405), Han Chinese ancestry (N=370) and a combined analysis consisting of cases of European, Han Chinese and other ancestries (N=801) consistently showed that patients with rare variants in FBN1 or FBN2 developed larger spinal curves. Asterisk indicates the following P-values: *<0.05, **<0.01.

Figure 3.4: Elevated pSMAD2 in AIS patients with rare FBN1 variants Paraspinous muscle from an unaffected control, AIS cases with rare variants in FBN1 or FBN2 and a Marfan syndrome patient were immunostained for pSMAD2. pSMAD2 was elevated in AIS cases with rare FBN1 variants at levels similar to Marfan syndrome. pSMAD2/actin ratio is shown normalized to the unaffected control.

Table 3.1: Rare variant frequencies for FBN1 and FBN2 in AIS cases and controls of European ancestry

AIS cases (≥40°) Controls (this study) Controls (ESP)

Variant Variant Odds ratio Variant Odds ratio N cMAF N cMAF P-value N cMAF P-value alleles alleles (95% CI) alleles (95% CI)

FBN1 311 13 0.021 489 5 0.005 4.2 (1.5-11.7) P=0.0041 4,300 44 0.005 4.2 (2.2-7.7) P=8.14×10-5

FBN2 316 11 0.017 427 5 0.006 3.0 (1.0-8.7) P=0.0307 4,300 56 0.007 2.7 (1.4-5.2) P=0.0054

FBN or 304 24 0.039 425 10 0.012 3.5 (1.6-7.3) P=5.46×10-4 4,300 100 0.012 3.5 (2.2-5.5) P=1.48×10-6 FBN2

cMAF, collapsed minor allele frequency; ESP, NHLBI Exome Sequencing Project

Table 3.2: Rare FBN1 and FBN2 variants identified in AIS cases (curve ≥10°) Scoliosis Genomic GVS Amino acid ESP Disease Case Ancestry Base change Gene curve position (hg19) function change frequencya association 6128001 84° EUR chr15:48902952 T>A FBN1 missense p.Ile107Leu - - 6340001 48° EUR chr15:48826300 T>G FBN1 missense p.Asn280Thr - - 6390001 60° AFR chr15:48796007 T>C FBN1 missense p.Gln697Arg - - 6377001 50° EUR chr15:48795990 T>G FBN1 missense p.Asn703His - - 6016001 36° Other chr15:48784766 C>T FBN1 missense p.Val916Met N=1/6,494 - 1044 70° EUR chr15:48777634* C>T FBN1 missense p.Gly1217Ser - - 6273001 48° AFR chr15:48777609* G>A FBN1 missense p.Pro1225Leu - - 1083 52° EUR chr15:48773879* C>T FBN1 missense p.Gly1313Ser - - 6226001 97° EUR chr15:48764870 A>C FBN1 missense p.Leu1405Arg - MFS [18] 6272001 90° EUR 6418001 65° EUR chr15:48760155 A>G FBN1 missense p.Met1576Thr - MFS [19,20] 6442001 35° EUR 450-14600 38° EUR chr15:48741087 C>T FBN1 missense p.Arg1850His - - 6320001 66° EUR chr15:48736768 C>T FBN1 missense p.Gly2003Arg - - 6368001 100° EUR chr15:48726873 A>T FBN1 missense p.Asn2178Lys - - 6111001 95° EUR chr15:48725128 T>A FBN1 missense p.Tyr2225Phe - - MFS 6386001 93° EUR chr15:48712949 A>G FBN1 missense p.Ile2585Thr - [18,21,22] 6005001 55° EUR chr15:48703201 C>T FBN1 missense p.Val2868Ile - - 6463001 57° AFR chr5:127873139 C>T FBN2 missense p.Gly53Asp N=2/6,489 - 6112001 48° EUR chr5:127872157 C>T FBN2 missense p.Arg92Lys - - 1157 45° EUR chr5:127782238 A>ACTGTA FBN2 frameshift - - - 6124001 75° EUR chr5:127713520 C>T FBN2 missense p.Val592Met - - 6279001 55° EUR chr5:127704904 G>T FBN2 missense p.Pro740His - - 1022 25° EUR chr5:127681205* G>A FBN2 missense p.Arg1021Cys - - 6191001 90° EUR chr5:127674750* A>C FBN2 missense p.Ile1116Ser - - 6129001 80° EUR chr5:127674724* G>C FBN2 missense p.Leu1125Val - - 6367001 70° Other chr5:127673755* C>T FBN2 missense p.Glu1178Lys - - 6229001 65° EUR chr5:127671182* C>G FBN2 missense p.Gly1271Ala - - 6206001 50° EUR chr5:127627260 G>T FBN2 missense p.Pro2085Thr - - 6383001 43° EUR chr5:127613647 T>C FBN2 missense p.Ile2466Val N=1/6,503 - 6418001 65° EUR chr5:127609564 A>G FBN2 missense p.Phe2603Ser N=2/6,503 - 6216001 50° EUR chr5:127607792 C>T FBN2 missense p.Gly2620Glu - - NA, not available; EUR, European; AFR, African; ESP, NHLBI Exome Sequencing Project; MFS, Marfan syndrome *Neonatal region aFrequency includes individuals of African and European ancestry

Table 3.3: Clinical characteristics of AIS cases (curve ≥10°) of European ancestry with rare variants in FBN1 and FBN2 No variant FBN1 P-value FBN2 P-value 86% 80% 83% Female P=0.35 P=0.51 (N=379) (N=15) (N=12) Spinal curve 49° ± 16 69° ± 23 58° ± 18 P=6.91×10-4 P=0.10 (avg ± sd) (N=364) (N=15) (N=12) Height percentile 55th ± 30 75th ± 31 56th ± 35 P=0.0035 P=0.41 (avg ± sd) (N=217) (N=15) (N=12) Weight percentile 57th ± 31 69th ± 30 50th ± 33 P=0.06 P=0.25 (avg ± sd) (N=217) (N=15) (N=12) Body mass index 22.2 ± 5.1 21.6 ± 4.0 20.2 ± 3.6 P=0.42 P=0.06 (avg ± sd) (N=217) (N=15) (N=12) First degree relative with 16% 33% 20% P=0.12 P=0.50 treated AIS (N=277) (N=12) (N=10) Beighton joint 1.4 ± 1.6 2.5 ± 2.8 1.0 ± 0.8 hypermobility score [5] P=0.09 P=0.45 (N=105) (N=10) (N=4) (avg ± sd) Ghent systemic features 2.4 ± 1.4 4.2 ± 2.6 3.5 ± 1.9 P=0.0072 P=0.10 score [4] (avg ± sd) (N=105) (N=10) (N=4) avg, average; sd, standard deviation

Table 3.4: Clinical features of AIS patients evaluated for Marfan syndrome

Case 6340001 6390001 6377001 6226001 6418001 6368001 6111001 6386001 6005001 Ancestry EUR AFR EUR EUR EUR EUR EUR EUR EUR Gender M F F F F F M F F Age (years) 13 14 17 18 14 10 13 15 24 Height (percentile) >99th 70th 57th 91st >99th >99th 93rd 98th 91st Weight (percentile) 99th 96th 62nd 85th 87th 88th 86th 92nd 95th Spinal curve 48° 60° 50° 97° 65° 100° 95° 93° 55° Curve type R-T R-T R-TC R-T R-T R-T L-T R-T R-T General patient information Treatment none surgery surgery surgery surgery surgery surgery surgery surgery

FBN1 variant p.N280T p.Q697R p.N703H p.L1405R p.M1576T p.N2178K p.Y2225F p.I2585T p.V2868I

FBN1 Variant reported in MFS No No No Yes Yes No No Yes No MFS diagnosis No No No No No No No Yes No Aortic root dilation - NA NA - - - - + - Ectopia lentis - NA NA NA - NA - - - Systemic features score 3/22 3/17 1/17 2/20 4/18 2/22 5/18 10/22 6/18

Pectus deformity ------Hindfoot deformity ------+ + Pes planus - + - - + - + + + Wrist/thumb sign ------+ - Upper/Lower segment 1.17 0.93 0.97 1.00 0.95 0.88 0.85 0.76 0.99 Arm span/Height 1.03 1.08 0.98 1.01 1.01 1.03 1.07 1.09 1.02 Reduced elbow extension - - - - + - - - - Dural ectasia - NA NA - NA - NA - NA Protrusio acetabuli - NA NA NA NA - NA - NA Marfan syndrome evaluation syndrome Marfan Skin striae + + - + + - + + + Facial features (3/5) ------Myopia + - - - - - + + + Pneumothorax ------Mitral valve prolapse - NA NA ------Hypermobility score 0/9 0/9 1/8 4/8 1/9 2/9 2/8 2/9 9/9 Hands flat on floor - - - NA - - - - +

L knee overextends ------+ - + R knee overextends ------+ - + L elbow bends >10° ------+ R elbow bends >10° ------+ L thumb touching - - + + - + - + + forearm R thumb touching - - NA + + + NA + +

Hypermobility evaluation Hypermobility forearm L little finger >90° - - - + - - - - + R little finger >90° - - - + - - - - + +, present; -, absent; L= left; R= right; T= thoracic; TC= thoracolumbar; EUR, European; AFR, African; MFS, Marfan syndrome; NA= not available

REFERENCES

1. Choi M, Scholl UI, Ji W, Liu T, Tikhonova IR, et al. (2009) Genetic diagnosis by whole exome capture and massively parallel DNA sequencing. Proc Natl Acad Sci U S A 106: 19096-19101. 2. Manolio TA, Collins FS, Cox NJ, Goldstein DB, Hindorff LA, et al. (2009) Finding the missing heritability of complex diseases. Nature 461: 747-753. 3. Cobb JR (1948) Outline for the study of scoliosis. American Academy of Orthopaedic Surgeons Instructional Course Lectures 5: 261-265. 4. Loeys BL, Dietz HC, Braverman AC, Callewaert BL, De Backer J, et al. (2010) The revised Ghent nosology for the Marfan syndrome. J Med Genet 47: 476-485. 5. Beighton P, Horan F (1969) Orthopaedic aspects of the Ehlers-Danlos syndrome. J Bone Joint Surg Br 51: 444-453. 6. Ramos E, Levinson BT, Chasnoff S, Hughes A, Young AL, et al. (2012) Population-based rare variant detection via pooled exome or custom hybridization capture with or without individual indexing. BMC Genomics 13: 683. 7. Bashiardes S, Veile R, Helms C, Mardis ER, Bowcock AM, et al. (2005) Direct genomic selection. Nat Methods 2: 63-69. 8. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, et al. (2009) The Sequence Alignment/Map format and SAMtools. Bioinformatics 25: 2078-2079. 9. Li B, Leal SM (2008) Methods for detecting associations with rare variants for common diseases: application to analysis of sequence data. Am J Hum Genet 83: 311-321. 10. Schneider CA, Rasband WS, Eliceiri KW (2012) NIH Image to ImageJ: 25 years of image analysis. Nat Methods 9: 671-675. 11. McCarthy MI, Abecasis GR, Cardon LR, Goldstein DB, Little J, et al. (2008) Genome-wide association studies for complex traits: consensus, uncertainty and challenges. Nat Rev Genet 9: 356-369. 12. Lenke L (2004) Idiopathic scoliosis. Philadelphia: Lippincot, Williams & Wilkins 3. 13. Sponseller PD, Hobbs W, Riley LH, 3rd, Pyeritz RE (1995) The thoracolumbar spine in Marfan syndrome. J Bone Joint Surg Am 77: 867-876. 14. Ramos Arroyo MA, Weaver DD, Beals RK (1985) Congenital contractural arachnodactyly. Report of four additional families and review of literature. Clin Genet 27: 570-581. 15. Sharma S, Gao X, Londono D, Devroy SE, Mauldin KN, et al. (2011) Genome-wide association studies of adolescent idiopathic scoliosis suggest candidate susceptibility genes. Hum Mol Genet 20: 1456-1466. 16. Putnam EA, Cho M, Zinn AB, Towbin JA, Byers PH, et al. (1996) Delineation of the Marfan phenotype associated with mutations in exons 23-32 of the FBN1 gene. Am J Med Genet 62: 233-242.

17. Faivre L, Collod-Beroud G, Loeys BL, Child A, Binquet C, et al. (2007) Effect of mutation type and location on clinical outcome in 1,013 probands with Marfan syndrome or related phenotypes and FBN1 mutations: an international study. Am J Hum Genet 81: 454-466. 18. Beroud C, Collod-Beroud G, Boileau C, Soussi T, Junien C (2000) UMD (Universal mutation database): a generic software to build and analyze locus-specific databases. Hum Mutat 15: 86-94. 19. Rommel K, Karck M, Haverich A, von Kodolitsch Y, Rybczynski M, et al. (2005) Identification of 29 novel and nine recurrent fibrillin-1 (FBN1) mutations and genotype- phenotype correlations in 76 patients with Marfan syndrome. Hum Mutat 26: 529-539. 20. Sheikhzadeh S, Kade C, Keyser B, Stuhrmann M, Arslan-Kirchner M, et al. (2012) Analysis of phenotype and genotype information for the diagnosis of Marfan syndrome. Clin Genet 82: 240-247. 21. Loeys B, Nuytinck L, Delvaux I, De Bie S, De Paepe A (2001) Genotype and phenotype analysis of 171 patients referred for molecular study of the fibrillin-1 gene FBN1 because of suspected Marfan syndrome. Arch Intern Med 161: 2447-2454. 22. Liu WO, Oefner PJ, Qian C, Odom RS, Francke U (1997) Denaturing HPLC-identified novel FBN1 mutations, polymorphisms, and sequence variants in Marfan syndrome and related connective tissue disorders. Genet Test 1: 237-242. 23. Sheikhzadeh S, Kade C, Keyser B, Stuhrmann M, Arslan-Kirchner M, et al. (2011) Analysis of phenotype and genotype information for the diagnosis of Marfan syndrome. Clin Genet. 24. Comeglio P, Johnson P, Arno G, Brice G, Evans A, et al. (2007) The importance of mutation detection in Marfan syndrome and Marfan-related disorders: report of 193 FBN1 mutations. Hum Mutat 28: 928. 25. Hung CC, Lin SY, Lee CN, Cheng HY, Lin SP, et al. (2009) Mutation spectrum of the fibrillin-1 (FBN1) gene in Taiwanese patients with Marfan syndrome. Ann Hum Genet 73: 559-567. 26. Callewaert BL, Loeys BL, Ficcadenti A, Vermeer S, Landgren M, et al. (2009) Comprehensive clinical and molecular assessment of 32 probands with congenital contractural arachnodactyly: report of 14 novel mutations and review of the literature. Hum Mutat 30: 334-341. 27. Green RC, Berg JS, Grody WW, Kalia SS, Korf BR, et al. (2013) ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing. Genet Med 15: 565-574. 28. Neptune ER, Frischmeyer PA, Arking DE, Myers L, Bunton TE, et al. (2003) Dysregulation of TGF-beta activation contributes to pathogenesis in Marfan syndrome. Nat Genet 33: 407-411. 29. Ng CM, Cheng A, Myers LA, Martinez-Murillo F, Jie C, et al. (2004) TGF-beta-dependent pathogenesis of mitral valve prolapse in a mouse model of Marfan syndrome. J Clin Invest 114: 1586-1592.

30. Matt P, Schoenhoff F, Habashi J, Holm T, Van Erp C, et al. (2009) Circulating transforming growth factor-beta in Marfan syndrome. Circulation 120: 526-532. 31. Kim KL, Yang JH, Song SH, Kim JY, Jang SY, et al. (2012) Positive Correlation Between the Dysregulation of Transforming Growth Factor-beta(1) and Aneurysmal Pathological Changes in Patients With Marfan Syndrome. Circ J. 32. Dietz HC, Cutting GR, Pyeritz RE, Maslen CL, Sakai LY, et al. (1991) Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 352: 337- 339. 33. Keane MG, Pyeritz RE (2008) Medical management of Marfan syndrome. Circulation 117: 2802-2813. 34. Ades LC, Holman KJ, Brett MS, Edwards MJ, Bennetts B (2004) Ectopia lentis phenotypes and the FBN1 gene. Am J Med Genet A 126A: 284-289. 35. Brautbar A, LeMaire SA, Franco LM, Coselli JS, Milewicz DM, et al. (2010) FBN1 mutations in patients with descending thoracic aortic dissections. Am J Med Genet A 152A: 413-416. 36. Francke U, Berg MA, Tynan K, Brenn T, Liu W, et al. (1995) A Gly1127Ser mutation in an EGF-like domain of the fibrillin-1 gene is a risk factor for ascending aortic aneurysm and dissection. Am J Hum Genet 56: 1287-1296. 37. Milewicz DM, Michael K, Fisher N, Coselli JS, Markello T, et al. (1996) Fibrillin-1 (FBN1) mutations in patients with thoracic aortic aneurysms. Circulation 94: 2708-2711. 38. Lemaire SA, McDonald ML, Guo DC, Russell L, Miller CC, 3rd, et al. (2011) Genome-wide association study identifies a susceptibility locus for thoracic aortic aneurysms and aortic dissections spanning FBN1 at 15q21.1. Nat Genet 43: 996-1000. 39. Sood S, Eldadah ZA, Krause WL, McIntosh I, Dietz HC (1996) Mutation in fibrillin-1 and the Marfanoid-craniosynostosis (Shprintzen-Goldberg) syndrome. Nat Genet 12: 209- 211. 40. Faivre L, Gorlin RJ, Wirtz MK, Godfrey M, Dagoneau N, et al. (2003) In frame fibrillin-1 gene deletion in autosomal dominant Weill-Marchesani syndrome. J Med Genet 40: 34- 36. 41. Le Goff C, Mahaut C, Wang LW, Allali S, Abhyankar A, et al. (2011) Mutations in the TGFbeta binding-protein-like domain 5 of FBN1 are responsible for acromicric and geleophysic dysplasias. Am J Hum Genet 89: 7-14. 42. Loeys BL, Gerber EE, Riegert-Johnson D, Iqbal S, Whiteman P, et al. (2010) Mutations in fibrillin-1 cause congenital scleroderma: stiff skin syndrome. Sci Transl Med 2: 23ra20. 43. Rybczynski M, Bernhardt AM, Rehder U, Fuisting B, Meiss L, et al. (2008) The spectrum of syndromes and manifestations in individuals screened for suspected Marfan syndrome. Am J Med Genet A 146A: 3157-3166. 44. Putnam EA, Zhang H, Ramirez F, Milewicz DM (1995) Fibrillin-2 (FBN2) mutations result in the Marfan-like disorder, congenital contractural arachnodactyly. Nat Genet 11: 456- 458.

45. Ades LC, Sreetharan D, Onikul E, Stockton V, Watson KC, et al. (2002) Segregation of a novel FBN1 gene mutation, G1796E, with kyphoscoliosis and radiographic evidence of vertebral dysplasia in three generations. Am J Med Genet 109: 261-270. 46. Milewicz DM, Grossfield J, Cao SN, Kielty C, Covitz W, et al. (1995) A mutation in FBN1 disrupts profibrillin processing and results in isolated skeletal features of the Marfan syndrome. J Clin Invest 95: 2373-2378. 47. Hadley-Miller N, Mims B, Milewicz DM (1994) The potential role of the elastic fiber system in adolescent idiopathic scoliosis. J Bone Joint Surg Am 76: 1193-1206. 48. Miller NH, Mims B, Child A, Milewicz DM, Sponseller P, et al. (1996) Genetic analysis of structural elastic fiber and collagen genes in familial adolescent idiopathic scoliosis. J Orthop Res 14: 994-999. 49. Lipton GE, Guille JT, Kumar SJ (2002) Surgical treatment of scoliosis in Marfan syndrome: guidelines for a successful outcome. J Pediatr Orthop 22: 302-307. 50. Sponseller PD, Bhimani M, Solacoff D, Dormans JP (2000) Results of brace treatment of scoliosis in Marfan syndrome. Spine (Phila Pa 1976) 25: 2350-2354. 51. Danielsson AJ, Hasserius R, Ohlin A, Nachemson AL (2007) A prospective study of brace treatment versus observation alone in adolescent idiopathic scoliosis: a follow-up mean of 16 years after maturity. Spine (Phila Pa 1976) 32: 2198-2207. 52. Sanders JO, Newton PO, Browne RH, Herring AJ (2012) Bracing in adolescent idiopathic scoliosis, surrogate outcomes, and the number needed to treat. J Pediatr Orthop 32 Suppl 2: S153-157. 53. Weinstein SL, Dolan LA, Wright JG, Dobbs MB (2013) Effects of bracing in adolescents with idiopathic scoliosis. N Engl J Med 369: 1512-1521. 54. Habashi JP, Judge DP, Holm TM, Cohn RD, Loeys BL, et al. (2006) Losartan, an AT1 antagonist, prevents aortic aneurysm in a mouse model of Marfan syndrome. Science 312: 117-121. 55. Brooke BS, Habashi JP, Judge DP, Patel N, Loeys B, et al. (2008) Angiotensin II blockade and aortic-root dilation in Marfan's syndrome. N Engl J Med 358: 2787-2795. 56. Groenink M, den Hartog AW, Franken R, Radonic T, de Waard V, et al. (2013) Losartan reduces aortic dilatation rate in adults with Marfan syndrome: a randomized controlled trial. Eur Heart J 34: 3491-3500. 57. Detaint D, Faivre L, Collod-Beroud G, Child AH, Loeys BL, et al. (2010) Cardiovascular manifestations in men and women carrying a FBN1 mutation. Eur Heart J 31: 2223-2229. 58. Davis MR, Summers KM (2012) Structure and function of the mammalian fibrillin gene family: implications for human connective tissue diseases. Mol Genet Metab 107: 635- 647. 59. Le Goff C, Cormier-Daire V (2012) From tall to short: the role of TGFbeta signaling in growth and its disorders. Am J Med Genet C Semin Med Genet 160C: 145-153.

60. Nistala H, Lee-Arteaga S, Smaldone S, Siciliano G, Carta L, et al. (2010) Fibrillin-1 and -2 differentially modulate endogenous TGF-beta and BMP bioavailability during bone formation. J Cell Biol 190: 1107-1121. 61. Maleszewski JJ, Miller DV, Lu J, Dietz HC, Halushka MK (2009) Histopathologic findings in ascending aortas from individuals with Loeys-Dietz syndrome (LDS). Am J Surg Pathol 33: 194-201. 62. Loeys BL, Chen J, Neptune ER, Judge DP, Podowski M, et al. (2005) A syndrome of altered cardiovascular, craniofacial, neurocognitive and skeletal development caused by mutations in TGFBR1 or TGFBR2. Nat Genet 37: 275-281. 63. Erkula G, Sponseller PD, Paulsen LC, Oswald GL, Loeys BL, et al. (2010) Musculoskeletal findings of Loeys-Dietz syndrome. J Bone Joint Surg Am 92: 1876-1883. 64. Ryzhkov, II, Borzilov EE, Churnosov MIP, Ataman AVP, Dedkov AA, et al. (2013) Transforming Growth Factor Beta 1 is a Novel Susceptibility Gene for Adolescent Idiopathic Scoliosis. Spine (Phila Pa 1976). 65. Yim AP, Yeung HY, Hung VW, Lee KM, Lam TP, et al. (2012) Abnormal skeletal growth patterns in adolescent idiopathic scoliosis--a longitudinal study until skeletal maturity. Spine (Phila Pa 1976) 37: E1148-1154. 66. Czaprowski D, Kotwicki T, Pawlowska P, Stolinski L (2011) Joint hypermobility in children with idiopathic scoliosis: SOSORT award 2011 winner. Scoliosis 6: 22. 67. Gurnett CA, Alaee F, Bowcock A, Kruse L, Lenke LG, et al. (2009) Genetic linkage localizes an adolescent idiopathic scoliosis and pectus excavatum gene to chromosome 18 q. Spine (Phila Pa 1976) 34: E94-100. 68. Hong JY, Suh SW, Park HJ, Kim YH, Park JH, et al. (2011) Correlations of adolescent idiopathic scoliosis and pectus excavatum. J Pediatr Orthop 31: 870-874. 69. Abul-Kasim K, Overgaard A, Ohlin A (2010) Dural ectasia in adolescent idiopathic scoliosis: quantitative assessment on magnetic resonance imaging. Eur Spine J 19: 754-759. 70. Isogai Z, Ono RN, Ushiro S, Keene DR, Chen Y, et al. (2003) Latent transforming growth factor beta-binding protein 1 interacts with fibrillin and is a microfibril-associated protein. J Biol Chem 278: 2750-2757.

Chapter 4:

Kinesin Family Member 6 (kif6) is Necessary for Normal Vertebral Development in Zebrafish

SUMMARY

Idiopathic scoliosis is a common condition in humans that causes spinal deformity in 2-

3% of the at risk population. The pathogenesis of idiopathic scoliosis remains poorly

understood, in part due to the lack of a relevant animal model. To identify a zebrafish model for

idiopathic scoliosis, we performed a forward mutagenesis screen and identified skolios, a

recessive zebrafish mutant with a spinal curvature phenotype. skolios develops an isolated spinal

deformity phenotype that parallels certain features of human idiopathic scoliosis, including a

medial-lateral curvature of the spine that arises independent of vertebral malformations. To

identify the genetic basis of skolios, we used meiotic mapping to narrow the locus to a 2.7 Mb

region on chromosome 17. Low-coverage whole genome sequencing was then used to identify

candidate mutations within the mapped locus. Twenty homozygous nonsynonymous mutations

were identified in the locus. Only one nonsynonymous change resulted in a nonsense mutation,

which caused a premature stop in kinesin family member 6 (kif6), a poorly characterized kinesin

of unknown function. To confirm that loss of kif6 is responsible for the skolios phenotype, we created additional mutant alleles using transcription activator-like effector nucleases (TALENs).

Three new TALEN-induced mutations causing frameshift mutations in kif6 were identified. All

zebrafish homozygous or compound heterozygous for kif6 frameshift mutations developed a

scoliosis phenotype that was indistinguishable from skolios mutants, confirming kif6 as the

causative gene. Although kif6 may play a role in cilia, no evidence for cilia dysfunction was

seen in skolios mutants. Overall, these findings demonstrate a novel role for kif6 in the

developing spine and identify a new candidate gene for human idiopathic scoliosis.

INTRODUCTION

Idiopathic scoliosis is a multifaceted genetic disorder that causes curvature of the spine in

2-3% of humans, with about 10% of patients progressing to deformity warranting bracing or

surgical treatment. Although idiopathic scoliosis is defined by a deviation of the spine from the

midline in the medial-lateral plane, curvature is typically complex and causes a three

dimensional deformity affecting all three planes (medial-lateral, dorsal-ventral and transverse).

Unlike secondary scoliosis that arises from congenital vertebral malformations or other

syndromic conditions, idiopathic scoliosis is poorly understood and no obvious structural

vertebral abnormalities are present. A major challenge in determining the pathogenesis of

idiopathic scoliosis has been the lack of a relevant animal model. Animal models of idiopathic

scoliosis have been primarily limited to experimental forms of scoliosis, where scoliosis is

induced by invasive surgical procedures, immobilization or the use of the systemic agents [1].

Because spinal deformity develops secondary to harsh interventions, experimentally induced

forms of scoliosis are unlikely to recapitulate many features of human idiopathic scoliosis.

With the exception of humans, naturally occurring scoliosis is relatively rare [2].

Humans ambulate in an upright position with their center of gravity over the pelvis, a

characteristic that is unique even among other primates [3]. This erect, bipedal posture

significantly alters the dorsal shear loads of the spine, which has been proposed as an important

prerequisite for scoliosis development [4]. In support of this model, pinealectomized rats

develop scoliosis when also made bipedal, whereas similarly treated quadrupedal rats do not [5].

Although pinealectomy and bipedalism may be important prerequisites for scoliosis in some species, pinealectomy in bipedal nonhuman primates does not cause spinal deformity [6].

Moreover, pinealectomy in guppy [7] and salmon [8] induces spinal curvature similar to bipedal

rats, demonstrating that scoliosis induced by pinealectomy does not require bipedalism.

A mutant guppy strain, curveback, is one of the few examples of an animal model with non-induced spinal deformity that parallels human idiopathic scoliosis [7]. curveback mutants develop a spinal curvature sporadically in the dorsal-ventral and medial-lateral plane that is not due to vertebral malformations. Sporadic spinal deformity has also been noted in other teleosts, including medaka and swordtail, suggesting that teleosts may be ideal animal models to identify and characterize naturally occurring scoliosis [9]. Of the teleosts, zebrafish (Danio rerio) are particularly well-suited model organisms, as zebrafish are highly tractable and have extensive genomic resources. Scoliosis in zebrafish has been largely unexplored. The zebrafish mutant leviathan is caused by recessive mutations in collagen type VIII alpha1a (col8a1a) and develops scoliosis, although mutants also develop notochord defects and congenital vertebral malformations [10]. It remains unclear if zebrafish are susceptible to naturally occurring scoliosis without vertebral malformations. Therefore, we sought to identify novel zebrafish scoliosis mutants that could serve as a model for human idiopathic scoliosis. Using an N-ethyl-

N-nitrosourea (ENU) mutagenesis forward genetic screen, we identify and describe skolios, a zebrafish mutant with recessively inherited scoliosis that shares several characteristics with human idiopathic scoliosis.

100

METHODS

Zebrafish maintenance

Zebrafish were maintained under standard conditions [11]. Zebrafish were bred by in

vitro fertilization using adults anesthetized in 0.16% tricaine methanesulfonate (3-amino benzoic acidethylester, Sigma-Aldrich) diluted in E3 embryo medium (5 mM NaCl, 0.17 mM KCl, 0.33 mM CaCl2, MgSO4 in H20) or by natural pairwise mating. With exception to WIK strains used

for meiotic mapping, all wild-type (WT) zebrafish were from an AB genetic background.

Experiments were carried out in accordance with the animal protocol guidelines at Washington

University.

Chemical Mutagenesis and Forward Genetic Screening

skolios was identified from an ENU mutagenesis screen performed on WT (AB)

zebrafish. The generation of mutagenized F1 female fish was carried out with ENU as previously described [12]. Gynogenetic diploids were subsequently derived by the early pressure method [13] and screened for spinal curvature.

Analysis of spinal and vertebral morphology

MicroCT imaging was performed using a VivaCt40 machine (ScanCo) as previously

described [14]. Zebrafish observations were imaged using an Olympus DP70 Digital

Microscope Camera fixed to an Olympus MVX10 MacroView fluorescence microscope with

MicroSuite Basic Edition software (Olympus). For bone and cartilage staining, zebrafish at

various time points were euthanized with tricaine methanesulfonate and fixed in 4% PFA/PBS.

101

Fixed zebrafish were immersed in 0.1% alcian blue in 7:3 EtOH:glacial acetic acid for 24-

48 hours and destained for 0.5-2 hours in 7:3 EtOH/glacial acetic acid to visualize cartilage.

Bones were then stained with alizarin red S in 0.5% KOH for 6-24 hours and destained in 0.5%

KOH. Stained zebrafish were stored in glycerol. The spinal curvature was quantified by imaging zebrafish positioned laterally. For consistency, spinal curvature was only quantified near the transition from abdominal (rib vertebrae) to caudal vertebrae (vertebrae posterior of ribs), which excluded the most anterior and posterior vertebrae. The apex of the curve was identified within the defined region and lines were drawn through the center of the two vertebrae

adjacent to the apex. The angle of the two intersecting lines was measured using Photoshop CS3

(Adobe). Measured angles were subtracted from 180° to calculate the spinal curve.

Meiotic mapping and whole genome sequencing

The skolios mutation was localized to chromosome 17 using centromeric linkage analysis

[10,15]. Fine mapping was achieved by mapcrossing to the polymorphic WIK strain and assessing for recombination along chromosome 17 by sequence length polymorphism analysis

[16]. For whole genome sequencing, genomic DNA from a skolios mutant was extracted using

Proteinase K (Roche) and 25:24:1 phenol:chloroform:isoamyl (Roche) according to the manufacturer’s instructions. Whole genome sequencing was performed by the Genome

Technology Access Center at Washington University. One lane of a HiSeq 2000 sequencer

(Illumina) was used to generate 101 bp paired-end reads. Raw sequencing reads were aligned to the Zv9 genome assembly using Novoalign (Novocraft Technologies). Variants were identified using SAMtools software [17] and annotated using SnpEff software [18]. Variants were

102 validated by Sanger sequencing using an ABI 3730 Sequencer (Life Technologies) and visualized using Sequencher DNA Sequencing software (Gene Codes).

RT-PCR of adult zebrafish tissues

Tissue samples were harvested from adult zebrafish euthanized in tricaine methanesulfonate. Tissue samples were dissected and pooled from 5-10 zebrafish and total RNA was extracted using Trizol Reagent (Invitrogen). RNA concentrations were measured using a

NanoDrop 2000 UV-Vis spectrophotometer. cDNA synthesis was performed using SuperScript

II Reverse Transcriptase (Invitrogen) with 1 µg RNA according to the manufacturer’s instructions. As a negative control, cDNA synthesis reactions were also performed with the reverse transcriptase omitted. PCR reactions were performed with Bullseye Taq Polymerase

(Midsci) using 1 µl cDNA and 35-40 amplification cycles with 58° annealing temperature and 30 second extension. The following primers were used:

kif6 (ex1-3) 5’- CGTGCAGGTAAACAGTAAAAGC -3’

5’- CAGCAACAGGTTTTGCAATG -3’

kif6 (ex16-18) 5’-GCGGAGAGAACACAGGAAAC-3’

5’-GTGCTGTTGAGCAGGTTGTG-3’

β-actin 5’-TACAATGAGCTCCGTGTTGC-3’

5’-AAGGAAGGCTGGAAGAGAGC-3’

Microinjection of morpholinos

Morpholinos (MOs) were purchased from Gene Tools, LLC to transiently knock-down zebrafish kif6. MOs were designed to target the translation start site (trans-block) and the

103

boundary between exon 1 and intron 1 (splice-block). An additional MO targeting the

translation start site but containing a 5 base mismatch served as a negative control (mismatch).

The 3’ end of the splice-blocking and mismatch MOs were modified with a fluorescein tag. The

sequences of the MOs are as follows:

trans-block 5’-TTGCTTTTACTGTTTACCTGCACGT-3’

splice-block 5’-ACAAAAGCAAAACACTCACCGAGGT-3’

mismatch 5’-TGCTTAACGATCTTAATCATCGCTT-3’

MOs were prepared at concentrations ranging from 0.05 mM to 3 mM and approximately 500 pl

were injected in 1-4 cell stage embryos.

Transcription activator-like effector nuclease (TALEN) design and assembly

Two TALENs (left and right) were designed to target the second exon of kif6. These

sites flanked a BccI restriction site that was used for mutation screening. TALENs were

generated using the Golden Gate method as previously described [19]. Plasmids were acquired

from the Golden Gate TALEN and TAL Effector Kit (AddGene) and RVD repeat arrays were cloned into pCS2TAL3DD and pCS2TAL3RR (AddGene). Plasmids were transformed into

DH5α competent cells (Invitrogen) and isolated using the QIAprep Spin Miniprep Kit (Qiagen).

The following sequence targets and RVD sites were used to generate kif6 TALENs:

Left target DNA TCCCTCTTATTGTTG

Left RVD sequence NG HD HD HD NG HD NG NG NI NG NG NN NG NG NN

Right target DNA GCATCTCTGGGAACC

Right RVD sequence NN NN NG NG HD HD HD NI NN NI NN NI NG NN HD

104

Injection of TALEN RNA in zebrafish embryos

Complete plasmids (containing the scaffold and RVD arrays) were linearized using XhoI

and 5’-capped mRNA was generated by in vitro transcription using the mMESSAGE mMACHINE SP6 Transcription Kit (Life Technologies). Capped mRNA was purified using the

RNeasy Mini Kit (Qiagen). Left and right TALENs were combined at equal concentrations.

WT embryos were collected and 40-50 pg of pooled TALENs were injected into the yolk of 1-4

cell stage embryos. Injected embryos and progeny were screened for TALEN-induced mutations

by digesting PCR amplified DNA with BccI (New England Biolabs). The following primers

were used for screening the BccI restriction site:

BccI 5’TGCAATAAATGGAAACAAGAACC-3’

5’-TTCGCACATTTGTTCAGTGAC-3’

Evaluation of cilia and cilia-related phenotypes

Defects in left-right asymmetry were determined by visualizing the position of the heart

in 30 hpf embryos. Abnormal left-right asymmetry was classified as ambiguous when the heart

was positioned at the midline and as situs inversus when the heart was positioned left of the

midline when viewed ventrally (the right side of the embryo). Otoliths were evaluated by

visualizing zebrafish laterally between 48-72 hpf. Immunofluorescence staining was performed

as previously described [20] using embryos collected at 28 hours post-fertilization (hpf).

Embryos were dechorionated in protease type XIV (Sigma) and fixed in 4% paraformaldehyde for immunostaining with anti-acetylated tubulin monoclonal antibody 6-11-B1 (Sigma). Fluid flow in the central canal was measured as previously stated [21]. Live 48 hpf zebrafish were

105

anesthetized in tricaine methanesulfonate and imaged at various time points after injection of 5%

tetramethylrhodamine conjugated to 70,000 molecular weight dextran into the brain ventricle.

Statistical analysis

A one-tailed Student’s t-test was used for all statistical analyses.

106

RESULTS

Identification of skolios

To identify zebrafish mutants with scoliosis, we performed an ENU mutagenesis screen

on wild-type (WT) AB zebrafish. From this screen, we identified a new zebrafish mutant,

skolios, which develops recessively inherited curvature of the body axis.

Body curvature first develops in skolios mutants during embryonic hatching stages (2-3

days post-fertilization [dpf]), causing a ventral curvature of the body axis that is frequently

limited to the distal end of the tail (Figure 4.1A). This early curvature arises without observable

defects in the notochord (e.g. kinking) and is incompletely penetrant with variable expressivity.

Approximately 20% of skolios embryos appear normal at 3 dpf, while the majority exhibit either

moderate (~70%) or severe (~10%) ventral body curvature (Figure 4.1A, Figure 4.1B).

Unlike the incompletely penetrant ventral body curvature first seen during hatching

stages, curvature in the medial-lateral plane becomes fully penetrant during early larval stages

(4-5 dpf) and progresses through later larval and juvenile stages (5-89 dpf) (Figure 4.1C).

Although there is large phenotypic variability in the appearance and severity of the curve, body

curvature defects are easily observed in both the dorsal-ventral and medial-lateral planes of all

adult skolios mutants (Figure 4.1D). Despite the severity of the deformity, adult skolios mutants

are both viable and fertile. Heterozygous mutants exhibit no overt embryonic or post-embryonic

phenotype.

107

Characterization of spinal curvature in skolios mutants

We performed microCT imaging of a skolios mutant and WT zebrafish at 50 dpf to determine if vertebral abnormalities underlie the curved body axis in skolios. Compared to WT,

the skolios mutant developed a marked curvature of the abdominal and caudal spine in both the dorsal-ventral and medial-lateral planes (Figure 4.1E). To evaluate the temporal onset of

curvature and to screen for vertebral defects, we performed a two-staining protocol using alcian blue, a positively charged dye that is thought to react with the acidic mucopolysaccharides in cartilage, and alizarin red, a dye that binds to mineralized bone [22]. Vertebral morphology was evaluated during several developmental time points for WT and skolios mutants. To track the progression of the spinal curve, curves were measured in the dorsal-ventral plane as shown in

Figure 4.2A. Although human idiopathic scoliosis is measured in the medial-lateral plane, individual zebrafish vertebrae were more easily observed when viewed laterally, which allowed for more accurate curve measurements. All curves were measured at the curve apex near the transition from abdominal to caudal vertebrae. By two months post-fertilization (mpf), the average spinal curve was similar in both male and female skolios mutants (Figure 4.2B). In males, spinal curvature was progressive and by 18 mpf, the average curve was 46° (SD=15.3°), compared to an average of 28° (SD=14.1°) at 2 mpf (P=0.043, Student’s t-test). Although the curves were similar at 2 mpf, males developed more severe spinal curves than females at both 7 mpf (36° versus 25°, P=0.042, Student’s t-test) and 18 mpf (46° versus 27°, P=0.037, Student’s t-test). Female curves were stable and did not change between 2 and 18 mpf. In contrast, curves measured in WT zebrafish were consistently small, although the spinal curves in males and females were highly variable. WT zebrafish measured at 2 and 7 mpf had an average spinal curve of 5.3° (SD=6.7°, N=8) in males and 3.4° (SD=2.1°, N=9) in females (not shown).

108

Spinal curvature in skolios mutants occurred in the absence of vertebral fusions or

malformations (Figure 4.2C). Vertebral fusions and malformations were noted in approximately

30% of all skolios mutants, which was comparable to frequencies observed in WT zebrafish

(Figure 4.3).

Meiotic mapping

To localize the gene for skolios, we performed meiotic mapping and narrowed the locus to a 2-5 Mb region prior to whole genome sequencing. Meiotic mapping localized the mutation to a region bounded by markers z8980 on one side and either z9692 or z24344 on the other

(z9692 but not z24344 was polymorphic in one WIK fish used for a mapcross, while z24344 but not z9692 was polymorphic in a second WIK fish used for a mapcross) (Figure 4.4A). The marker z24344 is not currently listed in Ensembl Zv9 but the sequence can be found in GenBank

(G47401.1) and can subsequently be located in Zv9. The smaller region, between z8980 and z9692, mapped the phenotype to a 2.7 Mb locus at chr17:47,500,198-50,216,501 (Zv9).

Whole genome sequencing

Although mapping limited the region harboring the mutation responsible for skolios to a

2.7Mb region on chromosome 17, approximately 20 genes remained within the locus. To

simultaneously investigate all genes within the interval, low-coverage whole genome sequencing

was performed on a single skolios mutant. Next generation sequencing reads were aligned to the

Zv9 Tübingen reference genome (Table 4.1). Nearly three-quarters of the reference genome was

covered at ≥3X, which we deemed sufficient to identify a homozygous base change. In the

mapped interval, 18,941 total variants were identified. To eliminate strain-specific

109

polymorphisms resulting from the alignment of skolios (generated on an AB background) to the

Zv9 sequence (Tübingen), variants that were present in sequence data from an unrelated zebrafish with a similar genetic background (AB) were removed. Only 20 of the remaining variants were homozygous and caused a nonsynonymous amino acid change. Among these, a single nonsense mutation was identified, causing a premature stop in kif6 (encoding kinesin family member 6). The G>T transversion at chr17:48958292 (Zv9) occurs in the second exon of kif6, which would result in a C>A change in the kif6 mRNA (kif6 is encoded on the minus strand) and produce premature stop early in the kif6 protein (Figure 4.4B). This variant was validated in an independent skolios mutant using Sanger sequencing (Figure 4.4C).

TALEN-induced mutations in kif6

Transient knock-down of kif6 with morpholinos targeting the translation start site or the first exon-intron boundary recapitulated the embryonic curvature defects (Figure 4.5); however, kif6 morphants developed into normal adults (not shown). Because the transient nature of morpholinos limits our ability to assess adult phenotypes, we sought to generate stable germline mutations in kif6 using TALENs (transcription activator-like effector nucleases), which have been shown to effectively induce new mutations in zebrafish [23]. TALENs are engineered proteins that can be designed to target a specific DNA sequence and introduce double-stranded breaks that are repaired through the error-prone nonhomologous end-joining pathway, thus introducing mutations near the targeted sequence [24].

We designed a TALEN pair targeting a region in the second exon of kif6, which flanked a

BccI restriction site and was adjacent to the G>T transversion at chr17:48958292 (Zv9) identified in skolios (Figure 4.6A). TALENs were combined at equal concentrations and

110 injected into 1-4 cell stage WT embryos. PCR amplification and subsequent digestion with BccI showed that all embryos injected with TALENs had some degree of undigested product, indicating that the BccI restriction site was disrupted by an induced mutation (Figure 4.6B).

Because the F0 (injected) zebrafish are mosaic, zebrafish were outcrossed with WT adults to create stable lines. Three new mutations were identified in the F1 generation: two different 8 bp deletions and a 4 bp deletion (Figure 4.6C). Sibling crosses generated F2 zebrafish that were either homozygous or compound heterozygous for TALEN-induced kif6 mutant alleles. All combinations produced zebrafish with curvature of the body axis in both embryonic and adult stages that were indistinguishable from skolios, strongly supporting kif6 as the genetic cause of skolios (Figure 4.7).

Characterization of kif6 in zebrafish

We performed RT-PCR at various embryonic time points and adult tissues to determine kif6 expression in WT zebrafish (Figure 4.8). In embryos, kif6 was expressed at low levels starting at 24 hpf, which were maintained through 120 hpf. In adult tissues, kif6 was expressed at low levels in brain, intestine, ovary and testis.

Assessment of cilia-related phenotypes in skolios

The function of kif6 is unknown, but it has been predicted to be involved in cilia/flagellum assembly or function [20,25,26], particularly in ciliated ependymal cells [27].

Several characteristic phenotypes are common in zebrafish with cilia defects (reviewed in

[28,29]), including ventral curvature of the embryo, left-right asymmetry defects, kidney cysts, hydrocephalus and otolith abnormalities. Because skolios mutants develop embryonic body

111

curvature, one of several cilia-related phenotypes, we wondered whether kif6 might be involved

in ciliary assembly/function. To address this we assessed skolios mutants for additional

ciliopathy-related phenotypes. To determine if skolios mutants exhibit an increased incidence of

left-right asymmetry defects, we evaluated the position of the heart in WT and skolios embryos

at 30 hpf (Figure 4.9A). Normal heart positioning was identified in 93.6% of WT embryos

(N=94) and 91.9% of skolios embryos (N=91.9%), suggesting that skolios is not associated with

an increased incidence of left-right asymmetry defects. Similarly, otoliths were evaluated and

WT (N=391) and skolios mutants (N=305) at 48-72 hpf and appeared normal in 98.5% and

98.7%, respectively (Figure 4.9B). Other easily observed phenotypes that are characteristic of

cilia defects, such as hydrocephalus and kidney cysts, were never observed in skolios.

To more closely examine cilia in skolios, we evaluated cilia structure and function in the

central canal. To evaluate gross ciliary structure, we employed whole-mount confocal immunostaining using an antibody labelling acetylated tubulin. No gross abnormalities in ciliary number or morphology were detected in 28 hpf embryos compared to WT (Figure 4.9C). To evaluate ciliary function, we assessed the movement of dye down the central canal in 48 hpf WT and skolios embryos using a previously described assay [21]. No differences in dye movement were observed in skolios mutants compared to WT (Figure 4.9D).

112

DISCUSSION

Despite the high frequency of idiopathic scoliosis in humans, the pathogenesis remains

elusive. Here, we describe a new zebrafish mutant called skolios which develops spinal

deformity that parallels many aspects of human idiopathic scoliosis.

First, skolios mutants develop a complex, three-dimensional curvature of the spine

without the presence of structural vertebral abnormalities. In humans, idiopathic scoliosis is

diagnosed only after other causes of scoliosis have been ruled out, including congenital vertebral

defects or other syndromic diagnoses that can cause scoliosis (e.g. cerebral palsy, Marfan

syndrome). Vertebral defects were noted in some skolios mutants but were present at the same frequency in WT, suggesting curvature is unlinked to these vertebral abnormalities in skolios

mutants. Moreover, spinal curvature is the only phenotype observed in skolios mutants,

suggesting that spinal curvature is not occurring secondary to a larger, systemic condition.

Second, like human idiopathic scoliosis, we noted sexual dimorphism in skolios for the

most severe curves. In humans, mild idiopathic scoliosis affects males and females equally, but females develop severe deformity ten times more frequently than males [30]. In skolios, both

genders were affected, but over time, adult male mutants developed a more severe spinal

deformity compared to adult females. Although this is opposite to the female bias observed in humans, it nevertheless demonstrates a role for gender in the progression of spinal deformity but not for overall susceptibility.

Using meiotic mapping and low-coverage whole genome sequencing, we identified a nonsense mutation in kif6, a poorly characterized member of the kinesin family, as a candidate mutation for skolios. Our results, along with others [13,31-33], demonstrate that low-coverage

113

whole genome sequencing is an effective approach to identify the genetic basis of a zebrafish

mutant. kif6 was confirmed as the causative gene by creating additional mutant alleles using

TALENs that recapitulated both the embryonic and post-embryonic skolios phenotypes.

Kinesins are motor proteins that have diverse function, but can be identified by their highly conserved motor domain that binds to microtubules and powers movement through the hydrolysis of ATP. Using degenerate primers designed to this conserved motor domain, kif6 was first identified in cDNA from mouse hippocampus [34]. In a large mouse cDNA library, kif6 was identified in spinal ganglion and cerebellum [35] and western blotting of mouse tissues has shown kif6 expression in various tissues, including brain and spinal cord [26]. Our expression analysis in zebrafish demonstrated similar results, with kif6 expressed in zebrafish brain, intestine, ovary and testis. To our knowledge, this is the first demonstration of kif6 expression in zebrafish and the first demonstration of a mutant phenotype resulting from a mutation in kif6.

Kif6 is poorly characterized and its function is unknown. Most investigations of kif6 have been limited to a common single nucleotide polymorphism (rs20455) in humans, which was associated with coronary artery disease [36-40], although a large meta-analysis failed to replicate the association [41] and the consequence of the common polymorphism in kif6 is currently being debated. A kif6 mutant mouse was found to have normal cardiac function, while additional phenotypes were not described [26]. Kif6 was originally identified as an orphan kinesin [35,42], but has also been included with kinesin-family member 9 (kif9) in the Kinesin-9 superfamily.

Although kif9 and kif6 share a region of conserved sequence near the kinesin motor domain that is specific to the Kinesin-9 superfamily [25], the proteins are phylogenetically distant [42].

Members of the Kinesin-9 superfamily have only been identified in vertebrates and protozoa, which suggests they may be involved in cilia and flagella [25].

114

Further evidence for a role of kif6 in cilia was suggested by its downregulation in

forebrain of forkhead box J1 (FoxJ1)-null mice [27], a transcription factor that is necessary and sufficient for motile cilia generation [43,44]. FoxJ1 is also required for the differentiation of

ependymal cells [17], which are ciliated cells that line the ventricles in the brain and the central

canal in the spinal cord. Ependymal cells are responsible for circulating cerebral spinal fluid,

although they may have additional functions. In the Allen Brain Atlas (http://mouse.brain- map.org/), kif6 expression is limited to the ependymal layer of the ventricle [17] and central

canal of the spinal cord. We did not find gross abnormalities in cilia formation or movement in

the central canal and we did not identify additional cilia related phenotypes in skolios mutants

(defects in left-right asymmetry, otolith abnormalities, hydrocephalus or kidney cysts), however

these findings cannot rule out other cilia-related functions for kif6. Curvature of the body axis is

a common phenotype in zebrafish embryos with cilia defects [28] and has also been observed in

adult cilia mutants [45], which is consistent with a cilia-related defect in skolios. Cilia defects

have been proposed as a mechanism for human idiopathic scoliosis [46], though scoliosis is not a

common in human ciliopathies. Future studies are needed to clarify the function of kif6 and the

mechanism by which it leads to spinal curvature in skolios. It is unclear whether kif6 mutations

are important in the genesis of human idiopathic scoliosis. Efforts aimed at identifying kif6

mutations in a DNA database of pediatric patients with musculoskeletal disorders, including

scoliosis, are currently underway (Christina Gurnett, personal communication).

Overall, this study identifies skolios, a zebrafish mutant with spinal deformity. The

spinal curvature phenotype, which occurs isolated from other features and is not secondary to

vertebral malformations suggests that skolios may serve as a model for human idiopathic

scoliosis, which could be a valuable resource to better understand this complex human condition.

115

Moreover, we have established zebrafish as a viable model for spinal deformity that may be of use to confirm genetic findings in human studies of idiopathic scoliosis. Using meiotic mapping and whole genome sequencing, we identify kif6 as genetic basis of skolios. These finding highlight a new role for kif6 in spinal development and stability, expanding upon the already diverse cellular functions of kinesins.

116

Figure 4.1: Skolios mutants develop recessively-inherited curvature of the body axis (A) Ventral curvature phenotypes observed in skolios embryos at 3 dpf. (B) Relative frequencies of ventral curvature phenotypes in WT and skolios embryos (3 dpf). (C) Dorsal view of embryonic (4 dpf) and juvenile (17 dpf) zebrafish showing development of sagittal body curvature in skolios mutants. (D) Severe body curvature in both the medial-lateral and dorsal- ventral planes of an adult skolios mutant. (E) MicroCT imaging at approximately 50 dpf shows marked curvature of the abdominal (rib vertebrae) and caudal (vertebrae posterior to ribs) spine in skolios mutants.

117

Figure 4.2: Spinal curvature in skolios mutants occurs independent of major vertebral abnormalities and is progressive through adult stages in males (A) Method of measuring spinal curve in zebrafish using skeletal histomorphology. Lines were drawn through the middle of the two vertebrae adjacent to the apex of the curve. The measured angle from the intersection was subtracted from 180° to calculate the spinal curve. (B) Spinal curves measured for female skolios mutants at 2 (N=11), 5 (N=9), 7 (N=5) and 18 (N=5) mpf and male skolios mutants at 2 (N=10), 5 (N=8), 7 (N=5) and 18 (N=5) mpf. The average spinal curve is shown for each time point. At 2 mpf, males and females are similarly affected; however, by later stages, males continue to progress in severity. *P<0.05. (C) Skeletal histomorphology of a 2 mpf female WT and a skolios mutant with a 35° spinal curve near the transition from abdominal to caudal vertebrae that developed in the absence of vertebral fusions or other major vertebral abnormalities. Box identifies region shown in A.

118

Figure 4.3: Vertebral fusions and malformations occur at similar frequencies in skolios and WT zebrafish (A) Zebrafish between 2-7 mpf were stained with Alizarin red to visualize individual vertebrae. Staining shows vertebral fusions (arrows) identified in WT and skolios. (B) The overall frequency of vertebral fusions in skolios is similar to WT.

119

Figure 4.4: Meiotic mapping and whole genome sequencing identifies a nonsense mutation in kif6 as candidate for skolios (A) Summary of mapping experiments. Two mapcrosses identified a ~3 Mb region on chromosome 17 harboring the causative genetic lesion for skolios. (B) Gene and protein schematic of kif6 with the corresponding location of the nonsense mutation. The G>T mutation (C>A in mRNA) in the second exon of kif6 causes a premature stop early in the kif6 protein sequence. Kif6 protein domains are drawn according to the simple modular architecture research tool (SMART) [47]. The kinesin motor domain is shown, with grey bars indicating coiled-coil regions. (C) Validation of the G>T mutation in skolios by Sanger sequencing.

120

Figure 4.5: Transient knockdown of kif6 using morpholinos WT zebrafish were injected with morpholinos at the 1-4 cell stage. A morpholino targeting the kif6 translation start site but containing a 5 base mismatch (mismatch MO) served as a negative control.

121

Figure 4.6: Design and efficiency of kif6 TALENs (A) Genomic sequence showing left and right TALENs that were designed in the second exon of kif6 near the site of the nonsense mutation identified in skolios. The TALEN pair flank a BccI restriction site. (B) BccI restriction digest of PCR amplified DNA from WT embryos. Embryos 1-3 were not injected with kif6 TALENs and BccI cleaves the PCR product completely. Embryos 4-13 were injected with kif6 TALENs at the 1-4 cell stage and all embryos had varying levels of uncut product, indicating a mutation was induced at the BccI restriction site. (C) Three mutations identified in TALEN-injected embryos.

122

Figure 4.7: TALEN-induced mutations in kif6 recapitulate the skolios phenotype Zebrafish with TALEN-induced mutations in kif6 develop both embryonic and adult body axis curvature, similar to skolios mutants.

123

Figure 4.8: RT-PCR expression of kif6 in WT embryos and adult tissues (A) RT-PCR in WT zebrafish embryos shows low levels of kif6 between 24-120 hpf. (B) RT- PCR of various zebrafish tissues shows kif6 expression primarily in brain, intestine and reproductive organs.

124

Figure 4.9: Cilia structure and function are normal in skolios mutants (A) Defects in left-right asymmetry were evaluated by visualizing the position of the heart in 30 hpf embryos for WT (N=94) and skolios mutants (N=173). Normally, zebrafish hearts are located to the left of the midline (right side when viewed ventrally) at 30 hpf. In zebrafish with ambiguous heart positioning, hearts appear at the midline and to the right side (left side when viewed ventrally) in zebrafish with situs inversus. Ambiguous and inversed orientations of the heart were observed at similar frequencies in WT and skolios embryos. (B) Otoliths were observed laterally in in 48-72 hpf zebrafish embryos. Normal otolith numbers (two) and abnormal otolith numbers (one or three) were quantified WT (N=371) and skolios mutants (N=305). (C) Whole-mount confocal immunostaining of 28 hpf WT and skolios embryos shows normal appearance of cilial in the central canal (arrows). (D) Live 48 hpf WT and skolios embryos were injected with 5% tetramethylrhodamine conjugated to 70,000 molecular weight dextran. Sixty minutes after injection, the dye has migrated similar distances in skolios mutants compared to WT embryos.

125

Table 4.1: Identification of candidate gene using low-coverage whole genome sequencing Sequencing and variant identification 101 bp paired-end reads 98,289,069 Aligned paired-end reads 65,990,507 ≥3X coverage 74% Total genomic variants 8,043,651 Variant filtering Total variants in 2.7 Mb mapped locus 18,941 and absent in unrelated zebrafish with similar 4,166 genetic background and homozygous 3,156 and causing a nonsynonymous amino acid 20 change and causing a nonsense mutation (gene) 1 (kif6)

126

REFERENCES

1. Janssen MM, de Wilde RF, Kouwenhoven JW, Castelein RM (2011) Experimental animal models in scoliosis research: a review of the literature. Spine J 11: 347-358. 2. Ouellet J, Odent T (2013) Animal models for scoliosis research: state of the art, current concepts and future perspective applications. Eur Spine J 22 Suppl 2: S81-95. 3. D'Aout K, Aerts P, De Clercq D, De Meester K, Van Elsacker L (2002) Segment and joint angles of hind limb during bipedal and quadrupedal walking of the bonobo (Pan paniscus). Am J Phys Anthropol 119: 37-51. 4. Castelein RM, van Dieen JH, Smit TH (2005) The role of dorsal shear forces in the pathogenesis of adolescent idiopathic scoliosis--a hypothesis. Med Hypotheses 65: 501- 508. 5. Machida M, Saito M, Dubousset J, Yamada T, Kimura J, Shibasaki K (2005) Pathological mechanism of idiopathic scoliosis: experimental scoliosis in pinealectomized rats. Eur Spine J 14: 843-848. 6. Cheung KM, Wang T, Poon AM, Carl A, Tranmer B, Hu Y, Luk KD, Leong JC (2005) The effect of pinealectomy on scoliosis development in young nonhuman primates. Spine (Phila Pa 1976) 30: 2009-2013. 7. Gorman KF, Tredwell SJ, Breden F (2007) The mutant guppy syndrome curveback as a model for human heritable spinal curvature. Spine (Phila Pa 1976) 32: 735-741. 8. Fjelldal PG, Grotmol S, Kryvi H, Gjerdet NR, Taranger GL, Hansen T, Porter MJ, Totland GK (2004) Pinealectomy induces malformation of the spine and reduces the mechanical strength of the vertebrae in Atlantic salmon, Salmo salar. J Pineal Res 36: 132-139. 9. Gorman KF, Breden F (2007) Teleosts as models for human vertebral stability and deformity. Comp Biochem Physiol C Toxicol Pharmacol 145: 28-38. 10. Johnson SL, Gates MA, Johnson M, Talbot WS, Horne S, Baik K, Rude S, Wong JR, Postlethwait JH (1996) Centromere-linkage analysis and consolidation of the zebrafish genetic map. Genetics 142: 1277-1288. 11. Westerfield M (1993) The zebrafish book : a guide for the laboratory use of zebrafish (Brachydanio rerio). Eugene, OR: M. Westerfield. 12. Gansner JM, Madsen EC, Mecham RP, Gitlin JD (2008) Essential role for fibrillin-2 in zebrafish notochord and vascular morphogenesis. Dev Dyn 237: 2844-2861. 13. Obholzer N, Swinburne IA, Schwab E, Nechiporuk AV, Nicolson T, Megason SG (2012) Rapid positional cloning of zebrafish mutations by linkage and homozygosity mapping using whole-genome sequencing. Development 139: 4280-4290. 14. Gray RS, Wilm TP, Smith J, Bagnat M, Dale RM, Topczewski J, Johnson SL, Solnica- Krezel L (2014) Loss of col8a1a function during zebrafish embryogenesis results in congenital vertebral malformations. Dev Biol 386: 72-85.

127

15. Johnson SL, Africa D, Horne S, Postlethwait JH (1995) Half-tetrad analysis in zebrafish: mapping the ros mutation and the centromere of linkage group I. Genetics 139: 1727- 1735. 16. Shimoda N, Knapik EW, Ziniti J, Sim C, Yamada E, Kaplan S, Jackson D, de Sauvage F, Jacob H, Fishman MC (1999) Zebrafish genetic map with 2000 microsatellite markers. Genomics 58: 219-232. 17. Li H, Handsaker B, Wysoker A, Fennell T, Ruan J, Homer N, Marth G, Abecasis G, Durbin R, Genome Project Data Processing S (2009) The Sequence Alignment/Map format and SAMtools. Bioinformatics 25: 2078-2079. 18. Cingolani P, Platts A, Wang le L, Coon M, Nguyen T, Wang L, Land SJ, Lu X, Ruden DM (2012) A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly (Austin) 6: 80-92. 19. Cermak T, Doyle EL, Christian M, Wang L, Zhang Y, Schmidt C, Baller JA, Somia NV, Bogdanove AJ, Voytas DF (2011) Efficient design and assembly of custom TALEN and other TAL effector-based constructs for DNA targeting. Nucleic Acids Res 39: e82. 20. Ha K, Buchan JG, Alvarado DM, McCall K, Vydyanath A, Luther PK, Goldsmith MI, Dobbs MB, Gurnett CA (2013) MYBPC1 mutations impair skeletal muscle function in zebrafish models of arthrogryposis. Hum Mol Genet 22: 4967-4977. 21. Kramer-Zucker AG, Olale F, Haycraft CJ, Yoder BK, Schier AF, Drummond IA (2005) Cilia-driven fluid flow in the zebrafish pronephros, brain and Kupffer's vesicle is required for normal organogenesis. Development 132: 1907-1921. 22. Walker MB, Kimmel CB (2007) A two-color acid-free cartilage and bone stain for zebrafish larvae. Biotech Histochem 82: 23-28. 23. Sander JD, Cade L, Khayter C, Reyon D, Peterson RT, Joung JK, Yeh JR (2011) Targeted gene disruption in somatic zebrafish cells using engineered TALENs. Nat Biotechnol 29: 697-698. 24. Miller JC, Tan S, Qiao G, Barlow KA, Wang J, Xia DF, Meng X, Paschon DE, Leung E, Hinkley SJ, Dulay GP, Hua KL, Ankoudinova I, Cost GJ, Urnov FD, Zhang HS, Holmes MC, Zhang L, Gregory PD, Rebar EJ (2011) A TALE nuclease architecture for efficient genome editing. Nat Biotechnol 29: 143-148. 25. Miki H, Okada Y, Hirokawa N (2005) Analysis of the kinesin superfamily: insights into structure and function. Trends Cell Biol 15: 467-476. 26. Hameed A, Bennett E, Ciani B, Hoebers LP, Milner R, Lawrie A, Francis SE, Grierson AJ (2013) No evidence for cardiac dysfunction in Kif6 mutant mice. PLoS One 8: e54636. 27. Jacquet BV, Salinas-Mondragon R, Liang H, Therit B, Buie JD, Dykstra M, Campbell K, Ostrowski LE, Brody SL, Ghashghaei HT (2009) FoxJ1-dependent gene expression is required for differentiation of radial glia into ependymal cells and a subset of astrocytes in the postnatal brain. Development 136: 4021-4031. 28. Malicki J, Avanesov A, Li J, Yuan S, Sun Z (2011) Analysis of cilia structure and function in zebrafish. Methods Cell Biol 101: 39-74.

128

29. Jaffe KM, Thiberge SY, Bisher ME, Burdine RD (2010) Imaging cilia in zebrafish. Methods Cell Biol 97: 415-435. 30. Lenke L (2004) Idiopathic scoliosis. Philadelphia: Lippincot, Williams & Wilkins 3. 31. Bowen ME, Henke K, Siegfried KR, Warman ML, Harris MP (2012) Efficient mapping and cloning of mutations in zebrafish by low-coverage whole-genome sequencing. Genetics 190: 1017-1024. 32. Voz ML, Coppieters W, Manfroid I, Baudhuin A, Von Berg V, Charlier C, Meyer D, Driever W, Martial JA, Peers B (2012) Fast homozygosity mapping and identification of a zebrafish ENU-induced mutation by whole-genome sequencing. PLoS One 7: e34671. 33. Leshchiner I, Alexa K, Kelsey P, Adzhubei I, Austin-Tse CA, Cooney JD, Anderson H, King MJ, Stottmann RW, Garnaas MK, Ha S, Drummond IA, Paw BH, North TE, Beier DR, Goessling W, Sunyaev SR (2012) Mutation mapping and identification by whole-genome sequencing. Genome Res 22: 1541-1548. 34. Nakagawa T, Tanaka Y, Matsuoka E, Kondo S, Okada Y, Noda Y, Kanai Y, Hirokawa N (1997) Identification and classification of 16 new kinesin superfamily (KIF) proteins in mouse genome. Proc Natl Acad Sci U S A 94: 9654-9659. 35. Miki H, Setou M, Hirokawa N, Group RG, Members GSL (2003) Kinesin superfamily proteins (KIFs) in the mouse transcriptome. Genome Res 13: 1455-1465. 36. Bare LA, Morrison AC, Rowland CM, Shiffman D, Luke MM, Iakoubova OA, Kane JP, Malloy MJ, Ellis SG, Pankow JS, Willerson JT, Devlin JJ, Boerwinkle E (2007) Five common gene variants identify elevated genetic risk for coronary heart disease. Genet Med 9: 682-689. 37. Iakoubova OA, Sabatine MS, Rowland CM, Tong CH, Catanese JJ, Ranade K, Simonsen KL, Kirchgessner TG, Cannon CP, Devlin JJ, Braunwald E (2008) Polymorphism in KIF6 gene and benefit from statins after acute coronary syndromes: results from the PROVE IT-TIMI 22 study. J Am Coll Cardiol 51: 449-455. 38. Iakoubova OA, Tong CH, Rowland CM, Kirchgessner TG, Young BA, Arellano AR, Shiffman D, Sabatine MS, Campos H, Packard CJ, Pfeffer MA, White TJ, Braunwald E, Shepherd J, Devlin JJ, Sacks FM (2008) Association of the Trp719Arg polymorphism in kinesin-like protein 6 with myocardial infarction and coronary heart disease in 2 prospective trials: the CARE and WOSCOPS trials. J Am Coll Cardiol 51: 435-443. 39. Shiffman D, Chasman DI, Zee RY, Iakoubova OA, Louie JZ, Devlin JJ, Ridker PM (2008) A kinesin family member 6 variant is associated with coronary heart disease in the Women's Health Study. J Am Coll Cardiol 51: 444-448. 40. Shiffman D, O'Meara ES, Bare LA, Rowland CM, Louie JZ, Arellano AR, Lumley T, Rice K, Iakoubova O, Luke MM, Young BA, Malloy MJ, Kane JP, Ellis SG, Tracy RP, Devlin JJ, Psaty BM (2008) Association of gene variants with incident myocardial infarction in the Cardiovascular Health Study. Arterioscler Thromb Vasc Biol 28: 173-179. 41. Assimes TL, Holm H, Kathiresan S, Reilly MP, Thorleifsson G, Voight BF, Erdmann J, Willenborg C, Vaidya D, Xie C, Patterson CC, Morgan TM, Burnett MS, Li M, Hlatky MA, Knowles JW, Thompson JR, Absher D, Iribarren C, Go A, Fortmann SP, Sidney S,

129

Risch N, Tang H, Myers RM, Berger K, Stoll M, Shah SH, Thorgeirsson G, Andersen K, Havulinna AS, Herrera JE, Faraday N, Kim Y, Kral BG, Mathias RA, Ruczinski I, Suktitipat B, Wilson AF, Yanek LR, Becker LC, Linsel-Nitschke P, Lieb W, Konig IR, Hengstenberg C, Fischer M, Stark K, Reinhard W, Winogradow J, Grassl M, et al. (2010) Lack of association between the Trp719Arg polymorphism in kinesin-like protein-6 and coronary artery disease in 19 case-control studies. J Am Coll Cardiol 56: 1552-1563. 42. Miki H, Setou M, Kaneshiro K, Hirokawa N (2001) All kinesin superfamily protein, KIF, genes in mouse and human. Proc Natl Acad Sci U S A 98: 7004-7011. 43. Yu X, Ng CP, Habacher H, Roy S (2008) Foxj1 transcription factors are master regulators of the motile ciliogenic program. Nat Genet 40: 1445-1453. 44. Stubbs JL, Oishi I, Izpisua Belmonte JC, Kintner C (2008) The forkhead protein Foxj1 specifies node-like cilia in Xenopus and zebrafish embryos. Nat Genet 40: 1454-1460. 45. Bachmann-Gagescu R, Phelps IG, Stearns G, Link BA, Brockerhoff SE, Moens CB, Doherty D (2011) The ciliopathy gene cc2d2a controls zebrafish photoreceptor outer segment development through a role in Rab8-dependent vesicle trafficking. Hum Mol Genet 20: 4041-4055. 46. Burwell RG, Dangerfield PH, Freeman BJ, Aujla RK, Cole AA, Kirby AS, Pratt RK, Webb JK, Moulton A (2006) Etiologic theories of idiopathic scoliosis: the breaking of bilateral symmetry in relation to left-right asymmetry of internal organs, right thoracic adolescent idiopathic scoliosis (AIS) and vertebrate evolution. Stud Health Technol Inform 123: 385-390. 47. Schultz J, Milpetz F, Bork P, Ponting CP (1998) SMART, a simple modular architecture research tool: identification of signaling domains. Proc Natl Acad Sci U S A 95: 5857- 5864.

130

Chapter 5:

Discussion and Future Directions

131

Scoliosis is a form of spinal deformity that causes the spine to curve abnormally in the

frontal plane. Although scoliosis occasionally arises as a secondary feature of a broader,

systemic condition, most cases of scoliosis develop in late childhood with no known etiology and

are classified as adolescent idiopathic scoliosis (AIS). AIS is generally a mild condition, but for

unknown reasons, about 10% of AIS patients develop progressive deformity and bracing or

surgery is warranted to prevent negative physical and psychological outcomes. Twin and family

studies indicate a role for genetic factors in AIS development, but the identification of

susceptibility loci using linkage, candidate gene and genome-wide association studies have yielded mixed and largely inconclusive findings. Overall, the genetic basis of AIS remains poorly understood, which has been further challenged by the lack of a relevant animal model.

In this dissertation, we explored the role of rare variation in AIS, which had been largely unaddressed prior to the commencement of this work. Specifically, we evaluated rare copy number variants (CNVs), chromosomal aneuploidy and rare coding variants (SNPs, INDELs) as

genetic risk factors for AIS. We also proposed that zebrafish (Danio rerio) is an excellent

animal model to identify genetic and molecular pathways involved in spinal development. While

there is still much to be learned about the genetic basis of AIS, several general conclusions can

be drawn from the results of the work presented here that will aid in future efforts to identify

genetic risk factors and better understand AIS.

1. AIS is complex disorder with complex genetics

A major challenge addressing the genetic underpinnings of AIS has been the variable

expressivity of the phenotype. Individuals with AIS have substantial phenotypic variation; curve

severity, age of onset, rates of progression and overall curve morphology can all differ

132

significantly among patients. It is unclear to what degree genetic and environmental variation

influences each of these features of AIS and if there are genetic risk factors that modify the AIS phenotype independent from overall susceptibility.

Our findings support a role for genetic risk factors that act as disease modifiers. In our exome sequencing study of rare coding variants, we identified rare missense mutations in FBN1

(encoding fibrillin-1) and FBN2 (encoding fibrillin-2) that were strongly associated with curve

severity in AIS patients. These results pose a challenge to future investigations of AIS, as phenotypic variability among patients could hamper genetic studies by reducing the power of

association or replication studies when case populations are highly heterogeneous. Moreover,

segregation analysis may be misinterpreted in such cases. For example, if a genetic variant

identified in a severely affected AIS patient is absent in a mildly affected family member, the

variant may be deemed unimportant due to the failed segregation, but this result could also be

consistent with the variant being a major disease modifier. The complexity of the AIS phenotype

should therefore be considered when assembling AIS patient cohorts to identify genetic risk

factors for overall AIS susceptibility versus genetic modifiers that contribute to other variable

features of AIS. Because we have shown that genetic modifiers can impact curve severity, it

would be interesting to evaluate other features of AIS that differ among patients (e.g. curve

morphology, rates of progression).

In addition to the interaction between genetic factors and a complex phenotype, we have

also shown that those genetic factors are also complex, with many types of genetic variation

contributing to AIS. Previous genome-wide association studies have addressed the ‘common

disease, common variant’ hypothesis (CDCV) and shown that common variants in LBX1

(encoding ladybird homeobox 1) are associated with AIS in individuals of Asian ancestry [1-4].

133

An opposing theory, the ‘common disease, rare variant’ (CDRV) hypothesis, states that common diseases are due to rare variation of higher penetrance [5]. We show that the CDRV hypothesis is also a likely disease model for AIS development. In our study of FBN1 and FBN2, we show that exceedingly rare coding variants were significantly associated with AIS. Similarly, our analysis of copy number variation identified rare duplications of chromosome 1q21.1 and trisomy X that were enriched in AIS patients compared to controls.

Collectively, these results indicate that many forms of genetic variation (e.g. CNVs, missense mutations), both common and rare, are significant genetic factors associated with AIS susceptibility and/or modifiers of AIS features (e.g. curve severity). Future studies will therefore need to address multiple forms of genetic variation when evaluating AIS cases and stratify case populations to reduce heterogeneity.

2. Several genes/loci conferring risk to AIS are implicated in scoliosis-associated

syndromes

Our results suggest that genes already linked to disorders with high rates of scoliosis are good candidates for AIS. As previously mentioned, we demonstrated an association with FBN1 and FBN2 in AIS cases with severe deformity. Mutations in FBN1 cause Marfan syndrome [6] and 63% of Marfan syndrome patients develop scoliosis [7]. However, Marfan syndrome also causes a variety of other symptoms, including aortic aneurysm and dissections, ectopia lentis, tall stature, pectus excavatum and carinatum, dolichostenomelia, arachnodactyly and joint hypermobility. Mutations in FBN2 cause congenital contractural arachnodactyly (also called

Beals syndrome) [8], and 50% of patients develop scoliosis [9], as well as “crumpled” ears, muscular hypoplasia, contractures of the major joints, pectus excavatum and carinatum, tall

134

stature, dolichostenomelia and arachnodactyly. Disease-causing mutations in FBN1 and FBN2

are thought to be highly penetrant and in FBN1, over 1,000 mutations have been described [10].

Because both FBN1 and FBN2 already were linked to Mendelian disorders, it was surprising to

discover an association with FBN1 and FBN2 and AIS. Since Marfan syndrome and congenital

contractural arachnodactyly both lead to scoliosis, we conclude that these specific mutations

confer risk to isolated skeletal phenotypes such as scoliosis without additional disease features.

A comparable finding was observed by Gao et al. [11] in CDH7 (encoding

chromodomain helicase DNA-binding protein 7). Mutations in CDH7 cause CHARGE

syndrome (coloboma of the eye, heart defects, atresia of the choanae, retardation of growth

and/or development, genital and/or urinary abnormalities, and ear abnormalities and deafness)

[12] and >60% of CHARGE adolescent and adult patients develop scoliosis [11,13]. Sequencing

CDH7 in 25 affected patients with AIS showed the minor allele of a common polymorphism

(rs4738824) was significantly over-transmitted [11]. Although additional studies are needed in a

larger case population to validate these findings, these results are consistent with variants

affecting a gene (CDH7) typically associated with syndromic forms of scoliosis (CHARGE

syndrome) also being a genetic risk factor for AIS.

Similarly, duplications and deletions of 16p11.2 have been recently linked to scoliosis in

individuals with a wide variety of other neurobehavioral and congenital abnormalities [14]. The

16p11.2 locus was also identified by linkage analysis of 202 idiopathic scoliosis families [15,16]

and Al-Kateb et al. [14] hypothesized that 16p11.2 could therefore be a novel susceptibility locus for idiopathic scoliosis. Prior to the work presented here, 16p11.2 CNVs had not been described in an individual with isolated scoliosis. We identified a duplication of chromosome 16p11.2 in a

single AIS patient, providing preliminary evidence that 16p11.2 rearrangements can occur in a

135

scoliosis patient without the additional neurobehavioral and congenital abnormalities common to

this genomic disorder. Although additional studies containing larger patient cohorts will be

needed to demonstrate an association with 16p11.2 rearrangements and AIS, this finding

supports a role for the 16p11.2 locus in AIS susceptibility, despite causing a genomic disorder

most often associated with many additional phenotypes.

Overall, our findings and others suggest that genes or loci causing systemic diseases

associated with high rates of scoliosis (and many other symptoms) can be involved in the development of scoliosis as an isolated phenotype. Future genetic studies of AIS should

therefore focus on genes and molecular pathways associated with neuromuscular or syndromic forms of scoliosis. For example, scoliosis is common to a number of Mendelian disorders, such as Ehlers-Danlos syndrome (caused by mutations in collagen genes) [17]. Loeys-Dietz syndrome is a dominant Mendelian disorder caused by mutations in the transforming growth factor-beta receptors, TGFβR1 and TGFβR2, that is also highly associated with scoliosis [18].

Because Loeys-Dietz syndrome, like Marfan syndrome and congenital contractural

arachnodactyly, alter the TGF-β pathway, TGFβR1 and TGFβR2 are especially good candidates

to evaluate in future studies.

3. A subset of AIS patients are undiagnosed for other conditions, but this is very rare

In our CNV study, we identified two AIS patients with trisomy of the X chromosome,

also called Triple X syndrome (47,XXX). Triple X syndrome generally causes only mild

phenotypes, which can include scoliosis [19]. As a result, many females with triple X syndrome

are unaware of the condition [20]. The two females we identified with triple X syndrome

represented <1% of all AIS females we evaluated for CNVs or chromosomal aneuploidy.

136

However, discovery of triple X syndrome allows clinicians to monitor other associated

phenotypes (e.g. premature ovarian failure and developmental delay), so this finding is of clinical

relevance despite its rarity.

When we identified patients with rare variants in FBN1, we considered the possibility

that these individuals were undiagnosed for Marfan syndrome. In some cases, clinical recognition of Marfan syndrome is challenging due to large phenotypic variability among patients and nonspecific features. Features that are most specific to Marfan syndrome (e.g. aortic dilation, ectopia lentis) are only identifiable though testing which is not routinely performed.

Therefore, we referred AIS patients with rare variants in FBN1 for a thorough clinical

examination by a clinical geneticist. One patient with a previously described mutation in FBN1

met the diagnostic criteria for Marfan syndrome after examination, representing only 0.3% of all

AIS patients of European ancestry evaluated (N=311). Marfan syndrome had not been

previously suspected in this individual, but our finding is likely to be of great impact to the future

health and management of this individual, as 90% of Marfan syndrome patients are affected with

cardiac valve disease, aortic dissection, or congestive heart failure [21].

Generally, the vast majority of AIS patients we evaluated were not found to have genetic or phenotypic findings suggestive of another diagnosis as a primary cause of scoliosis. However, we have found that scoliosis can be attributed to a previously undiagnosed, underlying condition in a very small number of AIS patients. This result was unexpected, as most syndromic forms of scoliosis occur simultaneously with additional, often easily recognized disease features. Because many underlying diagnoses necessitate additional clinical care, future clinical and research studies involving AIS patients should make every attempt to rule out scoliosis-associated disorders like triple X syndrome and Marfan syndrome that can be difficult to recognize.

137

4. Genetic factors may predict AIS patients at risk for severe deformity

Because AIS causes deformity in previously healthy children, a major long-term goal of

AIS research is to identify methods to predict and ultimately prevent severe deformity. One

feasible approach to achieve this would be to identify genetic risk factors for severe deformity

that could inform treatment strategies (e.g. bracing). Idiopathic spinal deformity is common in

the pediatric population, but not all individuals develop severe forms of disease and currently

there is no method of predicting curve progression. As a result, individuals with mild curvatures

may be unnecessarily monitored and treated, whereas treatment may be delayed individuals at

high risk of severe deformity.

We have shown through our evaluation of FBN1 and FBN2 that 7.6% of patients of

European ancestry had a rare variant in either gene and that these variants were significantly

correlated to curve severity in a large cohort of AIS patients of European ancestry (N=405),

which replicated in a cohort of Han Chinese AIS patients (N=370). This indicates that FBN1 and

FBN2 variants may be prognostic genetic markers to predict scoliosis progression and

demonstrates the potential of identifying AIS patients for proactive treatment. More studies will

be needed to assess the clinical value and predictive value of identifying rare fibrillin variants,

but our findings may be an important advance in our ability to identify patients at high risk of progressive deformity.

138

5. Zebrafish may be an ideal model to identify new genes involved in scoliosis/spine

development and model genetic findings in human AIS

A major challenge in understanding AIS has been the lack of a relevant animal model.

Because zebrafish (Danio rerio) are highly tractable and have extensive genomic resources, zebrafish are particularly well-suited model organisms to model scoliosis. Another teleost mutant, the guppy strain curveback, has been shown to develop non-induced spinal deformity

[22], yet, spinal deformity in zebrafish has been largely unexplored. We performed a forward genetic screen to identify zebrafish mutants with features of AIS. From this screen, we identified skolios. skolios mutants develop a medial-lateral curvature of the spine that arises independent of vertebral malformations, similar to humans with AIS. Using meiotic mapping, low-coverage whole genome sequence and TALENs, we identified kif6 as the genetic basis of skolios.

Overall, these results show that zebrafish are susceptible to AIS-like spinal curvature and therefore, zebrafish can be used as an animal model for AIS. Our mutagenesis screen (a forward genetic approach) successfully identified a zebrafish mutant with an AIS-like phenotype. We confirmed the genetic basis using the many useful tools available in zebrafish. Moreover, these results suggest that zebrafish are suitable for reverse genetic screens. In future studies, we hope to induce the genetic changes identified in AIS patients (e.g. FBN1 and FBN2 mutations) to determine their effect on the spinal column in zebrafish.

In summary, we have provided evidence that rare duplications and deletions are genetic risk factors for AIS. We have also shown that rare coding variants in FBN1 and FBN2 are associated with severe deformity and, as there are no reliable methods to predict progression in

AIS patients, we hope these results will aid in our ability to develop future personalized

139 prevention strategies that might eliminate the need for surgery in AIS patients at high risk of severe deformity and to eliminate unnecessary screening and treatment for patients at low risk of progression. Finally, we have demonstrate zebrafish as a viable animal model to study novel genetic and molecular pathways involved in spine development and to model genetic findings uncovered in human AIS populations. The work presented here represents a significant advance in our understanding of AIS, which will be a primer for future studies aimed at elucidating the genetic basis and pathogenesis of AIS.

140

REFERENCES

1. Takahashi Y, Kou I, Takahashi A, Johnson TA, Kono K, Kawakami N, Uno K, Ito M, Minami S, Yanagida H, Taneichi H, Tsuji T, Suzuki T, Sudo H, Kotani T, Watanabe K, Chiba K, Hosono N, Kamatani N, Tsunoda T, Toyama Y, Kubo M, Matsumoto M, Ikegawa S (2011) A genome-wide association study identifies common variants near LBX1 associated with adolescent idiopathic scoliosis. Nat Genet 43: 1237-1240. 2. Fan YH, Song YQ, Chan D, Takahashi Y, Ikegawa S, Matsumoto M, Kou I, Cheah KS, Sham P, Cheung KM, Luk KD (2012) SNP rs11190870 near LBX1 is associated with adolescent idiopathic scoliosis in southern Chinese. J Hum Genet 57: 244-246. 3. Jiang H, Qiu X, Dai J, Yan H, Zhu Z, Qian B, Qiu Y (2013) Association of rs11190870 near LBX1 with adolescent idiopathic scoliosis susceptibility in a Han Chinese population. Eur Spine J 22: 282-286. 4. Gao W, Peng Y, Liang G, Liang A, Ye W, Zhang L, Sharma S, Su P, Huang D (2013) Association between common variants near LBX1 and adolescent idiopathic scoliosis replicated in the Chinese Han population. PLoS One 8: e53234. 5. Schork NJ, Murray SS, Frazer KA, Topol EJ (2009) Common vs. rare allele hypotheses for complex diseases. Curr Opin Genet Dev 19: 212-219. 6. Dietz HC, Cutting GR, Pyeritz RE, Maslen CL, Sakai LY, Corson GM, Puffenberger EG, Hamosh A, Nanthakumar EJ, Curristin SM, et al. (1991) Marfan syndrome caused by a recurrent de novo missense mutation in the fibrillin gene. Nature 352: 337-339. 7. Sponseller PD, Hobbs W, Riley LH, 3rd, Pyeritz RE (1995) The thoracolumbar spine in Marfan syndrome. J Bone Joint Surg Am 77: 867-876. 8. Putnam EA, Zhang H, Ramirez F, Milewicz DM (1995) Fibrillin-2 (FBN2) mutations result in the Marfan-like disorder, congenital contractural arachnodactyly. Nat Genet 11: 456-458. 9. Ramos Arroyo MA, Weaver DD, Beals RK (1985) Congenital contractural arachnodactyly. Report of four additional families and review of literature. Clin Genet 27: 570-581. 10. Keane MG, Pyeritz RE (2008) Medical management of Marfan syndrome. Circulation 117: 2802-2813. 11. Gao X, Gordon D, Zhang D, Browne R, Helms C, Gillum J, Weber S, Devroy S, Swaney S, Dobbs M, Morcuende J, Sheffield V, Lovett M, Bowcock A, Herring J, Wise C (2007) CHD7 gene polymorphisms are associated with susceptibility to idiopathic scoliosis. Am J Hum Genet 80: 957-965. 12. Vissers LE, van Ravenswaaij CM, Admiraal R, Hurst JA, de Vries BB, Janssen IM, van der Vliet WA, Huys EH, de Jong PJ, Hamel BC, Schoenmakers EF, Brunner HG, Veltman JA, van Kessel AG (2004) Mutations in a new member of the chromodomain gene family cause CHARGE syndrome. Nat Genet 36: 955-957. 13. Doyle C, Blake K (2005) Scoliosis in CHARGE: a prospective survey and two case reports. Am J Med Genet A 133A: 340-343.

141

14. Al-Kateb H, Khanna G, Filges I, Hauser N, Grange DK, Shen J, Smyser CD, Kulkarni S, Shinawi M (2014) Scoliosis and vertebral anomalies: Additional abnormal phenotypes associated with chromosome 16p11.2 rearrangement. Am J Med Genet A. 15. Miller NH, Justice CM, Marosy B, Doheny KF, Pugh E, Zhang J, Dietz HC, 3rd, Wilson AF (2005) Identification of candidate regions for familial idiopathic scoliosis. Spine (Phila Pa 1976) 30: 1181-1187. 16. Miller NH, Justice CM, Marosy B, Swindle K, Kim Y, Roy-Gagnon MH, Sung H, Behneman D, Doheny KF, Pugh E, Wilson AF (2012) Intra-familial tests of association between familial idiopathic scoliosis and linked regions on 9q31.3-q34.3 and 16p12.3- q22.2. Hum Hered 74: 36-44. 17. Rabenhorst BM, Garg S, Herring JA (2012) Posterior spinal fusion in patients with Ehlers- Danlos syndrome: a report of six cases. J Child Orthop 6: 131-136. 18. Erkula G, Sponseller PD, Paulsen LC, Oswald GL, Loeys BL, Dietz HC (2010) Musculoskeletal findings of Loeys-Dietz syndrome. J Bone Joint Surg Am 92: 1876- 1883. 19. Olanders S (1977) Females with Supernumerary X Chromosomes; A Study of 39 Psychiatric Cases: Esselte Studium. 20. Gustavson KH (1999) [Triple X syndrome deviation with mild symptoms. The majority goes undiagnosed]. Lakartidningen 96: 5646-5647. 21. Ha HI, Seo JB, Lee SH, Kang JW, Goo HW, Lim TH, Shin MJ (2007) Imaging of Marfan syndrome: multisystemic manifestations. Radiographics 27: 989-1004. 22. Gorman KF, Tredwell SJ, Breden F (2007) The mutant guppy syndrome curveback as a model for human heritable spinal curvature. Spine (Phila Pa 1976) 32: 735-741.

142