Clinical and Genetic Studies of Three Inherited Skeletal ...216272/FULLTEXT01.pdf · radiological...

UMEÅ UNIVERSITY MEDICAL DISSERTATIONS

New Series No. 1259 ISSN 0346-6612 ISBN 978-91-7264-753-4

Clinical and Genetic Studies of Three

Inherited Skeletal Disorders

Eva-Lena Stattin

From the Department of Medical Biosciences Medical and Clinical Genetics Umeå University SE-901 85 Umeå Sweden 2009

Copyright© 2009 by Eva-Lena Stattin New Series No. 1259 ISBN: 978-91-7264-753-4 ISSN: 0346-6612 Cover: Pedigree, familial osteochondritis dissecans Printed by: Hemströms tryckeri, Härnösand, Sweden 2009

3

To Pär, Kalle and Oskar

4

Contents

CONTENTS .......................................................................................................................4

PUBLICATIONS...............................................................................................................6

THESIS AT A GLANCE ..................................................................................................7

ABSTRACT .......................................................................................................................8

ABBREVIATIONS............................................................................................................9

0BMEDICAL GENETICS ............................................................................................ 11 28BHuman genome................................................................................................................. 11 29BDNA structure................................................................................................................... 11 30BCharacteristics of Mendelian inheritance.......................................................................... 11 31BInter individual variations................................................................................................. 12

INTRODUCTION ...........................................................................................................14

15BThe Skeleton................................................................................................ 14 16BExtracellular matrix ...................................................................................... 14 17BCartilage....................................................................................................... 15

1BCOMPONENTS OF THE EXTRACELLULAR MATRIX NETWORK.................................... 16 18BThe large aggregating Proteoglycans.......................................................... 16 19BACAN-gene and Aggrecan .......................................................................... 16

32BMolecular gene structure, organization and function........................................................ 16 2BCHONDRODYSPLASIAS DUE TO AGGRECAN DEFECTS............................................. 18 3BCHONDRODYSPLASIAS DUE TO AGGRECAN DEFECTS - ANIMAL MODELS ................. 19

33BCartilage matrix deficiency............................................................................................... 19 34BCartilage matrix deficiency-Bc ......................................................................................... 19 35BNanomelia......................................................................................................................... 19 36BBulldog dwarfism ............................................................................................................. 19

4BCHONDRODYSPLASIAS DUE TO AGGRECAN DEFECTS – HUMAN DISORDERS............ 20 37BSpondyloepiphyseal dysplasia type Kimberley (SEDK)................................................... 20 38BSpondyloepimetaphyseal dysplasia (SEMD-aggrecan-type) ............................................ 20 39BFamilial osteochondritis dissecans.................................................................................... 20

Disorders associated with defective sulfation of proteoglycans .................. 21 20BDegenerative diseases involving aggrecan ................................................. 21

40BVNTR-region of the ACAN-gene and OA ....................................................................... 21 41BAge-related changes of aggrecan ...................................................................................... 22

5BCOMPONENTS OF THE EXTRACELLULAR MATRIX NETWORK.................................... 22 42BThe Collagens ................................................................................................................... 22

CHONDRODYSPLASIA DUE TO DEFECTS IN ECM-PROTEINS ................................... 23 43BMultiple epiphyseal dysplasia........................................................................................... 23

SKELETONENESIS ............................................................................................... 23 22BThe Indian Hedgehog (IHH)-gene; an important gene for skeletal development................................................................................................. 24

44BBrachydactyly type A1 ..................................................................................................... 25

AIMS OF THE THESIS .................................................................................................26

5

METHODOLOGY ..........................................................................................................27 45BEthics ................................................................................................................................ 27

23BIntroduction to field work .............................................................................. 27 46BCollecting clinical material ............................................................................................... 27 47BEvaluation of suspected skeletal dysplasias...................................................................... 27

24BField work..................................................................................................... 27 48BStudy I and II (Familial OCD).......................................................................................... 27 49BStudy III (MED) ............................................................................................................... 28 50BStudy IV (BDA1).............................................................................................................. 28

25BMethods ....................................................................................................... 29 51BStatistical analysis............................................................................................................. 29 52BLinkage analysis ............................................................................................................... 29

26BMolecular genetics methods ........................................................................ 29 53BDNA extraction................................................................................................................. 29 PCR .................................................................................................................................. 29 54BGenome-wide scan ........................................................................................................... 30 55BSequencing/Candidate gene analysis ................................................................................ 30 56BInterpretation of identified variations ............................................................................... 31 57BDHPLC............................................................................................................................. 31 58BRestriction endonuclease digestion of the PCR product ................................................... 32 59BProtein interaction analysis............................................................................................... 32 60BMass spectrometry ............................................................................................................ 33

RESULTS AND DISCUSSION ......................................................................................34

7BPAPER I ............................................................................................................. 34 8BPAPER II ............................................................................................................ 36 9BPAPER III ........................................................................................................... 39 10BPAPER IV ........................................................................................................... 41

27BConcluding remarks and future perspectives .............................................. 43

POPULÄRVETENSKAPLIG SAMMANFATTNING PÅ SVENSKA ......................44

11BINTRODUKTION ................................................................................................... 44 12BMANUSKRIPT I & II .............................................................................................. 44 13BMANUSKRIPT III .................................................................................................. 45 14BMANUSKRIPT IV.................................................................................................. 46

ACKNOWLEDGEMENTS ............................................................................................47

REFERENCES ................................................................................................................49

6

Publications This thesis is based on the following original publications, which will be referred to by their roman numerals: I Eva-Lena Stattin, Yelverton Tegner, Magnus Domellöf and

Niklas Dahl, ”Familial osteochondritis dissecans associated with early osteoarthritis and disproportionate short stature”, Osteoarthritis and Cartilage (2008) 16, 890-896

II Eva-Lena Stattin, Fredrik Wiklund, Karin Lindblom, Patrik

Önnerfjord, Björn-Anders Jonsson, Yelverton Tegner, Takako Sasaki, André Struglics, Stefan Lohmander, Niklas Dahl, Dick Heinegård and Anders Aspberg, “A mutation in the aggrecan C-type lectin domain disrupts extracellular matrix interactions and causes dominant familial osteochondritis dissecans with short stature and early osteoarthritis”. Submitted.

III Eva-Lena Stattin, Inger Cullman, Shiro Ikegawa, William G.

Cole, Outi Mäkitie and Niklas Dahl, “Mutations in the Aggrecan Gene (ACAN) are Associated with Multiple Epiphyseal Dysplasia”. Submitted.

IV Eva-Lena Stattin, Bjarne Lindén, Torsten Lönnerholm, Jens

Schuster and Niklas Dahl, “Brachydactyly type A1 associated with unusual radiological findings and a novel mutation in the Indian Hedgehog (IHH) gene”. European Journal of Medical Genetics, accepted for publication, pending minor revision.

The original paper was reprinted with kind permission from the publisher.

7

Thesis at a glance Topic Material &

metods Illustration Results Conclusion

I Unusual clinical and radiological characteristics within a large family with inherited skeletal mal-formations. A novel type of skeletal dysplasia?

Investigation of 53 members in a five-generation family. Clinical examination, anthropometric measurements, radiological imaging, questionnaire, blood samples.

X-ray of knee joint in OCD

15 affected members identified. OCD in knees, hips and elbows. Early osteo-arthritis (OA). Disproportional short stature with short arms and legs.

fOCD is asso-ciated with OCD in knees, hips and/or elbows. Disproportionate short stature - reduced normal growth spurt. OCD before epiphyseal closure, and early onset OA.

II What gene mutation is associated with fOCD? What are the functional consequences of the mutation?

19 affected family members investigated. DNA, joint fluid, bone marrow and cartilage. Genome wide linkage analysis. Mass-spectrometry. BIA-core. Interaction-, affinity- and protein studies.

DNA-sequence chromatogram illustrating mutation

Linkage to chromosome 15q26 with max lod score 6.36 (θ 0). ACAN-gene mutation p.Val2303Met. Loss of interaction. Low or no affinity to fibulin1, fibulin2 and tenascin-R. Secreted.

fOCD Val>Met substitution segregate with the disease in the family. The mutation alters normal aggrecan G3 domain function, binding surface and interaction with known ligands.

III

a. Search for new genes associated with Multiple Epiphyseal Dysplasia (MED) via a candidate gene approach. b. Do ACAN mutations cause MED?

DNA from 39 individuals with MED, screened for ACAN mutations. Sequencing, dHPLC, and restriction endonuclease digestion.

Gel separated PCR fragments showing mutations

Heterozygous mutation in exon 8 of ACAN-gene, in an adult male. Compound heterozygous mutations in exon 6 and 7, of a child. Parents heterozygous.

First ACAN-gene mutations identified associated with MED. These findings extend the spectrum of mutated genes that may cause MED

IV a. What are the clinical and radiological characteristics of OCD and brachymesophalangy? b. What mutation causes the disease?

Five-generation family, 7 affected and 9 healthy family members. Linkage-analysis, Sequencing and protein modeling.

3D structure of wild-type and mutant IHH

BDA1. Linkage to chromosome 2q35 marker. Novel IHH-gene mutation. Change in 3D structure of IHH identified.

BDA1 with unusual clinical findings. Disturbed interaction between IHH and its ligand (PTC). Novel genotype-phenotype correlations.

8

Abstract Mutations in genes of importance for cartilage development may lead to skeletal malformations, chondroskeletal dysfunction and increased susceptibility to degenerative joint disease. Characterization of these mutations and identification of molecular pathways for the corresponding gene products have contributed to our understanding of mechanisms regulating skeletal patterning, endochondral ossification and joint formation.

A five generation family segregating autosomal dominant osteochondritis dissecans (OCD) was identified. Affected family members presented with OCD in knees, hips and elbows, short stature, and early osteoarthritis. A genome wide scan and a multipoint linkage analysis identified aggrecan (ACAN) as a prime candidate gene. DNA sequence analysis of the ACAN-gene revealed heterozygosity for a missense mutation (c.6907G>A) in affected subjects, resulting in a p.V2303M substitution in the aggrecan G3 domain C-type lectin. This domain is important for the interaction with other proteins in the cartilage extracellular matrix. To determine the effect of the V2303M substitution on secretion and interaction, we performed binding studies with recombinant mutated and wild type G3 proteins. We found decreased affinity or complete loss of interaction between V2303M aggrecan and fibulin1, fibulin2 and tenascin-R. Analysis of articular cartilage from an affected family member confirmed that V2303M aggrecan is produced and present.

In search for gene mutations associated with multiple epiphyseal dysplasia (MED) we considered the ACAN-gene a likely candidate. The ACAN-gene was analysed in 39 individuals with MED and screened negative for mutations in six previously known MED genes. Sequence analysis revealed a heterozygous missense mutation (c.1448G>T) in one adult male and compound heterozygous missense mutations (c.1366T>C and c.836G>A) in a five year old boy with healthy parents, each of them carrier for one of the mutations.

A large family segregating autosomal dominant brachymesophalangia and OCD in finger joints was characterised. The clinical presentation in six affected family members was consistent with the diagnosis Brachydactyly type A1, in this family characterized by shortening of the middle phalanges, short ulnar styloid process, flattening of the metacarpal heads and mild osteoarthritis. The condition may be caused by mutations in the Indian hedgehog gene (IHH) or a yet unidentified gene on chromosome 5p13. Sequence analysis of the IHH-gene in affected individuals revealed a novel C to T transition (c.472C>T) leading to a p.158Arg>Cys substitution. Residue 158 in IHH is highly conserved throughout evolution and molecular structure modelling of IHH suggests that the R158C substitution leads to a conformational change at the site of interaction with the IHH-receptor. This supports that the substitution causes Brachydactyly type A1 in this family. In summary, we report on the clinical, radiological and molecular genetic characteristics of the three skeletal disorders OCD, MED and BDA1. Our results provide a novel molecular mechanism in the pathophysiology of familial osteochondritis dissecans confirming the importance of aggrecan C-type lectin for cartilage function. We also show that ACAN-gene mutations may be associated with MED extending the spectrum of skeletal dysplasias associated with the aggrecan gene. Finally, we report on a novel missense mutation in a conserved region of the IHH-gene associated with BDA1.

9

Abbreviations ACAN-gene The gene that encodes Aggrecan. ACG1B Achondrogenesis 1B. AD Autosomal dominant. adMED Autosomal dominant Multiple epiphyseal dysplasia. Allele One of two or more alternative forms of a gene. AO2 Atelosteogenesis type II. AR Autosomal recessive. BDA1 Brachydactyly type A1. bp Base pair. cM Centimorgan, a unit of genetic distance, 1/100th of a

Morgan, which is the distance over which we would expect to see one recombination per meiosis.

CLD C-type lectin domain. CMD Cartilage matrix deficiency. CMCI First carpometacarpal joint. COL2A1 Collagen 2 alpha 1 gene. COL9A1 Collagen 9 alpha 1 gene. COL9A2 Collagen 9 alpha 2 gene. COL9A3 Collagen 9 alpha 3 gene. COMP Cartilage oligomeric matrix protein. CS Chondroitin sulphate. CS1, CS2 CS attachment regions. DIP Distal interphalangeal joint. DNA Deoxyribonucleic acid. DTD Diastrophic dysplasia. DTDST Diastrophic dysplasia sulfate transporter gene. ECM Extracellular matrix. EGF Epidermal growth factor. FGF Fibroblast-growth-factor. fOCD Familial osteochondritis dissecans. GAG Glycosaminoglycan. GWS Genome-wide scan. HPLC high pressure liquid chromatography. IHH Indian hedgehog-gene. IGD Interglobular domain. Locus, loci Part(s) of a chromosome that most likely include a gene.

at least partly responsible for a disease. LOD Logarithm 10 of odds. MALDI-TOF MS Matrix-assisted laser desorption ionization time of the

flight mass spectrometry. Mb Million base pairs. MATN3 Matrillin 3 gene. MED Multiple epiphyseal dysplasia. MRI Magnetic resonance imaging. mRNA Messenger RNA. MS Mass spectrometry. OA Osteoarthritis.

10

OCD Osteochondritis dissecans. PCR Polymerase chain reaction. PIP Proximal interphalangeal joint. PTHrP Parathyroid hormone related peptide. PTR Proteoglycan tandem repeat. rMED Recessive inherited multiple epiphyseal dysplasia. SCR Short complement–or consensus repeat. SEDK Spondyloepiphyseal dysplasia type Kimberley. SEMD Spondyloepimetaphyseal dysplasia. SNP Single nucleotide polymorphism. VNTR Variable number of tandem repeat.

11

0BMedical Genetics 28BHuman genome

The function of the human genome is to transfer information stably from one generation to the next. It consists of two parts: the nuclear genome and the mitochondrial genome. The nuclear genome comprises approximately 3x10-9 nucleo-tides (bp) of DNA, stored in 23 chromosome pairs (IHGS consortium 2004). Twenty-two of these are autosomal chromosomal pairs and the remaining is a sex-determining pair. The mitochondrial genome is a circular DNA molecule of 37 genes located in the energy-generating organelles called mitochondria. The results from the human genome project, an international collaboration with a primary aim of determining the nucleotide sequence of the entire human nuclear genome, was presented in 2001 and the DNA sequence was estimated to compromise approxima-tely 30,000 genes (VENTER et al. 2001). The number of genes was less than expected and today the protein coding genes is estimated to be even less, about 20-25,000. Humans have roughly the same number of genes as the unquestionably less complex nematode, Caenorhabditis elegans. The number of genes was expected to be about 80,000- 100,000. The estimates were high because they were based on the supposition that, a single gene specifies a single mRNA and a single protein. The discovery that the number of genes is much lower indicates that differential splicing is more prevalent than earlier expected, and suggests that the link between mRNA and protein might be more complex than previously thought.

29BDNA structure Deoxyribonucleic acid (DNA) contains the genetic information for the synthesis of proteins. DNA is composed of three types of units: a five-carbon sugar, a deo-xyribose; a nitrogen-containing base; and a phosphate group. The bases are of two

types, purines and pyrimidines. In DNA there are two purine bases - adenine (A) and guanine (G) - and two pyrimidine bases - thymine (T) and cytosine (C). DNA is composed of two nucleotide chains held together by complementary pairing of A with T and G with C. The native state of DNA is a double helix (WATSON and CRICK 1953). Ribonucleic acid (RNA) is very similar to DNA, but differs in some important structural details; in the cell RNA is single stranded, whereas DNA is double stranded, RNA nucleotide contains ribose instead of deoxyribose, and RNA has the base uracil instead of thymine, which is present in DNA. Genetic information is preserved by DNA replication, a process that produces identical copies of DNA molecules during each cell division into two daughter cells. In the transcription process, genetic information is transferred from DNA into RNA and then RNA specifies the synthesis of polypeptides, which subsequently form proteins. The DNA – RNA – polypeptide (protein) flow of genetic information has been described as the central dogma of molecular biology. The genetic code is a three-letter code. There are four possible bases to choose from each of the three base positions (4)3 in a codon, producing 64 possible codons. There are only 20 major types of amino acids and therefore each amino acid is specified on average by about three different codons. At least 100,000 different kinds of proteins can be produced.

30BCharacteristics of Mendelian inheritance The science of genetics began with the work of Gregor Mendel who published the results of his experiments on crosses between strains having inherited variations in the garden pea, in what came to be called Mendel’s Laws (1865). All genetic variation originates from a change in DNA sequence, a mutation. Mutations can affect germline cells or somatic cells; mutations

12

in somatic cells can lead to cancer, and germline mutations can be transmitted from one generation to another. Single-gene disorders arise as a result of mutations in one or both alleles of a gene on an autosome or sex chromosome. Heritable characters are determined by genes, where different genes are responsible for the expression of different characteristics. A gene is a specific coding sequence of DNA. Each individual carries two copies of each gene, one of which was inherited from the mother and the other from the father. A gene may occur in different forms called alleles, each potentially having a different physical expression. When a person has a pair of identical alleles, he or she is said to be homozygous; when the alleles are different, he or she is heterozygous. Compound heterozygous describes a genotype in which two different mutant alleles of the same gene are present. Mendelian inheritance patterns shown by single-gene disorders depend on the chromosomal location, autosomal (located on an autosome) or x-linked (located on the x-chromosome), and whether the phenotype is dominante (expressed when only one chromosome of a pair carries the mutant allele despite there being a normal allele on the other chromosome) or recessive (expressed only when both chromosomes of a pair carry a mutant allele). There are several factors that make the inheritance pattern of an individual pedigree and the phenotype of individuals in the pedigree difficult to interpret. The inherited disorder can have a variable penetrance, meaning that the expression of the disease trait is not absolute even if you are a mutation-carrier. The disease can have a variable expression, frequently seen in dominant inherited disorders. The disease can have age depending onset, meaning that the age at onset can be late for mutation carriers. Some people exhibit the trait but do not have the trait genotype; these are called phenocopies. De novo (new) mutations and genetic heterogeneity can be present.

31BInter individual variations The human genome DNA sequence is identical between individuals in 99.5-99.8% (KIDD et al. 2004): the small percentages that differ make each individual unique. Sequencing of the human genome has revealed multiple inter individual variations. The most frequent changes in the DNA sequence involve single nucleotide substitutions, insertions and deletions. They are usually not associated with disease unless they alter the coding sequence or an important regulatory sequence. These variants in-clude single nucleotide polymorphisms (SNPs), variable number of tandem repeats (VNTR), microsatellites, and restriction fragment length polymorphisms (RFLP). These repeat markers are multiallelic and they can be used to identify the parental alleles of individuals as well as to define recombinants between marker loci. Repeat sequences have been extremely important for human genetics studies since they have been used as genetic markers and form the basis of the genetic map of the human genome. A SNP is a polymorphism arising by a change in one single nucleotide. SNPs have only two alleles and are the most common form of genetic variation: approximately ten million SNPs have been identified in the genome sequence (BENTLEY et al. 2008; FRAZER et al. 2007; WANG et al. 2008; WHEELER et al. 2008). Copy number variants (CNVs), deletions and insertions have recently been re-cognized as an important source of DNA polymorphisms (MCCARROLL 2008). CNVs are found everywhere, in every size all along the human genome and the deletions or duplications can change the dosage of a gene. CNVs have been reported in children with mental retardation, small head size, autism and schizophrenia (STEFANSSON et al. 2008). New technologies have made personal genome sequencing possible. The base pairs from single individuals have been decoded, the genome sequence of J.C. Venter, James Watson an Asian and an Nigerian individual were sequenced using

13

Sanger sequencing and new short-read sequencing technology (BENTLEY et al. 2008; Levy et al. 2007; WANG et al. 2008; WHEELER et al. 2008). Next-generation sequencing technologies have a very high throughput; a hundred million DNA fragments can be sequenced in parallel on

the chip. The time needed to decipher a human genome, as well as the costs of sequencing have been considerably reduced. Massive parallel sequencing technology makes it possible to consider whole genome sequencing as a clinical tool in the near future.

14

Introduction 15BThe Skeleton

The human skeleton is a complex organ consisting of 206 bones. It has many key functions including mechanical support for movement, protection of vital organs, and blood and mineral reservoir. The skeleton consists of two tissues; bone and cartilage, and three cell types; osteoblasts, osteo-clasts, and chondrocytes. Chondrocytes and osteoblasts are of mesenchymal origin and osteoclasts belong to the monocyte-macrophage cell lineage. Abnormalities in the development, growth, and maintenance of these components give rise to many and varied forms of skeletal dysplasias. The skeletal dysplasias are a large group of clinically distinct and genetically hetero-geneous disorders. Each skeletal dysplasia is rare, but the birth incidence is approximately 2-3/10,000 for the whole group (ORIOLI et al. 1986). Clinical manifestations range in severity from mild growth retardation, precocious osteo-arthritis to prenatal lethality (ALA-KOKKO et al. 1990;; SUPERTI-FURGA et al. 1996; UNGER et al. 2008). There were 372 different conditions included in the 2006 revision of genetic skeletal disorders, and these were placed in 37 different groups defined by molecular, biochemical and/or radiological criteria (SUPERTI-FURGA and UNGER 2007). Skeletal dysplasias must be differentiated from one another for specific genetic counselling, prognosis and treat-ment. There has been an explosive increase in the description of biochemical and molecular defects in the skeletal dysplasias. However, the gene locus or molecular defect is still unknown for over half of the disorders.

16BExtracellular matrix

Extracellular matrix (ECM) or connective tissues play important supportive and structural roles in complex organisms

connecting cells, organs and other tissues (Fig. 1), (GUILAK et al. 2006; HEINEGARD and OLDBERG 1989). The ECM fills the spaces between cells and contains adhesion proteins that link components of the matrix both to one another and to attached cells (HEINEGARD and OLDBERG 1989). ECM matrices are composed of fibrous proteins embedded in a gel-like polysaccharide ground substance. The mechanical pro-perties of these tissues are determined by the composition of the ECM, and in particular by the amount, kind, and spatial orientation of its fibers. Therefore, the function of the ECM varies with the tissue. A range of osteochondrodysplasias is caused by mutations in components of the ECM in cartilage and bone (GLEGHORN et al. 2005; UNGER 2002). Significant and exciting insights into all aspects of vertebrate skeletal development have been obtained through molecular and genetic studies of animal models and humans with inherited disorders of skeletal morpho-genesis, organogenesis, and growth (COHN 2001). Mutations in the genes encoding structural proteins of the ECM affect one or more steps in the diverse set of coordinated events in the process of skeletal development. Depending on the role of the gene product and the severity of the defect, disturbance of 1) endochondral ossification and linear growth, 2) structural integrity and stability of articular cartilage, 3) and mineralization is seen. Better under-standing of genetic interactions, patho-physiological mechanisms and the structure and function of cartilage ECM molecules could give us a key to new treatments or even cures for skeletal dis-orders. With this background, it becomes important to study and understand the mechanisms behind skeletal disorders. Here clinical features and genetic back-ground of familial osteochondritis dissecans, multiple epiphyseal dysplasia and brachy-dactyly type A1 are studied.

15

a

b

c

d

e

f

g

h

ij

k

l

f

m

a

b

c

d

e

f

g

h

ij

k

l

f

a

b

c

d

e

f

g

h

ij

k

l

f

m

Fig.1. Schematic presentation of extracellular matrix proteins found in cartilage. Hyaluronan, aggrecan and link protein building large aggregates. Collagen type II, XI and collagen IX become cross-linked in cartilage, and the resulting fibrils arranged in different orientations within the matrix. Collagen IX cross-linking itself. COMP is cross-linking the collagen IX and collagen II.

a) Chondrocyte b) Aggrecan c) Link protein d) Hyaluronan e) Collagen II f) Collagen IX g) Collagen XI h) Fibulin i) Tenascin j) COMP k) Biglycan l) Collagen VI m) Fibromodulin

Reprinted with permission from the publisher, J. Dudia, CMLS, Cell. Mol. Life Sci. Vol. 62, 2005, 2241-2256. ECM, collagen network (Publisher Birkhäuser Basel) Modified from a schematic of Lorenzo and Heinegård, Lund University, Sweden.

17BCartilage

Cartilage is widely distributed throughout the human body, in the joints, trachea, nose, intervertebral discs and ears. Cartilage is categorized into three general subgroups - hyaline, elastic and fibro-cartilage - based on morphologic criteria and content of collagen and elastin (ROSS

1995). Hyaline cartilage the most wide-spread subgroup in humans includes articular cartilage. Collagens are members of the most abundant protein family of the ECM accounting for more than two-thirds of the articular cartilage dry weight. Aggrecan the major proteoglycan of the ECM in the articular cartilage, forms approximately one-third of the tissue.

16

Together with collagen, aggrecan maintains the visco-elastic properties of cartilaginous tissue (HARDINGHAM and FOSANG 1992). Many other structural proteins and non-collagenous proteins are present at varying but low concentrations within the complex three-dimensional collagen network (Fig. 1) (HEINEGARD et al. 1988; HEINEGARD et al. 1989). Chondrocytes are the single cellular component of hyaline cartilage responsible for the synthesis and maintenance of ECM. In healthy cartilage, there is steady-state equilibrium between anabolic and catabolic activity. The macromolecules turn over slowly (MAROUDAS et al. 1998) and the ability to remodel and repair the matrix is limited.

1BComponents of the Extracellular matrix network 18BThe large aggregating Proteoglycans

There are four members of the family of large aggregating proteoglycans, also named lecticans or hyalecticans (IOZZO and MURDOCH 1996) referring to their features and function. One common feature of theses molecules is the N-terminal domain (G1 domain) that binds hyaluronan. The proteglycans are composed of a core protein and one or more covalently attached sulfated glycosaminoglycan (GAG) chains (KIANI et al. 2002). GAG chains are large extended structures with repeated dis-accharide units, sulfate and carboxylate groups. The negatively charged anionic groups on the GAG chain create a large difference in ion concentration between the cartilage and surrounding tissue and water is drawn into cartilage because of this osmotic imbalance and water trapped in the extracellular matrix resists compression.

19BACAN-gene and Aggrecan

32BMolecular gene structure, organization and function Aggrecan has fundamental functions in chondroskeletal morphogenesis during

development. Skeletal development in the embryo starts with the formation of a hyaline cartilage template that is replaced by bone through the process of endochondral bone formation (HARDINGHAM et al. 1990). Cartilage persists in the epiphyseal growth plate until puberty where the continuous proliferation, maturation, and hypertrophy of chondrocytes take place, resulting in long bone growth.

Aggrecan together with hyaluronan forms big aggregates that give cartilage its unique gel-like properties and its resistance, essential for distributing load in weight-bearing joints. Aggrecan is encoded by the ACAN-gene and expressed by chondrocytes. The human aggrecan gene, located at chromosome 15q26 (KORENBERG et al. 1993), consists of 19 exons ranging in size from 77 to 4224bp (VALHMU et al. 1995). The complete nucleotide sequence of human aggrecan was reported by Doege (DOEGE et al. 1991). Its core protein is composed of three globular domains, G1, G2 and G3, and an extended central interglobular domain between the G2 and G3 domain dedicated to glycosaminoglycan (GAG) chain attachment (Fig. 2).

Exons 3-6 encode the G1 region and exons 8-10 encode the G2 region. The large exon 12 encodes the CS-rich domain and part of the KS rich domain. The G3 domain is encoded by exons 13-18, including the alternatively spliced epidermal growth factor-like and complement regulatory protein-like domains (Fig. 2). The N-terminal of the protein is formed by the G1 domain, which interacts with hyaluronan and link protein to build multimolecular aggregates (HARDINGHAM and MUIR 1972). The G1-domain contains three regions (the B, B`and A) that are important for maximal binding to hyaluronan and link protein, two proteoglycan tandem repeats (B and B´) and an immunoglobulin fold (A) (PERKINS et al. 1991). The interaction between aggrecan G1 domain and link protein is mediated via the immunoglobulin fold (GROVER and ROUGHLEY 1994). In addition to its binding interactions, the G1 domain of aggrecan

17

Fig.2. Schematic presentation of aggrecan its domains, the link protein and hyaluronan. The core protein of aggrecan is composed of three globular domains (G1, G2 and G3) and a large extended interglobular domain between the G2 and G3 domain for glycosaminoglycan chain attachment. The G1 domain interacts with hyaluronan and link protein and forms large aggregates that give cartilage its unique gel-like properties and its resistance, essential for distributing load in weight-bearing joints. The C-terminal G3 consist the EGF-like domain, the CRP and the C-type lectin domain. The EGF and CRP repeats of the G3 domain are subject to alternative splicing. (Modified from Heinegård et al., and reprinted with kind permission from the publisher. Dick Heinegård et al, FASEB J. 3:2042-2051; 1989. Structure and biology of cartilage and bone matrix noncollagenous macromolecules).

inhibit GAG modification and product secretion and thus regulate product pro-cessing (KIANI et al. 2001; YANG et al. 2000).

The interglobular domain (IGD) between G1 and G2 is encoded by exon 7 of aggrecan. The IGD contains sites of proteolytic attack of metalloproteinases, (aggrecanases) that cleaves the inter-globular domain (Glu373-Ala374) of the aggrecan core protein resulting in a loss of the whole GAG attachment region and the G3 domain (FOSANG et al. 1996). OA is associated with an increased loss of aggrecan fragments due to the action of both the MMPs and aggrecanases with predominant cleavage occurring in the interglobular domain (IGD) of aggrecan (STRUGLICS et al. 2006). The G2-domain shares some of the structural features of the G1 domain. Although it consists of two proteoglycan tandem repeats, the G2 domain does not show any HA-binding function or interactions with other ECM

molecules (FOSANG and HARDINGHAM 1989). The exact function of the G2 domain is unknown, but it might inhibit aggrecan secretion (KIANI et al. 2001). It may also have a role in the aggrecan specific GAG chain attachment. The G1 and G2 domain work together to inhibit the secretion of aggrecan, whereas the G3 domain and the chondroitin sulphate-glycosylated core protein region promote secretion (KIANI et al. 2001). Each aggrecan contains approxima-tely 100 chondrotin sulfate chains and 60 keratan sulfate chains.

The C-terminal G3-domain of aggrecan consists of three modules: the EGF-like module, the CRP and the C-type lectin module. The G3 domain may be of importance for the intracellular synthesis and maturation of aggrecan, an assumption derived from the knowledge that truncated protein in nanomelic chick fails to reach the Golgi elements and appears to be targeted for ER-associated degradation (VERTEL et al. 1993). The G3-domain regulates the

18

attachment of GAG chains and functions in GAG modification thereby enhancing aggrecan secretion (CHEN et al. 2002; KIANI et al. 2001). The central portion of the G3 domain is required for GAG chain attachment and product secretion (DOMOWICZ et al. 2000). The C-type lectin module of the G3 domain is known to interact with high affinity with ECM proteins in the fibulin- , fibrilin- and tenascin-families (ASPBERG et al. 1999; ASPBERG et al. 1995; ISOGAI et al. 2002; OLIN et al. 2001). The interaction of aggrecan with other ECM proteins suggests that they are essential for the functional assembly of aggrecan. The CRP domain may function by promoting glycosaminoglycan chain attach-ment to the core protein during aggrecan synthesis (YANG et al. 2000).

The epidermal growth factor (EGF)-like domains have two motifs in tandem and bind Ca2+ ions, these EGF-like domains,

and the CRP domain exhibit alternative splicing in aggrecan (DOEGE et al. 1991; FULOP et al. 1993). Variations of the G3 domain are produced by alternative splicing during post-transcriptional processing. All three of the globular domains of aggrecan contain sequences that are highly conserved in aggrecan from different species. The amino acid sequence of human aggrecan and rat aggrecan are about 75% identical. In paper II and III we describe chondro-dysplasias due to/associated with ACAN-gene mutations.

2BChondrodysplasias due to aggrecan defects

The understanding of aggrecan structure and function is facilitated by the discovery of the genetically based diseases involving aggrecan. Cartilage matrix deficiency (cmd) (RITTENHOUSE et al. 1978) is the first known

Fig. 3. Chondrodysplasias due to aggrecan defects in chick and mice. The disease causing mutation, phenotype and biochemical defects are shown. (A). Skeletal structures stained with alizarin red and alcian blue (for calcified bone and glycosaminoglycans of the cartilage, respectively) of day-12 chicken embryos, nanomelic mutant chick (right) and normal, wt (left). (B). Newborn cmd heterozygous and homozygous mouse mutant. (C). 10-day old normal and bm mouse. Reprinted with kind permission from the publisher. Nancy B. Schwartz, Glycobiology vol. 12 no. 4pp. 57R-68R, 2002.

19

genetic disorder of proteoglycans identified in mammals (KIMATA et al. 1981). There are four condrodysplasias described in animal models, the cartilage matrix deficiency in mice, the nanomelia in chickens, the cmd-Bc mice and the bulldog dwarfism in cattle, which results from mutations in the aggrecan gene (Fig. 3). There are two human diseases previously associated with an ACAN-gene mutation, the dominantly inherited Spondyloepiphyseal dysplasia type Kimberley (GLEGHORN et al. 2005) and the recessive spondyloepimetaphyseal dysplasia (TOMPSON et al. 2009).

3BChondrodysplasias due to aggrecan defects - Animal models

33BCartilage matrix deficiency

Mouse cartilage matrix deficiency (cmd/ cmd) is an autosomal recessive lethal mutation in mice resulting in a syndrome including disproportionate dwarfism, short snout and cleft palate (Fig. 3) (RITTENHOUSE et al. 1978). Mouse cmd is caused to a 7-bp deletion in exon 5 of the aggrecan gene, which results in a frame shift and a premature stop codon in exon 6 (WATANABE et al. 1994). The ossification is normal but the aggrecan is absent from the cartilage ECM, and as a consequence, the homo-zygous embryos have gross skeletal abnormalities (KIMATA et al. 1981). The mutation is lethal; the affected homozygous cmd mice die shortly after birth due to respiratory difficulties arising from the collapse of the defective cartilage rings in the trachea. The heterozygous cmd mice appear normal at birth, but they develop dwarfism and spinal misalignment later in life and die after 12-15 months (WATANABE et al. 1997), whereas wild type mice live for 2-2.5 years. Histological examination of the cervical spine of the heterozygote mice revealed degeneration and herniation of vertebral discs, and the authors suggest that aggrecan gene defects may be involved in genetic predisposition to disc herniation in

humans (WATANABE et al. 1997).

34BCartilage matrix deficiency-Bc

Cartilage matrix deficiency-BC (cmd-Bc) mice is another spontaneous mutation with the same locus and phenotype as the cmd/cmd mice. The cmd-Bc mice is caused by a deletion of exons 2 to 18 in the aggrecan gene, resulting in a significantly shortened mRNA (KRUEGER et al. 1999).

35BNanomelia Chick embryos homozygous for the autosomal recessive gene nanomelia (nm) exhibit a lethal form of micromelia (LANDAUER 1965) with reduced trunk and head sizes and gross skeletal abnormalities including shortened, broad and malformed limbs (Fig. 3). The molecular defect of nanomelia is a G to T transversion in exon 12 leading to a premature stop codon and production of a truncated core protein (LI et al. 1993). The nanomelic chick synthesize small amounts of a truncated core protein that fails to progress through the secretory pathway. Therefore chicken mutant nanomelia lacks aggrecan in its cartilage matrix and there is no modification of the GAG chains (VERTEL et al. 1993).

36BBulldog dwarfism Bulldog dwarfism in Dexter cattle is one of the earliest Mendelian inherited disorders described in animals. Homozygous Dexter cattle fetuses display extreme dis-proportionate dwarfism and die around the seventh month of gestation. The phenotype consists of a short vertebral column, marked micromelia, a large abdominal hernia, and a relatively large head with a retruded muzzle, cleft palate, and protruding tongue (CAVANAGH et al. 2007). Bulldog dwarfism is caused by a 4-bp insertion in exon 11 the ACAN-gene (CAVANAGH et al. 2007). Heterozygous cattle show a milder form of dwarfism, most markedly having shorter legs.

20

4BChondrodysplasias due to aggrecan defects – Human disorders

37BSpondyloepiphyseal dysplasia type Kimberley (SEDK) The spondyloepiphyseal dysplasia (SED) is a heterogeneous group of chondro-dysplasias characterized by radiographic abnormalities of the vertebral bodies and proximal epiphyses (HORTON 2002). Spon-dyloepiphyseal dysplasia type Kimberley (SEDK; [MIM 608361]) is a mild autosomal dominant form of SED. SEDK is charac-terized by short stature, radiological signs of flattened vertebral bodies, and early onset OA (ANDERSON et al. 1990). SEDK is associated with a single base pair insertion in exon 12 of the ACAN gene (GLEGHORN et al. 2005). SEDK was the first reported chondrodysplasia due to an ACAN-gene mutation.

38BSpondyloepimetaphyseal dysplasia (SEMD-aggrecan-type) An autosomal recessive form of Spondylo-epimetaphyseal dysplasia (SEMD), charac-terized by severe short stature, relatively macrocephaly, midface hypoplasia, rela-tively prognathism, and low set ears has recently been described associated with an ACAN-mutation (TOMPSON et al. 2009). The disease is caused by homozygous missense mutations in the C-type lectin domain within the G3 domain of aggrecan. The authors describe mild proportionate short stature in heterozygous carriers (TOMPSON et al. 2009). 39BFamilial osteochondritis dissecans Familial osteochondritis dissecans (fOCD; MIM 165800) is a rare skeletal dysplasia characterized by short stature or dwarfism, osteochondritis dissecans in knees, hips and elbows, tibia vara, scoliosis and early osteoarthritis (OA) (ANDREW et al. 1981; AULD and CHESNEY 1979; FONSECA et al. 1990; HANLEY et al. 1967; KOZLOWSKI and MIDDLETON 1985; MUBARAK and CARROLL 1979; PAES 1989; PHILLIPS and GRUBB 1985;

PICK 1955; STATTIN et al. 2008; STOUGAARD 1964; TOBIN 1957). There are approximately 50 familial cases reported in the literature (ANDREW et al. 1981; AULD and CHESNEY 1979; FONSECA et al. 1990; GARDINER 1955; HANLEY et al. 1967; KOZLOWSKI and MIDDLETON 1985; MUBARAK and CARROLL 1979; PAES 1989; PHILLIPS and GRUBB 1985; PICK 1955; STOUGAARD 1964; TOBIN 1957). OCD is defined as a separation of cartilage and subchondral bone from the sur-rounding tissue (SMILLIE 1960). The etiology of OCD is unknown, but it is probably heterogeneous including trauma, inflammation, ischemia, and defects of ossification (SCHENCK and GOODNIGHT 1996). Some authors have suggested that anomalies within the subchondral bone may preexist, predisposing to development of OCD from apparently minor trauma (GREEN and BANKS 1953; HANLEY et al. 1967; PICK 1955; SMILLIE 1960). Disturbance of the endochondral ossification may cause areas of local ischemic necrosis, pre-disposing the joint to OCD after microtrauma (YTREHUS et al. 2007). The formation of an abnormal cartilage due to altered networking of molecules in the cartilage may predispose to OCD. A main hereditary factor has also been suggested, supported by reports of familial occurrence of OCD (ANDREW et al. 1981; AULD and CHESNEY 1979; FONSECA et al. 1990; HANLEY et al. 1967; KOZLOWSKI and MIDDLETON 1985; MUBARAK and CARROLL 1979; PAES 1989; PHILLIPS and GRUBB 1985; PICK 1955; SMILLIE 1960; STOUGAARD 1964; TOBIN 1957). Our data from paper II show that disturbed function in interactions of aggrecan, as one major component in the cartilage ECM, is a key factor in the etiology of familial OCD.

There are indicators for a difference in the long-term prognosis and the natural course of OCD on the condyles of the femur in sporadic OCD compared with familial OCD (FONSECA et al. 1990; LINDEN 1977). OCD in joints with a closed epiphyseal line, adult OCD, almost always develops into OA; mild to severe gonarthrosis was developed

21

in 43 of 53 joints diagnosed to have had sporadic OCD in adulthood (LINDEN 1977). While the vast majority of sporadic OCD occurring before closure of the epiphyseal lines, juvenile OCD, heals spontaneously; two of 23 joints in non-familial cases of juvenile OCD developed mild gonarthrosis (LINDEN 1977). A higher incidence of gon-arthrosis in juvenile OCD in familial cases than in sporadic cases indicates an entirely separate patho-genesis from non-familial OCD (FONSECA et al. 1990; LINDEN 1977; STATTIN et al. 2008), results that are supported by findings in paper I and II.

Disorders associated with defective sulfation of proteoglycans

Aggrecan together with hyaluronan traps water in the ECM due to counter ions bound to the sulfate and carboxyl groups carried on the glycosaminoglycan chains of each proteoglycan molecule. Thus, skeletal phenotypes similar to aggrecan deficiency also result from mutations affecting glycol-saminoglycan sulfation (SCHWARTZ and DOMOWICZ 2002). In normal cartilage, chondrocytes synthesize large amounts of sulfated proteoglycans and secrete them into the ECM. Mutations that depress components of the metabolic pathway involved in sulfation exert severe effects in cartilage because chondrocytes produce undersulfated aggrecan molecules. As a result, negative charge and hydration in the cartilage ECM is reduced. Brachymorphic mice (bm) homozygous for the bm/bm locus is a nonlethal growth disorder characterized by disproportionately short size at birth and postnatal growth retardation. Bm-mice cartilaginous matrix stains poorly for proteoglycans (Fig. 3) and is caused by a mutation in the SK2-gene, belonging to the PAPS synthetase gene family (KURIMA et al. 1998). The bm-mice phenotype is due to undersulfated GAG chains. Brachymorphic mice breed normally and have life spans comparable to wild-type mice. In humans Spondyloepimetaphyseal

dysplasia (SEMD) Pakistani type, is caused by a nonsense mutation in the SK2 gene (UL

HAQUE et al. 1998). The disease is characterized by short stature, bowed lower limbs, enlarged knee joints and early-onset degenerative joint disease in hands and knees. Other skeletal dysplasias due to defective sulfation is diastrophic dysplasia (DTD; [MIM 222600]), recessive inherited multiple epiphyseal dysplasia (rMED; [MIM 226900], achondrogenesis 1B (ACG1B [MIM 600972]), and atelosteogenesis type II (AO2; [MIM 256050]) (HASTBACKA et al. 1992; HASTBACKA et al. 1996; SUPERTI-FURGA et al. 1996; SUPERTI-FURGA et al. 1999), all homozygous for mutations in the diastro-phic dysplasia sulfate transporter (DTDST) gene. rMED is the mildest disease of the group (see MED above). DTD is charac-terized by congenital severe club foot, hitchhiker thumbs and cauliflower ears (HASTBACKA et al. 1994). AO2 is a perinatal lethal disease and characterized by severely shortened limbs, small chest, scoliosis, clubfoot, abducted thumbs and great toes (HASTBACKA et al. 1996). ACG1B seems likely to represents the null phenotype of the DTDST gene (SUPERTI-FURGA et al. 1996), AO2 seems to represent a partial loss of function, and the DTD appears to be even milder loss of function.

20BDegenerative diseases involving aggrecan

40BVNTR-region of the ACAN-gene and OA Exon 12 of the aggrecan gene encoding the chondroitin sulfate 1 (CS1) domain exhibits size polymorphism. This variable number of tandem repeat polymorphism results in individuals with different lengths of aggrecan core proteins, bearing different numbers of potential attachment sites for chondroitin sulfate. Thirteen different alleles have been identified with repeat numbers ranging from 13 to 33 (DOEGE et al. 1997). The most common alleles possess 26, 27 or 28 repeats. Association between the human aggrecan CS1 polymorphism and

22

articular cartilage degeneration has been reported (HORTON et al. 1998; KIRK et al. 2003). The first report of a correlation between aggrecan polymorphism and intervertebral disc degeneration showed a significant difference between the distribution of allele sizes and the severity of disc degeneration on MRI (KAWAGUCHI et al. 1999). In addition, researchers have found no correlation between aggrecan CS1 polymorphism and degenerative disc disease and no correlation with primary OA and a protective effect of a certain length of aggrecan core protein against hand OA (KAMARAINEN et al. 2006; ROUGHLEY et al. 2006). Further studies are needed to establish whether aggrecan polymorphism is directly linked to intervertebral disc degeneration or OA. 41BAge-related changes of aggrecan The turnover of normal articular cartilage results in age-related changes in the structure of proteoglycan aggregates. Aggrecan structure varies with increasing age with respect to both core protein, GAG size, and the modification of the sulphation pattern of the CS chains. There is an increase in size and number of the KS chains and a decrease of the CS chains (HARDINGHAM and BAYLISS 1990). There is also a decrease in thickness and cellularity of articular cartilage with age (VIGNON et al. 1976). There is a proteolytic trimming of the molecule starting with the G3 domain, further proteolytic cleavage shortens the CS attachment region, and the interglobular domain is particularly susceptible to proteolytic activity (FLANNERY et al. 1992; FOSANG et al. 1992). About half of the mature aggrecan molecules in the cartilage extra cellular matrix are missing the G3 domain (PAULSSON et al. 1987). Alternations in cartilage in normal aging are in-dependent of the changes that occur in degenerative joint disease, but it is probable that some age-related changes predispose cartilage to the development of OA. Further studies are needed to clarify to what extent age-related alternations in aggrecan and the

ECM networking coordinate the initiation of cartilage degeneration.

5BComponents of the Extracellular matrix network

42BThe Collagens The ECM consists predominantly of colla-gens and proteoglycans and in smaller amounts of other non-collagenous proteins. Collagens constitute a heterogeneous protein-family where all of them contain one or several unique stiff triple-stranded helical domains (EYRE 2002). Collagen fibrils provide tensile strength but do not resist compression or bending. Type II collagen is the major collagen in cartilage. The turnover of collagen is thought to be very slow; in bone collagen molecules persist for about 10 years before they are degraded and replaced. By contrast, most cellular proteins have half-lives of hours or days. Collagen type IX and type XI are also cartilage specific collagens and bind to the surface of type II fibrils. Because they are thought to link these fibrils to one another and to other components in the extra-cellular matrix, they may be important for the organization of the fibrils in the matrix (ERLEBACHER et al. 1995). Collagen II mutations cause a spectrum of phenotypes ranging from prenatal lethal achondrogenesis, to hypochondrogenesis, to the various spondyloepiphyseal dysplasias and early-onset OA (NISHIMURA et al. 2005). These disorders can be grouped under the term type II collagenopathies, the first group of skeletal dysplasias for which the underlying genetic defect was found (LEE et al. 1989).

Type IX collagen is composed of three alpha chains, encoded by three genes: COL9A1, COL9A2, and COL9A3. Mutations in collagen IX genes have been found to cause multiple epiphyseal dysplasia (MED) (UNGER et al. 2008). In paper III we present a mutation in a gene not previously associated with MED, which extends the spectrum of mutated genes behind the disease. 6B

23

Chondrodysplasia due to defects in ECM-proteins

43BMultiple epiphyseal dysplasia Multiple epiphyseal dysplasia (MED) was initially described by Sir Thomas Fairbank (FAIRBANK 1945). He described MED “as a rare congenital development error charac-terized by mottling or irregularity in density and outline of several of the developing epiphyses, dwarfism and stubby digits”. MED is a clinically and genetically heterogeneous skeletal dysplasia; although it is primarily autosomal dominant rare autosomal recessive forms have been described (BALLHAUSEN et al. 2003; HUBER et al. 2001; SHEFFIELD 1998; SUPERTI-FURGA et al. 1999; UNGER et al. 2008). MED is one of the most common skeletal dysplasias, affecting about 1:10,000 individuals (BRIGGS and CHAPMAN 2002). Dominant inherited MED is characterized by pain and stiffness in joints and delayed and irregular ossification of epiphyses (BRIGGS and CHAPMAN 2002; BRIGGS et al. 1995; LACHMAN et al. 2005). In contrast to other chondrodysplasias the disease is not always characterized by significant short stature (HAGA et al. 1998; SUPERTI-FURGA et al. 1999). In childhood or early adulthood, the condition results in premature OA of both weight bearing and non-weight bearing joints. Spinal alteration is absent or slight, but Schmorl bodies and irregular vertebral end plates may be observed. Radiological findings consist of predominantly epiphyseal involvement and radiographic abnormal-lities can be present before the onset of physical symptoms (UNGER et al. 2008). Autosomal recessive inherited MED (rMED) is characterized by normal stature, brachydactyly, club foot and bilateral double-layered patella (UNGER et al. 2008).

Causative mutations have been reported in six genes: COMP (MIM 600310), MATN3 (MIM 602109), COL9A1 (MIM 120210), COL9A2 (MIM 120260), COL9A3 (MIM 1202170) and DTDST (MIM 226900). These genes are responsible for approxi-mately 50% of MED mutations (BRIGGS and

CHAPMAN 2002; ROSSI and SUPERTI-FURGA 2001). It seems likely that several more disease-causing genes contribute to the overall prevalence of MED, or a single gene yet to be identified. In paper III we present ACAN-gene mutations associated with MED.

Skeletonenesis In the embryo skeletal development begins with a process that delineates the number, size, and shape of the individual skeletal elements. When the pattern is set, mesen-chymal precursor cells migrate to the site of skeletogenesis, condense, and differentiate into chondrocytes forming the cartilage anlagen, which serve as templates for the later bones (ERLEBACHER et al. 1995). Bone formation occurs by two major processes: intramembranous and endochondral ossify-cation (OLSEN et al. 2000). Intra-membranous ossification develops flat bones such as the facial bones and cranium. This process involves direct differentiation of mesenchymal progenitor cells into bone-forming osteoblasts. Long bones of the limbs as well as the ribs develop in a process called endochondral ossification (OLSEN et al. 2000). During enchondral ossification the growth of the skeleton takes place in the cartilaginous growth plates; the cartilage is replaced by bone through a sequential process of cell proliferation, extracellular matrix synthesis, cellular hypertrophy, matrix mineralization, and vascular invasion (ERLEBACHER et al. 1995; LEFEBVRE and SMITS 2005). The prolife-ration of chondrocytes in the growth plate is under the control of a local feedback loop that depends on temporal and spatial location and primarily involves signaling molecules synthesized by chondrocytes: parathyroid hormone-related peptide (PTHrP) and the Indian hedgehog (IHH). This feedback loop regulates the rate at which chondrocytes leave the proliferative zone and mature into differentiated hypertrophic cells (Fig. 4) (BALLOCK and O'KEEFE 2003; EHLEN et al. 2006; OLSEN et al. 2000).

24

Fig. 4. Schematic diagram showing the Ihh-PTHrP pathway. During endochondral ossification, mesenchymal cells differentiate into chondrocytes, which mature, differentiate, and are replaced by bone. The feedback loop of the Ihh-PTHrP pathway regulates the pace of chondrocyte differentiation within the growth plate. Ihh expressed in the prehypertrophic zone of the growth plate, A) controls bone formation in the bone collar, B) stimulates chondrocyte proliferation directly, and C) stimulates perichondrial cells to express PTHrP. PTHrP inhibits Ihh expression and prevents proliferative cells from differentiating. Hypertrophic chondrocytes express type II and type X collagen, before apoptosis they express vascular endothelial growth factor (VEGF), which promotes the invasion of blood vessels into the cartilage tissue. The blood vessels transport cells for bone tissue forming in the ossification centre. (Reprinted with kind permission from the publisher. Bjorn R. Olsen et al. Bone Development. Annu. Rev. Cell Dev. Biol. 2000. 16:191-220)

22BThe Indian Hedgehog (IHH)-gene; an important gene for skeletal development

The IHH-gene located on chromosome 2q33-35, is composed of 3 exons spanning 5.5 kb of genomic DNA and consists of a highly conserved N-terminal signalling domain and a C-terminal catalytic domain (MARIGO et al. 1995). The IHH-gene encodes a paracrine signaling molecule. Many signa-ling factors are expressed only for a limited period of time during embryogenesis, genes

are turned on and off. Defects of signalling molecules, which are only transiently expressed during early embryogenesis, results in finite organ malformations. Ihh acts as a stimulator of growth plate chondrocyte proliferation; it prevents chon-drocytes hypertrophy and regulates bone formation in the perichondrial collar and of trabecular bone below the growth plate (Fig. 4) (EHLEN et al. 2006). Ihh expressed and secreted by pre-hypertrophic chondrocyte stimulates PTHrP expression in peri-articular perichondrial cells (VORTKAMP et

25

al. 1996). PTHrP seems to prevent chon-drocyte hypertrophy in the growth plate and maintains a pool of cells above the hyper-trophic zone in a proliferative condition (KARP et al. 2000). Thus the Ihh-PTPHrP pathway acts in a negative feedback loop to regulate the pace of chondrocyte differ-rentiation within the growth plate (Fig. 4). Mutations in the IHH-gene have been linked to two inherited skeletal develop-ment defects: a heterozygous mutation that causes Brachydactyly type A1 and homozygous mutations that cause Acrocapitofemoral dysplasia (GAO et al. 2001; HELLEMANS et al. 2003). Acrocapitofemoral dysplasia is a recessive inherited skeletal dysplasia characterized by short stature of variable degree with short limbs, brachydactyly, and cone-shaped epiphyses in the hands and hips (HELLEMANS et al. 2003). In transgenic mice, loss of Ihh leads to severe dwarfism due to reduced chondrocytes proliferation and no enchondral bone formation, resulting in mice with unsegmented and uncalcified digits. These mice die at birth due to respiratory insufficiency (ST-JACQUES et al. 1999).

In paper IV we study a human phenotype associated with a novel IHH-gene mutation. 44BBrachydactyly type A1 Congenital hand anomalies have an estimated prevalence of 1/200 (IVY 1957). Brachydactyly type A1 (BDA1; MIM 112500), one type of hand anomaly, was the first identified human disease with Mendelian autosomal dominant inheritance (FARABEE 1903). A century later McCready et al. identified the causal mutation in the Indian Hedgehog gene within the original family described by Farabee (MCCREADY et al. 2005). BDA1 is characterized by broad hands with shortening of all digits. The most severely shortened bones are the

middle phalanges that may be absent or fused with the distal phalanges (BELL 1951; FARABEE 1903; FITCH 1979). The feet show a similar pattern of abnormalities - absent or rudimentary middle phalanges and proxi-mal phalanx of the big toe. The disorder shows a considerable clinical variation both within and between families, and may also be present with other manifestations such as shortness of the long bones, malformed or absent epiphyses, scoliosis, abnormal menisci, club feet and short stature (RAFF et al. 1998; SLAVOTINEK and DONNAI 1998). Although mutations in the Indian hedgehog (IHH) gene have been found causative (GAO et al. 2001), genetic heterogeneity has been demonstrated. A second gene locus for BDA1 has been mapped to chromosome 5p13.3-p13.2 (ARMOUR et al. 2002) and a third locus has been suggested from the exclusion of both the IHH and chromosome 5 loci in one BDA1 family (KIRKPATRICK et al. 2003). Six different heterozygous missense mutations and one deletion in the IHH-gene have previously been identified in 13 BDA1-families of various ethnic background (GAO et al. 2001; GAO and HE 2004; GIORDANO et al. 2003; LIU et al. 2006; LODDER et al. 2008; MCCREADY et al. 2002; ZHU et al. 2007). These mutations are all located in the highly conserved amino-terminal signalling domain of the IHH protein, which is important for interaction with the ligand Patched (HALL et al. 1995). Mutations may cause BDA1 through a haploinsuffiency of the wild type Ihh, resulting in a decreased chondrocyte proliferation and attenuation of the negative feedback loop that regulates chondrocyte maturation during enchondral ossification. In paper IV, we report the clinical and radiological characteristics of a five-generation family segregating BDA1 and a novel IHH missense mutation.

26

Aims of the thesis The general objective of this thesis was to contribute with new knowledge about familial osteochondritis dissecans, multiple epiphyseal dysplasia and brachy-dactyly type A1, i) by describing the clinical and radiographic characteristics in the three types of inherited skeletal disorders, ii) by identifying disease causing gene mutations behind the three disorders and iii) by clarifying the functional consequences at the protein level. Specific aims were:

Paper I

To describe clinical and radiological findings in a family with an autosomal dominant inherited form of osteochondritis dissecans.

Paper II

To identify a locus and the gene causing familial osteochondritis dissecans and to identify the disease causing gene mutation.

To determine if an ACAN-gene variant identified has any effect on aggrecan G3 domain secretion and/or interactions.

Paper III

To search for yet unidentified genes associated with MED by using a candidate gene approach.

To identify MED causing gene mutations in the ACAN-gene.

Paper IV

To describe clinical and radiological findings in a family with autosomal dominant inherited brachymesophalangy.

To localize the disease causing gene. To identify the disease causing mutation and to predict an effect on the

structure or function of the IHH protein.

27

Methodology A summary of the main methods used in this thesis is given below.

45BEthics Study I, II and IV were approved by the Regional Ethical Review Board at Umeå University. Dnr 01-244, 07-109M study I and II, Dnr 06-058M study IV. Informed consent was obtained from each parti-cipating family member and/or their legal guardian. Study III was granted ethical approval by the local ethical committee for scientific research in the respective city/ country. 23BIntroduction to field work

46BCollecting clinical material In planning a linkage analysis, the first step is to determine how many affected indivi-duals or families one needs to localize the gene underlying the autosomal dominant trait of interest. The answer will depend on various factors such as delineation of the phenotype, penetrance of the disease, if there is a variable expression, age-related penetrance in late-onset diseases, false paternity, and heterogeneity. The organisa-tion and collection of clinical material is very taxing, and it must yield an adequate number of participants with a well-defined phenotype. 47BEvaluation of suspected skeletal dysplasias More than 300 recognised forms of skeletal disorders have been identified although every individual condition is very rare. Therefore, determining a specific diagnosis is demanding and a challenge. Dispropor-tional short stature is the most frequent clinical complication, but is not necessarily present. Once a skeletal dysplasia is suspected, clinicians need to gather the history, including the prenatal and detailed

family history. Next a clinical examination with anthropometric measurements and radiographs is performed (UNGER 2002). A full skeletal survey is important to obtain because the distribution of affected and unaffected areas is a key to the specific diagnosis (OFFIAH and HALL 2003). Since growth parameters are essential information, the height, weight and head circumference should be measured and recorded. Deter-mining proportions is done by measure-ment of the sitting height and then subtracting the sitting height from the total height to determine the upper segment to lower segment ratio. The arm span to height ratio and the upper to lower segment ratio are used to determine the most shortened segment, either the spine or the limbs. When the limbs are the most shortened it is possible to classify it in ritzomelic, meso-melic or acromelic depending on which segment is most affected. After the team of clinicians and radiologists has delineated the pattern of radiographic abnormalities and clinical features, a search of the medical literature and radiographic atlases can produce a matching pattern. Molecular genetic investigations are considered when there are a limited number of differential diagnoses. The establishment of a correct diagnosis is important for numerous reasons, including appropriate surveillance, proper clinical treatment, prediction of height, correct recurrence risk, and for some diagnosis prenatal diagnostics. 24BField work

48BStudy I and II (Familial OCD) The index case was a 15-year-old boy referred to the department of clinical genetics (Umeå University Hospital) for genetic evaluation of suspected skeletal dysplasia. There were several differential diagnoses possible and one very interesting familial osteochondritis dissecans (fOCD),

28

since there were very few cases described of fOCD in the literature and nothing known about the genetic aetiology. A full skeletal survey evaluated by the skeletal dysplasia team in Uppsala found a suspected skeletal dysplasia, but not the name of the disease. To establish a diagnosis, we invited all family members for clinical examination. Two of the authors (E-LS and YT) examined 43 family members at Läkarhuset Hermelinen in Luleå. In total, 53 family members were examined; ten of the family members went to the local nurse to be measured and to give blood samples. All family members answered an interview-based questionnaire on joint pain, affected joints, age at onset, pharmacological or surgical treatments and function. A clinical examination was performed that included anthropometric measurements: height, weight, arm span, sitting height and head circumference. Blood samples were collected from all participants. Data of previous radiological investigations were collected from patient’s records and all old radiographs were collected; additional radiological investigations were performed in 13 subjects. Four family members underwent MRI and one X-ray of knees and hips. Radiological examinations of hands were made in eight affected individuals. Additional radiological investigations were planned, but some family members were to busy to participate in the radiological examination. Some years later (2007), we invited all participants to Läkarhuset Hermelinen to inform them about the disease and the study results. At this occation, we also collected additionally samples. Twenty-six participants brought a urine sample (second micturition for the day) to the meeting. Blood samples were collected from 23 family members (12 affected and 11 unaffected), and joint fluid from the knee was collected from 9 family members. Plasma and serum was separated and samples were frozen at -70C. DNA was extracted and stored at the Department of Clinical Genetics. One individual underwent knee replacement surgery in September

2008 and another in February 2009. Both these participants donated joint fluid, bone marrow and cartilage. Anonymous control chromosomes from 115 Swedish individuals were used for the study.

49BStudy III (MED) Thirty-nine individuals with the clinical and radiological diagnosis of multiple epi-physeal dysplasia (MED) were recruited for the study: 24 cases from Canada, 13 cases from Japan and 2 cases from Sweden. All the cases were screened negative for known-MED genes (COMP, COL9A1, COL9A2, COL9A3, MATN3, DTDST). Cases from Canada and Japan have previously been described (ITOH et al. 2006; JAKKULA et al. 2005). Additional blood samples from parents of one child were collected for segregation analysis of the identified mutation in the family. Anonymous control chromosomes from 60 Canadian and 46 Swedish individuals were used for the study.

50BStudy IV (BDA1) A family with osteochondritis dissecans and brachymesophalangia was described 1978 by one of the authors (ANDREN et al. 1978). Since OCD was found in finger joints, the paper was relevant to my study. Conse-quently, I contacted one of the authors of this earlier study, Bjarne Lindén. Bjarne provided information that allowed me to organize a meeting with family members. We reinvestigated this family 30 years later. In total, the family comprises 26 family members, including seven affected indivi-duals. Clinical examination was given to six of these subjects (four females at ages 57, 57, 59 and 89 years and two males at ages 35 and 59 years), an examination that included a full medical history and bilateral radiographs of hands and feet. Peripheral blood samples were obtained from all seven affected individuals as well as from nine healthy family members. Anonymous control chromosomes from 115 Swedish individuals were used for the study.

29

25BMethods

51BStatistical analysis In study I ANOVA was used to compare mean values and chi-square was used to compare proportions. P-values of <0.05 were considered significant. Pearson’s correlation coefficient is shown with two-tailed significance. Statistical analysis was adjusted for gender. For statistical analysis the software SPSS version 12.0.1 was used. 52BLinkage analysis Genetic mapping is based on the principles of inheritance. Genetic linkage occurs when two loci are inherited together and located so close on the same chromosome that it is unlikely that they will be separated by a crossover event (recombination) during meiosis. The two loci are said to show genetic linkage if alleles at these loci segregate together more often than would be expected by chance (in >50% of all meioses). The recombination fraction (theta - θ) is the probability for a recombination to occur. Theta is a measure of genetic distance between two loci; 1% of recombine-tion corresponds to approximately 1 centimorgan (cM) of genetic distance. The closer together two loci are, the smaller the recombination fraction. If theta equals less than 0.5, then the loci are linked. A method used to describe the statistical likelihood of linkage between a genetic marker and a trait is the logarithm of odds (LOD) score (OTT 2001). The parametric (or model based) method for LOD score calculation is chosen when a disease is monogenetic and show a clear mode of inheritance. Parametric linkage analysis is a powerful way to detect genes causing monogenetic disorders. Lod score >3 is the threshold for significant linkage, and a Lod score less than -2 excludes linkage. These values are equivalent to the odds of 1000 to 1 in favour of linkage or 100 to 1 against linkage. For a whole-genome search a genome-wide threshold of significance must be used although most often a Lod score of 3.3 is used (LANDER and KRUGLYAK 1995). Linkage

analysis was performed in study II and IV. In study II analysis was performed using MCLINK (Myriad Genetics Inc.), which allows the analysis of very large pedigrees (ABKEVICH et al. 2001). Multi-point linkage analysis uses the information from several markers to test linkage between the phenotype and a specific position on the marker map. The multipoint analysis is more powerful than single point analysis. Parametric linkage analysis requires specification of genetic parameters, such as penetrance, disease allele frequency, phenocopy and mutation rates that describe the mode of inheritance included in the calculations. An autosomal dominant in-heritance model was used in the analysis and the disease allele frequency was set to 0.0001. Affected cases were assumed to be carriers of the disease allele with a 15% phenocopy rate. Unaffected individuals were given an unknown phenotype. In study III genotypes were assessed with the Gene Mapper 2 software program (Applied Bio systems, Foster City, CA). Two-point LOD scores were calculated with the Gene hunter and MLINK programs using an autosomal dominant model, with a penetrance of 100% and a disease frequency of 10-6. We assumed equal recombination frequencies between males and females. 26BMolecular genetics methods

53BDNA extraction DNA was prepared from 10 ml whole blood, according to standard salting out methods and stored at the department of Clinical genetics in study II, III, and IV. The identified mutations were confirmed in an additional sample - either a new sample or using the stored sample at the department of clinical genetics - to verify the findings and reduce the risk of sample mix up or contamination. PCR Polymerase chain reaction (PCR) is a technique that amplifies a single or a few copies of your target DNA and generates

30

millions or more copies of the DNA sequence (SAIKI et al. 1985). The process uses 15-20 bases long primers, each binding to the 3´end of the opposite DNA strands to be amplified. A special heat-tolerant DNA polymerase - Taq polymerase (from the bacterium Thermus aquaticus) - is used to replicate the single-stranded DNA segments extending from the primer. The DNA is denatured by heat, resulting in single-stranded DNA molecules. The temperature is lowered and the primers will hybridize to their complementary DNA sequences. Taq polymerase synthesizes the first set of complementary partially double-stranded DNA sequences. The third cycle will produce two double-stranded molecules that contain only the target sequence as well as six longer molecules that contain the target DNA and flanking sequence. Repeated cycles of denaturation, annealing and synthesis results in exponentially increased number of target sequence molecules. PCR is a very sensitive technique and has many applications in biology. It can amplify target sequences that are present in extremely low copy numbers in a sample. However, the technique has limitations, for example that you need available enough information of the sequence for the target DNA so that you can design primers. 54BGenome-wide scan Genetic mapping uses genetic markers that are spaced throughout the genome to discover how often two loci are separated by meiotic recombination. This information is used to find markers that are inherited with the disease and find chromosomal regions that are shared between affected family members. A whole-genome search for linkage was performed in study II. Samples from 38 family members were analysed using the ABI linkage panel set of 400 microsatellite markers with an average marker density of 10cM (ABI Linkage Mapping Set 10cM version 2.5 [Applied Biosystems, Foster City, CA, USA]) spanning all chromosomes not including chromosome X and Y. The primers were amplified using multiplex

PCR reactions according to the manu-facturer’s instructions. PCR products were resolved through 36 cm capillary arrays using POP-4 polymer and an ABI PRISM 3170 DNA sequencer. Genome-wide linkage studies are limited by the recombination rate. A genomic region linked to the disease might contain several hundred genes. In study II the candidate region spans approximately 100 genes. The specific chromosomal region (candidate region) found to be linked to the disease was subsequently fine mapped with additional microsatellite markers to narrow down the region, but all affected family members shared a common haplotype of 10.5 Mb. If the region is reduced to a 1- to 2- cM interval, it is possible to identify all of the genes encoded in that region and sequence them in affected individuals. If the sequence of a DNA molecule is known, then the genes that it contains can be identified and the activities of those genes can be studied in detail.

55BSequencing/Candidate gene analysis Sequence analysis of candidate genes was performed in study II, III and IV. Amplified PCR products of samples form affected and unaffected family members were sequenced in both sense and antisense directions. The same primers for the ACAN gene were used in study II and III, primers published in paper II, Table 2. Amplicons with variant DHPLC elution peaks in study II, III and IV were sequenced for detection of sequence variants using the same primers used in the original amplification. The sequencing was per-formed with an ABI PRISM 3170 DNA sequencer following the standard protocols recommended by the manufacturer. Sequences obtained were aligned to and com-pared with published wild type sequence (GeneBank accession numbers NM_013227.2 ACAN-gene and NM_002181.2 IHH-gene) using Sequencher 4.7 analysis software. To determine the base sequence and identify variations we used Sanger sequencing or

31

dideoxy sequencing (ABI PRISM 3170), an automated DNA sequencer using fluore-scence labelling. The technique is based on deoxynucleotide (dNTP) and dideox-nucleotide (ddNTP) (SANGER et al. 1977). Four separate fluorescent dyes label the base-specific reactions. A ddNTP differ from the dNTP in that they lack the 3´-hydroxyl group as well as the 2´-hydroxyl group. For the DNA synthesis, the DNA polymerase must catalyse a condensation reaction between the 3´-hydroxyl group of the last nucleotide added to the growing DNA chain and the 5´-phosphate group of the next nucleotide to be added, releasing water and forming a phoshodiester linkage with the 3´-carbon atom of the adjacent sugar. Because a ddNTP lacks the 3´-hydroxyl group, this reaction cannot take place. Therefore, ddNTP blocks continued DNA synthesis. Each coloured peak represents a different-size fragment of DNA, ending with a fluorescent base; four different colours represent the four bases of DNA. During the electrophoresis run, a laser beam is focused at a specific position on the gel. As the individual DNA fragments migrate past this position, the dyes fluoresce in different wavelength. The information is recorded electronically and the sequence is stored in a computer database. Dideoxy sequencing is a fast and robust method to detect sequence variations. However, to determine the significance of the identified variation is difficult (see the section below). Failure to detect mutations may arise for a number of reasons. One reason could be that only the allele without the mutation was amplified in the PCR reaction due to a large deletion of one allele. In addition, polymorphisms in the primer can infer with the PCR reaction, and the disease causing mutation can be an unusual splice mutation located in the intron or close to the intron making it difficult to detect. Another reason for not detecting a mutation is that the disease can be genetically heterogeneous; that is, the wrong gene that has been identified for examination.

56BInterpretation of identified variations Sequencing will detect any nucleotide change present, and it becomes important to distinguish benign changes from pathologic ones, which is not as easy or straightforward as it may seem. Nonsense and frameshift mutations are usually pathologic. Missense mutations that lead to a conservative amino acid substitution may disrupt the transcription and change the protein structure. A major amino acid substitution may not be deleterious if it lies within a nonessential domain of the protein. To determine if an identified sequence variation can be of importance for the disease, bioinformatics programs are used to predict the significance of the change of amino acid to the protein function. SIFT is a sequence homology-based tool that sorts intolerant from tolerant substitutions, and classifies substitutions as tolerated or deleterious (NG and HENIKOFF 2003). SIFT predicts whether an amino acid substitution at a particular position in the protein will have a phenotypic effect. Mutation analysis should be performed in a normal popula-tion and the allele frequency of the mutation should be calculated and checked for whether the mutation segregates together with disease in the family. A gene that consistently has a mutation in the affected individuals but not in the unaffected individuals is most likely the disease gene. Functional assays can elu-cidate the effect of mutations. We used all these tools to interpret the identified mutation in paper II, and all the tools but functionally analyses were used in paper III and IV. In paper IV we also used protein structure prediction and modelling using the SWISS-MODEL (ARNOLD et al. 2006).

57BDHPLC

Allele frequencies of sequence variants detected in study II, III and IV were determined by PCR amplification of the respective amplicons with the same primers used for sequence analysis. A mutation-

32

positive as well as a confirmed normal DNA control was included for each gene frag-ment screened on the high-performance liquid chromatography (dHPLC). DNA from 230 control chromosomes was used in study II and IV and 212 Caucasian control chromosomes (120 Canadian and 92 Swedish) were used in study III and compared (using dHPLC, Transgenomic Wawe 3500HT system) to sequenced samples with presence/absence of the sequence variant. All amplicons with variant DHPLC elution peaks were sequenced for detection of sequence variants. DHPLC is a technology for screening of gene mutations through detecting heteroduplex DNAs (XIAO and OEFNER 2001). It is faster and cheaper than sequencing, but it cannot be used as the only method: any abnormality detected must be confirmed by DNA sequencing of the fragment for precise mutation location, identification, and interpretation. Chromato-graphic processes can be defined as separation techniques involving mass-transfer between a stationary and mobile phase. HPLC uses a liquid mobile phase to separate components of a mixture. These components are first dissolved in a solvent and then forced to flow through a chromatography column under high pressure. Heteroduplexes will be separated form homoduplexes. DHPLC detects genetic variations such as SNPs, deletions, insertions and missense mutations. Failure to detect a mutation can be due to poor quality PCR; that is, very low or high yields can affect the ability to see mutations. Mutations located in a high melt GC-rich pocket can be hard to detect and it may be necessary to redesign primers with the GC-rich regions near the ends of the fragments. Mutations in small fragments (< 110-120 bp) or fragments to large (> 500 bp) can remain undetected. 58BRestriction endonuclease digestion of the PCR product Restriction enzymes are bacterial enzymes that cut DNA at specific base sequences,

restriction sites, in the genome. Restriction endonuclease digestion was used in study III to evaluate the frequencies of the mutations in exons 6 and 7 (212+1 samples) on control chromosomes. Both mutations result in novel recognition sites for restriction enzymes. The mutant variant in exon 6 creates a restriction site for BcnI (CauII) (#ER0061, Fermentas) (ref: 108bp, 157bp and 177bp, and the mutation of 157bp and 285bp). The novel variant identified in exon 7 creates a restriction site for Mva1269I (BSMI) (#ER0962, Fermantas) (ref: 478bp, 323bp, and the mutation of 801bp). PCR amplicons across the location of the mutations were digested and the resulting fragments were separated and visualized on a 4% Agarose gel (NuSieve® GTG®). 59BProtein interaction analysis Biomolecular interaction analysis (BIA) is a technique that measures protein-protein interactions and binding affinity. BIAcore uses a technology based on surface plasmon resonance (SPR), an optical phenomenon that enables detection of unlabeled inter-actants in real time. Detection depends on changes in the mass concentration of macromolecules at the biospecific interface and does not require any labelling inter-actants. BIAcore can be used to characterize molecules in terms of specificity of interaction, kinetics, and binding strength, followed in real time, so that kinetic information is readily derived. One of the interacting molecules is immobilized on the surface of a sensor chip. A potential interaction partner is then injected in solution and flows over the sensor surface. As an interaction partner binds to immobilized molecules, refractive index at the surface alters in proportion to the change in mass concentration. This results in a change in SPR signal (the value of the resonance angel expressed in resonance units) that is detected in real time and presented as a sensorgram (presentation of resonance signal as a function of time). Sensorgrams display the formation and

33

dissociation of complexes over the entire course of an interaction; the kinetics (association and dissociation rates) are revealed by the shape of the binding curve. The affinity (strength) of the interaction is determined from steady-state responses or from the kinetic values. In study II we used Real-time BIA (Biacore AB).

60BMass spectrometry Mass spectrometry sorts molecules from each other depending on the relationship between the mass and ionization. There are different methods for ionization of the molecule: Matrix-assisted laser desorption

ionization (MALDI), Electrospray ionization (ESI), etc. Matrix-assisted laser desorption/ ionization time-of-flight mass spectrometry (MALDI-TOF MS) and Ion trap MSMS were used in study II. MALDI-TOF MS is an ionization technique used in mass spectro-metry allowing the analysis of bio molecules and large organic molecules. The DNA fragments are mixed with a carrier matrix and painted on the desorption target surface. Then a laser is used to desorb and ionize the DNA fragments and acceleration in a mass spectrometer allows fragment length determination by time of flight detection.

34

Results and discussion

7BPaper I We describe a five-generation family with an autosomal dominant inherited osteo-chondritis dissecans in knees and/or hips and/or elbows associated with dispropor-tional short stature and early osteoarthritis. The clinical and radiological features of afflicted family members indicated the diagnosis familial osteochondritis dissecans (STATTIN et al. 2008). The most important differential diagnosis - multiple epiphyseal dysplasia (MED) - was excluded when re-evaluating radiological investigations, since we found no radiological changes consistent with epiphyseal dysplasia or multiple epiphyseal dysplasia in our family. Because it can be difficult to exclude the diagnosis a skeletal dysplasia expert evaluated all old x-rays inclusive all affected children.

In total, 53 family members were examined, and we re-evaluated all radio-logical examinations available from these individuals, all together 31 family members, including spouses. We re-evaluated a total of 202 radiological examinations and 39 of these were from unaffected family members. Out of these 202 radiological examinations there are 13 (new) performed at the time of the study. Fifteen family members from whom we obtained information of joint and skeletal disease were confirmed as affected by familial osteochondritis dissecans after clinical examination and radiological evaluation showing characteristic changes in at least one joint. A proven case of OCD had either a positive history of radiographic OCD in the medical record, a surgical removal of a loose body, or a radiograph demonstrating an osteochondral fragment and/or a defect. Ten males and five female were affected; we believe the gender difference is coincidental.

The common features of the family members include OCD in knees and/or hips and/or elbows, early onset OA and

disproportional short stature with a long trunk in relation to body height. Familial OCD have previously been described in combination with dwarfism or variable short stature, multiple OCD lesions, OA and tibia vara (ANDREW et al. 1981; FONSECA et al. 1990; HANLEY et al. 1967; MUBARAK and CARROLL 1979; PAES 1989; PHILLIPS and GRUBB 1985; PICK 1955; TOBIN 1957), results that are consistent with our findings.

Thirteen affected family members had a body height of -2 standard deviations (SD) or less then -2 SD when compared to a Swedish reference population (WIKLAND et al. 2002). Two individuals had a more normal body height, one of which was a still growing child (height -1 SD) and the other a male who had been treated with growth hormone before puberty (height 168cm; -1.9 SD). The male who had been treated with GH during childhood had a body height of 168 cm, which is 11.5 cm taller than the mean height of the affected males. This observation suggests that children with familial OCD might benefit from GH treatment. However, this needs to be evaluated in further studies.

Growth charts were available for five subjects and height at birth was within ±2SD in all of them. The prepubertal longitudinal growth in these subjects was between ±0 and -2SD. The growth spurt appeared to be poor resulting in a final height of -2 to -4 SD. This finding suggests that the short stature among affected subjects is a result of reduced pubertal growth. However, there were only five growth charts available and further studies are required to find out the specific growth pattern in fOCD. Predictive testing in family members is possible, hence the mutation is known. However, we have not offered pre-symptomatic testing because there is no efficient treatment to offer.

Eleven out of 14 cases had an arm span

35

of ≤2SD compared to a reference population (GERVER and DE BRUIN 1996), but normal compared to their own height. Five out of 14 individuals had a sitting height of ≤2SD and the ratio sitting height/total height was ≥2SD in 10 out of 14 when compared to a reference population (FREDRIKS et al. 2000), indicating short legs and a long trunk. Affected family members had normal head circumference.

We found significant differences between affected and healthy subjects in height, weight, arm span, sitting height, head circumference and shoe number (Table III, paper I). When comparing data from the interview-based questionnaire between affected and healthy family members, we found a significant difference in joint pain, pain > 6 month, knee pain, surgical treatment, and knee-, hip-, elbow-, and tibia operation. We found no difference in ankle-, hip-, elbow-, tibia-, or pain in another joint. We found no difference in consumption of pain killers (Table IV, paper I).

Part of this family has been reported previously. There was a reference (STOUGAARD 1964) in a deceased family member’s medical record. Stougaard found that short stature in this family was associated with OCD in knees and elbows and tall stature was associated with OCD only in the elbows. In our study, we could not confirm this association between OCD in a specific joints and body height. In contrast to what was previously reported, our examination of all living family members showed that all affected adults had a significant shorter stature than unaffected family members. One explana-tion for Stougaard´s finding could be that one tall family member had OCD in one elbow for other reasons than fOCD (a phenocopy).

Fifteen affected family members had OCD at the lateral aspect of the medial femoral condyle and five were diagnosed with OCD in the femoral patellar joint. Five individuals had OCD located at the femoral head and seven had OCD in the elbow,

located in the humeral capitellum. None of the healthy family members had signs of OCD.

OA was observed in the medial knee joint in 12 cases, in the hip in six cases and in the elbow in five cases. OA was found in five unaffected family members, one of them underwent knee replacement and another underwent hip replacement. The other three unaffected family members with OA had hip arthritis, mild degenerative changes in the spine and OA in the caput of the metacarpal bone I.

The vast majority of common OCD lesions which appear when epiphyseal lines are open, will heal spontaneously and only rarely will they result in secondary OA (LINDEN 1977). The likelihood of develop-ment of OA depends on the location and size of the OCD (TWYMAN et al. 1991). The classically placed lesion at the medial femoral condyle seen in all cases of this family seldom results in OA and this area is non-weight-bearing and so less prone to OA (TWYMAN et al. 1991). On the other hand, almost all individuals afflicted by OCD at the time when the epiphyseal lines are closed develop OA (KOCHER et al. 2006; LINDEN 1977). Despite early onset of OCD, seven out of 12 affected family members with onset during childhood or adolescence had developed OA. This is in contrast to the previously described natural history of juvenile OCD and should be considered in the follow-up of familial OCD. OA is a common disease in elderly people and could be a confounding factor; some of the cases may be expected to develop OA regardless of OCD.

Mutations in genes encoding structural components of articular cartilage such as genes encoding types II, IX and XI collagens have been demonstrated to be associated with an early onset OA (ALA-KOKKO et al. 1990; SNEAD and YATES 1999; UNGER et al. 2001).

In our family, the inheritance pattern suggests a monogenetic predisposition to OCD, and candidate genes include those encoding extra cellular matrix proteins and

36

genes involved in cartilage and bone growth and maturation.

In this family linkage to collagen II, IX, COMP and MATN3 was excluded by using of micro satellite markers and linkage analysis (not published data).

In 1888, König coined “osteochondritis dissecans” as the term to describe a condition characterized by loose bodies in joints. However, osteochondritis dissecans is an inappropriate term to use because it is generally agreed that inflammation is not the aetiology of the primary lesion. The term osteochondritis has been replaced with osteochondrosis by many scientists and might be a more appropriate term. We used the term osteochondritis dissecans because it is still in use today and to my knowledge the most commonly used term. Furthermore we want to emphasize that the OCD is part of a dominant inherited phenotype in this family (fOCD; MIM 165800). One weakness of the study is that we did not take new radiographs of all affected family members at the time for the study.

To conclude study I, we show that this familial form of OCD is associated with an altered childhood longitudinal growth resulting in a relatively long trunk compared to total height as well as a reduced normal growth spurt resulting in short stature. Our observations also indicate that this inherited form of OCD, with onset before epiphyseal closure, is associated with the development of OA.

8BPaper II We describe the molecular mechanism for the aetiology of familial osteochondritis dissecans and confirm the importance of the aggrecan C-type lectin interactions for cartilage function in vivo.

A significant linkage to the disease locus for a marker on chromosome 15q26 was found using alleles generated from 382 markers spanning all autosome, except the chromosome X and Y. All family members shared a haplotype of 10.5 Mb restricted by

the SNP markers D15S205 and D15S130. Fine mapping of the locus, using 40 additional microsatellite markers, gave no further information. The locus contained approximately 100 genes, and among them, the ACAN-gene and the HAPLN3-gene, which encode aggrecan and proteoglycan link protein 3, respectively. Aggrecan was our prime candidate gene for fOCD based on the fundamental importance of aggrecan for skeletal development during embryo-genesis (HARDINGHAM et al. 1990). We performed sequencing of the coding region of the ACAN gene, including the alterna-tively spliced EGF2 repeat, in two affected and two unaffected family members. Due to technical difficulties associated with the high content of VNTRs, parts of exon 12 and 13 were not investigated. Therefore, we cannot exclude the possibility that the family can have another mutation in this region or in the regulatory elements in the gene.

Sequencing revealed a G to A transition in exon 17 leading to a Valine to Methionine amino acid substitution of the coding sequence. The mutation was identified in all affected family members (19 individuals) and excluded in all unaffected (34 indivi-duals). The carrier status of the mutation was confirmed by repeated sequence analysis of the exon in independent blood sample or independent stored DNA in all affected family members. The mutation was excluded on 230 control chromosomes derived from unrelated healthy military recruits.

Fifteen affected family members were described in paper I (STATTIN et al. 2008). In paper II, additional four affected family members were included. Two of these (pedigree number V:5 and V:7) developed osteochondritis dissecans in the knees at the age of thirteen years, after we had collected data and radiological examina-tions for paper I, whereas the remaining two family members (pedigree number IV:5 and IV:18) showed a milder phenotype but were found to carry the familial haplotype and mutation. Neither of these two indivi-

37

duals (pedigree number IV: 5 and IV:18) fulfilled the inclusion criteria, OCD in at least one joint; however, one did not go through radiological examination of knees and hips and the other had an exudation in one knee. Both of them have short stature, 160 cm (height -3SD, sitting height per-centage of total height+1SD) and 158 cm (height -1.5SD, sitting height percentage of total height+2.5SD). Pedigree number IV: 18 is 158 cm tall an her sister is 168 cm tall and their mothers height is 168 cm. This information supports the diagnosis of fOCD. Both of them considered themselves to have the familial disease, and one of them had knee pain since the age of 10-20 years of age. We suggest that there is a variable expression of the disorder, and that they have a mild phenotype. Autosomal dominant conditions are more variable than recessive ones, probably because the phenotype of the heterozygote involves a balance between the effects of the two alleles. Ytrehus et al. suggest a refinement of the terminology of OCD to include the modifiers latens, manifesta and dissecans (YTREHUS et al. 2007). The knee exudation in one patient could be the expression of osteochondritis manifesta, a focal failure of endochondral ossification, seen on the radiographic examination but not osteo-chondritis dissecans with a fissure that extends trough the articular cartilage.

We identified two other families in the county of Norrbotten with OCD in knees and elbows in two generations: a father and his son in one family and a mother and her son in the other. Mutation analysis revealed that all four carried the V2303M substitu-tion, and we assume that they are related to the five-generation family described and that they all belong to a common ancestor.

To find out if the Val to Met substitution also is a cause of sporadic OCD, we sequenced exon 17 together with a positive control in 26 cases with sporadic OCD (unpublished data); none of the sporadic OCD cases carried the mutation.

To determine whether the V2303M substitution has any effects on aggrecan G3

domain secretion and/or interactions, we produced wild type and mutant human aggrecan G3 domains of three different splice variants with and without the V2303M substitution. All three mutated G3 variants were expressed and secreted by the human embryonic kidney 293 cell line, although the protein yields of the shorter mutant recombinant proteins were lower than those of the corresponding wild type proteins.

The C-type lectin domain of aggrecan is known to interact with ECM proteins in the tenascin (ASPBERG et al. 1995; ASPBERG et al. 1997), fibrillin (ISOGAI et al. 2002), and fibulin (ASPBERG et al. 1999; OLIN et al. 2001) families. We found co-purification of the CLD ligand fibulin-1 with all three wild-type G3 variants, but not with any of the V2303M mutated G3 domains using his-tag affinity precipitation of the recombinant proteins from conditioned medium using NiNTA beads. The identity of fibulin-1 as well as that of the recombinant proteins was verified by MALDI-TOF mass spectrometry. The affinity for tenascin-R was decreased in the mutated G3 domain compared to wild type. Analyses of interactions by surface plasmon resonance using a BIAcore verified that tenascin-R, fibulin-1 and fibulin-2 binding were affected for all V2303M mutated G3 splice variants with weakened affinities or complete loss of interaction. These results clearly demonstrate that the familial osteochondritis dissecans-associated V2303M substitution affects normal aggrecan G3 domain function. Furthermore, the strictly conserved Val2303 residue is located in one of the beta strands at the known ligand interaction site of the aggrecan c-type lectin domain (LUNDELL et al. 2004) with its side chain buried within the tightly packed hydrophobic core of the protein fold. Molecular structure modelling of the ligand interaction site of the aggrecan c-type lectin domain suggests that the Val2303M substitution leads to a conforma-tional change of the aggrecan c-type lectin. From these findings, we suggest that the interaction between CLD and its ligands is

38

affected by the Val2303Met variant. We studied proteoglycans from articular

cartilage of an affected family member undergoing knee arthroplasty to determine whether aggrecan containing the V2303M mutation is actually produced and secreted in vivo. Mass spectrometry showed the presence of peptides with both wild-type sequence and the V2303M sequence. The sample was further analyzed by Ion-Trap MSMS to verify the peptide sequences, a test that confirmed the presence of both the wild type and the V2303M aggrecan sequences.

The total amount of proteoglycan was quantified using Alcian blue precipitation. The proteoglycan contents of a control osteoarthritic knee articular cartilage extract (pooled from 10 individuals) and of the patient extract as well as of purified aggrecan (D1 fraction) from the OA cartilage pool and the patient were similar (26.8 versus 24.8 mg/g and 13.1 versus 14.0 mg/g, respectively). These results indicate that there is no major alteration in aggrecan secretion or glycosaminoglycan sulfation in the patient cartilage.

The clinical features of this disease include osteochondritis dissecans, premature OA, and mild disproportionate short stature. All of these symptoms are consistent with an impaired matrix assembly and function due to the aggrecan G3 dysfunction.

The aetiology of osteochondritis disse-cans is likely to be heterogeneous including trauma and micro fractures resulting from the cumulative damage of repetitive stress or injury and/or fragility of the cartilage and bone (SCHENCK and GOODNIGHT 1996). Disturbance of the endochondral ossifica-tion may cause areas of local ischemic necrosis that predispose the joint to osteochondritis dissecans after micro trauma (YTREHUS et al. 2007). The forma-tion of an abnormal cartilage may predispose to osteochondritis dissecans that may result from an altered networking of molecules in the cartilage. Our results show that disturbed function in a domain of aggrecan interacting with other extra-

cellular matrix molecules is a key factor in the aetiology of familial osteochondritis dissecans.

Sporadic osteochondritis dissecans pre-dispose for early onset OA (LINDEN 1977; TWYMAN et al. 1991), but lesions occurring before the closure of the epiphysis usually heal spontaneously (LINDEN 1977). In contrast, almost all individuals with osteochondritis dissecans lesions occurring after growth plate closure develop OA (LINDEN 1977). Early onset OA is part of the phenotype in our family, although the lesions occur before growth plate closure (STATTIN et al. 2008). Premature OA is part of the pathology associated with many mutations affecting ECM assembly or stability, e.g., in Stickler syndrome (COL2A1, COL11A1, COL11A2) and multiple epiphyseal dysplasia (COL9A1, COL9A2, COL9A3, COMP, MATN3) (LI et al. 2007). Since sporadic osteochondritis dissecans before growth plate closure does not lead to OA, it appears probable that the early onset OA associated with familial osteochondritis dissecans is due to disorganized and weakened cartilage ECM caused by loss of G3 interactions in the V2303M aggrecan.

Another clinical feature of familial osteochondritis dissecans is dispropor-tionate short stature (STATTIN et al. 2008). The ACAN gene has been identified by genome-wide association analysis as a locus influencing adult height (WEEDON et al. 2008). Mutations in the aggrecan gene resulting in chondrodysplasias have been identified in several species.

The majority of described aggrecan mutations are either null (cmd-Bc and Dexter Bulldog cattle BD2) or functional null (cmd, nanomelia, Dexter Bulldog cattle BD1, SEDK). In the case of ACANBD1, this has been verified to be due to nonsense mediated decay (CAVANAGH et al. 2007), which is likely the case for all the functional null mutations. The remaining mutant aggrecan mRNA may be translated, but secretion of truncated aggrecan is prevented by proteosomal degradation (VERTEL et al. 1993). Homozygous mutations result in

39

lethal skeletal dysplasias; in heterozygous state, these mutations lead to aggrecan haploinsufficiency and dwarfism phenol-types in the case of SEDK with early onset OA in weight bearing joints (EYRE et al. 2002; GLEGHORN et al. 2005). Aggrecan attracts water that provides a swelling pressure to tissue ECMs all through osmotic attraction of water by counter ions bound to the sulfate and carboxyl groups carried on the glycosaminoglycan chains of each proteoglycan molecule. Thus, skeletal phenotypes similar to aggrecan deficiency also result from mutations affecting glycos-aminoglycan sulfation (SCHWARTZ and DOMOWICZ 2002). Our data are consistent with a normal V2303M aggrecan secretion and we have no reason to believe that sulfation is affected, because of the normal result of the Alcian blue assay. We suggest that the loss of G3 interactions in the mutant may contribute to a disorganized ECM in the growth plate during endochondral ossification and thus lead to decreased long bone growth in the affected individuals.

The importance of the G3 interactions for skeletal development is supported by the recent publication of an additional missense mutation (D2267N) in the human aggrecan CLD (TOMPSON et al. 2009). Clinical characteristics of the recessive D2267N spondyloepimetaphyseal dysplasia, are extreme short stature, brachydactyly and cranio-facial abnormalities. Unlike the familial osteochondritis dissecans patients, these patients showed no signs of OA, but they are still very young. Mild proportionate short stature was reported for the hetero-zygous parents and a half sibling for the D2267N mutation (TOMPSON et al. 2009).

To conclude study II, we provide a molecular explanation for autosomal dominant familial osteochondritis dissecans. Our results clearly demonstrate that the familial osteochondritis dissecans-linked V2303M substitution affects normal aggrecan G3 domain function.

9BPaper III We describe for the first time mutations

in the ACAN-gene associated with auto-somal dominant and recessive inherited MED.

The ACAN-gene was considered to be a strong candidate gene for MED because it encodes the core protein of aggrecan, a major component of the cartilage extra cellular matrix, and because of the essential importance for skeletal development and cartilage function (HARDINGHAM et al. 1990). ACAN-gene mutations have pre-viously been found to cause skeletal dysplasias in humans (GLEGHORN et al. 2005; TOMPSON et al. 2009), and we describe another one in paper II. A recessive syndrome of macrocephaly, MED and dysmorphic features has been mapped to chromosome 15q26, but the aggrecan gene was excluded as a candidate gene (BAYOUMI et al. 2001).

Sequence analysis revealed a G to T transversion (NM_001135.2:c.1448G>T, p.Arg483Leu) in exon 8 in an adult male of Canadian origin with a family history of autosomal dominant inherited MED. Unfortunately, we have no further samples from family members or information about phenotype of the parents. The R483L residue shows a very high degree of con-servation throughout evolution, suggesting functional significance. The mutation was excluded on 212 control chromosomes including 120 of Canadian origin. The patient had an onset at the age of 6 years with the last follow-up at the age of 21 years. He has a proportional but slightly short stature (10-25% percentile). Radio-logical findings revealed epiphyseal dys-plasia of the hips, knees, spine, and feet. The hips are most severely affected, and he has gone through surgical treatments with an innominate osteotomia (JAKKULA et al. 2005).

The ACAN-gene encodes aggrecan and is the most abundant proteoglycan in cartilage extra cellular matrix. Aggrecan consists of a core protein, with three globular domains - G1, G2, and G3 - separated by a long glycosaminoglycan attachment region (HARDINGHAM et al.

40

1994). The G2-domain consists of two proteoglycan tandem repeats similar to those of the G1 domain; however, the G2 domain does not show any hyaluronan-binding function. Interactions with other ECM molecules have not been found (FOSANG and HARDINGHAM 1989). The only function assigned to the G2 domain is that it inhibits product secretion (KIANI et al. 2001). It may also have a role in the aggrecan specific GAG chain attachment. Thus the G2 domain has an uncertain function and spans the Arg>Leu sub-stitution associated with the case of autosomal dominant MED in our study. We hypothesize that the G2 domain of aggrecan may help maintain the extracellular matrix (ECM) network and that the Arg483Leu substitution may disrupt the matrix ensemble and contribute to skeletal deformation.

We also identified compound hetero-zygote mutations in a five-year-old boy of Canadian origin and without family history of MED. One mutation is a G to A transition in exon 6 (NM_001135.2:c.836G>A) re-sulting in p.Arg279Gln. The second mutation is a T to C transition in exon 7 (NM_001135.2: c.1366T>C) resulting in p.Phe456Leu. Both codons 279 and 456 are highly conserved, and the amino acid substitutions are non-tolerant according to the SIFT database (NG and HENIKOFF 2003). Analysis of the father and the mother by sequence analysis and restriction enzyme digestion revealed that the father is heterozygous for the G to A transition in exon 6 and the mother is a carrier of the T to C transition in exon 7. Both parents carring mutations in exon 6 and exon 7 had normal stature without a chondroskeletal phenotype.

The G to A transition was identified in one of 120 Canadian control chromosomes (1/212 control chromosomes) and the maternal T to C transition was identified in two of the 46 Swedish control chromosomes (2/212 control chromosomes). The segrega-tion pattern and the absence of family history support both mutations to con-tribute to impaired Aggrecan function and

MED. The affected boy was diagnosed with

MED at the age of three years and the last follow-up was at the age of five years. He has a normal height (50-75% percentile) and coxa vara. The radiological findings showed severe hip involvement with marked deformity or complete absence of the proximal femoral secondary ossification centres. The femoral necks were wide and short with irregular metaphyses. The distal femoral epiphyses were slightly flat, short, wide and in varus position, while the metaphyses of the distal femurs and the distal tibias showed mild flaring and irregularity (JAKKULA et al. 2005). Interestingly, our patients with ACAN mutations presented with some clinical similarities. Both had a history of early onset bilateral dysplasia of the proximal femurs with abnormal secondary ossifica-tion centres and abnormal development of the femoral necks while the other joints were less severely affected and the heights were within normal limits.

The most amino terminal globular domain G1 is composed of three functional sub-domains, two proteoglycan tandem repeats (PTR) and an immunoglobulin fold. The G1-domain forms non-covalent stable interactions with hyaluronan and link protein, a configuration that immobilizes the aggrecan molecule in the ECM (HARDINGHAM and MUIR 1972). Exons 3-6 encode the G1 domain of aggrecan and exon 6 encodes one of the proteoglycan tandem repeats (PTR). The PTR of the G1 domain are comprised of two cysteine-rich motifs that form disulfide bonds. These disulfide bonds are involved in aggrecan interaction with hyaluronan (HA). The interaction between aggrecan G1 domain and link protein (LP) is mediated via the immuno-globulin folds of the protein (GROVER and ROUGHLEY 1994). Cartilage link protein also has hyaluronan binding functions (HASCALL and HEINEGARD 1974). It is unlikely that the mutation results in a loss of interaction between aggrecan G1 and HA, since we identified healthy carriers. Even if the

41

mutant G1 affinity to HA is reduced, the binding between LP and HA could stabilize the complex although this might result in a weaker aggregate.

The inter-globular domain (IGD) between G1 and G2 is encoded by exon 7 of aggrecan. The IGD domain is the site of proteolytic cleavage to a variety of proteinases on aggrecan during pathological cartilage degradation (FOSANG et al. 1992; STRUGLICS et al. 2006). Cleavage of aggrecan molecule in the IGD region near its G1 domain results in rapid loss of the whole GAG attachment region. The Phe456Leu substitution is located some 60 amino acids from the classic aggrecanase (-TEGE/ARGS-) cleavage site and some 80 amino acids from the MMP site (-IPEN/FFGV-). The carrier parents of the mutations in exon 6 and exon 7 have a normal stature without a chondroskeletal phenotype. However, it can not be excluded that carriers of mutant alleles in these functionally important domains of the ACAN-gene might have increased susceptibility for early joint degeneration.

Mutations in the ACAN-gene have recently been reported in two human osteo-chondrodysplasias, the Spondyloepiphyseal dysplasia type Kimberley (SEDK) (GLEGHORN et al. 2005) and spondyloepimetaphyseal dysplasia (SEMD) (TOMPSON et al. 2009). In paper II, we describe familial osteo-chondritis dissecans (fOCD) caused by an ACAN-mutation. SEDK, MED, and fOCD are characterized by mild short stature and early onset OA, whereas SEMD is charac-terized by extreme shortness. There are similarities between fOCD and MED; for example, osteochondritis dissecans is described in MED families (LOHINIVA et al. 2000). The autosomal recessive inherited SEMD, described in a single family, may have a mild expression in carriers since the parents and one carrier half sibling have a relatively short stature (TOMPSON et al. 2009). Consistent with the findings in SED and SEMD, our findings indicate that ACAN mutations may be associated with both dominant and recessive forms of MED. Our

results supports an extensive genetic heterogeneity behind MED and further studies are required to discover the genes responsible for the disease. One weakness of the study is the missing samples and data of parents and relatives to the affected male with family history of MED. Another is not having direct proof of the missense mutations functionally importance.

To conclude study III, we present for the first time mutations in the ACAN-gene associated with MED. The mutations were compatible with both autosomal dominant and recessive inherited MED. Thus the ACAN-gene is the second reported gene behind autosomal recessive MED, pre-viously reported only for the sulphate transporter, DTDST. Further studies are required to identify if other skeletal dysplasias may be associated with ACAN-mutations as well as to clarify the patho-physiology of the mutations identified in our study.

10BPaper IV We describe rare clinical and radiological features of BDA1 and a novel mutation in the IHH-gene associated with BDA1.

The clinical and radiological features of afflicted family members indicated the diagnosis brachydactyly type A1 (BDA1). Anthropometric measurements of six affected individuals showed ordinary height and head circumferences. The arm span was short (<-2SD) in four out of six cases. The disorder exhibits considerable clinical variation both within and among families. Some BDA1 patients present with extensive skeletal manifestations with shortness of the long bones, malformed or absent epiphyses, scoliosis, abnormal menisci, club feet and short stature (RAFF et al. 1998; SLAVOTINEK and DONNAI 1998). The clinical and radiological features in this family were consistent in all affected family members.

Radiography of hands revealed that the six individuals had consistent abnormalities with bilateral and marked shortening of all middle phalanges. Affected members show hypoplasia of the ulnar styloid processes, a

42

finding that has been reported previously in BDA1 (NISSEN 1933). All subjects had a flattening of the metacarpal head of digits II-V. Clinodactyly were present in the middle phalanges of digit II (ulnar deviation) and digit V (radial deviation). The proximal phalanges of the thumb were short in four out of the six patients. The lengths of the distal phalanges were normal.

Each patient had rounded or ovoid ossicles at the metacarpal heads in at least one position of ray II-V, previously described as osteochondritis dissecans (ANDREN et al. 1978). Small ossicles - i.e., sesamoid bones - have been observed earlier in a few BDA1 cases close to the metacarpo-phalangeal joints (BOE and LUCHT 1979). In our family, these sesamoid bones remained unchanged after 30 years.

Mild OA of the carpo-metacarpal joint II, III, IV, and V was observed in all but one affected. OA is not previously described as associated with BDA1 and it can not be excluded that, in this family, this is part of normal ageing.

Radiographic analysis of the feet revealed bilateral absence of the middle phalanx of digit IV and V in all six patients. Five out of six affected individuals pre-sented with aplasia of middle phalanx III-IV and in two individuals the aplasia involved also digit II. All affected subjects but one had a distinct shorter proximal phalanx for digit I when compared to its distal phalanx. The metacarpo-phalangeal pattern profile was consistent in all affected family members showing a marked shortening of the middle phalanx and a normal distal phalanx.

We obtained evidence for linkage to the disease with chromosome 2q markers, showing a peak lod score at D2S301 (LOD score of 3.43 at theta 0.00). Sequence analysis of the Indian hedgehog gene (IHH) revealed a novel C to T transition (NM_002181.2:c.472C>T) in all affected family members resulting in an Arg to Cys amino acid substitution of the coding sequence (p.Arg158Cys).

The Arg158Cys substitution is located in

the highly conserved amino-terminal sig-nalling domain of the IHH protein. Six different heterozygous missense mutations and one deletion in the IHH-gene have previously been identified in 13 BDA1-families of various ethnic origins (GAO et al. 2001; GAO and HE 2004; GIORDANO et al. 2003; LIU et al. 2006; LODDER et al. 2008; MCCREADY et al. 2002; ZHU et al. 2007). These mutations, as well as the mutation reported here, are all located in the con-served amino-terminal signalling domain of the IHH protein.

The mutated cysteine residue is strictly conserved between species and is located at the edge of a groove at the known inter-action site of the IHH and its receptor Patched (PTC) (HALL et al. 1995). Molecular structure modelling suggests that the R158C substitution leads to a conformational change of the interaction site. The pro-truding appearance of Arg158 is markedly changed by the Arg158Cys substitution (Fig.5, paper IV). This further supports that the substitution of this highly conserved amino acid causes the disease and that the interaction between IHH and the receptor, Patched is affected by the Arg158Cys variant. We suggest that the mutations probably cause BDA1 through a haplo-insuffiency of the wild type Ihh, resulting in a decreased chondrocytes proliferation and attenuation of the negative feed-back loop that regulates chondrocytes maturation during enchondral ossification. One weak-ness of the study is the absence of functional studies.

To conclude study IV, we report a novel missense mutation in the IHH-gene associated with BDA1. The mutation is associated with an unusual combination of clinical features including hypoplasia of the ulnar styloid processes, ulna minus, sesamoid bones, OA, and normal length of all distal phalanges. Our findings add to the geno-type-phenotype correlations known for BDA1 and may facilitate the counselling and follow-up of families with this rare con-dition.

43

27BConcluding remarks and future perspectives

Gene mutations affecting proteins in cartilage development may lead to skeletal dysplasias, skeletal malformations, dys-function, and increased susceptibility to disease or injury. Characterization of these mutations and investigation of the molecular pathways for the corresponding gene products have contributed to our understanding of mechanisms regulating skeletal patterning, endochondral ossifica-tion, and bone and joint formation. Mutations in genes encoding structural components of articular cartilage give rise to rare forms of highly penetrant autosomal dominant inherited skeletal dysplasias associated with early-onset OA (LI et al. 2007). These inherited forms of OA in humans represent rare cases of the disease. Still studies of these inherited forms may provide insights into the pathogenesis of more common forms of OA.

The world’s population is ageing, particularly within the developing world. An increase in the prevalence of OA can be predicted and the world-wide social and economic burden from this disorder will become substantial. OA is a complex disease with a multifactorial background. The major etiological determinants are ageing, mechanical, hormonal and genetic factors. Twin-pair, sibling-risk, and segregation studies have identified multiple gene variants, common polymorphisms, associated with an increased risk of OA. Studies of constituents of the cartilage and joint fluid and their alterations in disease are likely to provide important new informa-tion with both diagnostic and therapeutic applications.

Better insight into molecular mechanisms by which chondrocytes control the functio-nal integrity of the ECM components of adult articular cartilage is needed. Deman

ding tasks for the future is the development of gene-based detection to be used early in disease of cartilage. This may facilitate effective prevention and treatment strate-gies such as biologically active molecules and or drugs blocking tissue destruction and promoting tissue repair and, possibly, the establishment of gene therapy. Better understanding of genetic and environ-mental interactions and the structure and function of cartilage extracellular matrix molecules will be crucial for effective therapeutic strategies in degenerative joint disease.

In this thesis, three different skeletal disorders has been studied, we studied the phenotype and the genotype in fOCD, MED and BDA1. We studied the C-type lectin domain (CLD) of G3 aggrecan and the consequences of the loss of interactions between aggrecan and its ligands. These investigations have led to further studies. We will evaluate the heritability of OCD using the Swedish twin registry. Studies have been initiated to further explore aggrecan G2 function, where residues in the G2 sequence is altered by amino acid replacement, using findings from paper III. In vivo biochemical experiments using mutant joint fluid, cartilage and bone marrow designed to explore aggrecan G3 function further is also being initiated. In addition, we have initiated studies for evaluating early signs, predictive markers for OCD and OA development in serum and plasma. We will evaluate identified SNPs as risk factors for OA development in a genome-wide association study using a cohort of patients who went through arthroplastic surgery due to knee- or hip OA. The outcome of this together with conclusions from animal and in vitro studies may provide important insights into the functions of ECM proteins and espe-cially aggrecan.

44

Populärvetenskaplig sammanfattning på svenska

(Summary in Swedish)

I avhandlingsarbetet beskrivs kliniska och genetiska studier av tre skelettsjukdomar. I arbete I och II har jag studerat familjär osteokondritis dissekans (fOCD), i arbete III multipel epifysär dysplasi (MED) och i arbete IV en nedärvd form av korta ben i fingrar och tår kallad Brakydaktyli typ A1 (BDA1).

11BIntroduktion Den extracellulära matrisen eller extra-cellulär matrix (ECM), består av en mängd olika proteiner som utgör stödjevävnad i kroppen, fyller ut området mellan celler och binder ihop celler och vävnader. I brosk består byggmaterialet fram för allt av proteinfamiljen kollagener och stora flaskbortsliknande proteoglykaner (aggrecan) som ofta är kopplade till sockerstrukturer. Kondrocyter är de celler som i brosket producerar och utsöndrar de olika komponenterna i ECM. Proteinerna i broskets ECM organiseras och binds till varandra i stora fiber- eller nätliknande strukturer. Broskets tryck- och stöt-upptagande förmåga kommer av att proteo-gykanen aggrecan binder vatten och bildar ett nätverk mellan fibrerna av kollagen som står för den mekaniska hållfastheten i brosket genom att i sin tur bilda ett fiber-innehållande nätverk. Alla vävnader är ständigt under uppbyggnad eller nedbrytning. De broskbildande cellerna kommunicerar med varandra och omgivningen genom att proteiner binder cellen och signalerar och reglerar vad cellen skall göra. Störs balansen påverkas eller förloras vävnadens egenskaper vilket leder till allvarliga sjukdomstillstånd. Därför är det viktigt att känna till funktionen hos och relationen mellan olika proteiner i brosk. Hur organi-sationen skapas, hur proteinerna samverkar och hur funktionerna upprätthålls.

Människans totala arvsmassa har kartlagts med sekvensundersökning, vilket tillåter oss att överblicka alla arvsanlag och andra betydelsefulla element som finns i kodad form i vår arvsmassa. De molekylära verktyg som utvecklats ger helt nya förutsättningar att utreda den majoritet av sjukdomar som ännu inte har fått någon tillfredställande förklaring. Verktygen och ny medicinsk kunskap kommer att möjliggöra nya former av diagnostik och terapi.

Jag har identifierat och studerat några olika genförändringar (mutationer) vilka ger förändrade proteiner i extracellulär matrix och leder till olika former av skelett dysplasier.

12BManuskript I & II På mottagningen för Klinisk genetik träffade jag en 15-årig pojke som kom på remiss för utredning av en misstänkt skellettdysplasi. Han var kortvuxen och hade osteokondritier i knä- och höftleder. Många i hans släkt hade samma sjukdoms-bild. Osteokondriter är broskbitar som lossnar från ledytor. Den sporadiska formen av osteokondritier förekommer hos 15-29/100,000, medan den ärftliga formen av osteochondritis dissecans (OCD) är mycket ovanligt och finns endast beskriven bland ca 50 fall i litteraturen. Vi undersökte 53 medlemmar från denna familj och efter-granskade alla röntgenbilder som tagits av dessa personer genom åren. Femton familjemedlemmar hade sjukdomen som karakteriserades av att broskbitar lossnar i lederna med början före puberteten, en disproportionerlig kortvuxenhet med korta ben och en tidig artros-utveckling i drabbade leder. Sjukdomen fanns i alla generationer och var autosomal dominant nedärvd.

45

För att ta reda på orsaken till sjukdomen gjordes en genetisk screening med kopplings-analys och jag identifierade genetiska markörer i ett område på den långa armen av kromosom 15 som återfanns hos de familjemedlemmar som var sjuka. Området omfattade många gener, bla. ACAN-genen som kodar för aggrecan, en viktig beståndsdel i brosk med betydelse för broskets förmåga att motstå tryck och belastning. Dessa faktorer i kombination med kunskap om att det tidigare finns beskrivet skelettdysplasier orsakade av mutationer i den genen gjorde att jag valde ACAN-genen som utgångspunkt för vidare undersökning. Sekvensering av ACAN-genen påvisade en punktmutation (c.6907G>A) hos alla drabbade i familjen, mutationen leder till ett aminosyre utbyte från valin till metionin (p.V2303M). Den identifierade mutationen i ACAN-genen är lokaliserad till en starkt konserverad region av proteinet som kallas G3-domänen. Det är en funk-tionell enhet som i sin tur är uppbyggd av mindre enheter varav C-typ lektin modulen är den del där mutationen är belägen. C-typ lektin modulen är viktig för nätverks-bildning och binder bl.a. till proteinerna fibulin-1, fibulin-2 samt tenacin-R. För att ta reda på om det identifierade aminosyre utbytet V2303M har någon effekt på G3 domänens utsöndring eller interaktion med andra proteiner så gjorde vi bindnings-studier med rekombinant muterat G3 protein (ett ”tillverkat” protein inne-hållande mutationen) och normalt G3 protein. Resultaten visade på en svag bindningsförmåga eller avsaknad av bindning för proteinet innehållande den felaktiga aminosyran. För att ta reda på om det muterade proteinet utsöndras använde vi oss av brosk från en familjemedlem med fOCD som genomgick en knäleds-plastik. Undersökningen visade att V2303M aggre-can produceras och finns i brosket. Sammanfattningsvis så har vi hittat en mutation som ger den molekylärgenetiska förklaringen till den bakomliggande orsaken till familjär osteochondritis dissecans och bekräftar i funktionella studier hur viktig

aggrecan c-typ lectin är för interaktion med andra proteiner och för broskets funktion.

13BManuskript III Multipel epifysär dysplasia (MED) är en av de vanligaste skelettdysplasierna och före-kommer hos ca 1:10, 000 personer. Sjukdomsbilden varierar mellan olika individer, vissa får diagnosen som barn och hos andra upptäcks sjukdomen först i vuxen ålder och då många gånger p.g.a. artros. Sjukdomen kan vara autosomal dominant eller recessivt nedärvd. Den autosomal dominanta formen karakteriseras av smärtor och stelhet i lederna, mild kortvuxenhet, underutvecklade tillväxt-plattor, och tidig artros. Den recessiva formen karakteriseras bland annat av mild kortvuxenhet, korta breda händer, klump-fötter och dubbel knäskål vid röntgen-undersökning. Sjukdomen kan orsakas av mutationer i flera olika gener som kodar för extracellulär matrix proteiner. Hos ca 50 % av patienterna med MED identifieras en mutation i någon av följande sex olika gener COMP, MATN3, COL9A1, COL9A2, COL9A3 och DTDST. För att försöka identifiera andra gener av betydelse för sjukdomen använde jag en kandidatgen approach. Jag valde ACAN-genen som kandidatgen för MED, eftersom jag funnit en mutation i genen som förklarade fOCD och att man tidigare beskrivit mutationer i genen som förklarar två andra skelett-dysplasier, spondyloepiphyseal dysplasia typ Kimberly (SEDK) och spondyloepimeta-physeal dysplasia (SEMD). Dessutom är aggrecan av fundamental betydelse för broskets funktion. Jag sekvenserade ACAN-genen med material från 39 patienter med sjukdomen som alla tidigare varit screening-negativa för mutationer i kända gener av betydelse för MED. Vid sekvensering av ACAN-genen identifierade jag en punktmutation (c.1448G>T) hos en vuxen man. Jag identifierade också två mutationer (c.1366T>C and c.836G>A) hos en fem årig pojke med friska föräldrar där båda föräldrarna var bärare av var sin

46

mutation. Sammanfattningsvis så är detta den första rapporten om ACAN-gen mutationer associerade med MED, våra fynd utökar spektrat av muterade gener som kan förklara sjukdomen.

14BManuskript IV Ortopeden Bjarne Lindén och medförfattare undersökte och beskrev en stor familj med korta fingrar och osteokondriter i finger-lederna 1978. Vi, jag och Bjarne undersökte sex drabbade individer i denna familj 30 år senare. Vi lät röntga händer och fötter på samtliga. Alla drabbade i familjen hade en liknade sjukdomsbild med korta mellan-falanger, kort processus styloideus av ulna, platta metacarpal huvuden och mild artros. Röntgenfynden talade för diagnosen brakydaktyli typ A1 (BDA1). BDA1 karak-teriseras av avsaknad av eller korta mellanfalanger i händer och fötter. Sjuk-domen förklaras av mutationer i Indian hedgehog genen (IHH) eller en ännu okänd

gen på kromosom 5p13. IHH-genen har betydelse för reglering av längdtillväxt i tillväxtplattorna. Kopplingsanalys genom-fördes för kända BDA1 lokus och påvisade koppling till IHH-genen. Jag sekvenserade genen och identifierade en mutation (c.472C>T) som leder till aminosyre utbytet p.158Arg>Cys. För att försöka förstå mutationens betydelse så använde jag en datamodell av proteinets struktur och jämförde normal aminosyra med den muterade. Datamodellen visar att R158C utbytet leder till en förändring av platsen där interaktion sker mellan IHH och dess receptor. Att det sker en förändring av bindning-stället samt att aminosyre utbytet är beläget i en starkt konserverad region talar för att mutationen kan vara sjukdomsorsakande. Sammanfattningsvis beskriver jag i detta arbete kliniska och röntgenologiska karakteristika hos en familj med BDA1 associerad med en icke tidigare beskriven mutation i IHH-genen.

47

Acknowledgements

This research was carried out at the Department of Clinical Genetics, Umeå University Hospital, Sweden. I wish to thank everyone who has been involved in and contributed to this work, in particular: Göran Roos, my supervisor and Head of the Department of Clinical Genetics, for providing good working conditions and for support and advice. Thank you for encouraging me to engage in doing research even though the clinical work load was large. Niklas Dahl, my supervisor at Department of Clinical Genetics, Uppsala for creativity and deep knowledge of medical genetics and for your constructive, valuable and supportive criticism. It has been a pleasure to work with you. Eva Holmberg for initial and constant support and Gösta Holmgren (in memory) for introducing me to the fascinating field of medical genetics. Yelverton Tegner and Bjarne Lindén for support and for many fun moments and good laughs during field work. Anders Aspberg, Dick Heinegård and Stefan Lohmander for your deep knowledge, fruitful discussions, and for your creativity and enthusiasm. I really enjoyed spending time in Lund. Susanne Haralsson and Inger Cullman staff at the Department of Medical Biosciences, Medical and Clinical Genetics and Pathology for your enthusiasm and invaluable technical assistance. Björn-Anders Jonsson, for technical assistance and for sharing your deep knowledge in molecular genetics. Thanks to my co-authors (those not already mentioned) for important contributions: Karin Lindblom, Patrik Önnerfjord, André Struglics, Magnus Domellöf, Torsten Lönnerholm, Jens Schuster, Shiro Ikegawa, Outi Mäkitie, William Cole, Fredrik Wiklund and Takako Sasaki.

48

Irina Golovleva, Jenni Jonasson and Kristina Cederquist for always being ready to discuss new ideas and possibilities for good collaboration and for friendship. Carin Nilsson and Lena Östman for skilful administrative support. I would also like to thank everybody at the Department of Clinical Genetics in Umeå. It is inspiring and fun to work with you. Thank you all for good collaboration and for friendship. My colleagues at the Department of Clinical Genetics, Uppsala, for support and stimulating discussions. Gisela Dahlquist, professor of Paediatrics former supervisor, for teaching me basic scientific skills. Maggie Wendelius, my clinical supervisor during my paediatric training for attention, lateral thinking and support from the very first time we met. Solveig Isberg for skilful editorial support, always “sunny” and cool. My family Märta-Lis and Ture for giving me the “hard-working and never-give-up” genes, for endless support and for bringing me up in a caring and loving atmosphere. Roger for your generosity and kindness. Kerstin and Nohr for love, support and many encouraging “sms”. Pär for your inspiration, never-ending support, love and friendship. Kalle and Oskar, for bringing so much happiness and love into my life. And last but not least, I am very grateful to all the family members participating in the studies; without your cooperation this thesis would never have been written. This project received financial support from County council of Västerbotten, Northern County Councils Cooperation Committee “Visare Norr”, the Capio Research Foundation and Umeå University.

49

References

2004 Finishing the euchromatic sequence of the

human genome. Nature 431: 931-945. ABKEVICH, V., N. J. CAMP, A. GUTIN, J. M.

FARNHAM, L. CANNON-ALBRIGHT et al., 2001 A robust multipoint linkage statistic (tlod) for mapping complex trait loci. Genet Epidemiol 21 Suppl 1: S492-497.

ALA-KOKKO, L., C. T. BALDWIN, R. W. MOSKOWITZ and D. J. PROCKOP, 1990 Single base mutation in the type II procollagen gene (COL2A1) as a cause of primary osteo-arthritis associated with a mild chondrodysplasia. Proc Natl Acad Sci U S A 87: 6565-6568.

ANDERSON, I. J., P. TSIPOURAS, C. SCHER, R. S. RAMESAR, R. W. MARTELL et al., 1990 Spondyloepiphyseal dysplasia, mild auto-somal dominant type is not due to primary defects of type II collagen. Am J Med Genet 37: 272-276.

ANDREN, L., N. CARSTAM and B. LINDEN, 1978 Osteochondritis dissecans and brachy-mesophalangia: a hereditary syndrome. J Hand Surg [Am] 3: 117-122.

ANDREW, T. A., J. SPIVEY and R. H. LINDEBAUM, 1981 Familial osteochondritis dissecans and dwarfism. Acta Orthop Scand 52: 519-523.

ARMOUR, C. M., M. E. MCCREADY, A. BAIG, A. G. HUNTER and D. E. BULMAN, 2002 A novel locus for brachydactyly type A1 on chromosome 5p13.3-p13.2. J Med Genet 39: 186-188.

ARNOLD, K., L. BORDOLI, J. KOPP and T. SCHWEDE, 2006 The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling. Bioinformatics 22: 195-201

ASPBERG, A., S. ADAM, G. KOSTKA, R. TIMPL and D. HEINEGARD, 1999 Fibulin-1 is a ligand for the C-type lectin domains of aggrecan and versican. J Biol Chem 274: 20444-20449.

ASPBERG, A., C. BINKERT and E. RUOSLAHTI, 1995 The versican C-type lectin domain recognizes the adhesion protein tenascin-R. Proc Natl Acad Sci U S A 92: 10590-10594.

ASPBERG, A., R. MIURA, S. BOURDOULOUS, M. SHIMONAKA, D. HEINEGARD et al., 1997 The C-type lectin domains of lecticans, a family of aggregating chondroitin sulfate proteo-glycans, bind tenascin-R by protein-protein interactions independent of carbohydrate moiety. Proc Natl Acad Sci U S A 94: 10116-10121.

AULD, C. D., and R. B. CHESNEY, 1979 Familial osteochondritis dissecans and carpal tunnel syndrome. Acta Orthop Scand 50: 727-730.

BALLHAUSEN, D., L. BONAFE, P. TERHAL, S. L. UNGER, G. BELLUS et al., 2003 Recessive multiple epiphyseal dysplasia (rMED): phenotype delineation in eighteen homo-zygotes for DTDST mutation R279W. J Med Genet 40: 65-71.

BALLOCK, R. T., and R. J. O'KEEFE, 2003 The biology of the growth plate. J Bone Joint Surg Am 85-A: 715-726.

BAYOUMI, R., K. SAAR, Y. A. LEE, G. NURNBERG, A. REIS et al., 2001 Localisation of a gene for an autosomal recessive syndrome of macro-cephaly, multiple epiphyseal dysplasia, and distinctive facies to chromosome 15q26. J Med Genet 38: 369-373.

BELL, J., 1951 On Brachydactyly and Symphalangism. University Press, Cambridge, England.

BENTLEY, D. R., S. BALASUBRAMANIAN, H. P. SWERDLOW, G. P. SMITH, J. MILTON et al., 2008 Accurate whole human genome sequencing using reversible terminator chemistry. Nature 456: 53-59.

BOE, S., and U. LUCHT, 1979 Brachy-mesophalangy and loose bodies in the metacarpophalangeal joints. Acta Orthop Scand 50: 755-758.

BRIGGS, M. D., and K. L. CHAPMAN, 2002 Pseudoachondroplasia and multiple epiphyseal dysplasia: mutation review, molecular interactions, and genotype to phenotype correlations. Hum Mutat 19: 465-478.

BRIGGS, M. D., S. M. HOFFMAN, L. M. KING, A. S. OLSEN, H. MOHRENWEISER et al., 1995 Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nat Genet 10: 330-336.

CAVANAGH, J. A., I. TAMMEN, P. A. WINDSOR, J. F. BATEMAN, R. SAVARIRAYAN et al., 2007 Bulldog dwarfism in Dexter cattle is caused by mutations in ACAN. Mamm Genome 18: 808-814.

CHEN, L., Y. WU, V. LEE, C. KIANI, M. E. ADAMS et al., 2002 The folded modules of aggrecan G3 domain exert two separable functions in glycosaminoglycan modification and product secretion. J Biol Chem 277: 2657-2665.

COHN, D. H., 2001 Defects in extracellular

50

matrix structural proteins in the osteo-chondrodysplasias. Novartis Found Symp 232: 195-210; discussion 210-192.

DOEGE, K. J., S. N. COULTER, L. M. MEEK, K. MASLEN and J. G. WOOD, 1997 A human-specific polymorphism in the coding region of the aggrecan gene. Variable number of tandem repeats produce a range of core protein sizes in the general population. J Biol Chem 272: 13974-13979.

DOEGE, K. J., M. SASAKI, T. KIMURA and Y. YAMADA, 1991 Complete coding sequence and deduced primary structure of the human cartilage large aggregating proteoglycan, aggrecan. Human-specific repeats, and additional alternatively spliced forms. J Biol Chem 266: 894-902.

DOMOWICZ, M. S., E. W. PIROK, 3RD, T. E. NOVAK and N. B. SCHWARTZ, 2000 Role of the C-terminal G3 domain in sorting and secretion of aggrecan core protein and ubiquitin-mediated degradation of accumulated mutant precursors. J Biol Chem 275: 35098-35105.

EHLEN, H. W., L. A. BUELENS and A. VORTKAMP, 2006 Hedgehog signaling in skeletal development. Birth Defects Res C Embryo Today 78: 267-279.

ERLEBACHER, A., E. H. FILVAROFF, S. E. GITELMAN and R. DERYNCK, 1995 Toward a molecular understanding of skeletal development. Cell 80: 371-378.

EYRE, D., 2002 Collagen of articular cartilage. Arthritis Res 4: 30-35.

EYRE, S., P. ROBY, K. WOLSTENCROFT, K. SPRECKLEY, R. ASPINWALL et al., 2002 Identification of a locus for a form of spondyloepiphyseal dysplasia on chromo-some 15q26.1: exclusion of aggrecan as a candidate gene. J Med Genet 39: 634-638.

FAIRBANK, H. A. T., 1945 Dysplasia epiphysealis multiplex Proc. Roy. Soc. Med. 39: 15-317.

FARABEE, W., 1903 Hereditary and Sexual Influences in Meristic Variation: A Study of Digital Malformation in Man. Harward University.

FITCH, N., 1979 Classification and identification of inherited brachydactylies. J Med Genet 16: 36-44.

FLANNERY, C., V. STANESCU, M. MORGELIN, R. BOYNTON, J. GORDY et al., 1992 Variability in the G3 domain content of bovine aggrecan from cartilage extracts and chondrocyte cultures. Arch Biochem Biophys 297: 52-60.

FONSECA, A. S., D. KERET and G. D. MACEWEN, 1990 Familial osteochondritis dissecans.

Orthopedics 13: 1259-1262. FOSANG, A. J., and T. E. HARDINGHAM, 1989

Isolation of the N-terminal globular protein domains from cartilage proteoglycans. Identification of G2 domain and its lack of interaction with hyaluronate and link protein. Biochem J 261: 801-809.

FOSANG, A. J., K. LAST, V. KNAUPER, G. MURPHY and P. J. NEAME, 1996 Degradation of cartilage aggrecan by collagenase-3 (MMP-13). FEBS Lett 380: 17-20.

FOSANG, A. J., P. J. NEAME, K. LAST, T. E. HARDINGHAM, G. MURPHY et al., 1992 The interglobular domain of cartilage aggrecan is cleaved by PUMP, gelatinases, and cathepsin B. J Biol Chem 267: 19470-19474.

FRAZER, K. A., D. G. BALLINGER, D. R. COX, D. A. HINDS, L. L. STUVE et al., 2007 A second generation human haplotype map of over 3.1 million SNPs. Nature 449: 851-861.

FREDRIKS, A. M., S. VAN BUUREN, R. J. BURGMEIJER, J. F. MEULMEESTER, R. J. BEUKER et al., 2000 Continuing positive secular growth change in The Netherlands 1955-1997. Pediatr Res 47: 316-323.

FULOP, C., E. WALCZ, M. VALYON and T. T. GLANT, 1993 Expression of alternatively spliced epidermal growth factor-like domains in aggrecans of different species. Evidence for a novel module. J Biol Chem 268: 17377-17383.

GAO, B., J. GUO, C. SHE, A. SHU, M. YANG et al., 2001 Mutations in IHH, encoding Indian hedgehog, cause brachydactyly type A-1. Nat Genet 28: 386-388.

GAO, B., and L. HE, 2004 Answering a century old riddle: brachydactyly type A1. Cell Res 14: 179-187.

GARDINER, T. B., 1955 Osteochondritis dissecans in three members of one family. J Bone Joint Surg Br 37-B: 139-141.

GERVER, W. J., and R. DE BRUIN, 1996 Body composition in children based on anthropometric data. A presentation of normal values. Eur J Pediatr 155: 870-876.

GIORDANO, N., L. GENNARI, M. BRUTTINI, F. MARI, I. MELONI et al., 2003 Mild brachydactyly type A1 maps to chromosome 2q35-q36 and is caused by a novel IHH mutation in a three generation family. J Med Genet 40: 132-135.

GLEGHORN, L., R. RAMESAR, P. BEIGHTON and G. WALLIS, 2005 A mutation in the variable repeat region of the aggrecan gene (AGC1) causes a form of spondyloepiphyseal dysplasia associated with severe, premature

51

osteoarthritis. Am J Hum Genet 77: 484-490.

GREEN, W. T., and H. H. BANKS, 1953 Osteochondritis dissecans in children. J Bone Joint Surg Am 35-A: 26-47; passim.

GROVER, J., and P. J. ROUGHLEY, 1994 The expression of functional link protein in a baculovirus system: analysis of mutants lacking the A, B and B' domains. Biochem J 300 ( Pt 2): 317-324.

GUILAK, F., L. G. ALEXOPOULOS, M. L. UPTON, I. YOUN, J. B. CHOI et al., 2006 The pericellular matrix as a transducer of biomechanical and biochemical signals in articular cartilage. Ann N Y Acad Sci 1068: 498-512.

HAGA, N., K. NAKAMURA, K. TAKIKAWA, N. MANABE, S. IKEGAWA et al., 1998 Stature and severity in multiple epiphyseal dysplasia. J Pediatr Orthop 18: 394-397.

HALL, T. M., J. A. PORTER, P. A. BEACHY and D. J. LEAHY, 1995 A potential catalytic site revealed by the 1.7-A crystal structure of the amino-terminal signalling domain of Sonic hedgehog. Nature 378: 212-216.

HANLEY, W. B., V. A. MCKUSICK and F. T. BARRANCO, 1967 Osteochondritis dissecans with associated malformations in two brothers. A review of familial aspects. J Bone Joint Surg Am 49: 925-937.

HARDINGHAM, T., and M. BAYLISS, 1990 Proteoglycans of articular cartilage: changes in aging and in joint disease. Semin Arthritis Rheum 20: 12-33.

HARDINGHAM, T. E., and A. J. FOSANG, 1992 Proteoglycans: many forms and many functions. Faseb J 6: 861-870.

HARDINGHAM, T. E., A. J. FOSANG and J. DUDHIA, 1990 Domain structure in aggregating proteoglycans from cartilage. Biochem Soc Trans 18: 794-796.

HARDINGHAM, T. E., A. J. FOSANG and J. DUDHIA, 1994 The structure, function and turnover of aggrecan, the large aggregating proteoglycan from cartilage. Eur J Clin Chem Clin Biochem 32: 249-257.

HARDINGHAM, T. E., and H. MUIR, 1972 The specific interaction of hyaluronic acid with cartillage proteoglycans. Biochim Biophys Acta 279: 401-405.

HASCALL, V. C., and D. HEINEGARD, 1974 Aggregation of cartilage proteoglycans. I. The role of hyaluronic acid. J Biol Chem 249: 4232-4241.

HASTBACKA, J., A. DE LA CHAPELLE, I. KAITILA, P. SISTONEN, A. WEAVER et al., 1992 Linkage disequilibrium mapping in isolated founder

populations: diastrophic dysplasia in Finland. Nat Genet 2: 204-211.

HASTBACKA, J., A. DE LA CHAPELLE, M. M. MAHTANI, G. CLINES, M. P. REEVE-DALY et al., 1994 The diastrophic dysplasia gene encodes a novel sulfate transporter: positional cloning by fine-structure linkage disequilibrium mapping. Cell 78: 1073-1087.

HASTBACKA, J., A. SUPERTI-FURGA, W. R. WILCOX, D. L. RIMOIN, D. H. COHN et al., 1996 Atelosteogenesis type II is caused by mutations in the diastrophic dysplasia sulfate-transporter gene (DTDST): evidence for a phenotypic series involving three chondrodysplasias. Am J Hum Genet 58: 255-262.

HEINEGARD, D., P. ANTONSSON, E. HEDBOM, T. LARSSON, A. OLDBERG et al., 1988 Non-collagenous matrix constituents of cartilage. Pathol Immunopathol Res 7: 27-31.

HEINEGARD, D., K. HULTENBY, A. OLDBERG, F. REINHOLT and M. WENDEL, 1989 Macro-molecules in bone matrix. Connect Tissue Res 21: 3-11; discussion 12-14.

HEINEGARD, D., and A. OLDBERG, 1989 Structure and biology of cartilage and bone matrix noncollagenous macromolecules. Faseb J 3: 2042-2051.

HELLEMANS, J., P. J. COUCKE, A. GIEDION, A. DE

PAEPE, P. KRAMER et al., 2003 Homozygous mutations in IHH cause acrocapitofemoral dysplasia, an autosomal recessive disorder with cone-shaped epiphyses in hands and hips. Am J Hum Genet 72: 1040-1046.

HORTON, W., 2002 Chondrodysplasias: disorders of cartilage matrix proteins in connective tissue and its heritable disorders. Wiley-Liss, New York, New York.

HORTON, W. E., JR., M. LETHBRIDGE-CEJKU, M. C. HOCHBERG, R. BALAKIR, P. PRECHT et al., 1998 An association between an aggrecan polymorphic allele and bilateral hand osteoarthritis in elderly white men: data from the Baltimore Longitudinal Study of Aging (BLSA). Osteoarthritis Cartilage 6: 245-251.

HUBER, C., S. ODENT, S. RUMEUR, P. PADOVANI, C. PENET et al., 2001 Sulphate transporter gene mutations in apparently isolated club foot. J Med Genet 38: 191-193.

IOZZO, R. V., and A. D. MURDOCH, 1996 Proteoglycans of the extracellular environ-ment: clues from the gene and protein side offer novel perspectives in molecular diversity and function. Faseb J 10: 598-614.

ISOGAI, Z., A. ASPBERG, D. R. KEENE, R. N. ONO,

52

D. P. REINHARDT et al., 2002 Versican interacts with fibrillin-1 and links extracellular microfibrils to other connective tissue networks. J Biol Chem 277: 4565-4572.

ITOH, T., S. SHIRAHAMA, E. NAKASHIMA, K. MAEDA, N. HAGA et al., 2006 Comprehensive screening of multiple epiphyseal dysplasia mutations in Japanese population. Am J Med Genet A 140: 1280-1284.

IVY, R. H., 1957 Congenital anomalies; as recorded on birth certificates in the Division of Vital Statistics of the Pennsylvania Department of Health, for the period 1951-1955, inclusive. Plast Reconstr Surg (1946) 20: 400-411.

JAKKULA, E., O. MAKITIE, M. CZARNY-RATAJCZAK, G. C. JACKSON, R. DAMIGNANI et al., 2005 Mutations in the known genes are not the major cause of MED; distinctive phenotypic entities among patients with no identified mutations. Eur J Hum Genet 13: 292-301.

KAMARAINEN, O. P., S. SOLOVIEVA, T. VEHMAS, K. LUOMA, P. LEINO-ARJAS et al., 2006 Aggrecan core protein of a certain length is protective against hand osteoarthritis. Osteoarthritis Cartilage 14: 1075-1080.

KARP, S. J., E. SCHIPANI, B. ST-JACQUES, J. HUNZELMAN, H. KRONENBERG et al., 2000 Indian hedgehog coordinates endochondral bone growth and morphogenesis via para-thyroid hormone related-protein-dependent and -independent pathways. Development 127: 543-548.

KAWAGUCHI, Y., R. OSADA, M. KANAMORI, H. ISHIHARA, K. OHMORI et al., 1999 Association between an aggrecan gene polymorphism and lumbar disc degeneration. Spine 24: 2456-2460.

KIANI, C., L. CHEN, Y. J. WU, A. J. YEE and B. B. YANG, 2002 Structure and function of aggrecan. Cell Res 12: 19-32.

KIANI, C., V. LEE, L. CAO, L. CHEN, Y. WU et al., 2001 Roles of aggrecan domains in biosynthesis, modification by glycos-aminoglycans and product secretion. Biochem J 354: 199-207.

KIDD, K. K., A. J. PAKSTIS, W. C. SPEED and J. R. KIDD, 2004 Understanding human DNA sequence variation. J Hered 95: 406-420.

KIMATA, K., H. J. BARRACH, K. S. BROWN and J. P. PENNYPACKER, 1981 Absence of proteoglycan core protein in cartilage from the cmd/cmd (cartilage matrix deficiency) mouse. J Biol Chem 256: 6961-6968.

KIRK, K. M., K. J. DOEGE, J. HECHT, N. BELLAMY and N. G. MARTIN, 2003 Osteoarthritis of the hands, hips and knees in an Australian twin sample--evidence of association with the aggrecan VNTR polymorphism. Twin Res 6: 62-66.

KIRKPATRICK, T. J., K. S. AU, J. M. MASTROBATTISTA, M. E. MCCREADY, D. E. BULMAN et al., 2003 Identification of a mutation in the Indian Hedgehog (IHH) gene causing brachydactyly type A1 and evidence for a third locus. J Med Genet 40: 42-44.

KOCHER, M. S., R. TUCKER, T. J. GANLEY and J. M. FLYNN, 2006 Management of osteo-chondritis dissecans of the knee: current concepts review. Am J Sports Med 34: 1181-1191.

KORENBERG, J. R., X. N. CHEN, K. DOEGE, J. GROVER and P. J. ROUGHLEY, 1993 Assign-ment of the human aggrecan gene (AGC1) to 15q26 using fluorescence in situ hybridization analysis. Genomics 16: 546-548.

KOZLOWSKI, K., and R. MIDDLETON, 1985 Familial osteochondritis dissecans: a dysplasia of articular cartilage? Skeletal Radiol 13: 207-210.

KRUEGER, R. C., JR., K. KURIMA and N. B. SCHWARTZ, 1999 Completion of the mouse aggrecan gene structure and identification of the defect in the cmd-Bc mouse as a near complete deletion of the murine aggrecan gene. Mamm Genome 10: 1119-1125.

KURIMA, K., M. L. WARMAN, S. KRISHNAN, M. DOMOWICZ, R. C. KRUEGER, JR. et al., 1998 A member of a family of sulfate-activating enzymes causes murine brachymorphism. Proc Natl Acad Sci U S A 95: 8681-8685.

LACHMAN, R. S., D. KRAKOW, D. H. COHN and D. L. RIMOIN, 2005 MED, COMP, multilayered and NEIN: an overview of multiple epiphyseal dysplasia. Pediatr Radiol 35: 116-123.

LANDAUER, W., 1965 Nanomelia, a Lethal Nutation of the Fowl. J Hered 56: 131-138.

LANDER, E., and L. KRUGLYAK, 1995 Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat Genet 11: 241-247.

LEE, B., H. VISSING, F. RAMIREZ, D. ROGERS and D. RIMOIN, 1989 Identification of the molecular defect in a family with spondyloepiphyseal dysplasia. Science 244: 978-980.

53

LEFEBVRE, V., and P. SMITS, 2005 Transcriptional control of chondrocyte fate and differentiation. Birth Defects Res C Embryo Today 75: 200-212.

LEVY, S., G. SUTTON, P. C. NG, L. FEUK, A. L. HALPERN et al., 2007 The diploid genome sequence of an individual human. PLoS Biol 5: e254.

LI, H., N. B. SCHWARTZ and B. M. VERTEL, 1993 cDNA cloning of chick cartilage chondroitin sulfate (aggrecan) core protein and identification of a stop codon in the aggrecan gene associated with the chondrodystrophy, nanomelia. J Biol Chem 268: 23504-23511.

LI, Y., L. XU and B. R. OLSEN, 2007 Lessons from genetic forms of osteoarthritis for the pathogenesis of the disease. Osteoarthritis Cartilage 15: 1101-1105.

LINDEN, B., 1976 The incidence of osteo-chondritis dissecans in the condyles of the femur. Acta Orthop Scand 47: 664-667.

LINDEN, B., 1977 Osteochondritis dissecans of the femoral condyles: a long-term follow-up study. J Bone Joint Surg Am 59: 769-776.

LIU, M., X. WANG, Z. CAI, Z. TANG, K. CAO et al., 2006 A novel heterozygous mutation in the Indian hedgehog gene (IHH) is associated with brachydactyly type A1 in a Chinese family. J Hum Genet 51: 727-731.

LODDER, E. M., A. J. HOOGEBOOM, J. H. COERT and E. DE GRAAFF, 2008 Deletion of 1 amino acid in Indian hedgehog leads to brachydactylyA1. Am J Med Genet A 146A: 2152-2154.

LOHINIVA, J., P. PAASSILTA, U. SEPPANEN, O. VIERIMAA, S. KIVIRIKKO et al., 2000 Splicing mutations in the COL3 domain of collagen IX cause multiple epiphyseal dysplasia. Am J Med Genet 90: 216-222.

LUNDELL, A., A. I. OLIN, M. MORGELIN, S. AL-KARADAGHI, A. ASPBERG et al., 2004 Structural basis for interactions between tenascins and lectican C-type lectin domains: evidence for a crosslinking role for tenascins. Structure 12: 1495-1506.

MARIGO, V., D. J. ROBERTS, S. M. LEE, O. TSUKUROV, T. LEVI et al., 1995 Cloning, expression, and chromosomal location of SHH and IHH: two human homologues of the Drosophila segment polarity gene hedgehog. Genomics 28: 44-51.

MAROUDAS, A., M. T. BAYLISS, N. UCHITEL-KAUSHANSKY, R. SCHNEIDERMAN and E. GILAV, 1998 Aggrecan turnover in human articular cartilage: use of aspartic acid racemization as a marker of molecular age.

Arch Biochem Biophys 350: 61-71. MCCARROLL, S. A., 2008 Extending genome-

wide association studies to copy-number variation. Hum Mol Genet 17: R135-142.

MCCREADY, M. E., A. GRIMSEY, T. STYER, S. M. NIKKEL and D. E. BULMAN, 2005 A century later Farabee has his mutation. Hum Genet 117: 285-287.

MCCREADY, M. E., E. SWEENEY, A. E. FRYER, D. DONNAI, A. BAIG et al., 2002 A novel mutation in the IHH gene causes brachy-dactyly type A1: a 95-year-old mystery resolved. Hum Genet 111: 368-375.

MUBARAK, S. J., and N. C. CARROLL, 1979 Familial osteochondritis dissecans of the knee. Clin Orthop Relat Res: 131-136.

NG, P. C., and S. HENIKOFF, 2003 SIFT: Predicting amino acid changes that affect protein function. Nucleic Acids Res 31: 3812-3814.

NISHIMURA, G., N. HAGA, H. KITOH, Y. TANAKA, T. SONODA et al., 2005 The phenotypic spectrum of COL2A1 mutations. Hum Mutat 26: 36-43.

NISSEN, K., 1933 A study in inherited brachydactyly. Annals of Eugenics 5: 281-301.

OFFIAH, A. C., and C. M. HALL, 2003 Radiological diagnosis of the constitutional disorders of bone. As easy as A, B, C? Pediatr Radiol 33: 153-161.

OLIN, A. I., M. MORGELIN, T. SASAKI, R. TIMPL, D. HEINEGARD et al., 2001 The proteoglycans aggrecan and Versican form networks with fibulin-2 through their lectin domain binding. J Biol Chem 276: 1253-1261.

OLSEN, B. R., A. M. REGINATO and W. WANG, 2000 Bone development. Annu Rev Cell Dev Biol 16: 191-220.

ORIOLI, I. M., E. E. CASTILLA and J. G. BARBOSA-NETO, 1986 The birth prevalence rates for the skeletal dysplasias. J Med Genet 23: 328-332.

OTT, J., 2001 Major strengths and weaknesses of the lod score method. Adv Genet 42: 125-132.

PAES, R. A., 1989 Familial osteochondritis dissecans. Clin Radiol 40: 501-504.

PAULSSON, M., M. MORGELIN, H. WIEDEMANN, M. BEARDMORE-GRAY, D. DUNHAM et al., 1987 Extended and globular protein domains in cartilage proteoglycans. Biochem J 245: 763-772.

PERKINS, S. J., A. S. NEALIS, D. G. DUNHAM, T. E. HARDINGHAM and I. H. MUIR, 1991 Molecular modeling of the multidomain

54

structures of the proteoglycan binding region and the link protein of cartilage by neutron and synchrotron X-ray scattering. Biochemistry 30: 10708-10716.

PHILLIPS, H. O., and S. A. GRUBB, 1985 Familial multiple osteochondritis dissecans. Report of a kindred. J Bone Joint Surg Am 67: 155-156.

PICK, M. P., 1955 Familial osteochondritis dissecans. J Bone Joint Surg Br 37-B: 142-145.

RAFF, M. L., K. A. LEPPIG, J. C. RUTLEDGE, E. WEINBERGER and R. A. PAGON, 1998 Brachydactyly type A1 with abnormal menisci and scoliosis in three generations. Clin Dysmorphol 7: 29-34.

RITTENHOUSE, E., L. C. DUNN, J. COOKINGHAM, C. CALO, M. SPIEGELMAN et al., 1978 Cartilage matrix deficiency (cmd): a new autosomal recessive lethal mutation in the mouse. J Embryol Exp Morphol 43: 71-84.

ROSS, M. R., LJ AND KAYE, GI 1995 Histology, A Text and Atlas 3rd ed,.

ROSSI, A., and A. SUPERTI-FURGA, 2001 Muta-tions in the diastrophic dysplasia sulfate transporter (DTDST) gene (SLC26A2): 22 novel mutations, mutation review, asso-ciated skeletal phenotypes, and diagnostic relevance. Hum Mutat 17: 159-171.

ROUGHLEY, P., D. MARTENS, J. RANTAKOKKO, M. ALINI, F. MWALE et al., 2006 The involvement of aggrecan polymorphism in degeneration of human intervertebral disc and articular cartilage. Eur Cell Mater 11: 1-7; discussion 7.

SAIKI, R. K., S. SCHARF, F. FALOONA, K. B. MULLIS, G. T. HORN et al., 1985 Enzymatic amplification of beta-globin genomic sequences and restriction site analysis for diagnosis of sickle cell anemia. Science 230: 1350-1354.

SANGER, F., S. NICKLEN and A. R. COULSON, 1977 DNA sequencing with chain-terminating inhibitors. Proc Natl Acad Sci U S A 74: 5463-5467.

SCHENCK, R. C., JR., and J. M. GOODNIGHT, 1996 Osteochondritis dissecans. J Bone Joint Surg Am 78: 439-456.

SCHWARTZ, N. B., and M. DOMOWICZ, 2002 Chondrodysplasias due to proteoglycan defects. Glycobiology 12: 57R-68R.

SHEFFIELD, E. G., 1998 Double-layered patella in multiple epiphyseal dysplasia: a valuable clue in the diagnosis. J Pediatr Orthop 18: 123-128.

SLAVOTINEK, A., and D. DONNAI, 1998 A boy with

severe manifestations of type A1 brachydactyly. Clin Dysmorphol 7: 21-27.

SMILLIE, I., 1960 Osteochondritis dissecans. Loose Bodies in Joints. Etiology, Pathology, Treatment. Livingstone Ltd;1960.

SNEAD, M. P., and J. R. YATES, 1999 Clinical and Molecular genetics of Stickler syndrome. J Med Genet 36: 353-359.

ST-JACQUES, B., M. HAMMERSCHMIDT and A. P. MCMAHON, 1999 Indian hedgehog signaling regulates proliferation and differentiation of chondrocytes and is essential for bone formation. Genes Dev 13: 2072-2086.

STATTIN, E. L., Y. TEGNER, M. DOMELLOF and N. DAHL, 2008 Familial osteochondritis dissecans associated with early osteoarthritis and disproportionate short stature. Osteo-arthritis Cartilage 16: 890-896.

STEFANSSON, H., D. RUJESCU, S. CICHON, O. P. PIETILAINEN, A. INGASON et al., 2008 Large recurrent microdeletions associated with schizophrenia. Nature 455: 232-236.

STOUGAARD, J., 1964 Familial Occurrence of Osteochondritis Dissecans. J Bone Joint Surg Br 46: 542-543.

STRUGLICS, A., S. LARSSON, M. A. PRATTA, S. KUMAR, M. W. LARK et al., 2006 Human osteoarthritis synovial fluid and joint cartilage contain both aggrecanase- and matrix metalloproteinase-generated aggre-can fragments. Osteoarthritis Cartilage 14: 101-113.

SUPERTI-FURGA, A., J. HASTBACKA, W. R. WILCOX, D. H. COHN, H. J. VAN DER HARTEN et al., 1996 Achondrogenesis type IB is caused by mutations in the diastrophic dysplasia sulphate transporter gene. Nat Genet 12: 100-102.

SUPERTI-FURGA, A., L. NEUMANN, T. RIEBEL, G. EICH, B. STEINMANN et al., 1999 Recessively inherited multiple epiphyseal dysplasia with normal stature, club foot, and double layered patella caused by a DTDST mutation. J Med Genet 36: 621-624.

SUPERTI-FURGA, A., and S. UNGER, 2007 Nosology and classification of genetic skeletal disorders: 2006 revision. Am J Med Genet A 143: 1-18.

TOBIN, W. J., 1957 Familial osteochondritis dissecans with associated tibia vara. J Bone Joint Surg Am 39-A: 1091-1105.

TOMPSON, S. W., B. MERRIMAN, V. A. FUNARI, M. FRESQUET, R. S. LACHMAN et al., 2009 A recessive skeletal dysplasia, SEMD aggrecan type, results from a missense mutation affecting the C-type lectin domain of

55

aggrecan. Am J Hum Genet 84: 72-79. TWYMAN, R. S., K. DESAI and P. M. AICHROTH,

1991 Osteochondritis dissecans of the knee. A long-term study. J Bone Joint Surg Br 73: 461-464.

UL HAQUE, M. F., L. M. KING, D. KRAKOW, R. M. CANTOR, M. E. RUSINIAK et al., 1998 Mutations in orthologous genes in human spondyloepimetaphyseal dysplasia and the brachymorphic mouse. Nat Genet 20: 157-162.

UNGER, S., 2002 A genetic approach to the diagnosis of skeletal dysplasia. Clin Orthop Relat Res: 32-38.

UNGER, S., L. BONAFE and A. SUPERTI-FURGA, 2008 Multiple epiphyseal dysplasia: clinical and radiographic features, differential diagnosis and molecular basis. Best Pract Res Clin Rheumatol 22: 19-32.

UNGER, S. L., M. D. BRIGGS, P. HOLDEN, B. ZABEL, L. ALA-KOKKO et al., 2001 Multiple epiphyseal dysplasia: radiographic abnormalities correlated with genotype. Pediatr Radiol 31: 10-18.

VALHMU, W. B., G. D. PALMER, P. A. RIVERS, S. EBARA, J. F. CHENG et al., 1995 Structure of the human aggrecan gene: exon-intron organization and association with the protein domains. Biochem J 309 ( Pt 2): 535-542.

WANG, J., W. WANG, R. LI, Y. LI, G. TIAN et al., 2008 The diploid genome sequence of an Asian individual. Nature 456: 60-65.

WATANABE, H., K. KIMATA, S. LINE, D. STRONG, L. Y. GAO et al., 1994 Mouse cartilage matrix deficiency (cmd) caused by a 7 bp deletion in the aggrecan gene. Nat Genet 7: 154-157.

WATANABE, H., K. NAKATA, K. KIMATA, I. NAKANISHI and Y. YAMADA, 1997 Dwarfism and age-associated spinal degeneration of heterozygote cmd mice defective in aggrecan. Proc Natl Acad Sci U S A 94: 6943-6947.

WATSON, J. D., and F. H. CRICK, 1953 Molecular structure of nucleic acids; a structure for deoxyribose nucleic acid. Nature 171: 737-738.

WEEDON, M. N., H. LANGO, C. M. LINDGREN, C. WALLACE, D. M. EVANS et al., 2008 Genome-wide association analysis identifies 20 loci that influence adult height. Nat Genet 40: 575-583.

VENTER, J. C., M. D. ADAMS, E. W. MYERS, P. W. LI, R. J. MURAL et al., 2001 The sequence of the human genome. Science 291: 1304-1351.

VERTEL, B. M., L. M. WALTERS, B. GRIER, N. MAINE and P. F. GOETINCK, 1993 Nanomelic chondrocytes synthesize, but fail to translocate, a truncated aggrecan precursor. J Cell Sci 104 ( Pt 3): 939-948.

WHEELER, D. A., M. SRINIVASAN, M. EGHOLM, Y. SHEN, L. CHEN et al., 2008 The complete genome of an individual by massively parallel DNA sequencing. Nature 452: 872-876.

VIGNON, E., M. ARLOT, L. M. PATRICOT and G. VIGNON, 1976 The cell density of human femoral head cartilage. Clin Orthop Relat Res: 303-308.

WIKLAND, K. A., Z. C. LUO, A. NIKLASSON and J. KARLBERG, 2002 Swedish population-based longitudinal reference values from birth to 18 years of age for height, weight and head circumference. Acta Paediatr 91: 739-754.

VORTKAMP, A., K. LEE, B. LANSKE, G. V. SEGRE, H. M. KRONENBERG et al., 1996 Regulation of rate of cartilage differentiation by Indian hedgehog and PTH-related protein. Science 273: 613-622.

XIAO, W., and P. J. OEFNER, 2001 Denaturing high-performance liquid chromatography: A review. Hum Mutat 17: 439-474.

YANG, B. L., L. CAO, C. KIANI, V. LEE, Y. ZHANG et al., 2000 Tandem repeats are involved in G1 domain inhibition of versican expression and secretion and the G3 domain enhances glycosaminoglycan modification and pro-duct secretion via the complement-binding protein-like motif. J Biol Chem 275: 21255-21261.

YTREHUS, B., C. S. CARLSON and S. EKMAN, 2007 Etiology and pathogenesis of osteo-chondrosis. Vet Pathol 44: 429-448.

ZHU, G., X. KE, Q. LIU, J. LI, B. CHEN et al., 2007 Recurrence of the D100N mutation in a Chinese family with brachydactyly type A1: evidence for a mutational hot spot in the Indian hedgehog gene. Am J Med Genet A 143: 1246-1248.

Date post:	26-Apr-2020
Category:	Documents
Upload:	others
View:	8 times
Download:	0 times

Clinical and Genetic Studies of Three Inherited Skeletal ...216272/FULLTEXT01.pdf · radiological...

Documents