pmc logo imageJournal ListSearchpmc logo image
Logo of ajhgJournal URL: redirect3.cgi?&&auth=0FE8CQZK0oVxyRvxGPTbm9mx4yz2QGB384fQEF5Hm&reftype=publisher&artid=1274373&article-id=1274373&iid=123697&issue-id=123697&jid=203&journal-id=203&FROM=Article|Banner&TO=Publisher|Other|N%2FA&rendering-type=normal&&http://www.journals.uchicago.edu/AJHG/index.html
Am J Hum Genet. 2001 November; 69(5): 969–980.
Published online 2001 September 14.
PMCID: PMC1274373
A Mutation in COL9A1 Causes Multiple Epiphyseal Dysplasia: Further Evidence for Locus Heterogeneity
Malwina Czarny-Ratajczak,1,3,* Jaana Lohiniva,1,* Piotr Rogala,4 Kazimierz Kozlowski,5 Merja Perälä,1 Liisa Carter,6 Tim D. Spector,6 Lukasz Kolodziej,7 Ulpu Seppänen,2 Renata Glazar,3 Jan Królewski,7 Anna Latos-Bielenska,3 and Leena Ala-Kokko1,8
1Collagen Research Unit, Biocenter and Department of Medical Biochemistry, and 2Department of Radiology, University of Oulu, Oulu, Finland; Departments of 3Medical Genetics and 4Orthopaedics, Karol Marcinkowski University of Medical Sciences, Poznan, Poland; 5Department of Radiology, Royal Alexandra Hospital for Children, Sydney; 6Twin Research & Genetic Epidemiology Unit, St. Thomas' Hospital, London; 7Orthopaedic Clinic for Children, Pomeranian Medical University, Szczecin, Poland; and 8Center for Gene Therapy and Department of Medicine, Tulane University Health Sciences Center, New Orleans
Address for correspondence and reprints: Dr. Leena Ala-Kokko, Center for Gene Therapy, Tulane University Health Sciences Center, 1430 Tulane Avenue, SL-99, New Orleans, LA 70112. E-mail: lalako/at/tulane.edu
*The first two authors made equal contributions.
Received July 10, 2001; Accepted August 24, 2001.
Abstract
Multiple epiphyseal dysplasia (MED) is an autosomal dominantly inherited chondrodysplasia. It is clinically highly heterogeneous, partially because of its complex genetic background. Mutations in four genes, COL9A2, COL9A3, COMP, and MATR3, all coding for cartilage extracellular matrix components (i.e., the α2 and α3 chains of collagen IX, cartilage oligomeric matrix protein, and matrilin-3), have been identified in this disease so far, but no mutations have yet been reported in the third collagen IX gene, COL9A1, which codes for the α1(IX) chain. MED with apparently recessive inheritance has been reported in some families. A homozygous R279W mutation was recently found in the diastrophic dysplasia sulfate transporter gene, DTDST, in a patient with MED who had a club foot and double-layered patella. The series consisted of 41 probands with MED, 16 of whom were familial and on 4 of whom linkage analyses were performed. Recombination was observed between COL9A1, COL9A2, COL9A3, and COMP and the MED phenotype in two of the families, and between COL9A2, COL9A3, and COMP and the phenotype in the other two families. Screening of COL9A1 for mutations in the two probands from the families in which this gene was not involved in the recombinations failed to identify any disease-causing mutations. The remaining 37 probands were screened for mutations in all three collagen IX genes and in the COMP gene. The probands with talipes deformities or multipartite patella were also screened for the R279W mutation in DTDST. The analysis resulted in identification of three mutations in COMP and one in COL9A1, but none in the other two collagen IX genes. Two of the probands with a multipartite patella had the homozygous DTDST mutation. The results show that mutations in COL9A1 can cause MED, but they also suggest that mutations in COL9A1, COL9A2, COL9A3, COMP, and DTDST are not the major causes of MED and that there exists at least one additional locus.
Introduction

Multiple epiphyseal dysplasia (MED [MIM 132400]) is a disorder of the skeletal system that is manifested as a disturbance in the development of the epiphyses (Fairbank 1947), the major component of which is cartilage. The changes are usually observed in the majority of the growth centers, including those of the spine (Hoefnagel 1967). The joints are usually bilaterally—but not always symmetrically—affected in MED.

The first symptoms of MED occur in childhood, usually at age 2–14 years, and include waddling gait, restriction of joint mobility, and pain and stiffness in the weight-bearing joints. MED is clinically heterogeneous, consisting of the Fairbank, Ribbing, and unclassified types (International Working Group on Constitutional Diseases of Bone 1998). The Fairbank type is more severe than the Ribbing type and is characterized by shortness of stature; short, stubby fingers; and small epiphyses in several joints, including the knee, ankle, hand, and hip (Fairbank 1947; Silverman 1996). The Ribbing type (Ribbing 1937; Silverman 1996) is confined predominantly to the hip joints and is characterized by hands that are normal and stature that is normal or near-normal. The unclassified types represent combinations of the Fairbank and Ribbing phenotypes.

Radiological examination of the skeleton shows delayed, irregular mineralization of the epiphyseal ossification centers and of the centers of the carpal and tarsal bones (Silverman 1996). Early-onset osteoarthritis (OA) is also a very common finding. Some patients with MED present with mild spinal and patellar abnormalities. The spinal changes can include irregular end plates of the vertebral bodies, Schmorl’s nodes, wedging of the vertebral bodies, and narrowed disc spaces. Some patients may have ovoid vertebral bodies in the thoraco-lumbar spine in the first few years of life (Hulvey and Keats 1969; Silverman 1996). Radiographs of the knees show multipartite patellae in some MED cases (Sheffield 1998). The epiphyses are flat in the Ribbing type (Ribbing 1937) and are small in the Fairbank type (Silverman 1996).

MED is an autosomal dominantly inherited osteochondrodysplasia for which four loci have been identified so far: EDM1 (MIM 132400), EDM2 (MIM 600204), EDM3 (MIM 600969), and EDM5 (Chapman et al. 2001). EDM1 is located on chromosome 19p13.1 (Newton et al. 1994) and contains a gene for cartilage oligomeric matrix protein (COMP [MIM 600310]) (Oehlmann et al. 1994). Eleven mutations in COMP have been identified in the Ribbing, Fairbanks, and unclassified forms of the disease (Briggs et al. 1995, 1998; Ballo et al. 1997; Susic et al. 1997; Ikegawa et al. 1998; Loughlin et al. 1998; Deere et al. 1999; Délot et al. 1999), and ~40 mutations have been identified in pseudoachondroplasia (PSACH [MIM 177170]), an allelic disorder (Deere et al. 1998, 1999; Ikegawa et al. 1998; Délot et al. 1999; Newman et al. 2000). The other two MED loci contain genes for collagen IX—namely, COL9A2 (MIM 120260) on chromosome 1p32.3-p33 (Warman et al. 1994) and COL9A3 (MIM 120270) on chromosome 20q13.3 (Brewton et al. 1995). Four mutations have been identified in the COL9A2 gene (Muragaki et al. 1996; Holden et al. 1999; Spayde et al. 2000), and three have been identified in the COL9A3 gene (Paassilta et al. 1999a; Bonneman et al. 2000; Lohiniva et al. 2000). Recently, Chapman and others (2001) identified a new MED locus, EDM5. It is located on chromosome 2p24-p23 and contains a gene for matrilin-3 (MATN3 [MIM 602109]). Two different missense mutations were reported in the exon encoding the von Willebrand factor A domain of matrilin-3 in two unrelated families with MED (Chapman et al. 2001).

COMP is an extracellular matrix glycoprotein of 524 kD that belongs to the thrombospondin protein family (Hedbom et al. 1992; Mörgelin et al. 1992; Oldberg et al. 1992). It is a bouquet-shaped pentameric molecule composed of five identical arms (Mörgelin et al. 1992; Malashkevich et al. 1996), each containing a coiled-coil N-terminal domain responsible for pentamerization, four epidermal growth factor–like repeats, eight thrombospondin type 3 (T3) repeats, and a large globular C-terminal domain (Mörgelin et al. 1992; Hummel et al. 1998). Most of the mutations characterized in COMP in patients with MED or PSACH have been amino acid substitutions clustered in the T3 repeats (Briggs et al. 1995). It has been shown that such mutations may disturb calcium binding (Misenheimer and Mosher 1995; Chen et al. 2000; Maddox et al. 2000; Thur et al. 2001).

Collagen IX is a heterotrimer of α1(IX), α2(IX), and α3(IX) chains encoded by the COL9A1 (MIM 120210), COL9A2, and COL9A3 genes (Pihlajamaa et al. 1998; Paassilta et al. 1999b). It is a fibril-associated collagen with interrupted triple helices, consisting of three collagenous (COL1–COL3) and four non-collagenous (NC1–NC4) domains. It is a quantitatively minor cartilage component that is covalently cross-linked to collagen II fibrils via the COL2 domain (Diab et al. 1996). The NC3 domain functions as a hinge, allowing the other two more N-terminal domains, COL3 and NC4, to project away from the fibril surface. It has been suggested that these domains may mediate interactions between collagen IX and other matrix molecules. The three genes are almost identical in their genomic organization, with one exception—the COL9A1 gene contains 38 exons, whereas the others have 32.

Some families with MED have been reported to have autosomal recessive MED (EDM4 [MIM 226900]). A homozygous R279W mutation was found in the diastrophic dysplasia sulfate transporter gene (DTDST [MIM 222600]) in one individual with autosomal recessive MED characterized by club foot and double layered patella (Superti-Furga et al. 1999). The same homozygous mutation was recently found in two unrelated sibships with apparently isolated club foot (Huber et al. 2001).

Here we studied 41 probands with MED, of whom 16 were familial. Recombinations were observed between COMP, COL9A1, COL9A2, and COL9A3 and the phenotype in two families and between three of the genes and the phenotype in another two families. Since mutations could be identified in only 6 of 39 probands, we suggest that at least one additional locus must exist for MED.

Subjects and Methods

Subjects
Forty unrelated probands with MED (probands 1–40, table 1) were examined in the Departments of Orthopaedics and Medical Genetics at the University of Medical Sciences, Poznan, Poland, and in the Department of Orthopaedics at the University of Medical Sciences, Szczecin, Poland, and one proband (proband 41, table 1) was examined at the Department of Rheumatology, St. Thomas’ Hospital, London (proband 41, table 1). A detailed clinical and radiological examination was performed on all the probands and on most of the affected family members. Sixteen of the probands had a positive family history of disease, and 25 had no family history. Blood samples were obtained from the probands and the family members, for genomic DNA isolation. Signed informed consent was obtained from all subjects.
Table 1Table 1
Clinical and Radiological Findings in the Probands[Note]

Linkage Analysis
Intragenic markers or microsatellite markers were used for linkage analysis of COL9A1, COL9A2, COL9A3, and COMP. Three sequence variations were tested in COL9A1: T→C in IVS2−12, C→T in IVS19−88, and A→G in IVS20+61. The IVS2 variation was amplified by PCR with the primers E3F (5′-GTG GTC AAT TGC TAT TTT CTG GTT C) and E3R (5′-GCT TTA TCT ACC TGG AAC TGA G), and the IVS19 and IVS20 variations were amplified with the primers E20F (5′-CCA TCA GAA GAA TTC TCC TTG GAC) and E20R (5′-GAT AAA AAG TTA TGT TTA AAT GGC). Primers A19F (5′-TGG ATC TCA GTT TCC CTA CCT G) and A19R (5′-CAA GAG GTG GTG ATT GAG CAA GAG C) were used for PCR amplification of a region of the COL9A2 gene that contained three single-nucleotide variations: A→G in exon 19+23, C→G in exon 19+49, and C→G in IVS18−4. In addition, microsatellite markers, DIS211 and L-MYC, were used to analyze linkage to chromosome 1 (Mäkelä et al. 1992; Weissenbach et al. 1992). The primer pair FI19C (5′-CAG AAT GGC GTG CCA GGA CTC G) and RI19C (5′-CCA ACA TGG GCC ACT GAG C) was used to amplify four single-nucleotide variations in the COL9A3 gene by PCR; G→C in IVS18−90, C→T in IVS19+10, G→C in IVS19+12, and G→A in IVS19+58. The primer pair JC18F (5′-GGG CTG GCC ACT GAA GCT CTG AGA) and JC18R (GCC GCG GTG AGG GTG GCT GCT AT) was used to amplify a single-nucleotide variation in the COMP gene, T→C in IVS18+53. A tetranucleotide repeat in IVS9 of the COMP gene had been reported elsewhere (Briggs et al. 1995).

Mutation Screening and Sequencing
Exons and the boundaries of the COMP (Newton et al. 1994; GenBank accession number AC003107), COL9A1, COL9A2 (Pihlajamaa et al. 1998), and COL9A3 (Paassilta et al. 1999b) genes were amplified by PCR using 40 ng of genomic DNA, 0.25 μM of forward and reverse primers, 1.5 μM MgCl2, 0.2 mM dNTPs, and 1 U of AmpliTaq Gold polymerase (Perkin Elmer). The PCR conditions included an initial denaturation for 10 min at 95°C, followed by 35 cycles at 95°C for 30 s, at 54°C–63°C for 30 s, and at 72°C for 30 s, followed by 1 cycle at 72°C for 10 min. This was followed by denaturation at 98°C for 3 min and reannealing at 68°C for 30 min, to generate heteroduplexes. All PCR products were checked for quality and quantity on 2% agarose gels.

Conformation-sensitive gel electrophoresis (CSGE) was performed as described elsewhere (Körkkö et al. 1998), and the PCR products that contained heteroduplexes were sequenced using the dRhodamine Terminator Cycle Sequencing Ready Reaction Kit and an ABI Prism 377 Sequencer (Perkin Elmer), after prior treatment with exonuclease and shrimp alkaline phosphatase (Werle et al. 1994).

A region of the DTDST gene containing the R279W mutation was amplified by PCR with a primer pair that has been described elsewhere (Huber et al. 2001). The presence of the mutation was studied by digestion with StyI restriction endonuclease.

RNA Analysis
Total RNA was extracted from Epstein-Barr virus (EBV)–transformed lymphoblasts from proband 41. First-strand cDNA synthesis (Superscript Preamplification System [Gibco BRL]) was followed by PCR amplification with primers corresponding to exon 5 (D1F: 5′-GCA GCC TTT TCG AAT TTG TCC TCC TTG) and exon 19 (D1R: 5′-CTC CGA GTT CTC CCT GGT CAC CTT CTT CAC) of the COL9A1 gene. A second amplification was performed using nested primers from exon 6 (JV-9B: 5′-GAA ACT TGC CAT GAG CTG CCA) and exon 11 (JV-8R: 5′-ATC CAT CAG GTC CTG TTA AT). The products were analyzed on agarose gels and by sequencing.

Results

Diagnosis
Initially, 59 probands had been selected for examination, on the basis of their clinical histories and clinical findings, which were typical of MED. Radiographs were taken of the ankles, elbows, hands, hips, knees, shoulders, and spine and were evaluated independently by two expert radiologists. After evaluation of the clinical and radiological findings, a diagnosis of MED was confirmed in 41 cases; the remaining eighteen probands were excluded because they were found to have bilateral Legg-Calve-Perthes disease (MIM 150600), spondyloepiphyseal dysplasia tarda (MIM 184100), spondyloepimetaphyseal dysplasia (MIM 300106) or mild to severe forms of PSACH. The detailed clinical and radiological findings regarding the probands are presented in table 1, and radiographs of the knees and spine of proband 14 and of the hips of the proband’s affected brother are shown in figure 1A–E.
Figure  1Figure 1
Radiographs of the knees (A, anteroposterior (AP) view; B, lateral view) and thoraco-lumbar spine (C, AP view; D, lateral view) of proband 14 at age 14 years (table 1), and of the hips (E) of the proband’s affected brother at age 11 years (individual (more ...)

Linkage Analysis
Of the cases of disease in the 41 probands, 25 were sporadic and 16 were familial. Linkage to all four candidate genes, COL9A1, COL9A2, COL9A3 and COMP, was tested in the four largest families available (probands 12, 14, 20, and 32; table 1), yielding recombinations with the MED phenotype in two of the families, and recombinations between COL9A2, COL9A3 andCOMP and the phenotype in the other two (fig. 2).
Figure  2Figure 2
Pedigrees of four families with MED. Blackened symbols denote affected individuals. The probands are indicated by arrows ([NW pointing arrow]). The family pedigree of proband 12 is shown in panel A, that of proband 14 in panel B, that of proband 20 in panel C, (more ...)

Mutations and Polymorphisms in the COMP Gene
All probands except for numbers 12, 14, 20, and 32 (table 1 and fig. 2) were analyzed for mutations in COMP. All 19 exons and exon boundaries of the gene were amplified by PCR and were tested for sequence variations by CSGE. Products that contained heteroduplexes in CSGE analyses were sequenced, leading to the identification of three mutations (fig. 3). The rest of the observed sequence variations were likely to be neutral, because they did not change the amino acid encoded and were also found in controls (not shown).
Figure  3Figure 3
CSGE analysis of the PCR products for exons 8, 12, and 16 of the COMP gene, and for exon 8 of the COL9A1 gene. “E” denotes exon, “P” denotes proband, and “C” denotes control.

A novel A→C transversion was found in exon 12, which changed a GAT codon for Asp420 to a GCT codon for Ala in the sixth T3 repeat (fig. 3). This was found in proband 2 (table 1) and in her two affected daughters, who were aged 16 and 20 years, but not in any unaffected family members or in the 100 controls tested. The younger daughter had had knee pains since the age 7 years. Clinical examination revealed crepitation of the knees, laxity of the knee joints, and mild scoliosis, while radiological examination showed subchondral sclerosis, flattening of the femoral heads, valgus deformity, short femoral necks, and irregular and flat epiphyses of the knees (not shown). Arthroscopy revealed a bipartite patella and chondromalacia patellae. Her older sister had had hip pains since age 14 years and currently had knee and elbow pains as well. A clinical examination revealed valgus deformity of the forearms, hyperextension of the elbows, joint laxity, crepitation of the knees, mild scoliosis, and mild shortening of the toes. Radiographs showed valgus deformity of the femoral neck, mild flattening of the femoral heads, irregular and flat knee epiphyses, sclerosis of the acetabulum, and endplate irregularities of the lumbar vertebral bodies (not shown).

A second mutation, identified in proband 8 (table 1), consisted of a single-base change in exon 8 (fig. 3) that converted a CCG codon for Pro276 to a CGG codon for Arg in the first T3 repeat. Four other family members were also affected: the proband’s mother and maternal grandmother, an aunt, and an uncle. All of them had the mutation, which was not found in any of the unaffected family members. The mutation was not found in any of the 100 controls tested. The proband’s mother is 46 years old and 168 cm tall. She has had difficulties in walking since age 15 years and had hip pain since age 23 years. At the time of the present study, she also had pain in the knees, shoulders, and wrists. Clinical examination revealed crepitation of the shoulders, valgus deformity of the knees, and a limited range of motion of the hips. Radiological examination showed severe hip and knee OA, flattening of the femoral heads and knee epiphyses, shortening of the femoral neck, and endplate irregularities in the lumbar spine (not shown). The uncle is 44 years old and 172 cm tall. He has had a waddling gait and hip pain since age 3 years and has a limited range of movement in his hips and knees. Radiological examination revealed flattening of the knee epiphyses with moderately severe OA, flat femoral heads with multiple subchondral cysts, varus deformity of the hips, and short femoral necks (not shown). A diagnosis of MED was confirmed in the other two affected individuals by clinical and radiological examination.

The third mutation, affecting exon 16 in proband 10 (table 1, fig. 3), changed an ACG codon for Thr585 to an ATG codon for Met in the C-terminal globular domain. None of the proband’s family members were affected, nor did they carry the mutation, suggesting a de novo mutation. The proband had had knee pain since age 6 years and hip pain since age 12 years. The clinical and radiological findings are presented in table 1. There was no radiological evidence of metaphyseal or vertebral-body abnormalities.

Mutation Analysis of the COL9A1, COL9A2, and COL9A3 Genes
All the probands in whom no mutations were found or no recombinations were observed between the collagen IX genes and the phenotype were analyzed for sequence variations in the genes. Exons and the boundaries of the COL9A1, COL9A2, and COL9A3 genes were amplified by PCR and were analyzed for heteroduplexes by CSGE. In addition, exons 2, 3, and 4 of the COL9A2 and COL9A3 genes and exons 8, 9, and 10 of the COL9A1 gene were analyzed by sequencing.

Analysis of exon 8 of the COL9A1 gene in proband 41 identified a unique heteroduplex (table 1, fig. 3). Sequencing of the product showed insertion of a T at the donor splice site of IVS8+3 (data not shown). The presence of the insertion was confirmed by digestion with MseI restriction endonuclease (data not shown). The proband’s affected mother also had the insertion, whereas his unaffected sister did not. The insertion was not found in any of 600 control chromosomes tested. RNA isolated from the lymphoblasts of the affected son and from a control sample was reverse transcribed and amplified by PCR, using primers specific to exons 6 and 11. The products were first analyzed on agarose gels, where the control sample always had only one band, but the sample from the affected individual constantly showed two major bands and one or more minor ones. One of the major bands corresponded to the control band in size, whereas the other one was ~150 bp shorter (fig. 4). Sequencing of the PCR products indicated that the insertion resulted in at least three splicing defects (not shown): one lacking sequences for exons 8 and 10, and the other two lacking sequences for either exon 8 or exon 10. Skipping of exon 8, exon 10, or exons 8 and 10 leads to an in-frame deletion of 25, 21, or 46 amino acids, respectively, in the COL3 domain of the α1(IX) chain.

Figure  4Figure 4
Agarose gel electrophoresis of α1(IX) reverse-transcriptase PCR products. RNA was isolated from EBV-transformed lymphoblasts from proband 41 and was analyzed for splicing by PCR as indicated in the Subject and Methods section. The control sample (more ...)

The proband, now aged 30 years, has had knee pains and difficulty in walking since the age of 10 years, and pain and stiffness in the knees had persisted. Radiographs indicated evidence of early OA in the right knee, Schmorl’s nodes, endplate irregularities, and anterior osteophytes in the thoraco-lumbar vertebrae, including calcification of the intervertebral disc (fig. 5A and B). The hips were normal, but the sacro-iliac joints were ill defined. The proband’s mother had arthritis affecting her hands, feet (fig. 5C), hips (fig. 5D), knees (fig. 5E), and spine, with symptoms starting at age 45 years. She has had a total knee replacement. The proband’s sister was asymptomatic, but their maternal grandmother had bilateral hip replacements.

Figure  5Figure 5
Radiographs of the thoraco-lumbar spine (A, lateral view) and lumbar spine (B, lateral view) of proband 41 (table 1), and of the left foot (C, AP view), hips (D, AP view), and left knee (E, AP view) of the proband’s affected mother. Radiographs (more ...)

Several sequence variations were identified, although these were likely to be neutral, since they were also found in controls. The variations that resulted in amino acid changes are shown in table 2. One of the variations in COL9A3, G→A resulting in the conversion of Arg to Gln, was not found in any of the 109 controls tested. This variation was found in proband 21 and in her unaffected father. Thus, the variation most likely represents a rare neutral sequence variant.

Table 2Table 2
Amino Acid Variations in COL9A1, COL9A2, and COL9A3

Analysis of the DTDST gene
Because a homozygous R279W DTDST mutation has been reported in an individual with autosomal recessively inherited MED characterized by club foot and/or double-layered patella, we tested for the presence of the mutation in the probands who had either multipartite patella or talipes deformities (probands 3, 4, 5, 14, 17, 18, 22, 26, 36, and 39; table 1) by StyI restriction-enzyme digestion. No heterozygotes were found, but probands 3 and 36 were homozygous for the mutation (data not shown). The mutation was confirmed by sequencing (data not shown). The parents of probands 3 and 36 were unaffected. Samples from the parents were not available for mutation analysis.

Discussion

MED is, clinically and radiologically, a highly heterogeneous disease for which five loci have so far been identified. Eleven mutations in COMP have been found in the Ribbing, Fairbanks, and unclassified forms (Briggs et al. 1995, 1998; Ballo et al. 1997; Susic et al. 1997; Ikegawa et al. 1998; Loughlin et al. 1998; Deere et al. 1999; Délot et al. 1999), and all of these mutations have been different and have been located in different domains of the molecule, thus probably partially explaining the clinical heterogeneity. However, all of the mutations identified to date in the other two genes, COL9A2 and COL9A3, have had the same consequence—that is, skipping of exon 3 and a resulting in-frame deletion of 12 amino acids (Muragaki et al. 1996; Holden et al. 1999; Paassilta et al. 1999a; Bonneman et al. 2000; Lohiniva et al. 2000; Spayde et al. 2000). All individuals with collagen IX splicing mutations also share a very similar phenotype, typically consisting of normal to near-normal height, as well as epiphyseal dysplasia of the knees and some other joints in childhood, but generally only OA of the knees in adulthood.

The phenotypic heterogeneity of MED may also be due to locus heterogeneity, a hypothesis supported by the results of the present linkage analysis. The analysis of four families with MED showed recombinations between the four candidate genes and the phenotype in two of the families, and between three of the candidate genes and the phenotype in the other two families. The results of the mutation screening analysis further supported the possibility of locus heterogeneity. Surprisingly, screening of the exon sequences and exon boundaries for the mutations in the four candidate genes, COL9A1, COL9A2, COL9A3, and COMP, resulted in the identification of only four mutations in the probands, a finding that suggests that mutations in collagen IX and COMP do not explain the majority of MED cases. Even though the CSGE method has proved to be highly sensitive (Körkkö et al. 1998), it is possible that some mutations have gone undetected. One of the limitations of the method is that it does not detect large deletions, and it is true that only point mutations or small deletions have been reported in the genes for collagen IX and COMP in patients with MED.

Of the three mutations found in COMP, two were novel, and both were located in the T3 repeats, as has been the case with most of the previously characterized mutations (Briggs et al. 1995, 1998; Ballo et al. 1997; Susic et al. 1997; Ikegawa et al. 1998; Loughlin et al. 1998; Deere et al. 1999). The third mutation, Thr585 to Met in the C-terminal globular domain, has been detected previously in a family with a mild form of PSACH (Briggs et al. 1998). The finding that the same mutation can lead to different but overlapping phenotypes is not surprising. First, the various MED and PSACH forms make up a phenotypic spectrum, and second, there are several examples of the same mutations causing variable phenotypes in different individuals—for example, collagen I mutations in osteogenesis imperfecta (MIM 166210 and MIM 166200) (Körkkö et al. 1997). Phenotypic variation can be evident even among members of one family, as was the case in one family with MED family that had a COL9A3 mutation (Paassilta et al. 1999a).

The mutation in COL9A1 identified in one family represents the first human disease-causing mutation ever reported in connection with this gene. Since COL9A1 contains six additional exons coding for the longer NC4 domain compared with the two other collagen IX genes, exon 9 of COL9A1 corresponds to exon 3 in the others. For this reason, it was somewhat surprising that the mutation, the insertion of a T at the donor splice site of IVS8+3, did not result in skipping of exon 9. The finding that it led to a complex splice pattern involving mainly exons 8 and 10 was not in itself surprising, however, since mutations at donor splice sites have been shown to lead to variable splice forms (Schwarze et al. 1999).

Although the exact phenotype of the two individuals carrying the COL9A1 splicing mutation is unclear in terms of whether cartilage degeneration typical of primary OA or of chondrodysplasia is the predominant feature, the phenotype of MED in the proband is supported by a history of knee pain and walking difficulties in childhood, as well as by the nature of the mutation. It has been shown elsewhere that a diagnosis of MED is difficult to establish in adult patients, in the case of collagen IX mutations (Muragaki et al. 1996; Holden et al. 1999; Paassilta et al. 1999a; Bonneman et al. 2000; Lohiniva et al. 2000; Spayde et al. 2000). Consequently, these mutations should be considered when evaluating patients with familial OA primarily affecting the knees.

Two additional loci, EDM4 and EDM5, have been identified recently in patients with MED. EDM4 contains the DTDST gene, in which a homozygous R279W mutation was shown to cause recessively inherited MED associated with club foot and double-layered patella (Superti-Furga et al. 1999). The same homozygous mutation was also reported in two sibships with apparently isolated club foot (Huber et al. 2001). This mutation was detected here in two probands with multipartite patella. The EDM5 locus contains the MATN3 gene (Chapman et al. 2001). Two different mutations were reported very recently in this gene in two families with autosomal dominant MED (Chapman et al. 2001). Since this gene was not analyzed here, it is possible that some of the probands in the present study may have mutations in this gene.

Acknowledgments

We thank Ms. Aira Harju and Ms. Marika Oikarinen for their expert technical assistance. We are also indebted to Mr. Neil Watson for his efforts in coordinating the clinical data. This work was supported in part by grants from the Academy of Finland, the Louisiana Gene Therapy Research Consortium (New Orleans), and HCA-The Health Care Company (Nashville) (all to L.A.-K.), and by Polish State Committee for Scientific Research grants 4P05A00716 and 2266/1A/162/97 (to A.L.-B. and M.C.-R.).

Electronic-Database Information

Accession numbers and the URL for data in this article are as follows:

Online Mendelian Inheritance in Man (OMIM), http://www.ncbi.nlm.nih.gov/Omim/ (for MED [MIM 132400], EDM1 [MIM 132400], EDM2 [MIM 600204], EDM3 [MIM 600969], EDM4 [MIM 226900], COMP [MIM 600310], PSACH [MIM 177170], COL9A1 [MIM 120210], COL9A2 [MIM 120260], COL9A3 [MIM 120270], DTDST [MIM 222600], MATN-3 [MIM 602109], Legg-Calve-Perthes disease [MIM 150600], spondyloepiphyseal dysplasia tarda [MIM 184100], spondyloepimetaphyseal dysplasia [MIM 300106], and osteogenesis imperfecta [MIM 166210 and MIM 166200]).
References
Ballo R, Briggs MD, Cohn DH, Knowlton RG, Beighton P, Ramesar RS (1997) Multiple epiphyseal dysplasia, Ribbing type: a novel point mutation in the COMP gene in a South African family. Am J Med Genet 68:396–400 [PubMed].
Bonnemann CG, Fox GF, Shapiro F, Wu J-J, Feener CA, Thompson TG, Anthony DC, Eyre DR, Darras BT, Kunkel LM (2000) A mutation in the α3 chain of type IX collagen causes autosomal dominant epiphyseal dysplasia with mild myopathy. Proc Natl Acad Sci USA 97:1212–1217 [PubMed].
Brewton RG, Wood BM, Ren Z-X, Gong Y, Tiller GE, Warman ML, Lee B, Horton WA, Olsen BR, Baker JR, Mayne R (1995) Molecular cloning of the α3 chain of human type IX collagen: linkage of the gene COL9A3 to chromosome 20q13.3. Genomics 30:329–336 [PubMed].
Briggs MD, Hoffmann SMG, King LM, Olsen AS, Mohrenweiser H, Leroy JG, Mortier GR, Rimoin DL, Lachman RS, Gaines ES, Cekleniak JA, Knowlton RG, Cohn DH (1995) Pseudoachondroplasia and multiple epiphyseal dysplasia due to mutations in the cartilage oligomeric matrix protein gene. Nat Genet 10:330–336 [PubMed].
Briggs MD, Mortier GR, Cole WG, King LM, Golik SS, Bonaventure J, Nuytinck L, De Paepe A, Leroy JG, Biesecker L, Lipson M, Wilcox WR, Lachman RS, Rimoin DL, Knowlton RG, Cohn DH (1998) Diverse mutations in the gene for cartilage oligomeric matrix protein in the pseudoachondroplasia-multiple epiphyseal dysplasia disease spectrum. Am J Hum Genet 62:311–319 [PubMed].
Chapman KL, Mortier GR, Chapman K, Loughlin J, Grant ME, Briggs MD (2001) Mutations in the region encoding the von Willebrand factor A domain of matrilin-3 are associated with multiple epiphyseal dysplasia. Nat Genet 28:393–396 [PubMed].
Chen H, Deere M, Hecht JT, Lawler J (2000) Cartilage oligomeric matrix protein is a calcium-binding protein, and a mutation in its type 3 repeats causes conformational changes. J Biol Chem 275:26538–26544 [PubMed].
Deere M. Sanford T, Ferguson HL, Daniels K, Hecht JT (1998) Identification of twelve mutations in cartilage oligomeric matrix protein (COMP) in patients with pseudoachondroplasia. Am J Med Genet 80:510–513 [PubMed].
Deere M, Sanford T, Francomano CA, Daniels K, Hecht JT (1999) Identification of nine novel mutations in cartilage oligomeric matrix protein in patients with pseudoachondroplasia and multiple epiphyseal dysplasia. Am J Med Genet 85:486–490 [PubMed].
Délot E, King LM, Briggs MD, Wilcox WR, Cohn DH (1999) Trinucleotide expansion mutations in the cartilage oligomeric matrix protein (COMP) gene. Hum Mol Genet 8:123–128 [PubMed].
Diab M, Wu JJ, Eyre DR (1996) Collagen type IX from human cartilage: a structural profile of intermolecular cross-linking sites. Biochem J 314:327–332 [PubMed].
Fairbank T (1947) Dysplasia epiphysialis multiplex. Br J Surg 34:225–232.
Hedbom E, Antonsson P, Hjerpe A, Aeschlimann D, Paulsson M, Rosa-Pimentel E, Sommarin Y, Wendel M, Oldberg A, Heinegard D (1992) Cartilage matrix proteins. An acidic oligomeric protein (COMP) detected only in cartilage. J Biol Chem 267:6132–6136 [PubMed].
Hoefnagel D, Sycamore LK, Russell SW, Bucknall WE (1967) Hereditary multiple epiphyseal dysplasia. Ann Hum Genet 30:201–210 [PubMed].
Holden P, Canty EG, Mortier GR, Zabel B, Spranger J, Carr A, Grant ME, Loughlin JA, Briggs MD (1999) Identification of novel pro-α2(IX) collagen gene mutations in two families with distinctive oligo-epiphyseal forms of multiple epiphyseal dysplasia. Am J Hum Genet 65:31–38 [PubMed].
Huber C, Odent S, Rumeur S, Padovani P, Penet C, Cormier-Daire V, Minnich A, Le Merrer M (2001) Sulphate transporter gene mutations in apparently isolated club foot. J Med Genet 38:191–192 [PubMed].
Hulvey JT, Keats T (1969) Multiple epiphyseal dysplasia: a contribution to the problem of spinal involvement. Am J Roentgenol 106:170–177.
Hummel KM, Neidhart M, Vilim V, Hauser N, Aicher WK, Gay RE, Gay S, Hauselmann HJ (1998) Analysis of cartilage oligomeric matrix protein (COMP) in synovial fibroblasts and synovial fluids. Br J Rheumatol 37:721–728 [PubMed].
Ikegawa S, Ohashi H, Nishimura G, Kim KC, Fukushima Y, Nagai T, Nakamura Y (1998) Novel and recurrent COMP (cartilage oligomeric matrix protein) mutations in pseudoachondroplasia and multiple epiphyseal dysplasia. Hum Genet 103:633–638 [PubMed].
International Working Group on Constitutional Diseases of Bone (1998). International nomenclature and classification of the osteochondrodysplasias (1997). Am J Med Genet 79:376–382 [PubMed].
Körkkö J, Annunen S, Pihlajamaa T, Prockop DJ, Ala-Kokko L (1998) Conformation sensitive gel electrophoresis (GSGE) for simple and accurate detection of mutations. Comparison with denaturing gradient gel electrophoresis (DGGE) and nucleotide sequencing. Proc Natl Acad Sci USA 95:1681–1685 [PubMed].
Körkkö J, Kuivaniemi H, Paassilta P, Zhuang J, Tromp G, DePaepe A, Prockop DJ, Ala-Kokko (1997) Two new recurrent nucleotide mutations in the COL1A1 gene in four patients with osteogenesis imperfecta: about one-fifth are recurrent. Hum Mutat 9:148–156 [PubMed].
Lohiniva J, Paassilta P, Seppänen U, Vierimaa O, Kivirikko S, Ala-Kokko L (2000) Splicing mutations in the COL3 domain of collagen IX cause multiple epiphyseal dysplasia. Am J Med Genet 90:216–222 [PubMed].
Loughlin J, Irven C, Mustafa Z, Briggs MD, Carr A, Lynch S-A, Knowlton RG, Cohn DH, Sykes B (1998) Identification of five novel mutations in cartilage oligomeric matrix protein gene in pseudoachondroplasia and multiple epiphyseal dysplasia. Hum Mutat 1998 Suppl 1:S10–S17 [PubMed].
Maddox BK, Mokashi A, Keene DR, Bächinger HP (2000) A cartilage oligomeric matrix protein mutation associated with pseudoachondrodysplasia changes the structural and functional properties of the type 3 domain. J Biol Chem 275: 11412–11417 [PubMed].
Mäkelä TP, Hellsten E, Vesa J, Alitalo K, Peltonen L (1992) An Alu variable polyA repeat polymorphism upstream of L-myc at 1p32. Hum Mol Genet 1:217 [PubMed].
Malashkevich VN, Kammerer RA, Efimov VP, Schulthess T, Engel J (1996) The crystal structure of a five-stranded coiled coil in COMP: a prototype ion channel? Science 274:761–765 [PubMed].
Misenheimer TM, Mosher DF (1995) Calcium ion binding to thrombospondin 1. J Biol Chem 270:1729–1733 [PubMed].
Mörgelin M, Heinegård D, Engel J, Paulsson M (1992) Electron microscopy of native cartilage oligomeric protein purified from the swarm rat chondrosarcoma reveals a five-armed structure. J Biol Chem 267:6137–6141 [PubMed].
Muragaki Y, Mariman ECM, van Beersum SEC, Perälä M, van Mourik JBA, Warman ML, Olsen BR, Hamel BCJ (1996) A mutation in the gene encoding the α2 chain of the fibril-associated collagen IX, COL9A2, causes multiple epiphyseal dysplasia (EDM2). Nat Genet 12:103–105 [PubMed].
Newman B, Donnai D, Briggs MD (2000) Molecular diagnosis is important to confirm suspected pseudoachondroplasia. J Med Genet 37:64–65 [PubMed].
Newton G, Weremowics S, Morton CC, Copeland NG, Gilbert DJ, Jenkins NA, Lawler J (1994) Characterization of human and mouse cartilage oligomeric matrix protein. Genomics 24:435–439 [PubMed].
Oehlmann R, Summerville GP, Yeh G, Weaver EJ, Jimenez SA, Knowlton RG (1994) Genetic linkage mapping of multiple epiphyseal dysplasia to the pericentromeric region of chromosome 19. Am J Hum Genet 54:3–10 [PubMed].
Oldberg A, Antonsson P, Lindblom K, Heinegard D (1992) COMP (cartilage oligomeric matrix protein) is structurally related to the thrombospondins. J Biol Chem 267:22346–22350 [PubMed].
Paassilta P, Lohiniva J, Annunen S, Bonaventure J, Le Merrer M, Pai L, Ala-Kokko L (1999a) COL9A3: a third locus for multiple epiphyseal dysplasia. Am J Hum Genet 64:1036–1044 [PubMed].
Paassilta P, Pihlajamaa T, Annunen S, Brewton RG, Wood BM, Johnson CC, Liu J, Gong Y, Warman ML, Prockop DJ, Mayne R, Ala-Kokko L (1999b) Complete sequence of 23 kb human COL9A3 gene: detection of Gly-X-Y triplet deletions that represent neutral variants. J Biol Chem 274:22469–22475 [PubMed].
Pihlajamaa T, Vuoristo MM, Annunen S, Perälä M, Prockop DJ, Ala-Kokko L (1998) Two genes of 90 and 15 kb code for similar polypeptides of the same collagen molecule. Matrix Biol 17:237–241 [PubMed].
Ribbing S (1937) Studien über die hereditäre multiple Epiphysenstörungen. Acta Radiol (Stockholm) Suppl 34:77–107.
Schwarze U, Starman BJ, Byers PH (1999) Redefinition of exon 7 in the COL1A1 gene of type I collagen by an intron 8 splice-donor–site mutation in a form of osteogenesis imperfecta: influence of intron splice order on outcome of splice-site mutation. Am J Hum Genet 65:336–344 [PubMed].
Sheffield EG (1998) Double-layered patella in multiple epiphyseal dysplasia: a valuable clue in the diagnosis. J Pediatr Orthop 18:123–128 [PubMed].
Silverman FN (1996) C John Hodson Lecture. Reflections on epiphyseal dysplasias. Am J Roentgenol 167:835–842 [PubMed].
Spayde EC, Joshi AP, Wilcox WR, Briggs, Cohn DH, Olsen BR (2000) Exon skipping mutation in the COL9A2 gene in a family with multiple epiphyseal dysplasia. Matrix Biol 19:121–128 [PubMed].
Superti-Furga A, Neumann L, Riebel T, Eich G, Steinmann B, Spranger J, Kunze J (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 [PubMed].
Susic S, McGrory J, Ahier J, Cole WG (1997) Multiple epiphyseal dysplasia and pseudoachondroplasia due to novel mutations in the calmodulin-like repeats of cartilage oligomeric matrix protein. Clin Genet 51:219–224 [PubMed].
Thur J, Rosenberg K, Nitsche DP, Pihlajamaa T, Ala-Kokko L, Heinegård D, Paulsson M, Maurer P (2001) Mutations in cartilage matrix oligomeric protein causing pseudoachondroplasia and multiple epiphyseal dysplasia affect binding of calcium and collagen I, II, and IX. J Biol Chem 276:6083–6092 [PubMed].
Warman ML, McCarthy MT, Perälä M, Vuorio E, Knoll JH, McDaniels CN, Mayne R, Beier DR, Olsen BR (1994) The genes encoding α2(IX) collagen (COL9A2) map to human chromosome 1p32.3-p33 and mouse chromosome 4. Genomics 23:158–162 [PubMed].
Weissenbach J, Gyapay G, Dib C, Vignal A, Morissette J, Millasseau P, Vaysseix G, Lathrop M (1992) A second generation linkage map of the human genome. Nature 359:794–801 [PubMed].
Werle E, Schneider C, Renner M, Volker M, Fiehn W (1994) Convenient single-step, one tube purification of PCR products for direct sequencing. Nucleic Acids Res 22:4354–4355 [PubMed].