Evidence for locus heterogeneity in the Bethlem myopathy and linkage to 2q37
Evidence for locus heterogeneity in the Bethlem myopathy and linkage to 2q37Marcy C. Speer1,*, Rup Tandan2, P. Nagesh Rao3, Timothy Fries2, Jeffrey M. Stajich1, Pieter A. Bolhuis4, G. Joost Jöbsis4, Jeffery M. Vance1, Kristi D. Viles1, Karen Sheffield5, Christi James6, Stephen G. Kahler7, Mark Pettenati3, John R. Gilbert1, Peter H. Denton1, Larry H. Yamaoka1 and Margaret A. Pericak-Vance1
1Division of Neurology, Duke University Medical Center, Durham, NC 27710, USA, 2University Health Center, Burlington, Vermont, 05401 USA, 3Department of Pediatrics, Bowman Gray School of Medicine, Winston-Salem, NC 27157, USA, 4University of Amsterdam, Amsterdam, The Netherlands, 5Southwest Foundation for Biomedical Research, San Antonio, TX 78228, USA, 6Duke University Medical Center, Durham, NC 27710, USA and 7Department of Pediatrics, Duke University Medical Center, Durham, NC 27710, USA
Received January 16, 1996;Revised and Accepted April 11, 1996
The Bethlem myopathy, a childhood onset autosomal dominant myopathy with joint contractures, has recently been localized to 21q in a series of Dutch families and the [alpha]1 and [alpha]2 subunits of type VI collagen (COL6A1 and COL6A2) have been postulated as candidate genes. We investigate a large family of French Canadian descent (family 1489) in which the Bethlem myopathy is segregating. Family 1489 is unlinked to the region of interest on 21q, thus demonstrating locus heterogeneity within the Bethlem myopathy classification. In view of the localization of the genes coding the [alpha]1 and [alpha]2 subunits of type VI collagen on chromosome 21q, we carried out linkage analysis on chromosome 2q where the [alpha]3 subunit of type VI collagen has been localized. We demonstrate linkage to markers in this region, define the region of disease gene localization, and confirm by FISH analysis that COL6A3 is located within the interval of interest making COL6A3 a feasible candidate gene for the Bethlem myopathy.
Bethlem myopathy is a rare, childhood onset autosomal dominant proximal myopathy with widespread flexion contractures involving most frequently the elbows and ankles. Neither facial nor cardiac musculature are involved. Progression of the disorder is slow and the clinical course is generally mild with no decrease in life expectancy. The creatine kinase values range from normal to four times normal. Originally described in 1976 in a series of Dutch families (1 ), later reports described families of diverse ethnicity (2 -4 ) including a large family of French Canadian descent (5 ). All families appear to be phenotypically similar. Recently, several Bethlem myopathy families of Dutch background were linked to chromosome 21 (6 ,7 ).
We investigate this French Canadian family (DUK 1489) for linkage and show evidence for locus heterogeneity within the Bethlem myopathy. We further demonstrate that family 1489 is linked to an interval spanned by D2S336 and D2S395 on chromosome 2 and establish the [alpha]3 subunit of type VI collagen as a viable candidate for this disorder.
Evidence for linkage in a series of Dutch Bethlem myopathy families was reported to 21q (6 ,7 ) with several markers includingD21S212. Markers flanking the interval for disease gene localization in these families have not yet been determined. We have excluded family 1489 (LOD scores <= -2.0) for approximately 18 cM on either side of D21S212, with significant evidence for locus heterogeneity within this diagnostic classification ([chi]21 = 13.34, p <0.001).
For the chromosome 21 linked families, two candidate genes were postulated, COL6A1 and COL6A2, the [alpha]1 and [alpha]2 subunits of type VI collagen (6 ,7 ). A collagen gene is a feasible candidate because of the joint contractures associated with the disease phenotype. The third subunit of type VI collagen (COL6A3) is located on 2q, although not mapped by genetic means, and thus 2q was considered a likely area for investigation for linkage of family 1489.
To assess evidence for linkage to 2q, markers D2S336, D2S345, D2S338, and D2S395 were genotyped in family 1489. Evidence in favor of linkage (LOD scores >= 3.0) was demonstrated with D2S336, D2S345, and D2S338 (Table 1 ). There was no evidence for recombination between the disease gene and D2S345 or D2S338 with maximum LOD scores of 8.13 and 7.03, respectively, using the full pedigree analysis. Similar conclusions are drawn from the `low penetrance' analyses (Table 1 ).
. Two-point LOD scores for Bethlem myopathy and chromosome 2 markers
Penetrance = 0.95
`Low penetrance' analysis
One LOD
One LOD
Marker
0.00
0.05
0.10
0.15
0.20
0.30
0.40
z([theta])
[theta]
unit SI
z([theta])
[theta]
unit SI
D2S336
-[infinity]
3.04
3.49
3.48
3.27
2.45
1.26
3.52
0.12
0.02-0.30
2.57
0.10
NA*
D2S345
8.13
7.47
6.78
6.06
5.29
3.61
1.74
8.13
0.00
0.00-0.08
5.01
0.00
0.00-0.12
D2S338
7.03
6.45
5.84
5.20
4.53
3.08
1.47
7.03
0.00
0.00-0.09
4.58
0.00
0.00-0.13
D2S395
-[infinity]
2.00
2.21
2.16
1.98
1.43
0.69
2.22
0.11
NA
0.87
0.14
NA
*Not applicable.
Multipoint linkage analysis confirmed the results expected by two-point linkage analysis: the maximum multipoint LOD score occurred at when the disease locus demonstrated no recombination with D2S338/D2S345 with a support interval spanning from 3 cM proximal to D2S338 and 7 cM distal to D2S338. As expected, the low penetrance analysis suggested a slightly greater support interval (data not shown). The disease gene is localized to a 17 cM interval spanned by D2S336 and D2S395(Fig. 1 ); no other markers have been confirmed to be localized to this area to allow further sublocalization.
The initial investigation of COL6A3 as a candidate involved the genotyping of two RFLPs within COL6A3 in family 1489. These polymorphisms were relatively uninformative: only 7/26 available meioses were informative with respect to disease status. None of these informative meioses showed evidence for recombination with the COL6A3 gene. COL6A3 demonstrated recombination events suggesting its placement distal to D2S336 and proximal to D2S395(data not shown).
In order to determine whether COL6A3 was localized to the interval of interest spanned by D2S336 and D2S395, fluorescence in situ hybridization (FISH) was performed with YACs positive for COL6A3,D2S336, and D2S395 probes. This analysis revealed their positions to be at the 2q37 region on banded chromosomes. In the dual color FISH with pairwise combination of the three markers, the order of the probes on the metaphase chromosome was cen-D2S336-COL6A3-D2S395-tel. These results together with the linkage analysis are consistent with the hypothesis that the disease phenotype in this family linked to 2q and the Dutch families linked to 21q is caused by mutations in different subunits of type VI collagen.
COL6A3 is a compelling candidate for Bethlem myopathy. Mutations in collagen genes have proven to be responsible for a wide variety of disorders including osteogenesis imperfecta [COL1A1 (e.g., 8 ) and COL1A2 (e.g., 9 )], Ehlers-Danlos syndrome [COL1A1 (e.g., 10 ) and COL1A2 (e.g., 11 )], the X-linked Alport syndrome [COL4A5 (e.g., 12 )], and the autosomal recessive Alport syndrome [COL4A3 (e.g., 13 ) and COL4A4 (e.g., 13 )], among others.
In some cases, the distribution of a particular collagen protein accurately predicts the functional consequences of a mutation. For example, in Ehlers-Danlos syndrome type IV, patients may experience uterine rupture during labor, diverticuli, and arterial rupture after mild trauma, all due to failure of medium sized vessels. The genetic defect in Ehlers-Danlos syndrome type IV has been identified to be in COL3A1, which is primarily expressed in these medium sized vessels (e.g., 14 ). Type VI collagen, however, is ubiquitously distributed in essentially all tissues (15 ,16 ) and thus no obvious predictions with respect to the Bethlem myopathy phenotype are possible. Nonetheless, there is precedence for mutations in different sub-units of a collagen protein leading to similar phenotypes: mutations in COL1A1 or COL1A2 lead to various types of osteogenesis imperfecta (8 ,9 ); similarly, mutations in either COL1A1 or COL1A2 can lead to different forms of Ehlers-Danlos syndrome (10 ,11 ).
There is no clear mechanism for understanding the phenotypic variability in these disorders since the type or location of mutation does not usually predict the resultant phenotype. The triple helical domain would be one area to screen for mutations as many OI variants are due to changes in this domain of COL1A1 and COL1A2 (e.g., 17 ,18 ); similarly, deletions in the triple helical domain of COL3A1 have been reported in Ehlers-Danlos type IV (14 ). Another potential mechanism for mutation in this disorder would be alternative splicing, which has previously been observed in COL6A3 expression in normal and tumor cells (19 ). Should defects in COL6A3 be proven to be responsible for the Bethlem myopathy, this will be the first myopathic disorder in which a collagen gene has been implicated. Further investigations as to the validity of COL6A3 as a candidate gene for the Bethlem myopathy in this family are underway.
Thirty-six members of family 1489 are included in the present study. Blood was obtained for creatine kinase testing and DNA extraction. DNA was extracted from leukocytes on an ABI GenePure 341. The diagnostic criteria for Bethlem myopathy include proximal greater than distal extremity weakness, joint contractures, and childhood onset of symptoms. Of these 36 individuals, 19 are affected, 12 unaffected, one is of unknown diagnostic status, and four are unrelated spouses. Parents report the age of onset of symptoms in their children to be approximately 2-5 years of age. One asymptomatic individual at 50% a priori risk to carry the Bethlem myopathy gene had a mildly elevated creatine kinase value of 294 (normal range 61-200) and was considered to be of unknown diagnostic status for the linkage analysis.
This study was approved by the Institutional Review Board at Duke University Medical Center. Informed consent was obtained from all study participants.
All pedigree, clinical, DNA sample, and genotyping data is stored in the PEDIGENE database management system (20 ).
Dinucleotide (CA)n polymorphism genotypings were carried out according to methods as previously described (21 ) and genotypes for critical recombination events were confirmed by re-genotyping. To investigate linkage to 21q, D21S212 (22 ) was genotyped. Investigation of linkage to 2q was performed by genotyping the markers D2S336, D2S345, D2S338, and D2S395 (23 ). Two RFLPs were identified by the cDNA for COL6A3 (24 ,25 ). Genomic DNA from family members was digested with BglI and NcoI and hybridized with the COL6A3 cDNA according to methods previously described (21 ).
For the linkage analysis, individuals were considered to be affected when they met the diagnostic criteria as defined above. All spouses were considered to be normal with respect to clinical status. Linkage analysis was performed under the assumption that the Bethlem myopathy is an autosomal dominant trait with disease allele frequency of 0.0001. Since onset is in early childhood and no cases of obligate, asymptomatic carriers have been reported, penetrance was assumed to be nearly complete and assigned a value of 0.95. Two-point and multipoint linkage analyses were performed using the MLINK and LINKMAP modules of the LINKAGE package (version 5;1) (26 ) as implemented in the FASTLINK program (27 ,28 ). No evidence for recombination was found between the markers D2S345 and D2S338 and thus these were haplotyped for multipoint analysis to maximize the linkage information. The genetic map for the multipoint linkage analysis was limited to three loci with distances cen-D2S336-0.05-D2S345/D2S338-0.12- D2S395-tel (23 ). Allele frequencies were calculated from a series of at least 64 unrelated Caucasian controls. Frequencies for these markers and for others are available via anonymous ftp (site: dnadoc.mc.duke.edu in the /pub/ALLELE_FREQ directory). In addition, `low penetrance' analyses were performed using only phenotypic information from family members while retaining marker genotypic information on all family members, regardless of affection status. This approach utilizes only those individuals whose phenotypic classification is not subject to misclassification due to incorrect specification of disease allele penetrance and maximizes the available data for inferences of unavailable parental marker genotypes. Assessment of linkage heterogeneity was performed as described previously (21 ) after combining the LOD scores from family 1489 and those reported earlier as linked to 21q (6 ,7 ).
The CEPH YAC megabase library was the source for selection of YACs for D2S336, D2S395, and COL6A3. The following YACs were identified as positives for the individual primers: 773 H6 (D2S336), 787 B12 (COL6A3), and 780 A10 (D2S395). These YACs were grown up and the DNA extracted from each culture. The DNA from each YAC was re-tested with the corresponding primers to confirm positives. YAC selection and FISH analysis are as previously described (29 ).
The YACs were labeled either with digoxigenin-11-dUTP (Boehringer Mannheim) or biotin (Bionick, BRL) and hybridized to prometaphase chromosomes prepared from peripheral blood lymphocytes. The probes were hybridized singularly to determine the chimeric nature or co-hybridized with other probes when ordering their relative positions and detected using the dual-color technique (FITC and Rhodamine) (30 ). To determine specific band assignments, simultaneous detection of in situ hybridization signals and chromosome banding was performed by the technique described by (31 ) with some modifications. Briefly, three day old lymphocyte cultures were synchronized with BrdU (250 [mu]g/ml), washed twice in RPMI-1640 medium, and exposed to thymidine (2.5 [mu]g/mo) for 6 h prior to harvest. The slides were stained with 0.5 [mu]g/ml Hoechst 33258 for 15 min, mounted in MacIlvaine's buffer (pH 7.5), and irradiated with a 15 W black light source for 15 min at 50oC. The slides were counterstained with DAPI or Propidium Iodide (PI). At least 20 metaphases were examined per hybridization using a Zeiss Axiophot microscope equipped with a filter set simultaneously transmitting FITC-labeled probe sites and PI and a triple band pass filter (DAPI, FITC, and Rhodamine). Sublocalization of COL6A3 to chromosomal region 2q37 was unequivocally made on banded metaphase spreads. True images were captured and stored on TOMS Image Analysis System (TOMS, Columbus GA) and printed on a SONY color printer.
The authors would like to thank Deborah Gross, Helen Harbett, and Peggy Pate for data entry and management, Carol Haynes and Mark Peedin for database management, and Michelle Eyster and Tim Tucker for DNA extraction and preparation. The authors further acknowledge support from grants from the Muscular Dystrophy Association (MCS, MAP-V) and the National Institutes of Health (NS26630 MAP-V) and the Prinses Beatrix Fonds, The Netherlands (PAB, GJJ). The authors thank the members of family 1489 who agreed to participate in the study, without whose long-term commitment to this project this work would not be possible.
1 Bethlem, J., van Wijngaarden, G.K. (1976) Benign myopathy, with autosomal dominant inheritance. Brain 99, 91-100.MEDLINE Abstract
2 Arts, W.F., Bethlem, J., Volkers, W.S. (1978) Further investigations on benign myopathy with autosomal dominant inheritance. J. Neurol.,217, 201-206.MEDLINE Abstract
3 Tachi, N., Tachi, M., Sasaki, K., Imamura, S. (1989) Early onset benign autosomal dominant limb-girdle myopathy with contractures (Bethlem myopathy). Pediatr. Neurol.,5, 232-236.MEDLINE Abstract
4 Merlini, L., Morandi, L., Granata, C., Ballestrazzi, A. (1994) Bethlem myopathy: Early onset benign autosomal dominant myopathy with contractures. Description of two new families. Neuromusc. Disord.,4, 503-511.MEDLINE Abstract
7 Jöbsis, G.J., Barth, P.G., Boers, J.M., Bass, F., Bolhuis, P.A., de Visser, M. (1996) Bethlem myopathy: Clinical and genetic aspects. Neurology, 46, 779-782.MEDLINE Abstract
8 Pope, F.M., Nicholls, A.C., McPheat, J., Talmud, P., Owen, R. (1985) Collagen genes and proteins in osteogenesis imperfecta. J. Med. Genet.,22, 466-478.MEDLINE Abstract
9 Myers, J.C., Dickson, L.A., Pope, F.M., Korhonen, V.R., Nicholls, A., Prockop, D.J., Pihlajaniemi, T. (1985) A homozygous frameshift mutation in the pro-alpha-1(I) collagen COOH-propeptide results in osteogenesis imperfecta. Ann. N.Y. Acad. Sci., 460, 482-485.
10 Cole, W.G., Chan, D., Chambers, G.W., Walker, I.D., Bateman, J.F. (1986) Deletion of 24 amino acids from the pro-alpha-1(I) chain of type I procollagen in a patient with the Ehlers-Danlos syndrome type VII. J. Biol. Chem., 261, 5496-5503.MEDLINE Abstract
11 Minor, R. R., Sippola-Thiele, M., McKeon, J., Berger, J., Prockop, D.J. (1986) Defects in the processing of procollagen to collagen are demonstrable in cultured fibroblasts from patients with the Ehlers-Danlos and osteogenesis imperfecta syndromes. J. Biol. Chem., 261, 10006-10014.MEDLINE Abstract
12 Barker, D.F., Hostikka, S.LO., Zhou, J., Chow, L.T., Oliphant, A.R., Gerken, S.C., Gregory, M.C., Skolnick, M.H., Atkin, C.O., Tryggvason, K. (1990) Identification of mutations in the COL4A5 collagen gene in Alport syndrome. Science, 248, 1224-1227.MEDLINE Abstract
13 Mochizuki, T., Lemmink, H.H., Mariyama, M., Antignac, C., Gubler, M.-C., Pirson, Y., Verellen-Dumoulin, C., Chan, B., Schroder, C.H., Smeets, H.J., Reeders, S.T. (1994) Identification of mutations in the alpha-3(IV) and alpha-4(IV) collagen genes in autosomal recessive Alport syndrome. Nature Genet., 8, 77-81.MEDLINE Abstract
14 Superti-Furga, A., Steinmann, B. (1988) Impaired secretion of type III procollagen in Ehlers-Danlos syndrome type IV fibroblasts: correction of the defect by incubation at reduced temperature and demonstration of subtle alerations in the triple-helical region of the molecule. Biochem. Biophys. Res. Commun., 150, 140-147.MEDLINE Abstract
15 Bruns, R.R., Press, W., Engvall, E., Timpl, R., Gross, J. (1986) Type VI collagen in extracellular, 100 nm periodic filaments and fibrils: identification by immunoelectron microscopy. J. Cell. Biol., 103, 393-404.MEDLINE Abstract
16 Von der Mark, H., Aumailley, M., Wick, G., Fleischmajer, R., Timpl, R. (1984) Immunochemistry, genuine size and tissue localization of collagen VI. Eur. J. Biochem., 142, 493-502.MEDLINE Abstract
17 Valli, M., Mottes, M., Tenni, R., Sangalli, A., Gomez-Lira, M., Rossi, A., Antoniazzi, F., Cetta, G., Pignatti, P.F. (1991) A de novo G to T transversion in a pro-alpha-1(I) collagen gene for a moderate case of osteogenesis imperfecta: substitution of cysteine for glycine 178 in the triple helical domain. J. Biol. Chem., 266, 1872-1878.MEDLINE Abstract
18 Cole, W.G., Patterson, E., Bonadio, J., Campbell, P.E., Fortune, D.W. (1992) The clinicopathological features of three babies with osteogenesis imperfecta resulting from the substitution of glycine by valine in the pro alpha-1(I) chain of type I procollagen. J. Med. Genet., 29, 112-118.MEDLINE Abstract
19 Stokes, D.G., Saitta, B., Timpl, R., Chu, M.L. (1991) Human alpha 3(VI) collagen gene. Characterization of exons coding for the amino-terminal globular domain and alternative splicing in normal and tumor cells. J. Biol. Chem., 266, 8626-8633.MEDLINE Abstract
20 Haynes, C., Speer, M.C., Peedin, M., Roses, A.D., Haines, J.L., Vance, J.M., Pericak-Vance, M.A.(1995) PEDIGENE: A comprehensive data management system to facilitate efficient and rapid disease gene mapping. Am. J. Hum. Genet., 57, A193.
21 Speer, M.C., Yamaoka, L.H., Gilchrist, J.M., Gaskell, C.P., Stajich, J.M., Vance, J.M., Kazantsev, A., Lastra, A.A., Haynes, C.S., Beckmann, J.S., Cohen, D., Weber, J.L., Roses, A.D., Pericak-Vance, M.A. (1992) Confirmation of genetic heterogeneity in limb-girdle muscular dystrophy: Linkage of an autosomal dominant form to chromosome 5q. Am. J. Hum. Genet., 50, 1211-1217.MEDLINE Abstract
22 Warren, A.C., McInnis, M.G., Blaschak, J., Kaliatsidaki, M., Petersen, M.B., Chakravarti, A., Antonarakis, S.E. (1992) Dinucleotide repeat (GT)n markers on chromosome 21. Genomics,14, 818-819.MEDLINE Abstract
23 Gyapay, G., Morissette, J., Vignal, A., Dib, C., Fizames, C.,Millasseau, P., Marc, S., Bernardi, G., Lathrop, M., Weissenbach, J.(1994) The 1993-94 Généthon human genetic linkage map. Nature Genet,7, 246-339.MEDLINE Abstract
24 Weaver, E.J., Chu, M-L., Knowlton, R.G. (1988) A BglI polymorphism in the type VI ([alpha]3) procollagen gene [COL6A3]. Nucleic Acids Res.,16, 11386.MEDLINE Abstract
25 Brotherton, D., Chu, M-L., Humphries, S.E., Henney, A.M. (1989) An NcoI RFLP associated with the gene encoding the [alpha]-3 chain of human type VI collagen (COL6A3). Nucleic Acids Res.,17, 1274.MEDLINE Abstract
26 Lathrop, G.M., Lalouel, J.M., Julier, C., Ott, J. (1984) Strategies for multilocus linkage analysis in humans. Proc. Natl Acad. Sci. USA, 81, 3443-3446.MEDLINE Abstract
27 Schaffer, A.A., Gupta, S.K., Siriram, K., Cottingham Jr., R.W. (1994) Avoiding recomputation in linkage analysis. Human Heredity,44, 225-237.MEDLINE Abstract
29 Ben Othmane, K., Loeb, D., Hayworth-Hodge, R., Hentati, F., Rao, N., Roses, A.D., Hamida, B., Pericak-Vance, M.A., Vance, J.M. (1995) Physical and genetic mapping of the CMT4A locus and exclusion of PMP-2 as the defect in CMT4A. Genomics,28, 286-290.
30 Rothschild, C.B., Akots, G., Hayworth, R., Pettenati, M.J., Rao, P.N., Wood, P., Frank-Michael, S., Hansmann, I., Serino, K, Keith, T.P., Fajans, S.S., Bowden, D.W. (1993) A genetic map of chromosome 20q12-13.1: Multiple highly polymorphic microsatellite and RFLP markers linked to the maturity onset diabetes of the young (MODY) locus. Am. J. Hum. Genet.,52, 110-123.MEDLINE Abstract
31 Hirai, M., Suto, Y., Kanoh, M. (1994) A method for simultaneous detection of fluorescent G-bands and in situ hybridization signals. Cytogenet. Cell Genet., 66, 149-151.MEDLINE Abstract
*To whom correspondence should be addressed
This page is maintained by OUP admin. Last updated Thu Oct 31 15:25:15 GMT 1996. Part of the OUP Journals World Wide Web service.Copyright Oxford University Press, 1996
