0021-972X/99/$03.00/0 The Journal of Clinical Endocrinology & Metabolism Copyright © 1999 by The Endocrine Society
Vol. 84, No. 3 Printed in U.S.A.
Locus Heterogeneity of Autosomal Dominant Osteopetrosis (ADO)* KENNETH E. WHITE, DANIEL L. KOLLER, ISTVAN TAKACS, KENNETH A. BUCKWALTER, TATIANA FOROUD, AND MICHAEL J. ECONS Departments of Medicine (K.E.W., I.T., M.J.E.), Medical and Molecular Genetics (D.L.K., T.F., M.J.E.), and Radiology (K.A.B.), Indiana University School of Medicine, Indianapolis, Indiana 46202 ABSTRACT Autosomal dominant osteopetrosis (ADO), is a heritable disorder that results from a failure of osteoclast-mediated bone resorption. The etiology of the disorder is unknown. A previous linkage study of one Danish family mapped an ADO locus to chromosome 1p21. We have studied two families from Indiana with ADO. The present study sought to determine if the ADO gene in these families was also linked to chromosome 1p21. We used six microsatellite repeat markers, which demonstrated linkage to the 1p21 ADO locus in the Danish
A
UTOSOMAL dominant osteopetrosis (ADO), also known as Albers-Scho¨nberg disease, is an inherited, osteosclerotic disorder that results from inadequate osteoclast-mediated skeletal resorption (1, 2). ADO has variable penetrance (3, 4) and occurs with an estimated prevalence between 1 in 20,000 (4) and 1 in 100,000 (3). The disorder is referred to as the “benign” form of osteopetrosis, to distinguish it from the autosomal recessive “malignant” forms of the disease, which are typically fatal if untreated (3). Although some individuals with ADO are asymptomatic, patients may present with dense but brittle bones that are prone to fracture, bone pain, osteomyelitis (particularly of the mandible), and nerve entrapment syndromes (3). Anemia with extramedullary hematopoiesis may also develop as a result of insufficient marrow space (3). Individuals affected with ADO display several radiographic and biochemical hallmarks. Upon X-ray analysis, patients frequently have osteosclerosis; endobone (“bone within bone”) formation at the hips, spine, and extremities; as well as vertebral end-plate sclerosis (Rugger-Jersey spine) (5, 6). Biochemically, serum levels of the brain-specific isoform of creatine kinase (CK-BB) (7, 8) and tartrate-resistant alkaline phosphatase (TRAP) (3) may be elevated in affected individuals. Interestingly, obligate carriers of a defective ADO gene have been identified in multiple families and show no detectable radiographic or biochemical manifestations of the disorder (3, 4, 9).
Received October 16, 1998. Revision received December 30, 1998. Accepted January 5, 1999. Address correspondence and requests for reprints to: Michael J. Econs, M.D., Indiana University School of Medicine, 975 W. Walnut St. IB445, Indianapolis, Indiana 46202. E-mail:
[email protected] *This research was supported by a grant from the Charles E. Culpeper Foundation, by NIH Grants AR42228 and AG05793, and by a Department of Medical and Molecular Genetics PHS Training Grant T32HD07373 (D.L.K.).
study, to perform linkage analysis in the new kindreds. Multipoint analysis excluded linkage of ADO to chromosome 1p21 (logarithm of the odds score , 27.00) in both families. In addition, no haplotype segregated with the disorder in either family. In summary, the present investigation ruled out linkage of ADO to chromosome 1p21 in two families from Indiana. Our results demonstrate that there is locus heterogeneity of this disorder; therefore, mutations in at least two different genes can give rise to the ADO phenotype. (J Clin Endocrinol Metab 84: 1047–1051, 1999)
Although the disorder results from insufficient osteoclast mediated bone resorption, the etiology of ADO is presently unknown and may not be understood until the gene is identified. In an effort to determine the chromosomal location of the disease gene, a linkage study was undertaken by Van Hul and colleagues (10) in a single Danish ADO family. Several candidate regions were analyzed, and linkage to 1p21 was identified (10). The ADO locus mapped to an 8.5 cM region between the markers D1S486 and D1S2792 [logarithm of the odds (LOD) . 4.00 ]. Because only a single family was used in this linkage analysis, it is unknown whether all ADO cases result from mutations in the same gene, or if the disease can result from mutations in more than one gene (locus heterogeneity). Therefore, the goal of the present study was to determine if the ADO gene locus in two ADO families of non-Danish descent also linked to chromosome 1p21. Materials and Methods Patients Family number EOP1 was previously reported by Johnston et al. (3), and family number EOP3 was referred to us by the Department of Medical and Molecular Genetics at Indiana University School of Medicine. We obtained blood samples from 8 members of family EOP1 and 19 members of family EOP3. Phenotypic evaluation of family members continues to proceed as medical records and radiographs become available. Only individuals with radiographic evidence of vertebral end-plate sclerosis or endobones were considered affected. If radiographic evidence of ADO was not present in an individual, but they had an affected parent or sibling and an affected child, they were considered an obligate gene carrier. Radiographs were read by a radiologist specializing in metabolic bone disorders (K.A.B.), who was blinded to all other phenotypic and genotypic information. Radiographs from normal spouses were included as negative controls, when available, to assure consistency of film evaluation. The study was approved by the Indiana University School of Medicine Institutional Review Board, and all patients gave written, informed consent before participating.
1047
1048
WHITE ET AL.
JCE & M • 1999 Vol 84 • No 3
FIG. 1. Radiographs of an ADO patient. A, Lateral chest X-ray of an affected female at 15 yr of age from family EOP3 showing the characteristic vertebral end-plate sclerosis (arrow on left side of panel A), as well as endobone formation in the sternum (arrow on right side of panel A). B, Lateral chest x-ray of the same individual at 3 yr of age, in which signs of ADO in the spine and sternum are far less evident. C, Radiograph of the left foot of the same patient at 12 yr of age, displaying endobone formation in the heel.
PCR amplification of microsatellite repeat markers DNA extraction from whole blood was performed as described previously (11–15). PCR primer pairs for the microsatellite repeat markers (Research Genetics, Inc.; Huntsville, AL) were used to amplify 30 ng of genomic DNA. The PCR products were 32P radiolabeled by 59 labeling of the forward primer with T4 polynucleotide kinase (Gibco BRL; Gaithersburg, MD) and g[32P]ATP (DuPont NEN; Boston, MA). The PCR products were resolved by electrophoresis on standard acrylamide sequencing gels and visualized by autoradiography. The sequences for all primers are available from the Centre d’Etude du Polymorphisme Humain (CEPH) marker database (http://www.cephb.fr/cephdb/). Paternity was verified by examining an additional 19 highly polymorphic microsatellite markers on other chromosomes, in addition to the 6 markers evaluated within the ADO region on chromosome 1.
Linkage analysis In accordance with the previously published linkage study (10), an autosomal dominant disease model with 60% penetrance was used for all linkage analyses with a population disease allele frequency of 1 per 100,000 chromosomes. Maximum likelihood estimates of the population marker allele frequencies were obtained from the observed pedigree data with the USERM13 subroutine of the MENDEL (Ann Arbor, MI) computer package (14). USERM13 was also used to calculate polymorphic information content (PIC) for each marker. Two-point LOD scores were computed using the program MLINK from the FASTLINK implementation of the LINKAGE (New York, NY) package of programs (15, 16). Order and distances for all markers except AMY2B were taken from the Marshfield Center for Medical Genetics sex-averaged map of chromosome 1 (http://www.marshmed.org/genetics) (17). The interval flanked by the markers D1S486 and D1S221 included the entire 1p21 region from a previous linkage study (10), according the Marshfield Center map. Genetic mapping data from the Genetic Location Database (http://cedar. genetics.soton.ac.uk/public html) (18) placed marker AMY2B in the interval between D1S495 and D1S239; for lack of additional information, AMY2B was assumed to lie halfway between these two markers. No recombination was observed between D1S239 and D1S248 in the CEPH families used for mapping by Marshfield; therefore, a short arbitrary distance between them (0.1 cM) was assumed. The final chromosome 1p21 genetic map was tel-D1S486 —2.68 —D1S495 —1.07 —AMY2B
—1.07 —D1S239 — 0.10 —D1S248 —3.22 —D1S221-cen (distances in cM). The computer program VITESSE (Pittsburgh, PA) (19) was used for multipoint linkage calculations. Multipoint LOD scores were computed at each marker position and at five equally-spaced points in each interval, using information from all six markers for each calculation.
Results Radiographic evidence of ADO in the Indiana families
Radiographs from members of family EOP1 displayed typical features of ADO (9). Representative radiographs from the proband of family EOP3, an 18-yr-old female, are shown in Fig. 1, A, B, and C. Upon examination of a lateral chest X-ray, vertebral end-plate sclerosis as well as endobone formation in the spine was readily apparent at age 15 yr (Fig. 1A). In addition, marked endobone formation was evident in her sternum (Fig. 1A). Of note, lateral chest X-rays taken at 3 yr of age showed slight vertebral sclerosis that was not as obvious as in the film taken at age 15 (Fig. 1B). Left foot x-rays taken at 12 yr showed an onion-skin like appearance of endobone formation in the heel (Fig. 1C). Linkage analysis with 1p21 microsatellite repeat markers
To test our ADO families for linkage to chromosome 1p21, we used 6 dinucleotide microsatellite repeat markers (see Materials and Methods) shown to localize to the 8.5 cM ADO region defined in the Danish kindred (10). Two-point LOD scores did not support linkage with any of the 6 chromosome 1p21 markers (Table 1). More importantly, multipoint linkage analysis of the region, defined by the telomeric marker D1S486 and by the centromeric marker D1S221, produced LOD scores of less than 22.0 (at u 5 0) for the individual families and, when combined, yielded LOD scores below 27.0 throughout the entire region (Fig. 2). These LOD scores
MULTIPLE ADO LOCI
surpass the traditional criteria for exclusion of linkage (LOD of , 22.0, or 100:1 odds against linkage) by a factor of 105. In addition, the scores calculated were stable across orderof-magnitude changes in the disease allele frequency estimate, as well as changes in the penetrance function (not shown). These results demonstrated that the two Indiana ADO families do not link to the previously described ADO region on chromosome 1p21. ADO haplotype analysis
Chromosome 1 haplotypes were determined across the ADO interval and are shown in Fig. 3. Visual inspection of the markers revealed no haplotype segregating with the disorder in either family. Of note, sisters II:1 and II:3 in family EOP1 each carry a different 1p21 haplotype from the diseasetransmitting parent (I:1) (Fig. 3A). Furthermore, two brothers from family EOP3, individuals II:1 and II:3, who both must carry the ADO haplotype because they each have affected children, received completely different 1p21 chromosomal regions from their parents, individuals I:1 and I:2 (Fig. 3B). The origin of the ADO mutation, however, is unknown in this family as both I:1 and I:2 have insufficient radiographic evidence for diagnosis, and neither has a family history of ADO. These haplotypes provide further evidence that the ADO gene in these families is not localized between markers D1S495 and D1S221. In sum, our results demonstrate that ADO displays locus heterogeneity; therefore mutations in at least two different genes, one at the 1p21 locus and one at TABLE 1. Combined two-point LOD scores for families EOP1 and EOP3 for 1p21 markers Marker
PIC
u 5 0.0
u 5 0.01
u 5 0.05
u 5 0.1
u 5 0.2
D1S486 D1S495 AMY2B D1S239 D1S248 D1S221
0.31 0.83 0.46 0.63 0.62 0.68
24.39 217.12 212.58 213.44 24.46 28.84
21.40 27.21 24.56 24.23 21.42 24.58
20.72 23.85 22.37 22.19 20.74 22.45
20.44 22.43 21.44 21.37 20.46 21.54
20.19 21.11 20.60 20.63 20.20 20.69
u, Recombination fraction.
FIG. 2. Multipoint LOD score analysis of the ADO region. LOD score is shown on the Y-axis, map position on the Xaxis (in cM). The results for individual families EOP1 (dashed line) and EOP3 (hairline), as well as the combined multipoint analyses (heavy line), are shown. The standard LOD of 22.0 used to exclude linkage is present as a dotted line.
1049
another unmapped locus, can give rise to clinically indistinguishable ADO phenotypes. Discussion
Identifying genes that are responsible for osteosclerotic disorders provides an opportunity to discover novel proteins that regulate bone structure and function (20). The first step in a positional cloning strategy is to identify a region of the genome that contains the disease gene locus, and the second step is to identify potential candidate genes within the linked region. The characterization of natural or targeted gene knockout animals has led to a large number of models for the recessive forms of osteopetrosis (21–24). These animal models indicate that the resorptive capabilities of osteoclasts are regulated through multiple pathways and support the idea that aberrations in several genes could give rise to osteopetrosis. The linkage results and haplotype analysis (Table 1, Figs. 2 and 3) exclude mutations in the ADO gene locus on chromosome 1p21 as a cause of ADO in our families. Therefore our results indicate that ADO is caused by mutations in at least two different genes. The phenotype in the Indiana and Danish families is remarkably similar. Although speculative, the clinical similarity between the families supports the possibility that the ADO gene on chromosome 1p21 and the as yet unmapped ADO gene locus responsible for the disease in our kindreds may interact as part of a complex or may be part of the same biochemical pathway. Alternatively, it is plausible that one gene could code for a circulating hormone or factor and the other gene code for its receptor. Mutations in either locus may lead to decreased effective circulating concentration of the hormone or may cause insufficient target receptor number and thereby result in ADO. Another dominant disorder, autosomal dominant polycystic kidney disease (ADPKD) is caused by mutations in several distinct but homologous genes (25). By analogy, it is possible that the two ADO genes possess similar functions that regulate the resorptive activities of osteoclasts. If the two genes are related through function or primary structure, then isolation of the
1050
WHITE ET AL.
JCE & M • 1999 Vol 84 • No 3
FIG. 3. ADO pedigrees with haplotypes. Filled symbols indicate affected individuals; open symbols, unaffected or unknown. Circles represent females; squares, males. “C” within a symbol indicates an obligate carrier. Marker haplotypes are listed below each individual. Question marks (?) indicate an unknown genotype, and brackets indicate an inferred genotype. A, Family EOP1. B, Family EOP3. Phenotypic data for individual II:3 (*) is incomplete to definitively conclude whether he is a carrier or is mildly affected.
MULTIPLE ADO LOCI
ADO gene at one locus could potentially allow rapid identification of the second ADO gene. The variable penetrance of ADO (4) remains an interesting facet in the genetic and clinical investigation of the disorder. There are numerous examples of both asymptomatic gene carriers and severely affected individuals within the same family (3, 4, 9). The determination that multiple genes for ADO exist may help to shed light on the issue of clinical variability among family members who presumably carry the same disease allele. It is possible that if the ADO genes at both loci interact, functional polymorphisms at the nonmutated locus may influence the presence and/or severity of disease in individuals who have a mutant allele at the other locus. Although speculative at the present time, this idea may be addressed after the isolation of the ADO genes from both loci. In summary, we demonstrate in this report that locus heterogeneity exists for ADO; therefore, alterations in at least two genes can give rise to the same phenotype. Elucidation of both ADO genes will provide important insight into the pathogenesis of ADO and will improve our understanding of the regulation of the homeostatic mechanisms involved in osteoclast function and skeletal resorption. Cloning the genes may also reveal potential therapeutic targets and thus help to improve the treatment of ADO patients. References 1. Albers-Schonberg H. 1904 Rontgenbilder einer seltenen Knochenerkrankung. Muenchener Med Wschr. 51:365. 2. Bollerslev J. 1995 Autosomal dominant osteopetrosis: bone metabolism and epidemiological, clinical, and hormonal aspects: update 1995. Endocr Rev. 4:365–373. 3. Johnston Jr CC, Lavy N, Lord T, Vellios F, Merritt AD, Deiss Jr WP. 1968 Osteopetrosis. A clinical, genetic, metabolic, and morphologic study of the dominantly inherited, benign form. Medicine. 47:149 –167. 4. Bollerslev J. 1987 Osteopetrosis. A genetic and epidemiological study. Clin Genet. 31:86 –90. 5. El-Tawil T, Stoker DJ. 1993 Benign osteopetrosis: a review of 42 cases showing two different patterns. Skeletal Radiol. 22:587–593. 6. Bollerslev J, Mosekilde L. 1993 Autosomal dominant osteopetrosis. Clinical Orthopaedics and Related Research. 294:45–51.
1051
7. Whyte MP, Chines A, Silva Jr DP, Landt Y, Ladenson JH. 1998 Creatine kinase brain isoenzyme (BB-CK) presence in serum distinguishes osteopetrosis among the sclerosing bone disorders. J Bone Miner Res. 11:1438 –1443. 8. Yoneyama T, Fowler HL, Pendleton JW, et al. 1992 Elevated serum levels of creatine kinase BB in autosomal dominant osteopetrosis. Clin Genet. 42:39 – 42. 9. Takacs I, Cooper H, Weaver D, Econs MJ. 1998 Bone mineral density and laboratory evaluation of a Type II autosomal dominant osteopetrosis carrier. J Med Genet. In press. 10. Van Hul W, Bollerslev J, Gram J, et al. 1997 Localization of a gene for autosomal dominant osteopetrosis (Albers-Schonberg Disease) to chromosome 1p21. Am J Hum Genet. 61:363–369. 11. Econs MJ, Barker DF, Speer MC, Pericak-Vance MA, Fain PR, Drezner MK. 1992 Multilocus mapping of the X-linked hypophosphatemic rickets gene. J Clin Endocrinol Metab. 75:201–206. 12. Econs MJ, Pericak-Vance MA, Betz H, Bartlett RJ, Speer MC, Drezner MK. 1990 The human glycine receptor: a new probe that is linked to the X-linked hypophosphatemic rickets gene. Genomics. 7:439 – 441. 13. Econs MJ, Fain PR, Norman M, et al. 1993 Flanking markers define the X-linked hypophosphatemic rickets gene locus. J Bone Miner Res. 8:1149 –1152. 14. Boehnke M. 1991 Allele frequency estimation from data on relatives. Am J Hum Genet. 48:22–25. 15. Schaffer A, Gupta S, Shriram K, Cottingham R. 1994 Avoiding recomputation in linkage analysis. Hum Hered. 44:225–237. 16. Lathrop GM, Lalouel JM. 1984 Easy calculations of LOD scores and genetic risks on small computers. Am J Hum Genet. 36:460 – 465. 17. Broman KW, Murray JC, Sheffield VC, White RL, Weber JL. 1998 Comprehensive human genetic maps: individual and sex-specific variation in recombination. Am J Hum Genet. 63:861– 869. 18. Collins A, Frezal J, Teague J, Morton NE. 1996 A metric map of humans: 23,500 loci in 850 bands. Proc Natl Acad Sci USA. 93:14771–14775. 19. O’Connell JR, Weeks DE. 1995 The VITESSE algorithm for rapid exact multilocus linkage analysis via genotype set-recoding and fuzzy inheritance. Nat Genet. 11:402– 408. 20. Whyte MP. 1997 Searching for gene defects that cause high bone mass. Am J Hum Genet. 60:1309 –1311. 21. Yoshida H, Hayashi S, Kunisada T, et al. 1990 The murine mutation osteopetrosis is in the coding region of the macrophage colony stimulating factor gene. Nature. 345:442– 444. 22. Soriano P, Montgomery C, Geske R, Bradley A. 1991 Targeted disruption of the c-src proto-oncogene leads to osteopetrosis in mice. Cell. 64:693–702. 23. Wang Z-Q, Ovitt C, Grigoriadis AE, Mohle-Steinlein U, Ruther U, Wagner EF. 1992 Bone and haemotopoitic defects in mice lacking c-fos. Nature. 360:741–745. 24. Hodgkinson CA, Moore KJ, Nakayama A, et al. 1993 Mutations at the mouse microphthalmia locus are associated with defects in a gene encoding a novel basic helix-loop-helix-zipper protein. Cell. 74:395– 404. 25. Calvet JP. 1998 Molecular genetics of polycystic kidney disease. J Nephrol. 11:24 –34.
