BoneKEy-Osteovision | Meeting Reports

Genetics of human bone diseases. Meeting report from the 29th Annual Meeting of the American Society for Bone and Mineral ResearchSeptember 16-19, 2007 in Honolulu, Hawaii, USA

Serge Ferrari


The sessions dedicated to the genetics of bone diseases started with an outstanding symposium entitled “Novel Insights into Bone Metabolism through Genetics.” M. Whyte first reviewed disorders associated with mutations in the RANKL/OPG-RANK system. Juvenile Paget's Disease, JPD (also known as familial hyperphosphatasia, OMIM No #239000), is a recessive disorder caused by OPG (TNFRSF11B gene) null mutations. The skeletal deformities and high bone turnover characteristic of this disorder may be complicated by vascular calcifications, as seen in the OPG null mouse, manifest as retinal exudate and bleeding. As an alternative to bisphosphonates, recombinant OPG is a promising new therapy for these patients. In a few JPD cases, however, OPG mutations were not found. RANK (TNFRSF11A gene) mutations that affect the signal peptide region, resulting in increased signaling (activating mutations), cause familial expansive osteolysis, FEO (OMIM No #174810). Other constitutive RANK mutations were found in some rare cases of early-onset Paget's disease, PDB2 (OMIM No #602080). Most recently, Sobacchi et al. described RANKL (TNFSF11 gene) loss-of-function mutations in an osteoclast-poor form of autosomal recessive osteopetrosis, OPTB2 (OMIM No #259710, see Not To Be Missed, BoneKEy, 2007 Aug;4(8):204–8). Of note, an abstract at the ASBMR meeting () reported no differences in RANKL gene polymorphisms between subjects with/without sporadic Paget's disease, whereas OPG polymorphisms were reported to be associated with Paget's in this group.

E. Shore reviewed her own findings concerning the causes of fibrodysplasia ossificans progressiva, FOP (OMIM No #135100), and progressive osseous heteroplasia, POH (OMIM No #166350). These two disorders are characterized by new bone formation independent of the skeleton, i.e., new bone at the wrong place and at the wrong time: endochondral ossification in FOP, intramembranous ossification that occurs in the skin in POH. They are autosomal dominant disorders with a frequency of about 1/million individuals that in the case of FOP can be inherited from either the father or the mother, whereas POH is inherited exclusively from the father. From a phenotypic point of view, these two disorders differ by the presence of big toe malformation in FOP but not in POH. Linkage approaches revealed that FOP is due to mutations in the type 1 BMP receptor (ACVR1 gene, see S. Ferrari, BoneKEy, 2006 December;3(12):11–29). In contrast, linkage approaches have been difficult in POH due to the small number of known cases (about 50). The responsible gene has nevertheless been identified by a candidate gene approach based on the phenotypic similarities that may exist between POH, Albright's hereditary osteodystrophy, AHO, and pseudohypoparathyroidism PHP1a. In two-thirds of POH cases, this revealed heterozygous mutations of the G protein α s (GNAS) allele that is paternally-derived (XL-α-s protein). However, how this mutation and/or paternal imprinting explains the restricted expression of the mutant protein in the skin remains unclear.

J. Marini reviewed autosomal recessive, mostly lethal forms of osteogenesis imperfecta (OI) characterized by white (normal) sclerae, small head circumference, and a round face. In 10 to 15% of OI cases, Col1A1 and Col1A2 mutations are not found. In OI type VII and IIB, for instance, (OMIM No #610682 and #610854), homozygous and compound mutations were found in a cartilage-associated protein (CRTAP gene) that shares 53% homology and forms a complex with prolyl 3-hydroxylase 1 (P3H1 gene, also known as leprecan, LEPRE1 gene). This complex is responsible for hydroxylation of the pro986 residue of Col1A1 necessary for the assembly of the triple helix. A mouse model of the CTRAP null mouse published by Morello et al. () demonstrates that this protein is expressed in chondrocytes at the chondro-osseous junction. Interestingly hypomorphic (non-lethal) mutations do exist in both CRTAP and LEPRE1, the latter found in families of African origin with OI type VIII.

OI cases with unusual presentations were reported by a group in Sweden (), the most interesting being three girls with multiple fractures but high BMD (+3SD) by DXA (“dense OI” according to the authors) and normal bone turnover markers. In all three cases, novel Col1A1 or Col1A2 mutations were found, in particular at the splice site of the C-terminal peptide. The major limitation of this work was the fact that the pathogenesis of the identified mutations was not directly demonstrated whereas concomitant (HBM) mutations in other genes were not ruled out. Also related to OI was a very important communication () reviewing experience with IV pamidonate in children with OI in relation to the potential risk of osteonecrosis of the jaw (ONJ). As a reminder, dentinogenesis imperfecta is also often present in these children who therefore require a number of dental procedures during growth. Despite follow-ups of up to 10 years in some children, extraction sockets were never found to be complicated by ONJ.

A group in France () reported its experience with the highly variable response to treatment with active vitamin D (1α vitamin D, 1-2 microg/d) and phosphorus in a large series of patients (n=81) with X-linked hypophosphatemic rickets (XLH). Their results suggested that genetic variation in the vitamin D receptor (VDR), in particular a haplotype that they identified as Hap1, influenced the height and Ca-Pi metabolic response to treatment. Dr. M. Whyte, however, pointed out that the outcome of XLH is primarily influenced by the type of PHEX mutations that are present, i.e., deactivating vs. non-deactivating mutations.

Moving from monogenic disorders to osteoporosis genetics, LRP5 gene polymorphisms previously described to be associated with BMD and/or fractures in several independent cohorts were analyzed in relation to these phenotypes in an extraordinarily large collection of participants (n=37,000) from both European and US cohorts (). This prospective meta-analysis at the participant level confirmed previous findings in that LRP5 missense SNPs in exons 9 and 18 were associated with spine and femur aBMD and fracture risk, with a maximal 26% increase in risk of vertebral fracture per 667M allele. Another abstract () reminded us that heterozygous carriers of LRP5 OPPG mutations have a lower BMD and a higher prevalence of vertebral fractures than controls, further suggesting that heterozygous LRP5 mutations should be suspected whenever osteoporosis is detected in middle-aged subjects. This study also indicated a higher prevalence of diabetes and hypercholesterolemia in these subjects, thereby providing a possible molecular clue to explain the association of these disorders with osteoporosis.

An interesting approach to predict osteoporotic fractures using Col1A1 polymorphisms was taken by Australian investigators (). They developed models that indicated the TT genotype of Col1A1 (5% of the population) to more than double the risk of any fracture at all ages independently of femur neck BMD, so that homozygozity for the T allele would be equivalent to +20 yrs of age or -1SD of BMD. A new candidate gene to be associated with BMD is FNLB, filamin B, mutations in which cause osteochondrodysplasia (). This gene is in a region previously mapped for linkage with osteoporosis (3p14) by one group in England. Now the authors report association of FNLB SNPs with BMD in two independent cohorts, i.e., they were able to replicate the association they found. Large-scale genotyping methods to simultaneously analyze thousands of SNPs were presented. By analyzing 113 SNPs in 54 selected genes with microarray-based techniques, sex-specific associations with BMD were reported, such as LRP5 and ESR2 in men, and IL6 and ESR1 in women (). A genome-wide association study (GWA) using the 100k SNP gene chip method found numerous SNPs associated with BMD and bone geometry at the hip (). However virtually no SNP shared association with BMD and bone geometric indices, such as femoral neck width, suggesting that these traits are determined by separate genes. A region-wide association study based on 200 tag SNPs covering 6Mb on chromosome 3p21 identified a voltage-dependent calcium channel subunit, CACNA2D2, to be associated with BMD in Chinese subjects ().

Negative reports are sometimes also useful. By studying genetic variation in the PTH system (i.e., PTH, PTHrP, PTH1R and PTH2R genes) in a prospective study of 1000 elderly women, OPRA investigators found no association with BMD, fractures and/or PTH levels, except between some PTH haplotypes and fracture risk ().

A major component of skeletal strength, besides bone mineral density, is the microarchitecture of cortical and trabecular bone. Using high-resolution pQCT, the heritability (h2, %) of human bone microstructure was evaluated at the distal radius and tibia in more than 100 mother-daughter pairs (). H2 was generally high and similar to h2 for aBMD. However, some heritability estimates decreased sharply past the menopause, suggesting that intense bone remodeling may overcome additive genetic effects on peak bone microstructure. Also, interestingly, the mean heritability of trabecular and cortical microarchitecture was near 50% once adjusted for body size and BMD at the same site, adding to the evidence that genetic effects may specifically affect bone structure.

Although reviewing the plethora of abstracts on mouse genetics is beyond the scope of this summary, it is worth mentioning an interesting approach that used the Haplotype Association Mapping (HAM) program to extract information from SNPs and BMD in 18-week female mice from 30 mouse strains (). This technique identified a positional candidate gene, CER1, in which a non-synonymous SNP was later found to be associated with low BMD in 1000 Chinese premenopausal women as well. Finally, one of the most interesting abstracts in this field () identified a highly polymorphic region in the 3′-UTR of the Pparγ gene by comparing sequences from B6 and C3H mice. Studies in congenic mice revealed that this region regulates the Pparγ transcription level in response to a high fat diet. Moreover, SNPs in the human synthenic region of mouse Ppar 3′-UTR were also reported to interact with dietary fat in association with BMD.

Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.