High resolution linkage and linkage disequilibrium analyses of chromosome 1p36 SNPs identify new positional candidate genes for low bone mineral density.

UNLABELLED: A quantitative trait locus (QTL) for BMD maps to chromosome 1p36. We have analyzed a high density SNP panel from this region for linkage and association to BMD in 39 osteoporosis pedigrees. Our results support the presence of genes controlling BMD on 1p36 and indicate new candidates for further analyses. INTRODUCTION: Low BMD is one of the major risk factors for osteoporosis. Following a genome scan in a sample of Caucasian families recruited through probands with low BMD, a region on 1p36 near marker D1S214 received support as a QTL for BMD from linkage (maximum lod-score = 2.87) and linkage disequilibrium (LD) analysis (p < 0.01). METHODS: To better characterize the genetic risk factors for low BMD located in this genomic region, we have genotyped the same group of families for 1095 SNPs located across 11 Mb on 1p36. Linkage and LD analyses have been performed using the variance component approach. RESULTS: Multivariate linkage analysis indicated two QTLs for femoral neck BMD, lumbar spine BMD and trochanter BMD simultaneously on 1p36, with maximum lod-scores of 4.37 at 12 cM and 3.59 at 22 cM. LD analysis identified several SNPs potentially associated with BMD, including the RERE gene SNP rs11121179 (p = 0.000005 for lumbar spine BMD). Other candidate genes include G1P2, SSU72 and CCDC27 (each containing 1 SNP with p < 0.001 for at least one BMD trait). CONCLUSIONS: This study supports the presence in 1p36 of QTLs affecting BMD at multiple skeletal sites. Replication of our results in other independent cohorts is warranted.

Association analysis of bone mineral density and single nucleotide polymorphisms in two candidate genes on chromosome 1p36.

Two candidate genes for bone mineral density (BMD), tumor necrosis factor alpha receptor 2 (TNFRSF1B) and lysyl hydroxylase (PLOD1), have been scanned for single nucleotide polymorphisms (SNPs) within their coding and promoter regions. These two genes, separated by about 200 kb, are located within the chromosomal interval 1p36.2-1p36.3 that has been linked to femoral neck BMD. In a patient population (n = 104) of European descent, there were four SNPs within TNFRSF1B and six SNPs within PLOD1 that occurred with greater than 5% frequency. There was significant linkage disequilibrium within both genes. Single marker analysis revealed significant association for one SNP located in intron 6 of PLOD1 and lumbar spine BMD (P = 0.01). Allelic haplotypes that encompassed the four SNPs in TNFRSF1B or the six SNPs in PLOD1 were assigned using a Bayesian algorithm as implemented in the program Haplotyper. Association of TNFRSF1B haplotypes with femoral neck BMD was statistically significant (P = 0.01). Similarly, PLOD1 haplotypes demonstrated a statistically significant association with spinal BMD (P = 0.04). These findings strengthen the potential importance of chromosome 1p36.2-1p36.3 in contributing to BMD variation, and are consistent with genetic variation in either PLOD1, TNFRSF1B or nearby genes playing a role in the phenotype.

The leatherback turtle, Dermochelys coriacea, exhibits both polyandry and polygyny.

The leatherback turtle (Dermochelys coriacea) is an endangered species, and world-wide populations are declining. To understand better the mating structure of this pelagic and fragile species, we investigated paternity in nearly 1000 hatchlings from Playa Grande in Parque Marino Nacional Las Baulas, Costa Rica. We collected DNA samples from 36 adult female leatherbacks and assessed allele frequency distributions for three microsatellite loci. For 20 of these 36 females, we examined DNA from hatchlings representing multiple clutches, and in some cases assessed up to four successive clutches from the same female. We inferred paternal alleles by comparing maternal and hatchling genotypes. We could not reject the null hypothesis of single paternity in 12 of 20 families (31 of 50 clutches), but we did reject the null hypothesis in two families (eight of 50 clutches). In the remaining six families, the null hypothesis could not be accepted or rejected with certainty because the number of hatchlings exhibiting extra nonmaternal alleles was small, and could thus be a result of mutation or sample error. Successive clutches laid by the same female had the same paternal allelic contribution, indicating sperm storage or possibly monogamy. None of 20 females shared the same three-locus genotype whereas there were two instances of shared genotypes among 17 inferred paternal three-locus genotypes. We conclude that both polyandry and polygyny are part of the mating structure of this leatherback sea turtle population.

Variance component linkage analysis indicates a QTL for femoral neck bone mineral density on chromosome 1p36.

Osteoporosis is a common condition characterized by reduced skeletal strength and increased susceptibility to fracture. Eight million Americans over the age of 50 have osteoporosis of the femoral neck. The most important risk factor for osteoporosis is low bone mineral density (BMD), and several epidemiological studies have shown the importance of genetic factors in determining variability of BMD. An initial genome screen in seven large pedigrees suggested that a candidate region conferring susceptibility to low BMD of the femoral neck was located on chromosome 1p36. We have now confirmed and extended this finding by analyzing nine microsatellite markers spanning a 40 cM interval across the candidate region in an expanded sample of 42 families. Heritability of femoral neck BMD was estimated as 0.51 +/- 0.13 in these families, after accounting for the effects of age, sex, body mass index, height and weight. Variance component linkage analysis yielded a maximum multipoint LOD score of 3.53 for linkage of femoral neck BMD to a quantitative trait locus (QTL) located near marker D1S214. The associated empirical P-value by simulation analysis was equal to 0.0001. The results strongly support the hypothesis that a major QTL controlling femoral neck BMD is located on chromosome 1p36.2-p36.3, and further analysis of candidate genes in this region is warranted.

Association of a polymorphism in the TNFR2 gene with low bone mineral density.

Previous genetic linkage data suggested that a gene on chromosome 1p36.2-36.3 might be linked to low bone mineral density (BMD). Here, we examine the gene for tumor necrosis factor receptor 2 (TNFR2), a candidate gene within that interval, for association with low BMD in a group of 159 unrelated individuals. We assess two polymorphic sites within the gene, a microsatellite repeat within intron 4, and a three-nucleotide variation in the 3' untranslated region (UTR) of the gene. The latter has five alleles of which the rarest allele is associated with low spinal BMD Z score (p = 0.008). Lowest mean spinal BMD Z scores were observed for individuals having genotypes that were heterozygous for the rarest allele. No homozygotes for the rarest allele were observed. Preliminary analysis suggests that there is a difference in the genotype frequency distribution between the group with low BMD and a control group.

Expression of a new RNA-splice isoform of WT1 in developing kidney-gonadal complexes of the turtle, Trachemys scripta.

WT1 is a tumor suppressor gene encoding a zinc finger DNA-binding protein required for normal vertebrate kidney and gonad development. Although the sequence and function of this gene has been studied mostly in mammals, comparative analysis in other vertebrates may suggest regions of conservation of function as well as evolution of function. We have initiated a study of this gene in the freshwater turtle, Trachemys scripta, a species that demonstrates temperature dependent sex determination. The turtle WT1 amino acid sequence (GenBank Accession No. AF019779) is over 85% identical to that of other species overall, but there are some major differences. The greatest differences are in the N-terminal portion of the peptide which is thought to mediate transcriptional repression by interaction with other proteins. Turtle WT1, like those of the alligator, chicken, and Xenopus lacks the proline- and glycine-rich stretches that are present in mammalian WT1. Exon 5, which is alternatively spliced in mammals, is altogether absent in the non-mammalian vertebrates. In addition, turtle WT1 is alternatively spliced so that exon 4 is either present or absent. These differences suggest that the interaction of reptilian WT1 with other factor required for mediation of activity may be different than the interaction of mammalian WT1. It also suggests that alternative splicing is a conserved regulatory mechanism of vertebrate WT1. Expression of WT1 in turtle embryonic kidney-gonadal complexes begins after the mesonephroi have formed and continues at least until the bipotential gonad begins to differentiate. Although the proportions of the different splice isoforms are relatively constant during these stages of kidney development, the level of steady state expression is increased in embryos incubated at 26 degrees C, the testis-producing temperature.

First-stage autosomal genome screen in extended pedigrees suggests genes predisposing to low bone mineral density on chromosomes 1p, 2p and 4q.

Osteoporosis is characterized by low bone density, and osteopenia is responsible for 1.5 million fractures in the United States annually. In order to identify regions of the genome which are likely to contain genes predisposing to osteopenia, we genotyped 149 members of seven large pedigrees having recurrence of low bone mineral density (BMD) with 330 DNA markers spread throughout the autosomal genome. Linkage analysis for this quantitative trait was carried out using spine and hip BMD values by the classical lod-score method using a genetic model with parameters estimated from the seven families. In addition, non-parametric analysis was performed using the traditional Haseman-Elston approach in 74 independent sib pairs from the same pedigrees. The maximum lod score obtained by parametric analysis in all families combined was +2.08 (theta = 0.05) for the marker CD3D on chromosome 11q. All other combined lod scores from the parametric analysis were less than +1.90, the threshold for suggestive linkage. Non-parametric analysis suggested linkage of low BMD to chromosomes 1p36 (Zmax = +3.51 for D1S450) and 2p23-24 (Zmax = +2.07 for D2S149). Maximum multi-point lod scores for these regions were +2.29 and +2.25, respectively. A third region with associated lod scores above the threshold of suggestive linkage in both single-point and multi-point non-parametric analysis was on chromosome 4qter (Zmax = +2.95 for D4S1539 and Zmax = +2.48 for D4S1554). Our data suggest the existence of multiple genes involved in controlling spine and hip BMD, and indicate several candidate regions for further screening in this and other independent samples.

Sequence and expression analysis of WT1 and Sox9 in the red-eared slider turtle, Trachemys scripta.

Temperature-dependent sex-determination (TSD) is a phenomenon that has been characterized at the ecological, morphological, and endocrinological levels in some reptilian species. We have begun to investigate TSD at the level of molecular development by cloning, sequencing, and analyzing the expression of two genes, WT1 and Sox9, in the red-eared slider turtle Trachemys scripta. We obtained almost full-length cDNA clones for WT1 and Sox9 that were greater than 73% identical to the human homologues at the nucleotide level. WT1 was expressed in urogenital tissue at all developmental stages examined (Yntema stages 12-20) at incubation temperatures that produce males (26 degrees C) or females (32 degrees C). Sox9 was also expressed throughout these same stages, but some differences were observed. At both 26 degrees C and 32 degrees C Sox9 was expressed in the mesonephroi and the undifferentiated gonads until Yntema stage 20, when only the gonad from the 26 degrees C embryos expressed a high level. In addition, there were two transcripts of Sox9 at all stages, but the relative proportion of the two transcripts differed at the two temperatures. Although the similarities in gene expression between a TSD species and other species with genotypically determined sex probably reflect the common features of organogenesis, differences may illustrate unique mechanisms for TSD.

Genetic linkage of familial granulomatous inflammatory arthritis, skin rash, and uveitis to chromosome 16.

Blau syndrome (MIM 186580), first described in a large, three-generation kindred, is an autosomal, dominantly inherited disease characterized by multiorgan, tissue-specific inflammation. Its clinical phenotype includes granulomatous arthritis, skin rash, and uveitis and probably represents a subtype of a group of clinical entities referred to as "familial granulomatosis." It is the sole human model with recognizably Mendelian inheritance for a variety of multisystem inflammatory diseases affecting a significant percentage of the population. A genomewide search for the Blau susceptibility locus was undertaken after karyotypic analysis revealed no abnormalities. Sixty-two of the 74-member pedigree were genotyped with dinucleotide-repeat markers. Linkage analysis was performed under a dominant model of inheritance with reduced penetrance. The marker D16S298 gave a maximum LOD score of 3.75 at theta = .04, with two-point analysis. LOD scores for flanking markers were consistent and placed the Blau susceptibility locus within the 16p12-q21 interval.

Vitamin D receptor genotype is not associated with bone mineral density in three ethnic/regional groups.

We report a cross-sectional study of 48 men, 56 premenopausal women, and 80 postmenopausal women who were of three ethnic/regional backgrounds: southern European (Greek, Italian), eastern European (Jewish, Polish, Hungarian), and western European (French, British). We determined bone mineral density (BMD) at four skeletal sites and assessed the vitamin D receptor (VDR) genotype by the Bsml restriction site polymorphism. Age and body mass index had significant effects on BMD by multiple regression analysis. In addition, ethnic/regional group had a significant effect on spinal BMD in premenopausal females (P = 0.014) and in males (P = 0.039). However, VDR genotype had no significant effect on BMD in any of the three study groups.

Osteopenia in 37 members of seven families: analysis based on a model of dominant inheritance.

BACKGROUND: The genetic factors involved in determining bone mineral density (BMD) have not been fully elucidated. We have begun genetic linkage analysis of seven families in which many members are osteopenic, in order to identify chromosomal loci that are potentially involved in determining BMD. MATERIALS AND METHODS: Spine BMD was measured in 143 members of seven kindred with familial osteopenia. The absolute BMD values for the spine (L2-L4) were converted to the age-, gender-, and weight-adjusted Z scores, and this corrected value was used as the quantitative trait on which to base subsequent genetic analyses. Simulations of linkage were performed in order to determine the information content of the pedigree set, and actual linkage analysis was conducted using polymorphic markers either within or near three candidate loci: COL1A1, COL1A2, and vitamin D receptor (VDR). RESULTS: The distribution of the corrected Z scores was bimodal (p = 0.001) suggesting a monogenic mode of inheritance of the low BMD trait. Simulation of linkage analysis suggested that the family data set was sufficient to detect linkage under a single major gene model. Actual linkage analysis did not support linkage to the three candidate loci. In addition, the VDR genotype was not statistically associated with low bone density at the spine. CONCLUSIONS: Loci other than COL1A1, COL1A2 and VDR are very likely responsible for the low BMD trait observed in these families. These families are suitable for a genome-wide screen using microsatellite repeats in order to identify the loci that are involved in osteopenia.

Substitution of glycine-661 by serine in the alpha1(I) and alpha2(I) chains of type I collagen results in different clinical and biochemical phenotypes.

We have characterised a point mutation causing the substitution of serine for glycine at position 661 of the alpha1(I) chain of type I collagen in a child with a severe form of osteogenesis imperfecta. An identical glycine substitution in the alpha2(I) chain was previously detected in a woman with post-menopausal osteoporosis. Two of her sons were heterozygous for the mutation and the third son was homozygous as a result of uniparental isodisomy. Biochemical profiles of the type I collagen heterotrimers were studied in each of the patients and compared with a control. Medium and cell-layer collagens were overmodified in all patients. Overmodification was obvious in the patient with the alpha 1(I) mutation but mild in the patients with the alpha 2(I) mutation, being slightly less evident in the heterozygote than in the homozygote. Investigation of the melting curves of the mutant collagen trimers in all three patients showed the same slight decrease in thermal stability and, hence, a lack of correlation with phenotypic severity. In contrast, the degree of overmodification of the collagen alpha chains was correlated with the phenotypic severity. The clinical observations in these patients illustrate the possibly predominant role of mutations in the collagen alpha1(I) chains over the same mutations in the alpha2(I) chains in determining the clinical outcome.

Molecular mechanisms of TSD in reptiles: a search for the magic bullet.

Significant progress has been made in understanding mechanisms of genetic sex determination. The ZFY gene encodes a zinc finger protein but is not the primary signal in sex determination. The SRY gene is the testis determining gene in man, mouse, rabbit, and probably marsupial mouse and wallaby. Temperature dependent sex determination probably involves a modification of development of the indifferent gonad due to differential expression of one or more specific DNA sequences whose behavior is controlled by some temperature sensitive process or to differential action of a gene product such as a protein. There are ZFY and SRY-like genes in reptiles. We cloned and sequenced a portion of the ZFY gene (Zft) from snapping turtle (Chelydra serpentina) that is found in both sexes. We cloned and sequenced portions of SRY-like genes (Sra for SRY-related-autosomal) from snapping turtle. Similar genes are found in alligator (Alligator mississippiensis) and lizards. Cladistic analysis suggests that there are two or three major families of SRY-like genes in vertebrates in addition to sex specific SRY genes located on the Y chromosome of eutherian and marsupial mammals. When placed on a phylogenetic tree these data indicate that Sras were present in early tetrapods. Sequestering of the SRY gene on the Y chromosome probably happened only once and this may have been the defining moment that set the mammalian line of Therapsid reptiles apart from other reptilian groups.

Mutation analysis of coding sequences for type I procollagen in individuals with low bone density.

Mutations in one of the two genes encoding type I procollagen (COL1A1 and COL1A2) are frequently the cause of osteogenesis imperfecta (OI), a disorder characterized by brittle bones. Here we tested whether patients with low bone density also have mutations in these genes. The 26 patients studied had no apparent metabolic bone disease, but most had a positive family history of osteopenia or osteoporosis. Although a diagnosis of OI was considered by the clinician in some cases, the clinical criteria for OI were not satisfied. Our strategy for mutation analysis consisted of PCR amplification of cDNA made to fibroblast mRNA using primers specific for the coding regions of COL1A1 and COL1A2. The PCR products were then sequenced directly with primers located within each PCR product. We found that 3 of 26 patients had mutations that altered the encoded amino acid. One mutation, at position alpha 2(I)-661 has been reported (Spotila et al. 1991 Proc Natl Acad Sci USA PNAS 88:5423). The other 2 patients, who were not related to each other, had a mutation that altered the proline codon at alpha 1(I)-27 to alanine. This mutation was not found in 81 normal individuals or in 37 additional osteopenic individuals. However, its effect on the biologic function of type I collagen, as well as its role in osteopenia, is uncertain. In addition to the two mutations, we found a polymorphism in codon alpha 2(I)-459. Although this polymorphism involved an amino acid substitution, it was present with equal frequency in the patient and the normal population. By analyzing this and previously reported neutral sequence variants in the COL1A2 gene, we determined that all patients expressed both alleles of the COL1A2 gene. The 12 patients who were heterozygous for a COL1A1 neutral sequence variant also expressed both alleles. Here we present all PCR primer and sequencing primer information. The results suggest that surveying a larger group of similarly selected individuals may reveal additional mutations in the COL1A1 or COL1A2 genes.

Self-assembly of collagen I from a proband homozygous for a mutation that substituted serine for glycine at position 661 in the alpha 2(I) chain. Possible relationship between the effects of mutations on critical concentration and the severity of the phenotype.

Procollagen I was isolated from cultured skin fibroblasts from a proband who was homozygous for a mutation in the COL1A2 gene that substituted a serine codon for a glycine codon at position 661 of the alpha 2(I) chain. The procollagen I was cleaved to pCcollagen I by procollagen N-proteinase and the pCcollagen I was used as a substrate for assay of self-assembly of collagen I into fibrils. The mutated pCcollagen I was cleaved to collagen I by procollagen C-proteinase at the same rate as control pCcollagen I. However, self-assembly of the mutated collagen I had a lag period that was 15-fold greater than the lag period observed with normal collagen I under the same conditions. Also, self-assembly of the mutated collagen I had a propagation rate of about one-fourth of the propagation rate of normal collagen I. In addition, the critical concentration for fibril assembly was slightly increased. Rotary shadowing electron microscopy of the mutated procollagen I did not reveal any increased flexibility of the triple helix as was seen previously with two mutated procollagens I in which there were substitutions of cysteine for glycine residues in the alpha 1(I) chain (Vogel, B. E., Doelz, R., Kadler, K. E., Hojima, Y., Engel, J., and Prockop, D. J. (1988) J. Biol. Chem. 263, 19249-19255; Lightfoot, S. J., Holmes, D. F., Brass, A., Grant, M. E., Byers, P. H., and Kadler, K. E. (1992) J. Biol. Chem. 267, 25521-25528). However, morphometric analysis by dark-field light microscopy and electron microscopy showed that the fibrils formed from the mutated collagen I appeared thicker in diameter than the fibrils formed from the normal collagen I. Comparison of the results with similar data on four mutated procollagens previously studied raised the possibility that mutations which markedly increase the critical concentration of fibril assembly produce more severe phenotypes than mutations which change other parameters of fibril assembly.

Sequence analysis of the ZFY and Sox genes in the turtle, Chelydra serpentina.

We have sequenced regions of the ZFY and Sox genes in the turtle Chelydra serpentina, a reptile with temperature-dependent sex determination. The ZFY gene in mammals encodes a transcription factor with multiple zinc fingers that may be involved in spermatogenesis as well as other processes. The turtle homologue, Zft, is 92% identical to the ZFY gene at the nucleotide and amino acid levels in the region of zinc fingers 7-12. There are several Sox genes in the turtle that are only 57-70% identical at the nucleotide level and about 55% identical at the amino acid level to the human sex-determining SRY gene. However, the turtle Sox genes, termed TSox, have the conserved motif called the HMG-box (for high mobility group DNA-binding protein) that defines a probable DNA-binding region, and thus are in the same gene family as the Sox genes of other organisms from Drosophila to man. One TSox sequence is identical at the amino acid level to a sequence found in birds, and is 98% identical to a sequence encoded autosomally in mouse and in man. The extent of sequence conservation among the Sox genes suggests that some of their functions may be conserved. Phylogenetic analysis of available Sox sequences including SRY (Sry) sequences suggests that there was a high degree of divergence between any possible immediate common ancestor of the turtle Sox sequences and the SRY (Sry) sequences.

Partial isodisomy for maternal chromosome 7 and short stature in an individual with a mutation at the COL1A2 locus.

Uniparental disomy for chromosome 7 has been described previously in two individuals with cystic fibrosis. Here, we describe a third case that was discovered because the proband was homozygous for a mutation in the COL1A2 gene for type I procollagen, although his mother was heterozygous and his father did not have the mutation. Phenotypically, the proband was similar to the two previously reported cases with uniparental disomy for chromosome 7, in that he was short in stature and growth retarded. Paternity was assessed with five polymorphic markers. Chromosome 7 inheritance in the proband was analyzed using 12 polymorphic markers distributed along the entire chromosome. Similar analysis of the proband's two brothers established the phase of the alleles at the various loci, assuming minimal recombination. The proband inherited only maternal alleles at five loci and was homozygous at all loci examined, except one. He was heterozygous for an RFLP at the IGBP-1 locus at 7p13-p12. The results suggest that the isodisomy was not complete because of a recombination event involving the proximal short arms of two maternal chromosomes. In addition, the phenotype of proportional dwarfism in the proband suggests imprinting of one or more growth-related genes on chromosome 7.

Mutation in a gene for type I procollagen (COL1A2) in a woman with postmenopausal osteoporosis: evidence for phenotypic and genotypic overlap with mild osteogenesis imperfecta.

Mutations in the two genes for type I collagen (COL1A1 or COL1A2) cause osteogenesis imperfecta (OI), a heritable disease characterized by moderate to extreme brittleness of bone early in life. Here we show that a 52-year-old postmenopausal woman with severe osteopenia and a compression fracture of a thoracic vertebra had a mutation in the gene for the alpha 2(I) chain of type I collagen (COL1A2) similar to mutations that cause OI. cDNA was prepared from the woman's skin fibroblast RNA and assayed for the presence of a mutation by treating DNA heteroduplexes with carbodiimide. The results indicated a sequence variation in the region encoding amino acid residues 660-667 of the alpha 2(I) chain. Further analysis demonstrated a single-base mutation that caused a serine-for-glycine substitution at position 661 of the alpha 2(I) triple-helical domain. The substitution produced posttranslational overmodification of the collagen triple helix, as is seen with most glycine substitutions that cause OI. The patient had a history of five previous fractures, slightly blue sclerae, and slight hearing loss. Therefore, the results suggest that there may be phenotypic and genotypic overlap between mild osteogenesis imperfecta and postmenopausal osteoporosis, and that a subset of women with postmenopausal osteoporosis may have mutations in the genes for type I procollagen.