Genetic Association of MMP10, MMP14, and MMP16 with Dental Caries.

Matrix metalloproteinases (MMPs), which degrade extracellular proteins as part of a variety of physiological processes, and their inhibitors have been implicated in the dental caries process. Here we investigated 28 genetic variants spanning the MMP10, MMP14, and MMP16 genes to detect association with dental caries experience in 13 age- and race-stratified (n = 3,587) samples from 6 parent studies. Analyses were performed separately for each sample, and results were combined across samples by meta-analysis. Two SNPs (rs2046315 and rs10429371) upstream of MMP16 were significantly associated with caries in an individual sample of white adults and via meta-analysis across 8 adult samples after gene-wise adjustment for multiple comparisons. Noteworthy is SNP rs2046315 (p = 8.14 × 10-8) association with caries in white adults. This SNP was originally nominated in a genome-wide-association study (GWAS) of dental caries in a sample of white adults and yielded associations in a subsequent GWAS of surface level caries in white adults as well. Therefore, in our study, we were able to recapture the association between rs2046315 and dental caries in white adults. Although we did not strengthen evidence that MMPs 10, 14, and 16 influence caries risk, MMP16 is still a likely candidate gene to pursue.

Genetic Association of MPPED2 and ACTN2 with Dental Caries.

The first genome-wide association study of dental caries focused on primary teeth in children aged 3 to 12 yr and nominated several novel genes: ACTN2, EDARADD, EPHA7, LPO, MPPED2, MTR, and ZMPSTE24. Here we interrogated 156 single-nucleotide polymorphisms (SNPs) within these candidate genes for evidence of association with dental caries experience in 13 race- and age-stratified samples from 6 independent studies (n = 3600). Analysis was performed separately for each sample, and results were combined across samples via meta-analysis. MPPED2 was significantly associated with caries via meta-analysis across the 5 childhood samples, with 4 SNPs showing significant associations after gene-wise adjustment for multiple comparisons (p < .0026). These results corroborate the previous genome-wide association study, although the functional role of MPPED2 in caries etiology remains unknown. ACTN2 also showed significant association via meta-analysis across childhood samples (p = .0014). Moreover, in adults, genetic association was observed for ACTN2 SNPs in individual samples (p < .0025), but no single SNP was significant via meta-analysis across all 8 adult samples. Given its compelling biological role in organizing ameloblasts during amelogenesis, this study strengthens the hypothesis that ACTN2 influences caries risk. Results for the other candidate genes neither proved nor precluded their associations with dental caries.

Genetic and environmental factors associated with dental caries in children: the Iowa Fluoride Study.

Dental caries remains the most common chronic childhood disease. Despite strong evidence of genetic components, there have been few studies of candidate genes and caries. In this analysis we tried to assess genetic and environmental factors contributing to childhood caries in the Iowa Fluoride Study. Environmental factors (age, sex, race, tooth-brushing frequencies and water fluoride level) and three dental caries scores (d(2)fs-total, d(2)fs-pit/fissure, and d(2)fs-smooth surface) were assessed in 575 unrelated children (mean age 5.2 years). Regression analyses were applied to assess environmental correlates. The Family-Based Association Test was used to test genetic associations for 23 single nucleotide polymorphism (SNP) markers in 7 caries candidate genes on 333 Caucasian parent-child trios. We evaluated the associations between caries status and the level of both single and multiple SNPs (haplotype) respectively. Permutation procedure was performed for correction of inflated type I errors due to multiple testing. Age, tooth-brushing frequency and water fluoride level were significantly correlated to at least one carious score. Caries on pit and fissure surfaces was substantially higher than on smooth surfaces (61 vs. 39%). SNPs in three genes (DSPP, KLK4 and AQP5) showed consistent associations with protection against caries. Of note, KLK4 and AQP5 were also highlighted by subsequent haplotype analysis. Our results support the concept that genes can modify the susceptibility of caries in children. Replication analysis in independent cohorts is highly needed in order to verify the validity of our findings.

Genome-wide association scan for childhood caries implicates novel genes.

Dental caries is the most common chronic disease in children and a major public health concern due to its increasing incidence, serious health and social co-morbidities, and socio-demographic disparities in disease burden. We performed the first genome-wide association scan for dental caries to identify associated genetic loci and nominate candidate genes affecting tooth decay in 1305 US children ages 3-12 yrs. Affection status was defined as 1 or more primary teeth with evidence of decay based on intra-oral examination. No associations met strict criteria for genome-wide significance (p < 10E-7); however, several loci (ACTN2, MTR, and EDARADD, MPPED2, and LPO) with plausible biological roles in dental caries exhibited suggestive evidence for association. Analyses stratified by home fluoride level yielded additional suggestive loci, including TFIP11 in the low-fluoride group, and EPHA7 and ZMPSTE24 in the sufficient-fluoride group. Suggestive loci were tested but not significantly replicated in an independent sample (N = 1695, ages 2-7 yrs) after adjustment for multiple comparisons. This study reinforces the complexity of dental caries, suggesting that numerous loci, mostly having small effects, are involved in cariogenesis. Verification/replication of suggestive loci may highlight biological mechanisms and/or pathways leading to a fuller understanding of the genetic risks for dental caries.

Analysis of multigenerational families with thoracic aortic aneurysms and dissections due to TGFBR1 or TGFBR2 mutations.

BACKGROUND: Mutations in the transforming growth factor beta receptor type I and II genes (TGFBR1 and TGFBR2) cause Loeys-Dietz syndrome (LDS), characterised by thoracic aortic aneurysms and dissections (TAAD), aneurysms and dissections of other arteries, craniosynostosis, cleft palate/bifid uvula, hypertelorism, congenital heart defects, arterial tortuosity, and mental retardation. TGFBR2 mutations can also cause TAAD in the absence of features of LDS in large multigenerational families, yet only sporadic LDS cases or parent-child pairs with TGFBR1 mutations have been reported to date. METHODS: The authors identified TGFBR1 missense mutations in multigenerational families with TAAD by DNA sequencing. Clinical features of affected individuals were assessed and compared with clinical features of previously described TGFBR2 families. RESULTS: Statistical analyses of the clinical features of the TGFBR1 cohort (n = 30) were compared with clinical features of TGFBR2 cohort (n = 77). Significant differences were identified in clinical presentation and survival based on gender in TGFBR1 families but not in TGFBR2 families. In families with TGFBR1 mutations, men died younger than women based on Kaplan-Meier survival curves. In addition, men presented with TAAD and women often presented with dissections and aneurysms of arteries other than the ascending thoracic aorta. The data also suggest that individuals with TGFBR2 mutations are more likely to dissect at aortic diameters <5.0 cm than individuals with TGFBR1 mutations. CONCLUSION: This study is the first to demonstrate clinical differences between patients with TGFBR1 and TGFBR2 mutations. These differences are important for the clinical management and outcome of vascular diseases in these patients.

Heterogeneous basis of the type VIB form of Ehlers-Danlos syndrome (EDS VIB) that is unrelated to decreased collagen lysyl hydroxylation.

Skin fibroblasts from the majority of patients with the clinical diagnosis of Ehlers-Danlos syndrome type VI (EDS VI; kyphoscoliosis type), have significantly decreased lysyl hydroxylase (LH) activity due to mutations in the LH1 gene (classified as EDS VIA: OMIM no. 225400). A rare condition exists in which patients are clinically similar but have normal levels of LH activity (designated EDS VIB: OMIM no. 229200). To define the biochemical defect, we have examined cultured fibroblasts from four EDS VIB patients for changes in the levels of the mRNAs for LH1, LH2, and LH3, collagen cross-linking patterns, and the extent of lysine hydroxylation of type I collagen alpha chains. Although normal levels of LH1 mRNA were observed in all four patients, in two patients the levels of LH2 mRNA were decreased by >50%, and a similar decrease was observed in LH3 mRNA in the other two patients. A distinct pattern of collagen cross-links, indicative of decreased lysyl hydroxylation, could be identified in EDS VIA patients, but there was no clear correlation between collagen cross-link pattern and changes in the individual LH mRNAs in EDS VIB patients. Linkage to tenascin-X was excluded in these patients. This study suggests that the basis for this form of EDS VI is genetically heterogeneous, and that alternative pathways in addition to lysine hydroxylation of collagen may be affected.

Deletions and duplications of Gly-Xaa-Yaa triplet repeats in the triple helical domains of type I collagen chains disrupt helix formation and result in several types of osteogenesis imperfecta.

Triple helix formation is a prerequisite for the passage of type I procollagen from the endoplasmic reticulum and secretion from the cell to form extracellular fibrils that will support mineral deposition in bone. Analysis of cDNA from 11 unrelated individuals with osteogenesis imperfecta (OI) revealed the presence of 11 novel, short in-frame deletions or duplications of three, nine, or 18 nucleotides in the helical coding regions of the COL1A1 and COL1A2 collagen genes. Triple helix formation was impaired, type I collagen alpha chains were post-translationally overmodified, and extracellular secretion was markedly reduced. With one exception, the obligate Gly-Xaa-Yaa repeat pattern of amino acids in the helical domains was not altered, but the Xaa- and Yaa position residues were out of register relative to the amino acid sequences of adjacent chains in the triple helix. Thus, the identity of these amino acids, in addition to third position glycines, is important for normal helix formation. These findings expand the known repertoire of uncommon in-frame deletions and duplications in OI, and provide insight into normal collagen biosynthesis and collagen triple helix formation.

Disruption of one intra-chain disulphide bond in the carboxyl-terminal propeptide of the proalpha1(I) chain of type I procollagen permits slow assembly and secretion of overmodified, but stable procollagen trimers and results in mild osteogenesis imperfecta.

Type I procollagen is a heterotrimer comprised of two proalpha1(I) chains and one proalpha2(I) chain. Chain recognition, association, and alignment of proalpha chains into correct registration are thought to occur through interactions between the C-terminal propeptide domains of the three chains. The C-propeptide of each chain contains a series of cysteine residues (eight in proalpha1(I) and seven in proalpha2(I)), the last four of which form intra-chain disulphide bonds. The remaining cysteine residues participate in inter-chain stabilisation. Because these residues are conserved, they are thought to be important for folding and assembly of procollagen. We identified a mutation (3897C-->G) that substituted tryptophan for the cysteine at position 1299 in proalpha1(I) (C1299W, the first cysteine that participates in intra-chain bonds) and resulted in mild osteogenesis imperfecta. The patient was born with a fractured clavicle and four rib fractures. By 18 months of age he had had no other fractures and was on the 50th centile for length and weight. The proband's mother, maternal aunt, and grandfather had the same mutation and had few fractures, white sclerae, and discoloured teeth, but their heights were within the normal range. In the patient's cells the defective chains remained as monomers for over 80 minutes (about four times normal) and were overmodified. Some secreted procollagens were also overmodified but had normal thermal stability, consistent with delayed, but normal helix formation. This intra-chain bond may stabilise the C-propeptide and promote rapid chain association. Other regions of the C-propeptide thus play more prominent roles in chain registration and triple helix nucleation.

Physical activity and bone measures in young children: the Iowa bone development study.

OBJECTIVES: Physical activity has a beneficial effect on bone development in circumpubertal children, although its effect on younger children is uncertain. In this cross-sectional study, we examined associations between physical activity and bone measures in 368 preschool children (mean age: 5.2 years, range: 4-6 years). DESIGN: Physical activity was measured using 4-day accelerometry readings, parental report of children's usual physical activity, and parental report of children's hours of daily television viewing. Total body and site-specific bone mineral content and area bone mineral density (BMD) were measured by dual energy radiograph absorptiometry. RESULTS: After adjustment for age and body size, accelerometry measures of physical activity and parental report of usual physical activity were consistently and positively associated with bone mineral content and BMD in both boys and girls (r = 0.15-0.28). Television viewing was inversely associated with hip BMD in girls (r = -0.15). The proportion of variance in bone measures explained by physical activity in linear regression models ranged from r(2) = 1.5% to 9.0%. In all of these models except total body BMD, at least 1 and often several of the physical activity variables entered as independent predictors. Activity variables most likely to enter the regression models were vigorous physical activity (as determined by accelerometry) and parental ranking of child's usual physical activity. CONCLUSIONS: Findings indicate that there are statistically significant and, perhaps important, associations between physical activity and bone measures during early childhood, well ahead of the onset of peak bone mass. This would suggest that intervention strategies to increase physical activity in young children could contribute to optimal bone development.

COL5A1 haploinsufficiency is a common molecular mechanism underlying the classical form of EDS.

We have identified haploinsufficiency of the COL5A1 gene that encodes the proalpha1(V) chain of type V collagen in the classical form of the Ehlers-Danlos syndrome (EDS), a heritable connective-tissue disorder that severely alters the collagen-fibrillar structure of the dermis, joints, eyes, and blood vessels. Eight of 28 probands with classical EDS who were heterozygous for expressed polymorphisms in COL5A1 showed complete or nearly complete loss of expression of one COL5A1 allele. Reduced levels of proalpha1(V) mRNA relative to the levels of another type V collagen mRNA, proalpha2(V), were also observed in the cultured fibroblasts from EDS probands. Products of the two COL5A1 alleles were approximately equal after the addition of cycloheximide to the fibroblast cultures. After harvesting of mRNAs from cycloheximide-treated cultured fibroblasts, heteroduplex analysis of overlapping reverse transcriptase-PCR segments spanning the complete proalpha1(V) cDNA showed anomalies in four of the eight probands that led to identification of causative mutations, and, in the remaining four probands, targeting of CGA-->TGA mutations in genomic DNA revealed a premature stop at codon in one of them. We estimate that approximately one-third of individuals with classical EDS have mutations of COL5A1 that result in haploinsufficiency. These findings indicate that the normal formation of the heterotypic collagen fibrils that contain types I, III, and V collagen requires the expression of both COL5A1 alleles.

Nonsense mutations in the COL1A1 gene preferentially reduce nuclear levels of mRNA but not hnRNA in osteogenesis imperfecta type I cell strains.

Osteogenesis imperfecta (OI) is a heterogeneous disorder of type I collagen resulting in varying degrees of severity. The mildest form of OI (Type I) is associated with bone fragility, normal or near normal stature and blue sclerae. All forms of OI are the result of mutations in COL1A1 or COL1A2, the genes that encode the proalpha1(I) and proalpha2(I) chains of type I collagen, respectively. Mutations identified in patients with OI type I lead to premature termination codons and allele-specific reductions of nuclear mRNA (termed nonsense-mediated mRNA decay or NMD), resulting in a COL1A1 null allele. In mammals, this process primarily effects RNA that co-purifies with the nuclear fraction of the cell. Using a semi-quantitative RT-PCR assay, we compare the relative amounts of normal and mutant transcripts in unprocessed hnRNA and mature mRNA isolated from the nuclear fraction of cells from 11 OI type I individuals with previously identified mutations distributed throughout the COL1A1 gene. While we detect about equal amounts of normal and mutant hnRNA from each cell strain, there is preferential reduction in the relative amount of mutant mRNA when compared to normal; only the cell strain with a mutation in the last exon escapes the major effects of NMD. Our data indicate that NMD targets mRNA rather than hnRNA for degradation, and that this occurs either during or after splicing but prior to cytoplasmic translation.

Osteocalcin: genetic and physical mapping of the human gene BGLAP and its potential role in postmenopausal osteoporosis.

Osteocalcin is an abundant, highly conserved bone-specific protein that is synthesized by osteoblasts. Temporally, osteocalcin appears in embryonic bone at the time of mineral deposition, where it binds to hydroxyapatite in a calcium-dependent manner. A role for osteocalcin in bone resorption has been suggested because of its ability to influence recruitment and differentiation of osteoclasts at the bone surface. The human osteocalcin gene has been mapped to 1q25-1q31 by somatic cell hybridization. In this paper, we refine both the genetic map and the physical map of osteocalcin and describe a new microsatellite (CA) marker, D1S3737, which is tightly linked to the gene. This marker and two other closely linked markers were used to identify alleles of the osteocalcin gene in case and control samples of postmenopausal white Iowans with low and high bone mineral density (BMD), respectively. A significant difference (P = 0.007) was observed between allele frequency distributions of case and control women with one of the markers, D1S3737. Further, logistic regression analysis determined one allele of D1S3737 as associated with BMD status in this population (P = 0.03). Our data suggest that genetic variation at the osteocalcin locus impacts BMD levels in the postmenopausal period and may predispose some women to osteoporosis.

Mutations of the alpha2(V) chain of type V collagen impair matrix assembly and produce ehlers-danlos syndrome type I.

Ehlers-Danlos syndrome (EDS) is a heterogeneous connective tissue disorder that severely impairs the structure and function of the skin, joints, eyes and blood vessels. We have identified mutations of the COL5A2 gene, which encodes the alpha2(V) chain of type V collagen, in two unrelated patients with the severe type I form of EDS. The first proband was heterozygous for a 7 bp deletion that resulted in skipping of exon 27 while the second proband was heterozygous for a single nucleotide substitution that resulted in skipping of exon 28. Cultured dermal fibroblasts from both probands produced about equal amounts of the normal and mutant alpha2(V) mRNAs and protein chains. The dermis from the first proband contained a sparse collagen fibrillar network with great variability in collagen fibril sizes and shapes. The dermal collagens were also abnormally soluble. Bone cells from the first proband also produced about equal amounts of the normal and mutant alpha2(V) mRNAs. However, the collagen fibrillar architecture and collagen solubility of the bone matrix were normal. Our findings show that heterozygous mutations of the COL5A2 gene can produce the EDS type I phenotype. They also suggest that type V collagen plays a more important role in collagen fibrillogenesis of dermis than that of bone.

Determinants of bone mineral density in postmenopausal white Iowans.

BACKGROUND: Osteoporosis is a major health problem for older individuals. For women, development of osteoporosis is a function of the accretion of "peak" bone mass in the third decade, age at menopause, and rate of bone loss with aging. Low bone mineral density (BMD) is a major risk factor for osteoporosis and fracture. The purpose of this study was to identify life style, nutritional, medical, and genetic predictors of low BMD in postmenopausal Iowa women. METHODS: One hundred thirty-four postmenopausal White women ranging in age from 57 to 81 years were included in this case-control study. Bone mineral density was measured at the femoral neck, using dual photon X-ray absorptiometry (Hologic 2000 QDR). Sixty-six women with BMD measurements below 0.68 g/cm2 (the bottom quartile of the BMD distribution in the population from which participants were recruited), and 68 women with values at or above 0.83 g/cm2 (the top quartile of the BMD distribution in the same population) were included. Information about environmental, nutritional, medical, and life style modifiers of BMD was obtained by written questionnaire and telephone interview. To assess familial factors that might influence BMD, we obtained a detailed family history for each participant. In addition, we tested the hypothesis that allelic variation at the Vitamin D receptor (VDR), and the type I collagen gene (COL1A1 and COL1A2) loci influence BMD. RESULTS: Weight, loss of height, age, and age at menopause were strong predictors of BMD in our population. After adjustment for these differences, we found no effect of genotype at the COL1A1, COL1A2, and VDR loci on BMD. CONCLUSIONS: Bone mineral density is a complex trait that is influenced by several different modifiers; in the present study, weight was the best predictor of postmenopausal BMD. While several studies suggest that VDR genotype is an important determinant of BMD, we did not find this association in our population, nor did we identify an association between allelic variation at the type I collagen gene loci and BMD. Identification of genes that determine body mass index may provide additional insight into risk factors for low BMD, and osteoporotic fractures.

A splice-junction mutation in the region of COL5A1 that codes for the carboxyl propeptide of pro alpha 1(V) chains results in the gravis form of the Ehlers-Danlos syndrome (type I).

Type V collagen is a constituent of type I collagen-rich fibrils in many connective tissues and is a regulator of fibril diameter. In tissues, type V collagen is a heterotrimer with the molecular structure: alpha 1(V)2 alpha 2(V) or alpha 1(V) alpha 2(V) alpha 3(V). We report that genomic polymorphisms at the pro alpha 1(V) gene (COL5A1) locus cosegregated with the gravis form of Ehlers-Danlos syndrome (EDS) (type I) in a three generation family. Affected family members, who had classical features including joint hyperextensibility, fragile skin, and widened, atrophic scars, were heterozygous for a 4 bp deletion at positions from +3 to +6 of intron 65, which resulted in removal of exon 65 sequences from processed mRNAs. Since exon 65 encodes 78 residues of the carboxyl propeptide, the expected result of this mutation is reduced efficiency in incorporating mutant pro alpha 1(V) chains into type V collagen molecules and reduced type V collagen synthesis. These studies indicate that heterozygous mutations in COL5A1 can result in EDS type I. However, linkage studies in other EDS I families indicate the disorder is heterogeneous; linkage to both COL5A1 and COL5A2 was excluded in two other families with EDS I while a fourth family was concordant for linkage to COL5A1 (Z = 2.11; theta = 0.00).

Premature chain termination is a unifying mechanism for COL1A1 null alleles in osteogenesis imperfecta type I cell strains.

Nonsense and frameshift mutations, which predict premature termination of translation, often cause a dramatic reduction in the amount of transcript from the mutant allele (nonsense-mediated mRNA decay). In some genes, these mutations also influence RNA splicing and induce skipping of the exon that contains the nonsense codon. To begin to dissect how premature termination alters the metabolism of RNA from the COL1A1 gene, we studied nonsense and frameshift mutations distributed over exons 11-49 of the gene. These mutations were originally identified in 10 unrelated families with osteogenesis imperfecta (OI) type 1. We observed marked reduction in steady-state amounts of mRNA from the mutant allele in both total cellular and nuclear RNA extracts of cells from affected individuals, suggesting that nonsense-mediated decay of COL1A1 RNA is a nuclear phenomenon. Position of the mutation within the gene did not influence this observation. None of the mutations induced skipping of either the exon containing the mutation or, for the frameshifts, the downstream exons with the new termination sites. Our data suggest that nonsense and frameshift mutations throughout most of the COL1A1 gene result in a null allele, which is associated with the predictable mild clinical phenotype, OI type 1.

Absence of mutations in the promoter of the COL1A1 gene of type I collagen in patients with osteogenesis imperfecta type I.

Osteogenesis imperfecta type I results from decreased production of structurally normal type I collagen as a result of a COL1A1 "null" allele. Steady state amounts of COL1A1 mRNA are reduced in both the nucleus and cytoplasm of dermal fibroblasts from most affected subjects. Mutations involving key regulatory sequences in the COL1A1 promoter, such as the TATAAA and CCAAAT boxes, could alter steady state levels of mRNA, and therefore lead to this phenotype. To determine the frequency of such mutations in OI type I cell strains, we used PCR amplified genomic DNA in conjunction with denaturing gradient gel electrophoresis (DGGE) and SSCP, to screen the 5' untranslated domain, exon 1, and a small portion of intron 1 of the COL1A1 gene. In addition, direct sequence analysis was performed on an amplified genomic DNA fragment that included the TATAAA and CCAAAT boxes. Forty unrelated probands with OI type I, in whom no causative mutation was known, were included in the study. No mutations were included in the study. No mutations were identified in either the TATAAA or CCAAAT boxes in any of the affected people. In addition, there was little evidence of sequence diversity among any of the 40 subjects. These data suggest that mutations in the COL1A1 promoter do not play a significant role in the aetiology of OI type I.

Osteogenesis imperfecta type I: molecular heterogeneity for COL1A1 null alleles of type I collagen.

Osteogenesis imperfecta (OI) type I is the mildest form of inherited brittle-bone disease. Dermal fibroblasts from most affected individuals produce about half the usual amount of type I procollagen, as a result of a COL1A1 "null" allele. Using PCR amplification of genomic DNA from affected individuals, followed by denaturing gradient gel electrophoresis (DGGE) and SSCP, we identified seven different COL1A1 gene mutations in eight unrelated families with OI type I. Three families have single nucleotide substitutions that alter 5' donor splice sites; two of these unrelated families have the same mutation. One family has a point mutation, in an exon, that creates a premature termination codon, and four have small deletions or insertions, within exons, that create translational frameshifts and new termination codons downstream of the mutation sites. Each mutation leads to both marked reduction in steady-state levels of mRNA from the mutant allele and a quantitative decrease in type I procollagen production. Our data demonstrate that different molecular mechanisms that have the same effect on type I collagen production result in the same clinical phenotype.

Molecular heterogeneity in osteogenesis imperfecta type I.

Osteogenesis imperfecta (OI) type I is characterized by bone fragility without significant deformity, osteopenia, normal stature, blue sclerae, and autosomal dominant inheritance. Dermal fibroblasts from most affected individuals produce about half the expected amount of type I collagen, suggesting that the OI type I phenotype results from a variety of mutations which alter the apparent expression of either COL1A1 or COL1A2, the genes encoding the chains of type I collagen. Short-pulse labeling of dermal fibroblasts with [3H]proline from affected individuals in 19 families indicates that most have alterations in the expected 2:1 synthetic ratio of pro alpha 1(I): pro alpha 2(I), with most having decreased production of pro alpha 1(I). Ratios of COL1A1:COL1A2 mRNA from these individuals, using slot-blot hybridization, indicate that they fall into different groups, but that most have decreased COL1A1 mRNA levels, compared with controls. These data suggest that most of our OI I families have COL1A1 mutations. Copy number and size of the COL1A1 gene by restriction endonuclease analysis of genomic DNA from affected individuals are normal in the families examined. We have identified one 3 generation family in which all affected members have one normal COL1A1 allele and another with a 5 base-pair deletion near the 3' end of the gene. The deletion creates a shift in the translational reading-frame and predicts the synthesis of an elongated pro alpha 1(I) chain. In a second family, a father and a son have a single exon deletion that results from a splicing mutation. Chemical cleavage analysis of amplified cDNA from affected individuals in different regions of the COL1A1 gene, including the promoter, suggests that several individuals have point mutations within the coding region of the gene, while one individual may have a small deletion within the alpha 1(I) carboxyl-terminal propeptide region. Our data provide evidence for significant molecular heterogeneity within the OI type I phenotype and indicate that a variety of mutations can result in decreased synthesis of type I collagen.

Osteogenesis imperfecta type I is commonly due to a COL1A1 null allele of type I collagen.

Dermal fibroblasts from most individuals with osteogenesis imperfecta (OI) type I produce about half the normal amount of type I procollagen, as a result of decreased synthesis of one of its constituent chains, pro alpha 1 (I). To test the hypothesis that decreased synthesis of pro alpha (I) chains results from mutations in the COL1A1 gene, we used primer extension with nucleotide-specific chain termination to measure the contribution of individual COL1A1 alleles to the mRNA pool in fibroblasts from affected individuals. A polymorphic MnlI restriction endonuclease site in the 3'-untranslated region of COL1A1 was used to distinguish the transcripts of the two alleles in heterozygous individuals. Twenty-three individuals from 21 unrelated families were studied. In each case there was marked diminution in steady-state mRNA levels from one COL1A1 allele. Loss of an allele through deletion or rearrangement was not the cause of the diminished COL1A1 mRNA levels. Primer extension with nucleotide-specific chain termination allows identification of the mutant COL1A1 allele in cell strains that are heterozygous for an expressed polymorphism. It is applicable to sporadic cases, to small families, and to large families in whom key individuals are uninformative at the polymorphic sites used in linkage analysis, making it a useful adjunct to the biochemical screening of collagenous proteins for OI.