|
A correction has been published: N Engl J Med 2002;346(21):1678.
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
Background Progressive osseous heteroplasia (POH), an autosomal dominant disorder, is characterized by extensive dermal ossification during childhood, followed by disabling and widespread heterotopic ossification of skeletal muscle and deep connective tissue. Occasional reports of mild heterotopic ossification in Albright's hereditary osteodystrophy (AHO) and a recent report of two patients with AHO who had atypically extensive heterotopic ossification suggested a common genetic basis for the two disorders. AHO is caused by heterozygous inactivating mutations in the GNAS1 gene that result in decreased expression or function of the alpha subunit of the stimulatory G protein (Gs
Methods We tested the hypothesis that GNAS1 mutations cause POH, using the polymerase chain reaction to amplify GNAS1 exons and exon–intron boundaries in 18 patients with sporadic or familial POH.
Results Heterozygous inactivating GNAS1 mutations were identified in 13 of the 18 probands with POH. The defective allele in POH is inherited exclusively from fathers, a result consistent with a model of imprinting for GNAS1. Direct evidence that the same mutation can cause either POH or AHO was observed within a single family, in which the phenotype correlated with the parental origin of the mutant allele.
Conclusions Paternally inherited inactivating GNAS1 mutations cause POH. This finding extends the range of phenotypes derived from haploinsufficiency of GNAS1, provides evidence that imprinting is a regulatory mechanism for GNAS1 expression, and suggests that Gs
) of adenylyl cyclase. is a critical negative regulator of osteogenic commitment in nonosseous connective tissues.
|
AHO is caused by heterozygous inactivating mutations in GNAS1, the gene for guanine nucleotide-binding protein (G protein) alpha stimulating activity polypeptide 1, that result in decreased expression or function of the alpha subunit of the stimulatory G protein (Gs) of adenylyl cyclase.9 Although heterotopic ossification in patients with AHO is milder and more superficial than that in patients with POH, the recent description10 of inactivating GNAS1 mutations, decreased Gs activity, or both in two patients with features of both AHO and severe progressive heterotopic ossification provided a further rationale to investigate GNAS1 as a candidate gene for POH.
Methods
Subjects
We evaluated subjects from 18 unrelated kindreds with familial POH (6 cases) or sporadic POH (12 cases). The protocol was approved by the investigational review boards of the Children's Hospital of Philadelphia and the University of Pennsylvania, and written informed consent was obtained from the subjects or their parents. Patients from Families 3,6 5, 12, and 13,2 8,3 11,11 16,12 and 1813 have been previously described in case reports.
Polymerase-Chain-Reaction Amplification and Sequencing of Genomic DNA
Genomic DNA was isolated directly from blood or from Epstein–Barr virus–transformed lymphocyte lines with the use of DNA blood-isolation reagents (QIAamp, Qiagen, Valencia, Calif.); it was amplified with the use of oligonucleotide primers flanking exons 2 to 13 of the human GNAS1 gene14 under previously reported polymerase-chain-reaction (PCR) conditions.15 For exon 1, the primer pair 5'ATGGGCTGCCTCGGGAACAGTA (forward; beginning at the initiator codon) and 5'CCCTTACCCAGCAGCAGCAGGC (reverse) was used for 40 cycles: 1 minute at 94°C, 40 seconds at 70°C, and 1 minute at 72°C. Each PCR reaction contained 200 ng of genomic DNA, each primer at a concentration of 0.4 µM, 0.08 mM deoxynucleoside triphosphates (Pharmacia, Piscataway, N.J.), 1.5 mM magnesium chloride, and 1.25 U of Taq polymerase (Life Technologies, Gaithersburg, Md.). Products were sequenced by the DNA Sequencing Core Facility of the University of Pennsylvania School of Veterinary Medicine.
RNA Isolation and Reverse Transcriptase–PCR Analysis
The total RNA from lymphoblastoid cell lines was isolated with use of Trizol reagent (Life Technologies), and poly(A)+ RNA was isolated with use of Micro-FastTrack reagents (Invitrogen, Carlsbad, Calif.). First-strand complementary DNA (cDNA) was synthesized from 1 µg of total RNA with the use of 12.5 µg of oligo-dT primer per milliliter and 1 mg of RNasin (Promega, Madison, Wis.) per milliliter, 5 U of SuperScript II reverse transcriptase (Life Technologies) per milliliter, 5 mM magnesium chloride, 10 mM dithiothreitol, and 1 mM deoxynucleoside triphosphates (Pharmacia) in a final volume of 10 µl at 42°C for 1 hour, followed by heat inactivation at 95°C for 10 minutes. First-strand cDNA (5 µl) was amplified by PCR with 1.5 mM magnesium chloride, 2.5 units of Amplitaq (Perkin–Elmer, Norwalk, Conn.), and 50 ng of each forward and reverse primer in a final volume of 20 µl for 40 cycles: one minute at 94°C, one minute at 53° to 72°C (optimized for each primer pair), and two minutes at 72°C. The products were gel-purified and sequenced as above. Primer pairs to amplify GNAS1 cDNA were designed with MacVector (Accelrys, San Diego, Calif.).
The GenBank accession numbers are AH002748 for GNAS1 human genomic DNA and NM_000516 for GNAS1 human cDNA.
Results
Inheritance Patterns in Families with POH
Subjects with POH from 18 unrelated families were examined (Figure 2). In four families (Families 1, 2, 3, and 4), children inherited POH from an affected father, a finding consistent with autosomal dominant inheritance. Two additional families (Families 5 and 18) had affected siblings with no affected parent. The severity and distribution of heterotopic ossification varied among affected members in each family; at least one affected person in each family had severe progressive heterotopic ossification during childhood, whereas other affected family members had dermal and subcutaneous ossification later in life, with slower progression to deeper tissues. The remaining 12 families each had only one member with POH. Affected persons had no primary skeletal or developmental defects and no laboratory or clinical evidence of hormone resistance.
|
Subjects with POH were screened for mutations in GNAS1 by PCR amplification and DNA sequencing of exons 1 to 13, all splice junctions, and some small introns. Heterozygous mutations were identified in 13 of 18 families: in 12 families, affected persons had small insertions, deletions, or tandem duplications that resulted in frame shifts predicted to cause premature termination of translation of the messenger RNA (mRNA) and loss of protein function (Figure 3 and Table 1). In one family, Family 14, a single-nucleotide, G-to-C mutation altered the invariant AG dinucleotide of the intron 12 acceptor splice site (Figure 3). Mutation of the conserved AG dinucleotide to AC results in the use of a downstream AG as the acceptor site and production of a frame-shifted mRNA (data not shown).
|
|
Nonpenetrant GNAS1 Mutations in Families with POH
In four kindreds with POH, we examined genomic DNA from phenotypically normal parents of affected children for GNAS1 mutations. The clinically unaffected father in Family 5 (Figure 2) had the same single-nucleotide deletion in exon 10 as his three affected daughters, indicating the absence of penetrance of this mutation in the father. Similarly, the unaffected father in Family 18 (Subject I-1 in Figure 2) carried the same deletion in exon 7 that was identified in his five affected daughters. In both Family 5 and Family 18, the unaffected mother and the unaffected siblings did not have the mutation. By contrast, the single affected person in Family 6 (pedigree not shown) and Subject II-1 in Family 1 (Figure 2 and Table 1) presumably had new mutations, because none of their parents carried that mutation, although gonadal mosaicism in a parent is possible. With the exception of the unaffected fathers in Families 5 and 18, all affected subjects who were tested in families with inheritance of POH, but none of the unaffected subjects who were tested in such families, shared the same GNAS1 mutation.
Expression of GNAS1 mRNA in POH
According to reverse-transcriptase–PCR amplification of GNAS1 mRNA from 10 subjects with identified mutations, only 2 (from Families 2 and 14) had expression of both normal and mutant GNAS1 transcripts (Table 1). Both subjects had a mutation near the 3' end of the gene: one was a frame shift in exon 13, and the other was an altered intron 12 acceptor splice site. The remaining eight subjects had only the normal mRNA sequence. This result indicated that mRNA from the mutant GNAS1 allele was not synthesized or was rapidly degraded, probably through nonsense-mediated decay,16 a cellular mechanism causing the rapid turnover of mRNA that contains nonsense and frame-shift mutations.
Variable Expression of GNAS1 Mutations and Parental Inheritance Patterns
The same 4-bp deletion found in Families 1, 9, 11, and 18 has been identified as the basis for GNAS1 haploinsufficiency in several patients with AHO.17 The 348delC frame-shift mutation identified in two subjects with POH from Families 6 and 12 (Table 1) has also been reported in a patient with AHO.18 Thus, the identical mutations in GNAS1 can result in widely varying clinical phenotypes. All other GNAS1 mutations identified in patients with POH are unique.
Direct evidence of variable expression of the same GNAS1 mutation was observed within a single family, and the phenotype correlated with the parental origin of the mutant allele (Family 18 in Figure 2). The proband (Subject II-8) had classic features of POH (severe cutaneous and subcutaneous ossification and progressive heterotopic ossification of skeletal muscle). Four of her five sisters had a milder POH phenotype. The proband and her affected siblings had a 4-bp frame-shift deletion in GNAS1 exon 7. The clinically unaffected father (Subject I-1) carried the mutation, but the unaffected mother did not, indicating inheritance of the mutant GNAS1 allele from a father with a nonpenetrant mutation. Nonpenetrance was also evident in Subject I-1 in Family 5. In both cases of fathers with nonpenetrant mutations, the parental origin of the mutated allele could not be determined. In Family 18, three children of the proband's sisters (Subjects III-3, III-7, and III-10 in Figure 2) exhibited features of AHO, including traces of subcutaneous ossification. Subcutaneous ossification was also noted in a fourth child (Subject III-6), who died suddenly at five months of age.13 The two children available for examination (Subjects III-3 and III-7) carried the GNAS1 mutation. The findings in this family support the hypothesis that the parental origin of the mutant GNAS1 allele determines whether POH or AHO will develop in a child.
Discussion
G proteins are members of a superfamily of guanosine triphosphate–binding proteins.19 All members of this superfamily bind guanine nucleotides with high affinity and specificity, possess intrinsic guanosine triphosphatase activity, and function as molecular switches: they are "on" in the guanosine triphosphate–bound conformation and "off" after hydrolysis of guanosine triphosphate to guanosine diphosphate. The most clearly defined function of G proteins is in signal transduction. A large variety of extracellular first messengers, such as peptide and glycoprotein hormones, neurotransmitters, growth factors, chemotactic agents, and sensory signals, regulate cellular activity or function by binding to cell-surface receptors that are associated with G proteins. The G proteins act as timing switches to regulate the generation of intracellular second messengers, including cyclic AMP.
During the past decade, G proteins have been the focus of intense investigation, with particular interest in the investigation of Gs, the stimulatory G protein of adenylyl cyclase, as a basis for human disease.19 At least three disorders with features of dysregulated osteogenesis are associated with somatic or germ-line mutations in GNAS1, the gene that encodes Gs, the alpha subunit of Gs. McCune–Albright syndrome is caused by postzygotic mutations that activate Gs and cause constitutive (hormone-independent) activation of adenylyl cyclase,20 whereas AHO9 and platelike osteoma cutis21 are caused by inactivating germ-line GNAS1 mutations. We determined that another disorder of osteogenesis, POH, is due to inactivating mutations in the same gene.
Phenotypic expression of genetic disorders caused by single-gene mutations can be highly variable, most notably in autosomal dominant conditions.22 Many of these disorders vary, either in the severity of a solitary abnormality or in the characteristic set of seemingly unrelated abnormalities, or in both. Inactivating mutations in GNAS1 have both these effects in AHO. Affected subjects have a wide range of developmental and somatic defects of variable severity, including subcutaneous ossification, skeletal dysmorphology, obesity, mental retardation, and hormone resistance. By contrast, patients with POH have a single characteristic disease manifestation (progressive heterotopic ossification) that can vary in its degree of severity.
In addition to variable expression in POH, we observed nonpenetrance (i.e., the complete failure of disease expression in some carriers of a gene mutation) in two families. Carriers of nonpenetrant GNAS1 mutations could have new mutations with a mosaic distribution that excludes cells that would express disease characteristics. Alternatively, nonpenetrance or variable expression could be due to influences from other genetic loci, epigenetic modifications, or environmental factors.
For several human genetic disorders, expression of the disease phenotype is influenced by whether the disease has been inherited from the father or the mother. Recent studies showing that maternal and paternal alleles of some genes do not function identically have helped clarify this phenomenon, known as imprinting.23 A consistent difference between AHO and POH is the parental origin of the mutant GNAS1 allele. This is illustrated by Family 18, in which paternal inheritance of the mutant GNAS1 allele resulted in POH, and maternal inheritance resulted in AHO.
Previous clinical and molecular studies9 have revealed that maternal transmission of GNAS1 mutations leads to the complete form of AHO, termed pseudohypoparathyroidism type 1a (MIM 103580), in which there is resistance to parathyroid hormone and other hormones whose receptors require Gs to stimulate adenylyl cyclase. By contrast, paternal inheritance of a GNAS1 mutation causes a less severe form of AHO, termed pseudopseudohypoparathyroidism (MIM 300800), in which the somatic defects are accompanied by normal hormone responsiveness. As in pseudopseudohypoparathyroidism, paternal inheritance of a mutant GNAS1 allele in patients with POH is associated with a lack of hormone resistance, but the two clinical phenotypes are otherwise clearly distinguishable.8,9
Molecular studies of the GNAS1 gene show monoallelic transcript expression according to the parent of origin,24,25,26,27 a result consistent with parental imprinting. However, the manner of expression is far more complex than anticipated. In addition to a Gs-specific exon 1, three unique, alternative first exons are spliced to exons 2 through 13 of GNAS1. One of the alternative exons encodes the neurosecretory protein NESP55 and is expressed exclusively from the maternal allele, whereas the other two alternative first exons are transcribed only from the paternal allele.24,25,28 Expression of the Gs-specific exon 1 appears to be biallelic in most tissues that have been examined,24,25,29 although expression exclusively from the maternally derived allele has been documented in some tissues (e.g., murine renal cortex and adipose tissue30 and human pituitary31).
Given the complexity of transcript synthesis at the GNAS1 locus, it is not surprising that a wide variety of phenotypic effects results from GNAS1 mutations and that the phenotypes are influenced by the parent from whom the mutant allele is inherited. Hormone resistance (such as that in pseudohypoparathyroidism type 1a) is strongly correlated with mutations in the maternally derived allele, indicating that the maternal allele is critical (at least in some tissues) for cellular functions required for signal transduction. In contrast, severe, progressive heterotopic ossification, such as that found in POH, correlates with paternal inheritance of the GNAS1 mutation, suggesting that the paternal allele specifically influences progressive osteoblastic differentiation, proliferation of cells in soft connective tissues, or both.
Why paternal inheritance of a GNAS1 mutation should result in pseudopseudohypoparathyroidism in some families and POH in others is unclear. Other phenotypes that have been associated with GNAS1 inactivation (such as features of AHO — including mild dermal ossification — that occur in both pseudohypoparathyroidism type 1a and pseudopseudohypoparathyroidism) seem to be influenced by GNAS1 mutations in either the maternal or the paternal allele. Tissue and developmental specificity of imprinting at the GNAS1 locus is a plausible cause of the disparity among phenotypes. Overlapping patterns of variable expressivity (i.e., a range of phenotypic severity) by GNAS1 maternal, paternal, and biallelic transcripts may result in a complex set of phenotypes among patients with GNAS1-inactivating mutations.
Our findings expand the spectrum of phenotypic variability attributable to mutations in the GNAS1 gene and establishes GNAS1 as an important regulator of osteoblastic commitment in nonosteogenic connective tissues. Further investigation into imprinting effects and the regulation and function of the multiple transcripts of the GNAS1 locus will be required to understand the genotypic complexity and phenotypic variability associated with mutations at this locus.
Supported by grants from the Progressive Osseous Heteroplasia Association, the International Fibrodysplasia Ossificans Progressiva Association, the New Jersey Association of Student Councils, the Ian Cali Fund, the Roemex Fellowship, Shriners Hospitals for Children (85480, to Dr. Whyte), the National Institute of Arthritis and Musculoskeletal and Skin Diseases (R01-AR46831, to Dr. Shore, and R01-AR41916, to Dr. Kaplan), and the National Institute of Diabetes and Digestive and Kidney Diseases (R01-DK34281, to Dr. Levine) and by the Isaac and Rose Nassau Professorship of Orthopaedic Molecular Medicine.
We are indebted to R. Bufo, J.P. Chanoine, J.E. Clarkson, J.M. Connor, N. Esterly, J. Fairly, F. Gannon, L. Gordon, G. Lehine, G.D. MacEwen, T. Malpass, S.P. Robertson, S. Rosenfeld, E. Shinall Miller, and J.A. Urtizberea for valuable discussions and referrals of patients; to J.K. Fang of the University of Pennsylvania School of Veterinary Medicine DNA Sequencing Facility; and to F. Gardner, S. Roth, R. Elkinson, M. Vamvakis, and J. Peeper for encouragement and support. We dedicate this work to our patients with POH worldwide.
Source Information
From the Departments of Orthopaedic Surgery (E.M.S., J.A., M.L., M.X., F.S.K.), Genetics (E.M.S., M.A.Z.), and Medicine (F.S.K.), University of Pennsylvania School of Medicine, Philadelphia; the Ilyssa Center for Molecular and Cellular Endocrinology (S.J.B., M.A.L.) and the Departments of Medicine (S.J.B.) and Pediatrics (M.A.L.), Johns Hopkins University School of Medicine, Baltimore; Genetic Health Services, Victoria and Murdoch Children's Research Institute, Royal Children's Hospital, Melbourne, Australia (R.J.M.G.); and the Center for Metabolic Bone Disease and Molecular Research, Shriners Hospitals for Children, and the Division of Bone and Mineral Diseases, Washington University School of Medicine, St. Louis (M.P.W.).
Address reprint requests to Dr. Shore at the Department of Orthopaedic Surgery, University of Pennsylvania School of Medicine, 424 Stemmler Hall, 36th and Hamilton Walk, Philadelphia, PA 19104-6018, or at shore{at}mail.med.upenn.edu.
References
-subunit of the stimulatory G protein and severe extraskeletal ossification. J Bone Miner Res 2000;15:2074-2083. [CrossRef][ISI][Medline]
| |||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
GNAS1 Mutations and Progressive Osseous Heteroplasia
Ahmed S. F., Barr D. G.D., Bonthron D. T., Farfel Z., Shore E. M., Kaplan F. S., Levine M. A., Bastepe M., Jüppner H.
Extract |
Full Text |
PDF
N Engl J Med 2002;
346:1669-1671, May 23, 2002.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
C). IVS 12 denotes intron (intervening sequence) 12. In Panels A through F, the arrow indicates the start of the frame-shifted sequence. In Panels H and I, the dashed line indicates the sequence that is duplicated in the region indicated by the solid line.