Pseudohypoparathyroidism Type Ia From Maternalbut Not Paternal Transmission of a GsaGene Mutation

Jon M. Nakamoto,1* Anna T. Sandstrom,3 Arnold S. Brickman,2 Robert A. Christenson,4 andCornelis Van Dop1

1Department of Pediatrics, University of California, Los Angeles; Los Angeles, California2Department of Medicine, University of California, Los Angeles; Los Angeles, California3Department of Pediatrics, Permanente Medical Group, Hayward, California4Texas Tech University Health Sciences Center, El Paso, Texas

While loss-of-function mutations in Gsa areinvariably associated with the short statureand brachydactyly of Albright hereditaryosteodystrophy (AHO), the association withhormone resistance (to parathyroid hor-mone and thyrotropin) typical of pseudohy-poparathyroidism type Ia (PHP-Ia) is muchmore variable. Observational studies andDNA polymorphism analysis suggest thatmaternal transmission of the Gsa mutationmay be required for full expression of clini-cal hormone resistance. To test this hypoth-esis, we studied transmission of a frameshiftmutation in Gsa through three generationsof a pedigree affected by AHO and PHP-Ia.While all family members carrying this loss-of-function mutation in one Gsa allele hadAHO, neither the presence of the mutationnor the degree of reduction of erythrocyteGsa bioactivity allowed prediction of pheno-type (AHO alone versus AHO and PHP-Ia).Paternal transmission of the mutation (fromthe patriarch of the first generation to threemembers of the second generation) was notassociated with concurrent PHP-Ia, but ma-ternal transmission (from two women in thesecond generation to four children in thethird generation) was invariably associatedwith PHP-Ia. No expansion of an upstreamshort CCG nucleotide repeat region was de-tected, nor was there evidence of uniparen-tal disomy by polymorphism analysis. Thisreport, the first to document the effectsacross three generations of both paternal

and maternal transmission of a specific Gsamutation, strongly supports the hypothesisthat a maternal factor determines full ex-pression of Gsa dysfunction as PHP-Ia. Am.J. Med. Genet. 77:261–267, 1998.© 1998 Wiley-Liss, Inc.

KEY WORDS: Albright hereditary osteo-dystrophy; pseudohypopara-thyroidism type Ia; Gsa muta-tion

INTRODUCTION

The hormone resistance syndrome pseudohypopara-thyroidism type Ia (PHP-Ia) has long been consideredthe archetypal disease of guanine nucleotide bindingprotein (G-protein) underactivity. Identification of thespecific molecular defects responsible for this conditionin humans has been the result of a remarkable inter-play of clinical observation and basic research, notably:(1) the initial demonstrations of resistance to parathy-roid hormone [Albright et al., 1942; Tashjian et al.,1966] and the role of cyclic AMP as a hormonal secondmessenger [Rall et al., 1956]; (2) proof of decreased pro-duction of cyclic AMP (cAMP) in PHP-Ia [Chase et al.,1969], and the first notions of a GTP-binding protein(later known as Gs) which couples hormone-receptorinteraction to elevation of cellular cAMP [Rodbell et al.,1971; Salomon et al., 1975]; and (3) measurement ofdecreased Gs protein activity in cells from patients withPHP-Ia [Farfel et al., 1980; Levine et al., 1980] andpurification [Northup et al., 1980] and cloning [Harriset al., 1985] of the specific Gs subunit (Gsa) responsiblefor GTP binding and stimulation of adenylyl cyclase,with demonstration of reduced messenger RNA for Gsain most patients with PHP-Ia [Carter et al., 1987;Levine et al., 1988]. Most recently, specific loss-of-function mutations in the gene for Gsa [Patten et al.,1990; Weinstein et al., 1990] have been identified inAlbright hereditary osteodystrophy (AHO), the short

Contract grant sponsor: NIH; Contract grant numbers:DK01997, RR-00865.

*Correspondence to: Jon M. Nakamoto, M.D., Ph.D., MDCC22-315, UCLA Children’s Hospital, 10833 Le Conte Ave., Los An-geles, CA 90095-1752. E-mail: jnakamot@ucla.edu

Received 4 August 1997; Accepted 20 January 1998

American Journal of Medical Genetics 77:261–267 (1998)

© 1998 Wiley-Liss, Inc.

stature–brachydactyly syndrome which accompaniesPHP-Ia, but which may exist without clinical hormoneresistance as well (occasionally termed pseudo-pseudohypoparathyroidism). In the search to under-stand what determines whether loss-of-function Gsamutations are expressed as AHO alone versus AHOwith PHP-Ia, observational studies and polymorphismanalysis have suggested that maternal transmission ofthe mutation is a key factor in the full expression ofclinical hormone resistance [Davies and Hughes, 1993;Wilson et al., 1994]. We have identified a frameshiftmutation in Gsa that was transmitted through threegenerations of a family affected by AHO and PHP, isassociated with variable reductions in Gsa bioactivity,but only presents with clinical hormone resistance af-ter transmission of the mutation through a mother.

MATERIALS AND METHODSSubjects

The family studied (pedigree in Fig. 1) was identifiedafter referral of the four children in generation III af-fected with PHP-Ia. Solid symbols indicate the pres-ence of Albright hereditary dystrophy and clinical hor-mone resistance (in this case, hypothyroidism as indi-cated by a history of elevated serum thyrotropin[TSH]), and hypocalcemia–hyperphosphatemia–elevated parathyroid hormone (PTH). Stippled symbolsindicate the presence of AHO but without documentedabnormalities of serum calcium, phosphorus, or PTHlevels. Due to limited clinical supplies during the studyperiod, PTH infusion tests could not be performed onall subjects with AHO. All studies were performed withinformed consent and assent, according to a protocolapproved by the Human Subjects Protection Commit-tee at the University of California, Los Angeles.

Measurement of Gsa Bioactivity inErythrocyte Membranes

Measurement was performed by cyc− complementa-tion as previously described [Bourne et al., 1983; VanDop et al., 1984]. With the exception of one unaffected

relative, all subjects whose erythrocyte membraneswere tested for the ability to complement the cAMPresponse pathway in cyc− cells exposed to adrenergicstimuli were studied in at least two separate assays.Cyclic AMP produced by donor membranes stimulatedby isoproterenol in the presence of GTP was purifiedand measured using modifications of the method of Sa-lomon et al. [1974]. Results are expressed as a fractionof cAMP generated by a concurrently drawn controlsample.

Polymerase Chain Reaction (PCR) Amplificationof Specific Gsa Exon-Intron Fragments

With genomic DNA from peripheral leukocytes ofstudy subjects as the template, nine pairs of oligo-nucleotide primers were used with Taq DNA polymer-ase (Perkin-Elmer) to amplify (in nine separate frag-ments) exons 2 to 13 of the Gsa gene [Kozasa et al.,1988], along with their respective exon-intron bound-aries [Nakamoto et al., 1996]. For exon 1, due to the86% G+C content of the 58 untranslated region, no con-sistent amplification products were obtained for exon 1despite use of elevated, extended annealing, or dena-turation temperatures, co-solvents (glycerol, DMSO, ortetramethylammonium chloride), nested primers, and7-deaza-28 deoxyguanosine 58-triphosphate ([7-deaza-GTP] in a 3:1 ratio of GTP), and this gene fragment wastherefore not included in this analysis. G+C-clampeddenaturing gradient gel electrophoresis (DGGE) [Shef-field et al., 1989] and DNA sequencing of PCR-amplified DNA were performed as previously described[Nakamoto et al., 1996].

‘‘No-Blot’’ Hybridization of Radiolabeled probeand Genomic DNA

Three affected children (III-3, 4, 5), their motherwith AHO (II-4), and the clinically unaffected father(II-5), were studied for polymorphism in exon 1. To iso-late an approximately 808-bp DNA fragment contain-ing the extended region of CCG repeats in the 58-untranslated region of exon 1, genomic DNA (10 mg)was sequentially digested with AccI (3 U/mg DNA) at37°C overnight, in a final volume to keep glycerol con-centration ø5%, followed by TthlllI digestion (3 U/mgDNA) at 65°C overnight. Vertical 2% agarose gels (1.5mm thick) containing 1 mg/ml ethidium bromide werepoured between glass plates previously heated in a dry-ing oven, followed by loading of double-digested DNAand 1/10 volume 15% Ficoll/0.5% bromophenol blue/0.5% xylene cyanol. After electrophoresis at 25 V × 15min to improve stacking, the gel was run at 100 V × 3hr, followed by denaturation of the DNA in situ in thegel matrix in 0.5 M NaOH/150 mM NaCl × 20 min, tworinses with water, neutralization in 0.5 M Tris-Cl (pH8)/150 mM NaCl at 4°C (to increase gel firmness) × 20min, and then drying under vacuum × 60 min on What-man 3MM paper. Dried gels were stored in heat-sealedplastic bags at 4°C until hybridization.

Preparation of End-Labeled OligonucleotideProbe and PAGE Purification

Lower (antisense) oligonucleotide primer c13 (58-GTC CTT CTG CAG CTG CTT CTC GAT CTT TTT

Fig. 1. Pedigree of a three generation family with members havingAHO (stippled symbols) or AHO and PHP-Ia (filled symbols). Symbols andpedigree identification numbers (e.g., I-2, II-5) in all following figures referto this pedigree diagram. Open symbols represent unaffected persons, andplus or minus signs represent presence or absence of brachydactyly, re-spectively.

262 Nakamoto et al.

GTT GGC-38), which is predicted to bind solely to theapproximately 800-bp exon 1 fragment of Gsa isolatedby AccI/TthlllI double digestion, was 58-radiolabeledusing 3 ml 7.3 mM oligonucleotide (22 pmol 58 ends 4261 ng), freshly-prepared kinase buffer (final concen-tration 50 mM Tris-Cl (pH 7.6)/10 mM MgCl2/0.14%b-mercaptoethanol), 120 mCi [g-32P]ATP (6000 Ci/mmol; Dupont/NEN), and 2 ml T4 polynucleotide ki-nase in a final volume of 20 ml at 37°C × 2 hr undershielding. Stop/tracking dye (20 mM Tris-Cl (pH 8)/1mM EDTA/8 M urea/0.05% xylene cyanol/0.05% bromo-phenol blue) was added (5 ml), and the reactions loadedonto a 20% acrylamide:bis-acrylamide (19:1)/7.8 Murea sequencing gel and run at 2000 V × 4 hr. Afterremoval of the other glass plate, the wet gel was cov-ered with plastic wrap and Kodak XAR5 X-ray filmplaced against the gel with careful alignment × 309 toidentify the position of the radiolabeled band, whichwas then cut out directly from the gel, with removal ofthe band confirmed by a follow-up exposure. The gelslice was placed in a 1.5-ml tube with 400 ml 10 mMTris-Cl/0.1 mM EDTA at room temperature × 1.5 hr forelution.

Hybridization of Labeled, PurifiedOligonucleotide and Digested Genomic DNA

The dried agarose gel was placed in water and theWhatmann 3MM paper removed. Prehybridization so-lution (20 ml 5 × SSPE (pH 8)/0.3% SDS/1 mg/ml dena-tured salmon sperm DNA) was added to a heat-sealedplastic bag with the rehydrated gel and incubated for 1hr at 65°C. After assay of the oligonucleotide probespecific activity, sufficient radiolabeled probe wasadded to reach 2.7 × 106 cpm/ml hybridization solutionand the hybridization reaction allowed to proceed withshaking for 21 hr at 65°C. Washes consisted of 2 ×SSPE at 4°C × 20 min, 2 × SSPE/0.1% SDS at roomtemp × 45 min, then 2 × SSPE at 65°C × 5 min, followedby an additional wash at 65°C × 1 min. X-ray film wasexposed with an intensifying screen at −70°C.

PCR-Based Polymorphism Analysis

The following PCR oligonucleotide primer pairs wereemployed to amplify polymorphic regions for allele par-ent-of-origin analysis: GNAS1-A (58-ACG TTA GCTTGG TGG ATA AG-38) and GNAS1-B (58-AGA GCTAGA GTG GTT ACC TG-38) (annealing temperature55°C), which bracket a dinucleotide repeat in the thirdintron of the GNAS1 gene for Gsa [Granqvist et al.,1991]; src11B-A (58-TTC AAG TGG TTG CCT CTG GC-38) and src11B-B (58-AGC AAC TTG CCC AGG CTATGA-38) (annealing temperature 65°C), which amplifya 191–207 bp CA-rich repeat in the SRC gene in the20q11.2 region [Xiang et al., 1991]; and OL3 (58-CCAGAT CGC GCC ACT TCA CT-38) and OL4 (58-AGATGA GCA TAG ATA CGA GA-38), in a 1:10 ratio of OL3to OL4 (annealing temperature 60°C), which amplifiesa hypervariable AluI repeat region linked to the aden-osine deaminase gene in 20q13.1 [Economou et al.,1990]. All PCR products were analyzed by electropho-resis (5% denaturing PAGE), using either a 32P-end-labeled oligonucleotide primer and autoradiography for

detection, or when feasible, unlabeled primers and vi-sualization of ethidium bromide-stained PCR productsusing UV transillumination.

RESULTSMeasurement of Gsa Bioactivity in Erythrocyte

Membranes From Study Subjects

Figure 2 shows complementation of the cAMP sig-nalling pathway in cyc− cells by Gsa from erythrocytesof study subjects. The function of Gsa is expressed as apercentage of cAMP generation in a concurrently col-lected unaffected control subject sample. Clinically un-affected relatives had Gsa bioactivity similar to that ofcontrols, while those affected by AHO or PHP-Ia, all ofwhom carried one mutant Gsa allele (see below), hadreduced activity of erythrocyte Gsa in the cyc− comple-mentation assay. While all four children affected by

Fig. 2. Gsa bioactivity as measured by complementation of the cAMPsignaling pathway in S49 cyc− cells by Gsa from erythrocytes of studysubjects, expressed as a percentage of cAMP generation in a sample fromunaffected control subject. Open symbols, unaffected individuals, stippledsymbols, AHO; filled symbols; AHO and clinical PTH resistance.

Transmission of a Gsa Mutation in PHP-Ia 263

PHP-Ia in generation III had Gsa bioactivity reducedapproximately 50%, one of the three individuals af-fected by AHO alone in the second generation also hadsignificantly reduced Gsa activity (to 58 ± 7% that ofnormal controls) without biochemical evidence of hor-monal resistance. Her two sibs with AHO had interme-diate reductions in Gsa activity in the assay, to ap-proximately 85% that of controls.

G+C-Clamped DGGE

In affected relatives, the multiple bands represent-ing hetero- and homoduplex DNA indicated a nucleo-tide sequence change in one allele for Gsa in the am-plified exon 7 to 8 genomic fragment (Fig. 3). There wascomplete correlation of this sequence change with thepresence of AHO in this family and reduced Gsa activ-ity in the cyc− complementation assay. No other evi-dence of sequence changes was found in the remainderof the exons from 2 through 13.

DNA Sequencing of PCR-Amplified DNA

Direct sequencing of the pool of amplified genomicfragments (to minimize the effect of TaqDNA polymer-ase’s relatively high base misincorporation rate) re-vealed in all affected pedigree members a heterozygous4-bp deletion in exon 7 of the GNAS1 gene (Fig. 4),which produces a reading frame shift and predictsa nonsense amino acid sequence starting at residue189 (corresponding to exon 7) and premature termina-tion of the polypeptide chain, with lack of the criticalGTPase (encoded by exons 8 and 9), effector interaction(encoded by exons 10 to 12) and C-terminal receptorinteraction (exon 13) domains.

‘‘No-Blot’’ Hybridization of Radiolabeled Probeand Genomic DNA

With autoradiography, all DNA samples (Fig. 5)showed the expected 808-bp band (arrow) without anyevidence of a difference between the mother with AHO(II-4) and her three children who had both AHO andPHP-Ia (III-3, 4, 5) in the size of the gene region con-taining the area of CCG repeats.

PCR-Based Polymorphism Analysis

To analyze the parental origin of each allele, severalmicrosatellites on chromosome 20 (where the singlegene GNAS1 encoding all isoforms of Gsa is locatedwithin 20q12 to 20q13.2) were tested for informative-ness in this pedigree. In this pedigree neither the di-nucleotide repeat in the third intron of the GNAS1gene, nor the SRC gene CA repeat region analysisproved informative, which is not surprising given thatonly two different allele sizes make up 66 to 99% ofthese polymorphisms. The Alu repeat polymorphismwas informative for one branch of the family (II-1, 2and III-1; Fig. 6) and demonstrates that both paternaland maternal alleles are represented in the affectedchild (III-1), ruling out uniparental disomy as a mecha-nism by which the effect of the Gsa mutation couldchange across generations.

DISCUSSION

This report documents the inheritance of a specificloss-of-function mutation in the gene for Gsa acrossthree generations of a family affected with AHO andPHP-Ia, with both paternal and maternal transmissionof the nucleotide sequence change. This mutation is

Fig. 3. A: DGGE showing transmission of a Gsa DGGE polymorphismin exon 7–8 across three generations. Bands of slower mobility representheteroduplexes consisting of one wild-type DNA strand and its complemen-tary mutant DNA strand. B: DGGE result showing the same polymor-phism in remaining affected members in the last two generations.

264 Nakamoto et al.

identical to that previously described in five other kin-dreds at a deletion hot-spot in exon 7 of the Gsa gene[Weinstein et al., 1992; Yu et al., 1995; Yokoyama etal., 1996]. As with previous reports of mutations lead-ing to decreased Gsa activity [Patten et al., 1990; Wein-stein et al., 1990], there is an invariant correlation ofthis mutation with the clinical appearance of brachy-dactyly and short stature of AHO, but not necessarilywith the hormonal abnormalities typical of PHP-Ia.The variable reduction of erythrocyte Gsa activity inthe cyc− complementation assay (from approximately50 to 85% that of unaffected controls) found in relativeswith the specific mutation and AHO alone in somecases overlaps the results seen in patients with bothAHO and PHP-Ia, consistent with previous reports[Wilson et al., 1994]. Therefore, neither the degree ofreduction of Gsa bioactivity in patient erythrocytemembranes nor the presence of a specific mutation inthe Gsa gene can predict the clinical phenotype (AHOalone versus AHO and PHP-Ia).

The most consistent predictor of the full AHO andPHP-Ia phenotype is transmission of the Gsa gene de-

fect from a mother to her offspring. This parent-of-origin effect was suggested by a retrospective clinicalreview of published pedigrees affected by AHO [Daviesand Hughes, 1993], which showed that 66/66 subjectswith AHO and PHP had mothers with AHO, while insix reported cases of offspring born to a father with

Fig. 4. Detection of a 4-bp deletion in exon 7 of one Gsa allele in familymembers affected by AHO or PHP-Ia. DNA sequencing was performed onPCR-amplified Gsa exon 7–8 fragments. In members affected by AHO withor without PHP-Ia, heterozygosity at this locus leads to overlapping bandpatterns following the point at which one allele has a 4-bp deletion.

Fig. 5. Absence of length polymorphism in the imperfect CCG repeatregion of Gsa exon 1, as verified by hybridization of AccI-TthlllI-digestedgenomic DNA in situ in the agarose gel (‘‘no-blot’’) to a PAGE-purified,end-labeled antisense Gsa exon 1 oligonucleotide. A length polymorphismin the detected fragment is not present among the mother with AHO andher children with both AHO and PHP-Ia.

Fig. 6. Polymorphism analysis by PCR demonstrating inheritance ofboth paternal and maternal alleles at the 20q13.1 locus containing a hy-pervariable Alu repeat sequence.

Transmission of a Gsa Mutation in PHP-Ia 265

AHO, all six had AHO only (referred to as pseudo-pseudohypoparathyroidism, or PPHP, by the authors).Using a FokI polymorphism which is present in thegeneral population but not in itself indicative of a Gsaabnormality located in exon 5 of the Gsa gene, Wilsonet al. [1994] tracked the inheritance of the presumablydefective Gsa allele and found that only maternaltransmission of this allele was associated with the phe-notype of both AHO and PHP-Ia; unfortunately, no spe-cific Gsa gene mutation was reported in that family. Aletter to the editor of the Journal of Medical Genetics[Schuster et al., 1994] reporting a father with AHOwho reportedly also had offspring affected by PHP-Iadid not identify a specific Gsa mutation.

We think that identification of the specific Gsa genemutation is important in a study of determinants forthe phenotype of AHO alone versus AHO and PHP-Ia.In the family described here, the identified mutationpredicts a Gsa protein which lacks several critical do-mains, including the GTPase ‘‘on-off switch’’ region, aswell as domains interacting with both adenylyl cyclaseand receptors. The predicted result is a null mutation,with little interference with wild-type Gsa function. Onthe other hand, the Gsa missense mutation in two boyswith PHP-Ia whose mothers did not have AHO [Naka-moto et al., 1996] affected an amino acid residue that isevolutionarily universally conserved throughout thesuperfamily of G-proteins. This unique mutant Gsaprotein displays both temperature sensitivity and re-duced affinity for guanine nucleotide [Iiri et al., 1994].That mutant Gsa protein, which retains the ability tocouple to adenylyl cyclase but no longer responds tohormone occupancy of the luteinizing hormone or para-thyroid hormone receptors, may exert a dominantnegative effect on function of the wild-type Gsa, result-ing in a more severe effect on cAMP production thanthat due to heterozygous null mutations in the Gsagene. It will be interesting to see whether other domi-nant negative Gsa gene mutations are expressed asPHP-Ia without requiring maternal transmission ofthe mutation.

We investigated this family for evidence of twomechanisms by which parent-of-origin effects can ef-fect increasing phenotypic severity—expansion of un-stable DNA sequences and uniparental disomy. The 58untranslated region of the Gsa mRNA is very G+C rich(86%), and we postulated that might lead to an in-creased frequency of polymerase slippage in this regionduring DNA replication or RNA transcription, particu-larly in light of a number of short CCG nucleotide re-peats located immediately prior to the transcriptionstart site. Such slippage has been postulated to under-lie the expansion of trinucleotide repeats in a numberof medical conditions, including fragile X syndrome (a58 region CCG repeat) [Verkerk et al., 1991], myotonicdystrophy (38 untranslated region CTG repeat) [Brooket al., 1992; Fu et al., 1992; Mahadevan et al., 1992],spinobulbar muscular atrophy (CAG repeat in exon 1)[La Spada et al., 1991], Huntington disease (CAG re-peat) [Huntington’s Disease Collaboration Group,1993], and several other rare neurodegenerative condi-tions. However, we found no evidence of length poly-morphism of exon 1 between generations of individuals

affected by AHO and those offspring affected by bothAHO and PHP-Ia, either at the genomic level or incDNA generated from peripheral leukocytes. Thesefindings do not rule out expansion of other unstableDNA sequence regions outside of the known sequenceof the Gsa gene, but repeat expansion detection meth-ods have not identified nucleotide repeat regions in theimmediate vicinity of the Gsa gene.

Uniparental disomy [Engel and DeLozier-Blanchet,1991] is another mechanism by which the sex of thetransmitting parent may determine the severity of thedisease in the offspring. This phenomenon, where bothchromosomes or chromosomal regions are inheritedfrom one parent, has been associated with conditionssuch as Prader-Willi [Mascari et al., 1992], Angelman[Lalande, 1996], and Wiedemann-Beckwith [Henry etal., 1991] syndromes. Any gene which shows evidenceof imprinting (i.e., differential expression depending onwhether it is inherited from the mother or the father),is a candidate for increased severity of disease expres-sion due to uniparental disomy [Solter, 1988]. Themouse homolog of the murine Gsa gene is located ondistal chromosome 2 (2H) [Ashley et al., 1987], a regionknown to be imprinted and homologous to regions onhuman chromosome 20q [Searle et al., 1989; Hall1990], where the human Gsa gene is located. Althoughthere is no evidence of universal differences in tran-scription between maternal and paternal Gsa alleles[Campbell et al., 1994], subtle effects of imprinting inthe specific hormonal target tissues affected in PHP-Iawould allow uniparental disomy to produce a parent-of-origin effect of Gsa gene mutations. Using three dif-ferent sets of oligonucleotide primers spanning poly-morphic sequences on the long arm of chromosome 20in the general vicinity of the Gsa gene, we found onebranch of the family which was informative. Amongthese members, offspring affected by both AHO andPHP-Ia had both maternal and paternal alleles for Gsa,thus ruling out uniparental disomy as a mechanism forthe appearance of PHP-Ia in the offspring of motherswith AHO. Our method could not exclude a gene con-version affecting only the Gsa gene locus on chromo-some 20, but not the polymorphic region investigatedhere.

In summary, we identified a specific Gsa gene muta-tion present in three generations of a family where ma-ternal but not paternal transmission of the mutation isassociated with increased phenotypic severity, i.e., ap-pearance of both AHO and clinical resistance to para-thyroid hormone and thyrotropin, which are typical ofPHP-Ia. Neither uniparental disomy nor expansion ofthe closest intragenic nucleotide repeat sequence areinvolved. Other mechanisms associated with genomicimprinting, such as differential methylation of allelesor differential activity of genes which modulate Gsabioactivity, remain to be explored.

ACKNOWLEDGMENTSThis work was supported by National Institutes of

Health Grants DK01997 (to J.M.N.) and RR-00865 (toC.V.D.). We thank Otto Mehls, M.D. for assistance inobtaining clinical data and samples and Marilyn Scottfor technical assistance.

266 Nakamoto et al.

REFERENCESAlbright F, Burnett CH, Smith PH, Parson W (1942): Pseudohypoparathy-

roidism—an example of ‘‘Seabright-Bantam syndrome.’’ Report ofthree cases. Endocrinology 30:922–932.

Ashley PL, Ellison J, Sullivan KA, Bourne HR, Cox DR (1987): Chromo-somal assignment of the murine Gi and Gs genes. J Biol Chem 262:15299–15301.

Bourne HR, Steinberg F, Golbus MS, Farfel Z (1983): A convenient methodfor measuring receptor-cyclase coupling activity in whole blood: Appli-cation to Duchenne muscular dystrophy. Am J Med Genet 15:469–474.

Brook JD, McCurrach ME, Harley HG, Buckler AJ, Church D, AburataniH, Hunter K, Stanton VP, Thirion JP, Hudson T, Sohn R, Zemelman B,Snell RG, Rundle SA, Crow S, Davies J, Shelbourne P, Buxton J, JonesC, Juvonen V, Johnson K, Harper PS, Shaw DJ, Housman DE (1992):Molecular basis of myotonic dystrophy: Expansion of a trinucleotide(CTG) repeat at the 38 end of a transcript encoding a protein kinasefamily member. Cell 68:799–808.

Campbell R, Gosden CM, Bonthron DT (1994): Parental origin of transcrip-tion from the human GNAS1 gene. J Med Genet 31:607–614.

Carter A, Bardin C, Collins R, Simons C, Bray P, Spiegel A (1987): Reducedexpression of multiple forms of the alpha subunit of the stimulatoryGTP-binding protein in pseudohypoparathyroidism type Ia. Proc NatlAcad Sci USA 84:7266–7269.

Chase LR, Melson GL, Aurbach GD (1969): Pseudohypoparathyroidism:Defective excretion of 3858-AMP in response to parathyroid hormone. JClin Invest 48:1832–1842.

Davies AJ, Hughes HE (1993): Imprinting in Albright’s hereditary osteo-dystrophy. J Med Genet 30:101–103.

Economou EP, Bergen AW, Warren AC, Antonarakis SE (1990): Thepolydeoxyadenylate tract of Alu repetitive elements is polymorphic inthe human genome. Proc Natl Acad Sci USA 87:2951–2954.

Engel E, DeLozier-Blanchet CD (1991): Uniparental disomy, isodisomy,and imprinting: Probable effects in man and strategies for their detec-tion. Am J Med Genet 40:432–439.

Farfel Z, Brickman AS, Kaslow HR, Brothers VM, Bourne HR (1980): De-fect of receptor-cyclase coupling protein in pseudohypoparathyroidism.N Engl J Med 303:237–242.

Fu Y-H, Pizzuti A, Fenwick RG Jr, King J, Rajnarayan S, Dunne PW,Dubel J, Nasser GA, Ashizawa T, de Jong P, Wieringa B, Korneluk R,Perryman MB, Epstein HF, Caskey CT (1992): An unstable tripletrepeat in a gene related to myotonic muscular dystrophy. Science 255:1256–1258.

Granqvist M, Xiang K, Seino M, Bell GI (1991): Dinucleotide repeat poly-morphism in Gs-alpha subunit gene (GNAS1) on chromosome 20.Nucleic Acids Res 19:4569.

Hall JG (1990): Genomic imprinting: Review and relevance to human dis-eases. Am J Hum Genet 46:857–873.

Harris BA, Robishaw JD, Mumby SM, Gilman AG (1985): Molecular clon-ing of complementary DNA for the alpha subunit of the G protein thatstimulates adenylate cyclase. Science 229:1274–1277.

Henry I, Bonaıti-Pellie C, Chehensse V, Beldjord C, Schwartz C, UtermannG, Junien C (1991): Uniparental paternal disomy in a genetic cancer-predisposing syndrome. Nature 351:665–667.

Huntington’s Disease Collaboration Group (1993): A novel gene containinga trinucleotide repeat that is expanded and unstable on Huntington’sdisease chromosome. Cell 72:1–20.

Iiri T, Herzmark P, Nakamoto JM, Van Dop C, Bourne HR (1994): RapidGDP release from Gsa in patients with gain and loss of endocrinefunction. Nature 371:164–167.

Kozasa T, Itoh H, Tsukamoto T, Kaziro Y (1988): Isolation and character-ization of the human Gs-alpha gene. Proc Natl Acad Sci USA 85:2081–2085.

Lalande M (1996): Parental imprinting and human disease. Annu RevGenet 30:173–195.

La Spada AR, Wilson EM, Lubahn DB, Harding AE, Fischbeck KH (1991):Androgen receptor gene mutations in X-linked spinal and bulbar mus-cular atrophy. Nature 352:77–79.

Levine MA, Downs RW Jr, Singer MJ, Marx SJ, Spiegel AM (1980): Defi-cient activity of guanine nucleotide regulatory protein in erythrocytesfrom patients with pseudohypoparathyroidism. Biochem Biophys ResCommun 94:1319–1324.

Levine MA, Ahn TG, Klupt SF, Kaufman KD, Smallwood PM, Bourne HR,

Sullivan KA, Van Dop C (1988): Genetic deficiency of the alpha subunitof the guanine nucleotide-binding protein Gs as the molecular basis forAlbright hereditary osteodystrophy. Proc Natl Acad Sci USA 85:617–621.

Mahadevan M, Tsilfidis C, Sabourin L, Shutler G, Amemiya C, Jansen G,Nevill C, Narang M, Barcelo J, O’Hoy K, LeBlond S, Earle-MacDonaldJ, de Jong PJ, Wieringa B, Korneluk RG (1992): Myotonic dystrophymutation: An unstable CTG repeat in the 38 untranslated region of thegene. Science 255:1253–1255.

Mascari MJ, Gottlieb W, Rogan PK, Butler MG, Waller DA, Armour JAL,Jeffreys AJ, Ladda RL, Nichols RD (1992): The frequency of uniparen-tal disomy in Prader-Willi syndrome: Implications for molecular diag-nosis. N Engl J Med 326:1599–1607.

Nakamoto JM, Zimmerman D, Jones EA, Loke KY, Siddiq K, Donlan MA,Brickman AS, Van Dop C (1996): Concurrent hormone resistance(pseudohypoparathyroidism type Ia) and hormone independence (tes-totoxicosis) caused by a unique mutation in the Gas gene. BiochemMolec Med 58:18–24.

Northup JK, Sternweis PC, Smigel MD, Schleifer LS, Ross EM, Gilman AG(1980): Purification of the regulatory component of adenylate cyclase.Proc Natl Acad Sci USA 77:6516–6520.

Patten JL, Johns DR, Valle D, Eil C, Gruppuso PA, Steele G, SmallwoodPM, Levine MA (1990): Mutation in the gene encoding the stimulatoryG protein of adenylate cyclase in Albright’s hereditary osteodystrophy.N Engl J Med 322:1412–1419.

Rall TW, Sutherland EW, Wosilait WD (1956): The relationship of epineph-rine and glucagon to liver phosphorylase. III. Reactivation of liverphosphorylase in slices and in extracts. J Biol Chem 218:483–495.

Rodbell M, Birnbaumer L, Pohl SL, Krans HMJ (1971): The glucagon-sensitive adenylyl cyclase system in plasma membranes of rat liver. V.An obligatory role of guanyl nucleotides in glucagon action. J BiolChem 246:1877–1882.

Salomon Y, Londos C, Rodbell M (1974): A highly sensitive adenylate cy-clase assay. Anal Biochem 58:541–548.

Salomon Y, Lin MC, Londos C, Rendell B, Rodbell M (1975): The hepaticadenylyl cyclase. I. Evidence for transition states and structural re-quirements for guanine nucleotide activation. J Biol Chem 250:4239–4245.

Schuster V, Kress W, Kruse K (1994): Paternal and maternal transmissionof pseudohypoparathyroidism type Ia in a family with Albright heredi-tary osteodystrophy: no evidence of genomic imprinting (letter). J MedGenet 31:84.

Searle G, Peters J, Lyon MF, Hall JG, Evans EP, Edwards JH, Buckle VJ(1989): Chromosome maps of man and mouse. IV. Ann Hum Genet53:89–140.

Sheffield VC, Cox DR, Lerman LS, Myers FM (1989): Attachment of a40-base-pair G+C-rich sequence (GC-clamp) to genomic DNA frag-ments by the polymerase chain reaction results in improved detectionof single-base changes. Proc Natl Acad Sci USA 86:232–236.

Solter D (1988): Differential imprinting and expression of maternal andpaternal genomes. Annu Rev Genet 22:127–146.

Tashjian AH Jr, Frantz AG, Lee JB (1966): Pseudohypoparathyroidism:Assays of parathyroid hormone and thyrocalcitonin. Proc Natl Acad SciUSA 56:1138–1342.

Van Dop C, Bourne HR, Neer RM (1984): Father to son transmission ofdecreased Ns activity in pseudohypoparathyroidism type Ia. J Clin En-docrinol Metab 59:825–828.

Verkerk AJ, Pieretti M, Sutliffe JS, Fu YH, Kuhl DP, Pizzuti A, Reiner O,Richards S, Victoria MF, Zhang FP, Eussen BE, van Ommen G-JB,Blonden LAJ, Riggins GJ, Chastain JL, Kunst CB, Galjaard H, CaskeyCT, Nelson DL, Oostra BA, Warren ST (1991): Identification of a gene(FMR-1) containing a CGG repeat coincident with a breakpoint clusterregion exhibiting length variation in fragile X syndrome. Cell 65:905–914.

Weinstein LS, Gejman PV, Friedman E, Kadowaki K, Collins RM, GershonES, Spiegel AM (1990): Mutations of the Gs-alpha-subunit gene inAlbright hereditary osteodystrophy detected by denaturing gradientgel electrophoresis. Proc Natl Acad Sci USA 87:8287–8290.

Wilson LC, Oude Luttikhuis MEM, Fraser WD, Clayton PT, Trembath RC(1994): Parental origin of Gs-alpha gene mutations in Albright’s he-reditary osteodystrophy. J Med Genet 31:835–839.

Xiang K, Phillippe G, Seino M, Bonham K, Fugita DJ, Bell GI (1991):Dinucleotide repeat polymorphism in the human SRC gene on chromo-some 20. Nucleic Acids Res 19:6967.

Transmission of a Gsa Mutation in PHP-Ia 267