A Mouse Model of Albright Hereditary Osteodystrophy Generated by Targeted Disruption of Exon 1 of the Gnas Gene
http://www.100md.com
《内分泌学杂志》
Department of Pediatrics, Division of Pediatric Endocrinology (E.L.G.-L., J.L.C., R.Z.), The Ilyssa Center for Molecular and Cellular Endocrinology (E.L.G.-L., J.L.C., R.Z.), Departments of Neurosciences (L.S.Z.), Medicine and Psychiatry (G.W.), and Comparative Medicine (D.L.H.), The Johns Hopkins University School of Medicine, Baltimore, Maryland 21287-2520
Weis Center for Research (W.S.), The Geisinger Clinic, Danville, Pennsylvania 17822
Divisions of Endocrinology, Oncology, and Human Cancer Genetics (M.S., M.D.R.), The Ohio State University College of Medicine and Arthur G. James Comprehensive Cancer Center, Columbus, Ohio 43210
Department of Pediatric Endocrinology (M.A.L.), The Cleveland Clinic Foundation, Cleveland, Ohio 44195
Abstract
Albright hereditary osteodystrophy is caused by heterozygous inactivating mutations in GNAS, a gene that encodes not only the -chain of Gs (Gs), but also NESP55 and XLs through use of alternative first exons. Patients with GNAS mutations on maternally inherited alleles are resistant to multiple hormones such as PTH, TSH, LH/FSH, GHRH, and glucagon, whose receptors are coupled to Gs. This variant of Albright hereditary osteodystrophy is termed pseudohypoparathyroidism type 1a and is due to presumed tissue-specific paternal imprinting of Gs. Previous studies have shown that mice heterozygous for a targeted disruption of exon 2 of Gnas, the murine homolog of GNAS, showed unique phenotypes dependent on the parent of origin of the mutated allele. However, hormone resistance occurred only when the disrupted gene was maternally inherited. Because disruption of exon 2 is predicted to inactivate Gs as well as NESP55 and XLs, we created transgenic mice with disruption of exon 1 to investigate the effects of isolated loss of Gs. Heterozygous mice that inherited the disruption maternally (–m/+) exhibited PTH and TSH resistance, whereas those with paternal inheritance (+/–p) had normal hormone responsiveness. Heterozygous mice were shorter and, when the disrupted allele was inherited maternally, weighed more than wild-type littermates. Gs protein and mRNA expression was consistent with paternal imprinting in the renal cortex and thyroid, but there was no imprinting in renal medulla, heart, or adipose. These findings confirm the tissue-specific paternal imprinting of GNAS and demonstrate that Gs deficiency alone is sufficient to account for the hormone resistance of pseudohypoparathyroidism type 1a.
Introduction
ALBRIGHT HEREDITARY osteodystrophy (AHO) is an autosomal dominant disorder caused by heterozygous inactivating mutations in GNAS, the gene that encodes the -chain of Gs (Gs), the heterotrimeric G protein that couples receptors for many hormones and neurotransmitters to activation of adenylyl cyclase (reviewed in Refs.1 , 2). AHO is characterized by a constellation of somatic defects that include short stature, obesity, round face, delayed puberty, subcutaneous ossifications, brachydactyly, dental hypoplasia, and cognitive deficits (3). AHO patients with GNAS mutations on maternally inherited alleles (4) also manifest resistance to multiple hormones such as PTH, TSH, gonadotropins, GHRH, and glucagon (1, 2, 5, 6, 7, 8, 9), a condition referred to as pseudohypoparathyroidism (PHP) type 1a. By contrast, AHO patients with GNAS mutations on paternally inherited alleles have only the phenotypic features of AHO without hormonal resistance, a condition termed pseudopseudohypoparathyroidism (pseudoPHP) (1, 3, 10). This unusual inheritance pattern was first observed clinically by inspection of published pedigrees and was subsequently ascribed to genomic imprinting of the GNAS gene (4, 11).
The GNAS locus on chromosome 20q13.2 in humans and the syntenic mouse Gnas locus on chromosome 2 consist of 13 exons that encode Gs (Fig. 1). Upstream of exon 1 are three alternative first exons (12, 13, 14) that each splice onto exons 2–13 to create novel transcripts. These include XLs, which is expressed only from the paternal allele (12, 13) and encodes a signaling protein that stimulates adenylyl cyclase but lacks a known receptor (15, 16, 17); NESP55, which is expressed only from the maternal allele and encodes a secretory protein (12, 13, 18, 19); and exon 1A (associated first exon) transcript derived from the paternal allele that does not encode a known protein (14, 20, 21). These alternative first exons are reciprocally imprinted and are associated with promoters that contain differentially methylated regions that are methylated on the nonexpressed allele. By contrast, the promoter for exon 1 is within a CpG island but is unmethylated on both alleles (12, 13, 21, 22). Although Gs is biallelically transcribed in most tissues (1, 12, 13, 23, 24, 25, 26, 27), there is preferential expression from the maternal GNAS allele in some tissues (28, 29, 30, 31, 32), which accounts for the restricted pattern of hormone resistance found only in patients with PHP type 1a.
A murine model of AHO was previously developed through disruption of exon 2 of the Gnas gene and revealed tissue-specific imprinting in the renal cortex and brown and white adipose tissue (31). These mice demonstrate a neonatal phenotype that is similar to mice with uniparental disomy of distal chromosome 2 (33, 34). All mice heterozygous for the mutated allele exhibited a high rate of neonatal mortality (77–80%) and neurological dysfunction. Although the phenotypes of surviving heterozygotes varied markedly depending on the parental origin of the mutant allele, only mice with maternal inheritance of the targeted allele exhibited resistance to PTH. These data were consistent with the hypothesis that the PHP type 1a phenotype is caused by paternal imprinting of Gs, but altered expression of the alternative transcripts for Gnas could not be excluded as contributing to unique aspects of the mutant phenotypes.
Here we report the generation and characterization of mice that are heterozygous for targeted disruption of exon 1 of Gnas. Our report provides novel data on the effect of loss of only Gs and complements another recently described mouse model in which exon 1 was disrupted (35). We show that maternal inheritance of this disrupted allele replicates the principal features of PHP type 1a in humans and that tissue-specific paternal imprinting in the renal cortex and thyroid accounts for PTH and TSH resistance. Moreover, we found significant differences between the exon 1 disrupted mice and the previously reported exon 2 disrupted mice, suggesting that loss of NESP55 and/or XLs does cause at least some of the abnormalities observed in the exon 2-disrupted mice.
Materials and Methods
Generation and maintenance of mice
The generation of mice carrying a targeted disruption of exon 1 of Gnas was described previously (36). Mice heterogeneous for the exon 1 replacement were maintained on either a pure 129SvEv or a hybrid 129SvEv/CD-1 background (referred to as CD-1) and were genotyped by Southern blot analysis. All mice that carry a mutant maternal allele are hereafter referred to as –m/+ and those with a mutant paternal allele as +/–p. Wild-type mice are referred to as wt.
Mice were fed a standard diet of Prolab RMH2500 mouse chow, which contains 0.95% calcium and 0.96% phosphate, and water ad libitum. All mice were 3 months of age at the time of analysis unless otherwise stated. All mouse protocols were carried out in accordance with the standards of the Johns Hopkins Animal Care and Use Committee.
Length measurements
Adult mice (mean age 10.5 ± 2 months) were anesthetized by ip injection of 200 μl of a 20 mg/ml solution of Avertin (2,2,2-tribromoethanol) (Aldrich, St. Louis, MO) in tert-amyl alcohol (Aldrich). Four measurements of stretched length (base of tail to tip of nose) were taken for each mouse using a digital caliper with accuracy of 0.03 mm (Fisher, Suwanee, GA), and the average of these values was used. Mice recovered without difficulty.
RNA analysis
Gnas transcripts encoding NESP55, XLs, and Gs were analyzed by RT-PCR of total RNA extracted from brain frontal cortex RNA using TRIzol according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). Total RNA (5 μg) was treated with Superscript II (Gibco, Grand Island, NY) to produce first-strand cDNA by using random hexamers according to the manufacturer’s recommendations. We amplified transcript-specific PCR products using a common reverse primer that corresponds to nucleotide sequences in exon 2 of Gnas and specific forward primers that anneal to nucleotide sequences in alternative first exons encoding NESP55, XLs, and Gs. Table 1 lists the specific primer pairs. NESP55 and Gs were amplified using the following PCR conditions: 94 C for 4.5 min, 30 cycles of 94 C for 30 sec, 55 C for 1 min, and 72 C for 1 min and final extension at 72 C for 4 min. These conditions produced a single band of 111 bp for NESP55 and a single band of 152 bp for Gs. XLs was amplified using the following PCR conditions: 94 C for 6 min, 40 cycles at 94 C for 30 sec, 50 C for 1 min, and 72 C for 1 min and final extension at 72 C for 5 min. These conditions produced a single band of 226 bp.
Analysis of Gs by blot hybridization was performed using total RNA (Versagene RNA purification kit, Gentra, Minneapolis, MN) and a radiolabeled 600-bp cDNA probe corresponding to the C terminus of rat Gs (kindly provided by Dr. Randall Reed, Johns Hopkins University School of Medicine). Quantification was carried out by PhosphorImager densitometry (Molecular Dynamics PhosphorImager using ImageQuant software, Sunnyvale, CA) and standardized to S26 mRNA as a control for loading equivalence. For each sample we determined the ratio of Gs to S26 mRNA content and then calculated the mean of the ratios for each genotype as well as the SEM and P values.
Real-time RT-PCR
Sections of renal cortex and medulla were isolated from wild-type and heterozygous knockout mice by use of a dissecting microscope. We conducted real-time PCR analysis (LightCycler, Roche, Indianapolis, IN) of cDNA using the TaqMan reverse transcription kit (Applied Biosystems, Foster City, CA). Aliquots (275 ng) of RNA were treated with 2.5 U of DNase I (Roche) for 15 min at room temperature in a total volume of 9.5 μl. The reaction was stopped by cooling to 4 C and the addition of 1.5 μl of 20 mM EDTA, and DNase I was inactivated by subsequently heating the reaction at 65 C for 10 min. Eight microliters of DNase-treated RNA (200 ng) were reverse transcribed in a total volume of 22 μl for 30 min at 48 C followed by heat inactivation for 5 min at 95 C. Two and 5 μl of cDNA were used to amplify Gs and NESP55, respectively, in a 25-μl PCR volume (18.2 and 45.5 ng RNA/tube) using oligonucleotide primers that were fluorescently labeled with VIC-N,N,N,N-tetramethyl-6-carboxyrhodamine (TAMRA) or 5-carboxyfluorescein-TAMRA (Table 1). As a control for RNA integrity and assay normalization, 18S rRNA was amplified using the TaqMan rRNA control reagents kit (Applied Biosystems).
The PCR was performed with initial incubations at 55 C for 2 min and 95 C for 10 min followed by 40–45 cycles of 95 C for 15 sec and 60 C for 60 sec. The threshold cycle (CT) was chosen based on the amplification curve and was 0.02 for all assays. Initial template, primer, and probe dilution studies were performed to determine the optimal template amount and component concentrations to preserve assay sensitivity but also to allow for less than 5% difference in the slopes of the linear portion of the PCR amplification curve between the Gnas products and 18S. Normalized results for Gs and NESP55 were calculated using the mean CT (Gs or NESP55) of all reactions for each sample and the mean CT of 18S amplification for each sample by calculating 2 – (CT G – CT 18S) as recommended by the manufacturer (Roche). Several samples were electrophoresed through agarose gels, and all showed a single unique band at the expected size for each amplicon. All experiments were performed in duplicate on three separate occasions and no-template controls were included in all experiments as a negative control.
Membrane preparation and Gs protein analysis
Fragments of tissue were homogenized at 4 C in 10 volumes of buffer A [10 mM Tris HCl (pH 7.4), 1 mM EDTA, 1 mM dithiothreitol, 5 μg/ml leupeptin] with 10% sucrose with a Polytron homogenizer (Brinkmann, Luzerne, Switzerland) for 3 x 8 sec at a setting of 8. The homogenate was placed on a cushion of buffer A with 44.5% sucrose and centrifuged at 150,000 x g at 4 C for 30 min. The membrane fraction was collected from the interface layer, resuspended in buffer A without sucrose, and centrifuged at 150,000 x g at 4 C for 30 min. The pellet was recentrifuged and resuspended in buffer A for storage at –70 C. Protein assays were determined by the BCA assay (Pierce, Rockford, IL) according to the instructions of the manufacturer using BSA as a standard.
Immunoblot analysis was performed as previously described (37), using either polyclonal antisera directed against a carboxyl terminus decapeptide of Gs (NEI-805; NEN Life Science Products, Boston, MA) at a 1:1000 dilution or an amino terminal fragment of Gs (C584, a generous gift of Dr. Janet Robishaw, Weis Center for Research, The Geisinger Clinic) at a 1:2000 dilution. Antibody was detected by subsequent incubation with [125I]protein A and radioactivity was quantified by PhosphorImager densitometry.
Adenylyl cyclase assays
Adenylyl cyclase assays were carried out as previously described for kidney and heart membranes (38, 39). For kidney membranes the assay buffer consisted of 50 mM Tris HCl (pH 7.5), 25 mM KCl, 0.2 mM EDTA, 0.2 mM ATP, 2 mM MgCl2, 0.1 mM cAMP, and 1 mM 3-isobutyl-1-methylxanthine (38). For heart membranes the assay buffer contained 25 mM Tris·HCl (pH 7.5), 1 mM cAMP, 5 mM MgCl2, 1 mM ATP, and 1 mM EDTA (39). Basal activity was measured in the presence of 10 μM GTP. Sodium fluoroaluminate (10 mM NaF with 30 μM AlCl3) was used as a direct stimulator of G protein activation. Receptor-stimulated adenylyl cyclase activity was assayed in the presence of 10 μM (–)-isoproterenol (isoprenaline) plus 10 μM GTP for heart membranes or 100 nM [Nle8,18,Tyr34] bPTH(1–34) plus 10 μM GTP for kidney membranes. MnCl2 (10 or 20 mM) was used as a direct stimulator of adenylyl cyclase. The stop solution contained 3H-cAMP as the recovery label plus excess cold cAMP (40).
Immunohistochemistry
Paraffin-embedded mouse thyroid tissue was analyzed for Gs using the VectaStain Elite ABC kit (Vector Laboratories, Inc., Burlingame, CA) and staining dyes following the manufacturer’s protocol. Preimmune sections were incubated with 1:1000 dilution of IgG rabbit serum (ICN Pharmaceuticals, Aurora, OH); immune sections with 1:1000 dilution of Gs C terminus rabbit antiserum (NEN Life Science Products). Slides were stained with Vector VIP peroxidase substrate.
Results were interpreted independently by three blinded investigators (E.L.G.-L., J.L.C., and M.A.L.) and scored for the presence or absence of staining. Scoring was compared among the three investigators and agreement was required for a positive result to be determined.
Hormone assays
Calcium was determined on mouse serum diluted 1:51 in 0.5 M LiCl3 with a flame atomic absorptiometer (model 2380; PerkinElmer, Norwalk, CT) or on undiluted serum. PTH was measured in undiluted serum by immunoradiometric assay (Immutopics, Inc., San Clemente, CA). Serum phosphorus was determined by colorimetric assay with 0.5–10 mM sodium phosphate (Na2HPO4·7 H2O) as a standard. T4 was measured using serum from mice that were 1 yr of age; TSH was measured using serum from mice that were 3–6 months of age; and PTH was measured using serum from mice that were either 3 months (CD-1) or 1 yr (129SvEv) old. All other assays were performed using serum isolated from mice at 3 months of age. For the high-phosphate diet experiments, all mice were changed to diet TD 87132 (Teklad, Madison, WI) containing 0.9% phosphate for 3 wk and then either continued on this diet for another 3 wk or changed to TD 87133 containing 2.0% phosphate for 3 wk before measuring PTH, calcium, and phosphorus levels.
Skeletal analyses
Skeletal studies were carried out as described (41). For histomorphometry studies, mouse skeletons were labeled by ip injection of tetracycline (20 mg/kg, as a 2 mg/ml solution in PBS) both 5 and 2 d before collection (42).
Statistical analyses
Results are reported as mean ± SEM. ANOVA was employed to consider group differences based on genotype (Gnas –m/+, +/–p, wt) using the Bonferroni-Dunn correction where appropriate. For PTH analysis, significance was determined by 2 x 3 ANOVA by treatment (diet) and group (genotype) using the Bonferroni-Dunn correction. All t tests were subjected to the Bonferroni-Dunn correction as well. Differences were considered significant for all analyses if P < 0.05. 2 analysis was used for viability, perinatal mortality, and fertility data with significant differences considered as P < 0.05 for all analyses.
Results
Viability
The generation and characterization of embryonic stem cells carrying a disruption of the Gs-specific exon 1 was reported previously (36). After blastocyst injections of one of these targeted lines, we obtained chimeric mice that transmitted the targeted allele through the germ line. From matings of heterozygous mice, we were unable to recover any homozygous mutants (Gnas–/–) among 479 progeny that survived to weaning or among 12 pups that died before weaning. We were also unable to detect any homozygous mutants among eight viable embryos analyzed at embryonic d 15.5. Hence, homozygous disruption of Gnas via disruption of exon 1 likely results in early embryonic lethality, as has been suggested for the exon 2 disruption (31).
Viability was also affected in the heterozygous state. As shown in Table 2A, when heterozygous mice (denoted as –m/+ if they carry the mutant maternal allele and +/–p if they carry the mutant paternal allele) were mated with wt mice, the number of heterozygous offspring that survived to weaning was significantly reduced, compared with the number of wt offspring (P < 0.001). Of all possible mating combinations, only two populations provided adequate numbers of offspring for statistical analysis: heterozygous offspring of the +/–p females (which were therefore –m/+) and +/–p males (which were therefore +/–p) showed mortality rates of 66 and 31–40%, respectively. Therefore, this reduction in viability was more pronounced when the mutant allele was transmitted maternally, compared with paternally (P < 0.001), consistent with paternal imprinting of Gs expression. Analysis of 12 pups that died before weaning revealed that nine had maternal inheritance of the affected allele, one had paternal inheritance, and two were wild-type. Analysis of three –m/+ mice and one +/–p mouse revealed no abnormalities on complete necropsy including thyroid and bones. Heterozygous mice that survived to weaning appeared to have a normal life span.
Length and weight characteristics
There were no obvious differences in the sizes of heterozygous knockout pups at birth, but distinctions emerged over time. Specifically, adult male –m/+ and +/–p mice both had significantly shorter lengths than their wt littermates, but there was no difference between the –m/+ and +/–p mice themselves (Fig. 2A). Adult female –m/+ and +/–p mice also had significantly shorter lengths than their wt counterparts (Fig. 2B). Moreover, the –m/+ females were significantly shorter than the +/–p females.
Male 129SvEv –m/+ mice weighed significantly more than their wt littermates during the span of 3–9 months of age (Fig. 3A). Female –m/+ mice tended to weigh more than their wt littermates at ages 3–4 months, but this was no longer evident by 5 months of age, and by 6–9 months they weighed significantly less than their wt littermates (Fig. 3B).
Both the male and female paternally derived knockouts were lean, compared with wt littermates. Male +/–p mice weighed significantly less than their wt littermates at ages 3–4 months and 6–9 months and tended to weigh less at 5 months of age. When comparing the weights at all three age groups in the male +/–p mice, the differences were not statistically significant, indicating that the male +/–p mice had no weight gain from 3 months onward. Female 129 SvEv +/–p mice also weighed significantly less than their wt littermates at 3–4 months of age, but they then caught up in weight with time, reaching the same weight as wt mice by 7 months of age. These findings in +/–p mice were also confirmed in CD-1 mice. Weights were obtained on CD-1 mice (12–20 in each group) from 1 to 7 months of age. Male +/–p mice weighed significantly less than their wt male littermates up to 4 months of age and tended to weigh less up to 7 months of age. Female +/–p mice also tended to weigh less than their female wt littermates, but the difference was significant only from ages 2–4 months, after which point they also caught up in weight with time. Hence, in both 129SvEv and CD-1 mice, the male +/–p mice weighed significantly less than wt littermate mice throughout the span of 3–9 months, and the female +/–p mice weighed less than their wt littermates up to 4 months of age.
Expression of Nesp55 and Xls
Because we specifically targeted the Gs-specific exon 1 for disruption, we hypothesized that the targeting event would not disrupt generation of other transcripts that use alternative first exons. We therefore performed RT-PCR analysis using primers specific for the first exons of Nesp55 and Xls in heterozygous mice in which the mutant Gnas allele was either maternally or paternally derived. As shown in Fig. 4, we were able to detect expression of transcripts encoding both NESP55 and XLs in brain frontal cortex, a tissue that expresses both alternative transcripts at high levels (15, 17, 18, 43). We found predicted allele-specific expression regardless of whether the maternal or paternal allele had been disrupted.
Imprinting of Gs
We evaluated the effect of disruption of exon 1 on expression of Gs protein and mRNA in Gnas +/– mice. In the renal cortex (Fig. 5), levels of Gs mRNA were significantly reduced in knockout mice, compared with wild-type mice, but levels in –m/+ mice were significantly lower than levels in +/–p mice (43.2 ± 3.0 vs. 61.8 ± 2.4% of wt, P = 0.0028), consistent with partial paternal imprinting of Gs in the renal cortex (28). By contrast, similar analyses of Gs expression in the heart and adipose (Fig. 5) showed no evidence of paternal imprinting. Levels of Gs mRNA were similarly reduced in –m/+ and +/–p, compared with wt, consistent with biallelic expression of Gs in these tissues (heart: 44.7 ± 3.0 and 44.5 ± 3.7% of wt; adipose: 37.7 ± 7.0 and 38.2 ± 8.6%). There were no differences in expression based on gender in these tissues.
Real-time RT-PCR was also used to quantify Gs expression and showed that levels of Gs mRNA in the renal medulla and adipose of heterozygous mice were approximately 44–55% of corresponding levels from wt mice, regardless of the parental origin of the disrupted Gnas allele (Fig. 6).
Expression of Gs protein in various tissues was consistent with the results of the RNA analyses. Immunoblot analysis (Fig. 7) of membranes prepared from heart ventricles showed that Gs protein was reduced to 72 ± 3% of wt in –m/+ mice and 56 ± 5% of wt in +/–p mice. Similarly, in membranes prepared from gonadal fat pads, levels of Gs protein were 65 ± 9% of wt in –m/+ mice and 62 ± 6% of wt in +/–p. In both heart and adipose tissue, the effect of genotype was significant, but there was no influence of the parental origin of the disrupted allele (P = 0.083 and 0.090 in heart and adipose, respectively) or for gender.
By contrast, expression of Gs protein in kidney and thyroid was dependent on the parental origin of the disrupted Gnas allele. Membranes prepared from whole kidney contained significantly reduced levels of Gs protein, with 62 ± 6% of wt in –m/+ mice and 81 ± 7% of wt in +/–p mice. The levels in –m/+ mice were significantly less than those in +/–p mice.
Immunohistochemical analysis also revealed paternal imprinting of Gs in the thyroid. As shown in Fig. 8, there was little to no staining for Gs in follicular epithelial cells surrounding the colloid globules of –m/+ mice, compared with intense staining in the wt mice. The +/–p mice had intermediate amounts of staining.
Adenylyl cyclase activity
Because signaling through Gs is required for hormone activation of adenylyl cyclase, we analyzed membrane-bound adenylyl cyclase activity in tissues from knockout mice. Kidney membranes (Fig. 9A) from +/–p mice showed normal adenylyl cyclase activity in response to PTH as well as fluoroaluminate, which directly activates Gs. By contrast, –m/+ mice showed significantly reduced adenylyl cyclase responses to PTH and fluoroaluminate, compared with wt or +/–p mice. Similar trends were observed under basal conditions.
Adenylyl cyclase responsiveness to fluoroaluminate and the -adrenergic receptor agonist (–)-isoproterenol was reduced in cardiac membranes (Fig. 9B) from both –m/+ and +/–p mice, but there was no effect of parental origin of the disrupted Gnas allele. There was also a trend toward decreased adenylyl cyclase activity assayed under basal conditions.
Responses of adenylyl cyclase to Mn2+, a direct activator of catalytic activity, were similar in wt and heterozygous knockout mice, indicating that observed differences in adenylyl cyclase responsiveness to receptor activation and fluoroaluminate were not due to variation in the activity or expression of adenylyl cyclase protein.
In vivo responsiveness to PTH and TSH
Reduced Gs-coupled signaling in renal proximal tubules and thyroid follicular cells of maternal knockout mice would be expected to impair hormone responsiveness, leading to increased serum concentrations of PTH and TSH, respectively. Heterozygous knockout SvEv129 mice that were fed a standard diet (0.95% calcium and 0.96% phosphate content) had normal serum levels of calcium (Fig. 10A) and phosphorus (Fig. 10B). This was also found in CD-1 mice (data not shown). SvEv129 mice showed significantly elevated PTH levels, regardless of the parent of origin of the disrupted Gnas allele (2.4-fold greater in +/–p and 3-fold greater in –m/+ mice, compared with wt mice) (Fig. 11A), whereas –m/+ but not +/–p CD-1 mice showed increased PTH levels (Fig. 11B).
When SvEv129 wt mice were given a diet containing high phosphate content (2.0%), serum PTH levels were approximately 6-fold greater than in age-matched mice that were fed the standard phosphate diet. Under these conditions, –m/+ mice had significantly greater levels of PTH than their wt counterparts (2.9-fold). The +/–p mice had PTH levels that were intermediate in level, trending higher than their wt counterparts (2-fold) but lower than in the –m/+ mice (Fig. 11A). All mice that were given the high phosphate diet showed concomitant decreases in serum calcium concentrations and increases in serum phosphorus that were significantly different from corresponding concentrations in their counterparts fed the normal phosphate diet (Fig. 10, A and B). However, there were no differences in calcium and phosphorus levels between the heterozygous knockout and wt SvEv129 mice.
Reduction of Gs expression also affected responsiveness to TSH. As shown in Fig. 12A, TSH levels in SvEv129 mice were significantly elevated in –m/+ mice, compared with wt, with intermediate levels of TSH for the +/–p mice. Total T4 levels, however, were not different among the three groups (Fig. 12B).
Fertility and maternal successfulness in knockout mice
In addition to PTH and TSH, many other hormones and neurotransmitters bind to heptahelical receptors that signal through Gs, including the reproductive hormones LH and FSH. We therefore examined possible effects of Gs on fertility in our mouse model. As shown in Table 2B, when standardized for breeding opportunities, there was no difference in the number of pups born or weaned when either parent was +/–p, although the overall viability was lower, compared with wt. However, the number of pups born was significantly less (P < 0.0001) when either parent was –m/+ indicating impaired fertility of both the male and female –m/+ mice.
When the F2-F6 generations were analyzed over a span of 2 yr, we found that the litter sizes of 129SvEv mice were markedly smaller when either parent carried a disrupted Gnas allele, with the greatest reduction in litter size noted when the mother was –m/+ (Table 2C). When the offspring were standardized per breeding opportunity to examine viability, the percent that survived to the time of weaning of –m/+ females was significantly less than the other knockout mice (Table 2B). When examining the total number of pups born (both wt and knockout), there was an approximately 20% survival rate before weaning for pups produced by –m/+ females and a 73% survival rate of pups born to +/–p females. The corresponding survival rates of pups born to wt mothers who had bred with either –m/+ males or +/–p males were approximately 83 and 80%, respectively (Table 2B). By contrast, the reported survival rate for wt 129SvEv mice is approximately 92% (www.taconic.com/addinfo/repro.htm#129sve).
The reduced litter size and lower survival rate of pups born to females with maternal inheritance of the targeted allele (P < 0.0001), as well as the much lower number weaned per breeding opportunity, is consistent with reduced fertility and poor mothering by the –m/+ female. In addition, the very high proportion of pups that died before weaning born to the –m/+ mother, compared with the –m/+ father (Table 2A) supports poor mothering by the –m/+ females. In addition, the number of litters was markedly reduced for the –m/+ fathers such that even with a litter size comparable with that of the +/–p father and mother, there was a significant difference when standardized for breeding opportunities, indicating reduced fertility in the –m/+ fathers.
Absence of skeletal phenotype in knockout mice
Because skeletal defects and heterotopic ossifications are common in human patients with AHO, we examined the heterozygous mice for these abnormalities. We found no evidence of heterotopic ossifications: no sc nodules were palpated on 70 heterozygous mice, no ossifications were seen on serial hematoxylin and eosin sections of the skin of two –m/+ mice, and no soft tissue calcifications were observed by radiological examination of three additional –m/+ mice. Skeletons were prepared from 11 mice from both CD-1 and 129SvEv strains (three –m/+, three +/–p, five wt) and stained with Alizarin Red and Alcian Blue. No differences were noted in the extent of mineralization, and no disproportionate shortening of the long bones or metatarsals was noted.
Histomorphometry was performed on undecalcified sections of distal femur from CD-1 mice that had been double labeled with tetracycline. Static indices calculated for each mouse included bone volume, osteoblast surface, and trabecular thickness; structural indices included trabecular separation and trabecular number; and kinetic indices included mineralizing surface and mineral apposition rate. Osteoid surface was not calculated because no unmineralized osteoid was observed in young mice. Osteoclast surface was not calculated because too few osteoclasts were observed. There were no significant differences in any of the calculated indices between the heterozygous and wild-type mice at 5 months and 1 yr of age. There were significant differences in these indices between young and old mice that suggested a lower bone mass in older mice. Younger mice had higher osteoblast surface, bone volume, and trabecular number and lower trabecular separation than older mice (Table 3). Hence, unlike human patients with AHO, mice with targeted disruption of Gs did not have skeletal abnormalities.
Discussion
A long-standing clinical conundrum in patients with heterozygous inactivating mutations of GNAS has been the marked variability in the AHO phenotype. Clinical studies had first disclosed an association of phenotype with the parental origin of the defective GNAS allele, such that patients with mutations on maternally inherited alleles are resistant to multiple hormones and therefore have PHP type 1a, whereas patients with GNAS mutations on paternally inherited alleles have normal hormone responsiveness and only the somatic features of AHO, a condition termed pseudoPHP. These findings suggested that paternal imprinting of GNAS might be responsible for the variable phenotype. The subsequent observation that patients with PHP type 1a might have hormone resistance in some tissues (e.g. deficient response of proximal renal tubule to PTH, thyroid follicular cell to TSH, and pituitary somatotroph to GHRH) but not in others (e.g. normal response of distal renal tubule to vasopressin and adrenal cortex to ACTH) further refined the presumed molecular pathophysiology to tissue-specific paternal imprinting of Gs. Recent studies of GNAS expression in normal human tissues (12, 13, 24, 25, 26, 27, 28, 29, 30, 32) and mice have provided important confirmation of this hypothesis and have demonstrated both tissue-specific paternal imprinting of Gs and development of PTH resistance only when the maternal Gnas allele is disrupted or missing (31, 44).
A murine model of AHO was developed previously by other investigators through disruption of exon 2 of the Gnas gene (31). Homozygosity for the exon 2 disrupted allele led to embryonic lethality and heterozygosity led to extensive perinatal mortality. These heterozygotes had unique phenotypes depending on the parent of origin of the mutated allele: –m/+ offspring had wide bodies and higher birth weights, and 80% of them developed neurological symptoms with delayed cerebellar development and died after 6–21 d. The +/–p newborns had narrow bodies, lower birth weights, and a 77% mortality rate within 24 h of birth from lack of suckling. Those mice that survived weaning lived a normal life span.
The unusual development of neurological dysfunction in the –m/+ mice and the unexpectedly severe phenotype of the +/–p mice suggested that loss of other proteins encoded by Gnas might play an important role in determining the murine phenotype of Gnas in the exon 2 disrupted model. We therefore generated mice carrying a targeted disruption of exon 1 to study the effect of isolated deficiency of Gs. As described for mice with a disruption of exon 2, homozygosity resulted in lethality, and the –m/+ mice had a high perinatal mortality rate. However, unlike mice with targeted disruption of exon 2, the exon 1 –m/+ mice that we and Chen et al. (35) generated did not have severe neurological abnormalities or change in body habitus. This suggests that the predicted loss of NESP55 in exon 2 knockout mice may be the cause of these features because NESP55 is highly expressed in the brain (18, 43). Similar neurological features have not been described in humans with PHP type 1a who inherit a defective maternal GNAS allele with a mutation in exons 2–13, which constitute the 3' untranslated region of the NESP55 transcript (1, 2). Although most of these GNAS mutations are small and therefore may not disrupt the NESP55 transcript, a mutant NESP55 allele of maternal origin has recently been described in one phenotypically normal subject (45). However, in recently reported mice with targeted disruption of Nesp55, mice were found to have abnormal reactivity to new environments (46). Taken together, these observations suggest that NESP55 may be more critical for development of the brain in mice than in humans.
As with the –m/+ mice, the majority of the +/–p mice with the targeted disruption of exon 1 did not die within 24 h, had significantly higher perinatal survival, and had normal body shapes at birth, similar to the characteristics of the Gnas exon 1 knockout mice recently reported by Chen et al. (35). These features contrast with the exon 2 knockout and implicate the loss of XLs as the basis for unusual body habitus, poor suckling behavior, and high perinatal mortality, a hypothesis recently substantiated through the generation of mice with mutations in Gnasxl (47). These mice have poor postnatal growth and survival as well as diverse physiological defects, indicating that XLs is necessary for a variety of postnatal adaptations (47).
We found evidence for incomplete paternal imprinting of Gs in the renal cortex, with a corresponding decrease in PTH-stimulated adenylyl cyclase activity and elevated serum concentrations of PTH in –m/+ mice. Remarkably, serum levels of calcium and phosphorus remained normal despite evidence of PTH resistance in contrast to the findings in the exon 2 knockout mice in which the –m/+ mice, but not the +/–p mice, had decreased serum calcium and elevated serum phosphorus levels (31). The difference between the two mouse models may be secondary to strain variations. Although our mouse model did not exhibit any calcium and phosphorus abnormalities, this is still consistent with the findings in some patients with PHP type 1a. When the exon 1 knockout mice were fed a high-phosphate diet, they had lower calcium and higher phosphorus levels and exaggerated PTH responses, but there still were no significant differences between the calcium and phosphorus levels in wild-type and heterozygous knockout mice.
By contrast to the renal cortex, we found no evidence of imprinting in the renal medulla, consistent with intact vasopressin responsiveness in patients with PHP type 1a and the findings in the exon 2 disrupted mouse (31). Heterozygous +/–p mice had intermediate levels of PTH resistance, which differs from the findings in the exon 2 disrupted +/–p mouse (31). Given that recent studies suggest that XLs can antagonize Gs-dependent signaling pathways (47), it is tempting to speculate that the intermediate reduction in PTH sensitivity in our +/–p mice results from normal expression of XLs, whereas more normal sensitivity to PTH in the exon 2 disrupted +/–p mice results from the predicted absence of XLs expression.
In addition, we found that imprinting in the thyroid was accompanied by TSH resistance in our mouse model, consistent with human thyroid data (28, 30, 32). Although TSH levels were significantly elevated, serum T4 levels were not affected. However, there were no differences in serum TSH, T3, or T4 between exon 2 disrupted –m/+, +/–p, and wt mice (48). The features of our exon 1 knockout mice are consistent with the findings in most patients with PHP type 1a, who have elevated serum levels of TSH but normal levels of serum T4 (1, 2, 7, 49).
Imprinting was clearly tissue specific, similar to the findings of Yu et al. (31). We found no evidence for imprinting of Gs expression in the renal medulla, heart, or adipose. Although we found biallelic expression of Gs in fat tissue from the exon 1 disrupted mice, Yu et al. observed apparent paternal imprinting of Gs in both brown and white adipose tissue from their exon 2 disrupted mice (31) despite normal adrenergic responsiveness of adenylyl cyclase in both –m/+ and +/–p mice (48). Recent studies in normal human adipose tissue have shown that Gs is equivalently expressed from both parental alleles (27), thus supporting the lack of paternal imprinting of Gs in adipose that we observed in our mouse model.
In the recently reported Gnas exon 1 knockout mice (35), the greater severity of the obesity and metabolic abnormalities in the –m/+ mice, compared with the +/–p mice, is consistent with more significant effect of Gs deficiency in the –m/+ mice, possibly due to reduced hypothalamic melanocortin-Gs signaling in these mice (35).
Paternally derived (+/–p) heterozygous mice weighed less than wt mice, similar to the results described in exon 2 knockout mice (2, 48). Recent data in the exon 2 disrupted mice have shown that the +/–p mice have increased insulin sensitivity (50) and increased lipid clearance, and it has been hypothesized that XLs deficiency in the exon 2 knockout mice may be responsible (47, 51). This was substantiated in the Gnas exon 1 knockout mouse recently described (35). Both the –m/+ and +/–p mice exhibited insulin resistance and low metabolic rates. Interestingly, XLs expression is present in our –p/+ mice, which have the same lean phenotype as the exon 2 disrupted mice, thus implying that another mechanism, such as differences in strain and genetic background, may account for the differences between the exon 1 knockout models that we and Chen et al. (35) created.
Maternally derived (–m/+) heterozygous mice had marked weight gain in the males with a trend in the young females. This is consistent with the findings in the recently reported exon 1 knockout mouse (35) and with the phenotype of patients with PHP type 1a.
The Gs-specific targeted disruption of exon 1 resulted in shorter adult mice for both genders in –m/+ and +/–p mice, as in the exon 2 knockout mouse (26). These findings are consistent with the short stature seen in adults with PHP type 1a and pseudoPHP. The basis for short stature in patients with AHO is unknown but appears to be related to premature closure of epiphyses. It has become clear that Gs is a negative regulator of the differentiation of growth plate chondrocytes with intricate interactions with Indian hedgehog and PTHrP (26, 52, 53, 54, 55), and therefore a deficiency of Gs could lead to premature epiphyseal fusion as is seen with overactivation of Indian hedgehog and PTHrP signaling (52, 54, 55). Evidence in humans has revealed biallelic expression of GNAS in bone (27), and in mice, biallelic expression has been demonstrated in chondrocytes (26). Therefore, it is likely that haploinsufficiency of Gs is responsible for the similar occurrence of premature fusion and short stature in both PHP type 1a and pseudoPHP.
In addition, recent studies suggest a possible role for GH deficiency in the development of short stature due to GHRH resistance in some patients with PHP type 1a (6, 9). GH deficiency, however, is not present in pseudoPHP in which there is no hormonal resistance. We found that in the exon 1 knockout mice, the –m/+ females are significantly shorter than their +/–p counterparts. It is possible that patients with PHP type 1a who are GH deficient are shorter than their pseudoPHP counterparts, and this potential difference in phenotype needs to be evaluated to determine the significance of GH deficiency on height.
From analysis of heterozygous mice with targeted disruption of exon 2, there is no obvious skeletal phenotype consistent with AHO on gross examination (26, 31). We also found no skeletal phenotype in the heterozygous exon 1 knockout mouse. By contrast, some epiphyseal and growth plate abnormalities occur in a mouse in which both copies of Gnas were inactivated in chondrocytes via disruption of exon 2 (26, 53). This dosage discrepancy between mouse and human may reflect differences in bone biology because mouse epiphyses, unlike human epiphyses, do not fuse.
It is known that reproductive dysfunction is present in many women with PHP type 1a, and serum levels of LH and FSH may be either normal or elevated (7, 56, 57, 58). By contrast, very little is known about reproductive fitness in men with PHP type 1a. We found both male and female –m/+ mice had reduced fertility. Decreased fertility could be due to paternal imprinting of Gs in the ovary (30) as well as the testis (currently under investigation). The –m/+ females also exhibited very poor mothering skills, but the basis for this defect is presently unknown. The inability to react appropriately to novel stimuli in NESP55-deficient mice suggests that the lack of NESP55 may be responsible for the impaired ability to react to the new situation of motherhood (46).
Our findings indicate that loss of Gs alone is sufficient to cause a murine homolog of AHO. The mice with targeted disruption of exon 1 encoding Gs develop variable phenotypic features that are based on the parent of origin of the disrupted Gnas allele and that reflect the effect of both heterozygous loss of Gs and tissue-specific paternal imprinting. This paternal imprinting is partial and not an all-or-none phenomenon, as has already been shown in a variety of normal human tissues. Our data suggest that loss of NESP55 may lead to the abnormal neonatal phenotype and neurological abnormalities found in the –m/+ mice with targeted disruption of exon 2 and that loss of XLs may lead to the abnormal neonatal phenotype and high perinatal mortality rate in the +/–p exon 2 knockout mice. The physiological functions of NESP55 and XLs have not yet been fully elucidated, but our studies extend prior observations that suggest that these two proteins have more critical roles in mice than humans. The basis for this difference, and the role of imprinting in regulating expression of these two proteins, further extend the complexity of the GNAS locus.
Acknowledgments
The authors are grateful to Dr. John D. Gearhart and Dr. Ann M. Lawler for carrying out blastocyst injections and embryo transfers; Dr. Kimberly O. O’Brien for assistance in analysis of serum calcium levels; Tracey Hatcher for help with bone histomorphometry; and Dr. Amy Wisniewski for advice regarding statistical analyses.
Footnotes
This work was supported by United States Public Health Service Grants RO1 DK34281 and DK56178 (to M.A.L.), Supplement PA-99106 to RO1 DK56178 (to E.L.G.-L.), RO1 AA09000 (to G.W.), National Institutes of Health Research Resources RR00171 (to D.L.H.), National Institutes of Health/National Center for Research Resources Grant MOI RR00052 to the Johns Hopkins University School of Medicine General Clinical Research Center, and a generous donation by the Bosworth family.
Abbreviations: AHO, Albright hereditary osteodystrophy; CD-1, hybrid 129SvEv/CD-1 background; CT, threshold cycle; Gs, -chain of Gs; –m/+, mice with a mutant maternal allele; +/–p, mice with a mutant paternal allele; PHP, pseudohypoparathyroidism; wt, wild-type mice.
References
Levine MA 2002 Pseudohypoparathyroidism. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. San Diego: Academic Press; 1137–1163
Weinstein LS, Yu S, Warner DR, Liu J 2001 Endocrine manifestations of stimulatory G protein -subunit mutations and the role of genomic imprinting. Endocr Rev 22:675–705
Albright F, Burnett CH, Smith PH 1942 Pseudohypoparathyroidism: an example of "Seabright-Bantam syndrome." Endocrinology 30:922–932
Davies SJ, Hughes HE 1993 Imprinting in Albright’s hereditary osteodystrophy. J Med Genet 30:101–103
Brickman AS, Carlson HE, Levin SR 1986 Responses to glucagon infusion in pseudohypoparathyroidism. J Clin Endocrinol Metab 63:1354–1360
Germain-Lee EL, Groman J, Crane JL, Jan de Beur SM, Levine MA 2003 Growth hormone deficiency in pseudohypoparathyroidism type 1a: another manifestation of multihormone resistance. J Clin Endocrinol Metab 88:4059–4069
Levine MA, Downs Jr RW, Moses AM, Breslau NA, Marx SJ, Lasker RD, Rizzoli RE, Aurbach GD, Spiegel AM 1983 Resistance to multiple hormones in patients with pseudohypoparathyroidism. Association with deficient activity of guanine nucleotide regulatory protein. Am J Med 74:545–556
Bastepe M, Juppner H 2003 Pseudohypoparathyroidism and mechanisms of resistance toward multiple hormones: molecular evidence to clinical presentation. J Clin Endocrinol Metab. 88:4055–4058 (Editorial)
Mantovani G, Maghnie M, Weber G, De Menis E, Brunelli V, Cappa M, Loli P, Beck-Peccoz P, Spada A 2003 Growth hormone-releasing hormone resistance in pseudohypoparathyroidism type 1a: new evidence for imprinting of the Gs gene. J Clin Endocrinol Metab 88:4070–4074
Albright F, Forbes AP, Henneman PH 1952 Pseudopseudohypoparathyroidism. Trans Assoc Am Physicians 65:337–350
Wilson LC, Trembath RC 1994 Albright’s hereditary osteodystrophy. J Med Genet. 31:779–784 (Review)
Hayward BE, Kamiya M, Strain L, Moran V, Campbell R, Hayashizaki Y, Bonthron DT 1998 The human GNAS1 gene is imprinted and encodes distinct paternally and biallelically expressed G proteins. Proc Natl Acad Sci USA 95:10038–10043
Hayward BE, Moran V, Strain L, Bonthron DT 1998 Bidirectional imprinting of a single gene: GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc Natl Acad Sci USA 95:15475–15480
Swaroop A, Agarwal N, Gruen JR, Bick D, Weissman SM 1991 Differential expression of novel Gs signal transduction protein cDNA species. Nucleic Acids Res 17:4725–4729
Kehlenbach RH, Matthey J, Huttner WB 1994 XLs is a new type of G protein. Nature 372:804–808
Klemke M, Pasolli HA, Kehlenbach RH, Offermanns S, Schultz G, Huttner WB 2000 Characterization of the extra-large G protein -subunit XLs. II. Signal transduction properties. J Biol Chem 275:33633–33640
Pasolli HA, Klemke M, Kehlenbach RH, Wang Y, Huttner WB 2000 Characterization of the extra-large G protein -subunit XL s. I. Tissue distribution and subcellular localization. J Biol Chem 275:33622–33632
Ischia R, Lovisetti-Scamihorn P, Hogue-Angeletti R, Wolkersdorfer M, Winkler H, Fischer-Colbrie R 1997 Molecular cloning and characterization of NESP55, a novel chromogranin-like precursor of a peptide with 5-HT1B receptor antagonist activity. J Biol Chem 272:11657–11662
Leitner B, Lovisetti-Scamihorn P, Heilmann J, Striessnig J, Blakely RD, Eiden LE, Winkler H 1999 Subcellular localization of chromogranins, calcium channels, amine carriers, and proteins of the exocytotic machinery in bovine splenic nerve. J Neurochem 72:1110–1116
Jan de Beur SJ, Ding C, Germain-Lee E, Cho J, Maret A, Levine MA 2003 Discordance between genetic and epigenetic defects in pseudohypoparathyroidism type 1b revealed by inconsistent loss of maternal imprinting at GNAS1. Am J Hum Genet. 73:314–322.
Liu J, Litman D, Rosenberg MJ, Yu S, Biesecker LG, Weinstein LS 2000 A GNAS1 imprinting defect in pseudohypoparathyroidism type IB. J Clin Invest 106:1167–1174
Liu J, Yu S, Litman D, Chen W, Weinstein LS 2000 Identification of a methylation imprint mark within the mouse Gnas locus. Mol Cell Biol 20:5808–5817
Levine MA, Jap TS, Mauseth RS, Downs RW, Spiegel AM 1986 Activity of the stimulatory guanine nucleotide-binding protein is reduced in erythrocytes from patients with pseudohypoparathyroidism and pseudopseudohypoparathyroidism: biochemical, endocrine, and genetic analysis of Albright’s hereditary osteodystrophy in six kindreds. J Clin Endocrinol Metab 62:497–502
Campbell R, Gosden CM, Bonthron DT 1994 Parental origin of transcription from the human GNAS1 gene. J Med Genet 31:607–614
Zheng H, Radeva G, McCann JA, Hendy GN, Goodyer CG 2001 Gs transcripts are biallelically expressed in the human kidney cortex: implications for pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 86:4627–4629
Bastepe M, Weinstein LS, Ogata N, Kawaguchi H, Juppner H, Kronenberg HM, Chung UI 2004 Stimulatory G protein directly regulates hypertrophic differentiation of growth plate cartilage in vivo. Proc Natl Acad Sci USA 101:14794–14799
Mantovani G, Bondioni S, Locatelli M, Pedroni C, Lania AG, Ferrante E, Filopanti M, Beck-Peccoz P, Spada A 2004 Biallelic expression of the Gs gene in human bone and adipose tissue. J Clin Endocrinol Metab 89:6316–6319
Germain-Lee EL, Ding CL, Deng Z, Crane JL, Saji M, Ringel MD, Levine MA 2002 Paternal imprinting of Gs in the human thyroid as the basis of TSH resistance in pseudohypoparathyroidism type 1a. Biochem Biophys Res Commun 296:67–72
Hayward BE, Barlier A, Korbonits M, Grossman AB, Jacquet P, Enjalbert A, Bonthron DT 2001 Imprinting of the G(s) gene GNAS1 in the pathogenesis of acromegaly. J Clin Invest 107:R31–R36
Mantovani G, Ballare E, Giammona E, Beck-Peccoz P, Spada A 2002 The Gs gene: predominant maternal origin of transcription in human thyroid gland and gonads. J Clin Endocrinol Metab 87:4736–4740
Yu S, Yu D, Lee E, Eckhaus M, Lee R, Corria Z, Accili D, Westphal H, Weinstein LS 1998 Variable and tissue-specific hormone resistance in heterotrimeric Gs protein -subunit (Gs) knockout mice is due to tissue-specific imprinting of the Gs gene. Proc Natl Acad Sci USA 95:8715–8720
Liu J, Erlichman B, Weinstein LS 2003 The stimulatory G protein -subunit Gs is imprinted in human thyroid glands: implications for thyroid function in pseudohypoparathyroidism types 1A and 1B. J Clin Endocrinol Metab 88:4336–4341
Cattanach BM, Kirk M 1985 Differential activity of maternally and paternally derived chromosome regions in mice. Nature 315:496–498
Peters J, Beechey CV, Ball ST, Evans EP 1994 Mapping studies of the distal imprinting region of mouse chromosome 2. Genet Res 63:169–174
Chen M, Gavrilova O, Liu J, Xie T, Deng C, Nguyen AT, Nackers LM, Lorenzo J, Shen L, Weinstein LS 2005 Alternative Gnas gene products have opposite effects on glucose and lipid metabolism. Proc Natl Acad Sci USA 102:7386–7391
Schwindinger WF, Reese KJ, Lawler AM, Gearhart JD, Levine MA 1997 Targeted disruption of Gnas in embryonic stem cells. Endocrinology 138:4058–4063
Froehlich JC, Wand GS 1997 Adenylyl cyclase signal transduction and alcohol-induced sedation. Pharmacol Biochem Behav 58:1021–1030
Bellorin-Font E, Martin KJ 1981 Regulation of the PTH-receptor-cyclase system of canine kidney: effects of calcium, magnesium, and guanine nucleotides. Am J Physiol 241:F364–F373
Vatner DE, Asai K, Iwase M, Ishikawa Y, Wagner TE, Shannon RP, Homcy CJ, Vatner SF 1998 Overexpression of myocardial Gs prevents full expression of catecholamine desensitization despite increased -adrenergic receptor kinase. J Clin Invest 101:1916–1922
Salomon Y, Londos C, Rodbell M 1974 A highly sensitive adenylate cyclase assay. Anal Biochem 58:541–548
McPherron AC, Lawler AM, Lee SJ 1999 Regulation of anterior/posterior patterning of the axial skeleton by growth/differentiation factor 11. Nat Genet 22:260–264
McCarthy EF, Frassica FF1998 Pathology of bone and joint disorders. Philadelphia: W. B. Saunders Co.; 22–24
Lovisetti-Scamihorn P, Fischer-Colbrie R, Leitner B, Scherzer G, Winkler H 1999 Relative amounts and molecular forms of NESP55 in various bovine tissues. Brain Res 829:99–106
Williamson CM, Ball ST, Nottingham WT, Skinner JA, Plagge A, Turner MD, Powles N, Hough T, Papworth D, Fraser WD, Maconochie M, Peters J 2004 A cis-acting control region is required exclusively for the tissue-specific imprinting of Gnas. Nat Genet 36:894–899
Kim SJ, Gonen D, Hanna GL, Leventhal BL, Cook Jr EH 2000 Deletion polymorphism in the coding region of the human NESP55 alternative transcript of GNAS1. Mol Cell Probes 14:191–194
Plagge A, Isles AR, Gordon E, Humby T, Dean W, Gritsch S, Fischer-Colbrie R, Wilkinson LS, Kelsey G 2005 Imprinted Nesp55 influences behavioral reactivity to novel environments. Mol Cell Biol 25:3019–3026
Plagge A, Gordon E, Dean W, Boiani R, Cinti S, Peters J, Kelsey G 2004 The imprinted signaling protein XLs is required for postnatal adaptation to feeding. Nat Genet 36:818–826
Yu S, Gavrilova O, Chen H, Lee R, Liu J, Pacak K, Parlow AF, Quon MJ, Reitman ML, Weinstein LS 2000 Paternal versus maternal transmission of a stimulatory G-protein subunit knockout produces opposite effects on energy metabolism. J Clin Invest 105:615–623
Werder EA, Illig R, Bernasconi S, Kind H, Prader A 1975 Excessive thyrotropin-releasing hormone in pseudohypoparathyroidism. Pediatr Res 9:12–16
Yu S, Castle A, Chen M, Lee R, Takeda K, Weinstein LS 2001 Increased insulin sensitivity in Gs knockout mice. J Biol Chem 276:19994–19998
Chen M, Haluzik M, Wolf NJ, Lorenzo J, Dietz KR, Reitman ML, Weinstein LS 2004 Increased insulin sensitivity in paternal Gnas knockout mice is associated with increased lipid clearance. Endocrinology 145:4094–4102
Kobayashi T, Chung UI, Schipani E, Starbuck M, Karsenty G, Katagiri T, Goad DL, Lanske B, Kronenberg HM 2002 PTHrP and Indian hedgehog control differentiation of growth plate chondrocytes at multiple steps. Development 129:2977–2986
Sakamoto A, Chen M, Kobayashi T, Kronenberg HM, Weinstein LS 2005 Chondrocyte-specific knockout of the G protein G(s) leads to epiphyseal and growth plate abnormalities and ectopic chondrocyte formation. J Bone Miner Res 20:663–671
Tavella S, Biticchi R, Schito A, Minina E, Di Martino D, Pagano A, Vortkamp A, Horton WA, Cancedda R, Garofalo S 2004 Targeted expression of SHH affects chondrocyte differentiation, growth plate organization, and Sox9 expression. J Bone Miner Res 19:1678–1688
van der Eerden BC, Karperien M, Gevers EF, Lowik CW, Wit JM 2000 Expression of Indian hedgehog, parathyroid hormone-related protein, and their receptors in the postnatal growth plate of the rat: evidence for a locally acting growth restraining feedback loop after birth. J Bone Miner Res 15:1045–1055
Namnoum AB, Merriam GR, Moses AM, Levine MA 1998 Reproductive dysfunction in women with Albright’s hereditary osteodystrophy. J Clin Endocrinol Metab 83:824–829
Shima M, Nose O, Shimizu K, Seino Y, Yabuuchi H, Saito T 1988 Multiple associated endocrine abnormalities in a patient with pseudohypoparathyroidism type 1a. Eur J Pediatr 147:536–538
Wolfsdorf JI, Rosenfield RL, Fang VS, Kobayashi R, Razdan AK, Kim MH 1978 Partial gonadotrophin-resistance in pseudohypoparathyroidism. Acta Endocrinol 88:321–328(Emily L. Germain-Lee, Wil)
Weis Center for Research (W.S.), The Geisinger Clinic, Danville, Pennsylvania 17822
Divisions of Endocrinology, Oncology, and Human Cancer Genetics (M.S., M.D.R.), The Ohio State University College of Medicine and Arthur G. James Comprehensive Cancer Center, Columbus, Ohio 43210
Department of Pediatric Endocrinology (M.A.L.), The Cleveland Clinic Foundation, Cleveland, Ohio 44195
Abstract
Albright hereditary osteodystrophy is caused by heterozygous inactivating mutations in GNAS, a gene that encodes not only the -chain of Gs (Gs), but also NESP55 and XLs through use of alternative first exons. Patients with GNAS mutations on maternally inherited alleles are resistant to multiple hormones such as PTH, TSH, LH/FSH, GHRH, and glucagon, whose receptors are coupled to Gs. This variant of Albright hereditary osteodystrophy is termed pseudohypoparathyroidism type 1a and is due to presumed tissue-specific paternal imprinting of Gs. Previous studies have shown that mice heterozygous for a targeted disruption of exon 2 of Gnas, the murine homolog of GNAS, showed unique phenotypes dependent on the parent of origin of the mutated allele. However, hormone resistance occurred only when the disrupted gene was maternally inherited. Because disruption of exon 2 is predicted to inactivate Gs as well as NESP55 and XLs, we created transgenic mice with disruption of exon 1 to investigate the effects of isolated loss of Gs. Heterozygous mice that inherited the disruption maternally (–m/+) exhibited PTH and TSH resistance, whereas those with paternal inheritance (+/–p) had normal hormone responsiveness. Heterozygous mice were shorter and, when the disrupted allele was inherited maternally, weighed more than wild-type littermates. Gs protein and mRNA expression was consistent with paternal imprinting in the renal cortex and thyroid, but there was no imprinting in renal medulla, heart, or adipose. These findings confirm the tissue-specific paternal imprinting of GNAS and demonstrate that Gs deficiency alone is sufficient to account for the hormone resistance of pseudohypoparathyroidism type 1a.
Introduction
ALBRIGHT HEREDITARY osteodystrophy (AHO) is an autosomal dominant disorder caused by heterozygous inactivating mutations in GNAS, the gene that encodes the -chain of Gs (Gs), the heterotrimeric G protein that couples receptors for many hormones and neurotransmitters to activation of adenylyl cyclase (reviewed in Refs.1 , 2). AHO is characterized by a constellation of somatic defects that include short stature, obesity, round face, delayed puberty, subcutaneous ossifications, brachydactyly, dental hypoplasia, and cognitive deficits (3). AHO patients with GNAS mutations on maternally inherited alleles (4) also manifest resistance to multiple hormones such as PTH, TSH, gonadotropins, GHRH, and glucagon (1, 2, 5, 6, 7, 8, 9), a condition referred to as pseudohypoparathyroidism (PHP) type 1a. By contrast, AHO patients with GNAS mutations on paternally inherited alleles have only the phenotypic features of AHO without hormonal resistance, a condition termed pseudopseudohypoparathyroidism (pseudoPHP) (1, 3, 10). This unusual inheritance pattern was first observed clinically by inspection of published pedigrees and was subsequently ascribed to genomic imprinting of the GNAS gene (4, 11).
The GNAS locus on chromosome 20q13.2 in humans and the syntenic mouse Gnas locus on chromosome 2 consist of 13 exons that encode Gs (Fig. 1). Upstream of exon 1 are three alternative first exons (12, 13, 14) that each splice onto exons 2–13 to create novel transcripts. These include XLs, which is expressed only from the paternal allele (12, 13) and encodes a signaling protein that stimulates adenylyl cyclase but lacks a known receptor (15, 16, 17); NESP55, which is expressed only from the maternal allele and encodes a secretory protein (12, 13, 18, 19); and exon 1A (associated first exon) transcript derived from the paternal allele that does not encode a known protein (14, 20, 21). These alternative first exons are reciprocally imprinted and are associated with promoters that contain differentially methylated regions that are methylated on the nonexpressed allele. By contrast, the promoter for exon 1 is within a CpG island but is unmethylated on both alleles (12, 13, 21, 22). Although Gs is biallelically transcribed in most tissues (1, 12, 13, 23, 24, 25, 26, 27), there is preferential expression from the maternal GNAS allele in some tissues (28, 29, 30, 31, 32), which accounts for the restricted pattern of hormone resistance found only in patients with PHP type 1a.
A murine model of AHO was previously developed through disruption of exon 2 of the Gnas gene and revealed tissue-specific imprinting in the renal cortex and brown and white adipose tissue (31). These mice demonstrate a neonatal phenotype that is similar to mice with uniparental disomy of distal chromosome 2 (33, 34). All mice heterozygous for the mutated allele exhibited a high rate of neonatal mortality (77–80%) and neurological dysfunction. Although the phenotypes of surviving heterozygotes varied markedly depending on the parental origin of the mutant allele, only mice with maternal inheritance of the targeted allele exhibited resistance to PTH. These data were consistent with the hypothesis that the PHP type 1a phenotype is caused by paternal imprinting of Gs, but altered expression of the alternative transcripts for Gnas could not be excluded as contributing to unique aspects of the mutant phenotypes.
Here we report the generation and characterization of mice that are heterozygous for targeted disruption of exon 1 of Gnas. Our report provides novel data on the effect of loss of only Gs and complements another recently described mouse model in which exon 1 was disrupted (35). We show that maternal inheritance of this disrupted allele replicates the principal features of PHP type 1a in humans and that tissue-specific paternal imprinting in the renal cortex and thyroid accounts for PTH and TSH resistance. Moreover, we found significant differences between the exon 1 disrupted mice and the previously reported exon 2 disrupted mice, suggesting that loss of NESP55 and/or XLs does cause at least some of the abnormalities observed in the exon 2-disrupted mice.
Materials and Methods
Generation and maintenance of mice
The generation of mice carrying a targeted disruption of exon 1 of Gnas was described previously (36). Mice heterogeneous for the exon 1 replacement were maintained on either a pure 129SvEv or a hybrid 129SvEv/CD-1 background (referred to as CD-1) and were genotyped by Southern blot analysis. All mice that carry a mutant maternal allele are hereafter referred to as –m/+ and those with a mutant paternal allele as +/–p. Wild-type mice are referred to as wt.
Mice were fed a standard diet of Prolab RMH2500 mouse chow, which contains 0.95% calcium and 0.96% phosphate, and water ad libitum. All mice were 3 months of age at the time of analysis unless otherwise stated. All mouse protocols were carried out in accordance with the standards of the Johns Hopkins Animal Care and Use Committee.
Length measurements
Adult mice (mean age 10.5 ± 2 months) were anesthetized by ip injection of 200 μl of a 20 mg/ml solution of Avertin (2,2,2-tribromoethanol) (Aldrich, St. Louis, MO) in tert-amyl alcohol (Aldrich). Four measurements of stretched length (base of tail to tip of nose) were taken for each mouse using a digital caliper with accuracy of 0.03 mm (Fisher, Suwanee, GA), and the average of these values was used. Mice recovered without difficulty.
RNA analysis
Gnas transcripts encoding NESP55, XLs, and Gs were analyzed by RT-PCR of total RNA extracted from brain frontal cortex RNA using TRIzol according to the manufacturer’s instructions (Invitrogen, Carlsbad, CA). Total RNA (5 μg) was treated with Superscript II (Gibco, Grand Island, NY) to produce first-strand cDNA by using random hexamers according to the manufacturer’s recommendations. We amplified transcript-specific PCR products using a common reverse primer that corresponds to nucleotide sequences in exon 2 of Gnas and specific forward primers that anneal to nucleotide sequences in alternative first exons encoding NESP55, XLs, and Gs. Table 1 lists the specific primer pairs. NESP55 and Gs were amplified using the following PCR conditions: 94 C for 4.5 min, 30 cycles of 94 C for 30 sec, 55 C for 1 min, and 72 C for 1 min and final extension at 72 C for 4 min. These conditions produced a single band of 111 bp for NESP55 and a single band of 152 bp for Gs. XLs was amplified using the following PCR conditions: 94 C for 6 min, 40 cycles at 94 C for 30 sec, 50 C for 1 min, and 72 C for 1 min and final extension at 72 C for 5 min. These conditions produced a single band of 226 bp.
Analysis of Gs by blot hybridization was performed using total RNA (Versagene RNA purification kit, Gentra, Minneapolis, MN) and a radiolabeled 600-bp cDNA probe corresponding to the C terminus of rat Gs (kindly provided by Dr. Randall Reed, Johns Hopkins University School of Medicine). Quantification was carried out by PhosphorImager densitometry (Molecular Dynamics PhosphorImager using ImageQuant software, Sunnyvale, CA) and standardized to S26 mRNA as a control for loading equivalence. For each sample we determined the ratio of Gs to S26 mRNA content and then calculated the mean of the ratios for each genotype as well as the SEM and P values.
Real-time RT-PCR
Sections of renal cortex and medulla were isolated from wild-type and heterozygous knockout mice by use of a dissecting microscope. We conducted real-time PCR analysis (LightCycler, Roche, Indianapolis, IN) of cDNA using the TaqMan reverse transcription kit (Applied Biosystems, Foster City, CA). Aliquots (275 ng) of RNA were treated with 2.5 U of DNase I (Roche) for 15 min at room temperature in a total volume of 9.5 μl. The reaction was stopped by cooling to 4 C and the addition of 1.5 μl of 20 mM EDTA, and DNase I was inactivated by subsequently heating the reaction at 65 C for 10 min. Eight microliters of DNase-treated RNA (200 ng) were reverse transcribed in a total volume of 22 μl for 30 min at 48 C followed by heat inactivation for 5 min at 95 C. Two and 5 μl of cDNA were used to amplify Gs and NESP55, respectively, in a 25-μl PCR volume (18.2 and 45.5 ng RNA/tube) using oligonucleotide primers that were fluorescently labeled with VIC-N,N,N,N-tetramethyl-6-carboxyrhodamine (TAMRA) or 5-carboxyfluorescein-TAMRA (Table 1). As a control for RNA integrity and assay normalization, 18S rRNA was amplified using the TaqMan rRNA control reagents kit (Applied Biosystems).
The PCR was performed with initial incubations at 55 C for 2 min and 95 C for 10 min followed by 40–45 cycles of 95 C for 15 sec and 60 C for 60 sec. The threshold cycle (CT) was chosen based on the amplification curve and was 0.02 for all assays. Initial template, primer, and probe dilution studies were performed to determine the optimal template amount and component concentrations to preserve assay sensitivity but also to allow for less than 5% difference in the slopes of the linear portion of the PCR amplification curve between the Gnas products and 18S. Normalized results for Gs and NESP55 were calculated using the mean CT (Gs or NESP55) of all reactions for each sample and the mean CT of 18S amplification for each sample by calculating 2 – (CT G – CT 18S) as recommended by the manufacturer (Roche). Several samples were electrophoresed through agarose gels, and all showed a single unique band at the expected size for each amplicon. All experiments were performed in duplicate on three separate occasions and no-template controls were included in all experiments as a negative control.
Membrane preparation and Gs protein analysis
Fragments of tissue were homogenized at 4 C in 10 volumes of buffer A [10 mM Tris HCl (pH 7.4), 1 mM EDTA, 1 mM dithiothreitol, 5 μg/ml leupeptin] with 10% sucrose with a Polytron homogenizer (Brinkmann, Luzerne, Switzerland) for 3 x 8 sec at a setting of 8. The homogenate was placed on a cushion of buffer A with 44.5% sucrose and centrifuged at 150,000 x g at 4 C for 30 min. The membrane fraction was collected from the interface layer, resuspended in buffer A without sucrose, and centrifuged at 150,000 x g at 4 C for 30 min. The pellet was recentrifuged and resuspended in buffer A for storage at –70 C. Protein assays were determined by the BCA assay (Pierce, Rockford, IL) according to the instructions of the manufacturer using BSA as a standard.
Immunoblot analysis was performed as previously described (37), using either polyclonal antisera directed against a carboxyl terminus decapeptide of Gs (NEI-805; NEN Life Science Products, Boston, MA) at a 1:1000 dilution or an amino terminal fragment of Gs (C584, a generous gift of Dr. Janet Robishaw, Weis Center for Research, The Geisinger Clinic) at a 1:2000 dilution. Antibody was detected by subsequent incubation with [125I]protein A and radioactivity was quantified by PhosphorImager densitometry.
Adenylyl cyclase assays
Adenylyl cyclase assays were carried out as previously described for kidney and heart membranes (38, 39). For kidney membranes the assay buffer consisted of 50 mM Tris HCl (pH 7.5), 25 mM KCl, 0.2 mM EDTA, 0.2 mM ATP, 2 mM MgCl2, 0.1 mM cAMP, and 1 mM 3-isobutyl-1-methylxanthine (38). For heart membranes the assay buffer contained 25 mM Tris·HCl (pH 7.5), 1 mM cAMP, 5 mM MgCl2, 1 mM ATP, and 1 mM EDTA (39). Basal activity was measured in the presence of 10 μM GTP. Sodium fluoroaluminate (10 mM NaF with 30 μM AlCl3) was used as a direct stimulator of G protein activation. Receptor-stimulated adenylyl cyclase activity was assayed in the presence of 10 μM (–)-isoproterenol (isoprenaline) plus 10 μM GTP for heart membranes or 100 nM [Nle8,18,Tyr34] bPTH(1–34) plus 10 μM GTP for kidney membranes. MnCl2 (10 or 20 mM) was used as a direct stimulator of adenylyl cyclase. The stop solution contained 3H-cAMP as the recovery label plus excess cold cAMP (40).
Immunohistochemistry
Paraffin-embedded mouse thyroid tissue was analyzed for Gs using the VectaStain Elite ABC kit (Vector Laboratories, Inc., Burlingame, CA) and staining dyes following the manufacturer’s protocol. Preimmune sections were incubated with 1:1000 dilution of IgG rabbit serum (ICN Pharmaceuticals, Aurora, OH); immune sections with 1:1000 dilution of Gs C terminus rabbit antiserum (NEN Life Science Products). Slides were stained with Vector VIP peroxidase substrate.
Results were interpreted independently by three blinded investigators (E.L.G.-L., J.L.C., and M.A.L.) and scored for the presence or absence of staining. Scoring was compared among the three investigators and agreement was required for a positive result to be determined.
Hormone assays
Calcium was determined on mouse serum diluted 1:51 in 0.5 M LiCl3 with a flame atomic absorptiometer (model 2380; PerkinElmer, Norwalk, CT) or on undiluted serum. PTH was measured in undiluted serum by immunoradiometric assay (Immutopics, Inc., San Clemente, CA). Serum phosphorus was determined by colorimetric assay with 0.5–10 mM sodium phosphate (Na2HPO4·7 H2O) as a standard. T4 was measured using serum from mice that were 1 yr of age; TSH was measured using serum from mice that were 3–6 months of age; and PTH was measured using serum from mice that were either 3 months (CD-1) or 1 yr (129SvEv) old. All other assays were performed using serum isolated from mice at 3 months of age. For the high-phosphate diet experiments, all mice were changed to diet TD 87132 (Teklad, Madison, WI) containing 0.9% phosphate for 3 wk and then either continued on this diet for another 3 wk or changed to TD 87133 containing 2.0% phosphate for 3 wk before measuring PTH, calcium, and phosphorus levels.
Skeletal analyses
Skeletal studies were carried out as described (41). For histomorphometry studies, mouse skeletons were labeled by ip injection of tetracycline (20 mg/kg, as a 2 mg/ml solution in PBS) both 5 and 2 d before collection (42).
Statistical analyses
Results are reported as mean ± SEM. ANOVA was employed to consider group differences based on genotype (Gnas –m/+, +/–p, wt) using the Bonferroni-Dunn correction where appropriate. For PTH analysis, significance was determined by 2 x 3 ANOVA by treatment (diet) and group (genotype) using the Bonferroni-Dunn correction. All t tests were subjected to the Bonferroni-Dunn correction as well. Differences were considered significant for all analyses if P < 0.05. 2 analysis was used for viability, perinatal mortality, and fertility data with significant differences considered as P < 0.05 for all analyses.
Results
Viability
The generation and characterization of embryonic stem cells carrying a disruption of the Gs-specific exon 1 was reported previously (36). After blastocyst injections of one of these targeted lines, we obtained chimeric mice that transmitted the targeted allele through the germ line. From matings of heterozygous mice, we were unable to recover any homozygous mutants (Gnas–/–) among 479 progeny that survived to weaning or among 12 pups that died before weaning. We were also unable to detect any homozygous mutants among eight viable embryos analyzed at embryonic d 15.5. Hence, homozygous disruption of Gnas via disruption of exon 1 likely results in early embryonic lethality, as has been suggested for the exon 2 disruption (31).
Viability was also affected in the heterozygous state. As shown in Table 2A, when heterozygous mice (denoted as –m/+ if they carry the mutant maternal allele and +/–p if they carry the mutant paternal allele) were mated with wt mice, the number of heterozygous offspring that survived to weaning was significantly reduced, compared with the number of wt offspring (P < 0.001). Of all possible mating combinations, only two populations provided adequate numbers of offspring for statistical analysis: heterozygous offspring of the +/–p females (which were therefore –m/+) and +/–p males (which were therefore +/–p) showed mortality rates of 66 and 31–40%, respectively. Therefore, this reduction in viability was more pronounced when the mutant allele was transmitted maternally, compared with paternally (P < 0.001), consistent with paternal imprinting of Gs expression. Analysis of 12 pups that died before weaning revealed that nine had maternal inheritance of the affected allele, one had paternal inheritance, and two were wild-type. Analysis of three –m/+ mice and one +/–p mouse revealed no abnormalities on complete necropsy including thyroid and bones. Heterozygous mice that survived to weaning appeared to have a normal life span.
Length and weight characteristics
There were no obvious differences in the sizes of heterozygous knockout pups at birth, but distinctions emerged over time. Specifically, adult male –m/+ and +/–p mice both had significantly shorter lengths than their wt littermates, but there was no difference between the –m/+ and +/–p mice themselves (Fig. 2A). Adult female –m/+ and +/–p mice also had significantly shorter lengths than their wt counterparts (Fig. 2B). Moreover, the –m/+ females were significantly shorter than the +/–p females.
Male 129SvEv –m/+ mice weighed significantly more than their wt littermates during the span of 3–9 months of age (Fig. 3A). Female –m/+ mice tended to weigh more than their wt littermates at ages 3–4 months, but this was no longer evident by 5 months of age, and by 6–9 months they weighed significantly less than their wt littermates (Fig. 3B).
Both the male and female paternally derived knockouts were lean, compared with wt littermates. Male +/–p mice weighed significantly less than their wt littermates at ages 3–4 months and 6–9 months and tended to weigh less at 5 months of age. When comparing the weights at all three age groups in the male +/–p mice, the differences were not statistically significant, indicating that the male +/–p mice had no weight gain from 3 months onward. Female 129 SvEv +/–p mice also weighed significantly less than their wt littermates at 3–4 months of age, but they then caught up in weight with time, reaching the same weight as wt mice by 7 months of age. These findings in +/–p mice were also confirmed in CD-1 mice. Weights were obtained on CD-1 mice (12–20 in each group) from 1 to 7 months of age. Male +/–p mice weighed significantly less than their wt male littermates up to 4 months of age and tended to weigh less up to 7 months of age. Female +/–p mice also tended to weigh less than their female wt littermates, but the difference was significant only from ages 2–4 months, after which point they also caught up in weight with time. Hence, in both 129SvEv and CD-1 mice, the male +/–p mice weighed significantly less than wt littermate mice throughout the span of 3–9 months, and the female +/–p mice weighed less than their wt littermates up to 4 months of age.
Expression of Nesp55 and Xls
Because we specifically targeted the Gs-specific exon 1 for disruption, we hypothesized that the targeting event would not disrupt generation of other transcripts that use alternative first exons. We therefore performed RT-PCR analysis using primers specific for the first exons of Nesp55 and Xls in heterozygous mice in which the mutant Gnas allele was either maternally or paternally derived. As shown in Fig. 4, we were able to detect expression of transcripts encoding both NESP55 and XLs in brain frontal cortex, a tissue that expresses both alternative transcripts at high levels (15, 17, 18, 43). We found predicted allele-specific expression regardless of whether the maternal or paternal allele had been disrupted.
Imprinting of Gs
We evaluated the effect of disruption of exon 1 on expression of Gs protein and mRNA in Gnas +/– mice. In the renal cortex (Fig. 5), levels of Gs mRNA were significantly reduced in knockout mice, compared with wild-type mice, but levels in –m/+ mice were significantly lower than levels in +/–p mice (43.2 ± 3.0 vs. 61.8 ± 2.4% of wt, P = 0.0028), consistent with partial paternal imprinting of Gs in the renal cortex (28). By contrast, similar analyses of Gs expression in the heart and adipose (Fig. 5) showed no evidence of paternal imprinting. Levels of Gs mRNA were similarly reduced in –m/+ and +/–p, compared with wt, consistent with biallelic expression of Gs in these tissues (heart: 44.7 ± 3.0 and 44.5 ± 3.7% of wt; adipose: 37.7 ± 7.0 and 38.2 ± 8.6%). There were no differences in expression based on gender in these tissues.
Real-time RT-PCR was also used to quantify Gs expression and showed that levels of Gs mRNA in the renal medulla and adipose of heterozygous mice were approximately 44–55% of corresponding levels from wt mice, regardless of the parental origin of the disrupted Gnas allele (Fig. 6).
Expression of Gs protein in various tissues was consistent with the results of the RNA analyses. Immunoblot analysis (Fig. 7) of membranes prepared from heart ventricles showed that Gs protein was reduced to 72 ± 3% of wt in –m/+ mice and 56 ± 5% of wt in +/–p mice. Similarly, in membranes prepared from gonadal fat pads, levels of Gs protein were 65 ± 9% of wt in –m/+ mice and 62 ± 6% of wt in +/–p. In both heart and adipose tissue, the effect of genotype was significant, but there was no influence of the parental origin of the disrupted allele (P = 0.083 and 0.090 in heart and adipose, respectively) or for gender.
By contrast, expression of Gs protein in kidney and thyroid was dependent on the parental origin of the disrupted Gnas allele. Membranes prepared from whole kidney contained significantly reduced levels of Gs protein, with 62 ± 6% of wt in –m/+ mice and 81 ± 7% of wt in +/–p mice. The levels in –m/+ mice were significantly less than those in +/–p mice.
Immunohistochemical analysis also revealed paternal imprinting of Gs in the thyroid. As shown in Fig. 8, there was little to no staining for Gs in follicular epithelial cells surrounding the colloid globules of –m/+ mice, compared with intense staining in the wt mice. The +/–p mice had intermediate amounts of staining.
Adenylyl cyclase activity
Because signaling through Gs is required for hormone activation of adenylyl cyclase, we analyzed membrane-bound adenylyl cyclase activity in tissues from knockout mice. Kidney membranes (Fig. 9A) from +/–p mice showed normal adenylyl cyclase activity in response to PTH as well as fluoroaluminate, which directly activates Gs. By contrast, –m/+ mice showed significantly reduced adenylyl cyclase responses to PTH and fluoroaluminate, compared with wt or +/–p mice. Similar trends were observed under basal conditions.
Adenylyl cyclase responsiveness to fluoroaluminate and the -adrenergic receptor agonist (–)-isoproterenol was reduced in cardiac membranes (Fig. 9B) from both –m/+ and +/–p mice, but there was no effect of parental origin of the disrupted Gnas allele. There was also a trend toward decreased adenylyl cyclase activity assayed under basal conditions.
Responses of adenylyl cyclase to Mn2+, a direct activator of catalytic activity, were similar in wt and heterozygous knockout mice, indicating that observed differences in adenylyl cyclase responsiveness to receptor activation and fluoroaluminate were not due to variation in the activity or expression of adenylyl cyclase protein.
In vivo responsiveness to PTH and TSH
Reduced Gs-coupled signaling in renal proximal tubules and thyroid follicular cells of maternal knockout mice would be expected to impair hormone responsiveness, leading to increased serum concentrations of PTH and TSH, respectively. Heterozygous knockout SvEv129 mice that were fed a standard diet (0.95% calcium and 0.96% phosphate content) had normal serum levels of calcium (Fig. 10A) and phosphorus (Fig. 10B). This was also found in CD-1 mice (data not shown). SvEv129 mice showed significantly elevated PTH levels, regardless of the parent of origin of the disrupted Gnas allele (2.4-fold greater in +/–p and 3-fold greater in –m/+ mice, compared with wt mice) (Fig. 11A), whereas –m/+ but not +/–p CD-1 mice showed increased PTH levels (Fig. 11B).
When SvEv129 wt mice were given a diet containing high phosphate content (2.0%), serum PTH levels were approximately 6-fold greater than in age-matched mice that were fed the standard phosphate diet. Under these conditions, –m/+ mice had significantly greater levels of PTH than their wt counterparts (2.9-fold). The +/–p mice had PTH levels that were intermediate in level, trending higher than their wt counterparts (2-fold) but lower than in the –m/+ mice (Fig. 11A). All mice that were given the high phosphate diet showed concomitant decreases in serum calcium concentrations and increases in serum phosphorus that were significantly different from corresponding concentrations in their counterparts fed the normal phosphate diet (Fig. 10, A and B). However, there were no differences in calcium and phosphorus levels between the heterozygous knockout and wt SvEv129 mice.
Reduction of Gs expression also affected responsiveness to TSH. As shown in Fig. 12A, TSH levels in SvEv129 mice were significantly elevated in –m/+ mice, compared with wt, with intermediate levels of TSH for the +/–p mice. Total T4 levels, however, were not different among the three groups (Fig. 12B).
Fertility and maternal successfulness in knockout mice
In addition to PTH and TSH, many other hormones and neurotransmitters bind to heptahelical receptors that signal through Gs, including the reproductive hormones LH and FSH. We therefore examined possible effects of Gs on fertility in our mouse model. As shown in Table 2B, when standardized for breeding opportunities, there was no difference in the number of pups born or weaned when either parent was +/–p, although the overall viability was lower, compared with wt. However, the number of pups born was significantly less (P < 0.0001) when either parent was –m/+ indicating impaired fertility of both the male and female –m/+ mice.
When the F2-F6 generations were analyzed over a span of 2 yr, we found that the litter sizes of 129SvEv mice were markedly smaller when either parent carried a disrupted Gnas allele, with the greatest reduction in litter size noted when the mother was –m/+ (Table 2C). When the offspring were standardized per breeding opportunity to examine viability, the percent that survived to the time of weaning of –m/+ females was significantly less than the other knockout mice (Table 2B). When examining the total number of pups born (both wt and knockout), there was an approximately 20% survival rate before weaning for pups produced by –m/+ females and a 73% survival rate of pups born to +/–p females. The corresponding survival rates of pups born to wt mothers who had bred with either –m/+ males or +/–p males were approximately 83 and 80%, respectively (Table 2B). By contrast, the reported survival rate for wt 129SvEv mice is approximately 92% (www.taconic.com/addinfo/repro.htm#129sve).
The reduced litter size and lower survival rate of pups born to females with maternal inheritance of the targeted allele (P < 0.0001), as well as the much lower number weaned per breeding opportunity, is consistent with reduced fertility and poor mothering by the –m/+ female. In addition, the very high proportion of pups that died before weaning born to the –m/+ mother, compared with the –m/+ father (Table 2A) supports poor mothering by the –m/+ females. In addition, the number of litters was markedly reduced for the –m/+ fathers such that even with a litter size comparable with that of the +/–p father and mother, there was a significant difference when standardized for breeding opportunities, indicating reduced fertility in the –m/+ fathers.
Absence of skeletal phenotype in knockout mice
Because skeletal defects and heterotopic ossifications are common in human patients with AHO, we examined the heterozygous mice for these abnormalities. We found no evidence of heterotopic ossifications: no sc nodules were palpated on 70 heterozygous mice, no ossifications were seen on serial hematoxylin and eosin sections of the skin of two –m/+ mice, and no soft tissue calcifications were observed by radiological examination of three additional –m/+ mice. Skeletons were prepared from 11 mice from both CD-1 and 129SvEv strains (three –m/+, three +/–p, five wt) and stained with Alizarin Red and Alcian Blue. No differences were noted in the extent of mineralization, and no disproportionate shortening of the long bones or metatarsals was noted.
Histomorphometry was performed on undecalcified sections of distal femur from CD-1 mice that had been double labeled with tetracycline. Static indices calculated for each mouse included bone volume, osteoblast surface, and trabecular thickness; structural indices included trabecular separation and trabecular number; and kinetic indices included mineralizing surface and mineral apposition rate. Osteoid surface was not calculated because no unmineralized osteoid was observed in young mice. Osteoclast surface was not calculated because too few osteoclasts were observed. There were no significant differences in any of the calculated indices between the heterozygous and wild-type mice at 5 months and 1 yr of age. There were significant differences in these indices between young and old mice that suggested a lower bone mass in older mice. Younger mice had higher osteoblast surface, bone volume, and trabecular number and lower trabecular separation than older mice (Table 3). Hence, unlike human patients with AHO, mice with targeted disruption of Gs did not have skeletal abnormalities.
Discussion
A long-standing clinical conundrum in patients with heterozygous inactivating mutations of GNAS has been the marked variability in the AHO phenotype. Clinical studies had first disclosed an association of phenotype with the parental origin of the defective GNAS allele, such that patients with mutations on maternally inherited alleles are resistant to multiple hormones and therefore have PHP type 1a, whereas patients with GNAS mutations on paternally inherited alleles have normal hormone responsiveness and only the somatic features of AHO, a condition termed pseudoPHP. These findings suggested that paternal imprinting of GNAS might be responsible for the variable phenotype. The subsequent observation that patients with PHP type 1a might have hormone resistance in some tissues (e.g. deficient response of proximal renal tubule to PTH, thyroid follicular cell to TSH, and pituitary somatotroph to GHRH) but not in others (e.g. normal response of distal renal tubule to vasopressin and adrenal cortex to ACTH) further refined the presumed molecular pathophysiology to tissue-specific paternal imprinting of Gs. Recent studies of GNAS expression in normal human tissues (12, 13, 24, 25, 26, 27, 28, 29, 30, 32) and mice have provided important confirmation of this hypothesis and have demonstrated both tissue-specific paternal imprinting of Gs and development of PTH resistance only when the maternal Gnas allele is disrupted or missing (31, 44).
A murine model of AHO was developed previously by other investigators through disruption of exon 2 of the Gnas gene (31). Homozygosity for the exon 2 disrupted allele led to embryonic lethality and heterozygosity led to extensive perinatal mortality. These heterozygotes had unique phenotypes depending on the parent of origin of the mutated allele: –m/+ offspring had wide bodies and higher birth weights, and 80% of them developed neurological symptoms with delayed cerebellar development and died after 6–21 d. The +/–p newborns had narrow bodies, lower birth weights, and a 77% mortality rate within 24 h of birth from lack of suckling. Those mice that survived weaning lived a normal life span.
The unusual development of neurological dysfunction in the –m/+ mice and the unexpectedly severe phenotype of the +/–p mice suggested that loss of other proteins encoded by Gnas might play an important role in determining the murine phenotype of Gnas in the exon 2 disrupted model. We therefore generated mice carrying a targeted disruption of exon 1 to study the effect of isolated deficiency of Gs. As described for mice with a disruption of exon 2, homozygosity resulted in lethality, and the –m/+ mice had a high perinatal mortality rate. However, unlike mice with targeted disruption of exon 2, the exon 1 –m/+ mice that we and Chen et al. (35) generated did not have severe neurological abnormalities or change in body habitus. This suggests that the predicted loss of NESP55 in exon 2 knockout mice may be the cause of these features because NESP55 is highly expressed in the brain (18, 43). Similar neurological features have not been described in humans with PHP type 1a who inherit a defective maternal GNAS allele with a mutation in exons 2–13, which constitute the 3' untranslated region of the NESP55 transcript (1, 2). Although most of these GNAS mutations are small and therefore may not disrupt the NESP55 transcript, a mutant NESP55 allele of maternal origin has recently been described in one phenotypically normal subject (45). However, in recently reported mice with targeted disruption of Nesp55, mice were found to have abnormal reactivity to new environments (46). Taken together, these observations suggest that NESP55 may be more critical for development of the brain in mice than in humans.
As with the –m/+ mice, the majority of the +/–p mice with the targeted disruption of exon 1 did not die within 24 h, had significantly higher perinatal survival, and had normal body shapes at birth, similar to the characteristics of the Gnas exon 1 knockout mice recently reported by Chen et al. (35). These features contrast with the exon 2 knockout and implicate the loss of XLs as the basis for unusual body habitus, poor suckling behavior, and high perinatal mortality, a hypothesis recently substantiated through the generation of mice with mutations in Gnasxl (47). These mice have poor postnatal growth and survival as well as diverse physiological defects, indicating that XLs is necessary for a variety of postnatal adaptations (47).
We found evidence for incomplete paternal imprinting of Gs in the renal cortex, with a corresponding decrease in PTH-stimulated adenylyl cyclase activity and elevated serum concentrations of PTH in –m/+ mice. Remarkably, serum levels of calcium and phosphorus remained normal despite evidence of PTH resistance in contrast to the findings in the exon 2 knockout mice in which the –m/+ mice, but not the +/–p mice, had decreased serum calcium and elevated serum phosphorus levels (31). The difference between the two mouse models may be secondary to strain variations. Although our mouse model did not exhibit any calcium and phosphorus abnormalities, this is still consistent with the findings in some patients with PHP type 1a. When the exon 1 knockout mice were fed a high-phosphate diet, they had lower calcium and higher phosphorus levels and exaggerated PTH responses, but there still were no significant differences between the calcium and phosphorus levels in wild-type and heterozygous knockout mice.
By contrast to the renal cortex, we found no evidence of imprinting in the renal medulla, consistent with intact vasopressin responsiveness in patients with PHP type 1a and the findings in the exon 2 disrupted mouse (31). Heterozygous +/–p mice had intermediate levels of PTH resistance, which differs from the findings in the exon 2 disrupted +/–p mouse (31). Given that recent studies suggest that XLs can antagonize Gs-dependent signaling pathways (47), it is tempting to speculate that the intermediate reduction in PTH sensitivity in our +/–p mice results from normal expression of XLs, whereas more normal sensitivity to PTH in the exon 2 disrupted +/–p mice results from the predicted absence of XLs expression.
In addition, we found that imprinting in the thyroid was accompanied by TSH resistance in our mouse model, consistent with human thyroid data (28, 30, 32). Although TSH levels were significantly elevated, serum T4 levels were not affected. However, there were no differences in serum TSH, T3, or T4 between exon 2 disrupted –m/+, +/–p, and wt mice (48). The features of our exon 1 knockout mice are consistent with the findings in most patients with PHP type 1a, who have elevated serum levels of TSH but normal levels of serum T4 (1, 2, 7, 49).
Imprinting was clearly tissue specific, similar to the findings of Yu et al. (31). We found no evidence for imprinting of Gs expression in the renal medulla, heart, or adipose. Although we found biallelic expression of Gs in fat tissue from the exon 1 disrupted mice, Yu et al. observed apparent paternal imprinting of Gs in both brown and white adipose tissue from their exon 2 disrupted mice (31) despite normal adrenergic responsiveness of adenylyl cyclase in both –m/+ and +/–p mice (48). Recent studies in normal human adipose tissue have shown that Gs is equivalently expressed from both parental alleles (27), thus supporting the lack of paternal imprinting of Gs in adipose that we observed in our mouse model.
In the recently reported Gnas exon 1 knockout mice (35), the greater severity of the obesity and metabolic abnormalities in the –m/+ mice, compared with the +/–p mice, is consistent with more significant effect of Gs deficiency in the –m/+ mice, possibly due to reduced hypothalamic melanocortin-Gs signaling in these mice (35).
Paternally derived (+/–p) heterozygous mice weighed less than wt mice, similar to the results described in exon 2 knockout mice (2, 48). Recent data in the exon 2 disrupted mice have shown that the +/–p mice have increased insulin sensitivity (50) and increased lipid clearance, and it has been hypothesized that XLs deficiency in the exon 2 knockout mice may be responsible (47, 51). This was substantiated in the Gnas exon 1 knockout mouse recently described (35). Both the –m/+ and +/–p mice exhibited insulin resistance and low metabolic rates. Interestingly, XLs expression is present in our –p/+ mice, which have the same lean phenotype as the exon 2 disrupted mice, thus implying that another mechanism, such as differences in strain and genetic background, may account for the differences between the exon 1 knockout models that we and Chen et al. (35) created.
Maternally derived (–m/+) heterozygous mice had marked weight gain in the males with a trend in the young females. This is consistent with the findings in the recently reported exon 1 knockout mouse (35) and with the phenotype of patients with PHP type 1a.
The Gs-specific targeted disruption of exon 1 resulted in shorter adult mice for both genders in –m/+ and +/–p mice, as in the exon 2 knockout mouse (26). These findings are consistent with the short stature seen in adults with PHP type 1a and pseudoPHP. The basis for short stature in patients with AHO is unknown but appears to be related to premature closure of epiphyses. It has become clear that Gs is a negative regulator of the differentiation of growth plate chondrocytes with intricate interactions with Indian hedgehog and PTHrP (26, 52, 53, 54, 55), and therefore a deficiency of Gs could lead to premature epiphyseal fusion as is seen with overactivation of Indian hedgehog and PTHrP signaling (52, 54, 55). Evidence in humans has revealed biallelic expression of GNAS in bone (27), and in mice, biallelic expression has been demonstrated in chondrocytes (26). Therefore, it is likely that haploinsufficiency of Gs is responsible for the similar occurrence of premature fusion and short stature in both PHP type 1a and pseudoPHP.
In addition, recent studies suggest a possible role for GH deficiency in the development of short stature due to GHRH resistance in some patients with PHP type 1a (6, 9). GH deficiency, however, is not present in pseudoPHP in which there is no hormonal resistance. We found that in the exon 1 knockout mice, the –m/+ females are significantly shorter than their +/–p counterparts. It is possible that patients with PHP type 1a who are GH deficient are shorter than their pseudoPHP counterparts, and this potential difference in phenotype needs to be evaluated to determine the significance of GH deficiency on height.
From analysis of heterozygous mice with targeted disruption of exon 2, there is no obvious skeletal phenotype consistent with AHO on gross examination (26, 31). We also found no skeletal phenotype in the heterozygous exon 1 knockout mouse. By contrast, some epiphyseal and growth plate abnormalities occur in a mouse in which both copies of Gnas were inactivated in chondrocytes via disruption of exon 2 (26, 53). This dosage discrepancy between mouse and human may reflect differences in bone biology because mouse epiphyses, unlike human epiphyses, do not fuse.
It is known that reproductive dysfunction is present in many women with PHP type 1a, and serum levels of LH and FSH may be either normal or elevated (7, 56, 57, 58). By contrast, very little is known about reproductive fitness in men with PHP type 1a. We found both male and female –m/+ mice had reduced fertility. Decreased fertility could be due to paternal imprinting of Gs in the ovary (30) as well as the testis (currently under investigation). The –m/+ females also exhibited very poor mothering skills, but the basis for this defect is presently unknown. The inability to react appropriately to novel stimuli in NESP55-deficient mice suggests that the lack of NESP55 may be responsible for the impaired ability to react to the new situation of motherhood (46).
Our findings indicate that loss of Gs alone is sufficient to cause a murine homolog of AHO. The mice with targeted disruption of exon 1 encoding Gs develop variable phenotypic features that are based on the parent of origin of the disrupted Gnas allele and that reflect the effect of both heterozygous loss of Gs and tissue-specific paternal imprinting. This paternal imprinting is partial and not an all-or-none phenomenon, as has already been shown in a variety of normal human tissues. Our data suggest that loss of NESP55 may lead to the abnormal neonatal phenotype and neurological abnormalities found in the –m/+ mice with targeted disruption of exon 2 and that loss of XLs may lead to the abnormal neonatal phenotype and high perinatal mortality rate in the +/–p exon 2 knockout mice. The physiological functions of NESP55 and XLs have not yet been fully elucidated, but our studies extend prior observations that suggest that these two proteins have more critical roles in mice than humans. The basis for this difference, and the role of imprinting in regulating expression of these two proteins, further extend the complexity of the GNAS locus.
Acknowledgments
The authors are grateful to Dr. John D. Gearhart and Dr. Ann M. Lawler for carrying out blastocyst injections and embryo transfers; Dr. Kimberly O. O’Brien for assistance in analysis of serum calcium levels; Tracey Hatcher for help with bone histomorphometry; and Dr. Amy Wisniewski for advice regarding statistical analyses.
Footnotes
This work was supported by United States Public Health Service Grants RO1 DK34281 and DK56178 (to M.A.L.), Supplement PA-99106 to RO1 DK56178 (to E.L.G.-L.), RO1 AA09000 (to G.W.), National Institutes of Health Research Resources RR00171 (to D.L.H.), National Institutes of Health/National Center for Research Resources Grant MOI RR00052 to the Johns Hopkins University School of Medicine General Clinical Research Center, and a generous donation by the Bosworth family.
Abbreviations: AHO, Albright hereditary osteodystrophy; CD-1, hybrid 129SvEv/CD-1 background; CT, threshold cycle; Gs, -chain of Gs; –m/+, mice with a mutant maternal allele; +/–p, mice with a mutant paternal allele; PHP, pseudohypoparathyroidism; wt, wild-type mice.
References
Levine MA 2002 Pseudohypoparathyroidism. In: Bilezikian JP, Raisz LG, Rodan GA, eds. Principles of bone biology. San Diego: Academic Press; 1137–1163
Weinstein LS, Yu S, Warner DR, Liu J 2001 Endocrine manifestations of stimulatory G protein -subunit mutations and the role of genomic imprinting. Endocr Rev 22:675–705
Albright F, Burnett CH, Smith PH 1942 Pseudohypoparathyroidism: an example of "Seabright-Bantam syndrome." Endocrinology 30:922–932
Davies SJ, Hughes HE 1993 Imprinting in Albright’s hereditary osteodystrophy. J Med Genet 30:101–103
Brickman AS, Carlson HE, Levin SR 1986 Responses to glucagon infusion in pseudohypoparathyroidism. J Clin Endocrinol Metab 63:1354–1360
Germain-Lee EL, Groman J, Crane JL, Jan de Beur SM, Levine MA 2003 Growth hormone deficiency in pseudohypoparathyroidism type 1a: another manifestation of multihormone resistance. J Clin Endocrinol Metab 88:4059–4069
Levine MA, Downs Jr RW, Moses AM, Breslau NA, Marx SJ, Lasker RD, Rizzoli RE, Aurbach GD, Spiegel AM 1983 Resistance to multiple hormones in patients with pseudohypoparathyroidism. Association with deficient activity of guanine nucleotide regulatory protein. Am J Med 74:545–556
Bastepe M, Juppner H 2003 Pseudohypoparathyroidism and mechanisms of resistance toward multiple hormones: molecular evidence to clinical presentation. J Clin Endocrinol Metab. 88:4055–4058 (Editorial)
Mantovani G, Maghnie M, Weber G, De Menis E, Brunelli V, Cappa M, Loli P, Beck-Peccoz P, Spada A 2003 Growth hormone-releasing hormone resistance in pseudohypoparathyroidism type 1a: new evidence for imprinting of the Gs gene. J Clin Endocrinol Metab 88:4070–4074
Albright F, Forbes AP, Henneman PH 1952 Pseudopseudohypoparathyroidism. Trans Assoc Am Physicians 65:337–350
Wilson LC, Trembath RC 1994 Albright’s hereditary osteodystrophy. J Med Genet. 31:779–784 (Review)
Hayward BE, Kamiya M, Strain L, Moran V, Campbell R, Hayashizaki Y, Bonthron DT 1998 The human GNAS1 gene is imprinted and encodes distinct paternally and biallelically expressed G proteins. Proc Natl Acad Sci USA 95:10038–10043
Hayward BE, Moran V, Strain L, Bonthron DT 1998 Bidirectional imprinting of a single gene: GNAS1 encodes maternally, paternally, and biallelically derived proteins. Proc Natl Acad Sci USA 95:15475–15480
Swaroop A, Agarwal N, Gruen JR, Bick D, Weissman SM 1991 Differential expression of novel Gs signal transduction protein cDNA species. Nucleic Acids Res 17:4725–4729
Kehlenbach RH, Matthey J, Huttner WB 1994 XLs is a new type of G protein. Nature 372:804–808
Klemke M, Pasolli HA, Kehlenbach RH, Offermanns S, Schultz G, Huttner WB 2000 Characterization of the extra-large G protein -subunit XLs. II. Signal transduction properties. J Biol Chem 275:33633–33640
Pasolli HA, Klemke M, Kehlenbach RH, Wang Y, Huttner WB 2000 Characterization of the extra-large G protein -subunit XL s. I. Tissue distribution and subcellular localization. J Biol Chem 275:33622–33632
Ischia R, Lovisetti-Scamihorn P, Hogue-Angeletti R, Wolkersdorfer M, Winkler H, Fischer-Colbrie R 1997 Molecular cloning and characterization of NESP55, a novel chromogranin-like precursor of a peptide with 5-HT1B receptor antagonist activity. J Biol Chem 272:11657–11662
Leitner B, Lovisetti-Scamihorn P, Heilmann J, Striessnig J, Blakely RD, Eiden LE, Winkler H 1999 Subcellular localization of chromogranins, calcium channels, amine carriers, and proteins of the exocytotic machinery in bovine splenic nerve. J Neurochem 72:1110–1116
Jan de Beur SJ, Ding C, Germain-Lee E, Cho J, Maret A, Levine MA 2003 Discordance between genetic and epigenetic defects in pseudohypoparathyroidism type 1b revealed by inconsistent loss of maternal imprinting at GNAS1. Am J Hum Genet. 73:314–322.
Liu J, Litman D, Rosenberg MJ, Yu S, Biesecker LG, Weinstein LS 2000 A GNAS1 imprinting defect in pseudohypoparathyroidism type IB. J Clin Invest 106:1167–1174
Liu J, Yu S, Litman D, Chen W, Weinstein LS 2000 Identification of a methylation imprint mark within the mouse Gnas locus. Mol Cell Biol 20:5808–5817
Levine MA, Jap TS, Mauseth RS, Downs RW, Spiegel AM 1986 Activity of the stimulatory guanine nucleotide-binding protein is reduced in erythrocytes from patients with pseudohypoparathyroidism and pseudopseudohypoparathyroidism: biochemical, endocrine, and genetic analysis of Albright’s hereditary osteodystrophy in six kindreds. J Clin Endocrinol Metab 62:497–502
Campbell R, Gosden CM, Bonthron DT 1994 Parental origin of transcription from the human GNAS1 gene. J Med Genet 31:607–614
Zheng H, Radeva G, McCann JA, Hendy GN, Goodyer CG 2001 Gs transcripts are biallelically expressed in the human kidney cortex: implications for pseudohypoparathyroidism type 1b. J Clin Endocrinol Metab 86:4627–4629
Bastepe M, Weinstein LS, Ogata N, Kawaguchi H, Juppner H, Kronenberg HM, Chung UI 2004 Stimulatory G protein directly regulates hypertrophic differentiation of growth plate cartilage in vivo. Proc Natl Acad Sci USA 101:14794–14799
Mantovani G, Bondioni S, Locatelli M, Pedroni C, Lania AG, Ferrante E, Filopanti M, Beck-Peccoz P, Spada A 2004 Biallelic expression of the Gs gene in human bone and adipose tissue. J Clin Endocrinol Metab 89:6316–6319
Germain-Lee EL, Ding CL, Deng Z, Crane JL, Saji M, Ringel MD, Levine MA 2002 Paternal imprinting of Gs in the human thyroid as the basis of TSH resistance in pseudohypoparathyroidism type 1a. Biochem Biophys Res Commun 296:67–72
Hayward BE, Barlier A, Korbonits M, Grossman AB, Jacquet P, Enjalbert A, Bonthron DT 2001 Imprinting of the G(s) gene GNAS1 in the pathogenesis of acromegaly. J Clin Invest 107:R31–R36
Mantovani G, Ballare E, Giammona E, Beck-Peccoz P, Spada A 2002 The Gs gene: predominant maternal origin of transcription in human thyroid gland and gonads. J Clin Endocrinol Metab 87:4736–4740
Yu S, Yu D, Lee E, Eckhaus M, Lee R, Corria Z, Accili D, Westphal H, Weinstein LS 1998 Variable and tissue-specific hormone resistance in heterotrimeric Gs protein -subunit (Gs) knockout mice is due to tissue-specific imprinting of the Gs gene. Proc Natl Acad Sci USA 95:8715–8720
Liu J, Erlichman B, Weinstein LS 2003 The stimulatory G protein -subunit Gs is imprinted in human thyroid glands: implications for thyroid function in pseudohypoparathyroidism types 1A and 1B. J Clin Endocrinol Metab 88:4336–4341
Cattanach BM, Kirk M 1985 Differential activity of maternally and paternally derived chromosome regions in mice. Nature 315:496–498
Peters J, Beechey CV, Ball ST, Evans EP 1994 Mapping studies of the distal imprinting region of mouse chromosome 2. Genet Res 63:169–174
Chen M, Gavrilova O, Liu J, Xie T, Deng C, Nguyen AT, Nackers LM, Lorenzo J, Shen L, Weinstein LS 2005 Alternative Gnas gene products have opposite effects on glucose and lipid metabolism. Proc Natl Acad Sci USA 102:7386–7391
Schwindinger WF, Reese KJ, Lawler AM, Gearhart JD, Levine MA 1997 Targeted disruption of Gnas in embryonic stem cells. Endocrinology 138:4058–4063
Froehlich JC, Wand GS 1997 Adenylyl cyclase signal transduction and alcohol-induced sedation. Pharmacol Biochem Behav 58:1021–1030
Bellorin-Font E, Martin KJ 1981 Regulation of the PTH-receptor-cyclase system of canine kidney: effects of calcium, magnesium, and guanine nucleotides. Am J Physiol 241:F364–F373
Vatner DE, Asai K, Iwase M, Ishikawa Y, Wagner TE, Shannon RP, Homcy CJ, Vatner SF 1998 Overexpression of myocardial Gs prevents full expression of catecholamine desensitization despite increased -adrenergic receptor kinase. J Clin Invest 101:1916–1922
Salomon Y, Londos C, Rodbell M 1974 A highly sensitive adenylate cyclase assay. Anal Biochem 58:541–548
McPherron AC, Lawler AM, Lee SJ 1999 Regulation of anterior/posterior patterning of the axial skeleton by growth/differentiation factor 11. Nat Genet 22:260–264
McCarthy EF, Frassica FF1998 Pathology of bone and joint disorders. Philadelphia: W. B. Saunders Co.; 22–24
Lovisetti-Scamihorn P, Fischer-Colbrie R, Leitner B, Scherzer G, Winkler H 1999 Relative amounts and molecular forms of NESP55 in various bovine tissues. Brain Res 829:99–106
Williamson CM, Ball ST, Nottingham WT, Skinner JA, Plagge A, Turner MD, Powles N, Hough T, Papworth D, Fraser WD, Maconochie M, Peters J 2004 A cis-acting control region is required exclusively for the tissue-specific imprinting of Gnas. Nat Genet 36:894–899
Kim SJ, Gonen D, Hanna GL, Leventhal BL, Cook Jr EH 2000 Deletion polymorphism in the coding region of the human NESP55 alternative transcript of GNAS1. Mol Cell Probes 14:191–194
Plagge A, Isles AR, Gordon E, Humby T, Dean W, Gritsch S, Fischer-Colbrie R, Wilkinson LS, Kelsey G 2005 Imprinted Nesp55 influences behavioral reactivity to novel environments. Mol Cell Biol 25:3019–3026
Plagge A, Gordon E, Dean W, Boiani R, Cinti S, Peters J, Kelsey G 2004 The imprinted signaling protein XLs is required for postnatal adaptation to feeding. Nat Genet 36:818–826
Yu S, Gavrilova O, Chen H, Lee R, Liu J, Pacak K, Parlow AF, Quon MJ, Reitman ML, Weinstein LS 2000 Paternal versus maternal transmission of a stimulatory G-protein subunit knockout produces opposite effects on energy metabolism. J Clin Invest 105:615–623
Werder EA, Illig R, Bernasconi S, Kind H, Prader A 1975 Excessive thyrotropin-releasing hormone in pseudohypoparathyroidism. Pediatr Res 9:12–16
Yu S, Castle A, Chen M, Lee R, Takeda K, Weinstein LS 2001 Increased insulin sensitivity in Gs knockout mice. J Biol Chem 276:19994–19998
Chen M, Haluzik M, Wolf NJ, Lorenzo J, Dietz KR, Reitman ML, Weinstein LS 2004 Increased insulin sensitivity in paternal Gnas knockout mice is associated with increased lipid clearance. Endocrinology 145:4094–4102
Kobayashi T, Chung UI, Schipani E, Starbuck M, Karsenty G, Katagiri T, Goad DL, Lanske B, Kronenberg HM 2002 PTHrP and Indian hedgehog control differentiation of growth plate chondrocytes at multiple steps. Development 129:2977–2986
Sakamoto A, Chen M, Kobayashi T, Kronenberg HM, Weinstein LS 2005 Chondrocyte-specific knockout of the G protein G(s) leads to epiphyseal and growth plate abnormalities and ectopic chondrocyte formation. J Bone Miner Res 20:663–671
Tavella S, Biticchi R, Schito A, Minina E, Di Martino D, Pagano A, Vortkamp A, Horton WA, Cancedda R, Garofalo S 2004 Targeted expression of SHH affects chondrocyte differentiation, growth plate organization, and Sox9 expression. J Bone Miner Res 19:1678–1688
van der Eerden BC, Karperien M, Gevers EF, Lowik CW, Wit JM 2000 Expression of Indian hedgehog, parathyroid hormone-related protein, and their receptors in the postnatal growth plate of the rat: evidence for a locally acting growth restraining feedback loop after birth. J Bone Miner Res 15:1045–1055
Namnoum AB, Merriam GR, Moses AM, Levine MA 1998 Reproductive dysfunction in women with Albright’s hereditary osteodystrophy. J Clin Endocrinol Metab 83:824–829
Shima M, Nose O, Shimizu K, Seino Y, Yabuuchi H, Saito T 1988 Multiple associated endocrine abnormalities in a patient with pseudohypoparathyroidism type 1a. Eur J Pediatr 147:536–538
Wolfsdorf JI, Rosenfield RL, Fang VS, Kobayashi R, Razdan AK, Kim MH 1978 Partial gonadotrophin-resistance in pseudohypoparathyroidism. Acta Endocrinol 88:321–328(Emily L. Germain-Lee, Wil)