Abstract
Background
The stimulatory G-protein α-subunit encoded by GNAS exons 1-13 (GNAS-Gsα) mediates signal transduction of multiple G protein-coupled receptors, including arginine vasopressin receptor 2 (AVPR2). Various germline-derived loss-of-function GNAS-Gsα variants of maternal and paternal origin have been found in pseudohypoparathyroidism type Ia and pseudopseudohypoparathyroidism, respectively. Specific somatic gain-of-function GNAS-Gsα variants have been detected in McCune-Albright syndrome and may result in phosphate wasting. However, no germline-derived gain-of-function variant has been identified, implying that such a variant causes embryonic lethality.Methods
We performed whole-exome sequencing in two families with dominantly inherited nephrogenic syndrome of inappropriate antidiuresis (NSIAD) as a salient phenotype after excluding a gain-of-function variant of AVPR2 and functional studies for identified variants.Results
Whole-exome sequencing revealed two GNAS-Gsα candidate variants for NSIAD: GNAS-Gsα p.(F68_G70del) in one family and GNAS-Gsα p.(M255V) in one family. Both variants were absent from public and in-house databases. Of genes with rare variants, GNAS-Gsα alone was involved in AVPR2 signaling and shared by the families. Protein structural analyses revealed a gain-of-function-compatible conformational property for p.M255V-Gsα, although such assessment was not possible for p.F68_G70del-Gsα. Both variants had gain-of-function effects that were significantly milder than those of McCune-Albright syndrome-specific somatic Gsα variants. Model mice for p.F68_G70del-Gsα showed normal survivability and NSIAD-compatible phenotype, whereas those for p.M255V-Gsα exhibited severe failure to thrive.Conclusions
This study shows that germline-derived gain-of-function rare variants of GNAS-Gsα exist and cause NSIAD as a novel Gsα-mediated genetic disease. It is likely that AVPR2 signaling is most sensitive to GNAS-Gsα's gain-of-function effects.Free full text
Germline-Derived Gain-of-Function Variants of Gsα-Coding GNAS Gene Identified in Nephrogenic Syndrome of Inappropriate Antidiuresis
1 Significance Statement
The stimulatory G-protein α-subunit GNAS-Gsα mediates signal transduction of multiple G protein–coupled receptors, including arginine vasopressin receptor 2 (AVPR2). Specific gain-of-function variants in AVPR2 are known causes of nephrogenic syndrome of inappropriate antidiuresis (NSIAD), an arginine vasopressin–independent antidiuresis. In two families with NSIAD, after excluding AVPR2 gain-of-function variants, the authors identified two novel germline-derived variants of GNAS-Gsα. They also showed that both of the GNAS-Gsα variants had gain-of-function effects that were milder than those of specific somatic GNAS-Gsα variants reported in McCune–Albright syndrome, a condition that may result in renal phosphate wasting. The results refute the widely believed concept that a germline-derived GNAS-Gsα gain-of-function variant is absent because of embryonic lethality and reveal the genetic heterogeneity in NSIAD.
Visual Abstract
Abstract
Background
The stimulatory G-protein α-subunit encoded by GNAS exons 1–13 (GNAS-Gsα) mediates signal transduction of multiple G protein–coupled receptors, including arginine vasopressin receptor 2 (AVPR2). Various germline-derived loss-of-function GNAS-Gsα variants of maternal and paternal origin have been found in pseudohypoparathyroidism type Ia and pseudopseudohypoparathyroidism, respectively. Specific somatic gain-of-function GNAS-Gsα variants have been detected in McCune–Albright syndrome and may result in phosphate wasting. However, no germline-derived gain-of-function variant has been identified, implying that such a variant causes embryonic lethality.
Methods
We performed whole-exome sequencing in two families with dominantly inherited nephrogenic syndrome of inappropriate antidiuresis (NSIAD) as a salient phenotype after excluding a gain-of-function variant of AVPR2 and functional studies for identified variants.
Results
Whole-exome sequencing revealed two GNAS-Gsα candidate variants for NSIAD: GNAS-Gsα p.(F68_G70del) in one family and GNAS-Gsα p.(M255V) in one family. Both variants were absent from public and in-house databases. Of genes with rare variants, GNAS-Gsα alone was involved in AVPR2 signaling and shared by the families. Protein structural analyses revealed a gain-of-function–compatible conformational property for p.M255V-Gsα, although such assessment was not possible for p.F68_G70del-Gsα. Both variants had gain-of-function effects that were significantly milder than those of McCune–Albright syndrome–specific somatic Gsα variants. Model mice for p.F68_G70del-Gsα showed normal survivability and NSIAD-compatible phenotype, whereas those for p.M255V-Gsα exhibited severe failure to thrive.
Conclusions
This study shows that germline-derived gain-of-function rare variants of GNAS-Gsα exist and cause NSIAD as a novel Gsα-mediated genetic disease. It is likely that AVPR2 signaling is most sensitive to GNAS-Gsα’s gain-of-function effects.
The GNAS complex locus generates multiple gene products, including the stimulatory G-protein α-subunit encoded by GNAS exons 1–13 (GNAS-Gsα).1,2 GNAS-Gsα is biallelically expressed in most tissues, including the renal collecting duct, and it is predominantly expressed from the maternal allele in a few tissues, such as the renal proximal tubule, thyroid, gonad, and pituitary.1–5 G-protein α-subunit (Gsα) mediates signal transductions of multiple G protein–coupled receptors, including parathyroid hormone 1 receptor (PTH1R), thyroid-stimulating hormone receptor (TSHR), follicle-stimulating hormone receptor (FSHR), luteinizing hormone/choriogonadotropin receptor (LHCGR), growth hormone–releasing hormone receptor (GHRHR), and arginine vasopressin receptor 2 (AVPR2).6 To date, various germline-derived loss-of-function (LOF) variants of GNAS-Gsα of maternal and paternal origin have been found in pseudohypoparathyroidism type Ia (PHP-Ia) and pseudopseudohypoparathyroidism (PPHP), respectively,1,2,6,7 and specific somatic gain-of-function (GOF) variants of GNAS-Gsα, such as p.R201C and p.R201H, have been detected in McCune–Albright syndrome (MAS) and endocrine tumors.1,6,8 However, no germline-derived GOF variant of GNAS-Gsα has been identified, implying that such a variant causes embryonic lethality.9
Nephrogenic syndrome of inappropriate antidiuresis (NSIAD) is a rare disorder caused by impaired renal capacity to excrete a free water load into the urine under undetectable or very low plasma arginine vasopressin (AVP) values.10 Although this condition is associated with blood hypo-osmolality and hyponatremia as well as inappropriately elevated urine osmolality and sodium concentration as observed in the syndrome of inappropriate antidiuretic hormone secretion (SIADH),11 NSIAD and SIADH show a sharp contrast in the plasma AVP/ADH value. Consistent with such an AVP-independent antidiuresis, NSIAD has been shown to be caused by specific GOF variants of AVPR2, such as p.R137C, p.R137L, and p.F229V.10,12
Here, we report relatively mild germline-derived GOF variants of GNAS-Gsα identified in two families with NSIAD as a salient phenotype. This study demonstrates the presence of germline-derived GOF variants of GNAS-Gsα and establishes a novel Gsα-mediated genetic disease.
Methods
Ethical Approval
Human studies were approved by the Institutional Review Board Committees at Hamamatsu University School of Medicine (91–002) and National Research Institute for Child Health and Development (519), and they were performed after obtaining written informed consent. Mouse studies were approved by the Animal Care and Use Committee of National Research Institute for Child Health and Development (A2014–001), and they were conducted in accordance with the approved protocols.
Oligonucleotides
Oligonucleotides used in this study are shown in Supplemental Table 1.
Molecular Diagnostic Studies
We performed Sanger sequencing, genome-wide array comparative genomic hybridization (aCGH), and whole-exome sequencing (WES) using leukocyte genomic DNA samples.
Sanger sequencing was carried out for all of the coding exons and their splice sites of AVPR2 and for variant-positive GNAS-Gsα exons on the ABI 3130xl Genetic Analyzer (Thermo Fisher Scientific). To confirm heterozygous variants, sequencing was performed for subcloned wild-type (WT) and variant alleles.
Genome-wide aCGH was performed with a catalog human array (1×1M format; ID G4447A) according to the manufacturer’s instructions (Agilent Technologies). Obtained data were analyzed using the default settings of the Agilent Genomic Workbench 7.0. Copy number alterations were regarded as normal variants if they had been registered in Database of Genomic Variants (http://dgv.tcag.ca/dgv/app/home) or ClinVar (http://www.ncbi.nlm.nih.gov/clinvar/).
WES was carried out using SureSelect Human All Exon V6 (Agilent Technologies). Captured libraries were sequenced by NextSeq 500 (Illumina) with 150-bp paired end reads. Reads were aligned to the reference genome (Human GRCh37/hg19; http://genome.ucsc.edu/) using BWA-MEM (Version 0.7.12) with default parameters. Duplicated reads were removed by Picard (Version 2.9.2), and local realignment and base quality recalibration were performed by GATK Version 3.7. Variants were identified with the GATK HaplotypeCaller, and those with minor allele frequencies of <0.005 in all of the following public databases (the whole-genome and exome data for the East Asian population in the Genome Aggregation Database [gnomAD_genome_EAS and gnomAD_exome_EAS],13 the Human Genetic Variation Database,14 allele frequency data of 2049 Japanese individuals,15 and in-house database) were selected as rare variants. Final variants were annotated with Annovar.16 In silico pathogenicity prediction for identified rare variants was carried out with Combined Annotation–Dependent Depletion,17 Polyphen-2 HumVar,18 Sorting Intolerant From Tolerant,19 and MutationTaster.20 We further examined the involvement of genes with rare variants in the AVPR2 signaling pathway.21,22
Protein Structural Analyses
GNAS-Gsα variants were introduced into the crystal structures of the Gα-containing complexes, including the GDP-bound G(i)αβγ-heterotrimer (Protein Dank Bank [PDB] ID code 1gg2),23 the nucleotide-free G(s)αβγ-heterotrimer in complex with an agonist-occupied monomeric β2-adrenergic receptor (PDB ID code 3sn6),24 and the GTP analog (GTPγS)-bound G(s)α in complex with adenylate cyclase (PDB ID codes 1azs25 and 1tl726). Free energy changes caused by the found variants of the Gα-containing complexes were calculated with FoldX software (Version 4.0β).27 In this calculation, ligand molecules, including guanine nucleotides, were ignored.
In Vitro Functional Studies
Functional studies were performed for GNAS-Gsα variants identified in this study and p.R137L variant of AVPR2 identified in NSIAD10 using the Dual Luciferase Reporter Assay System (Promega) (Supplemental Material).
Model Mouse Experiments
Model mice for GNAS-Gsα variants were generated by genome editing with the CRISPR/Cas9 method (Supplemental Material), and they were studied for phenotypes after maintenance by backcross with a WT C57BL/6 strain for multiple generations. We first examined a general condition since birth, fertility at 8–10 weeks of age, and body weight and amount of free oral water intake (monitoring for 2 days) at 5 months of age. We next performed an acute water load test (a volume of 20 µl/g body wt water administrated via a gastric tube) at 5 months of age; blood samples were obtained by cardiac puncture under deep anesthesia in a nonwater-loaded condition and a postwater-loaded condition (1 hour after the water load) from different groups of mice (nonpaired samples), whereas urine samples were collected by forced excretion with handling stimulation in prewater-loaded and postwater-loaded conditions from the same group of mice (paired samples).
Statistical Analyses
The variables were expressed as the mean±SEM. For statistical analyses, we first examined normality by the chi-squared test. For nonpaired samples following normality, we used t tests for two groups with similar variances and Welch t tests for two groups with different variances after comparing variances by the F test. For nonpaired samples not following normality, we used Mann–Whitney U tests. We also used paired t tests for paired samples following normality and Wilcoxon signed rank tests for paired samples not following normality. Statistical significance of the frequency was analyzed by the chi-squared test. P<0.05 was considered significant.
Results
Clinical Report
We encountered two Japanese families with dominantly inherited NSIAD (Figure 1A). Representative water metabolism data are shown in Figure 1B, and clinical findings related to Gsα-mediated signal transductions are summarized in Table 1, with detailed information provided in Supplemental Table 2.
Clinical studies indicating NSIAD-compatible findings in six subjects (A-I-2, A-II-2, A-III-3, and A-III-4; B-II-2 and B-III-2), uncertain findings in three subjects (A-III-2; B-I-2 and B-III-1), and normal findings in two subjects (A-III-1; B-II-1). (A) The pedigrees of family A and family B. The arrows indicate the probands. (B) Representative water metabolism data. Serum sodium concentrations represent the data obtained by random samplings (in subjects who received multiple random samplings, the lowest values are shown); those obtained shortly after hyponatremic convulsions in A-III-3, A-III-4, and B-III-2 together with the effects of fluid restriction after the convulsions in A-III-3 and B-III-2; and those obtained before the water load test and at the time of the lowest serum osmolality during the water load test in A-I-2, A-II-2, A-III-1, A-III-2, and A-III-3. The relationships between plasma arginine vasopressin (AVP) and serum or urine osmolality are on the basis of the data obtained from the water load test in A-I-2, A-II-2, A-III-1, A-III-2, and A-III-3; shortly after hyponatremic convulsions in A-III-3, A-III-4, and B-III-2; and by random sampling in five subjects of family B. The light blue areas represent the reference ranges for serum sodium in healthy Japanese children (138–145 mEq/L), which are a bit different from those of healthy Japanese adults (136–146 mEq/L),37,38 and the relationship between plasma AVP and serum or urine osmolality (the ranges of low-plasma AVP values [<3.0 pg/ml] are illustrated).39 UV/WI (percentage) indicates the ratios of urine volume to water intake in the acute oral water load test.29 NSIAD, nephrogenic syndrome of inappropriate antidiuresis.
Table 1.
Summary of clinical findings
| Subject | A-I-2 | A-II-2 | A-III-1 | A-III-2 | A-III-3 | A-III-4 | B-I-2 | B-II-1 | B-II-2 | B-III-1 | B-III-2 |
|---|---|---|---|---|---|---|---|---|---|---|---|
| GNAS-Gsα | VTa | VTa | WT | VTa | VTa | VTa | NE | WT | VTb | VTb | VTb |
| Sex | W | W | M | W | W | W | W | M | W | M | W |
| Present age, yr | 77 | 50 | 31 | 28 | 22 | 13 | 69 | 54 | 42 | 15 | 9 |
| NSIAD (AVPR2 related) | Yes | Yes | No | Uncertain | Yes | Yes | Uncertain | No | Yes | Uncertain | Yes |
| Febrile and/or afebrile convulsions with hyponatremia | + | + | − | − | + | + | − | − | − | − | + |
| Hyponatremia | + | + | − | − | + | + | − | − | + | ± | + |
| Low serum osmolality under low plasma AVP | + | + | − | − | + | + | − | − | + | ± | + |
| High urine osmolality under low plasma AVP | + | + | − | + | + | + | ± | − | + | + | + |
| Amelioration of hyponatremia by fluid restrictionc | NE | NE | NE | NE | + | NE | NE | NE | NE | NE | + |
| Urine excretion failure in an acute oral water load testd | + | + | − | + | + | NE | NE | NE | NE | NE | NE |
| Growth pattern (GHRHR related) | WNR | WNR | WNR | WNR | WNR | WNR | WNR | WNR | WNR | WNR | WNR |
| Gonadal function (FSHR/LHCGR related) | WNR (fertile) | WNR (fertile) | NE | WNR | WNR | WNR | WNR (fertile) | WNR (fertile) | WNR (fertile) | NE | WNR |
| Calcium metabolism (PTH1R related) | WNR | WNR | NE | WNR | WNR | Hypocalciuria (subclinical) | NE | NE | NE | NE | WNR |
| Thyroid function (TSHR related) | WNR | Elevated (subclinical) | WNR | WNR | Elevated (subclinical) | Elevated (subclinical) | NE | NE | NE | NE | Elevated (subclinical) |
| MAS findings | − | − | − | − | − | − | − | − | − | − | − |
| Other abnormal phenotype | − | − | − | − | − | − | − | − | − | − | − |
Besides the G-protein-coupled receptors (GPCRs) shown in this table, Gsa mediates signal transductions of other GPCRs. GNAS-Gsα, G-protein α-subunit encoded by GNAS exons 1–13; VT, variant; WT, wild-type; NE, not examined; W, woman; M, man; NSIAD, nephrogenic syndrome of inappropriate antidiuresis; AVPR2, arginine vasopressin receptor 2; AVP, arginine vasopressin; GHRHR, growth hormone–releasing hormone receptor; WNR, within the normal range; FSHR, follicle-stimulating hormone receptor; LHCGR, luteinizing hormone/choriogonadotropin receptor; PTH1R, parathyroid hormone 1 receptor; TSHR, thyroid-stimulating hormone receptor; MAS, McCune–Albright syndrome; +, positive; +/−, equivocal; −, negative.
Family A has been reported previously28 before the birth of A-III-4. The proband (A-III-3) had five episodes of febrile and afebrile convulsions with hyponatremia up to 4 years of age and was found to have NSIAD-compatible blood and urine laboratory findings (obvious blood hypo-osmolality and hyponatremia and inappropriately high urine osmolality and sodium concentration under very low plasma AVP values). Hyponatremia was ameliorated by fluid restriction (60 ml/kg per day) for 2 days. A-III-3 was placed on fluid restriction as recommended for chronic hyponatremia11 and had no convulsions thereafter. Allegedly, A-I-2 and A-II-2 also had afebrile hyponatremic convulsions several times, and A-II-2 and A-III-2 sometimes felt mild finger trembling. Thus, an acute oral water load test (20 ml/kg p.o. over 30 minutes; blood and urine samplings at 60- and 30-minute intervals, respectively, for 4 hours)29 was performed, revealing unequivocal NSIAD findings and reduced urine excretion in A-I-2, A-II-2, and A-III-3; uncertain findings in A-III-2 (normal blood sodium concentration and osmolality, high urine osmolality and sodium concentration inappropriate for plasma AVP, and reduced urine excretion) (more information regarding the interpretation of the previous data in A-III-228 is given at the asterisk in Supplemental Table 2); and normal findings in A-III-1. Fluid restriction was also recommended for A-I-2 and A-II-2. Subsequently, A-III-4 was born and had nine episodes of febrile and afebrile hyponatremic convulsions up to 4 years of age. A-III-4 also received a diagnosis of NSIAD and was managed by fluid restriction. A-I-2, A-II-2, A-III-2, A-III-3, and A-III-4 seldom felt thirsty in their daily lives and almost never enjoyed drinking water. Nevertheless, NSIAD-compatible laboratory data were repeatedly obtained by random sampling in A-I-2, A-II-2, A-III-3, and A-III-4. Pituitary-adrenal, renin-aldosterone, and renal functions were normal in all of the subjects examined.
Family B was identified by NSIAD-compatible laboratory findings of the proband (B-III-2) when she had two episodes of afebrile convulsions at 6 months of age. The hyponatremia and low serum osmolality were ameliorated by fluid restriction (60 ml/kg per day) for 2 days. B-III-2 was placed on fluid restriction and had no convulsions thereafter. Allegedly, B-I-2, B-II-2, and B-III-1 took just a small amount of fluid in their daily lives, although they had no clinical symptoms, such as afebrile convulsions. Laboratory studies revealed NSIAD-compatible findings in B-II-2, uncertain findings in B-III-1 (borderline low blood sodium concentration and osmolality and high urine osmolality inappropriate for plasma AVP) and B-I-2 (borderline high urine osmolality for plasma AVP only), and normal findings in B-II-1. Fluid restriction was also performed for B-II-2 and B-III-1.
After the identification of GNAS-Gsα variants, we examined clinical and laboratory findings relevant to Gsα-mediated signal transductions in variant-positive subjects (Supplemental Table 2, Table 1).1,6 Growth patterns related to GHRHR signaling and gonadal functions related to FSHR/LHCGR signaling were normal in all of the variant-positive subjects examined, whereas subclinical hypocalciuria related to PTH1R signaling was detected in a single subject (A-III-4); subclinical nonautoimmune hyperthyroidism related to TSHR signaling was identified in four subjects (A-II-2, A-III-3, A-III-4, and B-III-2). Clinical features characteristic of MAS, such as peripheral precocious puberty, café au lait skin pigmentation, and polyostotic fibrous dysplasia, were absent from all of the variant-positive subjects.
Molecular Diagnostic Studies
Conventional Sanger direct sequencing revealed no variant on the coding exons and their splice sites of AVPR2, and genome-wide aCGH showed no pathogenic copy number variant in the probands of family A and family B.
Thus, we performed WES of six subjects in family A and four subjects in family B (Figure 1A). The results are summarized in Supplemental Table 3 (the URLs used are shown in the footnotes). Here, considering the possibility of variable expressivity and reduced penetrance, we first examined the WES data of obviously affected subjects only. Consequently, it was revealed that 23 rare variants, including four variants completely absent from the databases, were shared by A-I-2, A-II-2, A-III-3, and A-III-4 and that 101 rare variants, including 23 variants completely absent from the databases, were shared by B-II-2 and B-III-2 (the presence or absence of these variants in the remaining subjects examined is also shown in Supplemental Table 3). Of these rare variants, an in-frame deletion variant [NM_000516.5:c.201_209del, p.(F68_G70del)] on exon 2 and a missense variant [NM_000516.5:c.763A>G, p.(M255V)] on exon 10 of GNAS-Gsα were extracted as the most likely disease-causing variants in family A and family B, respectively (Figure 2). Indeed, both variants were completely absent from the databases, and although the pathogenicity of p.(F68_G70del) was unknown, that of p.(M255V) was assessed to be high. Furthermore, of genes with the rare variants, GNAS-Gsα alone was shared by family A and family B, and it was found to reside on the AVP signaling pathway. The GNAS-Gsα variants were present in subjects with uncertain clinical findings and absent from those with normal findings in both family A and family B (Figure 1A).
Variants of GNAS exons 1–13 encoding G-protein α subunit (GNAS-Gsα) identified in family A and family B. (A) Electrochromatograms of GNAS-Gsα variants identified in A-III-3 (deletion of a 9-bp segment is highlighted with a light yellow rectangle between normal sequences highlighted with light blue and light orange rectangles) and B-III-2 (a missense substitution at the position indicated with red asterisks). The p.(F68_G70del) variant has also been revealed in A-I-2, A-II-2, A-III-2, and A-III-4, and the p.(M255V) variant has also been detected in B-II-2 and B-III-1. (B) Assessment of the two variants. The American College of Medical Genetics and Genomics (ACMG) criteria (PS3, PM2, PM4, PP1, PP3, and PP4) are in the ACMG Standards and Guidelines.31 The p.(F68_G70del) variant can be classified as pathogenic if the segregation data of family A with multiple subjects are regarded as stronger (moderate) evidence (PP1*).
Protein Structural Analyses
The M255 residue at the edge of switch III was found to make van der Waals contacts between its side chain and the methylene parts of the side chain of the R265 residue, thereby contributing to the stabilization of the P-loop conformation via hydrogen bonding between the R265 residue and the P-loop (Figure 3A). Thus, the replacement of the M255 residue with a β-branched residue valine (p.M255V) would more or less compromise the GTPase activity by indirectly destabilizing the P-loop conformation that plays a critical role in the development of the GTPase activity of G proteins.30 Meanwhile, the free energy change caused by p.M255V was estimated to be small by FoldX software, suggesting that this variant affects neither structural integrity of the Gsα nor the adenylate cyclase activation (Figure 3B). Collectively, p.M255V-Gsα was considered to have a GOF effect leading to enhanced cAMP production (Supplemental Figure 1). By contrast, the effect of the p.F68_G70del on the structure of the G(s)α-protein could not be evaluated, because the position of p.F68_G70del mapped within the crystallographically invisible parts of the G(s)α-subunit in the complexed states (Figure 3C). Similar protein structural analyses indicated severe GOF effects for Gsα-variants identified in MAS and obvious LOF effects for those found in PHP-Ia/PPHP (Supplemental Figure 1).
Protein structural analyses indicating a gain-of-function effect of p.M255V-Gsα and an uncertain effect of p.F68_G70del-Gsα. (A) Mapping of the p.M255 residue on the GDP-bound G(i)αβγ heterotrimer (Protein Data Bank [PBD] ID code 1gg2; www.rcsb.org) [the GDP-bound G(s)αβγ heterotrimer is not available, and the M255 residue of the G(s)α-subunit corresponds to the L232 residue of the G(i)α-subunit], the G(s)αβγ-β2-adrenergic receptor (β2AR) complex (PDB ID code 3sn6), and the GTP analog (guanosine 5ʹ-O-(3-thio)triphosphate [GTPγS])-bound G(s)α-adenylate cyclase (AC) complex (PDB ID code 1azs). The nucleotide exchange and hydrolysis cycle of G proteins is shown at the top. The M255 residue of the G(s)α-subunit and the L232 residue of the G(i)α-subunit are shown as van der Waals spheres in red. The bound nucleotides are shown as orange sticks. In the rightmost inset, the side chain of the R265 residue is shown as sticks with translucent van der Waals spheres. The black dotted lines represent hydrogen bonds. (B) In silico estimation of the free energy change on the variant in each of the Gα-containing complexes. (C) Mapping of the p.F68_G70 residues on the secondary structures (the upper part; the arrows, cylinders, and thick lines represent sheets, helices, and loops, respectively) and the crystal structures (the lower part) of the G(s)αβγ-β2AR complex (PDB ID code 3sn6) and the GTP analog (GTPγS)-bound G(s)α-AC complexes (PDB ID codes 1azs and 1tl7). The p.F68_G70 residues (shown in red) are located within unassigned parts of the G(s)α-containing complexes due to missing electron density (shown with dotted lines).
In Vitro Functional Studies
The results of dual luciferase reporter assays are shown in Figure 4, and statistical data are described in Supplemental Table 4. Luciferase activity was similar between p.F68_G70del-Gsα and p.M255V-Gsα and between p.R201C-Gsα and p.R201H-Gsα at the three different AVP dosages, whereas it tended to be higher for p.M255V-Gsα than for p.F68_G70del-Gsα in the absence of AVP and for p.R201H-Gsα than for p.R201C-Gsα at the three different AVP dosages. Notably, luciferase activity was significantly higher for p.F68_G70del-Gsα/p.M255V-Gsα than for WT-Gsα in the absence of AVP, whereas it was similar between p.F68_G70del-Gsα/p.M255V-Gsα and WT-Gsα in the presence of low and high AVP dosages. Furthermore, luciferase activity was significantly higher for p.R201C-Gsα/p.R201H-Gsα than for p.F68_G70del-Gsα/p.M255V-Gsα at the three different AVP dosages, except for the similar activities between p.M255V-Gsα and p.R201C-Gsα in the absence of AVP and between p.F68_G70del-Gsα/p.M255V-Gsα and p.R201C-Gsα in the presence of a high AVP dosage. In addition, luciferase activity was increased in an AVP dosage–dependent manner, most clearly for WT-Gsα followed by p.F68_G70del-Gsα/p.M255V-Gsα and most mildly for p.R201C-Gsα/p.R201H-Gsα. Similar luciferase reporter assays showed that NSIAD-specific p.R137L-AVPR2 had a constitutive activation function in the absence of AVP and virtually lost responsiveness to AVP (Supplemental Figure 2).
In vitro functional studies showing mild gain-of-function effects of p.F68_G70del-Gsα and p.M255V-Gsα. (A) Schematic representation of the luciferase assays and the vectors used in this study. Arginine vasopressin (AVP) binding to arginine vasopressin receptor 2 (AVPR2) facilitates the coupling of AVPR2 with intracellular heterotrimeric G protein and the exchange of GDP for GTP on the G-protein α-subunit (Gsα)–subunit. The GTP-bound Gsα-subunit dissociates from the β/γ dimer and AVPR2 and activates adenylate cyclase (AC), thereby producing cAMP from ATP. Then, cAMP phosphorylates cAMP response element-binding protein (CREB) and promotes the expression of cAMP response element (CRE)–associated luciferase gene. CMV, cytomegalovirus promoter; HA, hemagglutinin-tag; Luc, luciferase gene. (B) Transactivation functions of various Gsα-proteins (p.F68_G70del and p.M255V identified in nephrogenic syndrome of inappropriate antidiuresis [NSIAD] and p.R201C and p.R201H reported in McCune–Albright syndrome [MAS]). AVP (+) and AVP (++) indicate AVP concentrations of 0.0001 and 0.001 IU/ml, respectively, in the culture media. Statistical data are shown for luciferase activities in the absence of AVP, reflecting constitutional activation functions of various Gsα-proteins (detailed statistical data are shown in Supplemental Table 4). The luciferase activities obtained with an empty expression vector lacking wild-type (WT) or variant (VT) GNAS-Gsα cDNA are predicted to be due to endogenous Gsα of COS7 cells. ns, not significant. *P<0.05; **P<0.01.
Model Mouse Experiments
We first studied model mice for p.F68_G70del-Gsα (Gnas-GsαΔ). Mice homozygous for Gnas-GsαΔ (Gnas-GsαΔ/Δ) and those heterozygous for Gnas-GsαΔ (Gnas-GsαWT/Δ) were apparently healthy with no discernible abnormality since birth as were WT mice (Gnas-GsαWT/WT). At 8–10 weeks of age, mating between heterozygous Gnas-GsαWT/Δ female and male mice produced offspring with a genotype frequency consistent with the Mendelian mode of inheritance (Supplemental Table 5), and Gnas-GsαΔ/Δ mice bred offspring after mating with Gnas-GsαWT/WT mice of different sex (not shown). At 5 months of age, body weight and daily water intake were similar among Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice (Figure 5A, Supplemental Table 6).
Model mouse for p.F68_G70del-Gsα exhibiting NSIAD-like findings. (A) Body weight and free oral water intake of Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice (mean±SEM). (B) Representative results of the acute water load test (mean±SEM). The data with statistically significant difference (P<0.05) are indicated (detailed statistical data are shown in Supplemental Table 7). WT, wild-type.
We next performed an acute water load test on Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice of age 5 months. The results are shown in Figure 5B (urine calcium values are not shown, because they were below the detection limit in multiple samples irrespective of the mouse genotype), and statistical data are described in Supplemental Table 7. In a nonwater- or prewater-loaded condition, most serum and urine biochemical concentrations were similar among Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice of both sexes. The acute water load resulted in significant alterations in several serum biochemical concentrations, including slightly but significantly decreased sodium concentrations, and significant reductions in most urine biochemical concentrations, including osmolality and sodium concentrations, in Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice of both sexes. The degree of reduction in urine biochemical concentrations was large in Gnas-GsαWT/WT mice, intermediate in Gnas-GsαWT/Δ mice, and small in Gnas-GsαΔ/Δ mice of both sexes. Consequently, in a postwater-loaded condition, although serum biochemical concentrations were similar among Gnas-GsαWT/WT, Gnas-GsαWT/Δ, and Gnas-GsαΔ/Δ mice of both sexes, female urine osmolality and male urine sodium concentration were significantly higher in Gnas-GsαWT/Δ mice than in Gnas-GsαWT/WT mice, and urine osmolality and potassium, inorganic phosphorus, and creatinine concentrations were significantly higher in Gnas-GsαΔ/Δ mice than in Gnas-GsαWT/Δ and Gnas-GsαWT/WT mice of both sexes. These findings indicated that Gnas-GsαΔ resulted in compromised renal capacity to excrete a free water load into the urine in an apparently Gnas-GsαΔ allele frequency–dependent manner.
We also generated model mice for p.M255V-Gsα (Gnas-GsαM255V). Although mating between heterozygous Gnas-GsαWT/M255V female and male mice ages 8–10 weeks old produced offspring with different genotypes, Gnas-GsαM255V/M255V mice showed severe failure to thrive and often died shortly after birth; such findings were less severely observed in Gnas-GsαWT/M255V mice, regardless of the parental origin of the Gnas-GsαM255V allele. Consequently, the genotype frequency of surviving pups did not follow the Mendelian mode of inheritance at approximately 24 and 48 hours of age (Supplemental Table 5). However, the number of embryos conceived after mating between heterozygous female and male mice was consistent with the Mendelian mode of inheritance at 18.5 days postcoitum (Supplemental Table 5), although Gnas-GsαM255V/M255V and Gnas-GsαWT/M255V embryos exhibited apparently severe and moderate growth failure, respectively. These findings indicated an apparently Gnas-GsαM255V allele frequency–dependent nonimprinted deleterious effect on survivability in the extrauterine environment. Thus, we could not perform a detailed examination, including an acute water load test, on Gnas-GsαM255V/M255V mice as well as Gnas-GsαWT/M255V mice.
Discussion
We identified two novel germline-derived rare variants of GNAS-Gsα in subjects with variably manifested NSIAD. Subsequently, protein structural analyses indicated a GOF-compatible conformational property for p.M255V-Gsα, although the effect of p.F68_G70del on the Gsα-protein structure remained unknown. Luciferase assays showed that both p.F68_G70del-Gsα and p.M255V-Gsα had constitutive activation functions in the absence of AVP, with luciferase activities being significantly milder in p.F68_G70del-Gsα than in MAS-specific p.R201C-Gsα/p.R201H-Gsα and in p.M255V-Gsα than in MAS-specific p.R201H-Gsα. Furthermore, model mice for p.F68_G70del showed normal survivability and reproduced NSIAD-compatible phenotype in an apparently Gnas-GsαΔ allele frequency–dependent manner, although those for p.M255V showed reduced extrauterine viability and failure to thrive. Collectively, these findings argue that the two GNAS-Gsα variants are true disease-causing variants with mild GOF effects and that such mild GOF variants of GNAS-Gsα permit survival in the human even if they are derived from the germline and cause NSIAD as the salient phenotype. In this regard, the variable expressivity of NSIAD would primarily be due to the effects of other genetic and environmental factors. According to the ACMG Standards and Guidelines,31 p.F68_G70del and p.M255V are regarded as a “pathogenic” variant and a “likely pathogenic” variant, respectively, if the segregation data of family A are regarded as stronger evidence (Figure 2).
NSIAD was the sole clinically recognizable phenotype in the two families, whereas detailed clinical and laboratory studies in the GNAS-Gsα variant-positive subjects revealed subclinical hyperthyroidism in four subjects and hypocalciuria in a single subject. This implies that, of Gsα-mediated signaling, AVPR2 signaling is most sensitive to the mild GNAS-Gsα GOF effects followed by TSHR signaling and PTH1R signaling and that other signaling, including MAS-related signaling, is almost, if not totally, nonresponsive to the mild GOF effects. Notably, NSIAD has not been found in MAS,1,6,8 although MAS-specific GNAS-Gsα variants showed marked luciferase activities for AVPR2 signaling. Thus, it might be possible that the MAS-specific severe somatic GOF variants of GNAS-Gsα cause embryonic lethality when present in vital tissues, including the collecting duct and other developmentally related tissues, and permit survival when primarily present in nonvital tissues, such as the skin, bone, and gonad. In this case, it is assumed that MAS-causing GOF variants of GNAS-Gsα exert more potent effects on AVPR2 signaling than apparently nonlethal NSIAD-causing GOF variants of AVPR2.10,12 The above notion remains purely speculative, and additional studies are necessary to clarify why NSIAD is absent from MAS.
The GNAS-Gsα variants identified in this study were derived from the mothers in all of the subjects in whom parental origin of the variants was identified. The maternal origin of the variants would be coincidental for the development of NSIAD. Because PHP-Ia and PPHP are free from polyuria in the presence of a single copy of functional GNAS-Gsα of paternal or maternal origin, respectively,1,2,6,7 this argues against predominant maternal expression of GNAS-Gsα in the renal collecting duct. Furthermore, mouse Gnas-Gsα is known to be biallelically expressed in the collecting duct.1,3 Rather, because clinical features related to GHRHR, FSHR/LHCGR, PHP1R, and TSHR signal transductions remained undetected or subclinical in the variant-positive subjects despite the preferential maternal expression of GNAS-Gsα in the pituitary, gonad, renal proximal tubule, and thyroid,1–5 this would provide additional support for low or nonresponsiveness of such signal transductions to the mild GNAS-Gsα GOF effects.
Several matters should be pointed out in this study. First, A-III-3 and A-III-4 had febrile as well as afebrile convulsions with hyponatremia. Although fever is known to increase AVP release via ΔN-TRPV1,32 luciferase assays indicated diminished responsiveness of p.F68_G70del-Gsα/p.M255V-Gsα to AVP. Thus, it is unknown whether the fever enhanced hyponatremia in A-III-3 and A-III-4. Second, the variant-positive subjects seldom felt thirsty and took just a small amount of fluid. This would reflect a physiologic response to hyponatremia and hypo-osmolality,33 although detailed neuroimaging-based studies for thirst34,35 have not been performed. Third, the two variants would also be present on other GNAS transcripts, including maternally expressed NESP55 and paternally expressed XLαs and A/B.2 In this context, lack of clinically discernible features other than NSIAD in subjects with maternally inherited variants would argue against a phenotypic effect of a probably altered NESP55 function. Fourth, p.F68_G70del-Gsα had a GOF effect. This implies that even a crystallographically invisible and structurally flexible region can have an important biologic function. Fifth, in the luciferase assays, an AVP dosage effect was remarkable for WT-Gsα, moderate for p.F68_G70del-Gsα/p.M255V-Gsα, mild for p.R201C-Gsα/p.R201H-Gsα, and absent for p.R137L-AVPR2. This would suggest that the AVP responsiveness is reduced as the constitutive activation function of Gsα increases and virtually abolished when the receptor itself becomes constitutively active. Sixth, Gnas-GsαΔ of maternal origin could have a certain effect in a few tissues, including the renal proximal tubule, where Gnas-Gsα shows maternal expression.1–5 This might more or less have contributed to the significantly increased urine inorganic phosphorus concentration in Gnas-GsαΔ/Δ mice, although increased urine inorganic phosphorus concentration is simply explained by exaggerated water resorption in the collecting duct. Seventh, mice with Gnas-GsαM255V exhibited reduced survivability and failure to thrive. In this regard, the luciferase activity in the absence of AVP was somewhat higher for p.M255V-Gsα than for p.F68_G70del-Gsα, and it was not significantly different between p.M255V-Gsα and p.R201C-Gsα. Such a degree of constitutive activation function, although it had no clinically recognizable effect other than NSIAD in variant-positive subjects of family B, could underlie the phenotypic consequences in mice with Gnas-GsαM255V.
Recently, Biebermann et al.36 have identified an identical de novo GNAS-Gsα p.F376V variant on the maternally derived alleles of two unrelated boys with clinical features suggestive of both GOF and LOF functions of GNAS-Gsα, including NSIAD, gonadotropin-independent precocious puberty, skeletal dysplasia, and PTH resistance in the proximal but not distal tubules. In vitro studies for the p.F376V-Gsα showed enhanced ligand-independent AVPR2, LHCGR, and PTH1R signaling and blunted ligand-dependent responses. The development of NSIAD in the two boys would provide additional support for the notion that relatively mild GOF variants of GNAS-Gsα permit survival and cause NSIAD when present in the collecting duct. Furthermore, the presence of the NSIAD-causing p.F376V variant on the maternally inherited allele may raise the possibility that the maternal transmission of the p.F68_G70del and p.M255V variants is not necessarily coincidental but has some specific effects on the water metabolism via a signal from tissues with maternal GNAS-Gsα expression as implicated by Biebermann et al.36
In summary, this study shows for the first time that germline-derived GNAS-Gsα GOF variants do exist and cause NSIAD as the salient phenotype. This study establishes a novel Gsα-mediated genetic disease and provides critical information on Gsα-mediated signal transductions.
Acknowledgments
We thank Prof. Yutaka Oki for his critical advice on the arginine vasopressin measurement; Drs. Fumiko Kato, Shinichi Nakashima, Kaori Yamoto, Yasuko Fujisawa, Noriyuki Takubo, Satoshi Narumi, Kenji Miyado, and Masayo Kagami for their technical assistance; Dr. Hiroshi Yoshida for introducing family B; and Dr. Midori Awazu for her supervision throughout this study.
This work was supported by Grant-in-Aid for Scientific Research on Innovative Areas JP17H06428 (to M.Fukami and T.Ogata) and Grant-in-Aid for Scientific Research (B) JP17H04204 (to T.Ogata) from the Japan Society for the Promotion of Science and by grants JP18ek1009278 (to Y. Matsubara), JP17ek0109297 (to H.Saitsu), and JP18ek0109301 (to T.Ogata) from the Japan Agency for Medical Research and Development.
Footnotes
Published online ahead of print. Publication date available at www.jasn.org.
See related editorial, “GNAS: A New Nephrogenic Cause of Inappropriate Antidiuresis,” on pages 722–725.
Supplemental Material
This article contains the following supplemental material online at http://jasn.asnjournals.org/lookup/suppl/10.1681/ASN.2018121268/-/DCSupplemental.
Supplemental Figure 1. Structural analyses for Gsα-variant proteins.
Supplemental Figure 2. Dual luciferase assays for the nephrogenic syndrome of inappropriate antidiuresis-causing p.R137L variant of AVPR2.
Supplemental Table 1. Oligonucleotides used in this study.
Supplemental Table 2. Clinical and biochemical findings.
Supplemental Table 3. Rare variants shared by subjects with obvious nephrogenic syndrome of inappropriate antidiuresis.
Supplemental Table 4. P values for the differences in luciferase reporter assays.
Supplemental Table 5. Numbers of pups and embryos produced by mating between heterozygous female and male mice.
Supplemental Table 6. P values for the differences in body weight and daily water intake.
Supplemental Table 7. P values for the differences in biochemical values obtained from an acute water load test in mice.
References
Articles from Journal of the American Society of Nephrology : JASN are provided here courtesy of American Society of Nephrology
Full text links
Read article at publisher's site: https://doi.org/10.1681/asn.2018121268
Read article for free, from open access legal sources, via Unpaywall:
https://jasn.asnjournals.org/content/jnephrol/30/5/877.full.pdf
Citations & impact
Impact metrics
Citations of article over time
Alternative metrics
Smart citations by scite.ai
Explore citation contexts and check if this article has been
supported or disputed.
https://scite.ai/reports/10.1681/asn.2018121268
Article citations
Choosing the Best Tissue and Technique to Detect Mosaicism in Fibrous Dysplasia/McCune-Albright Syndrome (FD/MAS).
Genes (Basel), 15(1):120, 18 Jan 2024
Cited by: 0 articles | PMID: 38255009 | PMCID: PMC10815810
V2 vasopressin receptor mutations: future personalized therapy based on individual molecular biology.
Front Endocrinol (Lausanne), 14:1173601, 24 May 2023
Cited by: 1 article | PMID: 37293495 | PMCID: PMC10244717
Review
Free full text in Europe PMCIntractable hyponatremia complicated by a reset osmostat: a case report.
J Med Case Rep, 17(1):13, 14 Jan 2023
Cited by: 0 articles | PMID: 36639685 | PMCID: PMC9840259
A complex pheotype in a girl with a novel heterozygous missense variant (p.Ile56Phe) of the GNAS gene.
Orphanet J Rare Dis, 17(1):83, 23 Feb 2022
Cited by: 1 article | PMID: 35197096 | PMCID: PMC8867619
Differential diagnosis between syndrome of inappropriate antidiuretic hormone secretion and cerebral/renal salt wasting syndrome in children over 1 year: proposal for a simple algorithm.
Pediatr Nephrol, 37(7):1469-1478, 01 Sep 2021
Cited by: 2 articles | PMID: 34468821 | PMCID: PMC9192468
Review
Free full text in Europe PMC
Go to all (11) article citations
Data
Data behind the article
This data has been text mined from the article, or deposited into data resources.
BioStudies: supplemental material and supporting data
Protein structures in PDBe (4)
-
(3 citations)
PDBe - 3sn6View structure
-
(3 citations)
PDBe - 1azsView structure
-
(2 citations)
PDBe - 1gg2View structure
-
(2 citations)
PDBe - 1tl7View structure
Similar Articles
To arrive at the top five similar articles we use a word-weighted algorithm to compare words from the Title and Abstract of each citation.
GNAS: A New Nephrogenic Cause of Inappropriate Antidiuresis.
J Am Soc Nephrol, 30(5):722-725, 08 Apr 2019
Cited by: 4 articles | PMID: 30962326 | PMCID: PMC6493987
A complex pheotype in a girl with a novel heterozygous missense variant (p.Ile56Phe) of the GNAS gene.
Orphanet J Rare Dis, 17(1):83, 23 Feb 2022
Cited by: 1 article | PMID: 35197096 | PMCID: PMC8867619
Nephrogenic syndrome of inappropriate antidiuresis - An ethnically, genetically and phenotypically diverse disorder: First report in a Chinese adult and review of published cases.
Clin Chim Acta, 519:214-219, 11 May 2021
Cited by: 0 articles | PMID: 33984325
Review
A disruptive mutation in exon 3 of the GNAS gene with albright hereditary osteodystrophy, normocalcemic pseudohypoparathyroidism, and selective long transcript variant Gsalpha-L deficiency.
J Clin Endocrinol Metab, 92(5):1764-1768, 13 Feb 2007
Cited by: 14 articles | PMID: 17299070
GNAS Spectrum of Disorders.
Curr Osteoporos Rep, 13(3):146-158, 01 Jun 2015
Cited by: 76 articles | PMID: 25851935 | PMCID: PMC4417430
Review Free full text in Europe PMC