Significance of GNAS mutations for morbid obesity in children

Abstract

Background

Hereditary forms of obesity are characterized by early severe heterogeneous manifestations of the phenotype along with a rapid progression to morbid obesity, primary due to pathogenic variants of certain genes. Most forms are characterized by moderate to severe neuropsychic developmental delays, dysmorphic features and organ-specific developmental anomalies.

Result

We searched for hereditary causes of morbid obesity in children by exome sequencing. As a result, we have identified 5 variants in the GNAS locus, two of which were identified for the first time: NM_000516.7(GNAS):c.201del, (p.Phe68Leufs*32) and NM_000516.7(GNAS):c.586 − 18_591del. Children showed tolerance to parathyroid hormone and thyroid-stimulating hormone. It has been observed that almost all the children with frameshift variants or nonsense mutations presented with subcutaneous ossifications.

Conclusions

The search for variants in a group of patients with morbid obesity, as conducted in our research, reaffirms the need for use molecular genetic testing to determine the main diagnosis and facilitate early detection of the disease. This is particularly relevant given the wide clinical variability of monogenic forms of obesity.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40246-025-00781-2.

Introduction

Hereditary forms of obesity are characterized by early severe heterogeneous manifestations of the phenotype along with a rapid progression to morbid obesity, primary due to pathogenic variants of certain genes. Most forms are characterized by moderate to severe neuropsychic developmental delay, dysmorphic features and anomalies in organ development. The clinical manifestations of obesity syndromes are often exhibit similarities, which complicates their diagnosis; as a rule, they are difficult to distinguish from convertional obesity, and the determination of genetic causes is important for genetic counselling and the selection of appropriate treatment [1].

The active study of the molecular genetic aspects of childhood obesity via modern technologies and exome/genome sequencing allows the identify the underlying causal variants previously undescribed in the literature that are involved in the pathogenesis of this the disease.

Recently, there has been an increasing number of publications on the identification of mutations in the GNAS in children with obesity [2, 3].

The GNAS gene encodes the Gα_s protein (stimulatory G protein alpha subunit), which mediates signaling by hormones and ligands that bind to G protein-coupled receptors (GPCRs) to generate cyclic AMP [4]. Mutations in GNAS cause developmental delay, short stature and skeletal abnormalities in a syndrome called Albright hereditary osteodystrophy. Imprinting at the GNAS locus results in tissue-specific suppression of the paternally inherited gene allele [5]. Pseudohypoparathyroidism (PHP) type Ia is caused by heterozygous, maternally inherited inactivating mutations. These mutations lead to Gα_s deficiency. There are some imprinted tissues where this signalling protein is derived only or predominantly from the maternal allele, including pituitary, ovaries, thyroid and proximal renal tubules, affected individuals develop PTH-resistance leading to hypocalcemia and hyperphosphatemia, as well as resistance towards TSH and, sometimes, other peptide hormones and ligands. In addition, patients develop features in non-imprinted tissues such as round face, short stature, brachymetacarpia, ectopic ossification, and mental retardation. Paternally inherited GNAS mutations lead to pseudo-PHP characterized by some features of Albright hereditary osteodystrophy (AHO) in the absence of hormone resistance [6]. GNAS mutations disrupt signaling through the melanocortin receptor type 4 (MC4R) which contributes to hyperphagia, impaired sympathetic nervous system activation, as well as accelerated growth and weight gain [7]. Even heterozygous partial loss-of-function MC4R mutations cause severe obesity. These results are explain why severe obesity develops in patients with missense mutations in GNAS before other features of classical PHP appear [3]. The development of obesity occurs only with maternal, but not paternal, Ga_s mutations. Obesity is often diagnosedd in the first years of life, leading to significant health complications in early childhood similar to other monogenic forms of obesity [8].

Despite the progress made in examining the molecular landscape of obesity, there is a lack of comprehensive data regarding the spectrum and frequency of pathogenic changes in various genes, as well as on genotypic-phenotypic correlations of identified variants with disorders affecting various body systems in relation to obesity in children. We evaluated the significance of pathogenic variants of the GNAS gene in children with severe obesity from Russia, and also performed a comparative analysis of the clinical and genetic characteristics of mutation variants.

Materials and methods

A total of 137 unrelated patients (58 girls and 79 boys) with morbid obesity with early onset of the disease (from 1 to 17 years) were included. They were examined at Endocrinology Research Centre, Moscow, Russia. The study was approved by the local ethics committee of the Endocrinology Research Centre, protocol No. 18 dated 10.10.2022. Written informed consent was obtained from a parent and/or legal guardian signing for children.

Inclusion criteria: children with morbid obesity and with onset before 7 years, and standart deviation score (SDS) for body mass index (BMI) index of more than 3.5. Exclusion criteria: the presence of organic pathology affecting the central nervous system.

Clinical data were obtained from medical records corresponding to the date of the initial hospitalization if there was more than one. The study protocol included a clinical examination of patients, a detailed collection of family history, physical examination, assessment of phenotypic features and anthropometric parameters. Weight SDS, Height SDS and BMI SDS were calculated using Auxology 1.0 software (Pfizer, USA). Laboratory methods included analysis of glycated haemoglobin (HbA1c), haemoglobin, total and ionized calcium, phosphorus, fasting blood glucose, total cholesterol, high-density lipoprotein cholesterol (HDL), low-density lipoprotein cholesterol (LDL), triglycerides (TG), alanine aminotransferase (ALT), and aspartate aminotransferase (AST). Hormonal profiling included measurement of insulin, parathyroid hormone (PTH), free thyroxine (FT4), thyroid-stimulating hormone (TSH), insulin-like growth factor 1 (IGF-1), calcitonin, and 25-hydroxyvitamin D. Instrumental diagnostic methods included an ultrasound examination of the abdominal organs, kidneys, and thyroid gland, as well as radiography of the hands and wrists. Genomic DNA was extracted from peripheral blood lymphocytes using the MagPure Blood DNA kit (Magen, China). Quantitative and qualitative analysis of the isolated DNA was assessed using a Nanodrop 2000 spectrophotometer (Thermo Fisher Scientific, Waltham, MA) and a Qubit 2.0 fluorometer (Invitrogen, Carlsbad, CA, USA) with the Qubit dsDNA HS Assay Kit. Whole exome libraries were prepared using the KAPA HyperPlus Kit (Roche, Basel, Switzerland) according to the manufacturer’s protocol. Library enrichment was performed using the KAPA HyperExome Kit (Roche, Basel, Switzerland).

Sequencing was conducted on an Illumina NovaSeq 6000 with the NovaSeq 6000 S4 Reagent Kit v1.5 (200 cycles) using a paired-end2 × 100 bp reading strategy. Variant validation was performed by Sanger sequencing on an ABI 3500 (Applied Biosystems) with specific primers designed to target fragments of the GNAS exons. To verify of the identified variants, a search was performed on the parents to determine whether the variants was inherited or occurred de novo. NGS data were processed via an automated algorithm that included read alignment to the human genome reference sequence (GRCh38), alignment postprocessing, variant identification and variant filtering for quality, and variant annotation on all known transcripts of each gene from the RefSeq database. Computer algorithms were used to predict the pathogenicity of variants based following the guidelines of the American College of Medical Genetics and Genomics (ACMG) [9]. The SpliceAI and AdaBoost programs were used to predict the impact of changes in splice sites and intron regions adjacent to the splice site. The clinical significance of the identified variants was studdied with OMIM and HGMD. The effects of the identified variants on the structure and function of the protein were assessed with the ANNOVAR, SIFT, and Mutation Tester, MutPred, 1000 Genomes, Exome Aggregation Consortium, dbSNP, HGMDB, etc.

Results

Whole-exome sequencing results revealed that 56 children (40.8%) harbored pathogenic or likely pathogenic variants, as well as variants of uncertain clinical significance, explaining the molecular genetic causes of obesity. Among the identified obesity variants, the most common variants were in the genes responsible for monogenic forms of obesity (POMC, PCSK1, MC4R, NTRK2, SIM1, SH2B1, LEP, LEPR, SEMA3A, NRP2, MC3R, ADCY3, KSR2, DYRK1B), PHP (GNAS), ciliopathies (Bardet-Biedl syndrome, Alstrom syndrome), and also single variants in genes for a wide range neurological diseases characterized by the presence of the symptom “obesity” (ADNP, ADRB2, BAP1, ENPP1, FAT1, FFAR4, KCNJ11, KCNJ15, METTL5, MLC1, NAA10, PACS1, SPEN, TNPO2, TRIP12). The age of onset of obesity in children with the identified variants ranged from 0 to 5 years old. In children with no causal variants, the age, at which obesity was noted, ranged from 0 to 17 years old [10]. The presence of polyphagia is noted in most patients who have been identified variants in the genes of the leptin-melanocortin pathway and their cofactors, which is the most common symptom in hereditary forms of obesity. Polyphagia is also noted in neurological diseases and requires further study [10].

Among the children with a confirmed genetic diagnosis, the highest number of variants was found in the GNAS gene (n = 5, 8.9%): three deletions and two single nucleotide substitutions in the heterozygous. Three variants were previously described as pathogenic: GNAS: c.565_568del (p.Asp189MetfsTer14), GNAS: c.85 C > T (p.Gln29Ter) (Table S1) and GNAS: c.493 C > T (p.Arg165Cys), and two variants: GNAS: c.201del, (p.Phe68LeufsTer32) and GNAS: c.586 − 18_591del were identified for the first time and classified as likely pathogenic variants according to ACMG criteria (Fig. 1). In all cases, maternal inheritance is assumed, despite the unavailability of family analysis in two cases (P1, P4) (Table 1).

Fig. 1.

Illustration of exonic and domain localization of the identified changes in the structure of GNAS and Ga_s. The arrows indicate the direction and parental origin of the transcripts. Coding exons are shown in black, noncoding exons in grey; cen, centromeric; Mat, maternal; Pat, paternal; tel, telomeric

Table 1.

Characteristics of identified variants in the GNAS gene

	Patient
	P1	P2	P3	P4	P5
Genetic Change	NM_000516.7(GNAS):c.85C>T, (p.Gln29Ter)	NM_000516.7(GNAS):c.201del, (p.Phe68LeufsTer32)	NM_000516.7(GNAS):c.493C>T, (p.Arg165Cys)	NM_000516.7(GNAS):c.565-568del, (p.Asp189MetfsTer14)	NM_000516.7(GNAS):c.586-18_591del
Clinical significance (ACMG)	Pathogenic	Likely pathogenic	Pathogenic	Pathogenic	Likely pathogenic
Sex, current age	M, 15y	W, 9y	W, 4y	W, 11y	M, 2y9m
Weight (kg)	52.0, SDS: -0.059	64.0, SDS: 3.089	25.6,  SDS: 4.331	36.0,  SDS: 0.675	30.0,  SDS: 8.025
Height (cm)	146.0, SDS: -2.354	133.5, SDS: 0.336	95.0, SDS: 1.032	127.5, SDS: 1.350	85.0, SDS: 0.787
BMI (kg/m2)	24.4, SDS: 1.346	35.9, SDS: 2.823	28.4, SDS: 4.653	22.1, SDS: 1.576	41.5, SDS: 7,11
Age at onset of disease/ first visit/ genetic diagnosis	1m/10y/14y	1m/1y2m/8y	1m/1y11m/2y7m	1m/1y6m/9y	1m/3m/2y5m
TSH, NR=0.64-5.76 mIU/L (Age)	5.65 (14y4m),8.87 (14y9m)	13.60 (1y2m),8.11 (6y),6.09 (6.5y),2.59 (7y),2.38 (8y)	12.45 (1y11m),14.18 (2y6m),7.014 (2y7m),2.75 (4y8m)	2.30 (?),1.03 (9y)	6.69 (3m),9.06 (1y1m),4.33 (1y7m)
PTH, NR=15-65 pg/ml (Age)	536.8 (14y4m),306.7 (14y9m)	498.8 (6y),117.5 (6.5y),259.9 (7y),277.7 (8y)	50.9 (2y7m),86.5 (4y8m)	166.0 (?),192.8 (9y)	20.9 (3m),75.7 (1y1m),68.3 (1y7m)
T4 free, NR=11.5-20.4 pmol/l (Age)	11.87 (14y4m),9.0 (14y9m)	11.41 (6y),13.14 (7y),14.89 (8y)	10.98 (2y7m),15.0 (4y8m)	15.20 (?)	10.53 (1y1m),14.84 (1y7m)
Phosphorus, NR=1.45-1.78 mmol/l (Age)	3.08 (14y4m),1.77 (14y9m)	2.05 (6y),2.04 9 (7y),1.98 (8y)	1.90 (2y6m),1.90 (2y7m),1.54 (4y8m)	1.99 (9y)	1.84(1y1m),1.6 (1y7m)
Calcium total, NR=2.25-2.75 mmol/l/Ionized calcium, NR=1.03-1.29 mmol/l (Age)	1.44/0.70 (14y4m),2.36/1.08 (14y9m)	2.41/1.10 (6y),2.44/1.09 (7y),2.39/1.08 (8y)	2.49/- (1y11m),2.5/1.16 (2y7m),2.49/1.17 (4y8m)	2.21/1.08 (9y)	2.49/1.18 (1y1m),2.29/1.12 (1y7m)
Speech Delay or Learning Difficulties	Yes	Yes	Yes	Yes	Yes
Brachymetacarpia/Ectopic Ossifications	Yes/Yes	Yes/Yes	None/None	None/Yes	None/Yes

Notes: M - man, W- woman, SDS – Standard Deviation Score, BMI – Body mass index, TSH – thyroid stimulating hormone, PTH – parathyroid hormone, NR – normal reference, (?) – unknown

Disruption of this gene leads to the development of PHP, characterized by a wide range of clinical manifestations. Based on the medical history, all children had a high rate of weight gain in the first months of life and, as a result, severe obesity in subsequent childhood. The majority of patients are characterized by short stature - the discrepancy between the observed growth and the target height.

Thus, at the time of the first visit, almost all patients had high TSH levels due to resistance (Table 1). After the taking of levothyroxine, a decrease in the TSH level in the patient’s blood serum was noted. In three patients, PTH levels significantly exceeded the reference values, while in two children the levels PTH were within the normal range with a tendency to increase hormone levels. This probably stems from a lack resistance to PTH at an earlier age. At the same time, the levels of free T4 were either within the normal range or were slightly below the norm. Phosphorus levels in all five patients were above normal, while the levels of total and ionized calcium were within normal limits (Table 1).

Four of the five patients had subcutaneous calcifications (Table 1). Brachymetacarpia was present only in two children. All the children experienced mental retardation. The biochemical parameters were characterized by high PTH. All the children had either delayed speech or learning skills. More detailed clinical information is provided in supplementary materials (Table S2).

The NM_000516.7(GNAS) variant: c.85 C > T, (p.Gln29Ter) results in the appearance of a premature stop codon and, as a consequence, translational arrest (Fig. 2, P1). The variant is not available in the gnomAD database. It is located in the region of the protein responsible for interaction with the receptor (GPCR) and can probably lead to disruption of hormonal signal transmission into the cell. It has been described many times in the literature (Table S1).

Fig. 2.

Pedigrees of the probands and Sanger sequencing variants in GNAS. Image of the deletion in Patient P5 in the IGV graphical browser [11]

The following variant is a single nucleotide deletion NM_000516.7(GNAS):c.201del p.Phe68LeufsTer32, which leads to a shift in the reading frame and the appearance of a premature terminating codon (Fig. 2, P2). It is not found in the gnomAD database and is not described in the literature. This variant was found in both the child and the mother, who had low growth, obesity, subcutaneous calcifications, and decreased intelligence.

Another variant is NM_000516.7(GNAS):c.493 C > T, (p.Arg165Cys), which is also very common among patients with PHP type 1a, has the impact of arginine-to-cysteine substitution. This variant has been repeatedly described by various researchers (Tabl. S1). The variant was found in both the child and the mother. At the same time, there are no clinical manifestations in the mother (Fig. 2, P3).

Variant NM_000516.7(GNAS):c.565_568del, (p.Asp189MetfsTer14) (Fig. 2, P4), leads to deletion of 4 nucleotides in exon 7 of the gene and a reading frame shift and is the most common variant found in the literature (Tabl. S1) and is probably a hotspot in the GNAS gene.

24-base pair deletion NM_000516.7(GNAS):c.586 − 18_591del covering the intron 7 region and a part of the exon 8. This alteration results in the loss of a fragment that includes the region of the splice acceptor site and exon 8 of the gene. Consequently, this disruption triggers nonsense-mediated transcript decay and ultimately leads to protein loss (Fig. 2, P5). Not found in the gnomAD database, not described in the literature. The variant is also described in a mother who has low height, body weight deficiency, subcutaneous calcifications, and decreased intelligence (Fig. 2, P5).

Discussion

Early manifestation and rapid progression of the disease is one of the features of morbid obesity. Morbid obesity is based on a genetic component, in contrast to alimentary obesity, where the leading role is played by lifestyle. The variability of clinical manifestations and severity of various hereditary diseases create challenges in making accurate diagnose, with the exception of a number of syndromes. Variants in the GNAS gene characterized by the heterogeneous of clinical symptoms and complications. In our work, five variants in the GNAS were identified in five unrelated patients, who accounted for 3.65% of the total number of examined in children with morbid obesity. Additionally, PHP type Ia was detected in 8.9% of children with an established genetic diagnosis. Two novel mutations are described for the first time. One variant is located in the region responsible for interaction with the receptor, while the remaining variants are in the hexahelical domain of the alpha subunit of the G protein. Clinically, patients have severe obesity, short stature, ossification, brachymetacarpia, impaired phosphorus-calcium metabolism and resistance to PTH. Phenotypic manifestations (short stature, subcutaneous calcifications, and mental retardation) were also observed in two of three mothers. Molecular genetic testing allowed all five patients to be diagnosed, especially in patients in infancy, when hormone values still do not exceed the reference ranges. Genetic testing is important for classifying the identified variants, although causal variants may not always be identified due to diagnostic approaches using targeted or full-exome sequencing, losing sight of microdeletions or methylation defects. Our data are comparable with the results of other researchers. In the study by Mendes de Oliveira et al. [3], exome sequencing was performed in 2548 children with severe obesity. The work described 22 children with 19 genetic variants in the GNAS: 16 missense mutations, 2 nonsense mutations, and 1 mutation leading to a shift in the reading frame. The effects of GNAS variants on signalling from melanocortin receptor type 4, which is associated with Gαs and critical for appetite and weight regulation, and, as a consequence, the regulation of energy homeostasis were studied in a cohort of 16 children. It is noteworthy that variants in MC4R result in a loss of protein function are the predominant cause of monogenic obesity. Thus, in 14 of 16 cases, impaired the interaction between G_αs and MC4R, MC4R-mediated cAMP accumulation, MC4R-independent cAMP accumulation or all of these factors combined were detected. Gα_s trigger one of two signalling pathways from the MC2R and MC4R, providing the effects of ACTH in the first case and facilitating the regulation of basal metabolism in the second case. These findings explain the mechanism of obesity development in PHP 1a, which is mediated by the resistance of these receptors [3].

Thiele S. et al. described a number of missense substitutions and their impact on protein function. All identified missense substitutions were divided into three classes. The first class included variants involved in disruption of the structure of the hydrophobic core of Gas, which can lead to instability or even lack of ordered three-dimentional structure of the protein. The second class included variants that amend cofactor binding in the region. The third class was represent by variants disrupting the forming salt bridges, thereby introducing instability of the protein structure [12]. In support of this, Hinrichs et al. reported that substitution of the highly conserved amino acid at position p.Arg165Cis with cysteine - also identified in one of our patients, resulted in a decreased GTP binding rate and lower GDP dissociation, which ultimately led to a decreased adenylate cyclase-activating capacity due to stabilization of the αD/αE loop in the hexahelical domain [13].

The GNAS-Ga subunit plays a critical role in transmitting signals from various ligands through the receptor into the cell, thereby influencing numerous processes, including cell differentiation, growth and metabolism. Failure of this mechanism leads to severe disturbances in the vital activities of cells and the whole organism. First, pathogenic variants in GNAS result in disruption of hormonal signalling through parathyroid receptors associated with the G-protein complex. This condition is expressed in PTH resistance [14]. One of the main features of PHP 1a is the increased probability of developing multihormonal resistance in tissues where biallelic expression of Gα_s is absent or weakened. This effect arises because Gα_s plays a crucial role in transmitting signals from the receptor to the cell, impacting not only PTH but also other hormones [15, 16]. Our clinical findings are consistent with the fact that patients with PHP 1a with frameshift mutations or nonsense mutations leading to truncated forms have a significantly higher incidence of subcutaneous calcifications than patients with missense mutations. In contrast, shortened metacarpals/tarsals are more common in patients with missense mutations, whereas obesity, short stature and neurocognitive defects are not correlated with different types of GNAS mutations [12, 14, 17]. However, not all the patients have subcutaneous calcifications and brachymetacarpia, which probably indicates that there is no direct correlation between PTH resistance, impaired phosphorus-calcium metabolism and skeletal structure features [14]. Heterozygous inactivation of Gnas in mice (by disruption of exon 1 in the Ga_s transcript) induced to dysregulation of the expression of the remaining transcripts (XLas and NESP55) in GNAS during osteoblast differentiation in wild-type adipose stromal cells. These transcriptional changes in Gnas+/− mice are accompanied by accelerated osteoblastic differentiation of adipose stromal cells in vitro. In vivo, the altered osteoblast differentiation in Gnas+/− mice manifests as subcutaneous heterotopic ossification. Collectively these data suggest that Gnas is a key regulator in adipose-derived mesenchymal progenitor cells, especially in relation to bone formation [18].

The search for variants in the group of patients with morbid obesity, as conducted in our work once again highlights the need to use molecular genetic testing to determine the main diagnosis. Even with a well-defined genetic diagnosis and an enhanced understanding of the disease, we continue to witness significant clinical variability in phenotypic manifestations among patients primarily arising from the intricate disruption of the complex imprinted GNAS locus. This wide phenotypic variability at initial follow-up and during follow-up allows us to recommend family screening for early detection of the disease. An early and interdisciplinary approach should be crucial in the follow-up of patients. We acknowledge that the results of our work are constrained by the number of patients in whom mutations in GNAS were detected, thereby preventing us from conducting a detailed analysis of genotypic correlations in patients with PHP. Another limiting factor was the unavailability of families for collecting family history and conducting genetic screening of the probands’ family members.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author contributions

This study was designed by R.S. and R.Kh.; writing—original draft preparation by R.S., R.Kh.; Data curation by R.S., R.Kh., O.V., D.K.; formal analysis by V. P., I.M., N.M., R.Kh.; Investigation by R.S., O.V., D.K., Y.K., B.Y. P.O.; Supervision by N.M.; writing—review and editing by R.S., R.Kh., O.V., I.M., N.M. All authors have read and agreed to the published version of the manuscript.

Funding

The research was carried out within the state assignment of the Ministry of Health of the Russian Federation (theme No. 124020700098-5).

Data availability

The data that support the findings of this study are not openly available due to reasons of sensitivity and are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

The study was approved by the local ethics committee of the Endocrinology Research Centre, protocol No. 18 dated 10.10.2022.

Consent for publication

Written informed consent to participate in the study was obtained from the parents and/or legal guardians of each patient.

Competing interests

The authors declare no competing interests.