Background

Galloway-Mowat syndrome (GAMOS) (OMIM #251300) is a rare autosomal recessive syndrome first described in 1968. It is a clinically heterogeneous condition characterized by early-onset nephrotic syndrome associated with microcephaly, gyral abnormalities of the brain, and delayed psychomotor development. Most patients also present dysmorphic facial features, including hypertelorism, ear abnormalities, and micrognathia. Most affected individuals die in early childhood [1, 2]. The estimated prevalence is < 1/1,000,000 but it is likely that many cases remain misdiagnosed or undiagnosed.

Truncating variants in WDR73 gene were described as the first monogenic cause of GAMOS in two families [3]. The protein encoded by this gene is implicated in the regulation of the microtubule network during cell cycle progression, proliferation, and survival [3,4,5]. Homozygous missense variants in the WDR73 gene were later reported [6, 7]. Recently, pathogenic variants in the OSGEP, LAGE3, TP53RK, and TPRKB genes have been identified as novel monogenic causes of GAMOS. These genes encode four subunits of the evolutionary highly conserved KEOPS (kinase, endopeptidase, and other proteins of small size) complex. This complex plays an important role in brain and renal development. To date, the genetic cause of GAMOS has been reported in 54 families: OSGEP in 26 (48.15%), WDR73 in 19 (35.18%), TP53RK in 4 (7.40%), LAGE3 in 3 (5.55%), and TPRKB in 2 (3.70%) [2,3,4,5,6,7,8,9,10,11].

Here, we report two unrelated patients with GAMOS carrying homozygous pathogenic variants in the newly identified OSGEP gene.

Case presentation

Patient 1 was a first-child male born to healthy consanguineous parents from Spain with no previous family history of kidney disease (Table 1, Fig. 1A). He was born at 39.4 weeks of gestation. Birth weight was 3400 g, height was 51 cm, and head circumference was 34 cm, with no dysmorphic features. Forty-five days after birth, the patient was diagnosed as having congenital nephrotic syndrome with severe proteinuria, hypertension, and hypothyroidism. He also presented edema, hyperkalemia, hyponatremia, and hypomagnesemia. Renal ultrasound showed poor corticomedullary differentiation in the right kidney. Renal biopsy showed diffuse mesangial sclerosis, tubular atrophy, and primitive glomeruli (Fig. 1C). The patient progressed to end-stage renal disease and required peritoneal dialysis. He also presented left eye evisceration, dry right eye, and gastroesophageal reflux. At 5 months of age, the neurological examination revealed microcephaly with a head circumference of 38.5 cm (percentile < 1, − 4.72 SD), severe psychomotor delay for his age, and axial hypotonia. Cranial magnetic resonance imaging (CMRI) revealed brain atrophy and absence of normal myelination of the brainstem, cerebellar white matter, bilateral hemispheric white matter, internal capsules, and corpus callosum as well as abnormal intensity signal in the dentate nucleus and thalamus (Fig. 1D). In view of the congenital nephrotic syndrome and microcephaly with brain anomalies, a clinical diagnosis of GAMOS was suspected. The patient presented progressive neurological deterioration and died at 8 months of age.

Table 1 Clinical data of patients 1 and 2
Full size table
Fig. 1
figure 1

a Pedigree of patient 1 with a likely pathogenic OSGEP variant, c.81C > G p.(Asn27Lys), in homozygosity while his consanguineous parents are healthy heterozygous carriers. b The identified missense variant c.81C > G p.(Asn27Lys) affects a totally conserved amino acid N27 in OSGEP orthologs. c Silver-stained renal biopsy from patient 1 showed glomerular collapse with mesangial matrix increase, atrophic tubules, and interstitial fibrosis on light microscopy. d CMRI performed in patient 1 at 8 months of age: sagittal 3 Dimensional Imaging T1 sequence (a), axial reconstructions (b–e), and axial Turbo Spin Echo (TSE) T2 (f–j). k–o: Sequence TSE T2 of normal control individual. MRI revealed craniofacial disproportion in relation to microcephaly (a); supratentorial cortico-subcortical atrophy with increased extra-axial space, prominence of the frontal horns, and thinning of the corpus callosum (red arrow in a); bilateral subdural frontoparietal hygromas (red asterisks in h–j); and atrophy of the basal ganglia (h). A decrease in the number and depth of the grooves was observed (h–j) and there was an absence of normal myelination of the brainstem (red arrow in f), cerebellar peduncles (blue arrow in f), internal capsules (red arrow in h), and white bihemispheric substance (arrows in i and j). Hypointense T2 signal of the thalamus was evident (blue arrow in h). Enucleation of the left eye is denoted by the yellow arrow in f. Finally, there was an increase in the thickness of the cranial and facial subcutaneous cellular tissue (green arrows in a and b)

Full size image

Patient 2 was a female infant with normal karyotype (46, XX) born to healthy consanguineous parents from Pakistan. The patient had two healthy sisters and there was no family history of kidney disease (Table 1, Fig. 2a). She was born at 40.3 weeks of gestation. Birth weight was 2940 g, height was 49 cm, and head circumference was 32 cm with signs of microcephaly. Screening for metabolic disorders and cerebral ultrasound were normal. The patient presented dysmorphic features (wide nasal bridge, aquiline nose and retrognathia, low set ears, and arachnodactyly) with axial hypotonia and poor eye contact. Seventy-five days after birth, she presented with nephrotic range proteinuria, hypoproteinemia with severe hypoalbuminemia, hypertriglyceridemia and hypercholesterolemia. Serum creatinine and urea were normal. Abdominal ultrasound showed normal-sized kidneys and correct corticomedullary differentiation with cortical hyperechogenicity, bilateral pleural effusion, and a discrete amount of fluid in the abdominal cavity. Renal biopsy showed one glomerulus with increased mesangial matrix and two normal glomeruli; fibrosis and tubular atrophy were absent. CMRI revealed severe brain atrophy with normal cerebellum and brainstem. Electroencephalogram showed normal brain activity with low-amplitude brain waves and occasional frontal left epileptiform activity. The patient evolved with failure to thrive, anemia, and electrolyte disorders and finally died from cardiorespiratory arrest in a sepsis context at 7 months of age.

Fig. 2
figure 2

a Pedigree of the family of patient 2 with a likely pathogenic OSGEP variant c.157A > T p.(Ile53Phe) in homozygosity in the proband (arrow) and heterozygosity in her healthy parents. b Conservation of I53 in OSGEP orthologs to C. elegans

Full size image

Genetic study

Variant analyses of patients 1 and 2 were performed by targeted massive parallel sequencing using an updated version of our kidney disease gene panel that includes more than 200 genes causative of or associated with inherited kidney diseases (including WDR73, TPRKB, TP53RK, LARGE3, and OSGEP genes) [12].

Briefly, libraries were prepared according to the manufacturer’s standard protocol, NimbleGen SeqCap EZ Library SR version 4.3. Patients’ DNAs were fragmented and hybridized to the custom NimbleGen SeqCap EZ Choice gene panel and sequenced on a NextSeq 500 instrument (Illumina). Sequence data analysis was performed using an open-source in-house bioinformatic pipeline, as previously reported [12,13,14]. The mean depth of coverage per exon of OSGEP ranged from 153 to 433, with 100% of the bases covered at least 100X. Prediction of pathogenicity was evaluated using different bioinformatic algorithms (DANN, GERP, dbNSFP.FATHMM, LRT, MetaLR, MetaSVM, MutationAssessor, PROVEAN, SIFT, and MutationTaster). Clinical interpretation of variants was based on American College of Medical Genetics (ACMG) recommendations [15]. All candidate pathogenic variants were validated by conventional polymerase chain reaction amplification and Sanger sequencing. Familial segregation analysis was assessed. Analysis of copy number variations (CNVs) was performed using CoNVaDING (copy number variation detection in next-generation sequencing gene panels) software [16].

Patient 1 carried a homozygous missense variant c.81C > G p.(Asn27Lys) in exon 1 of the OSGEP gene (NM_017807), not previously described in the literature. This variant was predicted to be pathogenic by seven prediction tools (DANN, GERP, LRT, MutationAssessor, MutationTaster, SIFT, and PROVEAN) and benign by three (dbNSFP.FATHMM, MetaLR, and MetaSVM). This variant altered an evolutionarily highly conserved residue and was absent from the population databases Genome Aggregation Database (gnomAD) and 1000 Genomes (Fig. 1B). Segregation analysis showed that both parents were heterozygous carriers of this OSGEP variant (Fig. 1A). We concluded that this variant was likely pathogenic (Table 2).

Table 2 Information of the two identified variants in the OSGEP gene
Full size table

Patient 2 carried a homozygous missense variant c.157A > T p.(Ile53Phe) localized in exon 2 of the OSGEP gene. This variant has not been previously reported in literature. The variant c.157A > T p.(Ile53Phe) was predicted to be pathogenic by five prediction tools (DANN, LRT, MutationAssessor, MutationTaster, and PROVEAN) and benign by five (GERP, dbNSFP.FATHMM, MetaLR, SIFT, and MetaSVM). This variant was conserved in OSGEP orthologs to C. elegans and is extremely rare in the general population (Fig. 2b), with a minor allele frequency in South Asians is 0.00009799 (3 of 30,782 sequenced alleles, no homozygous individuals) in the gnomAD database. The global allele frequency was lower than the 0.0001 threshold for recessive gene OSGEP. The parents were confirmed to be heterozygous carriers (Fig. 2a). We classified this variant as likely pathogenic (Table 2).

Discussion and conclusions

We report two patients who presented with nephrotic syndrome with onset at < 3 months old, primary microcephaly, and developmental delay, which are hallmarks of GAMOS. Both patients carried homozygous likely disease-causing variants in the OSGEP gene. This gene was recently identified as causative of GAMOS in a large cohort of 907 individuals with nephrotic syndrome [2]. Pathogenic variants in one of the four genes TP53RK, TPRKB, LAGE3, and OSGEP, encoding KEOPS complex subunits, were found in 37 out of 91 patients with GAMOS. Independently, a homozygous pathogenic variant in the OSGEP gene was reported in two siblings with a similar renal-neurological phenotype, also by whole exome sequencing [8].

The OSGEP gene encodes the O-sialoglycoprotein endopeptidase enzyme, which regulates the second biosynthetic step in the formation of N-6-threonylcarbamoyladenosine in the cytosol, essential for mRNA translational initiation and efficiency. The highly conserved KEOPS complex is implicated in several cell processes, such as control of telomere length, telomere-associated DNA damage response signaling, and genome maintenance. Zebrafish larvae knockout of the osgep gene resulted in primary microcephaly, with increased apoptosis in the brain compared with controls and early lethality. Knockout mouse embryos also showed microcephaly compared with wild-type embryos. Neither mutant fish nor mice showed any renal phenotype, possibly due to embryonic early lethality [2].

Great strides have been made in the understanding of GAMOS disease over the past 4 years, with the identification of its genetic bases in some patients. However, the genetic etiology of more than three-quarters of patients with a clinical diagnosis of GAMOS remains elusive, suggesting that additional causative genes remain to be identified. Currently, the principal known causative genes of GAMOS are OSGEP and WDR73.

A review of the literature based on 31 patients (26 families) bearing OSGEP pathogenic variants and 23 patients (13 families) with WDR73 pathogenic variants indicates that OSGEP causes earlier onset of nephrotic syndrome than WDR73 [2,3,4, 6,7,8,9,10]. Eighty percent (25/31) of patients with OSGEP pathogenic variants developed nephrotic syndrome with a mean age at onset of 10.36 months (ranging onset from birth to 13 years). In comparison, 35% (8/23) of patients with WDR73 pathogenic variants presented nephrotic syndrome at a mean age of 7.7 years (ranging from 0.5 to 16 years). Our two patients carrying OSGEP pathogenic variants presented with nephrotic syndrome before 3 months of age.

Renal manifestations described in GAMOS patients vary from isolated proteinuria to steroid-resistant nephrotic syndrome, and some patients even have no renal alterations during follow-up period [2,3,4,5,6,7,8]. Intrafamilial clinical variability has also been described in GAMOS. For instance, two siblings carrying a WDR73 pathogenic variant manifested contrasting renal phenotype [3]. One of the affected siblings presented with nephrotic syndrome at the age of 5 years, rapidly developed chronic renal insufficiency, and died after a month, while the other had no renal symptoms at the age of 7 years [3]. A homozygous OSGEP pathogenic variant, c.974A > G p.(Arg325Gln), has also been associated with renal tubular anomalies [10]. It was detected in a girl with magnesium-wasting tubulopathy and partial Fanconi syndrome with a normal glomerular filtration rate who never developed nephrotic syndrome [10]. Interestingly, this variant was previously identified in two siblings with severe hypomagnesemia, hypercalciuria, and proteinuria but normal albumin levels [8]. The authors raised the question of whether these patients should be considered to be affected by a different clinical entity [8, 10].

The review of the literature also indicates that patients with OSGEP pathogenic variants have a shorter life expectancy than those with WDR73 pathogenic variants. Seventy-one percent (22/31) of patients with OSGEP pathogenic variants died at a mean age of 1.5 years (ranging from 6 weeks to 8 years). In line with these reported cases, our patients died at 8 and 7 months of age. However, seven patients with OSGEP pathogenic variants were alive at 13 (2), 10.5 (1), 7 (1), 3.5 (1), and 2 (1) years and at 7 (1) months [2, 10]. It should be noted that four of them carried the above-mentioned OSGEP variant, c.974A > G p.(Arg325Gln), associated with renal tubular anomalies [10]. Twenty-two percent (5/23) of patients carrying pathogenic variants in the WDR73 gene died at a mean age of 8.1 years (ranging from 2.5 to 17 years).

Nearly all OSGEP variants reported as causative of GAMOS are missense, except for two splicing variants [2, 8]. These variants are located throughout the OSGEP gene. By contrast, different types of variant in WDR73 causative of GAMOS have been reported in the literature, including nonsense (3), frameshift (3), and missense (4). No correlation seems to exist between the type or position of the variant and particular clinical features. Identification of the causative pathogenic variant in patients 1 and 2 confirmed the initial clinical suspicion of GAMOS and allowed precise genetic counseling to their parents. In particular, it allowed prenatal diagnosis of a baby girl without GAMOS for the parents of patient 1.

In conclusion, we report two patients with GAMOS caused by OSGEP pathogenic variants. These two cases, in conjunction with the reported cases in the literature, add evidence that OSGEP pathogenic variants are the most prevalent cause of GAMOS and are associated with a more severe phenotype than WDR73 pathogenic variants. For these reasons, OSGEP variant analysis should be considered as the first step in genetic diagnosis of patients with clinical suspicion of GAMOS; this is especially true for those labs that do not perform massive parallel sequencing and for those cases with early and severe onset of the disease. Genetic diagnosis of GAMOS is of paramount importance for genetic counseling and family planning and allows prenatal or preimplantation genetic diagnosis for future pregnancies.