Genetic analysis and outcomes of Omani children with steroid‐resistant nephrotic syndrome

Abstract

Background

Nephrotic syndrome (NS) is one of the most common kidney disorders seen by pediatric nephrologists and is defined by the presence of heavy proteinuria (>3.5 g/24 h), hypoalbuminemia (<3.5 g/dL), edema, and hyperlipidemia. Most children with NS are steroid‐responsive and have a good prognosis following treatment with prednisolone. However, 10%–20% of them have steroid‐resistant nephrotic syndrome (SRNS) and fail to respond to treatment. A significant proportion of these children progress to kidney failure.

Methods

This retrospective study aimed to determine the underlying genetic causes of SRNS among Omani children below 13 years old, over a 15‐year period and included 77 children from 50 different families. We used targeted Sanger sequencing combined with next‐generation sequencing approaches to perform molecular diagnostics.

Results

We found a high rate of underlying genetic causes of SRNS in 61 (79.2%) children with pathogenic variants in the associated genes. Most of these genetically solved SRNS patients were born to consanguineous parents and variants were in the homozygous state. Pathogenic variants in NPHS2 were the most common cause of SRNS in our study seen in 37 (48.05%) cases. Pathogenic variants in NPHS1 were also seen in 16 cases, especially in infants with congenital nephrotic syndrome (CNS). Other genetic causes identified included pathogenic variants in LAMB2, PLCE1, MYO1E, and NUP93.

Conclusion

NPHS2 and NPHS1 genetic variants were the most common inherited causes of SRNS in Omani children. However, patients with variants in several other SRNS causative genes were also identified. We recommend screening for all genes responsible for SRNS in all children who present with this phenotype, which will assist in clinical management decisions and genetic counseling for the affected families.

Nephrotic syndrome is one of the most common kidney disorders seen by pediatric nephrologists. This study in 77 Omani children determined the underlying genetic causes of steroid‐resistant nephrotic syndrome (SRNS). Pathogenic variants in NPHS2 were the most common cause of SRNS in our study seen in almost half of all cases. We recommend screening for all genes responsible for SRNS in all children who present with this phenotype, which will assist in clinical management decisions and genetic counseling for the affected families.

1. INTRODUCTION

Idiopathic nephrotic syndrome (INS) is one of the most common kidney disorders seen in the pediatric clinic with an estimated incidence of approximately 2–6.5 per 100,000 children per year, depending on ethnicities and regions (McKinney et al., 2001). The presence of heavy proteinuria (protein excretion >3.5 g/24 h), hypoalbuminemia (<3.5 g/dL), edema, and hyperlipidemia are the main clinical manifestations (Eddy & Symons, 2003). Most children with NS are steroid‐responsive and have a good prognosis, while 10%–20% of them exhibit steroid resistance and do not achieve complete remission within 4–6 weeks of glucocorticoid treatment (Trautmann et al., 2020), hence are labeled as steroid‐resistant nephrotic syndrome (SRNS). Outcomes for children with non‐genetic SRNS are generally good with around two‐thirds of patients achieving complete remission (Trautmann et al., 2023). For the genetic subgroup of SRNS, complete remission is more uncommon. Around 20% of genetic SRNS patients have chronic kidney disease (CKD), progress to kidney failure (KF) (Trautmann et al., 2023) and no response to immunosuppression in the first year is a predictor of poor kidney function outcomes (Ying et al., 2021). The most common histological diagnoses in children with NS include minimal change disease (MCD), focal segmental glomerulosclerosis (FSGS), membranoproliferative glomerulonephritis and membranous glomerulonephritis (Churg et al., 1970). A histological picture of FSGS is commonly seen in children with SRNS (Churg et al., 1970; McCarthy & Saleem, 2011).

The etiology of idiopathic SRNS in children is either due to a single gene disorder leading to abnormal expression of a podocyte specific protein (Machuca et al., 2009) or due to immune dysregulation which probably leads to production of circulating glomerular permeability factors (Machuca et al., 2009; Shalhoub, 1974). Several genetic causes are known to be involved in the pathogenesis of SRNS (OMIM PS256300) (Ha, 2017; Koziell et al., 2002). Mutations in NPHS1 (OMIM 602716), NPHS2 (OMIM 604766), WT1 (OMIM 256370), LAMB2 (OMIM 614199) and PLCE1 (OMIM 610725) are the most common causes of genetic forms of NS in children (Ha, 2017). These genes encoded different proteins contributing to the glomerular filtration barrier and defects lead to proteinuria and deterioration of kidney function (Wiggins, 2007) and eventually development of KF (Ha, 2017; Koziell et al., 2002). Recent advances in next‐generation sequencing (NGS) have enabled the identification of more than 50 podocyte‐related genes associated with different monogenic forms of NS (Bierzynska et al., 2017; Preston et al., 2019) including MYO1E (OMIM 614131) and NUP93 (OMIM 616892).

For children with SRNS due to a genetic cause, a molecular genetic diagnosis is essential for decision‐making related to treatment and predicting prognosis (Ha, 2017; Shalhoub, 1974). It enables clinicians to start appropriate management and treatment. In addition, patients with disease‐causing variants in some of the genes encoding enzymes of the coenzyme Q10 (CoQ10) pathway (e.g., COQ2 (OMIM 607426) and COQ6 (OMIM 614650)) can be treated using supplementation with CoQ10 (Preston et al., 2019). Furthermore, identification of disease‐causing mutations can facilitate the process of kidney donor selection and the prediction of post‐transplant recurrence of NS (Morello et al., 2023).

In Oman, inherited monogenic kidney diseases are relatively common, leading to a significant healthcare burden (Al Alawi, Al Salmi, Al Mawali, Al Maimani, & Sayer, 2017; Al Alawi, Al Salmi, Al Mawali, & Sayer, 2017). Molecular genetic studies of patients with inherited cystic kidney disease and renal ciliopathies from this population have previously been reported (Al Alawi et al., 2019, 2020, 2021). This present study describes, for the first time, the demographic characteristics, clinical features, and genetic findings of children with SRNS who were seen in the pediatric nephrology department at the Royal Hospital, Oman. The identification of the genetic components of these children with SRNS is essential given the tribal structure of the Omani population and the high rate of consanguineous marriages. This report gives insight into the type and frequency of genetic mutations detected in Omani children with SRNS and is anticipated to support pediatricians and pediatric nephrologists in management decisions and genetic counseling.

2. METHODS

2.1. Ethical compliance

This study was ethically approved by the Royal Hospital Research Ethical Committee, Ministry of Health (MOH/CSR/21/24412). For genetic studies of patients, written informed consent was provided by the patients' families. Genomic DNA was isolated from whole blood of patients and the available family members at the time of presentation using Hamilton's Microlab® STAR™, according to the manufacturer protocol.

2.2. Patient identification and recruitment

This retrospective study involved reviewing 77 pediatric patients (≤13 years old) from 50 Omani families diagnosed with primary SRNS at Royal Hospital, Oman, over a period of 15 years (2005 to 2020). The hospital electronic medical records of all 77 patients were retrospectively reviewed. The following variables were collected: age at onset of NS, gender, parental consanguinity, family history of disease (including previous genetic diagnosis in a sibling), and initial symptoms. Data on kidney function at presentation and last follow‐up, serum albumin, kidney biopsy findings, immunosuppressive treatment, renal replacement therapy including dialysis/transplant, and outcome were also collected.

This study was ethically approved by the Royal Hospital Research Ethical Committee, Ministry of Health (MOH/CSR/21/24412). For genetic studies of patients, written informed consent was provided by the patients' families. Genomic DNA was isolated from whole blood of patients and the available family members including other affected siblings and parents at the time of presentation using Hamilton's Microlab® STAR™, according to the manufacturer protocol. Genetic analysis, described below, was performed as a clinical service on incident cases.

Patients were diagnosed with SRNS based on the following criteria: failure to respond (absence of complete remission, defined as resolution of proteinuria) to prednisolone 60 mg/m²/day for 4–6 weeks with or without three doses of intravenous methylprednisolone. Patients were diagnosed with congenital nephrotic syndrome (CNS) if symptoms presented on the first 3 months of life, whereas they were classified with infantile nephrotic syndrome if presented between 3 and 12 months of age. Patients presenting with NS between 1 and 13 years of age and not responding to steroids were labeled as childhood SRNS. Patient were diagnosed with multidrug resistance based on absence of complete remission after 12 months of treatment with steroid and another two immunosuppressive medications. Prolonged high‐dose steroid was not given to patients presenting with SRNS in the context of a sibling with a known genetic diagnosis of SRNS or if there was established KF.

High blood pressure was defined according to the 4th report on diagnosis, evaluation, and treatment of high blood pressure in children and adolescents (Hogg et al., 2003).

Glomerular filtration rate (GFR) was calculated from the Schwartz equation; normal GFR was defined if GFR ≥90 mL/min/1.73 m². CKD was staged according to published CKD classification (Eckardt et al., 2009; Hogg et al., 2003).

Regarding the treatment of SRNS children in our center, the following were usually practiced:

1. The majority of patients with CNS were managed by daily intravenous albumin infusion in addition to receiving anti‐proteinuric measures using ACE inhibitor and indomethacin, nutrition support and management of complications including infection, anemia, and hypothyroidism.

2. Patients diagnosed with childhood SRNS were started on anti‐proteinuric measures using ACE inhibitors.

3. In cases where there was an absence of genetic testing or failure to detect disease‐causing variants in NPHS1 or NPHS2, children were treated with calcineurin inhibitor cyclosporine or tacrolimus for a minimum of 6 months, and some of them received mycophenolate and or rituximab before being labeled as multidrug resistance NS.

2.3. Molecular testing of NPHS1 and NPHS2

Molecular analysis of NPHS1 and NPHS2 genes was performed using targeted Sanger sequencing of patient DNA at the National Genetic Centre, Royal Hospital, Oman. All coding exons, adjacent exon‐intron boundaries and up to 700 bp of the 5′UTR end from both NPHS1 and NPHS2 genes were amplified using polymerase chain reaction (PCR). Primer3 was utilized to design primer sequences (http://primer3.ut.ee/), which are available upon request. PCR amplification was performed using AmpliTaq Gold PCR master mix (Qiagen) kit, as per the manufacturer instructions. The amplified amplicons were verified on 1% gel and purified using ExoSAP‐IT PCR clean‐up reagent (Applied Biosystems). Bi‐directional fluorescence‐based sequencing was performed on an ABI 3730 XL sequencer using BigDye Terminator V3.1 Cycle Sequencing kit (Applied Biosystems). The obtained sequences were assembled and aligned compared with a reference sequence using the SequencePilot 4.2.2 software (JSI Medical Systems GmbH).

2.4. Genetic variant detection and annotation

Genetic variants were assessed as clinically significant by using public databases including Human Gene mutation database (HGMD), LOVD and variants experimentally evidenced in the literature to be associated with disease phenotype. Variants commonly seen in healthy population with frequency ≥1% were excluded from analysis using databases 1000 Genomes project, Exome Aggregation Consortium (ExAC), Genome Aggregation Database (gnomAD), and NHLBI ESP. In addition, ACMG variant classification was performed using Varsome. GenBank Reference Sequences used were NPHS1 (NM_004646.4); NPHS2 (NM_014625.4); LAMB2 (NM_002292.4); MYO1E (NM_004998.4); PLCE1 (NM_016341.4); and NUP93 (NM_014669.5).

2.5. Whole exome sequencing

Ten patients, in whom no genetic mutations in NPHS1 and NPHS2 were detected, underwent further genetic testing using whole exome sequencing (WES) via CENTOGENE, Germany, or the Translational and Clinical Research Institute, Newcastle University, UK.

2.6. In silico modeling of PLCE1 and NUP93 alleles

Alphafold sequence AF‐Q9P212‐F1 was imported to PyMOL and labeled according to available UniProtKB data. PyMOL Mutation Wizard was used to model the most likely structural impact of c4301G>T, p.(Arg1434Leu) in the 1‐phosphatidylinositol 4,5‐bisphosphate phosphodiesterase epsilon‐1 protein, encoded from the PLCE1 gene on position 23.33 the q arm of chromosome 10 and has 39 exons. Due to the availability of human protein crystal structure informed by nuclear magnetic resonance and X‐ray crystallography, the mutational region of interest is predicted with >90% accuracy and the mutation effects are predicted with 93.6% accuracy.

Alphafold sequence AF‐Q8N1F7‐F1 was imported to PyMOL and labeled according to available UniProtKB data. PyMOL Mutation Wizard was used to model the most likely structural impact of c.1319T>C, p.(Phe440Ser) in the Nucleoporin 93 protein, encoded from the NUP93 gene on chromosome 16q13. Due to the availability of human protein crystal structure informed by electron microscopy, the mutational region of interest is predicted with >90% accuracy. The most likely rotamer has been illustrated above with mutation effects predicted with 41.9% accuracy.

3. RESULTS

3.1. Patients characteristics

In total, 77 children were included in this study from different regions of Oman, who were followed by the pediatric nephrology unit at Royal hospital, Oman. It comprised 41 females and 36 males from 50 different families. The mean age at disease onset was 17.5 months (range 0–120 months). Consanguineous marriages were reported in 58 (75%) of the total cohort and in 48 patients who had a positive family history of SRNS. Of these 77 patients, 24 (31.1%) were diagnosed with CNS, 9 (11.7%) were diagnosed with INS and the remaining 44 (57.1%) were diagnosed with childhood SRNS.

All children had nephrotic range proteinuria. Haematuria at presentation was reported in 60 (77.9%) of our patients. CKD at disease onset was seen in 23 (29.9%) patients. Hypothyroidism and hypertension were reported in 47 (61.0%) and 21 (28.0%), respectively. Mean eGFR at disease onset was 124 mL/min/1.73 m² while mean eGFR at last follow‐up was 53 mL/min/1.73 m². Mean serum albumin at disease onset and last follow‐up were 12.3 g/L and 26 g/L, respectively.

Before 2014, most patients who had SRNS underwent a diagnostic kidney biopsy, but subsequently, biopsy rates declined as genetic testing for SRNS was introduced. After 2014, only patients who were negative for disease‐causing variants in NPHS1 and NPHS2 proceeded to kidney biopsy. In this cohort, a kidney biopsy was performed in 43 (55.8%) children. The most common histopathological diagnosis was FSGS (n = 31; 40.4%), followed by CNS of Finnish type (n = 5; 6.5%), diffuse mesangial sclerosis (3; 3.9%), MCD (2; 2.6%), and severe interstitial fibrosis and tubular atrophy (2; 2.6%) (Figure 1a).

FIGURE 1.

Summary of genetic and clinical studies of our steroid resistance nephrotic syndrome (SRNS) cohort. (a) Distribution of genetic findings among different age groups of patients. Histograms indicate the number of patients with causative gene detected per age group. (b) Gene‐specific distribution of identified disease‐causing variants and without genetic causes “no mutation”. (c) Histopathological analysis of patients with SRNS. FSGS; Focal segmental glomerulosclerosis. FT NS: Finish type nephrotic syndrome. DMS: diffuse mesangial sclerosis. IFTA: interstitial fibrosis and tubular atrophy.

Disease‐causing genetic variants were identified in 60 (77.9%) children using a combination of direct Sanger sequencing of NPHS1 and NPHS2 (52 patients) and WES (8 patients) (Table 1 and Figure 1b). In these 60 patients, 46 patients had homozygous disease‐causing variants, consistent with the high rates of consanguinity.

TABLE 1.

Demographic data, phenotypic presentation and genetic analysis of the Omani SRNS patient cohort.

Family no.	Sex	Consanguinity	Age at diagnosis	Phenotype	Renal biopsy	Gene	Nucleotide change	Zygosity	Amino acid change and ACMG criteria	Patient and kidney outcomes
1a	Male	No	1 month	CNS	Not done	NPHS1	c.514_516delACC; c.2501T>A	Compound heterozygous	p.(Thr172del) (Pathogenic); p.(Val834Asp) (Likely Pathogenic)	Deceased
1b	Male	No	1 month	CNS	Not done	NPHS1	c.514_516delACC; c.2501T>A	Compound heterozygous	p.(Thr172del) (Pathogenic); p.(Val834Asp) (Likely Pathogenic)	Deceased
2	Female	Yes	1 month	CNS	FT	NPHS1	c.515_517 delCCA; c.1379G>A	Compound heterozygous	p.(Thr172del) (Pathogenic); p.(Arg460Gln) (Pathogenic)	KT
3	Female	Yes	1 month	CNS	Not done	NPHS1	c.515_517delCCA	Homozygous	p.(Thr172del) (Pathogenic)	Deceased
4a	Female	Yes	2 month	CNS	Not done	NPHS1	c.515_517delCCA	Homozygous	p.(Thr172del) (Pathogenic)	PD
4b	Female	Yes	1 month	CNS	FT	NPHS1	c.515_517delCCA	Homzygous	p.(Thr172del) (Pathogenic)	Deceased
4c	Male	Yes	1 year	INS	Not done	NPHS1	c.515_517delCCA	Homozygous	p.(Thr172del) (Pathogenic)	PD
5	Female	Yes	1 month	CNS	Not done	NPHS1	c.515_517delCCA	Homozygous	p.(Thr172del) (Pathogenic)	Deceased
6	Female	Yes	1 month	CNS	Not done	NPHS1	c.614_621 delCACCCCGGinsTT	Homozygous	p.(Thr205_Arg207delinsIle (Pathogenic)	Deceased
7	Female	Yes	1 month	CNS	Not done	NPHS1	c.515_517delCCA; c.3478C>T	Compound heterozygous	p.(Thr172del) (Pathogenic); p.(Arg160*) (Pathogenic)	CKD
8	Female	Yes	1 month	CNS	Not done	NPHS1	c.515_517 delCCA	Homozygous	p.(Thr172del) (Pathogenic)	Deceased
9	Male	Yes	1 month	CNS	Not done	NPHS1	c.515_517 delCCA	Homozygous	p.(Thr172del) (Pathogenic)	Normal GFR
10a	Female	Yes	3 month	CNS	Not done	NPHS1	c.2663G>A	Homozygous	p.(Arg88Lys) (Likely Pathogenic)	Deceased
10b	Female	Yes	1 month	CNS	FT	NPHS1	c.2663G>A	Homozygous	p.(Arg88Lys) (Likely Pathogenic)	HD
11	Female	Yes	2.5 month	CNS	FT	NPHS1	c.1134G>A	Homozygous	p.(Trp378*) (Likely Pathogenic)	KT
12	Male	Yes	1 month	CNS	FT	NPHS1	c.515_517delCCA	Homozygous	p.(Thr172del) (Pathogenic)	PD
13	Female	Yes	4 year	SRN	Not done	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	PD
14a	Female	Yes	2 year	SRN	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
14b	Female	Yes	5 year	SRN	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
14c	Female	Yes	3 year	SRN	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	Normal GFR
15a	Male	Yes	1 year	INS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	KT
15b	Male	Yes	1 year	INS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	KT
15c	Female	Yes	1 year	INS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	Deceased
16a	Female	Yes	2 year	SRNS	FSGS	NPHS2	c.467dup/c.709G>C	Compound heterozygous	p.(Leu156Phefs*11) (Pathogenic); p.(Glu237Gln) (Uncertain Significance PM1 PP1 BS2 BP4 BP6)	HD
16b	Male	Yes	2 year	SRNS	FSGS	NPHS2	c.467dup/c.709G>C	Compound heterozygous	p.(Leu156Phefs*11) (Pathogenic); p.(Glu237Gln)) (Uncertain Significance PM1 PP1 BS2 BP4 BP6)	HD
16c	Male	Yes	3 year	SRNS	FSGS	NPHS2	c.467dup/c.709G>C	Compound heterozygous	p.(Leu156Phefs*11) (Pathogenic); p.(Glu237Gln)) (Uncertain Significance PM1 PP1 BS2 BP4 BP6)	HD
17	Male	Yes	3 year	SRNS	Not done	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	PD
18	Male	Yes	3 year	SRNS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	Normal GFR
19a	Male	Yes	18 month	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
19b	Female	Yes	1 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	PD
20a	Male	Yes	1 year	INS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
20b	Female	Yes	5 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	HD
20c	Female	Yes	2 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	Deceased
20d	Male	Yes	3 year	SRNS	MCD	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
20e	Male	Yes	2 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	HD
20f	Male	Yes	5 month	INS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	Normal GFR
21a	Female	Yes	6 year	SRNS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	KT
21b	Female	Yes	2 year	SRNS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	KT
21c	Female	Yes	3 year	SRNS	Not done	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	Normal GFR
22a	Male	Yes	1.5 year	SRNS	FSGS	NPHS2	c.685C>T	Homozygous	p.(Arg229*) (Pathogenic)	PD
22b	Male	Yes	1 year	SRNS	FSGS	NPHS2	c.685C>T	Homozygous	p.(Arg229*) (Pathogenic)	PD
22c	Female	Yes	2 year	SRNS	Not done	NPHS2	c.685C>T	Homozygous	p.(Arg229*) (Pathogenic)	Normal GFR
23	Female	Yes	9 month	INS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	PD
24	Male	No	5 year	SRN	FSGS	NPHS2	c.686G>A; c.935dup	Compound heterozygous	p.(Arg229*) (Pathogenic); p.(Ser313Valfs*33) (Pathogenic)	Normal GFR
25a	Female	Yes	3 year	SRNS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
25b	Male	Yes	18 month	SRNS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
26	Female	Yes	3 year	SRNS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
27a	Female	Yes	3 year	SRNS	FSGS	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	KT
27b	Male	Yes	2 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	Normal GFR
28a	Male	Yes	4 year	SRNS	FSGS	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	Deceased
28b	Male	Yes	2 year	SRNS	Not done	NPHS2	c.467dup	Homozygous	p.(Leu156Phefs*11) (Pathogenic)	Normal GFR
29	Female	no	4 year	SRNS	Not done	NPHS2	c.779T>A	Homozygous	p.(Val1260Glu) (Pathogenic)	Normal GFR
30a	Male	Yes	1 month	CNS	Not done	LAMB2	c.1405+1G>A	Homozygous	Splice donor lost (Pathogenic)	Deceased
30b	Female	Yes	1 month	CNS	Not done	LAMB2	c.1405+1G>A	Homozygous	Splice donor lost (Pathogenic)	Deceased
31	Female	No	5 year	SRNS	FSGS	MYO1E	c.505C>T	Homozygous	p.(Arg169*) (Pathogenic)	Normal GFR
32a	Male	Yes	7 month	INS	DMS	PLCE1	c.4301G>T; c.4306del	Compound heterozygous	p.(Arg1434Leu) (Pathogenic); p.(Val1436*) (Pathogenic)	KT
32b	Male	Yes	2 year	SRNS	DMS	PLCE1	c.4301G>T; c.4306del	Compound heterozygous	p.(Arg1434Leu) (Pathogenic); p.(Val1436*) (Pathogenic)	KT
33	Male	Yes	7 year	INS	DMS	PLCE1	c.4306del	Homozygous	p.(Val1436*) (Pathogenic)	PD
34a	Female	Yes	4 year	SRNS/CKD	IFTA	NUP93	c.1319T>C	Homozygous	p.(Phe440Ser) (Uncertain Significance PP3 PM2 PP1 BP1)	PD
34b	Female	Yes	4 year	SRNS/CKD	IFTA	NUP93	c.1319T>C	Homozygous	p.(Phe440Ser) (Uncertain Significance PP3 PM2 PP1 BP1)	PD
35a	Male	Yes	1 month	CNS	Not done	No				Deceased
35b	Male	Yes	2 month	CNS	Not done	No				Deceased
36	Female	No	1 month	CNS	Not done	No				Deceased
37	Female	No	1 month	CNS	Not done	No				Deceased
38	Female	No	3 month	CNS/CKD	Not done	No				PD
39	Female	No	1 month	CNS	Not done	No				CKD
40	Male	No	3 year	SRNS	FSGS	No				Deceased
41	Male	No	4 year	SRNS	FSGS	No				Normal GFR
42	Female	No	4 year	SRNS	FSGS	No				HD
43	Male	No	10 year	SRNS	FSGS	No				KT
44	Male	Yes	10 year	SRNS	FSGS	No				Normal GFR
45	Male	No	2 year	SRN/CKD	FSGS	No				PD
46	Male	No	3 year	SRNS	FSGS	No				Normal GFR
47	Male	No	4 year	SRNS	MCD	No				CKD
48	Female	No	1 year	INS/CKD	FSGS	No				Deceased
49	Female	No	5 month	INS/CKD	FSGS	No				Deceased
50	Female	No	5 year	SRNS	FSGS	No				CKD

Note: Reference Sequences: NPHS1 NM_004646.4; NPHS2 NM_014625.4; LAMB2 NM_002292.4; MYO1E NM_004998.4; PLCE1 NM_016341.4; NUP93 NM_014669.5.

Abbreviations: CKD, chronic kidney disease; CNS, congenital nephrotic syndrome; Comp het, compound heterozygous; DMS, diffuse mesangial sclerosis; FSGS, focal segmental glomerulosclerosis; FT, finish type nephrotic syndrome; HD, hemodialysis; Het, heterozygous; Homo, homozygous; IFTA, interstitial nephritis with tubular atrophy; INS, infantile nephrotic syndrome; KT, kidney transplantation; M, month; MCD, minimal change disease; NGFR, normal glomerular filtration rate; PD, peritoneal dialysis; SRNS, steroid resistance nephrotic syndrome; Y, year. Shading in pale green is for NPHS1 variants; pale blue for NPHS2 variants; peach for LAMB2 variants; orange for MYO1E variants; green for PLCE1 variants; blue for NUP93 variants.

NPHS2 pathogenic variants were the most common, detected in 36 patients; pathogenic variants in NPHS1 were detected in 16 patients, pathogenic variants in PLCE1 were detected in three patients, pathogenic variants in LAMB2 and NUP93 were detected in two patients each, and a pathogenic variant in MYO1E was detected in one patient (Figure 1b). NPHS1 was the most frequent genetic cause of CNS whilst NPHS2 had a range of ages of presentation (Figure 1c).

3.2. NPHS1 genetic findings

Sixteen (21.6%) patients had disease‐causing variants in NPHS1 and all of them presented with CNS. The mean age at presentation was 1.25 ± 0.57 months (range 0–3 months). Consanguinity was reported in 14 patients and 9 patients had a family history of CNS. Kidney biopsy was performed for 5 patients, all of which showed histopathological features of Finnish‐type NS. Molecular analysis of NPHS1 showed that 12 patients had homozygous disease‐causing variants, while 4 had compound heterozygous disease‐causing variants. In total, seven different pathogenic variants were detected in NPHS1 gene (Table 1). The c.515_517 delCCA, p.(Thr172del) was the most common variant, detected in homozygous state in eight patients from five different families. All detected NPHS1 variants were already known (Heeringa et al., 2008; Schoeb et al., 2010), except for c.2663G>A, p.(Arg888Lys), that was detected in two siblings (family 10) in a homozygous state. Dialysis was performed for six patients, and two of these patients underwent kidney transplantation.

3.3. NPHS2 genetic findings

Thirty‐six patients from 18 different families had disease‐causing variants in NPHS2; of these, 28 patients had childhood SRNS and 8 patients had INS, none presented with CNS. The mean age at presentation was 29.8 months. Consanguinity was seen in 34 of these patients and 30 patients had a family history of kidney disease. Kidney biopsy was performed for 21 patients and the histopathologic findings were FSGS in 20 patients and MCD in 1 patient. Thirty‐two of these patients had homozygous causative variant, whereas only four patients had compound heterozygous pathogenic variants. There were six different pathogenic variants detected in NPHS2 all of which have been already reported previously in HGMD. The most common was c.779T>A, p.(Val1260Glu) seen in homozygous state in 18 patients. Dialysis was performed in 21 patients, and 12 of these patients received a kidney transplant (Table 3).

TABLE 3.

Summary of the clinical follow‐up, treatment, and outcome of the studied cohort.

	Total	NPHS1	NPHS2	Other genetic causes	No identified genetic cause
Mean follow‐up period (month)	59.3	53.1	74.9	31.1	39.4
Immunosuppressive medications	25	0	16	2	7
KF requiring dialysis	45	6	21	5	13
Kidney transplantation	17	2	12	2	1
Death	21	7	4	2	8

3.4. LAMB2 , MYO1E , PLCE1 , and NUP93 genetic findings

Eight patients from four different families were genetically diagnosed using WES. The first family (Family 30) involved two siblings who presented with severe CNS from birth and developed progressive CKD and died within 3 months of age. In addition, both siblings presented with eye phenotypes, including absence of a red reflex, hypopigmented iris with a pinpoint iris. They were found to have a known homozygous splicing variant (c.1405+1G>A) in LAMB2 that was anticipated to disrupt a highly conserved donor splice site (Bredrup et al., 2008).

A homozygous nonsense variant c.505C>T, p.(Arg169*) in MYO1E was detected in a 5‐year‐old female with SRNS (Family 31) with a kidney biopsy demonstrating features of FSGS. The c.505C>T allele has been classified as pathogenic (Feltran et al., 2017; Guaragna et al., 2020). This patient was resistant to prednisolone, mycophenolate and tacrolimus, and immunosuppressive medications were discontinued following the molecular genetic diagnosis, and she was continued on ACE inhibitor that resulted in the preservation of kidney function, with a normal eGFR at the last follow‐up aged 9 years.

PLCE1 pathogenic variants were detected in three children from two different families. In two affected siblings (Family 32), who presented with INS and childhood SRNS, compound heterozygous variants (c.4306del, p.(Val1436*); c.4301G>T, p.(Arg1434Leu)) in PLCE1 gene were detected. The c.4306del, p.(Val1436*) allele is novel and predicted to be disease causing. The c.4301G>T, p.(Arg1434Leu) missense variant is currently classified as a variant of uncertain significance according to the ACMG criteria but has been previously reported in Chinese patients with SRNS (Wang et al., 2017). In silico modeling of this allele demonstrates it is located within the PI‐PLC X‐box domain, important for its catalytic function and that the missense allele is predicted to cause a moderate disruption (Figure 2). Another child (Family 33), presented with INS, and was found to carry the same novel pathogenic allele c.4306del, p.(Val1436*), this time in its homozygous state. A kidney biopsy was performed for these three patients revealing diffuse mesangial sclerosis. All of them developed progressive CKD leading to KF and led to renal replacement therapy in the form of kidney transplantation (Family 32) and dialysis (Family 33).

FIGURE 2.

In silico modeling of PLCE1 p.(Arg1434Leu) missense variant. (a) The PLCE1 protein has several known domains: a 260 amino acid Ras‐GEF domain (p.531–790) (red), 149 amino acid PI‐PLC X‐box domain (p.1392–1540) (yellow), a 117 amino acid PI‐PLC Y‐box domain (p.1730–1846) (dark blue), 101 amino acid C2 domain (p.1856–1956) (pink), 103 amino acid Ras‐associating 1 domain (p.2012–2114) (orange) and a 104 amino acid Ras‐associating 2 domain (p.2135–2238) (cyan). There is a 79 amino acid region (p.1686–1764) that is required for PLCE1 to be activated by RHOA, RHOB, GNA12, GNA13, and G‐beta gamma (purple). (b) The missense p.(Arg1434Leu) variant in the PLCE1 protein (green) occurs in the PI‐PLC X‐box domain (yellow), important for the catalytic function of the protein. (c) Zoom in of the missense p.(Arg1434Leu) variant. The dashed black lines demonstrate the likely loss of the canonical geometry between 2 atoms in this loop with an overlap of each (numerically labeled in Angstroms). The small green disk represents atoms that are almost in contact or slightly overlapping. The red disks represent the significant pairwise overlap of atomic van der Waals radii causing a likely structural ‘clash’ (size of the disk is proportional to the size of the overlap) and the yellow disk lies in between the severity of overlap of red and green. The amino acid volume change is moderate (190.3 to 163.1 Angstroms).

Family 34 included an affected female and her paternal cousin, both presenting with hypertension, iron deficiency anemia, and reaching KF. WES identified a novel homozygous missense allele c.1319T>C, p.(Phe440Ser) in NUP93 (NM_014669.5). Pathogenic variants in NUP93 are associated with NS, type 12 (OMIM 614351) (Braun et al., 2016). Both homozygous missense and compound heterozygous truncating variants with missense variants have been reported in NUP93 in individuals with SRNS (Braun et al., 2016). The c.1319T>C variant has not previously reported in any databases and in silico analysis including structural modeling suggested it is pathogenic, affecting a highly conserved amino acid residue. Segregation of this causative allele within family members was confirmed (Figure 3) adding weight to its disease causality.

FIGURE 3.

Identification and modeling of a novel NUP93 missense variant. (a) Pedigree diagram of family 34, with proband and paternal cousin with SRNS diagnosis. (b) Sanger sequencing identified a missense mutation (p.Phe440Ser) in NUP93 gene (NM_014669.5) in the proband, which was also confirmed in her affected paternal cousin. (c) The predicted three‐dimensional structure of human Nucleoporin 93protein, encoded by NUP93 on chromosome 16, generated using AlphaFold Protein Structure Database (https://alphafold.ebi.ac.uk) and UniProtKB (https://www.uniprot.org/uniprot/) with associated codes: AF‐Q8N1F7‐F1 and Q8N1F7 respectively. (d) Region (labeled in blue) in which the c1319T>C p.(Phe440Ser) lies in an alpha helix near the middle of the protein. (e) Missense SNV c1319T>C, p.(Phe440Ser) modeled with the most probable (41.9% accuracy) rotamer demonstrated. The dashed black lines demonstrate the likely loss of the canonical geometry between two atoms in this loop with an overlap of each (numerically labeled in Angstroms). The red disks represent the significant pairwise overlap of atomic van der Waals radii causing a likely structural ‘clash’ (size of the disk is proportional to the size of the overlap). The amino acid volume change is very significant (190.8 to 93.5 angstroms) suggesting a significant impact of the mutation on the side‐chain structure.

3.5. Children without a genetic diagnosis

We were unable to genetically diagnose 17 patients (mean age at presentation 31.7 months) using Sanger sequencing screening of NPHS1 and NPHS2 (n = 15) and WES (n = 2). Six of these patients presented with CNS, nine had childhood SRNS, and the remaining two presented with INS. Nine of them have CKD at presentation. A kidney biopsy was performed in nine patients, of which eight had FSGS and one had MCD (Table 1, Figure 1a). KF requiring dialysis was performed in 13 patients, and one patient received a kidney transplant. Eight patients died (Table 3); of those five had CNS, two had INS, and one had childhood SRNS.

3.6. Genotype/phenotype correlations

In this study, patients with SRNS in whom we detected disease‐causing genetic variants were slightly younger (median age ~ 25 months) compared with children without genetic causes (median age 32 months). The median age for those with CNS and diagnosed with NPHS1 was 1.25 months (Table 2). Consanguinity was common within this study cohort, and a family history of NS was more commonly seen in children with a genetic cause (NPHS1: 56%, NPHS2: 81.1% and other genes: 87.3%) compared with children without an identified disease‐causing genetic variant (12.5%). CKD at presentation was more common in patients who had an identified genetic cause in LAMB2, PLCE1, MYO1E, and NUP93 and in those without mutation compared with those with NPHS1 or NPHS2 disease‐causing variants.

TABLE 2.

Clinical features of studied patients with SRNS.

	Total	NPHS1	NPHS2	Other genes	No pathogenic variant defined
Number of patients	77	16	36	8	17
Gender (Female/Male)	41/36	11/5	18/18	5/3	8/9
Mean age (month)		1.25	29.83	24.6	31.7
Consanguinity	58	14	34	7	3
Family history of nephrotic syndrome	48	9	30	7	2
Hematuria	60	9	32	7	12
CKD at presentation	23	2	5	7	9
Hypothyroidism	47	15	18	6	8
Hypertension at presentation	21	0	9	6	6
Mean eGFR at presentation (mL/min/1.73 m2)	124	87	165	18	85
Mean eGFR last follow‐up (mL/min/1.73 m2)	53	40	53	12	76
Mean albumin (g/L) at presentation	12.3	7.7	12.5	7.0	16.3
Mean albumin (g/L) last follow‐up	26.0	21.3	28.7	12.5	26.7

Abbreviations: CKD, chronic kidney disease; eGFR, estimated glomerular filtration rate.

3.7. Modality of treatment, follow‐up, and outcomes

The 24 patients presenting with CNS were not given immunosuppression treatment, including all of the patients with NPHS1 mutations. Of the remaining 53 patients, 25 patients received immunosuppressive medication (including steroids with or without other immunosuppressive medication; ciclosporin, tacrolimus, mycophenolate, or rituximab) during the mean follow‐up period of 59.3 months, (Table 3). Of those 25, 16 patients had NPHS2 variants, 2 had genetic diagnosis in other genes and 7 were without a genetic cause detected. Aside from a CNS presentation other reasons for withholding / avoiding immunosuppression included a known family history of NPHS2 mutation (11 cases) and KF at presentation. KF requiring dialysis was necessary for 45 patients, while kidney transplantation was performed for 17 patients. During follow‐up period, 21 patients died, 7 of them had NPHS1 disease‐causing variants, 4 had NPHS2 disease‐causing mutation, 2 had LAMB2 gene disease‐causing variants, and 8 had no identified disease‐causing variants detected (Table 3).

4. DISCUSSION

This study is the first to describe clinical features, genetic analysis, and outcomes of Omani children with SRNS. It comprised 77 children from 50 different families who presented over a 15‐year period. In this study, 60 out of 77 patients had a single gene defect, whereas 17 remained genetically undiagnosed. However, 15 of the undiagnosed cases were only screened for NPHS1 and NPHS2 using direct Sanger sequencing, meaning that an alternative molecular genetic diagnosis may have been missed in these cases. Just 2 cases remained undiagnosed following both targeted Sanger sequencing followed by WES approaches.

NPHS1 and NPHS2 were the two major genetic causes identified in this cohort of SRNS children from Omani. Other genetic causes identified were LAMB2, PLCE1, NUP93, and MYOE1. These data may be compared with other reports of SRNS patients. In Saudi Arabia, the most commonly identified genes causing SRNS were NPHS2 (22%), NPHS1 (12%), PLCE1 (8%), and MYOE1(6%) (Al‐Hamed et al., 2013). In a Japanese SRNS population, the most common genetic causes were seen in WT1 (25%), NPHS1 (12%), INF2 (12%), TRPC6 (10%), and LAMB2 (9%) (Nagano et al., 2020), while in China the genetic causes were ADCK4 (6.67%), NPHS1 (5.83%), WT1 (5.83%), and NPHS2 (3.33%) (Wang et al., 2017). European studies demonstrate a comparable spread of genetic variants in SRNS patients with the most common causes being NPHS1 (17%), WT1 (14%), and NPHS2 (11%) (Morello et al., 2020).

Our study revealed a high rate of molecular genetic diagnoses of SRNS, which is consistent with previous reports (Al‐Hamed et al., 2013; Santín et al., 2011). Most of our patients with genetic causes were born to consanguineous parents and consistent with this, most of the disease‐causing variants were detected in their homozygous state.

In our cohort, disease‐causing variants in NPHS2 were the most common causes of SRNS accounting for 47%, higher than the rate reported in Europe (40%) (Sadowski et al., 2015) and Turkey (29.9%) (Berdeli et al., 2007). We also observed that NPHS2 disease‐causing variants were detected in infants and children ranging between 5 months and 6 years at disease presentation, but were not identified in any patients presenting with CNS. In comparison, previous studies of CNS in non‐Finnish patients reported both NPHS1 and NPHS2 disease‐causing variants, contributing up to 75% of the genetic causes, but with patients with NPHS2 being less likely to present before first month of life (Machuca et al., 2010; Sadowski et al., 2015).

The most common disease‐causing NPHS2 variant in this study was the homozygous missense allele c.779T>A, p.(Val1260Glu), which was found in 18 of our patients. This allele was previously reported in Saudi Arabian patients (Al‐Hamed et al., 2013) and in four patients from Europe and North Africa by Weber et al. (2004). The second most common NPHS2 allele, c.467dup, p.(Leu156Phefs*11) was seen in 11 patients.

In our study, disease‐causing mutations have been identified in 75% of our CNS cohort, where NPHS1 was the most common causative gene. Our findings are consistent with the reported diagnostic rate of 75%–100% of cases of CNS (Preston et al., 2019). Most of our patients with NPHS1 variants had a homozygous disease‐causing variant, whereas 4 had compound heterozygous alleles. None of the Fin _major or Fin _minor variants were detected in our cohort, which was consistent with a previous study that showed the rarity of these mutations in non‐Finnish patients (Al‐Hamed et al., 2013; Beltcheva et al., 2001; Heeringa et al., 2008). Other studies in non‐Finnish patients found only two patients had Fin _major or Fin _minor mutations (Machuca et al., 2009). A kidney biopsy was performed for five CNS patients and all showed features of Finnish‐type NS. Most of our patients presented within 1 month of life and progressed to KF at the mean age of 31 months.

A homozygous pathogenic variant in LAMB2 gene was identified in 2 siblings (Family 30) who presented with CNS confirming the diagnosis of autosomal recessive NS type 5, this variant was previously reported by Bredrup et al. (2008) and Matejas et al. (2010). These 2 children progressed rapidly to KF and died within 3 months, along with minor eye abnormalities, consistent with a unifying diagnosis of Pierson syndrome. Pierson syndrome (OMIM 609049) is an autosomal recessive disorder secondary to variants in LAMB2 and is characterized by CNS, neurodevelopment abnormalities and eyes anomalies including microcoria and hypoplasia of the ciliary and pupillary muscles (Zenker et al., 2004). Usually, these symptoms present in utero or within the first 3 months of life, as in our identified patients. In contrast, Kagan et al. (2008) reported a mild presentation of an infant with CNS, who had minor eye abnormalities without any neurological abnormalities and developed KF at 16 months of age.

A novel PLCE1variant was identified in our cohort (Family 32 and 33) that is a frameshift deletion variant c.4306del, p.(Val1436*). All patients with PLCE1 pathogenic variants developed progressive CKD and started renal replacement therapy (dialysis/kidney transplantation). Pathogenic variants in PLCE1 gene are associated with NS type 3 and patients typically have early onset and severe NS with rapid progression to CKD, where most of them illustrate DMS and some FSGS on kidney biopsy (Hinkes et al., 2006). Although previous reports showed a partial response of NS patient with disease‐causing mutations in PLCE1 to tacrolimus therapy, further studies are still needed to determine the reno‐protective effect of such proteinuria reduction (Lin et al., 2014).

In this study, we identified only one child with a homozygous nonsense variant in MYO1E which was reported previously (Feltran et al., 2017; Guaragna et al., 2020). This patient's NS was resistant to immunosuppressant medications. MYO1E gene encodes myosin 1E, which is a non‐muscle class 1 myosin, that regulates the podocyte cytoskeleton (Kaplan et al., 2000; Kim et al., 2003). The MYO1E variants are associated with childhood onset autosomal recessive FSGS (Mele et al., 2011). The MYO1E is uncommon etiology for childhood FSGS but is seen more frequently in consanguineous families (Al‐Hamed et al., 2013; Mele et al., 2011).

The genetic findings of our study have a valuable impact on the early clinical management of studied children as well as future genetic counseling for affected families. The detection of causative variants in NS‐associated genes clarifies the accurate mode of inheritance and facilitates proper counseling of family members as well as guides the setting for kidney transplant and selection of living related kidney donors. Our study has some limitations including the small cohort size as well as the genetic screening methodologies used. Initially, we focused on the targeted screening of two main genes NPHS1 and NPHS2 and only selective cases were taken forward for WES, due to local resource limitations. To improve detection rates, searching for disease‐causing variants in all monogenic NS‐causing genes is required through modern NGS approaches including targeted gene panels, WES and whole‐genome sequencing. Such strategies of molecular genetic diagnosis are anticipated to improve our understanding of etiology of SRNS in distinct populations and enhance personalized medicine in the term of individualizing management and avoiding immunosuppressive medications.

5. CONCLUSION

NPHS1 and NPHS2 disease‐causative variants were the most commonly detected molecular genetic causes of SRNS in Omani children. Using WES increases the detection ability of the other rarer genetic causes of SRNS. In view of this, NGS screening of all known genes underlying NS should be implemented in all children who present with SRNS to allow optimization of treatment and prediction of kidney survival outcomes.

AUTHOR CONTRIBUTIONS

MSAR and IAA conceived the study and wrote the first draft of the manuscript. Clinical and molecular data analysis was performed by MSAR, IAA, BAG, AAM, NAK, NAAH, AAB, MAS, SAS, MAB, and FAH. In silico modeling was performed by HM. JAS coordinated the work and edited the manuscript. All authors read and approved the final manuscript.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

ETHICS STATEMENT

This study was ethically approved by the Royal Hospital Research Ethical Committee, Ministry of Health (MOH/CSR/21/24412). For genetic studies of patients, written informed consent was provided by the patients' families.

Al Riyami, M. S. , Al Alawi, I. , Al Gaithi, B. , Al Maskari, A. , Al Kalbani, N. , Al Hashmi, N. , Al Balushi, A. , Al Shahi, M. , Al Saidi, S. , Al Bimani, M. , Al Hatali, F. , Mabillard, H. , & Sayer, J. A. (2023). Genetic analysis and outcomes of Omani children with steroid‐resistant nephrotic syndrome. Molecular Genetics & Genomic Medicine, 11, e2201. 10.1002/mgg3.2201

DATA AVAILABILITY STATEMENT

The data are not available for public access because of patient privacy concerns but are available from the corresponding author on reasonable request.