Mutation Analysis of NPHS1 in a Worldwide Cohort of Congenital Nephrotic Syndrome Patients

Abstract

Background

Congenital nephrotic syndrome (CNS) is defined as nephrotic syndrome that manifests within the first 3 months of life. Mutations in the NPHS1 gene encoding nephrin, are a major cause for CNS. Currently, more than 173 different mutations of NPHS1 have been published as causing CNS, affecting most exons.

Methods

We performed mutation analysis of NPHS1 in a worldwide cohort of 20 families (23 children) with CNS. All 29 exons of the NPHS1 gene were examined using direct sequencing. New mutations were confirmed by demonstrating their absence in 96 healthy control individuals.

Results

We detected disease-causing mutations in 9 of 20 families (45%). Seven of the families showed a homozygous mutation, while two were compound heterozygous. In another 2 families, single heterozygous NPHS1 mutations were detected. Out of 10 different mutations discovered, 3 were novel, consisting of 1 splice site mutation and 2 missense mutations.

Conclusion

Our data demonstrate that the spectrum of NPHS1 mutations is still expanding, involving new exons, in patients from a diverse ethnic background.

INTRODUCTION

Congenital Nephrotic Syndrome

Congenital nephrotic syndrome (CNS) is defined as nephrotic syndrome manifesting by the 90^th day of life. CNS of the Finnish type (CNF; MIM#256300) is a recessively inherited disorder first described in highly inbred Finnish communities [1, 2]. CNF is characterized by massive proteinuria at birth, a large placenta and marked edema occurring within the first 3 months of life [1, 3–5]. Renal histology shows mesengial hypercellularity and matrix expansion that progresses with age towards complete mesangial sclerosis and capillary obliteration [2]. Irregular microcystic dilatation of the proximal tubules is the most typical histologic feature, but is not observed in all cases [6–8]. Ultrastructural analysis of the glomerular capillary loops show complete foot process effacement and swelling of endothelial cells [9]. The course of the disease is progressive, leading to end-stage renal disease by 2–3 years of age.

NPHS1

By positional cloning, CNF was shown to be caused by mutations in NPHS1 [10]. The Fin_major mutation (nt-121delCT, L41fsX91) and Fin_minor mutation (c.3325 C>T,R1109X) in the NPHS1 gene were the first mutations to be discovered and the most prevalent mutations of CNF in the Finnish population (98% of cases) [10]. However, these mutations are also found in other ethnic groups [11, 12]. Screening for NPHS1 mutations in patients of non-Finnish origin has shown that the frequency of NPHS1 mutations is lower than that in Finnish patients, accounting for 39–50% of non-Finnish cases with CNS [13, 14]. On the other hand, rare cases with a manifestation beyond the age of 90 days have also been published, indicating that different mutations in NPHS1 might cause a spectrum of clinical severity [15, 16]. To date, 173 different mutations in NPHS1 have been described (http://www.biobase-international.com). One striking finding among patients with CNS has been the detection of mutations in the NPHS2 gene, encoding podocin, which has been implicated in early-onset steroid-resistant nephrotic syndrome [17]. NPHS2 was shown to be mutated in almost 50% of cases with CNS who are of European origin [13]. In addition to the mutations in the NPHS1 and NPHS2 genes, further genetic heterogeneity has been demonstrated in CNS cases: PLCE1 and WT1 cause CNS and diffuse mesangial sclerosis (DMS) [18–21]. Mutations of LAMB2 cause Pierson syndrome, as a part of a syndromic entity [22] with nephrotic syndrome and microcoria, or as isolated nephrotic syndrome [23, 24].

Mitochondropathies in which the coenzyme Q10 biosynthesis pathway is disrupted may cause monogenic CNS along with neuromuscular symptoms as in mutations of the COQ2 [25], COQ6 [26] and PDSS2 genes [27].

Nephrin

NPHS1 codes for the nephrin protein, an essential component of the interpodocyte-spanning slit diaphragm [28]. Nephrin is a transmembrane protein of the Ig superfamily characterized by eight C2-type Ig-like domains and a fibronectin type III-like module in the extracellular region, a single transmembrane domain and a cytosolic C-terminal end [10]. Mutations in NPHS1 lead to disruption of the filtration barrier and cause massive protein loss. Nephrin plays a significant role in signaling between podocytes by interacting with CD2AP and podocin [29, 30].

METHODS

Patient and Data Ascertainment

Within a worldwide cohort of children with nephrotic syndrome referred to us since May 2008 for mutational analysis, we selected all the patients who had nephrotic syndrome onset within the first 90 days of life. These were a total of 25 patients from 22 families.

Patients with mutations in the other genes known to cause CNS were excluded from the study. The frequency of mutations, clinical signs, renal and extra-renal signs, and the results of the renal biopsy make up the basis for the choice of genes to be tested. First, we performed mutation analysis for NPHS2 and WT1 for all 25 patients since these are the most frequent monogenic causes of childhood NS [13]. One patient (A3318 II-1) was found to have a homozygous mutation in NPHS2 gene Ex2: c.353 C 1 T (H) (p.P118L) [31] and another patient (A3194 II-1) was revealed to have a novel heterozygous mutation in WT1 gene Ex8: c.1097 G>A (h) (p.R366H). These two patients were excluded from the cohort. Additionally, screening for all 31 exons of PLCE1 was performed in 2 patients (A3205 II-1 and A3360 II-1) with CNS because they revealed renal histology of DMS [18, 19]. However, none of these patients had a mutation in the PLCE1 gene. There were no additional signs for other CNS causing genes to be tested in our cohort; therefore, we performed mutation analysis for NPHS1 for the remaining 23 patients with CNS from 20 families.

Human subject research was approved by the University of Michigan Institutional Reviews Board and the Ethics Commission of the University of Freiburg, Germany. The diagnosis of CNS was made by pediatric nephrologists in specialized centers based on published criteria [32]. Following informed consent, detailed clinical and pedigree information was obtained by a standardized questionnaire available on www.renalgenes.org. Nephrotic range proteinuria was defined as proteinuria >40 mg/m²/h. When evaluating the frequency of mutations, we relate them to families rather than patients because siblings have identical mutations. When evaluating clinical data, we relate them to patients because siblings might differ in their clinical phenotypes. It was shown in one of our previous studies that out of the two siblings with the homozygous missense mutation in NPHS1 gene Ex14: c.1760 T 1 G (H) (p.L587R), only one developed nephrotic syndrome before the age of 90 days, while the other did not manifest until the age of 2 years [33].

Mutation Analysis

Genomic DNA was isolated from blood samples using the Puregene® DNA purification kit (Gentra, Minneapolis, Minn., USA) following the manufacturer’s guidelines. Mutation analysis by direct exon sequencing was performed using exon-flanking primers and by direct sequencing of all the exons for NPHS1, NPHS2, PLCE1 and PAX2. WT1 analysis was limited to exons 8 and 9 since mutations of this gene that account for isolated NS has almost exclusively been reported in these two exons [21, 34]. Exon primers for NPHS1, NPHS2, WT1 and PLCE1 have been published previously [14, 19, 21, 34, 35]. For sequence analysis the software SEQUENCHER 3.8 TM (Gene Codes, Ann Arbor, Mich., USA) was used. The published reference sequence of NPHS1 (NM_004646) was used as the relevant wild-type gene sequence. Sequencing of both DNA strands was performed for all detected mutations and other sequence variants. If parental samples were available, segregation of the variants was confirmed by direct sequencing of parental samples. For each novel mutation, its absence was demonstrated in 96 healthy control individuals of matched ethnic origin by direct sequencing. We here define ‘disease-causing mutations’ as the presence of both alleles of a recessive-disease gene (NPHS1 or NPHS2) and one allele of a dominant disease gene (WT1) that are absent from 96 healthy control individuals and from the ‘1,000 genomes’ database www.1000genomes.org).

Results

Patient Characteristics of the CNS Cohort

In this study, a worldwide cohort of 23 patients from 20 families with CNS was included. All patients were examined for NPHS1 mutations. Families were from the following ethnicities: 7 Caucasian, 2 Turkish, 4 Arabic, 3 Indian, 2 Pakistan, 1 Vietnamese and 1 Hispanic. Eight patients from 8 families were from consanguineous parents. Renal biopsy was performed in 7 of them and showed CNF (3 patients), DMS (2 patients), glomerular mesangial proliferation (1 patient) and mesangioprolifarative glomerulonephritis (1 patient). Because traditionally CNS is considered treatment refractory, 18 patients (78%) did not receive any therapy. In 3 patients (13%) steroid therapy was administered: 2 of them (A3319 II-1 and A3358 II-1) did not respond to the steroid therapy (steroid-resistant nephrotic syndrome). A3358 II-1 was then started on CPA and CsA therapy. The third patient (A3449 II-1) died 24 hours after the administration of steroid therapy so no data on the response was available. One patient (A3360 II-1) was partially responsive to cyclosporine A (CsA). One patient (A3325 II-1) was on antiproteinuric therapy with angiotensin-converting enzyme inhibitors.

NPHS1 Mutations

Mutation analysis by direct exon sequencing of all 29 exons of NPHS1 was performed for a total of 23 patients from 20 families. Both causative NPHS1 mutations were detected in 9 of the 20 families (45%; table 1); therefore, the CNS phenotype is fully explained. NPHS1 mutations represent a recessive single-gene cause of CNS. Recessive single-gene disease causes convey full penetrance of a disease. They are thus distinct from genetic variants that are found only to be associated with disease because associated variants usually explain only a low percentage of the phenotypic variance, as are the cases for instance in the MYH9/APOL1 [36] and HLA [37–42] variants that have been found in nephrotic syndrome.

Table 1.

Clinical and mutation information for 11 families with NPHS1 mutations detected

Patient number	Origin	Known Consan- guinity	Age of onset	Gender	Renal biopsy	Treatment	Other clinical features	NPHS1 mutationa (Exon: nucleotide change; aminoacid change)		Origin of mutation	Initial phenotype
Homozygous mutations
A3205 II-1	Caucasian	No	53 d	M	DMS	No treatment	none	EX18:c.2491C>T;p.R831C	Lenkkeri et al. 1999	North America	CNS, Finnish type
A3235 II-3	Arabic	Yes	2 mo	F	Not done	No treatment	none	EX2: C.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3236 II-1	Indian Subcontinent	Yes	1 mo	F	Not done	Conservative treatment	none	IVS 7+1 G>T; splice errorb	This study
A3325 II-1	Pakistan	Yes	2 mo	M	Not done	Albumin infusion, Lisinopril	Grand mal seizures; brother 4 y normal.	EX6:c.614-621delinsTT; p.T205,P206,R207>1205	Lenkkeri et al. 1999	Turkey	CNS, Finnish type
A3337 II-3	Arabic	Yes	1 mo	F	Not done	No treatment	Edema at birth, low set ears, depressed nasal bridge, high arched palatet; two deceased brothers (sample not available).	EX2: C.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3416 II-1	Indian Subcontinent	Yes	13 d	M	Not done	No treatment	Premature (34 weeks).	EX2: c.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3442 II-2	Indian Subcontinent	No	1 mo	M	CNF	ND	Microcephaly, aminoaciduria, 3+ glycosuria and acidosis suggesting proximal tubular defect. Died at 6 mo of age, his older sister died at age 3.5 y. Mother had oligohydramnios during pregnancy.	Ex9:c.1099 C>T; p.R367C	Lenkkeri et al. 1999	France	CNS, Finnish type
Compound heterozygous mutations
A3322 II-2	Caucasian	No	no data	M	Not done	No treatment	Degrees of proteinuria, not frank NS, hypothyroidism, hypertension, acidosis.	Ex22: c.2930A>G; p.Y977Cc	This study
								EX27: c.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3322 II-3	Caucasian	No	no data	F	Not done	No treatment	Degrees of proteinuria, not frank NS.	Ex22: c.2930 A>G; p.Y977Cc	This study
								EX27: c.3478 C>T; p R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3322 II-4	Caucasian	No	no data	F	Not done	No treatment	Degrees of proteinuria, not frank NS.	Ex22: c.2930 A>G; p.Y977Cc	This study
								EX27: c.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3322 II-5	Caucasian	No	no data	M	Not done	No treatment	Degrees of proteinuria, not frank NS.	Ex22: c.2930 A>G; p.Y977Cc	This study
								EX27: c.3478 C>T; p.R1160X	Lenkkeri et al. 1999	Italy	CNS, Finnish type
A3326 II-1	Hispanic	No	1 mo	F	CNF	No treatment	Unilateral nephrectomy (3/2009) with normal renal function afterwards, left inguinal hernia.	Ex2: c.139delG; p.E46fsX127	Heeringa et al. 2008	Hispanic	Nephrotic syndrome
								Ex13: c.1701 C>A; p.C567X	Beltcheva et al. 2001	Non-Finnish	CNS, Finnish type
Sinqle Heterozygous mutations
A3237 II-1	Caucasian	No	3 d	M	Not done	ND	CNS, born preterm (33+3), unexplained cardiorespiratory arrest day 5.	EX6; c. 644 T>G; p. L215R	This study
A3319 II-1d	Turkish	No	42 d	F	Glomerular mesangial proliferation	SRNS	none	Ex 15: c.2014 G>A; p.A672T	Machuca et al. 2010	France	CNS, Finnish type

CNS = congenital nephrotic syndrome; CSA = cyclosporin-A; d = days; DMS = diffuse mesangial sclerosis; Ex = exon; F = female; FSGS = focal segmental glomerulosclerosis; M = male; mo = months; MPGN = mesangial proliferative glomerulonephritis; ND = No data; NS = nephrotic syndrome; SRNS = steroid resistant nephrotic syndrome; SSNS = steroid sensitive nephrotic syndrome; y = years.

^aAll novel mutations were absent from 96 Turkish control individuals and from the 1,000 genomes project (http://www.1000genomes.org). Novel mutations are printed in bold. Novel missense mutations were conserved through evolution at least down to Danio rerio. RefSeq NM_004646 was used as relevant wild type gene sequence for human NPHS1.

^bThe novel mutation is shown to be segregating from mother and father.

^cThe novel mutation is shown to be segregating from father and the known mutation from mother.

^dA3319 II-1 also has a single heterozygous mutation in NPHS2 Ex5: c.729G>C; p.E273Q.

Nine families had disease-causing mutations in NPHS1. The affected individuals of 7 families were harboring the mutation homozygously (table 1). In these, we discovered one novel homozygous splice site mutation (IVS 7 + 1 G>T) in family A3236 (fig. 1). The affected individuals of 2 families (A3322 and A3326) were found to have compound heterozygous mutations (table 1). Family A3322 is compound heterozygous for a novel mutation in Ex22: c.2930 A>G (p.Y977C) along with the known mutation Ex27: c.3478 C>T (p.R1160X) [11] (fig. 1). p.Y977C is conserved down to Danio rerio with Poly-Phen1 [43] and PolyPhen2 [44] scores that are classified to be ‘probably damaging’ (online suppl. table 1, www.karger.com/doi/10.1159/000337379).

Fig. 1. Novel NPHS1 mutations found in this study.

One novel homozygous splice site mutation was found in NPHS1 (IVS 7+1 G>T) in patient A3236 II-1. Additionally, two novel heterozygous mutations, Ex22: c.2930 A>G; p.Y977C and Ex6: c.644 T>G; p.L215R, were detected in family A3322 (all affected) and A3237 II-1, respectively. For each mutation, sequence from the patient and a healthy control individual is shown.

Two patients from 2 families were found to have a single heterozygous mutation only (A3237 and A3319). One of these families (A3237) was carrying a novel single heterozygous mutation in Ex6: c.644 T>G (p.L215R) (fig. 1). p.L215R is conserved down to Caenorhabditis elegans with PolyPhen1 [43] and PolyPhen2 [44] scores that are classified to be ‘probably damaging’ (online suppl. table 1). Therefore, we speculate that our exon sequencing may have missed the second recessive mutation, e.g. a deletion or duplication or intronic mutations or mutations in the promotor region.

Overall, we discovered 10 different mutations, 3 of them novel, consisting of 1 splice site mutation (IVS 7 + 1 G>T) and 2 missense mutations (p.Y977C and p.L215R). We thereby extended the current NPHS1 mutation spectrum of 173 mutations (http://www.biobase-international.com) by 3 novel mutations.

DISCUSSION

In this study, we were able to define the disease-causing mutation in both alleles of NPHS1 in 9 of 20 families and in one allele only in 2 families. Three mutations were novel.

Biopsies were performed in 4 cases out of 11 CNS families in whom we detected a mutation in the NPHS1 gene. The results were CNF (2 patients), DMS (1 patient) and glomerular mesangial proliferation (1 patient). These data confirm the previous findings that NPHS1 mutations can cause a somewhat broader variety of histological phenotypes other than CNF [13, 14, 33].

Although CNF is classically known to be steroid resistant, several cases of steroid-sensitive patients with NPHS1 mutations have been reported previously [14, 45]. In our CNS cohort of patients, A3319 II-I, who carries the NPHS1 mutation in one allele only, was given steroid therapy. This individual had no response to therapy. Another finding was that patient A3325 II-1, who had a homozygous insertion-deletion in NPHS1, was clinically stable on lisinopril only. Previously, 1 patient compound heterozygous for Fin _minor and a missense mutation was shown to respond to enalapril [46]. In another study, patients with homozygous missense mutations or patients with compound heterozygosity for a missense mutation and a frameshift mutation (or a small homozygous deletion causing a nonframeshift mutation) were shown to have a partial response to antiproteinuric therapy rarely [33].

We previously reported a CNS case of Hispanic origin (A1893) explained by the mutations in NPHS1 gene Ex2: c.139delG (h) (p.E46fsX127) and c.3482–2 A>G (h) (splice site) [14]. In the current study, we found the same mutation (p.E46fsX127), but this time in compound heterozygosity with Ex13: c.1701 C>A (h) (p.C567X) [47] in patient A3326 II-1, who was also of Hispanic origin. A3326 II-1 was biopsied and proven as CNF and underwent unilateral nephrectomy. The patient had normal renal function of his unilateral kidney afterwards. A previous study showed that CNS management with captopril and indomethacin therapy in combination with unilateral nephrectomy achieves significant improvements in plasma albumin and reduces the need for albumin infusions and time in hospital; therefore, second nephrectomy, dialysis and transplantation can be delayed until the 3rd year of life or longer [48].

Another finding in this study was that family A3416 from the Indian subcontinent was found to have the mutation p.R1160X. In a previous study, p.R1160X was shown to be suggestive of a founder effect and therefore commonly known as the ‘Maltese mutation’ [11]. This mutation was also detected in CNF cases of Indian/Bangladesh origin, but associated with a different allele [11]. In the same study, p.R1160X resulted in an unexpectedly mild CNF phenotype in about half of the cases [11]. In our study at least a part of the patients with this mutation have a very early onset of CNS (13 days, 1 month and 2 months, respectively).

The classical notion that NPHS1 mutations are seen in nephrotic syndrome cases with age of onset in the first 90 days of life was changed by the recent discovery that NPHS1 mutations may cause onset beyond the first 3 months [15]. Previously, it was demonstrated that homozygosity mapping is a useful tool for screening for homozygous disease causing mutations in NPHS1 [33, 49]. In our study, we also screened 9 families for NPHS1 mutations (A3113, A3191, A3310, A3317, A3321, A3323 (2 sibs), A3327, A3329, A3377) that had the onset of NS symptoms beyond 90 days of life with single nucleotide polymorphism arrays (Gene Chip®) from Affymetrix, Inc. with a resolution of 250K (Human Mapping 250K Styl Array). The method has been described in detail previously [33, 49]. However, none of these families were found to have a disease-causing mutation in the NPHS1 gene by direct sequencing (data not shown).

In previous studies, it was shown that approximately one half of CNS cases are caused by recessive mutations in NPHS1 [13, 14, 33]. The NPHS1 mutation rate in our cohort was 45%, accordingly.

Of the mutations described in this study, R1160X was the most frequent. This mutation was found homozygously in 3 families and was found in a compound heterozygous state with the novel mutation Y977C in one family with 4 affected siblings (A3322), accounting for 7 of 40 alleles (17.5%).

Regarding the families in which we did not detect disease-causing mutations in NPHS1, NPHS2, PLCE1 or WT1, we cannot exclude mutations in regulatory elements or introns or heterozygous whole exon deletion as the missing allele since we used an exon sequencing approach that might not detect these mutations. We speculate that mutations affecting other essential slit diaphragm proteins or interaction partners of nephrin may cause the disease in these patients.