Nephrotic syndrome (NS) is frequently seen as an idiopathic primary disease that represents a heterogeneous group of glomerular disorders occurring mainly in children. It is generally divided into steroid-sensitive and steroid-resistant forms, depending upon the patient's response to steroid therapy. While most patients show a favorable outcome after steroid treatment, 10% of children with idiopathic NS do not respond to steroids or to any other immunosuppressive therapy, and progress to end-stage renal disease (ESRD). Renal histology typically shows focal segmental glomerulosclerosis (FSGS). Furthermore, in approximately one third of transplanted patients, the clinical course is complicated by a recurrence of proteinuria, histologically characterized by reappearance of FSGS in the transplanted kidney, which may ultimately compromise allograft function1. Several pathogenic factors have been postulated in these cases of recurrent disease after transplantation, among these a proteinuric circulating permeability factor, so far unidentified, produced in the setting of T-cell dysfunction2,3.
Diffuse mesangial sclerosis (DMS) is a separate histologic entity, clinically presenting with SRNS in early childhood and rapidly leading to ESRD in the first years of life. Renal histology shows enhanced production of mesangial matrix, thickening of the glomerular basement membrane (GBM), and podocyte hypertrophy.
Over the last decade, mutations in genes encoding podocyte proteins have been identified in several forms of hereditary NS. Mutations in the Wilm's tumor suppressor gene WT1 were discovered in patients with Denys-Drash, Frasier syndrome, and DMS4,5,6,7. NPHS1, which encodes nephrin, has been identified as the causative gene in congenital NS of the Finnish type (CNF)8. Mutations in ACTN4 encoding 
-actinin-4 were discovered in several families with autosomal-dominant FSGS, a form of hereditary NS with late onset and progression to ESRD9. Very recently, a heterozygous splice-site mutation in CD2AP, encoding CD2-associated protein, a component of the glomerular filtration complex, was described in two unrelated patients with primary FSGS, pointing to additional genes which might be involved in human susceptibility to glomerular disease10. By positional cloning, a novel podocyte-specific protein, podocin, encoded by NPHS2 on chromosome 1q25-31 and associated with autosomal-recessive forms of SRNS, was described by Boute et al11. Podocin is a 383-amino acid membrane protein that shares sequence homology with proteins of the band-7 stomatin family. It has a predicted hairpin-like structure and is expressed exclusively at the slit diaphragm, with both protein termini on the cytosolic side12. Podocin has been shown to accumulate in an oligomeric form in the slit diaphragm, where it interacts with CD2AP and nephrin13. It has been postulated that podocin may serve a scaffolding function in lipid rafts, and it has also been shown to direct the targeting of nephrin to lipid rafts, where it augments nephrin signaling14,15,16.
NPHS2 mutations were first detected in familial autosomal-recessive SRNS, which manifests early in childhood and presents histologically with FSGS11. Thereafter, NPHS2 mutations were also identified in sporadic cases of SRNS17,18,19, late onset FSGS20, and, surprisingly, in a few patients with recurrence of proteinuria after renal transplantation21. Since the discovery of podocin, we have performed mutational analysis and clinical follow-up of a large cohort of patients with either familial or sporadic SRNS originating mainly from France and North African countries. We present our results, correlating the age of onset and likelihood of post-transplantation recurrence in these patients with the nature of the underlying NPHS2 mutation/variant, and compare our results with other published series17,18,19,20,21,22,23,24.
METHODS
Patients
A total of 338 patients belonging to 272 families were included in the present study. They originated from a large number of countries worldwide, most of them, however, from Europe (mainly France) and from North Africa. Of these, 319 presented with SRNS, defined as lack of response to four weeks of treatment with prednisone followed by three bolus injections of methylprednisolone. Among them, 147 patients belonging to 81 families (44 being consanguineous) were classified as autosomal-recessive (AR) SRNS, defined as SRNS in families with either two or more affected children, or one (or more) affected individual in consanguineous families. This group included 16 families that were previously described in our initial study11. One hundred seventy-two patients presented with sporadic SRNS. Other distant family members were affected with proteinuria or NS in 7 cases. As the patterns of inheritance in these families were not suggestive of autosomal-recessive disease, the index cases were included in the group of sporadic SRNS. In addition, 19 patients (of 19 families) with DMS were also included in the present study.
Mean age at manifestation of NS was recorded in 286 patients and was 57.6 
6.8 months in AR SRNS (N = 107), whereas it was higher in sporadic cases (102.9 8.9 months; N = 169), especially in the cases with recurrence after transplantation (134.3 27.3 months; N = 24). Manifestation was early in the patients affected by DMS (22.1 5.8 months; N = 10). In total, as most of the patients were recruited through pediatric nephrology departments, only 37 developed NS after 15 years (4 patients with AR SRNS and 33 patients with sporadic SRNS, 7 of them with post-transplantation recurrence).
For 261 patients, the results of at least one kidney biopsy were provided to us (106 patients with AR SRNS and 155 sporadic cases). These evaluations showed that 65 patients had predominantly minimal changes (23 patients with AR SRNS, 42 sporadic cases) and 172 patients had FSGS (68 patients with AR SRNS, 104 sporadic cases). In a few cases, mesangial proliferation was observed (15 patients with AR SRNS, 9 sporadic cases) without glomerulosclerosis.
Information on renal transplantation was available for 115 patients with AR SRNS and 170 patients with sporadic SRNS. Transplantation was performed in 44 of the 115 patients with AR SRNS, and no early recurrence was observed in these cases. However, one patient developed proteinuria and FSGS 2 years after receiving an allograft. Among the group of sporadic SRNS, 74 underwent kidney transplantation and 25 of them had a recurrence of proteinuria. In 16 of these, early recurrence was diagnosed immediately after transplantation. Six patients showed a recurrence 14 days to 18 months after transplantation. No post-transplant data were available for 3 patients.
Genotyping and haplotype analysis
Peripheral blood samples were obtained after informed consent of the patients or their parents. Genomic DNA was extracted according to standard laboratory protocols. Five polymorphic microsatellite markers (D1S3758, D1S3760, D1S215, D1S3759, D1S2883) spanning 1.1 cm and flanking the NPHS2 locus were tested. Genotyping was performed after PCR amplification and electrophoresis as previously described in Fuchshuber et al25. Cyrillic 2.1 (Cherwell Scientific, Oxford, UK) was used to construct pedigrees and to perform haplotype analysis.
Mutation screening
Mutation analysis for NPHS2 was carried out by single-strand conformation polymorphism (SSCP) analysis for exons 2–7, as described previously in Boute et al11, and by direct sequence analysis for exons 1 and 8. Exons of the NPHS1 gene were amplified by PCR using flanking intronic primers26 and subsequently sequenced. Direct sequencing was carried out using the Big Dye-terminator method (Applied Biosystems, Foster City, CA, USA). Sequences were evaluated with the Sequencher software (Gene Codes, Ann Arbor, MI, USA). Segregation analysis of mutations in families was performed by direct sequencing, SSCP analysis, or specific restriction enzyme digestion.
Genotyping by single-base extension (SBE) and electrophoresis
In order to discriminate mutations from polymorphisms, all detected missense mutations were screened in 320 control chromosomes by single-base extension using SNaPshot Multiplex kit (Applied Biosystems). We expanded the control cohort to 75 individuals (150 chromosomes) of African descent for the screening of the A242V variant. Exons were amplified using flanking intronic primers. The PCR products from 4 to 6 exons were then mixed with 1.5 
L of ExoSAP-IT (Amersham Bioscience) and incubated for 1 hour at 37°C to remove primers and dNTP, followed by 15 minutes at 80°C for enzyme inactivation. Primer extension reactions27 were performed according to manufacturer's protocol. To clean up the single-base extension reaction, one unit of shrimp alkaline phosphatase (Amersham Bioscience) was added to the reaction and then incubated for 1 hour at 37°C, followed by 15 minutes at 80°C for enzyme inactivation. The extension products were denatured for 5 minutes at 95°C, and electrophoresis was performed using ABI Prism 373 A DNA Sequencer (Applied Biosystems). The results were analyzed using Genescan Analysis and Genotyper (Applied Biosystems).
Statistical analyses
Values reported are the mean standard error of the mean (SEM). Statistical comparisons were performed using one-way analysis of variance (ANOVA). P values of < 0.05 were considered significant.
RESULTS
Linkage analysis
In 62 families suggestive of AR SRNS, linkage analyses were performed with markers flanking the NPHS2 gene locus on chromosome 1q25-31. Results were compatible with linkage to this locus in 36 families, and pathogenic NPHS2 mutations were identified in 25 of them. In the remaining 11 families (all consanguineous, 10 with only one affected child), linkage was compatible as evidenced by homozygosity for the polymorphic makers flanking NPHS2, but no pathogenic NPHS2 mutations were identified. Linkage to the NPHS2 locus was excluded in 26 families, and we performed NPHS2 mutation screening in 17 of these families in parallel. No NPHS2 mutations were detected, as was expected, and further mutation analysis in the 9 additional families was not performed.
Homozygous, compound heterozygous, and single heterozygous pathogenic NPHS2 mutations
NPHS2 mutation screening was thus performed in 322 patients from 263 families. Pathogenic mutations were defined as variants that (1) were not found in 320 healthy control chromosomes, (2) segregated with the disease in families, and (3) were predicted to severely alter protein sequence and/or expression levels. Homozygous or compound heterozygous mutations were identified in 73 patients (42 families), of whom 62 patients presented clinically with AR SRNS (31 families), and 11 with sporadic SRNS. We did not identify pathogenic NPHS2 mutations in any patient with isolated DMS Table 1.
In addition, only one pathogenic NPHS2 mutation was identified in 2 patients from one family with AR SRNS and in 6 patients with sporadic SRNS. The entire coding region and all exon-intron boundaries of NPHS2 were sequenced in order to identify the second mutation, but yielded no additional mutation. Although some mutations may have been missed, such as heterozygous deletions, intronic mutations, or mutations in regulatory regions, it is unlikely that they could account for all the unidentified mutations. In addition to these 8 patients, 7 patients from 4 families (3 AR SRNS and 1 sporadic case) presented with compound heterozygosity of one pathogenic NPHS2 mutation and the R229Q polymorphism on the other allele.
The detection rate of NPHS2 mutations in the homozygous or compound heterozygous states for AR SRNS was 38% (31 out of 81 families studied by mutation screening and/or linkage analysis). On the contrary, mutation detection rate for sporadic SRNS was five times lower at 6.4% (11 affected patients of 172 patients studied). If patients with heterozygous pathogenic NPHS2 mutation with or without R229Q in compound heterozygosity were also included, the mutation detection rates rose to 43% for AR SRNS and to 10.5% for sporadic cases, respectively.
In total, 26 different NPHS2 mutations were detected in the present study; among these, 13 were novel mutations Table 2. These newly identified mutations include missense, nonsense, frameshift and splice-site mutations, and one in-frame deletion of 9 nucleotides, but no large or complete gene deletions. Some mutations were found in several unrelated families. In particular, the R138Q mutation (in exon 3) was observed in 32% of all affected alleles. A large majority of mutations were located in exons 4, 5, and 7, encoding the C-terminal part of podocin. NPHS2 missense mutations cluster in the early C-terminal cytoplasmic domain of the podocin protein, the region conserved among members of the stomatin protein family (see Figure 1). Only one mutation (P118L) was detected in the membrane domain, and 4 (A29T, G35fsX59; G92C; 275–2 A>C) in the N-terminal part of podocin. No missense mutations occurred in predicted N- or O-glycosylation, phosphorylation, or myristolation sites.
Figure 1.
Podocin structure and distribution of mutations. Predicted two-dimensional protein structure of podocin with both termini on the cytosolic side. The region of homology with stomatin is depicted in black. Variants of unknown significance and polymorphisms are shown in italics, nonsense mutations are underlined, and small insertions/deletions are depicted by arrowheads. One splice-site mutation (275–2 A>C) was identified affecting the acceptor splice-site of intron 1.
Full figure and legend (30K)Sequence variations with unknown phenotypic effects and polymorphisms in NPHS2
Several amino acid substitutions (P20L, A61V, L172V, E237Q, and A242V) were identified that either do not markedly alter the predicted protein structure (i.e., conservative amino acid change), and/or affect residues that are not highly conserved among members of the stomatin family. These amino acid substitutions were detected in the heterozygous state in AR and sporadic cases of SRNS, as well as in two DMS patients. Subsequently, the frequencies of these sequence changes were screened by SNaPshot analysis in a control cohort of 320 chromosomes. Screening results revealed that none of the healthy controls showed a P20L, A61V, or L172V sequence variant, indicating that these substitutions are probably not polymorphisms. However, in support of a minor role of these variants, a heterozygous L172V variant was identified in an AR SRNS family without linkage to the NPHS2 gene locus. A homozygous P20L variant was sequenced along with a homozygous pathogenic R168H mutation in a consanguineous family with two affected children. Finally, the A61V variant was observed in a case of sporadic SRNS, as well as in a DMS patient from a consanguineous family.
On the contrary, two amino acid substitutions, E237Q and A242V, were also found in healthy control patients. The E237Q substitution was identified in 2 sporadic cases and in 1 control individual (0.3%). This overall frequency in the healthy population is less than 1%, thereby not fulfilling the definition of an allelic polymorphism. Heterozygous A242V substitutions were sequenced in 7 patients: 2 with AR SRNS, 4 with sporadic SRNS, and 1 with DMS. It was also detected in 2 out of 320 control chromosomes (0.6%). Because most patients and both controls were of African origin, we expanded the control cohort to 75 individuals (150 chromosomes) of African descent and detected 6 individuals heterozygous for the A242V variant, demonstrating that it represents a polymorphism in the African population with an allele frequency of 4%.
In addition, the R229Q polymorphism was identified in 12 out of 320 healthy control chromosomes, defining a frequency of 3.75%, but never in the homozygous state in any control individual. This variant has recently been described by Tsukagushi et al20 as a non-neutral polymorphism. In the present study, a R229Q substitution was detected in 3 families with AR SRNS and in one patient with sporadic SRNS, in addition to a pathogenic mutation on the other allele. Furthermore, homozygous R229Q mutations were identified in 3 AR SRNS (2 of them being consanguineous) and in 2 sporadic cases. No additional NPHS2 mutations were detected in these patients. In 13 patients (11 with sporadic SNRS and 2 with DMS), we identified a heterozygous R229Q polymorphism without any other sequence variation in NPHS2. In total, the screening of 263 unrelated nephrotic patients (526 alleles) revealed 17 with heterozygous R229Q exchange (3.23%) and 5 with homozygous R229Q exchange (1.90%), resulting in an overall allele frequency in the nephrotic cohort of 5.13% (P = 0.35 when compared with the frequency of 3.75% in healthy control individuals). In contrast to what was observed with pathogenic mutations, these variants and polymorphisms were equally distributed either in the N-terminus (P20L and A61V) or the C-terminus part of the protein (L172V, R229Q, E237Q, A242V).
Age at onset
Analysis of the age at discovery of NS was performed among patients with and without pathogenic NPHS2 mutations for all cases in whom such data was available. Patients with pathogenic mutations in homozygous or compound heterozygous states manifested earlier (41.2 5.9 months, N = 63) than patients without two pathogenic mutations in NPHS2 for AR SRNS patients (69.9 11.5 months, N = 55; P = 0.019) and for sporadic cases (108.3 9.4 months, N = 158; P < 0.001). Similar results were found when only patients developing NS before 15 years of age were analyzed (41.5 6.1 months, N = 60 vs. 57.9 7.5 months, N = 53; P = 0.053 for AR SRNS and vs. 56.4 4.1 months, N = 127; P = 0.002 for sporadic cases). With respect to a possible genotype-phenotype correlation, some trends were apparent. Patients with frameshift or nonsense mutations in the homozygous or compound heterozygous state had an early mean age of onset (21.9 5.6 months, N = 13), as was the case with homozygous R138Q patients (11.8 3.2 months, N = 15). Interestingly, in 3 patients with homozygous R138Q mutations, the first clinical symptoms appeared at birth, resembling an intermediate phenotype between congenital and infantile NS. We therefore performed NPHS1 mutation analysis and found a third (heterozygous) mutation in NPHS1 (275–2 A>C acceptor splice-site mutation) in 1 patient. NPHS1 and NPHS2 analysis in the parents of this patient revealed a heterozygous R138Q-NPHS2 mutation in both parents, but no mutations were identified in NPHS1, thereby reflecting a de novo mutation in the patient.
On the contrary, some missense mutations were associated with milder clinical course. In particular, the V180M and the R238S variants were identified in patients with a later onset of SRNS with a median age of 128.7 12.4 months (N = 7). Also, patients with a heterozygous R229Q exchange in addition to one pathogenic NPHS2 mutation and patients with just one heterozygous mutation or variant of unknown significance have a later onset in our study cohort (88.1 12.6 months, N = 7 and 147.4 50.0 months, N = 11, respectively).
NPHS2 mutations and recurrence of proteinuria after renal transplantation
In total, 32 patients with NPHS2 pathogenic mutations, either in the homozygous or compound heterozygous state, underwent kidney transplantation. Among them, only one developed recurrence of proteinuria two years after its third renal transplantation. This patient belongs to a consanguineous family with 3 affected children, and harbors the R138X nonsense mutation. The first and second renal allografts were lost immediately due to acute rejection and acute allograft failure, respectively. Two years after the third allograft (with mother as donor), nephrotic-range proteinuria was detected with renal biopsy disclosing both FSGS and tacrolimus toxicity, but no evidence of rejection. Having shown previously that no NPHS2 mRNA could be detected in this patient's native kidney11, we assayed for serum antipodocin antibodies by indirect immunofluorescence on normal human kidney. However, no antipodocin antibodies were found. In addition, NPHS2 mutation analysis in 25 patients with sporadic SRNS and a recurrence of proteinuria after renal transplantation revealed heterozygous variants in 3 patients, namely T326fsX345, P20L, and E237Q (vs. 2 heterozygous variants/mutations out of 49 transplanted patients with sporadic SRNS and no recurrence of proteinuria after transplantation). The patient with a heterozygous pathogenic T326fsX345 mutation developed proteinuria 18 days after transplantation, and had FSGS on biopsy one year later, ultimately losing the allograft after four years. The second patient with heterozygous P20L variant had a rapid reappearance of proteinuria and allograft rejection immediately after receiving the first transplant. Because renal biopsy showed thrombotic microangiopathy, the contribution of the proteinuric state to allograft loss is likely minimal. However, three months after receiving a second transplant, the patient once again developed massive proteinuria, requiring plasmapheresis in addition to immunosuppressive therapy, including tacrolimus. The patient bearing the E237Q substitution presented with proteinuria and NS 18 months after renal transplantation, which resolved after treatment with iv cyclosporine A and plasmapheresis. In addition, a fourth patient with heterozygous R229Q exchange developed FSGS and proteinuria 22 months after renal transplantation. In all patients, subsequent direct sequencing of the entire NPHS2 coding region and adjacent exon-intron boundaries revealed no second mutation/variant.
DISCUSSION
Our study involving 338 individuals presenting with SRNS represents the largest cohort of patients published to date, and includes both familial autosomal-recessive and sporadic cases. We established a significant contribution of mutations of the NPHS2 gene encoding podocin because more than half (36/62) of families with AR SRNS were compatible with linkage. Pathogenic variants were discovered in 25 of these families, but not in those in whom linkage was not established. We would, therefore, advocate the use of haplotype analysis, using NPHS2 flanking polymorphic markers as an initial diagnostic tool in these families, followed by more rigorous mutation detection when linkage is compatible. The detection of 
40% of mutations in the families with autosomal-recessive inheritance demonstrates the preponderance of NPHS2 mutations in familial SRNS and confirms genetic heterogeneity in SRNS. This is a finding highly reminiscent of autosomal-dominant FSGS28, and emphasizes the need for identifying additional genes involved in familial forms of SRNS. We have extended previously published analyses of sporadic cases of SRNS17,19 by analyzing 172 patients. These patients have a significantly later onset of NS than familial cases. We report the identification of pathogenic mutations in either the homozygous or compound heterozygous or heterozygous state in 10.5% of cases, in agreement with results reported in the two groups of pediatric patients17,19. The lower mutation detection rate among sporadic cases of SRNS than familial cases, and the delayed onset in those in whom mutations where not found, point to complex inheritance patterns (as evidenced by the presence of NS in distant relatives in seven cases), and to potential gene-environment interactions in patients with sporadic NS.
Furthermore, we have shown that NPHS2 mutations are not causative in patients with WT1-negative cases of DMS. We included 19 patients with DMS in this cohort based on previous observations that patients with WT1 mutations can present either with FSGS (Frasier syndrome) or DMS (Denys-Drash syndrome)6,7. Furthermore, Nphs2-deficient mice develop renal failure secondary to DMS29. Our data, however, do not support a pathogenic role for podocin in this subset of patients.
We have identified 32 NPHS2 variants, 26 of which are clearly pathogenic mutations on the basis of deleterious effects on gene expression or predicted structural alteration of the protein. Several of these mutations were identified in unrelated families. The majority of these are missense mutations (15/26), but also include frameshifting (7/26), in-frame deletion (1/26), nonsense (2/26), and splice-site (1/26) mutations. Of these, we here report 13 novel mutations in addition to four nucleotide substitutions of unclear significance based on a conservative change in the amino acid and the lack of residue conservation among stomatin family members. We have also identified the previously reported A242V variant19 as a polymorphism enriched among patients of African origin. In addition, we sequenced the R229Q polymorphism with an overall frequency of 5.12% among our cohort of nephrotic patients.
The mutations are distributed throughout the entire gene, most affecting the C-terminal cytosolic domain. Interestingly, 21/26 variants affect the region of homology with the stomatin protein family11. The accumulation of mutations in this region correlates with the important functional roles ascribed to this domain, namely homo-oligomerization and interaction with other slit diaphragm proteins like nephrin, CD2AP, and the NEPH family14,15,30. Interestingly, genome database analyses point to a novel predicted gene (FLJ32940) expressed in human testis that partly overlaps with exon 6 of the NPHS2 gene, but is oriented in the opposite direction. Fifteen nucleotides are shared by both genes, but no NPHS2 mutation/variant identified so far affects any of the shared nucleotides. Were such mutations to occur, as would be the case with complete NPHS2 deletions, patients might present with a different phenotype.
The nature of the mutation correlates to some extent with the age of onset of nephrotic syndrome. Patients with frameshifting and protein-truncating mutations present earlier with NS. Among missense variants, severity of the disease appears to be determined by the impact of the amino acid substitution on specific functional domains and on the intracellular trafficking of podocin. Our recent functional studies of the intracellular trafficking of several podocin mutants demonstrated that certain missense variants manage to reach the plasma membrane (V180M and R238S), while others, including R138Q, are retained in the endoplasmic reticulum (ER)31. Huber et al have similarly shown that the R138Q mutation, while retaining the ability to homo-oligomerize, is retained in the ER and fails to recruit nephrin to lipid rafts16. The R138Q mutation, which accounts for 32% of mutant alleles in our cohort, predicts mean onset of proteinuria before the first year of life when present in the homozygous state. In contrast, patients with the V180M and R238S mutations have a mean onset of NS much later than patients with the homozygous R138Q mutations (128 vs. 12 months). It will be interesting to see whether pharmacologic approaches that ameliorate the mistrafficking of the missense podocin variants will allow them to function properly, as has recently been shown for the 
F508 CFTR protein32.
With the enlargement of our cohort, we excluded the A242V substitution as a true mutation, as previously reported19. We sequenced this variant mostly in the African population, with an allele frequency of 4%. Additional studies will be required to determine whether its presence imparts an increased susceptibility to proteinuric renal disease. Likewise, we found the E237Q variant in two patients and in one control in our large cohort, therein, highlighting the difficulty of discriminating true mutations from rare polymorphisms. An equally difficult task is to ascribe pathogenicity to three missense variants seen in the heterozygous state, including the P20L substitution, which we identified in two cases (as the sole variant in one and in conjunction with a homozygous R168H mutation in another). This variant affects a nonconserved residue and in our functional analysis reaches the plasma membrane, while the R168H protein does not31.
The identification of two pathogenic NPHS2 alleles in SRNS patients predicts a more severe outcome characterized by earlier onset of disease. This underscores the crucial role of podocin in the maintenance of the filtration barrier. Among cases where only a single mutation/variant is detected, NS tends to occur later in life. Caridi et al reported heterozygous variants in patients with frequently relapsing NS19, while we found one such variant (A61V) in a patient with DMS. One possibility is that the existence of one hypomorphic podocin allele predisposes individuals to glomerular dysfunction in the presence of additional renal insults, as has been recently suggested by Kim et al regarding CD2AP haploinsufficiency10. We have observed the spontaneous development of proteinuria in some aged Nphs2 heterozygous knock-out mice (C. Antignac, unpublished data). Alternatively, these podocin variants may act as modifiers in the setting of lesions in yet undefined genetic loci. Our own studies of Nphs2 null mice have revealed a critical role for genetic modifiers evidenced by disparate rates of progression of renal failure and survival based on the underlying mouse strain29.
Less severe phenotypes have also been noted in familial cases of FSGS involving a pathogenic mutation on one allele and the common polymorphism R229Q on the other allele20. Likewise, we detected seven patients belonging to four families with similar genotypes and a later age of onset of nephrotic syndrome. In addition, we found homozygous R229Q variant in five patients belonging to three unrelated families and in two sporadic cases.
Interestingly, despite a polymorphism frequency of 0.031, the homozygous R229Q variant has not been identified so far in control subjects. Although it has been shown in vitro to encode an altered protein with decreased binding affinity to nephrin20, it remains difficult to imagine that the R229Q substitution alone, even in the homozygous state, might be responsible for a severe disease. It is likely that additional sequence mutations in NPHS2 regulatory regions or introns, or variants in other genes, might explain the phenotype in these patients. It is, however, noteworthy that we found a higher frequency, though not significant, of the R229Q variant among nephrotic patients (5.13% vs. 3.75% in control patients), confirming previous findings of Tsukagushi et al20. Studies involving greater numbers of patients will be needed to determine the significance of this finding.
The severity of SRNS might also be modulated by the presence of additional mutations in genes encoding proteins interacting with podocin. NPHS2 mutations associated with NPHS1 variants have already been reported19,22, and we here describe a patient with very early onset SRNS in whom a heterozygous de novo splice mutation in NPHS1 and a homozygous NPHS2 R138Q mutation were detected. This is reminiscent of the triallelic inheritance found in some cases of Bardet-Biedl syndrome (BBS)33,34.
Contrary to previously published data by Bertelli et al21, and as would be expected with the resulting structural alteration of the glomerular filter resulting from the podocin gene defect, we have found that recurrent disease after renal transplantation is very unlikely in patients with two NPHS2 mutations. Out of 32 transplanted patients with two pathogenic NPHS2 mutations, we observed only one patient, bearing the R138X mutation in the homozygous state, with recurrent FSGS and a delayed onset. This rate (1/32) is far below the current rate of recurrence of 25% in FSGS1. Actually, Bertelli et al21 detected only two cases with recurrence of proteinuria after transplantation in patients with NPHS2-R138Q homozygous mutations, and in both cases, the patients had a mild episode with a favorable outcome. The mechanisms underlying these findings are not well defined. Pattraka et al recently described the presence of antiglomerular and antinephrin antibodies in the sera of transplanted patients with CNF who suffered from recurrent NS35. Although the development of antibodies directed against podocin is an attractive explanation for the recurrence of FSGS in our familial case of SRNS, our data and Bertelli's findings21 do not support this. However, in agreement with the findings of Bertelli et al, we also found that 3 out of 5 sporadic SRNS patients bearing heterozygous mutations who were transplanted developed recurrence21. In these patients, the theory of antipodocin antibodies seems rather unlikely because podocin expression will not be completely abolished. The development of SRNS when it recurs early after transplantation is commonly regarded as arising from one or more suggested circulating permeability factor(s) that alter renal permeability, leading to proteinuria2,3. As long as these circulating plasma factors remain unidentified, no clear conclusions can be drawn from genetic and clinical studies. Nevertheless, it can be suggested that NPHS2 mutations/variants may play a role in modulating the progression of renal disease before transplantation.
CONCLUSION
The present study shows that the phenotypic spectrum of patients affected by two clearly pathogenic NPHS2 mutations is restricted to severe early-onset SRNS and lack of recurrence of proteinuria after renal transplantation. This emphasizes the importance of NPHS2 mutation screening in children with early-onset SRNS to provide accurate genetic counseling and clinical prognostication. However, heterozygous NPHS2 mutations, sequence variants, and polymorphisms may play a role in atypical cases of SRNS with a later onset, mild clinical course, and recurring disease after renal transplantation. It can be expected that further genetic studies on families with NPHS2-negative SRNS will lead to the identification of new causative genes for SRNS.
References
| 1. | NIAUDET P. Steroid-resistant idiopathic nephrotic syndrome,. inPediatric Nephrology 1999; edited by Barrat TM, Avner ED, Harmon WE Baltimore, Lippincott Williams and Wilkins pp 749−763. |
| 2. | CARRARO M, CARIDI G & BRUSCHI M et al. Serum glomerular permeability activity in patients with podocin mutations (NPHS2) and steroid-resistant nephrotic syndrome. J Am Soc Nephrol 2002; 13: 1946−1952. | PubMed | |
| 3. | SAVIN VJ, SHARMA R & SHARMA M et al. Circulating factor associated with increased glomerular permeability to albumin in recurrent focal segmental glomerulosclerosis. N Engl J Med 1996; 334: 878−883. | Article | PubMed | ISI | ChemPort | |
| 4. | PELLETIER J, BRUENING W & LI FP et al. WT1 mutations contribute to abnormal genital system development and hereditary Wilms' tumour. Nature 1991; 353: 431−434. | Article | PubMed | ISI | ChemPort | |
| 5. | PELLETIER J, BRUENING W & KASHTAN CE et al. Germline mutations in the Wilms' tumor suppressor gene are associated with abnormal urogenital development in Denys-Drash syndrome. Cell 1991; 67: 437−447. | Article | PubMed | ISI | ChemPort | |
| 6. | BARBAUX S, NIAUDET P & GUBLER MC et al. Donor splice-site mutations in WT1 are responsible for Frasier syndrome. Nat Genet 1997; 17: 467−470. | Article | PubMed | ISI | ChemPort | |
| 7. | JEANPIERRE C, DENAMUR E & HENRY I et al. Identification of constitutional WT1 mutations, in patients with isolated diffuse mesangial sclerosis, and analysis of genotype/phenotype correlations by use of a computerized mutation database. Am J Hum Genet 1998; 62: 824−833. | PubMed | |
| 8. | KESTILA M, LENKKERI U & MANNIKKO M et al. Positionally cloned gene for a novel glomerular protein−nephrin−is mutated in congenital nephrotic syndrome. Mol Cell 1998; 1: 575−582. | Article | PubMed | ISI | ChemPort | |
| 9. | KAPLAN JM, KIM SH & NORTH KN et al. Mutations in ACTN4, encoding alpha-actinin-4, cause familial focal segmental glomerulosclerosis. Nat Genet 2000; 24: 251−256. | Article | PubMed | ISI | ChemPort | |
| 10. | KIM JM, WU H & GREEN G et al. CD2-associated protein haploinsufficiency is linked to glomerular disease susceptibility. Science 2003; 300: 1298−1300. | Article | PubMed | ISI | ChemPort | |
| 11. | BOUTE N, GRIBOUVAL O & ROSELLI S et al. NPHS2, encoding the glomerular protein podocin, is mutated in autosomal recessive steroid-resistant nephrotic syndrome. Nat Genet 2000; 24: 349−354. | Article | PubMed | ISI | ChemPort | |
| 12. | ROSELLI S, GRIBOUVAL O & BOUTE N et al. Podocin localizes in the kidney to the slit diaphragm area. Am J Pathol 2002; 160: 131−139. | PubMed | |
| 13. | SCHWARZ K, SIMONS M & REISER J et al. Podocin, a raft-associated component of the glomerular slit diaphragm, interacts with CD2AP and nephrin. J Clin Invest 2001; 108: 1621−1629. | Article | PubMed | ISI | ChemPort | |
| 14. | HUBER TB, HARTLEBEN B & KIM J et al. Nephrin and CD2AP associate with phosphoinositide 3−OH kinase and stimulate AKT-dependent signaling. Mol Cell Biol 2003; 23: 4917−4928. | Article | PubMed | ISI | ChemPort | |
| 15. | HUBER TB, KOTTGEN M & SCHILLING B et al. Interaction with podocin facilitates nephrin signaling. J Biol Chem 2001; 276: 41543−41546. | Article | PubMed | ISI | ChemPort | |
| 16. | HUBER TB, SIMONS M & HARTLEBEN B et al. Molecular basis of the functional podocin-nephrin complex: Mutations in the NPHS2 gene disrupt nephrin targeting to lipid raft microdomains. Hum Mol Genet 2003; 12: 3397−3405. | PubMed | |
| 17. | KARLE SM, UETZ B & RONNER V et al. Novel mutations in NPHS2 detected in both familial and sporadic steroid-resistant nephrotic syndrome. J Am Soc Nephrol 2002; 13: 388−393. | PubMed | |
| 18. | CARIDI G, BERTELLI R & CARREA A et al. Prevalence, genetics, and clinical features of patients carrying podocin mutations in steroid-resistant nonfamilial focal segmental glomerulosclerosis. J Am Soc Nephrol 2001; 12: 2742−2746. | PubMed | ISI | ChemPort | |
| 19. | CARIDI G, BERTELLI R & DI DUCA M et al. Broadening the spectrum of diseases related to podocin mutations. J Am Soc Nephrol 2003; 14: 1278−1286. | PubMed | |
| 20. | TSUKAGUCHI H, SUDHAKAR A & LE TC et al. NPHS2 mutations in late-onset focal segmental glomerulosclerosis: R229Q is a common disease-associated allele. J Clin Invest 2002; 110: 1659−1666. | Article | PubMed | ISI | ChemPort | |
| 21. | BERTELLI R, GINEVRI F & CARIDI G et al. Recurrence of focal segmental glomerulosclerosis after renal transplantation in patients with mutations of podocin. Am J Kidney Dis 2003; 41: 1314−1321. | PubMed | |
| 22. | KOZIELL A, GRECH V & HUSSAIN S et al. Genotype/phenotype correlations of NPHS1 and NPHS2 mutations in nephrotic syndrome advocate a functional inter-relationship in glomerular filtration. Hum Mol Genet 2002; 11: 379−388. | Article | PubMed | ISI | ChemPort | |
| 23. | FRISHBERG Y, RINAT C & MEGGED O et al. Mutations in NPHS2 encoding podocin are a prevalent cause of steroid-resistant nephrotic syndrome among Israeli-Arab children. J Am Soc Nephrol 2002; 13: 400−405. | PubMed | ISI | ChemPort | |
| 24. | CARIDI G, BERTELLI R & SCOLARI F et al. Podocin mutations in sporadic focal-segmental glomerulosclerosis occurring in adulthood. Kidney Int 2003; 64: 365. |
| 25. | FUCHSHUBER A, JEAN G & GRIBOUVAL O et al. Mapping a gene (SRN1) to chromosome 1q25-q31 in idiopathic nephrotic syndrome confirms a distinct entity of autosomal recessive nephrosis. Hum Mol Genet 1995; 4: 2155−2158. | PubMed | ISI | ChemPort | |
| 26. | LENKKERI U, MANNIKKO M & MCCREADY P et al. Structure of the gene for congenital nephrotic syndrome of the Finnish type (NPHS1) and characterization of mutations. Am J Hum Genet 1999; 64: 51−61. | Article | PubMed | ISI | ChemPort | |
| 27. | MAKRIDAKIS NM & REICHARDT JK. Multiplex automated primer extension analysis: Simultaneous genotyping of several polymorphisms. Biotechniques 2001; 31: 1374−1380. | PubMed | |
| 28. | WINN MP, CONLON PJ & LYNN KL et al. Linkage of a gene causing familial focal segmental glomerulosclerosis to chromosome 11 and further evidence of genetic heterogeneity. Genomics 1999; 58: 113−120. | Article | PubMed | ISI | ChemPort | |
| 29. | ROSELLI S, HEIDET L & SICH M et al. Early glomerular defect and severe renal disease in podocin-deficient mice. Mol Cell Biol 2004; 24: 550−560. | Article | PubMed | ChemPort | |
| 30. | SELLIN L, HUBER TB & GERKE P et al. NEPH1 defines a novel family of podocin interacting proteins. FASEB J 2003; 17: 115−117. | PubMed | ChemPort | |
| 31. | ROSELLI S, MOUTKINE I & GRIBOUVAL O et al. Plasma membrane targeting of podocin through the classical exocytic pathway: Effect of NPHS2 mutations. Traffic 2004; 5: 37−44. | PubMed | |
| 32. | EGAN ME, GLOCKNER-PAGEL J & AMBROSE C et al. Calcium-pump inhibitors induce functional surface expression of Delta F508-CFTR protein in cystic fibrosis epithelial cells. Nat Med 2002; 8: 485−492. | Article | PubMed | ISI | ChemPort | |
| 33. | BEALES PL, BADANO JL & ROSS AJ et al. Genetic interaction of BBS1 mutations with alleles at other BBS loci can result in non-Mendelian Bardet-Biedl syndrome. Am J Hum Genet 2003; 72: 1187−1199. | Article | PubMed | ISI | ChemPort | |
| 34. | KATSANIS N, ANSLEY SJ & BADANO JL et al. Triallelic inheritance in Bardet-Biedl syndrome, a Mendelian recessive disorder. Science 2001; 293: 2256−2259. | Article | PubMed | ISI | ChemPort | |
| 35. | PATRAKKA J, RUOTSALAINEN V & REPONEN P et al. Recurrence of nephrotic syndrome in kidney grafts of patients with congenital nephrotic syndrome of the Finnish type: Role of nephrin. Transplantation 2002; 73: 394−403. | PubMed | |
Acknowledgments
We thank the patients, the families, and the physicians who have contributed to this work. We are grateful to L. Heidet and M-C Gubler for helpful discussion and critical reading of the manuscript. This study was supported by grants from Inserm/Association Française contre les Myopathies/French Ministry of Research, the French National Program for Clinical Research (AOM 02 123), and the Association pour l'Utilisation du Rein Artificiel. S.W. was supported by the Deutsche Forschungsgemeinschaft (WE 2724–1), and E.L.E. is funded through an NIH F32-DK065409-01 grant.
MORE ARTICLES LIKE THIS
These links to content published by NPG are automatically generated
RESEARCH
Kidney International Original Article
Kidney International Original Article
Kidney International Original Article
Analysis of NPHS1, NPHS2, ACTN4, and WT1 in Japanese patients with congenital nephrotic syndrome
Kidney International Original Article
