Panel sequencing distinguishes monogenic forms of nephritis from nephrosis in children

Abstract

Background

Alport syndrome (AS) and atypical hemolytic–uremic syndrome (aHUS) are rare forms of chronic kidney disease (CKD) that can lead to a severe decline of renal function. Steroid-resistant nephrotic syndrome (SRNS) is more common than AS and aHUS and causes 10% of childhood-onset CKD. In recent years, multiple monogenic causes of AS, aHUS and SRNS have been identified, but their relative prevalence has yet to be studied together in a typical pediatric cohort of children with proteinuria and hematuria. We hypothesized that identification of causative mutations by whole exome sequencing (WES) in known monogenic nephritis and nephrosis genes would allow distinguishing nephritis from nephrosis in a typical pediatric group of patients with both proteinuria and hematuria at any level.

Methods

We therefore conducted an exon sequencing (WES) analysis for 11 AS, aHUS and thrombotic thrombocytopenic purpura-causing genes in an international cohort of 371 patients from 362 families presenting with both proteinuria and hematuria before age 25 years. In parallel, we conducted either WES or high-throughput exon sequencing for 23 SRNS-causing genes in all patients.

Results

We detected pathogenic mutations in 18 of the 34 genes analyzed, leading to a molecular diagnosis in 14.1% of families (51 of 362). Disease-causing mutations were detected in 3 AS-causing genes (4.7%), 3 aHUS-causing genes (1.4%) and 12 NS-causing genes (8.0%). We observed a much higher mutation detection rate for monogenic forms of CKD in consanguineous families (35.7% versus 10.1%).

Conclusions

We present the first estimate of relative frequency of inherited AS, aHUS and NS in a typical pediatric cohort with proteinuria and hematuria. Important therapeutic and preventative measures may result from mutational analysis in individuals with proteinuria and hematuria.

INTRODUCTION

Alport syndrome (AS) is a rare, progressive hereditary nephropathy that accounts for 1.6% of chronic kidney disease (CKD) manifesting before age 25 years [1, 2]. It is characterized by hematuria, proteinuria and extrarenal manifestations such as ocular and cochlear abnormalities. Patients with AS usually display severe decline of renal function, with 50% of males reaching end-stage renal disease (ESRD) by age 25 years and 15% of females reaching ESRD by age 40 years [3]. In contrast, thin basement membrane nephropathy (TBMN) is characterized by largely asymptomatic hematuria that is rarely associated with proteinuria and ESRD. Recessive mutations have been identified in the COL4A3 and COL4A4 genes as causative for AS and dominant mutations in the COL4A3 and COL4A4 genes have been identified as causative for TBMN [4–6]. Mutations in the COL4A5 gene cause X-linked AS [4–6]. To date, >1000 different mutations in COL4A3, COL4A4 and COL4A5 have been described. In patients with COL4A4 mutations, the ability to distinguish early which patients will exhibit severe renal disease and which patients will display benign hematuria can help reduce or delay the decline of renal function [7–10]. Previously, other groups have been able to identify disease-causing mutations in genes encoding for type IV collagen in 80% of patients with AS [11].

Atypical hemolytic–uremic syndrome (aHUS) is another rare nephropathy, characterized by microangiopathic hemolytic anemia, thrombocytopenia and acute kidney injury. It accounts for ∼2% of CKD cases that manifest before age 25 years [1, 2]. Mutations in nine genes have been identified as causative for aHUS [1, 2, 12–15]. The approval of eculizumab in 2011 has opened the door for new therapeutic approaches to the treatment of chronic aHUS. Thus, distinguishing between hereditary and nonhereditary forms of HUS has major implications for treatment approaches. In contrast, steroid-resistant nephrotic syndrome (SRNS) is a more common form of nephropathy that is characterized by proteinuria, hypoalbuminemia and edema. SRNS accounts for ∼10% of all CKD manifesting before age 25 years [1, 2, 16]. A monogenic cause of SRNS can be identified in ∼30% of cases that manifest before age 25 years [10, 16, 17].

In the setting of a pediatric nephrology clinic, presentation with a combination of proteinuria and hematuria poses a common diagnostic challenge. Molecular screening approaches for hereditary forms of CKD have been applied to well-defined disease cohorts, but few studies have tested for hereditary forms of CKD in such a typical pediatric patient population with both proteinuria and hematuria. We therefore hypothesized that identification of causative mutations by whole exome sequencing (WES) in known nephritis and nephrosis (NS) genes would allow distinguishing nephritis from nephrosis in a typical pediatric group of patients with proteinuria and hematuria. To examine the prevalence of hereditary forms of AS, aHUS and NS in a pediatric cohort of 371 CKD patients with proteinuria and hematuria manifesting before 25 years of age, we sequenced the coding regions of 11 AS-, aHUS- and thrombotic thrombocytopenic purpura (TTP)-causing genes and in parallel sequenced the coding regions of 23 common SRNS-causing genes. Causative mutations in AS-, aHUS- or NS-causing genes could be identified in 14.1% of individuals with childhood-onset proteinuria and hematuria and mutation analysis provides a safe approach for arriving at an etiologic diagnosis that can help distinguish nephritis from nephrosis in a pediatric population.

MATERIALS AND METHODS

Human subjects

This study was approved by the institutional review boards of Boston Children’s Hospital and the University of Michigan. DNA samples were collected from 2854 individuals between 2003 and 2014 after obtaining informed consent, clinical data and pedigree information (www.renalgenes.org). Inclusion criteria were defined by the clinical presentation of both any level of proteinuria and any level of hematuria. The majority of patients had nephrotic-range proteinuria as defined by >2.5 g of proteinuria per day or a urine protein:creatinine ratio >2 g/g of creatinine [18]. The subjects had an onset of proteinuria and hematuria before an age of 25 years. It has previously been reported that the overall prevalence of monogenic CKD is >20% in patients manifesting before age 20 years [1], with 29.5% of nephrotic syndrome cases caused by single gene mutations [16]. A separate, previous study molecularly solved 83% of patients with AS with an average age of molecular diagnosis of 26 years [11]. Based on these previous studies, there is good evidence to support the use of an age cutoff of 25 years, with a high likelihood of monogenic CKD etiology in patients presenting before age 25 years. Thus a total of 362 families (371 patients) who met the inclusion criteria were included in this study, which consisted of 193 male and 178 female subjects. There was a bias against inclusion of patients positive for WT1 mutations due to initial prescreening of patients with phenotypically described Denys–Drash syndrome, Frasier syndrome or Wilms tumor. Our patient cohort had partial overlap with a previously published cohort, as discussed below [16].

Mutation analysis

In order to screen patients for monogenic forms of AS, aHUS and SRNS, we took a two-pronged approach (Supplementary data, Figure S1). For monogenic forms of AS, aHUS and TTP, we screened all 362 families using barcoded multiplex polymerase chain reaction (PCR) and next-generation sequencing (NGS), even if they had previously undergone WES screening for mutations in SRNS-causing genes before the start of this study. This was done to ensure that we had thoroughly and uniformly screened every single patient for any mutations in any of the AS-, aHUS- and TTP-causing genes by the time of completion of this study. For monogenic forms of SRNS, some of our patients had been screened previously for pathogenic mutations in SRNS-causing genes in a previously published barcoded multiplex PCR and NGS study [16]. Of the 362 families, 315 had been previously screened for monogenic forms of SRNS using either WES or barcoded multiplex PCR and NGS, as alluded to above [16]. Thus there were 47 remaining families who had never been screened for monogenic forms of SRNS who were subsequently screened for monogenic forms of SRNS in this study using barcoded multiplex PCR and NGS. In summary, by the end of this study all 362 families were newly screened for mutations in AS-, aHUS- and TTP-causing genes and 47 families that had never been screened for mutations in NS-causing genes were newly screened for NS-causing mutations (Supplementary data, Figure S1).

High-throughput mutation analysis by array-based multiplex PCR and NGS

We designed 358 target-specific primer pairs for 300 coding exons and the adjacent splice sites of 11 genes that are known to cause AS, aHUS or TTP when mutated. The genes sequenced were ADAMTS13, C3, CD46, CFH, CFHR5, CFI, COL4A3, COL4A4, COL4A5, DGKE and THBD (Supplementary data, Table S1). For the 47 families who had not been screened previously for monogenic forms of SRNS, we used 524 target-specific primer pairs for 460 coding exons and the adjacent splice sites of 23 genes that are known to cause NS when mutated [16]. These 524 primer pairs were the same as those used for multiplex PCR and NGS in the patients previously screened for monogenic forms of NS [16]. The genes screened by multiplex PCR and NGS were ACTN4, ADCK4, ARHGAP24, ARHGDIA, CD2AP, COQ2, COQ6, CRB2, CUBN, INF2, ITGA3, ITGB4, LAMB2, LMX1B, MYO1E, NPHS1, NPHS2, PDSS2, PLCE1, PTPRO, SMARCAL1, TRPC6 and WT1 (Supplementary data, Table S2). In all multiplex PCRs, amplicon sizes ranged from 200 to 300 base pairs (primer sequences are available from the authors upon request). The use of barcoded multiplex PCR (48.48 Access Arrays system, Fluidigm, South San Francisco, CA, USA) allowed parallel amplification of all 358 amplicons in 362 families while screening AS, aHUS, and TTP genes and all 524 amplicons in the 47 families not previously screened for monogenic causes of SRNS. Subsequently the pooled barcoded PCR product libraries were sequenced on a MiSeq system (Illumina, San Diego, CA, USA) using the v2 chemistry. Sequence reads were aligned to the human reference sequence using CLC Genomics Workbench (CLC bio, Aarhus, Denmark) [19]. Prior to further evaluation, we excluded synonymous variants and variants that occur with a minor allele frequency >1% in the Short Genetic Variations database (dbSNP, version 138).

Homozygosity mapping

For genome-wide homozygosity mapping the GeneChip Human Mapping 250k StyI Array from Affymetrix (Santa Clara, CA, USA) was used. Nonparametric logarithm of odds scores were calculated using a modified version of the program GENEHUNTER 2.1 [20, 21] through stepwise use of a sliding window with sets of 110 single-nucleotide polymorphisms and the program ALLEGRO [22] in order to identify regions of homozygosity as described [23, 24] using a disease allele frequency of 0.0001 and Caucasian marker allele frequencies.

WES

WES and variant burden analysis were performed as described previously [25]. In brief, genomic DNA was isolated from blood lymphocytes and subjected to exome capture using SureSelect human exome capture arrays (Agilent Technologies, Santa Clara, CA, USA) followed by NGS on the HiSeq sequencing platform (Illumina) as previously described.

Mutation calling

Sequence reads were mapped against the human reference genome (National Center for Biotechnology Information build 37/hg19) using the CLC Genomics Workbench (version 6.5.1; CLC bio). Variants with minor allele frequencies <1% in the dbSNP (version 138) were selected and annotated for impact on the encoded protein and for conservation of the reference base and amino acid among orthologs across phylogeny. All patients were evaluated for mutations in genes known to cause AS or aHUS when mutated (Supplementary data, Table S1) and for genes known to cause SRNS when mutated (Supplementary data, Table S2). In all patients with any potentially pathogenic heterozygous variant in NPHS2, we further verified for the presence or absence of a second heterozygous c.686 G > A (p.R229Q) mutation [26]. This was done because the allele frequency of the NPHS2 p.R229Q mutation exceeds the 1% cutoff used in this study and thus would have been missed during our initial analysis unless we explicitly checked for it [26].

Validation of variants

Variants were validated as previously described [16]. Briefly, all variants previously reported as pathogenic in individuals with AS, aHUS, TTP or SRNS were considered as likely disease causing. Novel variants were ranked based on their likelihood to be deleterious for the function of the encoded protein. We considered protein truncation and obligatory splice site mutations as likely disease causing. For missense alleles, evolutionary conservation among orthologues and across phylogeny and bioinformatics prediction programs PolyPhen-2 [27], SIFT [28] and MutationTaster [29] were taken into consideration. All variants that were frequently present in the homozygous state for recessive genes (>1%) heterozygous state for dominant genes (>0.1%), or hemizygous state for X-linked dominant genes in healthy control cohorts [1000 Genomes Browser, Exome Aggregation Consortium (ExAC), Exome Variant Server (EVS) and Genome Aggregation Database (gnomAD)] were excluded unless previous studies demonstrated concrete loss of function or incomplete penetrance for the specific variant. Variants were confirmed in patient DNA using Sanger sequencing. Whenever parental DNA was available, segregation analysis was performed. Final calling of variant pathogenicity was performed by geneticists together with physician scientists who had knowledge of the clinical phenotypes and pedigree structure.

Coverage statistics for multiplex PCRs

While sequencing AS-, aHUS- and TTP-causing genes, we achieved a median sequencing coverage of 200× per individual and 400× per amplicon. Only 11 individuals (3%) and 27 amplicons (7.5%) had a median coverage <20×. While sequencing SRNS-causing genes in the 47 previously unscreened families in our cohort, we achieved a median sequencing coverage of 200× per individual and 200× per amplicon. No individuals (0%) and 43 amplicons (8.2%) had a median coverage <20×. Coverage statistics for our patients previously sequenced for mutations in SRNS-causing genes were previously reported [16].

Web Resources

• 1000 Genomes Browser, http://browser.1000genomes.org

• Biobase, https://portal.biobase-international.com/hgmd/pro/search_gene.php?

• Ensembl Genome Browser, http://www.ensembl.org

• Exome Aggregation Consortium, exac.broadinstitute.org

• Exome Variant Server, http://evs.gs.washington.edu/EVS/

• Genome Aggregation Database, http://gnomad.broadinstitute.org

• Human Gene Nomenclature Committee, http://www.genenames.org/

• MutationTaster, http://www.mutationtaster.org/ [29]

• Online Mendelian Inheritance in Man (OMIM), http://www.omim.org

• PolyPhen2, http://genetics.bwh.harvard.edu/pph2/ [27]

• Primer3, http://primer3.ut.ee/

• Sorting Intolerant from Tolerant (SIFT), http://sift.jcvi.org/ [28]

• UCSC Genome Browser, http://genome.ucsc.edu/cgi-bin/hgGateway

Software

• CLC Genomics Workbench, version 6.5.1 (CLC bio, Aarhus, Denmark)

• Alamut Visual, version 2.7, revision 1) (Interactive Biosoftware, Rouen, France)

RESULTS

Mutation detection

In a pediatric cohort of 371 patients (362 families) who had proteinuria and hematuria with an onset before 25 years of age, we examined for mutations in 11 genes that are known monogenic causes of AS (3 genes), aHUS (7 genes) or TTP (1 gene) if mutated (Supplementary data, Table S1) and for 23 genes that are known as monogenic causes of SRNS (Supplementary data, Table S2). Consanguinity was present in 56 of the 362 families screened (15.5%). We detected mutations in three of the three AS-causing genes and in three of the seven aHUS-causing genes (Table 1). We did not detect any mutations in the TTP-causing gene ADAMTS13. We detected causative mutations in 12 of the 23 SRNS-causing genes (Table 1). Mutations that likely explained the molecular etiology of disease were detected 51 of 362 unrelated families (14.1%) (Table 1).

Table 1.

Distribution of causative mutations detected in 18 of 34 sequenced genes for AS, aHUS, TTP and SRNS in 56 families presenting with proteinuria and hematuria before age 25 years

Gene symbol (n = 34)	Number of families with molecular genetic diagnosis (n = 51)	Percentage of total families (=100%)
Alport syndrome
COL4A5	10	2.76
COL4A3	6	1.66
COL4A4	1	0.28
aHUS
CFHR5	3	0.84
CFH	1	0.28
CFI	1	0.28
C3	0	0
CD46	0	0
DGKE	0	0
THBD	0	0
TTP
ADAMTS13	0	0
Nephrotic syndrome
NPHS1	5	1.38
NPHS2	5	1.38
LMX1B	4	1.10
PLCE1	4	1.10
LAMB2	3	0.82
SMARCAL1	2	0.56
ACTN4	1	0.28
ARHGDIA	1	0.28
COQ2	1	0.28
CUBN	1	0.28
INF2	1	0.28
TRPC6	1	0.28
ADCK4	0	0
ARHGAP24	0	0
CD2AP	0	0
COQ6	0	0
CRB2	0	0
ITGA3	0	0
ITGB4	0	0
MYO1E	0	0
PDSS2	0	0
PTPRO	0	0
WT1	0	0
Total	51	14.1

Genes with pathogenic variants

Variants were validated as previously described in the methods and in Sadowski et al. [16]. Mutations were detected in three AS-causing genes in 17 families: COL4A5 (10 families), COL4A3 (6 families) and COL4A4 (1 family) (Tables 2 and 3). Mutations were detected in three aHUS-causing genes in five families: CFHR5 (three families), CFH (one family) and CFI (one family) (Tables 2 and 3).

Table 2.

Pathogenic variants detected by multiplex PCR and NGS in 362 families (371 patients) with proteinuria and hematuria with an age of onset <25 in 11 genes that if mutated, cause AS, aHUS or TTP

Family	Exon (Zygosity)	Nucleotide change	Amino acid change	dbSNP (rs #)	Conserved to	Poly- phen2	SIFT	Mutation taster	Sex	Ethnicity (consan- guinity)	gnO-MAD general population	gnO-MAD closest ethnicity	Bio- base
COL4A3
A2041	4 (Het)	c.272G>A	p.Gly91Asp	−	Dr	0.994	Del	DC	m	Euro (N)	1/30960 (0 hom)	0/14998 (0 hom)	[30]
	48 (Het)	c.4421T>C	p.Leu1474Pro	200302125	Dr	1	Del	DC			735/276998 (0 hom)	620/126548 (0 hom)	[31, 32]
A1916	4 (Het)	c.279 + 6T>C	Splice	770953670	NA	−71%	−15.4%	−2.4%	f	Slavic (N)	1/245742 (0 hom)	0/30758 (0 hom)	Novel
A2490	23 (Het)	c.1504 + 6A>C	Splice	760718271	NA	+2.8%	+0.9%	+0.2%	f	Kazakh (N)	1/246078 (0 hom)	1/111616 (0 hom)	Novel
A1479	27 (Het)	c.1978C>A	p.Pro660Thr	773674552	Gg	0.661	Del	DC	f	Turk (Y)	5/277150 (0 hom)	5/126688 (0 hom)	Novel
A2358	29 (Hom)	c.2162del	p.Gly721Val*26	−	NA	NA	NA	NA	m	Asian (Y)	3/240354 (0 hom)	2/29504 (0 hom)	Novel
A2609	49 (Het)	c.4487G>A	p.Arg1496Gln	776086781	Gg	0.989	Del	DC	m	Euro (N)	11/276940 (0 hom)	1/126550 (0 hom)	Novel
COL4A4
B789	2 (Hom)	c.71 + 1G>A	Splice	−	NA	−100%	−100%	−100%	f	Cauc (N)	−	−	[33]
COL4A5
A5192	3 (Hem)	c.231 + 3A>G	Splice	376366035	NA	−32.2%	−0.4%	−11.1%	m	Turk (Y)	2/178120 (2 hem)	2/79762 (2 hem)	Novel
A965	4 (Hem)	c.274_279del	p.Arg92_Gly93del	−	Dm	NA	NA	NA	m	Euro (Y)	−	−	Novel
A3676	6 (Hem)	c.322-1G>A	Splice	−	NA	−100%	−100%	−100%	m	Indian (Y)	−	−	[34]
A2917	20 (Hem)	c.1217del	p.Gly406Val*68	−	NA	NA	NA	NA	m	Slavic (N)	−	−	Novel
A3933	24 (Hem)	c.1634G>A	p.Gly545Asp	−	Dm	1	Del	DC	m	Indian (N)	−	−	Novel
	24 (Hem)	c.1634G>A	p.Gly545Asp	−	Dm	1	Del	DC	m	Indian (N)	−	−	Novel
B711	25 (Hem)	c.1781G>A	p.Gly594Asp	−	Dm	0.355	Del	DC	m	Euro (N)	−	−	[35]
A1963	25 (Hem)	c.1931G>A	p.Gly644Asp	−	Dm	1	Del	DC	m	Hisp (Y)	−	−	[11, 36]
B28	46 (Hem)	c.4063del	p.Glu1355Asn*22	−	NA	NA	NA	NA	m	Arabic (Y)	−	−	Novel
A4926	48 (Hem)	c.4309C>G	p.Gln1437Glu	143778018	Dm	0.407	Del	DC	m	Filipino (N)	19/197664 (3 hom)	2/4713 (0 hom)	Novel
A169	49 (Hem)	c.4439del	p.Pro1480His*74	−	NA	NA	NA	NA	m	Turk (Y)	−	–	Novel
	49 (Hem)	c.4439del	p.Pro1480His*74	−	NA	NA	NA	NA	m	Turk (Y)	−	–	Novel
CFH
A4035	10 (Het)	c.1507C>G	p.Pro503Ala	570523689	Ci	0.746	Del	PMP	f	Euro (N)	4/245456 (0 hom)	4/111226 (0 hom)	[37]a
CFHR5
A4967	2 (Het)	c.232T>C	p.Ser78Pro	146025130	NA	0.986	Tol	PMP	m	African (N)	32/277182 (0 hom)	27/24038 (0 hom)	Novel
A2351	4 (Het)	c.486dup	p.Glu163Arg*35	565457964	NA	NA	NA	NA	f	Cauc (N)	564/276160 (0 hom)	395/126014 (0 hom)	[38, 39]
A3422	10 (Het)	c.1615T>G	p.Phe539Val	111989094	NA	0.998	Del	DC	m	Arabic (?)	5/277086 (0 hom)	0/126608 (0 hom)	Novel
CFI
A2336	13 (Het)	c.1558 + 5G>T	Splice	1114013791	NA	−21.2%	−11.9%	−12.3%	f	Asian (N)	2400/276954 (19 hom)	0/18862 (0 hom)	[40, 41]

The genes sequenced were ADAMTS13, C3, CD46, CFH, CFHR5, CFI, COL4A3, COL4A4, COL4A5, DGKE and THBD.

^aVariant reported in Biobase for nonrenal disease. The splice site prediction scores were derived left to right from the MaxEnt, NNSPLICE and HSF prediction programs, respectively.

Cauc, Caucasian; Ci, Ciona intestinalis; DC, disease causing; Del, deleterious; Dm, Drosophila melanogaster; Dr, Danio rerio; dup, duplication; Euro, European; F, female; Gg, Gallus gallus; Hem, hemizygous; Het, heterozygous; Hisp, Hispanic; Hom, homozygous; M, male; N, no; NA, not applicable; PMP, polymorphism; Tol, tolerated; Turk, Turkish; Y, yes.

Table 3.

Phenotypes of patients with pathogenic variants in 1 of 11 genes that if mutated cause AS, aHUS or TTP after multiplex PCR and NGS in 362 families (371 patients) with proteinuria and hematuria with an age of onset <25 years

Family	Sex	Age of onset (ESRD)	Proteinuria	Hematuria	Biopsy	Family history	Ethnicity (consanguinity)	Extrarenal treatment (response)	Biobase
COL4A3
A2041	M	18 months	4 +(dipstick)	Microscopic	FSGS	None	Euro (N)	Azospermia, ACEi(UR)	[30]
									[31, 32]
A1916	F	6 years	6 g/day	Microscopic	MPGN type 1	None	Slavic (N)	HTN, HepB, SST, BAS, SS(PR), CS(NR), MMF(UR)	Novel
A2490	F	9 years	1 g/day	Microscopic	ND	None	Kazakh (N)	HTN, SS(CR)	Novel
A1479	F	4 years	1.08 g/day	Microscopic	ND	Mother	Turk (Y)	SS(CR)	Novel
A2358	M	10 years	4+	Microscopic	FSGS	2 siblings (deceased)	Asian (Y)	SR	Novel
A2609	M	10 years	UPC 4 mg/mg	Microscopic	ND	None	Euro (N)	SS(CR)	Novel
COL4A4
B789	F	4 years	UPC 2.57 mg/mg	Microscopic	FSGS	None	Cauc (N)	−	[33]
COL4A5
A5192	M	3 years (3 years)	5 g/day	Microscopic	FSGS, TubAt	None	Turk (Y)	SR, CP(NR)	Novel
A965	M	15 years	1.88 g/day	Macroscopic	Alport, BM lamellation	Mother	Euro (Y)	SR, CS(CR)	Novel
A3676	M	5 years	15.5 g/day	Y, type N/A	ND	None	Indian (Y)	Syncope, SR	[34]
A2917	M	2 years	2 g/day	Macroscopic	Non-specified	Mother	Slavic (N)	−	Novel
A3933	M	6 years	9.5 g/day	Microscopic	MCNS, MesP, TBM	Brother, paternal cousin, paternal grandma	Indian (N)	−	Novel
B711	M	15 years	2 g/day	Macroscopic	Alport, FSGS	Mother, father	Euro (N)	SR	[35]
A1963	M	14 years	7 g/day	Y, type N/A	ND	Uncle	Hisp (Y)	SR, CP(PR)	[11, 36]
B28	M	1 year	3 g/day	Microscopic	ND	None	Arabic (Y)	−	Novel
A4926	M	1 year	2 g/day	Microscopic	ND	None	Filipino (N)	Cough, Pnm, FR, capto, cerf/erythro, albmn trans	Novel
A169-21	M	7 years	UPC 1.37 mg/mg	Microscopic	Crescentic GN	Brother	Turk (Y)	SR, CP(NR)	Novel
A169-22	M	1 year	UPC 0.1 mg/mg	Microscopic	Crescentic GN	Brother	Turk (Y)	SR, TX	Novel
CFH
A4035	F	13 years	UPC 2.57 mg/mg	Y, type N/A	FSGS, TMA	None	Euro (N)	HTN, SS(PR)	[37]a
CFHR5
A4967	M	7 years (7 years)	3 g/day	Microscopic	Active SLN	None	African (N)	HTN, SS(PR), ESRD	Novel
A2351	F	14 years	4 g/day	Microscopic	Diffuse MesP, TBM	None	Cauc (N)	HTN, SR, CP(UR)	[38, 39]
A3422	M	4 years	5 g/day	Microscopic	MPGN	None	Arabic (N)	−	Novel
CFI
A2336	F	9 years	3.3 g/day	Microscopic	ND	Sister, father	Asian (N)	−	[40, 41]

The genes sequenced were ADAMTS13, C3, CD46, CFH, CFHR5, CFI, COL4A3, COL4A4, COL4A5, DGKE and THBD.

^aVariant reported in Biobase for nonrenal disease.

ACEi, angiotensin-converting enzyme inhibitor; Albmn trans, albumin transfusions; BAS, bronchial asthma; BM, basement membrane; Capto, captopril; Cauc, Caucasian; Cerf/erythro, cefuroxime and erythromycin; CP, cyclophosphamide; CR, complete response; CS, cyclosporine; Euro, European; F, female; FSGS, focal segmental glomerulosclerosis; GN, glomerulonephritis; HepB, hepatitis B; Hisp, Hispanic; HTN, hypertension; M, male; MCNS, minimal change nephrotic syndrome; MesP, mesangial proliferation; MMF, mycophenolate mofetil; MPGN, membrane proliferative glomerulonephritis; N, no; ND, not done; NR, no response; Pnm, pneumonia; PR, partial response; SLN, sclerosing lobular nephritis; SR, steroid resistant; SS, steroid sensitive; SST, short stature; TBM, thin basement membrane; TMA, thrombotic microangiopathy; TubAt, tubular atrophy; Turk, Turkish; TX, transplant; type N/A, hematuria type unknown; UPC, urine protein:creatinine ratio; UR, unknown response; Y, yes.

In addition, mutations were detected in 12 SRNS-causing genes in 29 families: NPHS1 (5 families), NPHS2 (5 families), LMX1B (4 families), PLCE1 (4 families), LAMB2 (3 families), SMARCAL1 (2 families), ACTN4 (1 family), ARHGDIA (1 family), COQ2 (1 family), CUBN (1 family), INF2 (1 family) and TRPC6 (1 family) (Tables 4 and 5). No pathogenic variants were found in the following 16 genes: ADAMTS13, ADCK4, ARHGAP24, C3, CD2AP, CD46, COQ6, CRB2, DGKE, ITGA3, ITGB4, MYO1E, PDSS2, PTPRO, THBD and WT1. Of the 55 different disease-causing mutations detected in this study, 19 (34.5%) were novel variants that had never previously been reported in databases containing human disease-causing mutations.

Table 4.

Pathogenic variants detected in 362 families (371 patients) with proteinuria and hematuria with an age of onset <25 years in 23 genes that cause nephrotic syndrome if mutated

Family	Exon (Zygo- sity)	Nucleotide change	Amino acid change	dbSNP (rs #)	Conserved to	Poly- phen 2	SIFT	Muta- tion taster	Sex	Ethnicity (consan- guinity)	gnOMAD general population	gnO-MAD closest ethnicity	Bio- base	Method
ACTN4
A1055	14 (Het)	c. 1606C>A	P.Arg536Ser	−	Dr	0.933	Del	DC	M	Kurd (Y)	−	−	Novel	EXM
ARHGDIA
A1432	5 (Hom)	c.518G>T	p.Gly173Val	−	Sc	1	Del	DC	F	Jewish (Y)	1/246016 (0 hom)	1/9840 (0 hom)	[42]	PCR
	5 (Hom)	c.518G>T	p.Gly173Val	−	Sc	1	Del	DC	M	Jewish (Y)	1/246016 (0 hom)	1/9840 (0 hom)	[42]	PCR
COQ2
A103	3 (Het)	c.683A>G	p.Asn228Ser	121918232	Ce	0.918	Tol	DC	F	Euro (N)	32/276228 (0 hom)	0/111262 (0 hom)	[16, 43]	PCR
	5 (Het)	c.856C>T	p.Leu286Phe	776124921	Dm	0.997	Del	DC			2/245656 (0 hom)	2/111470 (0 hom)	[16]	PCR
CUBN
A1213	19 (Hom)	c. 2613_2614del	p.Asp872Leu*3	386833777	NA	NA	NA	NA	M	Balkan (N)	6/276980 (0 hom)	0/126554 (0 hom)	[44]	EXM
	19 (Hom)	c. 2613_2614del	p.Asp872Leu*3	386833777	NA	NA	NA	NA	M	Balkan (N)	6/276980 (0 hom)	0/126554 (0 hom)	[44]	EXM
	19 (Hom)	c. 2613_2614del	p.Asp872Leu*3	386833777	NA	NA	NA	NA	M	Balkan (N)	6/276980 (0 hom)	0/126554 (0 hom)	[44]	EXM
INF2
A675	2 (Het)	c.37G>A	p.Ala13Thr	201383094	Dr	0.982	Tol	DC	F	Euro (N)	85/240928 (1 hom)	40/106632 (0 hom)	[45]a	PCR
LAMB2
A1757	2 (Hom)	c.143A>C	p.Tyr48Ser	776905329	Dr	1	Del	DC	M	Hisp (N)	9/239510 (0 hom)	0/33384 (0 hom)	Novel	EXM
	2 (Hom)	c.143A>C	p.Tyr48Ser	776905329	Dr	1	Del	DC	F	Hisp (N)	9/239510 (0 hom)	0/33384 (0 hom)	Novel	EXM
A2356	7 (Hom)	c.736C>T	p.Arg246Trp	121912488	Ce	1	Del	DC	M	Asian (Y)	3/240414 (0 hom)	0/17184 (0 hom)	[16, 46]	PCR
A1613	10 (Hom)	c.1405 + 1G>A	Splice	780041521	NA	−100%	0%	−100%	M	Euro (N)	4/244056 (0 hom)	1/109858 (0 hom)	[16, 47]	PCR
LMX1B
A200	4 (Het)	c.737G>A	p.Arg246Gln	−	Ce	1	Del	DC	F	Turk (Y)	−	−	[16, 48]	PCR
A2175	4 (Het)	c.737G>A	p.Arg246Gln	−	–	1	Del	DC	M	Euro (N)	−	−	[16, 48]	PCR
A3180	4 (Het)	c.737G>A	p.Arg246Gln	−	Ce	1	Del	DC	F	Euro (N)	−	−	[16, 48]	PCR
A4009	7 (Het)	c.929C>G	p.Thr310Arg	−	Gg	0.701	Del	DC	F	Arabic (Y)	−	−	Novela	PCR
NPHS1
A1803	2 (Hom)	c.139delG	p.Ala47Pro*81	386833882	NA	NA	NA	NA	M	Cauc (N)	2/241678 (0 hom)	1/107982 (0 hom)	Novela	PCR
A3775	7 (Het)	c.840 + 1G>T	Splice	−	NA	−100%	−100%	−100%	F	Indian (N)	1/245036 (0 hom)	1/30778 (0 hom)	[49]a	PCR
	21 (Het)	c.2903G>T	p.Gly968Val	771798618	Ce	1	Del	DC			1/236378 (0 hom)	0/30014 (0 hom)	Novela	PCR
A3380	8 (Het)	c.928G>A	p.Asp310Asn	763972372	Dm	0.99	Del	DC	M	Asian (N)	3/241700 (0 hom)	3/17218 (0 hom)	[50, 51]	PCR
	21 (Het)	c.2816-3T>G	Splice	−	NA	−91.9%	−77.2%	−2.9%			−	−	[16, 51]	PCR
B115	12 (Hom)	c.1555C>T	p.Pro519Ser	−	Dr	0.984	Tol	PMP	F	Cauc (N)	−	−	[52]	EXM
A1500	20 (Hom)	c.2728T>C	p.Ser910Pro	−	Dr	0.959	Del	DC	F	Afr-Am (N)	−	−	[53, 54]	PCR
NPHS2
A4681	1 (Hom)	c.1A>T	p.Met1*	−	NA	NA	NA	NA	F	Arabic (Y)	−	−	[16]	EXM
A4624	4 (Hom)	c.467dup	p.Leu156Phe*11	−	NA	NA	NA	NA	F	Arabic (Y)	−	−	[17, 51, 55, 56]	PCR
B188	7 (Hom)	c.855-856del	p.Arg286Thr*17	749740335	NA	NA	NA	NA	F	Hisp (Y)	18/275798 (0 hom)	0/34342 (0 hom)	[57, 58]a	PCR
A1616	7 (Hom)	c.868G>A	p.Val290Met	200482683	Dr	0.998	Del	DC	F	Cauc (N)	33/276038 (0 hom)	32/126036 (0 hom)	[59–61]	PCR
A2239	8 (Hom)	c.926C>T	p.Ala309Val	−	Ce	0.742	Del	DC	M	Turk (N)	−	−	[62]a	PCR
PLCE1
A3233	10 (Hom)	c.3169C>T	p.Arg1057*	−	NA	NA	NA	NA	F	Arabic (Y)	1/245964 (0 hom)	0/111494 (0 hom)	[16]	PCR
A3617	11 (Hom)	c.3379_3380del	p.Asn1127*	−	NA	NA	NA	NA	F	Arabic (Y)	−	−	[16]	PCR
A3510	20 (Hom)	c.4600A>G	p.Lys1534Glu	−	Sc	0.998	Del	DC	F	Turk (Y)	−	−	[16]	PCR
A3869	26 (Hom)	c.5521A>G	p.Lys1841Glu	−	Sc	1	Del	DC	M	Arabic (Y)	−	−	[16]	PCR
SMARCAL1
A3146	4 (Het)	c.49C>T	p.Arg17*	119473034	NA	NA	NA	NA	F	Euro (N)	2/246210 (0 hom)	1/111680 (0 hom)	[63]a	PCR
	5 (Het)	c.836T>C	p.Phe279Ser	775057827	Ci	0.985	Tol	DC			28/277230 (0 hom)	23/126712 (0 hom)	[64]a	PCR
A4162	12 (Hom)	c.1736C>A	p.Ser579*	−	NA	NA	NA	NA	F	Euro (N)	−	−	[16]	PCR
TRPC6
A4685	2 (Het)	c.523C>T	p.Arg175Trp	−	Dr	1	Del	DC	F	Arabic (N)	−	−	[17]	EXM

The genes sequenced were ACTN4, ADCK4, ARHGAP24, ARHGDIA, CD2AP, COQ2, COQ6, CUBN, INF2, ITGA3, ITGB4, LAMB2, LMX1B, MYO1E, NPHS1, NPHS2, PDSS2, PLCE1, PTPRO, SMARCAL1, TRPC6 and WT1.

^aSolved in this study using multiplex PCR. The splice site prediction scores were derived left to right from the MaxEnt, NNSPLICE and HSF prediction programs, respectively.

Afr-Am, African American; Cauc, Caucasian; Ce, Caenorhabditis elegans; Ci, Ciona intestinalis; DC, disease causing; Dr, Danio rerio; Del, deleterious; Dm, Drosophila melanogaster; dup, duplication; Euro, European; F, female; EXM, homozygosity mapping and whole exome sequencing; Gg, Gallus gallus; Het, heterozygous; Hisp, hispanic; Hom, Homozygous; M, male; N, no; NA, not applicable; PCR, Fluidigm multiplex PCR + NGS; PMP, polymorphism; Sc, Saccharomyces cerevisiae; Tol, tolerated; Turk, Turkish; Y, yes.

Table 5.

Phenotypes of patients with pathogenic variants in 1 of 23 genes that if mutated cause nephrotic syndrome after multiplex PCR and NGS in 362 families (371 patients) with proteinuria and hematuria with an age of onset <25 years

Family	Sex	Age of onset (ESRD)	Proteinuria	Hematuria	Biopsy	Family history	Ethnicity (consanguinity)	Extrarenal treatment (response)	Biobase
ACTN4
A1055	M	10 years	UPC 4 mg/mg	Y, type N/A	FSGS	None	Kurd (Y)	SS(CR), CS(CR)	Novel
ARHGDIA
A1432	F	2 years (3 years)	ACR 0.85 mg/mg	Y, type N/A	DMS	None	Jewish (Y)	DF, SST, ID, CS(NR), TX	[42]
	M	1 year (1 year)	17.3 g/day	Y, type N/A	ND	None	Jewish (Y)	SST, TX	[42]
COQ2
A103	F	1 year	4 g/day	Y, type N/A	FSGS	None	Euro (N)	SR, CS(PR)	[16, 43]
CUBN
A1213	M	12	ACR 0.28 mg/mg	Microscopic	ND	None	Balkan (N)	Hypocalcemia	[44]
	M	5 years	ACR 0.26 mg/mg	Microscopic	ND	None	Balkan (N)	−	[44]
	M	<1× month	ACR 0.9 mg/mg	Microscopic	ND	None	Balkan (N)	−	[44]
INF2
A675	F	16 years	UPC 7.5 mg/mg	Y, type N/A	FSGS	None	Euro (N)	Steroids(UR)	[45]a
LAMB2
A1757	M	13 years	2.8 g/day	Y, type N/A	FSGS	Aunt	Hisp (N)	HTN, SR, Cellcept(NR)	Novel
	F	13 years	Y	Y, type N/A	FSGS	Aunt	Hisp (N)	−	Novel
A2356	M	1 year	4+	Microscopic	Finnish type	2 Brothers	Asian (Y)	Blindness	[16, 46]
A1613	M	<1 month	9.8 g/day	Y, type N/A	Dilated tubules, Microcysts	None	Euro (N)	FTT	[16, 47]
LMX1B
A200	F	8 years (9 years)	1.5 g/day	Y, type N/A	FSGS	Mother	Turk (Y)	SR, HemDi	[16, 48]
A2175	M	4 years (43 years)	3.1 g/L	Y, N/A	FSGS	None	Euro (N)	Factor XII Deficiency, SR, CS(PR), Acei(UR), HemDi, TX	[16, 48]
A3180	F	18 years (43 years)	UPC 2 mg/mg	Y, type N/A	FSGS	None	Euro (N)	HemDi, TX	[16, 48]
A4009	F	<1 month	UPC 4 mg/mg	Y, type N/A	MCNS	None	Arabic (Y)	SR, CS(NR)	Novela
NPHS1
A1803	M	<1 month	UPC 38 mg/mg	Microscopic	MCNS	None	Cauc (N)	CS(NR)	Novela
A3775	F	1 year	0.95 g/day	Y, type N/A	Diffuse MesP	None	Indian (N)	SS(CR)	[49]a
A3380	M	<1 month	Y	Y, type N/A	ND	None	Asian (N)	−	[50, 51]
B115	F	<1 month	50 mg/mg	Y, type N/A	ND	None	Cauc (N)	Hypothyroidism	[52]
A1500	F	1 year	UPC 20 mg/mg	Y, type N/A	MCNS	None	Afr-Am (N)	HTN	[53, 54]
NPHS2
A4681	F	7 years	UPC 8 mg/mg	Y, type N/A	FSGS	None	Arabic (Y)	SR	[16]
A4624	F	1 year	UPC 3.9 mg/mg	Y, type N/A	MPGN	None	Arabic (Y)	HTN, SR, CP(NR), CS(NR)	[17, 51, 55, 56]
B188	F	3 years (15 years)	3+	Macroscopic	MCNS	None	Hisp (Y)	TX	[57, 58]a
A1616	F	5 months	2.2 g/day	Y, type N/A	ND	None	Cauc (N)	Presentation post infection	[59–61]
A2239	M	<18	5.1 g/day	Microscopic	FSGS	Cousin	Turk (N)	−	[62]a
PLCE1
A3233	F	2	UPC 8.4 mg/mg	Y, type N/A	ND	None	Arabic (Y)	HTN, SR, CP(NR)	[16]
A3617	F	7 months	UPC 12 mg/mg	Y, type N/A	FSGS	Cousin, aunt	Arabic (Y)	−	[16]
A3510	F	1 year	UPC 2.7 mg/mg	Microscopic	ND	None	Turk (Y)	HTN	[16]
A3869	M	7 months	UPC 4.5 mg/mg	Y, type N/A	FSGS	None	Arabic (Y)	−	[16]
SMARCAL1
A3146	F	9 years	5 g/day	Y, type N/A	FSGS	None	Euro (N)	HTN, ID	[63]a
A4162	F	4 years	UPC 7.5 mg/mg	Y, type N/A	FSGS	None	Euro (N)	Celiac, VSD, PDA, SR, CS(UR)	[16]
TRPC6
A4685	F	17 years	UPC 9.8 mg/mg	Y, type N/A	FSGS	None	Arabic (N)	HTN, Allergies, CS(NR)	[17]

The genes sequenced were ACTN4, ADCK4, ARHGAP24, ARHGDIA, CD2AP, COQ2, COQ6, CUBN, INF2, ITGA3, ITGB4, LAMB2, LMX1B, MYO1E, NPHS1, NPHS2, PDSS2, PLCE1, PTPRO, SMARCAL1, TRPC6 and WT1.

^aSolved in this study using multiplex PCR.

ACEi, angiotensin-converting enzyme inhibitor; ACR, albumin:creatinine ratio; Afr-Am, African American; Cauc, Caucasian; CP, cyclophosphamide; CR, complete response; CS, cyclosporine; DF, deafness; DMS, diffuse mesangial sclerosis; Euro, European; F, female; FSGS, focal segmental glomerulosclerosis; FTT, failure to thrive; HemDi, hemodialysis; Hisp, Hispanic; HTN, hypertension; ID, intellectual disability; M, male; MCNS, minimal change nephrotic syndrome; MesP, mesangial proliferation; MPGN, membrane proliferative glomerulonephritis; N, no; ND, not done; NR, no response; PDA, persistent ductus arteriosus; PR, partial response; SR, steroid resistant; SS, steroid sensitive; SST, short stature; Turkish; TX, transplant; type N/A, hematuria type unknown; UPC, urine protein:creatinine ratio; UR, unknown response; VSD, ventricular septal defect; Y, yes.

DISCUSSION

Rate of mutation detection in AS and aHUS genes versus previous studies

Here we screened 362 families of 371 patients presenting with proteinuria and hematuria before age 25 years. We sequenced the coding regions of 34 genes known to cause monogenic AS, aHUS or SRNS and identified a causative mutation in 51 of 362 families (14.1%). AS, aHUS and SRNS often progress to ESRD, placing a significant health burden on patients and their families and adversely impacting quality of life [2]. Multiple monogenic causes of AS, aHUS and SRNS have been identified [4–6, 12, 13, 15, 43, 65, 44–46, 57, 58, 63, 66–89]. The frequency of single-gene mutations in typical, pediatric patients with proteinuria and hematuria has not yet been studied systematically in these genes. Our detection rate of AS-causing genes in 17 of 362 families (4.7%) was much lower than the 80% previously reported in a cohort of 101 patients with suspected or diagnosed AS [11]. The difference in mutation detection rates between our study and the previous study is likely due to the preselection of a cohort highly enriched for AS phenotypes in previous studies. Our mutation detection rate for AS-causing genes of 4.7% was reflective of broad inclusion criteria of proteinuria and hematuria instead of a distinct AS phenotype.

Mutation detection rate in SRNS genes versus previous studies

In 29 (8.0%) of 362 families, we identified pathogenic mutations in SRNS-causing genes. This percentage is much lower than previously reported mutation detection rates of 29.5% [16], 28.3% [17] and 31.7% [90] in three pediatric NS cohorts. This discrepancy is partially due to our exclusion of patients with Denys–Drash syndrome, Frasier syndrome or Wilms tumor phenotypes during patient cohort selection prior to beginning the study. Yet in the three previous cohort studies, mutations in WT1 only explained 4.8%, 5.8% and 2.6% of solved NS cases, respectively [16, 17, 90]. In addition, our cohort differed from previous NS cohorts in two major ways: first by the additional required inclusion criteria of the presence of hematuria, and second by the fact that the previous NS cohort studies only screened patients with SRNS, while our cohort included both steroid-sensitive and steroid-resistant NS patients. In addition, one of the previous studies actively excluded patients diagnosed with AS from their cohort [17]. Despite our broad inclusion criteria of proteinuria plus hematuria, we were still able to molecularly solve 14.1% of patients for monogenic forms of AS, aHUS or SRNS, further highlighting the power of using genetic screening to obtain a molecular etiology of disease.

Two of the previous NS cohort studies excluded consanguineous cases from their cohorts [17, 90]. The third previous SRNS cohort study had an overall consanguinity rate of 20.9% and solved 59.5% of consanguineous cases and 25% of nonconsanguineous cases [16]. In our study with an overall consanguinity rate of 15.5%, we detected a causative mutation in 20 of the 56 (35.7%) consanguineous families and we detected a causative mutation in 31 of the 306 (10.1%) nonconsanguineous families. This approximately matches previous reported rates of identification of causative monogenic variants in SRNS genes [16].

Study limitations

In this study we did not sequence CFB, CFHR1 and CFHR3, which have previously been described to cause aHUS when mutated [91, 92]. Currently 50 genes have been identified to cause SRNS when mutated [93]. Lastly, only patients who underwent WES were sequenced for all monogenic forms of NS. The majority of patients were sequenced for mutations in NS-causing genes using multiplex PCR limited to 23 genes. Therefore only 23 SRNS-causing genes were systematically screened in all patients. This contributed to the low rate at which a causative mutation was identified (Supplementary data, Tables S1 and S2).

Future directions

Although we used a minor allele frequency cutoff of 0.1% for dominant alleles, we detected heterozygous, pathogenic mutations in two patients in which the minor allele frequency was >0.1%. In patient A2336, we detected a heterozygous CFI mutation with a minor allele frequency of 0.9% in the general population. This exact allele, however, had been published twice [40, 41]. The first study identified this allele in 1 of 45 patients screened for mutations in aHUS-causing genes [40]. The second report of this allele was in a patient with sporadic aHUS not secondary to human immunodeficiency virus (HIV) or shiga toxin etiology [41]. In both reports, no additional phenotypic details were reported, but given the convergence of evidence from multiple research groups, we believe that the c.1558 + 5 G > T splice site mutation may cause aHUS.

In patient A2351, we detected a heterozygous mutation in CFHR5 with a minor allele frequency of 0.2% in the general population. As was the case with the CFI allele in patient A2336, the CFHR5 allele in patient A2351 was also published twice previously [38, 39]. It was first reported in a 7-year-old presenting with proteinuria, hematuria and loss of corticomedullary differentiation on ultrasound [39]. A biopsy 9 months later showed mesangial hypercellularity, CFHR5 deposits and a thickened glomerular basement membrane [39]. Like our patient, this patient carried the same CFHR5 allele heterozygously, but the healthy mother and sister were also heterozygous carriers, suggesting incomplete penetrance for this allele [39]. The patient had significantly reduced serum CFHR5 when compared with healthy controls [39]. In the second report of this allele, an 11-year-old male presented with gross hematuria, edema, proteinuria and hypoalbuminemia [38]. The authors reported dense deposits on the glomerular basement membrane and Bowman’s capsule and ophthalmic drusen [38]. This second patient also carried the CFHR5 allele heterozygously [38]. Given the above evidence, we believe the CFHR5 allele in patient A2351 can cause aHUS. We detected AS-causing mutations in some patients without a documented familial history of renal disease or copresentation with ocular abnormalities. Future investigations should focus on early detection of ocular abnormalities with more careful follow-up by health care providers in these patients.

CONCLUSIONS

To our knowledge, this is the first study undertaken to determine the prevalence of mutations in AS-, aHUS- and NS-causing genes in a pediatric cohort with inclusion criteria of proteinuria and hematuria, which are common and practically relevant findings in any pediatric nephrology clinic. We detected disease-causing mutations in 14.1% of families sequenced, highlighting the utility of using genetics to obtain a definitive molecular etiology of disease. Clinicians should continue to utilize DNA sequencing technologies to further clinical practice, especially in cases of monogenic aHUS, where recently developed therapies such as eculizumab can be used to chronically treat the disease. Early angiotensin blockade can delay the onset of ESRD in patients with AS and an early molecular genetic etiology of disease is critical to helping delay a severe decline of renal function in pediatric patients [94].