Kidney International. 2016 Feb;89(2):476-86.
Prioritization and burden analysis of rare variants in 208 candidate genes suggest they do not play a major role in CAKUT.
Nicolaou N, Pulit SL, Nijman IJ, Monroe GR, Feitz WF, Schreuder MF, van Eerde AM, de Jong TP, Giltay JC, van der Zwaag B, Havenith MR, Zwakenberg S, van der Zanden LF, Poelmans G, Cornelissen EA, Lilien MR, Franke B, Roeleveld N, van Rooij IA, Cuppen E, Bongers EM, Giles RH, Knoers NV, and Renkema KY.
Department of Genetics, Center for Molecular Medicine, University Medical Center Utrecht, Utrecht, The Netherlands.
Department of Urology, Radboudumc Amalia Children’s Hospital, Radboud University Medical Center, Nijmegen, The Netherlands.
Department of Pediatrics, Radboudumc Amalia Children’s Hospital, Radboud University Medical Center, Nijmegen, The Netherlands.
Department of Urology, University Medical Center Utrecht, Utrecht, The Netherlands.
Department for Health Evidence, Radboud University Medical Center, Nijmegen, The Netherlands.
Department of Molecular Animal Physiology, Donders Institute for Brain, Cognition and Behaviour, Radboud Institute for Molecular Life Sciences (RIMLS), Radboud University, Nijmegen, The Netherlands.
Department of Human Genetics, Radboud University Medical Center, Nijmegen, The Netherlands.
Department of Cognitive Neuroscience, Donders Institute for Brain, Cognition and Behaviour, Radboud University Medical Center, Nijmegen, The Netherlands.
Department of Pediatric Nephrology, Wilhelmina Children’s Hospital, University Medical Center Utrecht, Utrecht, The Netherlands.
Department of Nephrology and Hypertension, University Medical Center Utrecht, Utrecht, The Netherlands.
The leading cause of end-stage renal disease in children is attributed to congenital anomalies of the kidney and urinary tract (CAKUT). Familial clustering and mouse models support the presence of monogenic causes. Genetic testing is insufficient as it mainly focuses on HNF1B and PAX2 mutations that are thought to explain CAKUT in 5-15% of patients. To identify novel, potentially pathogenic variants in additional genes, we designed a panel of genes identified from studies on familial forms of isolated or syndromic CAKUT and genes suggested by in vitro and in vivo CAKUT models. The coding exons of 208 genes were analyzed in 453 patients with CAKUT using next-generation sequencing. Rare truncating, splice-site variants, and non-synonymous variants, predicted to be deleterious and conserved, were prioritized as the most promising variants to have an effect on CAKUT. Previously reported disease-causing mutations were detected, but only five were fully penetrant causal mutations that improved diagnosis. We prioritized 148 candidate variants in 151 patients, found in 82 genes, for follow-up studies. Using a burden test, no significant excess of rare variants in any of the genes in our cohort compared with controls was found. Thus, in a study representing the largest set of genes analyzed in CAKUT patients to date, the contribution of previously implicated genes to CAKUT risk was significantly smaller than expected, and the disease may be more complex than previously assumed.
Congenital anomalies of the kidney and urinary tract (CAKUT) refer to a spectrum of structural malformations. Phenotypes include kidney agenesis, multicystic dysplastic kidneys, kidney dysplasia, kidney hypoplasia, duplex collecting system, ureteropelvic junction obstruction, vesico-ureteral reflux, megaureter, ectopic ureter, horseshoe kidney, and posterior urethral valves, although it is debatable whether the latter belongs to the same spectrum of malformations due to a different developmental origin of the lower urinary tract. The prevalence of CAKUT is 3-6 per 1,000 live births. CAKUT can occur in combination with other organ malformations. The development of the kidney and urinary tract begins around day 22 of gestation in humans. An orchestrated interplay of growth factors, signaling factors, transcription factors, and other regulatory factors is known to be essential for normal kidney and urinary tract development.
Disturbances in normal kidney and urinary tract development result in CAKUT and, thereby, affect kidney function. CAKUT is the most common cause of end-stage renal disease in children that accounts for ~40% of the children that receive renal replacement therapies, i.e. dialysis or kidney transplantation. By providing insight into the etiological factors that contribute to CAKUT, we aim to improve diagnostics, counseling, and personalized care for CAKUT patients and their relatives.
Table 1 – The 208 human CAKUT candidate genes that were included in the gene panel design for targeted next-generation sequencing in CAKUT patients, in alphabetical order.
Genetic and environmental factors are proposed to be involved in CAKUT etiology. Familial clustering in ~10-20% of CAKUT cases indicates a strong genetic contribution to pathogenesis. Furthermore, single gene defects were previously shown to be causal in CAKUT-related multi-organ syndromes and in numerous animal models displaying renal malformations, suggesting a monogenic cause for CAKUT. However, a dominant inheritance pattern with incomplete penetrance is often observed in CAKUT families. Moreover, the clinical phenotype and severity of CAKUT vary among patients, even within and among families with the same underlying mutation. These findings suggest a complex etiological background involving multiple genetic, epigenetic, and environmental factors.1 Previous investigations on human CAKUT etiology included single candidate gene studies, linkage analysis, copy number variation analysis, whole exome sequencing, whole genome sequencing, and environmental studies. Still, for the majority of CAKUT cases the underlying molecular mechanisms of disease are unknown. In the study described here, we aimed to provide an in-depth analysis of 208 candidate genes for CAKUT in a phenotypically heterogeneous cohort of 453 CAKUT patients.2 We used targeted next generation sequencing to identify potentially pathogenic variants in CAKUT candidate genes. The gene panel design comprised 0.9 mega base pairs, based on the coding regions of genes that were 1) suggested from in vitro and in vivo models of CAKUT, 2) known from multi-organ syndromes including kidney or urinary tract malformations, and 3) previously implicated in human CAKUT (Table 1).
In total, 11,885 variants were identified in 453 CAKUT patients by targeted next generation sequencing. We performed a multistep filtering and prioritization approach to select the variants that were potentially disease-causing (Figure 1). These selected variants included rare variants with a minor allele frequency (MAF) < 0.01 in control databases that were either truncating (frameshift, nonsense, canonical splice site variants, or in frame insertions/deletions) or previously reported as disease-causing. This resulted in 256 variants that were included in a validation experiment using Sanger sequencing. Eventually, 180 variants were confirmed and subsequently classified into four groups (Figure 2). The groups were based on available evidence for a deleterious effect.
Figure 1 – Workflow to filter and prioritize variants identified in CAKUT patients. MAF, minor allele frequency; HGMD, human gene mutation database.
Groups 1-3 comprised 148 candidate variants in 82 different genes in 151/453 (33%) of CAKUT patients. The variants were prioritized for follow-up studies based on their population frequency, evolutionary sequence conservation, and multiple in silico methods that predict a deleterious effect on the biological function of the gene. Five heterozygous variants in four different genes (HNF1B, PAX2, SIX5, UMOD) were detected in six patients and were classified as causal mutations for autosomal dominant diseases (group 1). Fifteen candidate variants of pathogenicity were found in 14 genes among 20 patients (group 2). Likely deleterious variants comprised 101 missense variants that were in silico predicted to be deleterious and not seen before (group 3). More than 98% of the variants in group 3 were predicted to be among the 10% most deleterious of all possible substitutions of the human genome, using Combined Annotation Dependent Depletion (CADD; http://cadd.gs.washington.edu/).3 Of note, 32 variants, found in 69 patients were previously reported as disease-causing in HGMD Pro (group 4). After thorough investigation, we classified the variants in group 4 as variants of uncertain significance, based on their prevalent occurrence in population database, the lack of functional evidence for a pathogenic effect, or no reported significant association with the disease.
To determine whether any of the selected 208 candidate genes exhibited a significant excess burden of rare coding variants in the patients, we compared variant data of 434 patients of Dutch ancestry from our CAKUT cohort with 498 external healthy controls from the Genome of the Netherlands project.4 We performed five different burden tests: T1, which considers variants with frequency <1%; T5, which considers variants with frequency <5%; the variable threshold test, which considers multiple frequency cutoffs for tested variants and then adjusts for multiple-test correction; the weighted-sum test, which weights variants in a manner inversely proportional to their frequency; and SKAT, which allows for both protective and risk mutations in a single gene. However, none of the genes surpassed exome-wide significance (p < 5*10-7). Identical burden testing was performed using data from the 1000 Genomes Phase I, which replicated these findings.
Figure 2 – Classification of 180 validated variants in CAKUT patients. Four groups of variants were defined, based on available evidence from databases and literature indicating deleterious/damaging effects: 1) causal mutations, 2) candidates of pathogenicity, 3) likely deleterious variants and 4) variants of uncertain significance. CAKUT, congenital anomalies of the kidney and urinary tract.
In this study, we sequenced 208 genes, suspected to have a role in CAKUT pathogenesis, in 453 patients with CAKUT. We identified 148 variants in 82 genes in one third of the patient cohort. Variants have been prioritized for follow-up studies. Approximately 98% of the variants were predicted to be among the 10% most deleterious of all possible substitutions possible in the human genome. Thus, promising variants have not been detected in two-thirds of our patients, indicating that the genetic heterogeneity of this disease is likely much higher than previously assumed. Comparing the excess of rare variants in each gene with those identified in ancestry-matched controls in burden testing did not reveal significant associations, showing that rare variants in these 208 genes are not more prevalent in the patients compared to controls.
To conclude, our results indicate that the genetic architecture of CAKUT is far more complex than previously implicated. We underline that detection of novel and likely deleterious variants in any known genes does not by itself imply pathogenicity, as erroneously reported in many recent publications. Future investigations on CAKUT etiology should focus on this complex side of the story, including association and whole genome sequencing studies in large cohorts. Large collaborative efforts that cross borders and disciplines are indispensable. Diagnostic procedures for CAKUT patients should involve multiple gene screening, resulting in improved diagnostic accuracy for this heterogeneous disease and an essential step towards appropriate recurrence risk estimations and genetic counseling.
- Nicolaou N, Renkema KY, Bongers EM, Giles RH, Knoers NV. Genetic, environmental, and epigenetic factors involved in CAKUT. Nat Rev Nephrol 11, 720-731 (2015).
- Nicolaou N, Pulit SL, Nijman IJ, Monroe GR, Feitz WF, Schreuder MF, van Eerde AM, de Jong TP, Giltay JC, van der Zwaag B, Havenith MR, Zwakenberg S, van der Zanden LF, Poelmans G, Cornelissen EA, Lilien MR, Franke B, Roeleveld N, van Rooij IA, Cuppen E, Bongers EM, Giles RH, Knoers NV, Renkema KY. Prioritization and burden analysis of rare variants in 208 candidate genes suggest they do not play a major role in CAKUT. Kidney Int 89, 476-486 (2016).
- Kircher M, Witten DM, Jain P, O’Roak BJ, Cooper GM, Shendure J. A general framework for estimating the relative pathogenicity of human genetic variants. Nat Genet 46, 310-315 (2014).
- Boomsma DI, Wijmenga C, Slagboom EP, Swertz MA, Karssen LC, Abdellaoui A, Ye K, Guryev V, Vermaat M, van Dijk F, Francioli LC, Hottenga JJ, Laros JF, Li Q, Li Y, Cao H, Chen R, Du Y, Li N, Cao S, van Setten J, Menelaou A, Pulit SL, Hehir-Kwa JY, Beekman M, Elbers CC, Byelas H, de Craen AJ, Deelen P, Dijkstra M, den Dunnen JT, de Knijff P, Houwing-Duistermaat J, Koval V, Estrada K, Hofman A, Kanterakis A, Enckevort D, Mai H, Kattenberg M, van Leeuwen EM, Neerincx PB, Oostra B, Rivadeneira F, Suchiman EH, Uitterlinden AG, Willemsen G, Wolffenbuttel BH, Wang J, de Bakker PI, van Ommen GJ, van Duijn CM. The Genome of the Netherlands: design, and project goals. Eur J Hum Genet 22, 221-227 (2014).
This study was financially supported by the European Community’s Seventh Framework Programme (FP7/2009) under grant agreements 305608 (EURenOmics) and the Dutch Kidney Foundation under grant agreement KSTP.10.004.
Kirsten Renkema, PhD
Department of Genetics, STR.1.305
Center for Molecular Medicine
University Medical Center Utrecht
P.O. Box 85060
3508 AB Utrecht