Molecular diagnostic in fetuses with isolated congenital anomalies of the kidney and urinary tract by whole‐exome sequencing

Abstract Background In prenatal care, accumulating evidences has demonstrated that whole‐exome sequencing (WES) expedites the genetic diagnosis of fetal structural anomalies. However, the clinical value of WES in the diagnosis of prenatal isolated congenital anomalies of the kidney and urinary tract (CAKUT) is unknown. Methods Forty‐one fetuses with unexplained isolated CAKUT, normal karyotype and negative chromosomal microarray analysis (CMA) results, underwent WES and were accordingly grouped as (a) Group 1: complex cases with bilateral renal abnormalities (N = 19); and (b) Group 2: cases with isolated unilateral fetal renal abnormalities (N = 22). Results The detection rate of WES for pathogenic variants and incidental variants was 7.32% (3/41) and 2.4% (1/41), respectively. The three pathogenic variants were identified in the genes ACTA2 (multisystem smooth muscle dysfunction syndrome), PKHD1 (autosomal recessive form of polycystic kidney disease), and PKD1 (autosomal dominant polycystic kidney disease type 1). The incidental variants were detected in genes PPM1D (syndromic neurodevelopmental disorders). Furthermore, all above fetuses carrying pathogenic variants came from bilateral kidney anomalies. Thus, the detection rate was 0 for fetuses with unilateral fetal renal abnormalities and 15.7% (3/19) for bilateral renal abnormalities. Conclusion This cohort shows that prenatal WES is a supplementary approach for the etiologic diagnosis of unexplained isolated CAKUT with negative CMA, especially for fetuses with bilateral renal abnormality.


| INTRODUC TI ON
Congenital anomalies of the kidney and urinary tract (CAKUT), the incidence of which is around 0.2% by prenatal sonography, have been reported to accounts for 40%-50% of the global pediatric end-stage renal disease (ESRD) as well as 7% of global adult ESRD. 1 Accumulating evidences from animal and human studies suggest that genetic factors contribute to CAKUT. 2 To date, more than 200 clinical syndromes were described phenotypes of CAKUT, 3 and approximately 40 monogenic genes were identified to cause renal and urinary abnormality if mutated, which explained 5%-20% of CAKUT cases. 4 In prenatal care, microarray approaches are currently most recommended as the first-line genetic test for fetuses with CAKUT, achieving over 3.8% incremental yield of detecting pathogenic copy number variants (CNVs) in cases with normal karyotype. 5 However, most definitive diagnoses are challenging as part of the monogenic origin and the phenotype heterogeneity during an ongoing pregnancy. 6 Whole-exome sequencing (WES) has been widely applied as a first-line clinical diagnosis, especially in children's metabolic and neurodevelopmental disorders. In prenatal care, two large-scale prenatal studies suggested that WES facilitates the diagnosis of fetal structural anomalies, enabling fetal prognosis with higher accuracy while reducing the risk of recurrence in future pregnancies. 7,8 However, previous studies have reported many challenges regarding the clinical application of WES in prenatal care, besides prenatal phenotyping limitations and technical problems, the counseling was also challenged with incidental or secondary findings, as well as variants of uncertain significance. 9,10 Meanwhile, it was suggested that diagnostic yields were higher in multiple fetal anomalies. 10 However, the observations of most CAKUT by prenatal sonography were isolated and sporadic, which suggests that genetic origin that potentially contributes to this cohort should be considered. 11 In the field of prenatal study, the utility of WES for the diagnosis of fetal CAKUT remains uncertain and the effectiveness and feasibility of WES in prenatal urinary malformation, especially isolated sporadic cohort, needs to be further investigated.
In the present study, we performed WES analysis upon 41 isolated sporadic CAKUT cases with negative CMA results, exploring the application value of WES in the perinatal management of the cohort by means of evaluating the phenotypic, genotypic, and prognosis outcomes.

| Participant recruitment and sample collection
This is an observational cohort study of pregnant women attending the Nanjing Maternity and Child Health Care Hospital (Jiangsu, China) for prenatal diagnosis from January 1, 2012, to June 1,

This study was approved by the Medical Ethics Committee
of Nanjing Maternity and Child Health Care Hospital. All data were prospectively collected from our patient files. The inclusion criteria for fetuses were isolated renal abnormalities detected by the 22-24 weeks anomaly screening as prenatal ultrasound showed no signs of other fetal structural anomalies. Informed consent was obtained from parents. For fetuses whose tests revealed aneuploidy or CNVs which justified the fetal anomalous phenotypes, they were excluded from subsequent analyses. Then, a total of 41 subjects with negative CMA results (CMA results presenting no pathogenic genetic cause) were subsequently subjected to WES, including 19 cases with bilateral renal abnormalities and 22 cases with unilateral renal abnormality (URA), while unilateral kidney anomalies in combination with oligohydramnios which suggested possible kidney affection were also considered to be bilateral kidney anomalies. The inclusion criteria for URA were unilateral renal agenesis, multicystic dysplastic kidney, or kidney hypoplasia. We made clear and professional communications with parents about the expectations and worries, thus ensuring the parents with a thorough understanding before taking part into such testing. Participants were informed that only CAKUT-

| Extraction of genomic DNA
Invasive sampling was carried out by amniocentesis. Genomic DNA of the fetuses and parents was, respectively, collected from amniotic fluid and peripheral blood by routine methods. A Qubit Fluorometer (Thermo Fisher Scientific) and electrophoresis were utilized to conduct the quality and quantity control of genomic DNA.

| Variant validation by PCR and Sanger sequencing
The pathogenic and likely pathogenic variations in fetal phenotyperelated genes were validated by Sanger sequencing. All primer designs were generated with the Primer Premier 5 software (Premier Biosoft International, PCR conditions available on request). Sanger sequencing was completed on the ABI 3730xl DNA analyzer (Applied Biosystems).

| Variation analysis
All candidate variants were defined as pathogenic, likely pathogenic, of uncertain significance, likely benign or benign following the instructions of the ACMG guidelines, 14 and the results were subsequently reviewed by pediatric nephrologist, clinical geneticist, and pediatric radiologist, respectively. The pathogenic and likely pathogenic variants in fetal phenotype-related genes were reported to the parents. Meanwhile, the secondary findings and the incidental findings which were unrelated to the fetal CAKUT phenotypes but known to cause moderate to severe childhood onset disorders were only reported when parents requested for the results.

| Demographic characteristics and pregnancy outcome
The median age of pregnant women in this study was 31 years old, and the gestational age (GA) varied between 22 and 32 weeks (median GA = 23 weeks). The fetuses were from healthy non-consanguineous parents. All cases with no family history of CAKUT, only one case whose mother had congenital left kidney agenesis. Of the 41 fetuses, 19 exhibited bilateral prenatal kidney anomalies, including bilateral enlarged or hyperechogenic kidneys, bilateral kidney hypoplasia and bilateral multicystic dysplastic kidney (MCDK) et al Detailed information about prenatal findings in each fetus is presented in Table 1. 89% (17/19) of bilateral cases were combined with oligohydramnion and underwent termination of pregnancy (TOP). Additionally, two cases were delivered at full term, which were diagnosed with bilateral enlarged or hyperechogenic kidneys by prenatal ultrasound. Accordingly, the postnatal ultrasound detected kidneys abnormality including 1 case of left polycystic kidney disease and 1 case of diffuse renal disease. No developmental abnormalities were detected in the follow-up interviews.

| Pathogenic variations
DNA samples from 41 fetuses without causal anomalies on CMA were detected by WES. Pathogenic or likely pathogenic variants in PKD1, ACTA2, and PKHD1 genes were detected, and the diagnostic rate of WES was 3/41 (7.3%) ( Table 2). Moreover, the above-men-   (ARPKD). 16 The pregnancy of Case 3 was terminated due to oligohydramnios at the 27th week.

| Incidental findings
We detected incidental findings in Case 4 with unilateral renal agenesis (Family 38) and a de novo frame shift variant (c.1434delC) in PPM1D, suggesting a loss-of-function mechanism ( Table 2). This variant was defined as "pathogenic" according to the ACMG guidelines.
Previous reports demonstrated the presence of PPM1D gene in fetal and adult human brain tissues, 17 which is integrally involved in the mild to severe intellectual disability (ID), developmental delay, behavioral problems such as anxiety disorders, hypotonia, and periods of fever and vomiting, 18 while no renal phenotype was described.
Since the above-mentioned features were not compatible with the prenatal findings in our study, we therefore defined the variant as   23 Obviously, the yield was higher when the fetuses were affected by bilateral renal abnormalities.

| D ISCUSS I ON
The value of prenatal WES in genetic diagnosis of CAKUT with/ without other anomalies was thoroughly discussed in several recent studies (Table 3)  were affected by ARPKD. 25 Furthermore, Leire Gondra reported that HNF1B mutation represents the leading cause of polyhydramnios associated with hyperechogenic (and sometimes enlarged) kidneys. 26 In this study, a heterozygous missense mutation (PKD1:  31 However, it is controversial that the ACMG guidelines exclude prenatal WES, which has been discussed in some publications. Sarah Harris agreed that patients must be informed of the possibility of these results in advance. For variants located in genes which were reported to cause moderate to severe childhood onset disorders but not found to be related to phenotypes in fetus. 9 Kristin G recommended that these variants should be included in the returned results if found to be highly penetrant pathogenic. Since many fetal disorders, including nonsyndromic intellectual disability, neurodevelopmental disorders, and metabolic disorder, cannot be diagnosed by fetal imaging, adequate pretest and post-test counseling are therefore supplemented as a commonly acknowledged practice in this area. 12 Experts also suggest performing WES on those highly selected cohorts when a specific syndrome is suspected.

| CON CLUS ION
Based on our current findings and previous reports, WES offers a supplementary approach for prenatal genetic diagnosis of isolated CAKUT with negative CMA result. The prenatal WES is more suitable for the etiologic diagnosis of diseases with higher risk of monogenic causes, such as bilateral renal malformation. The supplementary approach may benefit genetic counseling, decision-making on pregnancy outcome, and future family planning.

ACK N OWLED G M ENTS
This study was sponsored by the National Natural Science