De novo variants identified by trio whole exome sequencing of bladder exstrophy epispadias complex
Abstract
Bladder exstrophy epispadias complex (BEEC) encompasses a spectrum of conditions ranging from mild epispadias to the most severe form: omphalocele—bladder exstrophy—imperforate anus—spinal defects (OEIS). BEEC involves abnormalities related to anatomical structures that are proposed to have a similar underlying etiology and pathogenesis. In general, BEEC, is considered to arise from a sequence of events in embryonic development and is believed to be a multi-etiological disease with contributions from genetic and environmental factors. Several genes have been implicated and mouse models have been generated, including a knockout model of p63, which is involved in the synthesis of stratified epithelium. Mice lacking p63 have undifferentiated ventral urothelium. MNX1 has also been implicated. In addition, cigarette smoking, diazepam and clomid have been implied as environmental factors due to their relative association. By in large, the etiology and pathogenesis of human BEEC is unknown. We performed de novo analysis of whole exome sequencing (WES) of germline samples from 31 unrelated trios where the probands have a diagnosis of BEEC syndrome. We also evaluated the DECIPHER database to identify copy number variants (CNVs) in genes in individuals with the search terms “bladder exstrophy” in an attempt to identify additional candidate genes within these regions. Several de novo variants were identified; however, a candidate gene is still unclear. This data further supports the multi-etiological nature of BEEC.
1 INTRODUCTION
The acronym BEEC (OMIM 600057) bladder exstrophy and epispadias complex is umbrella terminology for a series of conditions that arise from a proposed common defect in blastogenesis. The variable phenotype is attributed to embryological timing (Keppler-Noreuil, 2001). Bladder exstrophy is the most common phenotype (1/30,000) (Cervellione et al., 2015) with the most severe and rarest form resulting in cloacal and gastrointestinal malformations in cases with OEIS occurring in 1/200,000–1/400,000 pregnancies. BEEC is described as etiologically heterogenous with proposed genetic influences (MNX1 OMIM #176450, p63 OMIM *603273) and environmental factors (cigarette smoke [Boyadjiev et al., 2004], diazepam [Lizcano-Gil et al., 1995]) interplaying to create this defect. The vast majority of cases have an unknown etiology. Recent genomic studies are emerging. Reutter performed whole exome sequencing on 8 trios (patient–mother–father) with cloacal exstrophy and identified 5 de novo variants. However, BEEC remains a rare condition making it challenging to identify large sample sizes and thus limiting the ability to identify novel monogenic genes.
The underlying clinical characteristics of BEEC are mixed. The majority of BEEC cases are sporadic without Mendelian inheritance. There are rare reports of familial cases; however, a genetic etiology is supported by highly concordant monozygotic twins. The male to female ratio is 6:1 (Boyadjiev et al., 2004). Young maternal age is reported by some, but not by others (Boyadjiev et al., 2004). There does appear to be an association with advanced paternal age, suggesting a possible role for de novo mutations (Boyadjiev et al., 2004). MNX1 is a well investigated gene that causes a uro-rectal fistula in Currarino syndrome; however, it has not been reported to cause the ventral defect identified in BEEC (Boyadjiev et al., 2004). Several possible etiological mechanisms have been proposed including cloacal membrane rupture and a vascular hypothesis. This hypothesis links OEIS to an early vascular event, supported by the common finding of a single umbilical artery (Stevenson, 2021).
Our objective is to identify candidate, monogenetic etiologies of BEEC. We hypothesize that novel BEEC related genes will be identified via whole exome sequencing (WES) and analysis of the DECIPHER database.
2 SUBJECTS AND METHODS
2.1 Subjects
2.1.1 Editorial policies and ethical considerations
Probands with BEEC syndrome and their parents were recruited by Johns Hopkins Pediatric Urology and sequenced as part of the Baylor Johns Hopkins Center for Mendelian Genetics (BHCMG) under Johns Hopkins IRB #NA-00091907. Probands and or parents provided written, informed consent as required per the IRB.
Probands were phenotyped through a high-volume pediatric urology service by experts experienced in the management of BEEC. For the construct of this article, the probands were divided into cases with classic bladder exstrophy (CBE) and those with OEIS. The average age of the CBE and the OEIS cohorts were 10.1 years and 5 months, respectively, at the time of enrollment (Table 1).
| Bladder exstrophy | OEIS | |
|---|---|---|
| N = 21 (%) | N = 10 (%) | |
| Gender | ||
| Female | 5/21 (23.8) | 3 (30) |
| Male | 16/21 (76.2) | 7 (70) |
| Race | ||
| White | 17 (81.0) | 10 (100) |
| Black or African American | 4 (19.0) | 0 (0) |
| Ethnicity | ||
| Not Hispanic or Latino | 21 (100) | 10 (100) |
| Age (days) | 3694.9 (505–9719, SD 3033.7) | 145.6 (1–570, SD 194.5) |
2.2 Germline whole exome sequencing and de novo analysis pipeline
WES was performed on germline DNA from 31 unrelated probands with BEEC syndrome and their parents through the BHCMG. WES was performed utilizing the Agilent SureSelect HumanAllExonV5Clinical_ S06588914 for library preparation. Libraries were sequenced on the HiSeq2500 platform using 125 bp paired end runs and sequencing chemistry kits HiSeq Rapid PE Cluster Kit v4 and HiSeq SBS Kit v4. Fastq files were aligned with BWA (Li & Durbin, 2009) version 0.7.8 to the 1000 genomes phase 2 GRCh37/hg19 human genome reference. Duplicate molecules were flagged with Picard version 1.74. Local realignment around indels and base call quality score recalibration was performed using the Genome Analysis Toolkit (GATK) (McKenna et al., 2010) version 3.0-0. Variant filtering was done using the Variant Quality Score Recalibration (VQSR) method (DePristo et al., 2011). Analysis of annotated files were performed using the PhenoDB analysis tool (Sobreira et al., 2015) to identify rare (MAF <0.01 in the 1000 Genomes Project, Exome Variant Server, gnomAD, and our in-house database), functional variants (missense, nonsense, frameshift, splicing, or stoploss) that fit recessive or dominant inheritance patterns.
2.3 Variant prioritization
- Variants present in two or more probands;
- Variants classified as pathogenic, likely pathogenic or variant of uncertain significance (VUS) in ClinVar (Landrum & Kattman, 2018);
- Variants not present in gnomAD;
- Variants in genes with OMIM phenotypes (https://omim.org) that have overlapping features with BEEC;
- Variants in genes with a known mouse models (http://www.informatics.jax.org/) with overlapping features with BEEC;
- Variants with CADD >20 (Rentzsch et al., 2019); and,
- Variants in genes with pLI >0.9 or missense Z-score >3.09.
2.4 DECIPHER analysis
The search term “bladder exstrophy” was queried in DECIPHER (Firth et al., 2009) to identify any previously published CNVs present in an individual with bladder exstrophy. We compared the genes in these regions to the genes with rare de novo variants identified by our WES analysis.
3 RESULTS
3.1 Demographics
We present 21 cases of CBE and 10 cases of OEIS. There were more males than females in both cohorts. The probands were predominately white and non-Hispanic with 4 cases of CBE in black or African American probands (Table 1).
3.2 Germline whole exome sequencing and de novo analysis pipeline
We identified a total of 26 rare de novo missense, nonsense, frameshift, splicing, or stoploss in our cohort of 31 trios. One proband had four rare de novo variants, one had three, three probands had two, 14 had one, and 13 probands had no rare de novo variant. A total of 16 variants were identified in probands with classic bladder exstrophy (one frameshift and 15 missense) (Table 2), while one frameshift and 10 missense variants were identified in probands with OEIS (Table 3).
| RefgeneGene name | Patient | Variant (hg37) | Variant consequence | gnomAD allele frequency | pLI score | Missense Z score | CADD score | OMIM phenotypes | Knockout mouse model |
|---|---|---|---|---|---|---|---|---|---|
| ANTXR1 | 1 | 2:69409770:G:A | p.R444Q | 2.03e-05 | 0.98 | 1.6 | 1.65 | GAPO syndrome; {?Hemangioma, capillary infantile, susceptibility to} AD/AR | Increased extracellular matrix in uterus and ovaries, normal fertility |
| CCDC179 | 1 | 11:22881943:C:G | p.E9Q | 7.767e-06 | n/a | n/a | 2.14 | n/a | n/a |
| EPHA1 | 2 | 7:143098593:T:A | p.I86F | n/a | 5.3e-11 | −0.33 | 4.04 | n/a | Vaginal atresia, infertility |
| KDM5B | 3 | 1:202698968:C:T | p.R1455H | 3.655e-05 | 5.09e-0.5 | 1.99 | 2.45 | n/a | Decreased body weight, premature mortality decreased serum estradiol levels |
| LMO7 | 1 | 13:76382302:C:G | p.R628T | 4.081e-06 | 0.04 | −1.72 | 1.76 | n/a | Growth defects, skeletal muscle, retinal |
| OR51A7 | 4 | 11:4928663:C:T | p.H22Y | 4.066e-05 | 1.66e-06 | −2.11 | 0.62 | n/a | n/a |
| PKD1 | 5 | 16:2164832:G:A | p.P731L | 0.0003 | 0.99 | −0.67 | −0.41 | Polycystic kidney disease 1, AD | Heterozygous, cystic kidneys |
| PRAMEF2 | 6 | 1:12921404:C:T | p.R399C | 4.09e-05 | 0.0006 | −1.26 | 0.58 | n/a | n/a |
| RAB23 | 7 | 6:57075162:A:C | p.M6R | n/a | 0.039 | 0.064 | 2.52 | Carpenter syndrome, AR | Homozygous, axial defects, exencephaly, retinal |
| SOX18 | 3 | 20:62679908:G:A | p.R256W | n/a | n/a | n/a | 2.00 | Hypotrichosis-lymphedema-telangiectasia, AD/AR | Homozygotes cardiac anomalies, abnormal hair, edema. |
| SPINT3 | 8 | 20:44141412:G:A | p.T50M | 1.322e-05 | n/a | n/a | −1.58 | n/a | n/a |
| STK38 | 9 | 6:36483153:T:- | p.H210fs | n/a | 0.85 | 1.77 | n/a | n/a | Susceptibility to bacterial infection |
| TCHHL1 | 10 | 1:152058629:A:G | p.V510A | 4.062e-06 | 2.2e-08 | −2.37 | −1.35 | n/a | n/a |
| TPSD1 | 4 | 16:1306825:G:C | p.R94S | 4.077e-05 | 8.52 e-09 | −2.22 | 1.79 | n/a | n/a |
| TSHR | 1 | 14:81534618:C:T | p.T88I | n/a | 5.8e-06 | −0.09 | 2.42 | Hyperthyroidism, familial gestational; AD/AR | Homozy: Thyroid, growth restriction, hearing |
| RefgeneGene name | Patient | Variant (hg37) | Variant consequence | gnomADallele frequency | pLI score | Missense Z score | CADD score | OMIM phenotypes | Knockout mouse model |
|---|---|---|---|---|---|---|---|---|---|
| EGFL6 | 11 | X: 13645187: G:A |
p.R449Q | 3.92 e-05 | 2.66 e-08 | 0.45 | 3.12 | n/a | Hemizygous mice are normal |
| EIF6 | 12 | 20: 33867824: T:C |
p.N156S | 1.6e-0.05 | 0.8 | 0.9 | 3.58 | n/a | Translation protein, homozygous lethal, hetero reduced body weight |
| FANCE | 13 | 6: 35425721:C:A |
p.P310Q | 0.0002 | 0.003 | −0.43 | 1.86 | Fanconi anemia, AR | n/a |
| FLCN | 14 | 17: 17117088:C:T |
p.A541T | 4.06 e-06 | 0.96 | 1.53 | 3.26 | Birt-Hogg-Dube syndrome, AD | Knockout mice are prenatal lethal |
| HSD11B1 | 15 | 1: 209907666:-:T |
p.V227fs | 0.0006 | 0.69 | 0.36 | n/a | Cortisone reductase deficiency 2, AD | Slim versus obese depending on variant |
| INTS2 | 14 | 17: 59958490:G:A |
p.P719L | n/a | 1.00 | 0.87 | 4.19 | n/a | n/a |
| RPL4 | 13 | 15: 66793778:C:A |
p.R204L | n/a | 0.99 | 0.98 | 4.30 | n/a | n/a |
| TCP11L1 | 16 | 11: 33078755:A:G |
p.I131V | 4.06e-06 | 0.09 | 0.62 | 3.56 | n/a | n/a |
| TMEM120B | 14 | 12: 122208866:T:C |
p.V202A | n/a | 0.26 | 1.52 | 2.72 | n/a | n/a |
| WDR81 | 17 | 17: 1630025:G:A |
p.R591H | n/a | n/a | n/a | 3.70 | Cerebellar ataxia, MR, AR | n/a |
| ZNF280B | 18 | 22: 22842787:C:T |
p.V313M | 0.0024 | 0.28 | −0.35 | 1.50 | n/a | n/a |
3.3 Variant prioritization
- None of these variants were identified in more than one proband.
- None of the de novo variants were classified as pathogenic or likely pathogenic in ClinVar. The FANCE p.P310Q, was the only variant classified as VUS in ClinVar.
- There were 9 variants that were absent from gnomAD. These included 4 missense variants (EPHA1, RAB12, SOX18, TSHR) and 1 frameshift variant (RAB23) in probands with bladder exstrophy. The 4 variants identified in probands with OEIS were all missense variants in genes: INTS2, RPL4, TMEM120B, and WDR81.
- There were 5 variants in the CBE cohort in genes associated with a phenotype in OMIM. These included missense heterozygous variants in ANTXR1 (GAPO syndrome; OMIM #230740), PKD1 (polycystic kidney disease; OMIM #601313), RAB23 (Carpenter syndrome; OMIM #201000), SOX18 (hypotrichosis-lymphedema-a-telangiectasia; OMIM #607823) and TSHR (hyperthyroidism; OMIM #609152).
In the OEIS cohort we identified a frameshift variant in HSD11B1 (Cortisone reductase deficiency 2; OMIM #604931) and 3 missense, heterozygous variants in FANCE (Fanconi anemia; OMIM #600901), FLCN (Birt-Hogg-Dube syndrome; OMIM #135150), and WDR81 (cerebellar ataxia; OMIM #610185).
- We then identified variants in 5 genes with a knockout mouse model and no human OMIM phenotype. These included 1 frameshift and 3 heterozygous missense variants in the CBE cohort. A frameshift variant was identified in STK38 for which a knockout mouse is susceptible to bacterial infection. The missense variants were in EPHA1 (vaginal atresia), KDM5B (decreased body weight), and LMO7 (growth defects in muscle).
A heterozygous missense variant in EIF6 which is lethal in the homozygous knockout mouse; was identified in a proband with OEIS.
- There were no variants with CADD scores greater than 20.
- There were 2 variants in bladder exstrophy probands (ANTXR1, PKD1) and 3 variants in OEIS probands (FLCN, INTS2, RRL4) with pLI scores of >0.9. No missense Z-scores were >3.09.
3.4 DECIPHER analysis
Only three individuals with bladder exstrophy are identified in DECIPHER. All three of them had additional complex findings in addition to bladder exstrophy. One proband had a pathogenic deletion (16:29417210–30178708, 761.5 kb) (DECIPHER ID); one had a pathogenic duplication (16:46407633–62056029, 15.65 Mb) (DECIPHER ID); and one had a pathogenic triplication (8:81148983–83598327, 2.45 Mb) and multiple benign CNVs (DECIPHER ID).
None of the copy number variations identified in these individuals affected genes that we selected in the de novo analysis of our cohort.
4 DISCUSSION
BEEC remains a complex series of congenital defects that likely arise from an embryological insult. Cases appear to have a targeted origin without affecting neurological cognition or global neuro-motor function. Our analysis of rare, de novo variants in 31 probands with BEEC syndrome identified 26 genes. None of the variants were clinically significant in ClinVar, they all had CADD scores less than 20 and mZ less than 3.09; thus, analysis of these variants was not strongly suggestive of causation.
Several of the genes with de novo variants had OMIM phenotypes unrelated to the phenotype of our cohort. Interestingly, in the AR state FANCE presents as VACTERL which has an overlapping BEEC phenotype. Unfortunately, our case only had a heterozygous missense mutation that was not classified in ClinVar and the neonate had no signs/symptoms of anemia as would be expected with homozygous FANCE variants. Another candidate gene, EPHA1, is of interest due to the mouse phenotype of vaginal atresia in knockout mice. This missense mutation identified has a low CADD score and heterozygous mice appear phenotypically normal.
There were no overlapping rare de novo germline variants from our cohort that also overlapped with the pathogenic CNVs that were identified in DECIPHER. All of the cases in DECIPHER had large pathogenic CNVs in conjunction with additional syndromic features. The lack of a CNV association with cases of isolated bladder exstrophy further strengthens the argument for a heterogenous rather than single gene etiology of BEEC.
One important limitation of our study is that many of our candidate genes are implicated only in a single proband. None of our top candidates have exonic variants in more than one proband. Another limitation is that we did not have access to the identified DECIPHER individuals in order to confirm their phenotype and inquire about additional features.
In conclusion, we identified several de novo variants in genes from probands with BEEC. Unfortunately, the majority of these variants are missense mutations that are of unknown significance and appear to be benign based on variant in silico prediction scores, thus no strong candidate gene was identified.
AUTHOR CONTRIBUTIONS
Angie C. Jelin: formal analysis, funding, writing original draft. Elizabeth Wohler: formal analysis, review and editing. Renan Martin: formal analysis, review and editing. Heather Di Carlo: sample acquisition, conceptualization, review and editing. William Isaacs: conceptualization, sample storage review and editing. Joan Ko: conceptualization, sample acquisition, review and editing. Jason Michaud: conceptualization, sample acquisition, review and editing. Karin Blakemore: conceptualization, review and editing. David Valle: conceptualization, methodology, funding, review and editing. Nara Sobreira: formal analysis, methodology, conceptualization, writing, editing. John Gearhart: conceptualization, sample acquisition, review and editing.
ACKNOWLEDGMENTS
We thank all of the patients and parents who participated in this study. This study makes use of data generated by the DECIPHER community. A full list of centers who contributed to the generation of the data is available from https://decipher.sanger.ac.uk and via email from [email protected].
FUNDING INFORMATION
Dr. Jelin is supported by grant 5K23DK119949-02 from the National Institutes of Health (NIH). Dr. Valle, Dr. Sobreira, and Renan Martin are supported by 5UM1HG006542-08. The contents of the publication are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
CONFLICT OF INTEREST STATEMENT
The authors declare no conflicts of interest.
Open Research
DATA AVAILABILITY STATEMENT
The data that support the findings of this study are openly available in ClinVar at https://www.ncbi.nlm.nih.gov/clinvar/, reference number SCV004101633–SCV004101658.
