Skip to main content
Journal of the American Society of Nephrology : JASN logoLink to Journal of the American Society of Nephrology : JASN
. 2024 Sep 26;36(2):256–263. doi: 10.1681/ASN.0000000503

Systematic Screening of Autosomal Dominant Tubulointerstitial Kidney Disease–MUC1 27dupC Pathogenic Variant through Exome Sequencing

Ilias Bensouna 1,2, Thomas Robert 3,4, Xavier Vanhoye 5, Marine Dancer 5, Laure Raymond 5, Pierre Delaugère 6, Pascale Hilbert 7, Hugues Richard 8,9, Laurent Mesnard 1,2,10,11,✉
PMCID: PMC11801747  PMID: 39325540

Visual Abstract

graphic file with name jasn-36-256-g001.jpg

Keywords: CKD, genetic kidney disease, human genetics, kidney tubule, molecular genetics

Abstract

Key Points

  • MUC1 is associated with autosomal dominant tubulointerstitial kidney disease, a genetic disorder progressing to kidney failure.

  • Variations in this gene are not easily diagnosed by conventional methods due to the MUC1 architecture, which contains a variable number of tandem repeats.

  • Using dedicated bioinformatics tools, we systematically detected the presence of 27dupC most common MUC1 pathogenic variant from exome sequencing data.

Background

The MUC1 gene is associated with autosomal dominant tubulointerstitial kidney disease (ADTKD), leading to CKD. Current methods of sequencing, such as exome sequencing, rarely detect MUC1 pathogenic variants because of the variable number of tandem repeats (VNTR) in MUC1 exon2. We demonstrated that combining fast read filtering with a sensitive VNTR genotyping strategy enables systematic screening of 27dupC pathogenic MUC1 variant from exome data.

Methods

We initially validated our bioinformatics pipeline in a proof-of-concept cohort incorporating exome data from 33 participants with a known MUC1 pathogenic variant identified by Snapshot PCR and confirmed by 54 MUC1-negative individuals for negative control. We then retrospectively analyzed exome sequencing data from January 2019 to October 2023 from 3512 adult participants with nephropathy of unknown origin. Finally, we prospectively validated our pipeline in 825 additional participants enrolled from November 2023.

Results

SharkVNTyper accurately identified MUC1 variants in 32 of 33 participants and excluded its presence in all the 54 negative controls in the proof-of-concept cohort (sensitivity of 97%, specificity of 100%). Integration of the Shark tool with VNTyper significantly reduced running time from 6–12 hours to 5–10 minutes per sample, allowing both retrospective and prospective analyses. In the retrospective cohort, SharkVNTyper identified 23 additional positive participants who were not suspected clinically and had been missed in the initial exome analysis; 18 of these participants were confirmed as carrying the MUC1 27dupC mutation by low-throughput Snapshot PCR. In the prospective cohort of 825 participants with CKD, systematic screening discovered 13 positive participants, with 12 confirmed by PCR. Overall, of 63 participants (1.4% of 4653) with molecularly confirmed ADTKD-MUC1, comprehensive diagnoses and descriptions of the disease were available for 24 participants. The median age of kidney failure was 50 years, 38% exhibited bilateral multiple kidney cysts, 8% had early-onset gout, and 58% had arterial hypertension.

Conclusions

SharkVNTyper enabled the analysis of highly repeated regions, such as the MUC1 VNTR, and facilitated the systematic screening of ADTKD-MUC1 from exome data, fostering 27dupC variation identification.

Podcast

This article contains a podcast at https://dts.podtrac.com/redirect.mp3/www.asn-online.org/media/podcast/JASN/2024_11_15_KTS_November2024.mp3

Introduction

The improvement of next-generation sequencing, mostly on the basis of high-throughput sequencing of short reads, has revolutionized genomic analysis.1 Inherent to the length of reads produced, next-generation sequencing methods face limitations in detecting copy number variants or variants in regions with highly recurrent motifs or repeats.2 This poses technical challenges for some clinical genetic diagnoses, particularly in diseases with genes involving variable number of tandem repeat (VNTR) regions, such as the MUC1 gene.

MUC1, encoding the transmembrane glycoprotein mucin-1, has been implicated in autosomal dominant tubulointerstitial kidney diseases (ADTKDs) and is currently referred to as ADTKD-MUC1.3–5 The MUC1 gene contains a distinctive region of 25–120 guanine-cytosine–rich VNTRs of 60 base pairs (20 amino acids). Notably, the number of VNTRs can significantly vary among individuals from 30 to 80. In addition, VNTR expansion and sequence may exhibit notable divergence across different participants. Therefore, significant differences may exist compared with a linear reference genome because healthy individuals with MUC1 repeats in a family complicate the picture.6 Nevertheless, a cytosine insertion typically occurs in individuals with ADTKD-MUC1, resulting in a 8C-homopolymer specifically located between positions 53 and 59 of the VNTR unit. So far, this variation, 27dupC, has been found in more than 80% of participants with ADTKD-MUC1.7 This insertion leads to a frameshift in the reading frame, generating a premature stop codon shortly after the VNTR domain and consequently a truncated form of the mucin-1 protein.8 Kirby et al. and Wenzel et al. have studied several affected family members with the cytosine insertion within MUC1 and revealed a common unit or pattern of sequences shared between all individuals, indicating that the structure of this VNTR is not completely random.9,10 Cases of MUC1 variant detection by exome sequencing (ES) have been described where the variation affects the very first VNTR, facilitating the alignment of sequencing reads to the reference genome, but such occurrences seem to be somewhat anecdotal.11

Currently, available diagnostic methods for ADTKD-MUC1 include Snapshot PCR or immunohistochemistry of truncated proteins on kidney biopsies. These methods are expensive, impractical for routine clinical use, or not tailored for systematic molecular screening.12–14 Recently, the VNTyper and HaloPlex pipelines have been developed for fine-calling variants from the MUC1 coding VNTR region from targeted next-generation sequencing panels. However, their applicability to larger datasets such as ES or genome sequencing (GS), both usually associated with lower read depth, is under debate, partly because of issues related to computing scalability.3 Authors have pointed out that, in addition to the quality of reads, the applicability of VNTyper to ES or GS data can be limited because of an average coverage level around 80× to 30×, respectively, compared with the optimal 300–500× coverage for which VNTyper has been built. The same limitations apply to the HaloPlex, which requires even higher read depth (>1000×) with specific HaloPlex probes.15

Current next-generation sequencing targeted trends rely on in silico analysis, exemplified by the Genomic England project PanelApps. These in silico approaches are cost-effective, particularly when applied to ES or GS data. This strategy seems relevant for the first-hand management of adult participants with unknown nephropathies. To bridge the gap between ES and GS data and the detection of MUC1 variations, we further adapted the open-source VNTyper pipeline by adding a filtering step using Shark software.16,17 As opposed to VNTyper, this novel pipeline that we called SharkVNTyper allows computing scalability, with similar sensitivity and specificity. We prospectively tested its reliability for systematic detection of MUC1 variation from ES data.

Methods

SharkVNTyper Pipeline for Detecting MUC1 VNTR Pathogenic Variants

The VNTyper pipeline, developed by Saei et al. in 2023, can detect the MUC1 VNTR pathogenic variant from targeted sequencing data when the average sequence coverage is above 500×.18 This pipeline was built on the basis of the open-source Kestrel mapping-free variant caller for short-read sequencing data.19 We fine-tuned VNTyper to scale it up, from panel sequencing data files (a few MB) to much larger sequencing files (up to 100 GB) such as the one generated from ES (coverage of 80×) or GS (30×). To achieve this, we used Shark (Denti et al., 2021, available at: https://github.com/AlgoLab/shark), a tool that uses Bloom filters to efficiently retrieve short reads corresponding only to a given set of gene sequences (in our case, the MUC1 gene). This filtering step significantly reduces the processing time of the variant-calling steps.16,17 Given the variability of MUC1 sequences among individuals, we curated a comprehensive set of high-quality MUC1 alleles from previous publications and used it as a basis for filtering.9,10 This approach ensured that only reads relevant to MUC1 were retained for analysis. To validate the effectiveness of this workflow, combining Shark and VNTyper (called SharkVNType), we compared results obtained from both unfiltered fastq files and “sharked” files, in terms of both accuracy and running time. The workflow of SharkVNTyper is illustrated in Supplemental Figure 1.

Study Population

We initially established a proof-of-concept cohort by screening 333 individuals with a clinical suspicion of ADTKD-MUC1 between January 2019 and April 2024. All participants included in this cohort had been previously analyzed negative by ES. After Snapshot PCR detecting the 27dupC insertion, 33 individuals of the 333 tested were positive, confirming that they were participants with ADTKD-MUC1 belonging to 19 families, and serving as positive controls (Figure 1 and Supplemental Figure 2). The negative control group included 54 participants with a negative 27dupC result using Snapshot PCR and a classical ES pipeline.

Figure 1.

Figure 1

Workflow for SharkVNTyper development, proof-of-concept cohort in 333 participants with clinically suspected ADTKD, retrospective cohort of 3512 participants with CKD, and prospective cohort of 825 participants. (a) SharkVNTyper uses both Shark and VNTyper tools to detect the 27dupC pathogenic variant from fastq files of ES. (b) For the proof-of-concept cohort, we applied VNTyper alone (without Shark) and SharkVNTyper (the tool combining both Shark and VNTyper) to control participants to evaluate both methods. VNTyper alone was positive for all positive control participants, but when using Shark with VNTyper (SharkVNTyper), one of the participants tested negative. (c) After the proof-of-concept phase, SharkVNTyper was retrospectively applied to any ES participant prescribed for unknown nephropathy from August 2021 to September 2023 (excluding 333 participants with ADTKD allocated to a proof-of concept cohort from Sorbonne University). Confirmatory snapshot PCR was performed for any SharkVNTyper-positive participants. (d) From October 2023 and onward, SharkVNTyper was prospectively integrated with ES analysis to concurrently identify pathogenic 27dupC MUC1 variant. Again, a confirmatory snapshot PCR was performed for each positive sample. Snapshot PCR was initially indicated on the basis of clinical suspicion, yielding a diagnostic rate of 10% (33/333 positive snapshot PCRs) between January 2019 and April 2024. In the retrospective cohort, despite being depleted in patients with ADTKD, SharkVNTyper still achieved a diagnostic yield of 78% (18/23 positive snapshot PCRs over 3512 participants), whereas in the prospective cohort, it reached 92% (12/13 positive snapshot PCRs over 825 participants). ADTKD, autosomal dominant tubulointerstitial kidney disease; ES, exome sequencing.

To further validate our methodology, we conducted a retrospective analysis from any ES data generated from adult participants (n=3512) with nephropathy of unknown origin, sequenced from January 2019 to October 2023, excluding participants who already had a positive Snapshot PCR for suspected ADTKD-MUC1 during the initial clinical screening described above in the proof-of-concept cohort.20 Indeed, the integration of SharkVNTyper reduced the processing time per sample by more than 100-fold, enabling the complete retrospective analysis of a large cohort of exomes. We then initiated a prospective use of our pipeline from October 2023 to December 2023 (n=825). Samples exhibiting insertions were further confirmed for positivity through Snapshot PCR, verifying the presence of the 27dupC pathogenic variant. The participants' flow chart is described in Figure 1.

Among the participants diagnosed with ADTKD-MUC1, those followed at Sorbonne University and La Conception Hospital constituted the cohort for phenotypic description.

The study was performed in accordance with the ethical standards of the Declaration of Helsinki. Written informed consent was obtained from all participants regarding deidentified clinical and personal patient data collection (including demographic data: age, sex, self-declared ethnicity), analysis, and publication. The study was approved by an institutional review board (Direction de la Recherche Clinique et de l’Innovation [APHP220461]) and the Ethics Board of Sorbonne Université (CER-2022-009). Phenotype data were prospectively collected from the clinician when ES had been prescribed for unknown nephropathy with suspicion of a genetic origin.

ES and Variant Analysis

DNA was extracted from peripheral blood using the QIAsymphony DSP DNA Mini Kit on a QIAsymphony instrument following the manufacturer's (QIAGEN) guidelines. Library preparation and capture was performed with Twist reagents (Human Comprehensive Exome or Human Exome 2.0 Plus Comprehensive Exome Spike-in). Sequencing was performed on the Illumina NovaSeq6000 in paired-end mode (2×150 bp reads). Raw data (bcl format) were converted to fastq format using BCL Convert. Reads were aligned to the human reference genome (UCSC Genome Browser build hg38) with Burrows–Wheeler Aligner for maximal exact matches aligner.

Statistical Analysis

All tests were two-sided, and a P value < 0.05 was considered statistically significant. Analyses were performed using R statistical software version 4.3.0. Non-normally distributed continuous variables are expressed as median (interquartile range) and compared using two-tailed paired t tests.

To assess positive predictive value (PPV), a Snapshot PCR was performed using an orthogonal method for the participants with positive SharkVNTyper.

Results

Comparison of VNTyper Performance with and without Shark Software for Short-Read ES or GS Data

From the proof-of-concept cohort (n=333), 33 participants were ultimately confirmed positive for the 27dupC variant by Snapshot PCR, accounting for 10% of the cohort. With VNTyper alone, all 33 participants with known 27dupC insertion in the tested cohort were also positive, demonstrating 100% sensitivity. All 54 negative control participants yielded negative results, indicating 100% specificity. This demonstrates the validity of using ES data instead of panel sequencing for MUC1 typing. However, the running time ranged from 6 to 12 hours for each sample, impeding systematic large cohort screening. Integration of the Shark tool with VNTyper reduced the running time per sample to an average of 7.5 minutes (±2.5 minutes SD). The use of VNTyper from those filtered files seemed to be much faster, requiring less than a minute per sample. Using SharkVNTyper, only one patient of 33 positive controls was missed (sensitivity of 97%). Nevertheless, all 54 negative controls remained negative, with 100% specificity. Thus, there is an element of trade-off for systematic MUC1 ES detection with SharkVNTyper (Table 1). To note, six of the 33 positive control participants had 27insCCCC instead of the classical 27dupC. This atypical insertion was also found with SharkVNTyper, confirming that the pipeline can identify both variants. Performance was also tested on GS to assess its feasibility across different datasets: We tested ten GS samples with one positive control. The computing time is presented in Table 2. SharkVNTyper was negative for the positive control on GS, and this is likely due to the reduced sequence coverage (approximately 30× for GS against approximately 80× for ES). We did not perform extensive testing in a large GS dataset on the capability of our pipeline to detect MUC1 with positive controls.

Table 1.

Performance of VNTyper and SharkVNTyper on exome sequencing and genome sequencing data in adult participants with suspected autosomal dominant tubulointerstitial kidney disease

Running Time Detection of MUC1 Variant Rejection of MUC1 Variant
VNTyper on ES data 6–12 h 33 of 33 positive controls 54 of 54 negative controls
SharkVNTyper on ES data 10 min 32 of 33 positive controls 54 of 54 negative controls
SharkVNTyper on GS data 55 min 0 of 1 positive control 9 of 9 negative controls

Running time is expressed in minutes, except for VNTyper whose running time shows huge variability, depending on the sample. ES, exome sequencing; GS, genome sequencing.

Table 2.

Patient characteristics from the autosomal dominant tubulointerstitial kidney disease–MUC1 cohort followed at Sorbonne University Hospitals and La Conception Hospital with available clinical data

Characteristica MUC1
Not Suspected (N=10) Suspected (N=14) Total (N=24)
Age at diagnosis, yr 51 (47–54) 43 (36–69) 51 (40–68)
Male, n (%) 8 (80) 4 (29) 12 (50)
Geographic origin, n (%)
 African 2 (20) 0 (0) 2 (8)
 European 5 (50) 13 (93) 18 (75)
 North African 3 (30) 1 (7) 4 (17)
Severity of CKD, n (%)
 Mild (stage 1 to 3) 3 (30) 4 (29) 7 (29)
 Severe (stage 4 or 5) 7 (70) 10 (71) 17 (71)
Kidney failure age, yr 46 (38–50) 54 (37–65) 50 (36–59)
Kidney cyst, n (%) 3 (30) 6 (43) 9 (38)
Arterial hypertension, n (%) 8 (80) 6 (43) 14 (58)
Early gout onset, n (%) 0 (0) 2 (14) 2 (8)
Positive family history of nephropathy, n (%) 7 (70) 14 (100) 21 (88)
Suspected mode of transmission, n (%)
 Dominant 5 (100) 14 (100) 19 (100)
 Undetermined 0 (0) 0 (0) 0 (0)
 Recessive 0 (0) 0 (0) 0 (0)
Kidney biopsy, n (%) 4 (40) 6 (43) 10 (42)
History of TMA, n (%) 1 (10) 0 (0) 1 (4)

TMA, thrombotic microangiopathy.

a

For quantitative variables, values are expressed as median (interquartile ranges).

SharkVNTyper Validation in ES Cohort

We screened 3512 participants with CKD mainly affected by unknown nephropathies. Processing every sample from the retrospective cohort took 11 days of computing in total. In comparison, using the original version of VNTyper would have taken between 2.5 and 4 years for the whole cohort. SharkVNTyper identified 23 participants considered positive for the presence of MUC1 variants. To assess PPV, we conducted orthogonal Snapshot PCR tests on these 23 participants, of whom 18 were eventually confirmed as having the well-known deleterious MUC1 27dupC variant, whereas five were tested negative, meaning that the PPV was 78%. Although negative Snapshot PCR results could potentially indicate other or yet unknown minor MUC1 genetic variants, for the sake of our analysis, we considered them false positives. In the prospective element of the study, we implemented systematic SharkVNTyper for any exome performed for CKD and identified the presence of a relevant variant in 13 prospective participants. Among these, 12 were confirmed as having a MUC1 27dupC variant using the orthogonal method, yielding a PPV of 92%. This allowed us to establish a minimum estimate of specificity within our cohort. In total, considering that any negative Snapshot PCR results with positive SharkVNTyper were true false positives, a worst-case scenario of six false positives of 4301 negative samples led to a specificity of 99.9%.

Description of Participants with ADTKD-MUC1 from ES for Unknown Kidney Diseases Cohort Followed at Sorbonne University Hospitals and La Conception Hospital

Of the 63 participants with ADTKD-MUC1, 35 from 24 families were followed at Tenon Hospital and La Conception Hospital, with descriptions of the disease available for the 24 index participants. Of these, 14 were initially clinically suspected, whereas ten participants in whom there was no clinical suspicion were retrospectively identified by SharkVNTyper test screening and confirmed by Snapshot PCR (Table 2).

Of the 24 index participants, there was an equal male-to-female ratio, with the majority being White (75%). The median age for kidney failure onset was 51 (40–68) years, consistent with previous studies.9,10 Phenotypically, 38% of participants exhibited bilateral multiple kidney cysts with only one exception. In addition, 8% had a history of early gout onset (before the age of 45 years), and 58% had associated arterial hypertension. Family history of nephropathy was reported for 88% of the participants, meaning that some of the participants had ADTKD-MUC1 even without any known familial nephropathy. The main suspected mode of inheritance was dominant, although a significant proportion of participants without suspected ADTKD-MUC1 had a family history suggestive of an undetermined or recessive mode of inheritance (Table 1). There was significant variability in disease severity among participants, even within the same family.

Despite a large proportion of participants presenting with tubulointerstitial nephropathy (10/24 participants, 42%), the initial presentation was nevertheless heterogeneous. Nine of 24 participants presented with undetermined nephropathy (38%), two with glomerular nephropathy (8%), and two with vascular nephropathy (8%), while one patient presented with cystic nephropathy (4%) (Supplemental Figure 3). Of these participants, ten (42%) underwent kidney biopsy, revealing tubular lesions in most cases (6/10 participants). A diagnosis workflow is presented in Figure 2.

Figure 2.

Figure 2

Proposed diagnostic framework for ADTKD-MUC1 with ES. (A) Initial two-stage decision tree, particularly for retrospective analysis in participants with negative ES. (B) Preferred alternative approach allowing for simultaneous MUC1 analysis alongside initial ES analysis, enabling dual diagnosis.

Atypical ADTKD-MUC1 Presentation

In our ADTKD-MUC1 cohort, some participants exhibited atypical clinical presentation. A 21-year-old woman was initially referred for familial cystic kidney disease with kidney failure. Her mother, maternal aunt, and maternal grandfather all had multicystic kidney disease. Her mother experienced early-onset kidney failure at the age of 27 years. The patient had no associated early-onset gout, but a cleft palate was reported. Despite negative results from both ES and GS, a targeted search for a pathogenic variant in MUC1 led to a molecular diagnosis. Similarly, a 51-year-old man was referred for a nephrogenomics consultation to confirm a suspicion of Alport syndrome. He presented glomerular proteinuria; stage 4 CKD; bilateral cysts; hearing impairment; and kidney biopsy consistent with rare mesangial IgA deposits, focal segmental glomerulosclerosis lesions, and nephroangiosclerosis. Electron microscopy analysis did not reveal any lesions suggestive of Alport syndrome. ES was finally found to be negative, and the retrospective use of our pipeline enabled the identification of a pathogenic variant in MUC1. This case underscores the value of systematic research in addressing challenging kidney phenotypes.

Atypical clinical presentations were also associated with a second pathogenic variant in a second gene for two families in our cohort. In the first family, a kidney biopsy of a 58-year-old woman confirmed tubulointerstitial nephropathy of unknown origin, with a single kidney cyst, leading to kidney failure at the age of 56 years. This was although her family had been initially labeled as autosomal dominant polycystic kidney disease (ADPKD) because of a brother having a pathogenic variant in PKD1 and a typical ADPKD phenotype. The patient's tubulointerstitial nephropathy without associated enlarged multicystic kidney disease prompted further investigation. This led to the identification of a pathogenic variant in MUC1, explaining her nephropathy. Her brother with ADPKD was negative for MUC1 variation, and a third brother with the ADTKD phenotype had a variant in MUC1 but not in PKD1. The last brother, who died, had ADTKD with a variant in MUC1. The patient's two sisters, who had multicystic kidney disease and kidney failure, both had the familial variant in MUC1 and PKD1 (Supplemental Figure 4). On the other hand, a 68-year-old patient presented with glomerular nephropathy characterized by microscopic hematuria, focal segmental glomerulosclerosis lesions on kidney biopsy, bilateral kidney cysts, and hearing impairment, raising suspicion of Alport syndrome. Although the presence of a heterozygous pathogenic variant in the COL4A3 gene confirmed Alport syndrome, the severity of the phenotype, with kidney failure onset at age 36 years, prompted further investigation. SharkVNTyper revealed the concomitant presence of a pathogenic variant in MUC1, confirmed by Snapshot PCR, explaining the early onset of kidney failure with both clinical features of Alport syndrome and ADTKD-MUC1 syndrome. Two other participants in the cohort had a second variant of interest in a second gene, which warranted genetic counseling. One patient had a pathogenic variant in the BRCA1 gene, leading to oncogenetic follow-up and discussion of preventive mastectomy. Another patient had a heterozygous variant in the CFTR gene. The second variants found in another gene are presented in Supplemental Table 1.

Discussion

To our knowledge, we are the first to use ES data to systematically screen for MUC1 variation. MUC1 presents a challenge for sequencing because of its VNTR in exon 2, making conventional sequencing methods demanding and difficult outside of Snapshot PCR. Moreover, the duplicated VNTR sequence varies among individuals. Within this VNTR, the pathogenic insertion of a cytosine (27dupC) results in a frameshift and a truncated protein, observed in over 90% of participants. Olinger et al. reported a 94% 27dupC pathogenic variant prevalence in participants with ADTKD-MUC1 with a molecular diagnosis in a cohort of 93 families.21 This insertion can be detected by Snapshot PCR, but it is a costly and low-throughput method that can detect only 27dupC variants. As we increasingly transition to in silico panels on the basis of ES or GS data, developing a rapid tool to detect pathogenic variations in MUC1 directly from exome data becomes crucial.

VNTyper was developed for gene panel analysis data, with greater sequencing depth on target regions compared with ES studies. This method seems sensitive, but its performance was assessed with very high-depth targeted sequencing data. We therefore sought to validate this tool using files from ES. The reduction in sequencing depth did not seem to affect the tool's diagnostic performance. We observed a slight decrease in sensitivity for the SharkVNTyper tool but without any loss of specificity. One significant challenge faced with VNTyper alone was the extended running time per sample with ES data, averaging 10 hours per sample. To mitigate this, we adopted Shark software, which significantly reduced the running time per sample by focusing only on the MUC1 region of interest. The running time per sample decreased from the previous 6–12 hours to <10 minutes on average. This computational improvement allows large patient cohort screening or even routine diagnosis in participants with undetermined nephropathy. The SharkVNTyper tool was also able to detect a variant other than 27dupC, namely 27insCCCC. This could be attributed to the fact that 27insCCCC shares the same sequence of eight consecutive cytosines. However, the ability of SharkVNTyper to detect variants in MUC1 beyond these two remains unknown and should be addressed as MUC1 datasets become more comprehensive.

SharkVNTyper is an easy-to-use complementary tool, even with ES data. We propose a diagnostic algorithm integrated into the bioinformatics processing of ES data. Concurrent with basic analysis, SharkVNTyper could be systematically included in the postsequencing pipeline as a screening test to detect the potential presence of a pathogenic variant in MUC1. If a variant is detected within the run, an alert could be raised, enabling the biologist in charge to validate the presence of this variant directly using an orthogonal approach, even before interpreting the exome data. This would, therefore, be complementary information obtained alongside classic ES data.

SharkVNTyper is an effective systematic screening test for the 27dupC variation in patients with unknown kidney disease explored through ES. Indeed, among the 4337 participants of the prospective and retrospective cohorts, 4301 were negative with our test, and of these, no participants had known ADTKD-MUC1. For participants with clinical suspicion of ADTKD-MUC1 who tested negative with SharkVNTyper, Snapshot PCR was performed to ensure the absence of 27dupC; all results were negative. This screening test should always be confirmed by a diagnostic test, such as Snapshot PCR. In participants with suspected ADTKD-MUC1, the diagnostic test should always take precedence over a negative screening test. Currently, we cannot validate the use of this method for detecting variants other than 27dupC. Further investigation with participants known to have other types of variants, confirmed by a secondary method, would be required to establish this capability.

Adding Shark to VNTyper, detection of the 27dupC variant resulted in a negative detection in one of the 33 positive controls of the proof-of-concept cohort, although it was positive using VNTyper alone. In our opinion, the most probable explanation is that this participant's variant is included in a pattern that is not yet described and, therefore, not yet included in our pattern dictionary used as a reference by Shark. To improve SharkVNTyper, we continue to add new patterns to this dictionary as we discover them.

Using a tool tailored to ES data for identifying the 27dupC variant in MUC1 demonstrates increased efficacy for systematic screening. This approach combines the advantages of ES, especially with families with complex phenotypes or multiple variants of interest, while simultaneously highlighting the presence of a pathogenic variant in MUC1. It also enables the detection of this specific variant in patients with atypical presentation or with another variant previously found with ES. Importantly, a significant proportion of our cohort of participants with ADTKD-MUC1 had undergone kidney biopsy. Efficient genetic research could potentially obviate the need for this invasive procedure, with its associated high risk of bleeding complications.

Despite the historical challenges of ES associated with the VNTR structure of the MUC1 gene, we demonstrate for the first time, the feasibility of identifying the 27dupC variation in routine analysis. We showed that SharkVNTyper is an effective systematic screening tool for detecting the 27dupC variation from short-read next-generation sequencing exome data. It shows promise for routine clinical molecular diagnosis, offering the potential for uncovering previously unidentified molecular diagnoses in adult participants with CKD. Our findings highlight the utility of bioinformatics tools in identifying pathogenic variants in challenging genetic contexts, bypassing alignment to a reference genome. However, the detection of minor MUC1 variants may still require additional exploration, particularly in families with such suspected variants.

Supplementary Material

jasn-36-256-s001.pdf (1.4MB, pdf)
jasn-36-256-s002.pdf (531.8KB, pdf)

Acknowledgments

We would like to thank the Genetics Department of Pitie-Salpêtrière Hospital (Anne-Sophie Lebre, Jean-Madeleine De Sainte Agathe, Julie Bogoin) for their technical support. We would also like to thank Edouard Henrion, Boris Lipinski, Fanny Ponce, Vincent Gassend, and Barbara Adde from the Bioinformatics Department of Biomnis Laboratory for their technical support.

Footnotes

See related letter to the editor, “Using VNtyper from Whole Exome Sequencing Data to Detect Pathogenic Variants in the MUC1 Gene,“ on pages 327–328 and reply, “Authors’ Reply: Using VNtyper from Whole Exome Sequencing Data to Detect Pathogenic Variants in the MUC1 Gene,” on pages 329–330.

Disclosures

Disclosure forms, as provided by each author, are available with the online version of the article at http://links.lww.com/JSN/E864.

Funding

I. Bensouna: Assistance Publique - Hôpitaux de Paris.

Author Contributions

Conceptualization: Laurent Mesnard.

Data curation: Ilias Bensouna.

Formal analysis: Ilias Bensouna.

Investigation: Ilias Bensouna, Marine Dancer, Pascale Hilbert, Laure Raymond, Hugues Richard, Thomas Robert, Xavier Vanhoye.

Methodology: Ilias Bensouna, Pierre Delaugere, Laurent Mesnard, Hugues Richard.

Project administration: Laurent Mesnard.

Supervision: Laurent Mesnard.

Validation: Laurent Mesnard.

Writing – original draft: Ilias Bensouna.

Writing – review & editing: Marine Dancer, Laurent Mesnard, Laure Raymond, Hugues Richard, Thomas Robert, Xavier Vanhoye.

Data Sharing Statement

Partial restrictions to the data and/or materials apply. Restrictions related with genetic information. Selected further anonymous data can be shared by principal investigator (laurent.mesnard@aphp.fr) on demand.

Supplemental Material

This article contains the following supplemental material online at http://links.lww.com/JSN/E863.

Supplemental Figure 1. Schematic of VNTyper and SharkVNTyper pipeline integration workflow.

Supplemental Figure 2. Snapshot PCR for detection of 27dupC.

Supplemental Figure 3. Initial presentation of participants with ADTKD-MUC1.

Supplemental Figure 4. Pedigree of a family with combined variants in MUC1 and PKD1, demonstrating the utility of systematic screening for MUC1.

Supplemental Table 1. Second hit or variant found in another gene for participants with ADTKD-MUC1 in the ES-UKD cohort.

References

  • 1.Mallett AJ McCarthy HJ Ho G, et al. Massively parallel sequencing and targeted exomes in familial kidney disease can diagnose underlying genetic disorders. Kidney Int. 2017;92(6):1493–1506. doi: 10.1016/j.kint.2017.06.013 [DOI] [PubMed] [Google Scholar]
  • 2.Marsili L, Duque KR, Bode RL, Kauffman MA, Espay AJ. Uncovering essential tremor genetics: the promise of long-read sequencing. Front Neurol. 2022;13:821189. doi: 10.3389/fneur.2022.821189 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Eckardt KU Alper SL Antignac C, et al. Autosomal dominant tubulointerstitial kidney disease: diagnosis, classification, and management—a KDIGO consensus report. Kidney Int. 2015;88(4):676–683. doi: 10.1038/ki.2015.28 [DOI] [PubMed] [Google Scholar]
  • 4.Devuyst O Olinger E Weber S, et al. Autosomal dominant tubulointerstitial kidney disease. Nat Rev Dis Primers. 2019;5(1):60. doi: 10.1038/s41572-019-0109-9 [DOI] [PubMed] [Google Scholar]
  • 5.Ayasreh N Bullich G Miquel R, et al. Autosomal dominant tubulointerstitial kidney disease: clinical presentation of patients with ADTKD-UMOD and ADTKD-MUC1. Am J Kidney Dis. 2018;72(3):411–418. doi: 10.1053/j.ajkd.2018.03.019 [DOI] [PubMed] [Google Scholar]
  • 6.Okada E Morisada N Horinouchi T, et al. Detecting MUC1 variants in patients clinicopathologically diagnosed with having autosomal dominant tubulointerstitial kidney disease. Kidney Int Rep. 2022;7(4):857–866. doi: 10.1016/j.ekir.2021.12.037 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7.Živná M Kidd K Přistoupilová A, et al. Noninvasive immunohistochemical diagnosis and novel MUC1 mutations causing autosomal dominant tubulointerstitial kidney disease. J Am Soc Nephrol. 2018;29(9):2418–2431. doi: 10.1681/ASN.2018020180 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 8.Yamamoto S Kaimori JY Yoshimura T, et al. Analysis of an ADTKD family with a novel frameshift mutation in MUC1 reveals characteristic features of mutant MUC1 protein. Nephrol Dial Transplant. 2017;32(12):2010–2017. doi: 10.1093/ndt/gfx083 [DOI] [PubMed] [Google Scholar]
  • 9.Kirby A Gnirke A Jaffe DB, et al. Mutations causing medullary cystic kidney disease type 1 lie in a large VNTR in MUC1 missed by massively parallel sequencing. Nat Genet. 2013;45(3):299–303. doi: 10.1038/ng.2543 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 10.Wenzel A Altmueller J Ekici AB, et al. Single molecule real time sequencing in ADTKD-MUC1 allows complete assembly of the VNTR and exact positioning of causative mutations. Sci Rep. 2018;8(1):4170. doi: 10.1038/s41598-018-22428-0 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.De Haan A Van Eerde AM Eijgelsheim M, et al. Novel MUC1 variant identified by massively parallel sequencing explains interstitial kidney disease in a large Dutch family. Kidney Int. 2023;103(5):986–989. doi: 10.1016/j.kint.2023.02.021 [DOI] [PubMed] [Google Scholar]
  • 12.Ekici AB Hackenbeck T Morinière V, et al. Renal fibrosis is the common feature of autosomal dominant tubulointerstitial kidney diseases caused by mutations in mucin 1 or uromodulin. Kidney Int. 2014;86(3):589–599. doi: 10.1038/ki.2014.72 [DOI] [PubMed] [Google Scholar]
  • 13.Blumenstiel B DeFelice M Birsoy O, et al. Development and validation of a mass spectrometry–based assay for the molecular diagnosis of mucin-1 kidney disease. J Mol Diagn. 2016;18(4):566–571. doi: 10.1016/j.jmoldx.2016.03.003 [DOI] [PubMed] [Google Scholar]
  • 14.Gale DP, Kleta R. MUC1 makes me miserable. J Am Soc Nephrol. 2018;29(9):2257–2258. doi: 10.1681/ASN.2018070742 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 15.Fages V Bourre F Larrue R, et al. Description of a new simple and cost-effective molecular testing that could simplify MUC1 variant detection. Kidney Int Rep. 2024;9(5):1451–1457. doi: 10.1016/j.ekir.2024.01.058 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 16.Bonizzoni P Ceccato T Vedova GD, et al. Shark: fishing in a sample to discard useless RNA-seq reads. Bioinformatics. 2019. doi: 10.1101/836130 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 17.Denti L Pirola Y Previtali M, et al. Shark: fishing relevant reads in an RNA-seq sample. Bioinformatics. 2021;37(4):464–472. doi: 10.1093/bioinformatics/btaa779. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 18.Saei H Morinière V Heidet L, et al. VNtyper enables accurate alignment-free genotyping of MUC1 coding VNTR using short-read sequencing data in autosomal dominant tubulointerstitial kidney disease. iScience. 2023;26(7):107171. doi: 10.1016/j.isci.2023.107171 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Audano PA, Ravishankar S, Vannberg FO. Mapping-free variant calling using haplotype reconstruction from k-mer frequencies. Bioinformatics. 2018;34(10):1659–1665. doi: 10.1093/bioinformatics/btx753 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20.Doreille A Lombardi Y Dancer M, et al. Exome-first strategy in adult patients with CKD: a cohort study. Kidney Int Rep. 2023;8(3):596–605. doi: 10.1016/j.ekir.2022.12.007 [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 21.Olinger E Hofmann P Kidd K, et al. Clinical and genetic spectra of autosomal dominant tubulointerstitial kidney disease due to mutations in UMOD and MUC1. Kidney Int. 2020;98(3):717–731. doi: 10.1016/j.kint.2020.04.038 [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

Partial restrictions to the data and/or materials apply. Restrictions related with genetic information. Selected further anonymous data can be shared by principal investigator (laurent.mesnard@aphp.fr) on demand.


Articles from Journal of the American Society of Nephrology : JASN are provided here courtesy of American Society of Nephrology

RESOURCES