VNtyper enables accurate alignment-free genotyping of MUC1 coding VNTR using short-read sequencing data in autosomal dominant tubulointerstitial kidney disease

•Detecting pathogenic variants in VNTRs is challenging due to poor read mappability•VNtyper detects VNTR the MUC1 gene using Kestrel algorithm•VNtyper improves ADTKD-MUC1 diagnosis by accurate genotyping of coding VNTR•Integration VNtyper with SRS panel testing identified new overlooked patients The human genome comprises approximately 3% tandem repeats variable length (VNTR), a few which have been linked rare diseases. Autosomal dominant tubulointerstitial kidney disease—MUC1 (ADTKD-MUC1) caused specific frameshift gene. Calling from short-read sequencing (SRS) mappability. We developed computational pipeline, VNtyper, for reliable detection and demonstrated its clinical utility two distinct cohorts: (1) historical cohort including 108 families ADTKD (2) replication naive comprising 2,910 previously tested on genes involved monogenic renal In all cases known carry were re-identified, 25bp-frameshift insertion an additional mislaid family was detected. cohort, we discovered validated 30 patients. disease (ADTKD, OMIM: 174000) hereditary condition characterized progressive fibrosis or without tubular dilation atrophy, eventually leading end-stage failure.1Eckardt K.-U. Alper S.L. Antignac C. Bleyer A.J. Chauveau D. Dahan K. Deltas Hosking A. Kmoch S. Rampoldi L. et al.Autosomal disease: diagnosis, classification, management—A KDIGO consensus report.Kidney Int. 2015; 88: 676-683https://doi.org/10.1038/ki.2015.28Abstract Full Text PDF PubMed Scopus (216) Google Scholar affected individuals present mild-to-negative proteinuria bland urinary sediment abnormalities, normal size.2Devuyst O. Olinger E. Weber Eckardt disease.Nat. Rev. Dis. Primers. 2019; 5: 60https://doi.org/10.1038/s41572-019-0109-9Crossref (61) A positive history commonly reported this disease. Molecular genetics play key role diagnosing classifying given non-specific presentation. Pathogenic different genes, MUC1, UMOD, HNF1B, REN, SEC61A1, are be responsible ADTKD.1Eckardt Scholar,3Bolar N.A. Golzio Živná M. Hayot G. Van Hemelrijk Schepers Vandeweyer Hoischen Huyghe J.R. Raes al.Heterozygous Loss-of-Function SEC61A1 Mutations Cause Autosomal-Dominant Tubulo-Interstitial Glomerulocystic Kidney Disease Anemia.Am. J. Hum. Genet. 2016; 99: 174-187https://doi.org/10.1016/j.ajhg.2016.05.028Abstract (100) Scholar,4Ayasreh N. Bullich Miquel R. Furlano Ruiz P. Lorente Valero García-González M.A. Arhda Garin I. Dominant Tubulointerstitial Disease: Clinical Presentation Patients With ADTKD-UMOD ADTKD-MUC1.Am. 2018; 72: 411-418https://doi.org/10.1053/j.ajkd.2018.03.019Abstract (31) absence mutation does not rule out as other loci remain identified.2Devuyst recent advancements high-throughput technology pan-genome graph pipelines improved diagnostic rate Mendelian multigenic complex diseases.5Markello Huang Rodriguez Carroll Chang P.-C. Eizenga Markello T. Haussler Paten B. complete pedigree-based workflow candidate variant analysis.Genome Res. 2022; 32: 893-903https://doi.org/10.1101/gr.276387.121Crossref (0) Scholar,6Hickey Heller Monlong Sibbesen J.A. Sirén Dawson E.T. Garrison Novak A.M. Genotyping structural pangenome graphs vg toolkit.Genome Biol. 2020; 21: 35https://doi.org/10.1186/s13059-020-1941-7Crossref (79) Scholar,7Eggertsson H.P. Kristmundsdottir Beyter Jonsson H. Skuladottir Hardarson M.T. Gudbjartsson D.F. Stefansson Halldorsson B.V. Melsted GraphTyper2 enables population-scale variation graphs.Nat. Commun. 10: 5402https://doi.org/10.1038/s41467-019-13341-9Crossref (56) However, some challenges accurately detecting structures. For instance, number repeat (VNTR) region (encoding mucin-1 protein) (OMIM: real issue. Mucin-1 transmembrane glycoprotein widely expressed segments nephron kidney, thick ascending limb loop Henle collecting duct.8Patton Gendler S.J. Spicer A.P. epithelial mucin, milk, mammary gland tissues.Biochim. Biophys. Acta. 1995; 1241: 407-423https://doi.org/10.1016/0304-4157(95)00014-3Crossref (250) maps chromosome 1 exhibits large exon 2, embraces combination at least 34 motifs 60-mer that 20 125 times random patterns each allele. Each motif encodes highly glycosylated amino acids (aa) block (see Figure 1). initial five (1-2-3-4(or 4p)-5(or 5C)) final four (6(or 6p)-7-8-9) likewise unique sequence. Any these except 7 could found conventional pipelines. Between 5(or 5C) 6(or 6p), any remaining 22 conceivable.9Kirby Gnirke Jaffe D.B. Barešová V. Pochet Blumenstiel Ye Aird Stevens Robinson J.T. al.Mutations causing medullary cystic type lie missed massively parallel sequencing.Nat. 2013; 45: 299-303https://doi.org/10.1038/ng.2543Crossref (193) Scholar,10Ekici A.B. Hackenbeck Morinière Pannes Buettner Uebe Janka Wiesener Hermann Grupp al.Renal common feature autosomal diseases mutations mucin uromodulin.Kidney 2014; 86: 589-599https://doi.org/10.1038/ki.2014.72Abstract (70) Detailed information regarding published sequences their orientation highlighted S1. Therefore, complexity hotspot, finding has proven challenging. Indeed, single nucleotide C stretch seven recurrent VNTR.9Kirby Scholar,11Olinger Hofmann Kidd Dufour Belge Schaeffer Kipp Bonny Demoulin al.Clinical genetic spectra UMOD MUC1.Kidney 98: 717-731https://doi.org/10.1016/j.kint.2020.04.038Abstract (40) It established one 3n+1 bases, deletion 3n+2 nucleotides (all same frameshift), associated ADTKD-MUC19Kirby Scholar,12Yamamoto Kaimori J.-Y. Yoshimura Namba Imai Kobayashi Imamura Ichimaru Kato Nakaya al.Analysis novel reveals characteristic features mutant protein.Nephrol. Dial. Transplant. 2017; 2010-2017https://doi.org/10.1093/ndt/gfx083Crossref (21) Scholar,13Okada Morisada Horinouchi Fujii Tsuji Miura Katori Kitagawa Morozumi Toriyama al.Detecting Variants Clinicopathologically Diagnosed Having Disease.Kidney Rep. 7: 857-866https://doi.org/10.1016/j.ekir.2021.12.037Abstract production toxic neo-protein.14Dvela-Levitt Kost-Alimova Emani Kohnert Thompson Sidhom E.-H. Rivadeneira Sahakian Roignot Papagregoriou al.Small Molecule Targets TMED9 Promotes Lysosomal Degradation Reverse Proteinopathy.Cell. 178: 521-535.e23https://doi.org/10.1016/j.cell.2019.07.002Abstract (82) cases, creates frame it alters translational pattern aa block. reaches stop codon 85 acid after last motif. result neo-protein containing many copies sequence lacking C-terminal domain, accumulates cytoplasm activates unfolded protein response14Dvela-Levitt (Figure 1B). Former studies used alternative techniques, probe extension assays followed mass spectrometry9Kirby Scholar,15Blumenstiel DeFelice Birsoy Carter T.A. Rehm H.L. Ronco al.Development Validation Mass Spectrometry–Based Assay Diagnosis Disease.J. Mol. Diagn. 18: 566-571https://doi.org/10.1016/j.jmoldx.2016.03.003Abstract (23) ddNTP (dideoxynucleotide) Mwo1 digestion (SNaPshot),10Ekici targeted analysis use Illumina system,16Živná Přistoupilová Harden Conlon Lavin Connaughton D.M. al.Noninvasive Immunohistochemical Novel Causing Am. Soc. Nephrol. 29: 2418-2431https://doi.org/10.1681/ASN.2018020180Crossref (29) long-read (SMRT: molecule real-time sequencing), find causal variations; however, approaches either technically demanding expensive.13Okada Scholar,17Mantere Kersten Long-Read Sequencing Emerging Medical Genetics.Front. 426https://doi.org/10.3389/fgene.2019.00426Crossref (189) Scholar,18Wenzel Altmueller Ekici Popp Stueber Thiele Staubach Salido Nuernberg al.Single time allows assembly exact positioning causative mutations.Sci. 8: 4170https://doi.org/10.1038/s41598-018-22428-0Crossref Using SNaPshot approach,10Ekici only (dupC) investigated 1C). Other variations than dupC affecting restriction site non-conserved experimentally utilization technology. method named code-adVNTR19Park Bakhtiari Bafna Detecting regions code-adVNTR.iScience. 25104785https://doi.org/10.1016/j.isci.2022.104785Abstract Park al. recently indels VNTRs. small three dupC-positive 271 SNaPshot-negative individuals.19Park Scholar,20Popp Knaup K.X. Schneider Meiselbach Schiffer al.Prevalence German Chronic study.Eur. 30: 1413-1422https://doi.org/10.1038/s41431-022-01177-9Crossref Here pipeline algorithm based k-mer approach21Audano P.A. Ravishankar Vannberg F.O. Mapping-free calling haplotype reconstruction frequencies.Bioinformatics. 34: 1659-1665https://doi.org/10.1093/bioinformatics/btx753Crossref (17) call data. Our first goal evaluate our well-described before applying almost 3,000 (renome cohort) identify MUC1-positive approach study index (n = 108) relatives (symptomatic asymptomatic individuals) within 2B ). allowed detect 31 (from families) so classified “MUC1 positive”. Beside 62/63 symptomatic displayed assay. relative (NTIH_140) atypical signal (very high undigested motifs) positive” family, confirmed bearing risk individual linkage concluded uninterpretable SNaPshot. All 24 negative case, abnormal (strong 7C + shorter PCR product) led identification 5bp MwoI enzyme. event subcloning product 4E This also detected family. Altogether, investigation 32 MUC1-ADTKD 1/97 event.Figure 3Schematic overview pipelineShow full captionThe (SRS)-based deleterious shown.View Large Image ViewerDownload Hi-res image Download (PPT)Figure 4Modified explaining events siteShow caption(A–C) Examples migration shown. (B) case (REN6122000742), 8C signals present. modified protocol direct feasible, confirms results. (C) very strong observed SNaPshots if failed sites (due variants). aid potentially (NTI6120004559) polymorphic (HYP2316, NTI195).(D E) patient clinics ADTKD, product.View (PPT) (A–C) NTI195). (D product. 76 (109 symptomatic/116 individuals), assay negative. Such labeled families” 2B). applied VNtyper-Kestrel 237 re-identify determine members (considered false negatives positives VNtyper). successfully re-identified 97) 2C). described (NTIH_140), dupC, confirming segregation MUC1-negative group (76 cases; 116 six members, 25bp VNTR, investigation. absent relatives. results showed applicability cohort. but cannot method. concordant true-positive true-negative independent cluster discordant 5A – red dots below dotted line). While them SNaPshot, statistically lower depth score relevant (respective mean ± SD: 0.0025 0.0005 vs. 0.013 0.007, p 2.10−15). calculated sensitivity (ability true positives) 100% specificity exclude negatives) 76.11% VNtyper-Kestrel. To best depth-score threshold (sensitivity) while excluding aforementioned (specificity), points 5C). determined 0.00469 optimal because both point had (red 5A). Since there no AltDepth 20, decided further zone renome caution labeling <20 low confidence. Also, 10% above [0.00469–0.00515], label confidence analysis. Taken together, us set up parameters filter false-positive then compared code-adVNTR method, included tool. re-called SNaPshot-positive (NTIH155, NTIH153, NTIH377, NTI383, NTI188). contrast, called since evaluates likelihood calls owing errors filters value >0.001 obtained statistical test. Another discrepancy between VNtyper-code-adVNTR insertion. 23bp supporting reads larger coverage (p 0), By extracting indicated fastq files, predicted Kestrel. No pattern. shown Table S3. report herein able detect. extensive apply data undiagnosed estimate VNtyper. studied 6C displays distribution groups according diagnosis. unrelated (REN_6122GM002870 NTI_6121GM005428) 4bp duplication 4, respectively, (NTI1129, NTI_6121GM003097, NPH1908593) early conserved 5. IGV22Robinson Thorvaldsdóttir Winckler W. Guttman Lander E.S. Getz Mesirov J.P. Integrative genomics viewer.Nat. Biotechnol. 2011; 24-26https://doi.org/10.1038/nbt.1754Crossref (8013) visualization bam files S5. MUC1-related already available Among Kestrel-VNtyper 2,208 702 (0.00469) 662 individuals, samples thus filtered positives. 40 patients, (Figures 6A–6E). ensure validity results, categorized into groups. contained 0.00515 (the +10%) considered second [0.00469–0.00515]. separation differentiate high-confidence those 6A 6B - gray band), requires validation assays. threshold, 28 12 low-confidence and/or threshold. result, 25 variant, (an NTI_6120004559), 6E). These duplications well insertions 5, located NGS, S5). As expected, motif, around half total depth. analyses conducted. 25/25 cases. (NTI_6120004559), dupA 4C). label, (CC) 3, where ratio 50% expected. exhibited expected therefore Regarding 11 2/11 validated: HYP4100 (by SNaPshot) NTI1179 (with NGS). AltDepth. 0.00437. particular NGS (AltDepth equal 537 yielded 0.00612). Two NTI1168 PK432, (performed twice, confidence) SNaPshots. personal compat