PT  - JOURNAL ARTICLE
AU  - Sara Althari
AU  - Laeya A. Najmi
AU  - Amanda J. Bennett
AU  - Ingvild Aukrust
AU  - Jana K. Rundle
AU  - Kevin Colclough
AU  - Janne Molnes
AU  - Alba Kaci
AU  - Sameena Nawaz
AU  - Timme van der Lugt
AU  - Neelam Hassanali
AU  - Anders Molven
AU  - Sian Ellard
AU  - Mark I. McCarthy
AU  - Lise Bjørkhaug
AU  - Pål Rasmus Njølstad
AU  - Anna L. Gloyn
TI  - Unsupervised clustering of missense variants in the <em>HNF1A</em> gene using multidimensional functional data aids clinical interpretation
AID  - 10.1101/19010900
DP  - 2019 Jan 01
TA  - medRxiv
PG  - 19010900
4099  - http://medrxiv.org/content/early/2019/11/02/19010900.short
4100  - http://medrxiv.org/content/early/2019/11/02/19010900.full
AB  - Background Exome sequencing in diabetes presents a diagnostic challenge as depending on frequency, functional impact and genomic and environmental contexts, HNF1A variants can cause Maturity-onset Diabetes of the Young (MODY), increase type 2 diabetes risk, or be benign. A correct diagnosis matters as it informs on treatment, progression, and family risk. We describe a multi-dimensional functional dataset of 73 HNF1A missense variants identified in exomes of 12,940 individuals. Our aim was to develop an analytical framework for stratifying variants along the HNF1A phenotypic continuum to facilitate diagnostic interpretation.Methods HNF1A variant function was determined by 4 different molecular assays. Structure of the multi-dimensional dataset was explored using principal component analysis, k-means, and hierarchical clustering. Weights for tissue-specific isoform expression and functional domain were integrated. Functionally annotated variant subgroups were used to re-evaluate genetic diagnoses in national MODY diagnostic registries.Findings HNF1A variants demonstrated a range of behaviours across the assays. The structure of the multi-parametric data was shaped primarily by transactivation. Using unsupervised learning methods, we obtained high-resolution functional clusters of the variants which separated known causal MODY variants from benign and type 2 diabetes risk variants and led to reclassification of 4% and 9% of HNF1A variants identified in the UK and Norway MODY diagnostic registries, respectively.Interpretation Our proof-of-principle analyses facilitated informative stratification of HNF1A variants along the continuum, allowing improved evaluation of clinical significance, management and precision medicine in diabetes clinics. Transcriptional activity appears a superior readout supporting pursuit of transactivation-centric experimental designs for high-throughput functional screens.Funding Wellcome Trust, National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC), European Research Council, Norwegian Research Council, Stiftelsen Kristian Gerhard Jebsen, Western Norway Regional Health Authority, Novo Nordisk Fonden, Royal Norwegian Diabetes Foundation.Evidence before the study Molecular characterisation pipelines for studying the function of transcription factors consist primarily of in vitro cellular assays which interrogate transcriptional activity, protein abundance, localisation of the transcription factor to the nucleus, and binding to relevant DNA recognition sequences. The experimental techniques used to explore these mechanisms in vitro vary in robustness and reliability. There exist a wide variety of reported functional consequences of HNF1A variants in the literature, a gene causing the most common form of Maturity-onset Diabetes of the Young (HNF1A-MODY). The standard approach for analysing multi-tiered functional datasets has been to evaluate each functional parameter independently. Data from functional characterisation efforts of the HNF-1A protein encoded by the HNF1A gene, support that the degree of HNF-1A disruption tends to correlate positively with phenotypic severity: MODY-causing protein-altering variants impair HNF-1A transcriptional activity more severely (≤30% vs. wild-type) than HNF1A variants associated with increased risk for developing type 2 diabetes in population-specific contexts (40-60% vs. wild-type). Rare variants which demonstrated intermediate function (between MODY-casual and wild-type) in transactivation and nuclear localisation assays were shown to be associated with a 6-fold increase in type 2 diabetes predisposition.Added value of this study We have developed a proof-of-principle analytical framework for robust and unbiased variant stratification using multi-dimensional functional follow-up data from a large number of exome-identified missense variants in HNF1A. Through our analytical approach we were able to perform a comprehensive assessment of molecular function by utilising data from as many mechanistic dimensions as possible, avoiding arbitrarily determined cut-offs based on 1D functional data. Our method facilitated informative spatial organization of variants along the HNF1A molecular-phenotypic spectrum and an exploration of the contributions of each in vitro molecular mechanism on meaningful functional, and therefore clinical, stratification. Further, we were able to perform sensitive mapping of variant effects on molecular function with phenotypic outcome using clinical and genetic data from national MODY diagnostic registries of UK and Norway. This effort allowed us to annotate functional clusters with clinical knowledge and identify discordant classifications between functional genotype and clinical phenotype.Implications of all the available evidence Our novel approach towards analysing large functional datasets enables sensitive variant-phenotype mapping and multi-layered variant annotation. It also assists in prioritisation of functional elements and signatures for Multiplexed Assays of Variant Effects (MAVEs) whilst they largely remain limited to a single functional readout. Indeed, comprehensively annotated HNF1A variant clusters can aid in the interpretation and clinical classification of variants, and can also be utilised to calibrate supervised variant classification models built with high-throughput-derived experimental data.Competing Interest StatementThe views expressed in this article are those of the authors and not necessarily those of the NHS, the NIHR, or the Department of Health. McCarthy has served on advisory panels for Pfizer, Novo Nordisk and Zoe Global, received honoraria from Merck, Pfizer, NovoNordisk and Eli Lilly, and research funding from Abbvie, Astra Zeneca, Boehringer Ingelheim, Eli Lilly, Janssen, Merck, Novo Nordisk, Pfizer, Roche, Sanofi Aventis, Servier and Takeda. As of June 2019, MMcC is an employee of Genentech, and a holder of Roche stock. Dr. Gloyn reports grants from Wellcome Trust, grants from NIHR Oxford Biomedical Research Centre, grants from Horizon 2020, grants from NIDDK, grants from MRC, during the conduct of the study; personal fees from Novo Nordisk, personal fees from Merck, outside the submitted work.Clinical TrialNot a clinical trialFunding StatementALG is a Wellcome Senior Fellow in Basic Biomedical Science. SE and MIM are Wellcome Senior Investigators. This work was funded in Oxford by the Wellcome (095101 [ALG], 200837 [ALG], 098381 [MIM], 106130 [ALG, MIM], 203141 (ALG, MIM], 203141 [MIM]), Medical Research Council (MR/L020149/1) [MIM, ALG], European Union Horizon 2020 Programme (T2D Systems) [ALG], and NIH (U01-DK105535; U01-DK085545) [MIM, ALG]. The research was funded by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre (BRC) [ALG, MIM]. This work was funded in Bergen by grants from the European Research Council (#293574 [PRN]), the Norwegian Research Council (#240413/F20 [PRN]), Stiftelsen Kristian Gerhard Jebsen [PRN], the Novo Nordisk Fonden [PRN], the Western Norway Health Authorities, (#911745 [PRN]), and the University of Bergen [IA, LN, PRN].Author DeclarationsAll relevant ethical guidelines have been followed; any necessary IRB and/or ethics committee approvals have been obtained and details of the IRB/oversight body are included in the manuscript.YesAll necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived.YesI understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance).YesI have followed all appropriate research reporting guidelines and uploaded the relevant EQUATOR Network research reporting checklist(s) and other pertinent material as supplementary files, if applicable.YesSummary statistics is available for all data in the article.