Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation
<p>Maturity-Onset Diabetes of the Young (MODY) is a monogenic form of hyperglycaemia that is both genetically and clinically heterogeneous. The majority of MODY cases (80%) are caused by heterozygous damaging alleles in clinically actionable genes (<em>HNF1A, HNF4A and GCK</em>). F...
| Autor principal: | |
|---|---|
| Altres autors: | |
| Format: | Thesis |
| Publicat: |
2018
|
| _version_ | 1826286064421044224 |
|---|---|
| author | Althari, S |
| author2 | Gloyn, A |
| author_facet | Gloyn, A Althari, S |
| author_sort | Althari, S |
| collection | OXFORD |
| description | <p>Maturity-Onset Diabetes of the Young (MODY) is a monogenic form of hyperglycaemia that is both genetically and clinically heterogeneous. The majority of MODY cases (80%) are caused by heterozygous damaging alleles in clinically actionable genes (<em>HNF1A, HNF4A and GCK</em>). For carriers of these alleles, an accurate molecular diagnosis informs prognosis, management and treatment strategies, and is contingent upon understanding the functional consequences of known and novel variants. High-fidelity molecular assays, which enable comprehensive and contextual functional characterisation, at scale and speed, are needed to address the interpretive challenge. Herein, such efforts, with varying degrees of throughput, are described in the context of missense variants in <em>HNF1A</em> and <em>GCK</em>.</p> <p>Using multi-dimensional functional data from 73 <em>HNF1A</em> missense alleles identified through exome-sequencing of &Tilde;13K T2D cases and controls, and a combination of unsupervised learning methods, I produced high resolution variant clusters along the HNF1A phenotypic spectrum. To demonstrate robustness of the approach, I used the functionally annotated allele subgroups to re-evaluate molecular diagnoses in a national MODY diagnostic registry (Exeter, UK). This resulted in confident reclassification of n=7 alleles from pathogenic/likely pathogenic to VUS/likely benign in the clinical database. I also established a systematic pipeline for building a transactivation-centric Multiplexed Assay of Variant Effects (MAVE) to characterize an <em>HNF1A</em> saturation mutagenesis library. Using TM4SF4 as an endogenous functional readout in flow cytometric analyses, I was able to segregate a series of <em>HNF1A</em> alleles (n=8) based on transactivation-dependent distributions of TM4SF4-positive HepG2 cell populations: 2-8% from MODY-causal loss-of-function alleles, 30% from a low frequency allele associated with T2D risk, and 60-70% from neutral alleles and wild-type. This discriminatory capacity suggests the assay is sufficiently robust to score a pooled HNF1A variant library using TM4SF4 expression as a molecular signature. For variants in GCK, I characterised the function of two previously unannotated missense variants (hypomorphic homozygous p.E157K/p.E157K and heterozygous p.F133L/N) identified through genetic testing to aid in their clinical interpretation. Both showed catalytic and/or stability defects in <em>in vitro</em> kinetic assays consistent with GCK-MODY (p.E157K relative activity index = 0.15; p.F133L relative stability index = 0.53). </p> <p>Together, these functional genomic annotation efforts help resolve the complex relationship between genotype, molecular dysfunction and clinical phenotype for alleles in <em>HNF1A</em> and <em>GCK</em> and assist in medical diagnostic interpretation.</p> |
| first_indexed | 2024-03-07T01:38:14Z |
| format | Thesis |
| id | oxford-uuid:95ed8d80-358e-4359-9dfe-cc50b8c96d92 |
| institution | University of Oxford |
| last_indexed | 2024-03-07T01:38:14Z |
| publishDate | 2018 |
| record_format | dspace |
| spelling | oxford-uuid:95ed8d80-358e-4359-9dfe-cc50b8c96d922022-03-26T23:49:38ZFunctional annotation of variants in monogenic diabetes genes to support diagnostic interpretationThesishttp://purl.org/coar/resource_type/c_db06uuid:95ed8d80-358e-4359-9dfe-cc50b8c96d92ORA Deposit2018Althari, SGloyn, AMcCarthy, M<p>Maturity-Onset Diabetes of the Young (MODY) is a monogenic form of hyperglycaemia that is both genetically and clinically heterogeneous. The majority of MODY cases (80%) are caused by heterozygous damaging alleles in clinically actionable genes (<em>HNF1A, HNF4A and GCK</em>). For carriers of these alleles, an accurate molecular diagnosis informs prognosis, management and treatment strategies, and is contingent upon understanding the functional consequences of known and novel variants. High-fidelity molecular assays, which enable comprehensive and contextual functional characterisation, at scale and speed, are needed to address the interpretive challenge. Herein, such efforts, with varying degrees of throughput, are described in the context of missense variants in <em>HNF1A</em> and <em>GCK</em>.</p> <p>Using multi-dimensional functional data from 73 <em>HNF1A</em> missense alleles identified through exome-sequencing of &Tilde;13K T2D cases and controls, and a combination of unsupervised learning methods, I produced high resolution variant clusters along the HNF1A phenotypic spectrum. To demonstrate robustness of the approach, I used the functionally annotated allele subgroups to re-evaluate molecular diagnoses in a national MODY diagnostic registry (Exeter, UK). This resulted in confident reclassification of n=7 alleles from pathogenic/likely pathogenic to VUS/likely benign in the clinical database. I also established a systematic pipeline for building a transactivation-centric Multiplexed Assay of Variant Effects (MAVE) to characterize an <em>HNF1A</em> saturation mutagenesis library. Using TM4SF4 as an endogenous functional readout in flow cytometric analyses, I was able to segregate a series of <em>HNF1A</em> alleles (n=8) based on transactivation-dependent distributions of TM4SF4-positive HepG2 cell populations: 2-8% from MODY-causal loss-of-function alleles, 30% from a low frequency allele associated with T2D risk, and 60-70% from neutral alleles and wild-type. This discriminatory capacity suggests the assay is sufficiently robust to score a pooled HNF1A variant library using TM4SF4 expression as a molecular signature. For variants in GCK, I characterised the function of two previously unannotated missense variants (hypomorphic homozygous p.E157K/p.E157K and heterozygous p.F133L/N) identified through genetic testing to aid in their clinical interpretation. Both showed catalytic and/or stability defects in <em>in vitro</em> kinetic assays consistent with GCK-MODY (p.E157K relative activity index = 0.15; p.F133L relative stability index = 0.53). </p> <p>Together, these functional genomic annotation efforts help resolve the complex relationship between genotype, molecular dysfunction and clinical phenotype for alleles in <em>HNF1A</em> and <em>GCK</em> and assist in medical diagnostic interpretation.</p> |
| spellingShingle | Althari, S Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title | Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title_full | Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title_fullStr | Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title_full_unstemmed | Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title_short | Functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| title_sort | functional annotation of variants in monogenic diabetes genes to support diagnostic interpretation |
| work_keys_str_mv | AT altharis functionalannotationofvariantsinmonogenicdiabetesgenestosupportdiagnosticinterpretation |