Diabetes Care 1

Shivani Misra,1 Neelam Hassanali,2 Homozygous Hypomorphic Amanda J. Bennett,2 Agata Juszczak,2 Richard Caswell,3 Kevin Colclough,3 HNF1A Alleles Are a Novel Cause Jonathan Valabhji,4 Sian Ellard,3 of Young-Onset and Nicholas S. Oliver,1,4 and Anna L. Gloyn2,5,6 Result in Sulphonylurea-Sensitive Diabetes https://doi.org/10.2337/dc19-1843

OBJECTIVE Heterozygous loss-of-function mutations in HNF1A cause maturity-onset diabetes of the young (MODY). Affected individuals can be treated with low-dose sulpho- nylureas. Individuals with homozygous HNF1A mutations causing MODY have not been reported.

RESEARCH DESIGN AND METHODS We phenotyped a kindred with young-onset diabetes and performed molecular genetic testing, a mixed meal tolerance test, a sulphonylurea challenge, and in vitro assays to assess variant function.

RESULTS A homozygous HNF1A variant (p.A251T) was identified in three -treated 1 family members diagnosed with diabetes before 20 years of age. Those with the Diabetes, Endocrinology and Metabolism, Im- DIABETES IN COMMUNICATIONS NOVEL homozygous variant had low hs-CRP levels (0.2–0.8 mg/L), and those tested dem- perial College London, London, U.K. 2Oxford Centre for Diabetes, Endocrinology and onstrated sensitivity to sulphonylurea given at a low dose, completely transitioning off Metabolism, University of Oxford, Oxford, U.K. insulin. In silico modeling predicted a variant of unknown significance; however, in vitro 3Institute of Biomedical and Clinical Science, Uni- studies supported a modest reduction in transactivation potential (79% of that for the versity of Exeter Medical School, Exeter, U.K. 4 wild type; P < 0.05) in the absence of endogenous HNF1A. Diabetes and Endocrinology, Imperial College Healthcare NHS Trust, London, U.K. 5 CONCLUSIONS Wellcome Centre for Human Genetics, Univer- sity of Oxford, Oxford, U.K. Homozygous hypomorphic HNF1A variants are a cause of HNF1A-MODY. We thus 6Oxford NIHR Biomedical Research Centre, expand the allelic spectrum of variants in dominant causing diabetes. Churchill Hospital, Oxford, U.K. Corresponding author: Shivani Misra, s.misra@ Heterozygous loss-of-function mutations in the hepatocyte nuclear factor 1-a imperial.ac.uk (HNF1A) cause maturity-onset diabetes of the young (MODY) (1). Young age at onset, Received 16 September 2019 and accepted 7 autosomal dominant inheritance, and an excellent glycemic response to sulphonylurea January 2020 monotherapy characterize HNF1A-MODY (2). A spectrum of heterozygous HNF1A This article contains Supplementary Data online variants have been identifieddfrom classical MODY-causing mutations that impair at http://care.diabetesjournals.org/lookup/suppl/ insulin secretion (3,4), to milder alleles that increase the risk of . doi:10.2337/dc19-1843/-/DC1. Homozygous loss-of-function germline mutations are thought to be embryonically N.S.O. and A.L.G. jointly supervised this work. lethal (5). Establishing the pathogenicity of missense variants is challenging (6,7): © 2020 by the American Diabetes Association. hypomorphicvariantsintheheterozygous statemaybedismissedasbenign,butwhen Readers may use this article as long as the work is properly cited, the use is educational and not for inherited recessively, they may mimic loss-of-function heterozygous mutations (8). profit, and the work is not altered. More infor- We describe an insulin-treated individual in whom a homozygous HNF1A missense mation is available at http://www.diabetesjournals variant, p.A251T (c.751G.A), was identified. We metabolically characterized both .org/content/license. Diabetes Care Publish Ahead of Print, published online January 30, 2020 2 Homozygous HNF1A Alleles Cause Diabetes Diabetes Care

heterozygous carriers of the p.A251T variant and individuals homozygous for the p.A251T variant, performed in vitro functional studies to test our hypothesis that p.A251T is a hypomorphic HNF1A mutation, and assessed the transition to sulphonylurea therapy.

RESEARCH DESIGN AND METHODS In the proband, recruited to the MYDIABETES (MODY in Young-Onset Diabetes in Dif- ferent Ethnicities) study (clinical trial no. NCT02082132, ClinicalTrials.gov), we se- quenced HNF1A promoters, , and key flanking intronic sequences, along with 22 genes implicated in monogenic diabetes (9), and we assessed copy number var- iation using multiplex ligation-dependent probe amplification (9).

— In Silico Modeling of Pathogenicity Figure 1 Composite graph depicting results of the MMTT and the sulphonylurea challenge in the proband after morning premixed insulin was omitted. Glucose, C-peptide, and insulin during We assessed predicted pathogenicity us- a MMTT, and the glibenclamide test dose, at indicated time points are displayed. Glucogel 25 g, an ing established in silico tools, in line with oral glucose preparation B (BI Healthcare, Bridgend, U.K.), was administered sublingually. the American College of Medical Genetics classification (10) and as previously de- scribed (11). For structural analysis, we individuals (proband, his identical twin, prescription for 1.25-mg glibenclamide used FoldX (12), introducing variants and a sister) presented with type 1 di- once daily (which was subsequently in- p.A251T and a nearby MODY-causing abetes as teenagers and were treated creased to twice daily); insulin was mutation, p.V246 L (13), into the struc- with insulin (Supplementary Fig. 1 and stopped. Continuous glucose monitoring ture of the HNF1A DNA binding domain Supplementary Table 1). The father ful- performed before and after the treat- complexed with DNA (Research Collabo- filled criteria for prediabetes and the ment change demonstrated the reso- ratory for Structural Bioinformatics Pro- mother developed lution of , maintenance tein Data Bank identifier 1IC8) (14). during her third pregnancy at age 29 years; of glucose levels (Supplementary Fig. her diabetes persisted postpartum. 2), and HbA1c was reduced to 6.9% Clinical Studies In the proband, no additional mutations (52 mmol/mol). The proband’sidentical We assessed clinical features and hs-CRP, a or variants were detected, and anti-GAD65 twin also transitioned from insulin to marker of HNF1A protein function (15), in and IA2 antibodies were negative. Although sulphonylurea therapy. theproband andhis family.We also assessed consanguinity was not reported, both pa- in the proband b-cell function and glucose rents were from the same region in Eastern response to a mixed meal tolerance test Europe. In Silico Characterization of the (MMTT) (16), and his response to a 2.5-mg Levels of hs-CRP were low (#0.2–0.8 mg/L) p.A251T Variant The p.A251T variant is located in a region glibenclamide challenge. The proband un- in all those who were homozygous for of HNF1A that encodes a highly conserved derwent continuous glucose monitoring for p.A251T and in the heterozygous mother POU homeodomain DNA binding domain, 7-days before and after the switch to the (see Supplementary Table 1). The het- which is crucial for factor sulphonylurea. erozygous father displayed a higher function (14). Neither p.A251T nor other value (1.4 mg/L). rare variants at this position were present In Vitro Functional Characterization During the MMTT (see Fig. 1), blood in the Genome Aggregation Database The impact of p.A251T on HNF1A function glucose plateaued at 12.5 mmol/L; this (examined July 2019). A variant of un- was assessed by using a suite of in vitro was accompanied by a rise in C-peptide. certain significance was predicted by assays, as described previously (11). This After the proband ingested glibenclamide, in silico approaches (see Supplementary impact on function was compared with C-peptide increased further, beyond the Table 2). Detailed structural analysis of that of wild-type HNF1A and two DNA bind- levels measured during the MMTT. This p.A251T compared with that of a MODY- ing domain mutations that cause MODY. increase was sustained and decreased causing mutation (p.V246 L) at a similar only after glucose levels began to fall. location revealed the formation of a novel RESULTS Glucose dropped beyond fasting levels (to hydrogen bond (Supplementary Fig. 3). Clinical Characteristics of p.A251T 2.9 mmol/L), however, causing hypogly- Variant Carriers cemia. Oral glucose was administered, Young-onset diabetes cosegregated with resulting in another, higher C-peptide In Vitro Functional Assessment the homozygous, but not heterozygous, increase that restored normoglycemia. In vitro functional assays demonstrated a HNF1A p.A251T carriers; three homozygous The proband was discharged with a modest reduction in transactivation activity care.diabetesjournals.org Misra and Associates 3

and DNA binding for p.A251T-HNF1A, which was diagnosed at age 50 years, and a was also funded by the NIHR Oxford and BRC is less severe than bona fide mutations sister, diagnosed at age 53 years, had (to A.L.G.). fl (p.P112L and p.R203H) that cause early- only impaired fasting glycemia when Duality of Interest. No potential con icts of interest relevant to this article were reported. onset diabetes (Supplementary Fig. 4). studied (11,18). Thus the age of hetero- The views expressed are those of the author(s) zygous carriers at diagnosis has been and not necessarily those of the funding bodies markedly older (11) than that of individ- or oganizations involved in this research. CONCLUSIONS uals who are homozygous (12, 15, and Prior Presentation. Part of this work was presented as a poster at the 76th Scientific We report a homozygous hypomorphic 16 years). This is consistent with incom- Sessions of the American Diabetes Association, HNF1A mutation in a family with young- plete penetrance, and reported data New Orleans, LA, 10–14 June 2016. onset diabetes. The clinical phenotype is suggestthatahighBMImightberequired consistent with classical heterozygous to unmask the effect of the variant in the References HNF1A-MODY: individuals have low heterozygous state. 1. Frayling TM, Evans JC, Bulman MP, et al. beta- hs-CRP and demonstrate sensitivity to In our study, two homozygous indi- cell genes and diabetes: molecular and clinical sulphonylureas, enabling a transition off viduals had hs-CRP levels #0.2 mg/L, characterization of mutations in transcription – insulin. Importantly, young-onset diabe- whereas the heterozygous father had a factors. Diabetes 2001;50(Suppl. 1):S94 S100 2. Shepherd MH, Shields BM, Hudson M, et al.; tes cosegregated with homozygous indi- level of 1.4 mg/Ldwithin the range ob- UNITED study. A UK nationwide prospective study viduals, not heterozygous carriers. In vitro served for individuals with type 2 diabetes of treatment change in MODY: genetic subtype studies support p.A251T as being a hy- and those without diabetes (15,19). This and clinical characteristics predict optimal glycae- pomorphic variant with modest transac- higher level contrasts data from previous mic control after discontinuing insulin and met- formin. Diabetologia 2018;61:2520–2527 tivation defects only in the absence of reports in which hs-CRP levels were 3. Bjørkhaug L, Sagen JV, Thorsby P, Søvik O, endogenous HNF1A. This defect is likely all ,0.3 mg/L intheheterozygousstate Molven A, Njølstad PR. Hepatocyte nuclear factor- driven by low DNA-binding potential. (17), but it could be accounted for by 1 alpha gene mutations and diabetes in Norway. The defects observed, however, were his age and BMI. J Clin Endocrinol Metab 2003;88:920–931 4. Galan´ M, Garc´ıa-Herrero C-M, Azriel S, et al. less pronounced than those of reported The most compelling evidence favoring a fi Differential effects of HNF-1 mutations associ- bona de heterozygous MODY-causing pathogenicity of the biallelically inherited ated with familial young-onset diabetes on target mutations in a similar location. p.A251T variant is the successful transition gene regulation. Mol Med 2011;17:256–265 Our in silico studies provide a putative from insulin to low-dose sulphonylurea 5. Harries LW, Brown JE, Gloyn AL. Species- explanation for why p.A251T could have therapy after .10 years of diabetes. It specific differences in the expression of the modest structural effects despite being would not be expected that alternative HNF1A, HNF1B and HNF4A genes. PLoS One 2009;4:e7855 located in the POU homeodomain. We diabetes types could be effectively man- 6. Althari S, Gloyn AL. When is it MODY? Chal- showed that p.A251T effects are atten- aged with low-dose glibenclamide repli- lenges in the interpretation of sequence variants uated by the predicted formation of a cating the observed responses to a in MODY genes. Rev Diabet Stud 2015;12:330– novel hydrogen bond and that structural sulphonylurea in individuals with typical 348 changes are less deleterious than those heterozygous HNF1A mutations (20). 7. Sagen JV, Bjørkhaug L, Haukanes BI, et al. The HNF1A mutant Ala180Val: clinical challenges in of a similarly located MODY-causing determining causality of a rare HNF1A variant in mutation. The hypomorphic status of Conclusion familial diabetes. Diabetes Res Clin Pract 2017; this variant is further supported by the We conclude that biallelic hypomorphic 133:142–149 phenotype observed in heterozygous car- variants of HNF1A may present with a 8. Monies D, Maddirevula S, Kurdi W, et al. typical heterozygous MODY phenotype. Autozygosity reveals recessive mutations and riers, which is atypical for heterozygous novel mechanisms in dominant genes: implica- loss-of-function HNF1A mutations. Hypomorphic alleles causing MODY may tions in variant interpretation. Genet Med 2017; The p.A251T HNF1A variant has been be encountered more frequently as di- 19:1144–1150 reported in two studies describing het- verse populations, particularly those with 9. Ellard S, Lango Allen H, De Franco E, et al. erozygous carriers (17,18), bringing autozygosity, are studied using whole- Improvedgenetictestingformonogenicdiabetes exome or whole-genome approaches. using targeted next-generation sequencing. Dia- the total number of reported heterozy- betologia 2013;56:1958–1963 gous carriers to eight. In the study by 10. Richards S, Aziz N, Bale S, et al.; ACMG Thanabalasinghamet al. (17),theproband, Laboratory Quality Assurance Committee. Stan- who presented with diabetes at the age Funding. This work was undertaken with funds dards and guidelines for the interpretation of of 43 years, was managed with metformin from the Diabetes Research & Wellness Foun- sequence variants: a joint consensus recommen- dation (through a Sutherland-Earl Fellowship dation of the American College of Medical Genetics and sulphonylurea therapy for 25 years. 2013–2016) and the Imperial College Healthcare and Genomics and the Association for Molecular A heterozygous relative of the proband Charity, and with infrastructure support from Pathology. Genet Med 2015;17:405–424 developed diabetes later in life, at age the NIHR Imperial Biomedical Research Centre 11. Juszczak A, PavicT,Vu´ ckoviˇ c´ F, et al. Plasma 80 years. Although hs-CRP levels were (BRC), ImperialClinical ResearchFacility, andClinical fucosylated glycans and C-reactive protein as low in the three carriers studied, the Research Network. S.M. is currently supported by biomarkers of HNF1A-MODY in young adult-onset a Future Leaders Mentorship Award from the nonautoimmune diabetes.DiabetesCare 2019;42: level of fucosylated glycan GP30 (low European Association for the Study of Diabetes. 17–26 levels of which are predictive of HNF1A A.J. was a Diabetes UK George Alberti Clinical 12. Schymkowitz J, Borg J, Stricher F, Nys R, mutations) was above the cutoff in two Research Fellow when contributing to this study. Rousseau F, Serrano L. The FoldX web server: an of the three (18). In another study, a S.E. received a Senior Investigator Award from online force field. Nucleic Acids Res 2005;33: WellcomeTrust.A.L.G.isaWellcomeSeniorFellow W382–W388 heterozygous proband received a diag- in Basic Biomedical Science. Part of this work was 13. Ellard S, Colclough K. Mutations in the genes nosis at age 39 years and had a BMI of funded in Oxford by the Wellcome Trust (grants encoding the transcription factors hepatocyte 27 kg/m2. The proband’sfather,who 095101 and 200837 [both to A.L.G.]). The research nuclearfactor1alpha (HNF1A) and 4 alpha 4 Homozygous HNF1A Alleles Cause Diabetes Diabetes Care

(HNF4A) in maturity-onset diabetes of the 16. Oram RA, Jones AG, Besser REJ, et al. The HNF1A variants in Croatia. Biochem Med (Zagreb) young. Hum Mutat 2006;27:854–869 majority of patients with long-duration type 1 2018;28:020703 14. Chi Y-I, Frantz JD, Oh B-C, Hansen L, Dhe- diabetes are insulin microsecretors and have 19. McDonaldTJ,ShieldsBM, LawryJ, et al. High- Paganon S, Shoelson SE. Diabetes mutations functioning beta cells. Diabetologia 2014;57: sensitivity CRP discriminates HNF1A-MODY from delineate an atypical POU domain in HNF-1alpha. 187–191 other subtypes of diabetes. Diabetes Care 2011; Mol Cell 2002;10:1129–1137 17. Thanabalasingham G, Huffman JE, Kattla JJ, 34:1860–1862 15. Owen KR, Thanabalasingham G, James TJ, et al. Mutations in HNF1A result in marked 20. Pearson ER, Flechtner I, Njølstad PR, et al.; et al. Assessment of high-sensitivity C-reactive alterations of plasma glycan profile. Diabetes Neonatal Diabetes International Collaborative protein levels as diagnostic discriminator of 2013;62:1329–1337 Group. Switching from insulin to oral sulfonylur- maturity-onset diabetes of the young due to 18. Pavic´ T, Juszczak A, Pape Medvidovic´ E, et al. eas in patients with diabetes due to Kir6.2 HNF1A mutations. Diabetes Care 2010;33:1919–1924 Maturity onset diabetes of the young due to mutations. N Engl J Med 2006;355:467–477