Congenital hyperinsulinaemic hypoglycaemia (HH) is a rare disorder where episodes of hypoglycaemia are caused by unregulated insulin secretion. Variants in UCP2 (NM_003355.2) have been reported to cause HH (González-Barroso, et al., 2008) however, since this publication large scale population data have become available which provide important information on the prevalence of rare variants within the population. One such resource is the Genome Aggregation Database (gnomAD) (Lek, et al., 2016) which provides sequencing data from 138,632 individuals. Datasets such as this have revolutionised our ability to interpret variants, as rarity is an important criterion for pathogenicity – the frequency of a variant in a population not selected for the condition should not exceed the prevalence of the condition.
Three studies have reported UCP2 variants in 9 HH patients (table 1). The original paper (González-Barroso, et al., 2008) describes variants in two probands, Snider et al. identified variants in 2 patients out of 417 HH patients screened (Snider, et al., 2013), and Ferrara et al. presented 5 patients with UCP2 variants in a cohort of 211 patients with HH (Ferrara, et al., 2016). Whilst functional studies on the c.803C>G/p.Ala268Gly variant demonstrated a reduction in UCP2 protein activity (Vozza, et al., 2014), these findings alone are insufficient to assert a causal role of the variant in the aetiology of this disease.
Table 1. UCP2 (NM_003355.2) variants reported in patients with hyperinsulinaemic hypoglycaemia, and their frequency in the gnomAD database (gnomad.broadinstitute.org).
| Variant | Patients | Published in | gnomAD database | ||
|---|---|---|---|---|---|
| Number of individuals n= 138,632 | Frequency | Popmax ‡ frequency | |||
| p.Ser47Asn | 1 | Ferrara et al. | 42 | 1 in 6592 | 1 in 616 |
| p.Gly61Ser | 3 | Ferrara et al., Snider et al. | 21 | 1 in 13190 | 1 in 1601 |
| p.Gly174Asp | 1* | González-Barroso et al. | 0 | 0 | 0 |
| p.Leu175Val | 1* | González-Barroso et al. | 0 | 0 | 0 |
| p.Ala187Asp | 1* | González-Barroso et al. | 0 | 0 | 0 |
| p.Ala268Gly | 3 | Ferrara et al., González-Barroso et al., Snider et al. | 227 | 1 in 1216 | 1 in 128 |
| p.Arg279His◊ | 1 | This study◊ | 27 | 1 in 10208 | 1 in 1274 |
| c.816-2A>G | 1 | Ferrara et al. | 33 | 1 in 8157 | 1 in 3824 |
defined by Lek et al 2016 "popmax" is "the highest allele frequency in any one population"
The 3 variants were all found in the same patient
Variant submitted to LOVD http://databases.lovd.nl/shared/variants/0000163118#00022323
All variants in the original study were inherited from a parent without congenital HH and although Ferrara et al. reported that 4 relatives with a UCP2 variant had unusual glucose and insulin responses to oral glucose, none of the family members had received a clinical diagnosis of HH prior to the study. In these cases the lack of co-segregation could be attributed to age dependent expressivity or incomplete penetrance, as is seen with some other dominantly acting HH mutations (Glaser, et al., 1998).
We investigated the role of UCP2 in the aetiology of this disorder by analysing data from the gnomAD database and by sequencing the gene in a cohort of 206 individuals with HH by targeted next generation sequencing (Ellard, et al., 2013). Informed consent was obtained from the probands or their parents/guardians and the study was approved by the North Wales Research Ethics Committee, UK.
Whilst no protein truncating variants in UCP2 were found in our HH cohort, a single missense change (c.836G>A/p.Arg279His – see Supp. Table S1 for in silico predictions) was identified in one patient (1/206 (0.5%)). The frequency of rare UCP2 missense variants was not enriched in patients with HH in our cohort compared to their frequency in gnomAD (OR 0.4, 95% CI 0.009 – 2, P=0.5). The patient with a UCP2 missense variant was of Indian ethnicity and as the variant is present at 1 in 1274 in the gnomAD South Asian population (estimated HH prevalence in outbred populations is 1 in 50,000 (Bruining, 1990)), our interpretation is that this is a relatively common polymorphism in the Indian population and not causative of the patient’s HH. The likelihood that this is a benign variant is further supported by the presence of the variant in the proband’s clinically unaffected mother.
Four of the UCP2 variants reported as pathogenic in individuals with HH (Ferrara, et al., 2016; González-Barroso, et al., 2008; Snider, et al., 2013) are present at a high frequency in gnomAD – a dataset where individuals with severe paediatric disease have been excluded (see Table 1).These four variants account for eight of the nine patients reported to have HH caused by variants in UCP2. Ferrara et al. reported five probands with UCP2 variants; four were of African American ethnicity and had variants that have a frequency of greater than 1 in 2000 in the gnomAD African population. The fifth proband, of Western European ancestry, was heterozygous for the c.816-2A>G/p.? variant which is present in the non-Finnish European gnomAD population at a frequency of 1 in 3824.
The p.Ala268Gly variant reported as causative of HH in three separate studies (Ferrara, et al., 2016; González-Barroso, et al., 2008; Snider, et al., 2013) is present in 227 people in gnomAD, including 188 Africans. If this variant were pathogenic its frequency in the gnomAD African population would equate to an incidence of HH of 1 in 128 which is significantly higher than the known incidence of HH which is 1 in 50,000 in outbred populations (Bruining, 1990)(binomial test, P= <1x10-7). Lek et al. used ExAC to demonstrate that the presence of a variant at a high frequency in a public database is grounds for re-defining it as non-pathogenic despite supporting functional evidence (Lek, et al., 2016). While there is in vitro evidence that the p.Ala268Gly variant causes a reduction in UCP2 protein function (Ferrara, et al., 2016; Vozza, et al., 2014), this does not mean that it is sufficient to cause a clinical phenotype. We suggest that p.Ala268Gly could be a functional polymorphism and that the role of UCP2 in the aetiology of HH should be reconsidered.
In outbred populations estimates of HH range from 1 in 27,000 in Ireland (Glaser, et al., 2000), 1 in 40,000 in Finland (Otonkoski, et al., 1999) to 1 in 50,000 in the Netherlands (Bruining, 1990). While these estimates are likely an underrepresentation of the true prevalence as they likely miss milder cases of HH, all estimates are greatly in excess of the frequency observed for the published variants in UCP2. It is possible that variants in UCP2 could be involved in the aetiology of HH but as low effect risk factors in a similar manner to the role played by the KCNJ11 c.67G>A/p.E23K variant in type 2 diabetes where it is a known risk factor (OR 1.23) (Gloyn, et al., 2003) despite a frequency of 1 in 1.5 in gnomAD. However, if variants in UCP2 are acting as low effect risk factors or only cause a very mild phenotype then UCP2 should still not be screened as part of genetic diagnostic panels as it would not be a monogenic cause of severe HH.
In conclusion, our data does not support the role of variants in UCP2 as a monogenic cause of HH. We therefore urge caution in the diagnostic testing of UCP2 until its role in the aetiology of HH is proven.
Supplementary Material
Acknowledgements
The authors thank Matthew Johnson (University of Exeter, Exeter, UK), Anna-Marie Bussell, Rebecca Ward and Garan Jones (Department of Molecular Genetics, Royal Devon & Exeter NHS Foundation Trust, Exeter, UK) for their technical assistance and James Harrison (University of Exeter, Exeter, UK) for useful discussion of the manuscript.
Funding
SEF has a Sir Henry Dale Fellowship jointly funded by the Wellcome Trust and the Royal Society (Grant Number: 105636/Z/14/Z). SE is a Wellcome Trust Senior Investigator.
Footnotes
Potential conflicts of Interest
The authors have no conflicts of interest to disclose.
Contribution statement
KH, SE and SEF recruited the cohort. RC designed the targeted capture assay. TWL and MNW carried out the analysis. TWL, SE and SEF wrote the manuscript. All authors were involved in discussion of the manuscript and gave final approval of the version to be published.
References
- Bruining J. Recent advances in hyperinsulinism and the pathogenesis of diabetes mellitus. Current Opinion in Pediatrics. 1990;2(4):758–765. [Google Scholar]
- Ellard S, Lango Allen H, De Franco E, Flanagan SE, Hysenaj G, Colclough K, Houghton JAL, Shepherd M, Hattersley AT, Weedon MN, Caswell R. Improved genetic testing for monogenic diabetes using targeted next-generation sequencing. Diabetologia. 2013;56(9):1958–1963. doi: 10.1007/s00125-013-2962-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ferrara CT, Boodhansingh KE, Paradies E, Giuseppe F, Steinkrauss LJ, Topor LS, Quintos JB, Ganguly A, De Leon DD, Palmieri F, Stanley CA. Novel Hypoglycemia Phenotype in Congenital Hyperinsulinism Due to Dominant Mutations of Uncoupling Protein 2 (UCP2) J Clin Endocrinol Metab. 2016 doi: 10.1210/jc.2016-3164. jc20163164. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Glaser B, Kesavan P, Heyman M, Davis E, Cuesta A, Buchs A, Stanley CA, Thornton PS, Permutt MA, Matschinsky FM, Herold KC. Familial Hyperinsulinism Caused by an Activating Glucokinase Mutation. New England Journal of Medicine. 1998;338(4):226–230. doi: 10.1056/NEJM199801223380404. [DOI] [PubMed] [Google Scholar]
- Glaser B, Thornton P, Otonkoski T, Junien C. Genetics of neonatal hyperinsulinism. Archives of Disease in Childhood. Fetal and Neonatal Edition. 2000;82(2):F79–F86. doi: 10.1136/fn.82.2.F79. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Gloyn AL, Weedon MN, Owen KR, Turner MJ, Knight BA, Hitman G, Walker M, Levy JC, Sampson M, Halford S, McCarthy MI, et al. Large-Scale Association Studies of Variants in Genes Encoding the Pancreatic β-Cell KATP Channel Subunits Kir6.2 (KCNJ11) and SUR1 (ABCC8) Confirm That the KCNJ11E23K Variant Is Associated With Type 2 Diabetes. Diabetes. 2003;52(2):568–572. doi: 10.2337/diabetes.52.2.568. [DOI] [PubMed] [Google Scholar]
- González-Barroso MM, Giurgea I, Bouillaud F, Anedda A, Bellanné-Chantelot C, Hubert L, de Keyzer Y, de Lonlay P, Ricquier D. Mutations in UCP2 in Congenital Hyperinsulinism Reveal a Role for Regulation of Insulin Secretion. PLoS ONE. 2008;3(12):e3850. doi: 10.1371/journal.pone.0003850. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Lek M, Karczewski KJ, Minikel EV, Samocha KE, Banks E, Fennell T, O’Donnell-Luria AH, Ware JS, Hill AJ, Cummings BB, Tukiainen T, et al. Analysis of protein-coding genetic variation in 60,706 humans. Nature. 2016;536(7616):285–291. doi: 10.1038/nature19057. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Otonkoski T, Ammälä C, Huopio H, Cote GJ, Chapman J, Cosgrove K, Ashfield R, Huang E, Komulainen J, Ashcroft FM, Dunne MJ, et al. A point mutation inactivating the sulfonylurea receptor causes the severe form of persistent hyperinsulinemic hypoglycemia of infancy in Finland. Diabetes. 1999;48(2):408–415. doi: 10.2337/diabetes.48.2.408. [DOI] [PubMed] [Google Scholar]
- Snider KE, Becker S, Boyajian L, Shyng SL, MacMullen C, Hughes N, Ganapathy K, Bhatti T, Stanley CA, Ganguly A. Genotype and Phenotype Correlations in 417 Children With Congenital Hyperinsulinism. The Journal of Clinical Endocrinology and Metabolism. 2013;98(2):E355–E363. doi: 10.1210/jc.2012-2169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Vozza A, Parisi G, De Leonardis F, Lasorsa FM, Castegna A, Amorese D, Marmo R, Calcagnile VM, Palmieri L, Ricquier D, Paradies E, et al. UCP2 transports C4 metabolites out of mitochondria, regulating glucose and glutamine oxidation. Proceedings of the National Academy of Sciences of the United States of America. 2014;111(3):960–965. doi: 10.1073/pnas.1317400111. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.