Hyperglycemia and Hypoglycemia via the GCK Gene

Defects of glucokinase (hexokinase 4) can cause either hyperglycemia or hypoglycemia depending on underlying mutations of its encoding gene GCK (Osbak et al. Hum Mutat 30(11):1512-1526, 2009). Inactivating (loss-of-function) GCK mutations lead to two types of hyperglycemia: maturity-onset diabetes of the young (MODY) and permanent neonatal diabetes mellitus (PNDM). On the other hand, activating (gain-of-function) GCK mutations have been associated with hyperinsulinemic hypoglycemia, also known as congenital hyperinsulinism (CHI).

MODY is the most common type of monogenic diabetes, accounting for up to 5% of young adults diagnosed with diabetes (Owen. Clin Med 13(3):278-281, 2013; McDonald et al. Ann Clin Biochem 50(Pt 5):403-415, 2013). MODY typically presents in lean young adults before 25 years with an autosomal dominant family history. MODY patients have continuous production of endogenous insulin, absence of beta-cell autoimmunity and absence of signs of insulin resistance. As this non-insulin dependent type of diabetes is frequently misdiagnosed as Type 1 or Type 2 diabetes, a timely and accurate molecular diagnosis of MODY is essential to treatment decisions, prognosis, family screening and obstetric management of gestational diabetes (Pihoker et al. J Clin Endocrinol Metab 98(10):405540-62, 2013; Ellard et al. Diabetologia 51(4):546-553, 2008). Mutations in the GCK, HNF1A and HNF4A genes account for up to 80% of all MODY cases. GCK-MODY (MODY, type II or MODY2) represents a milder form of disease with a mild, asymptomatic and non-progressive fasting hyperglycaemia from birth without the requirement of pharmacological treatment. GCK-MODY is the most common form of MODY in children, accounting for about 30% of total MODY cases in the UK (Owen, 2013).

Neonatal diabetes mellitus, another type of monogenic diabetes, is defined as insulin-requiring hyperglycemia within the first 6 months (or even beyond 6 months) of life and is often associated with intrauterine growth retardation (Njølstad et al. N Engl J Med 344(21):1588-1592, 2001; Njølstad et al. Diabetes 52(11):2854-2860, 2003; Massa et al. Hum Mutat 25(1):22-27, 2005; Colombo et al. J Clin Invest. 118(6):2148-2156, 2008). About half of affected neonates have transient diabetes while the rest have permanent neonatal diabetes mellitus (PNDM; OMIM# 606176). Some severe PNDM patients may have neurological complications including developmental delay, epilepsy and neonatal diabetes (also termed DEND syndrome) (Molven et al. Expert Rev Mol Diagn 11(3):313-320, 2011). GCK-PNDM is caused by complete deficiency of glucokinase and thus represents the severe end of GCK- associated hyperglycemia.

Congenital hyperinsulinism (CHI) is a clinically and genetically heterogeneous condition characterized by hypoglycemia (Glaser et al. GeneReviews, 2003; Arnoux et al. Early Hum Dev 86(5):287-294, 2010). The age of disease onset ranges from the neonatal period with severe forms to infancy or childhood with milder forms. Severe patients typically have extremely low serum glucose while milder cases present with variable hypoglycemia. Affected newborns also develop nonspecific symptoms including seizures, apnea, hypotonia, and poor feeding. Severity of disease manifestations can vary within the same family.

The GCK gene has 10 coding exons that encode glucokinase (hexokinase 4), which phosphorylates glucose to produce glucose-6-phosphate in glucose metabolism pathways (Osbak et al., 2009). Playing a crucial role in the regulation of insulin secretion, glucokinase has been termed the glucose sensor in pancreatic beta-cells. GCK defects throughout the whole gene include missense, nonsense, regulatory, splicing site mutations, small deletion/insertions while intragenic exon-level deletions/duplications are rare (Human Gene Mutation Database).

MODY is inherited in an autosomal dominant manner. So far, there are about 12 different genes associated with MODY-like phenotypes (Molven et al., 2011). Alongside HNF1A, HNF4A and HNF1B, GCK is among the most frequently involved genes in MODY (Owen, 2013). Approximately 85% of documented pathogenic GCK mutations are associated with MODY (Human Gene Mutation Database). Among these MODY-associated GCK mutations, about 60% are missense changes. Partial or whole gene deletions have been found to be rare in GCK-MODY (Ellard et al. Diabetologia 50(11):2313-2317, 2007; Garin et al. Clin Endocrinol 68(6):873-878, 2008). Phenotypes are notably similar for all MODY-associated GCK mutations (Ellard et al., 2008).

PNDM is commonly caused by mutations in the genes ABCC8 and KCNJ11, which encode the subunits of the beta-cell ATP-sensitive potassium (KATP) channel (Molven et al., 2011). Another relatively frequent cause of PNDM are mutations in the insulin gene (INS). Defects of GCK represent a rare cause of PNDM. GCK-PNDM is inherited recessively and only limited cases have been reported. PNDM-associated GCK defects include missense, nonsense and frameshift mutations, either homozygous or compound heterozygous which result in complete deficiency of glucokinase activity (Osbak et al., 2009).

CHI is genetically caused by defects in genes involved in regulation of insulin secretion from pancreatic beta-cells (Kapoor et al. Eur J Endocrinol 168(4):557-564, 2013; Snider et al. J Clin Endocrinol Metab 98(2):E355-363, 2013). GCK-related congenital hyperinsulinism (OMIM# 602485) is inherited in an autosomal dominant manner. Dominant activating GCK mutations are the third most common genetic cause of CHI (Snider et al., 2013; Glaser et al. N Engl J Med 338(4):226-230, 1998). GCK mutations have been found in both diazoxide-responsive (Christesen et al. Eur J Endocrinol 159(1):27-34, 2008) and diazoxide-unresponsive CHI patients (Snider et al., 2013). De novo GCK mutations in CHI are very common. All germline GCK mutations found in CHI patients are missense substitutions with a clustering at a confined region termed the allosteric activator site, which is remote to the substrate-binding site. Notably, a postzygotic mosaic GCK mutation (c.1361_1363dupCGG/p.Ala454dup) has been found in a patient’s pancreatic, but not peripheral blood, DNA (Snider et al., 2013). Eight other genes have also been associated with CHI including two KATP genes ABCC8 and KCNJ11, GLUD1, HADH, SLC16A1, HNF4A, HNF1A and UCP2 (Glaser et al., 2003; Kapoor et al.; Snider et al., 2013).

The GCK mutation detection rate is widely variable depending on the age and clinical selection criteria of the target population (Osbak et al. Hum Mutat 30(11):1512-1526, 2009; Pihoker et al. J Clin Endocrinol Metab 98(10):405540-62, 2013). In a study of 82 children with incidental hyperglycaemia, 43% were found to carry GCK mutations (Feigerlová et al. Eur J Pediatr 165(7):446-452, 2006). The GCK mutation detection rate in a large cohort of patients with PNDM is unavailable because mutations have been only reported in limited individual cases. Defects of GCK represent a rare cause of PNDM. In a cohort of 298 diazoxide-nonresponsive CHI patients studied at the Hyperinsulinism Center in The Children’s Hospital of Philadelphia (CHOP), approximately 2% (7 of 298) were identified to have GCK mutations via DNA sequencing (Snider et al. J Clin Endocrinol Metab. 98(2):E355-63, 2013). In another cohort of 300 CHI patients studied in the United Kingdom, no GCK mutations were found in 105 diazoxide-nonresponsive or 183 diazoxide-responsive patients (Kapoor et al. Eur J Endocrinol 168(4):557-564, 2013). In a study of CHI patients enrolled from three European referral centers, the overall prevalence of GCK-related non-syndromic CHI was approximately 1%. In the medically responsive group, the prevalence of GCK-caused CHI was approximately 7% (Christesen et al. Eur J Endocrinol 159(1):27-34, 2008).

This test provides full coverage of all coding exons of the GCK gene plus 10 bases of flanking noncoding DNA in all available transcripts along with other non-coding regions in which pathogenic variants have been identified at PreventionGenetics or reported elsewhere. We define full coverage as >20X NGS reads or Sanger sequencing. PGnome panels typically provide slightly increased coverage over the PGxome equivalent. PGnome sequencing panels have the added benefit of additional analysis and reporting of deep intronic regions (where applicable).

It also includes targeted testing of a promoter regulatory mutation c.-557G>C (Gasperíková et al. Diabetes 58(8):1929-1935, 2009).

Dependent on the sequencing backbone selected for this testing, discounted reflex testing to any other similar backbone-based test is available (i.e., PGxome panel to whole PGxome; PGnome panel to whole PGnome).