Constitutive Activation of AKT2 in Humans Leads to Hypoglycemia Without Fatty Liver or Metabolic Dyslipidemia

Abstract

Context: The activating p.Glu17Lys mutation in AKT2, a kinase mediating many of insulin's metabolic actions, causes hypoinsulinemic hypoglycemia and left-sided hemihypertrophy. The wider metabolic profile and longer-term natural history of the condition has not yet been reported.

Objective: To characterize the metabolic and cellular consequences of the AKT2 p.Glu17Lys mutation in two previously reported males at the age of 17 years.

Design and intervention: Body composition analysis using dual-energy X-ray absorptiometry, overnight profiling of plasma glucose, insulin, and fatty acids, oral glucose tolerance testing, and magnetic resonance spectroscopy to determine hepatic triglyceride content was undertaken. Hepatic de novo lipogenesis was quantified using deuterium incorporation into palmitate. Signaling in dermal fibroblasts was studied ex vivo.

Results: Both patients had 37% adiposity. One developed hypoglycemia after 2 hours of overnight fasting with concomitant suppression of plasma fatty acids and ketones, whereas the other maintained euglycemia with an increase in free fatty acids. Blood glucose excursions after oral glucose were normal in both patients, albeit with low plasma insulin concentrations. In both patients, plasma triglyceride concentration, hepatic triglyceride content, and fasting hepatic de novo lipogenesis were normal. Dermal fibroblasts of one proband showed low-level constitutive phosphorylation of AKT and some downstream substrates, but no increased cell proliferation rate.

Conclusions: The p.Glu17Lys mutation of AKT2 confers low-level constitutive activity upon the kinase and produces hypoglycemia with suppressed fatty acid release from adipose tissue, but not fatty liver, hypertriglyceridemia, or elevated hepatic de novo lipogenesis. Hypoglycemia may spontaneously remit.

Figure 1.

Quantification of visceral and subcutaneous adipose tissue. Representative abdominal fat distribution determined by magnetic resonance imaging at the L4 vertebral level is shown with pseudocolored adipose depots for P1, P2, and two healthy sex- and age-matched controls (C1, C2). Visceral adipose tissue (VAT) is indicated in yellow, and subcutaneous adipose tissue (SCAT) is indicated in red.

Figure 2.

Overnight profiling of plasma glucose, insulin, and free fatty acid concentrations. Plasma concentrations of glucose, insulin, and free fatty acids are shown during fasting from 11 pm for (a) P1 and for (b) P2. Midnight is represented by 0 hours. Samples taken during spontaneous hypoglycemia in P2 are indicated with red icons, whereas administration of an oral “rescue” bolus of 34 g glucose solution is indicated by red arrows. NEFA, nonesterified fatty acids.

Figure 3.

Oral glucose tolerance testing. Plasma glucose and insulin concentrations after a 75-g oral glucose load are shown for (a and b) P1 and (c and d) P2. Testing was undertaken following the overnight fast illustrated in Fig. 2. Sex-matched control data from a large European white control population from eastern England are shown as 5th, 50th, and 95th percentiles.

Figure 4.

Insulin and IGF-1 signaling in primary dermal fibroblasts from P1. Phosphorylation of AKT at (a) threonine 308/309 (Thr309 in AKT2) and (b) serine 473/474 (Ser474 in AKT2) in dermal fibroblasts from P1 and three healthy controls assessed after 18 hours of serum starvation with or without insulin or IGF-1 stimulation. Results shown are means of three independent experiments, with error bars representing standard deviations. Differences in protein phosphorylation between patient cells and the corresponding control cells under the same conditions of ligand stimulation were tested by unpaired t test. *P < 0.05, **P < 0.01, ***P < 0.001. A.U., arbitrary units.