Diabetes is one of the fastest growing diseases in the western world, and so finding a cure is a pressing need for the millions of familes worldwide that are affected by the various defects in sugar metabolism that this disease encompasses. Individuals affected by lipoatrophic diabetes mellitus, for example, have high levels of sugar in the blood due to insulin resistance. The genetic predisposition to diabetes is probably as varied as the disease itself, therefore, animal models offer an opportunity to understand the molecular and physiological basis of this class of disorder as well as providing a testbed for possible cures. Laustsen, Michael and colleagues have now created a mouse model of lipoatrophic diabetes and have investigated a fruitful route for its cure by gene therapy.

The study stemmed from the authors' interest in four mouse proteins — the insulin receptor substrates (IRS) 1–4 — which relay the insulin signal within cells. Single knockout mice for each Irs gene had shown that Irs1 and Irs2 have important and non-redundant roles in post-natal growth and glucose homeostasis, respectively; by contrast, Irs3 and Irs4 knockout mice had few abnormalities. But the Irs3 and Irs4 null phenotypes could have been compensated for by Irs1 and Irs2 or, conversely, the Irs1 and Irs2 mutant phenotypes could have been ameliorated by Irs3 and Irs4. Double knockout mice showed no redundancy between Irs1 and Irs4; however, the phenotype of Irs1,3 double knockout mice bore a striking resemblance to the features of lipoatrophic diabetes in humans: the mice had high levels of glucose and insulin in the blood, had reduced white adipose tissue (lipoatrophy) and developed non-insulin-dependent diabetes. Furthermore, the diabetes phenotype could be reversed by injecting the Irs1,3−/− mice with an adenoviral vector carrying the leptin gene, which had previously been shown to alleviate the insulin resistance of the other mouse models of lipoatrophic diabetes.

Irs1 and Irs3, therefore, have complementary physiological roles. So, what promise does this hold for human diabetes sufferers? As humans lack the IRS3 protein, they might be more dependent on the function of IRS1 , and so sequence variants of this gene should be worth a closer look.