This is part of my last year in science series. Click here for the rationale.
Insulin signalling passes through a cascade of signaling proteins starting with the recruitment of the Insulin Receptor Substrate (IRS) to the Insulin Receptor. Insulin resistance, or impaired insulin signaling is a hallmark of obesity and diabetes. One of the ways in which was thought to happen was through phosphorylation of Serine 307 on IRS, a phosphorylation event which has been proposed to be inhibitory to insulin signaling. As an example, it has been proposed that inflammation (via JNK or IKKbeta), overnutrition (via S6K) or several other factors can lead to the phosphorylation of this protein. Since several of these factors correlate with reduced insulin signaling, and the ablation of these factors leads to both increased insulin signaling and reduced serine phosphorylation, the obvious hypothesis was that serine phosphorylation of IRS is causative of insulin signaling.
Correlation does not equal causation, so in order to test this hypothesis in vivo, Morris White's group at Harvard Medical School generated knockin mice, in which Serine 307 of IRS1 is mutated to an alanine (and is therefore unable to be phosphorylated). The idea would be that these mice would be unable to phosphorylated IRS1 on Serine 307, and therefore would be resistant to the deleterious effects of this phosphorylation. This serine to alanine knockin model is considered the gold standard for translating an observational protein phosphorylation site into an in vivo phenotype. Unfortunately for the prevailing hypothesis, the opposite was true.
The paper, from Copps et al. published in January of 2010 show that on a normal diet S307A mice showed modest reductions in insulin sensitivity, and increased fasted insulin levels. Both of these effects were amplified by high fat diet, and were associated with a reduction in weight gain, in contradiction to the previous hypothesis, that these effects would be ameliorated. Mechanistically, when on a high fat diet (or coupled with liver specific IRS2 knockout) the S307A mouse had decreased tyrosine phosphorylation of IRS, but no effect of the downstream insulin targets Akt and S6K. Therefore it is unclear exactly how the insulin intolerance is propagated into effects on glucose homeostasis.
In the past year this article has been cited 10 times (according to Google Scholar), mostly in review articles, but the major upshot here, is that the models which showed effects on IRS Serine 307 phosphorylation and insulin resistance, and concluded that insulin resistance was mediated by increased phosphorylation may need re-interpretation. IRS is phosphorylated on several other sites, so the general hypothesis that serine phosphorylation of IRS causes insulin resistance could still be true, but that again might need to wait until such a knockin model can be generated. This work also points out the risks of correlating phenotypes with incompletely characterized phosphorylation sites.
Copps KD, Hancer NJ, Opare-Ado L, Qiu W, Walsh C, & White MF (2010). Irs1 serine 307 promotes insulin sensitivity in mice. Cell metabolism, 11 (1), 84-92 PMID: 20074531 DOI