The real test of eIF-2α(P)'s role was to see what happened to protein synthesis if you got rid of it. For reasons that needn't concern us in-depth at the moment, we hypothesized that insulin might induce dephosphorylation of eIF2α(P). Our experiments confirmed that very high dose insulin (20U/kg!) administered at the onset of reperfusion did, in fact, clear eIF2α(P) from neurons. Moreover--and here's the kicker--removing eIF2α(P) with insulin restored protein synthesis.

That pretty much sealed it. Clearly, the massive phosphorylation of eIF2α during early reperfusion was responsible for the suppression of translation initiation in vulnerable neurons.

Insulin to the rescue! The images above are autoradiographs of protein synthesis in CA1 paired with immunohistochemistry of the same areas for eIF2α(P) (like that you saw on the previous page). On the left, the autoradiographs show protein synthesis as little black dots overlying the cells like so much pepper. These are silver grains "developed" by radiation from S-35 methionine given to the animals at 1 hour of reperfusion. The radioactive methionine was incorporated into protein only in those cells that had intact translation. Note that in control (nonischemic) animals, the neurons are busily making protein and are clear of eIF2α(P). After ischemic injury, the neurons are full of eIF2α(P) and protein synthesis is offline. But if the animals receive 20U/kg of insulin at onset of reperfusion, by 90 minutes the neurons are clear of eIF2α(P) and protein synthesis is back online.  

Of course, the beautiful thing about science is that, as soon as you answer one question, another one takes its place. Having found that phosphorylation of eIF2α(P) was the critical event, the question immediatley became: who's the kinase?