It's well known that heat shock causes the repression of pre-messenger RNA splicing. Reporting in Nature, James Manley and colleagues have now identified the repressor responsible for this effect — a general splicing repressor known as SRp38.

SRp38 is activated by dephosphorylation and, indeed, the authors showed that SRp38 became dephosphorylated when HeLa cells were heat shocked in vivo. When the cells were allowed to recover, SRp38 reverted to its phosphorylated form. Heat treatment of HeLa nuclear extracts in vitro also resulted in the rapid dephosphorylation of SRp38 and inhibited splicing activity. When nuclear extracts were left to recover, splicing activity was restored.

Direct evidence came from depletion studies, in which the depletion of SRp38 from the nuclear extracts of heat-treated cells, or from heat-treated nuclear extracts, restored splicing activity. Subsequent addition of dephosphorylated SRp38 to the depleted nuclear extract restored splicing inhibition.

Glutathione-S-transferase (GST) 'pull-down' assays revealed a mechanistic clue. The U1 small nuclear ribonucleoprotein (snRNP) bound strongly to the GST–SRp38 fusion protein, but only when SRp38 was in its dephosphorylated form. Previous studies had shown that the interaction between another splicing factor, ASF/SF2, and U1 snRNP stabilizes the U1 snRNP recognition of 5′ splice sites. Manley and colleagues now show that SRp38 in its dephosphorylated form results in a strong decrease in ASF/SF2–U1-snRNP complex formation at 5′ splice sites. This suggests that dephosphorylated SRp38 represses splicing by interacting with U1 snRNP, and thereby interferes with U1 snRNP recognition of 5′ splice sites.

So, is SRp38-mediated repression of splicing important for cell survival under stress conditions? The answer is yes — even though SRp38-deficient cells are viable, they are temperature sensitive and recover more slowly from heat shock than wild-type cells. So, the authors conclude that SRp38 “...represses splicing following heat shock, at least in part, to prevent the possible accumulation of inaccurately spliced mRNA”.