Spliceosomal Catalytic Activation Through Orchestrated Function of tri-snRNP and B Complex Proteins

Splicing is catalyzed by secondary structures involving U6 snRNA, and catalysis is downregulated by a mutually exclusive base-pairing interaction with U4 snRNA. During spliceosome assembly, inactive U6 is recruited to the substrate via the U4/U6.U5 tri-snRNP. U4 is then removed from U6 by the ATP-dependent RNA helicase Brr2 to catalytically activate the spliceosome. Because Brr2 is a core component of the tri-snRNP, regulation of Brr2 is anticipated as necessary to couple its unwinding activity to substrate recognition. Indeed, we show that the purified conventional budding yeast tri-snRNP, in which Brr2 is already active, fails to function in splicing. Still, the mechanism for Brr2 regulation remains unresolved. Cryo-EM structures of the human tri-snRNP, in which Brr2 is repressed, have implicated a mechanism of repression that involves sequestration of Brr2 away from U4 by Prp28, which mediates U1 release from the 5’ splice site, and Sad1. In the conventional yeast tri-snRNP, in which Brr2 is not repressed, both factors are conspicuously absent. However, by pull-down and mass spectrometry of tri-snRNPs isolated from yeast under low salt, we have found both Prp28p and Sad1p associate with the tri-snRNP. Further, these tri-snRNPs are repressed for Brr2p-mediated U4/U6 unwinding, consistent with a role for Sad1p and Prp28p in repressing Brr2p. However, we found that salt-washing Sad1 and Prp28 off of tri-snRNPs, though sufficient to alter the RNA secondary structure around the U4 central domain, was insufficient to activate Brr2-mediated unwinding. Conventional, activated yeast tri-snRNPs not only lack Sad1p and Prp28p but also include B-complex proteins (BCP), Prp38p, Snu23p and Spp381p. Indeed, we found addition of these factors to tri-snRNPs washed free of Prp28p and Sad1p activated U4/U6 unwinding. Further, mass spectrometry and biochemical analysis showed that binding of Prp28p and Sad1p to the tri-snRNP was mutually exclusive with the binding of BCP factors, consistent with mutual exclusivity inferred from structures where BCP interact with the U6-5’ splice site duplex. These results imply that Prp28 prepresses Brr2p in part by disallowing the premature recruitment of these activating BCP factors. Contrary to previous interpretations, these findings support a mechanism for spliceosome activation that is common between yeast and humans. Our data point to a two-step mechanism to activate Brr2p. In the first step, the release of U1 from the 5’ splice site dissociates the repressive Prp28p and Sad1p, and in the second step, the binding of U6 to the 5’ splice site recruits the activating BCP factors. In this way, spliceosome activation is tightly coupled to 5’ splice site recognition.