UC Santa Cruz

Branchpoint (bp) recognition is a critical step in spliceosome assembly and influences 3' (ss) selection. Improper splice site choices can produce aberrant proteins and targets of nonsense mediated decay (NMD). The U2 snRNP is stably bound to the bp in the first ATP dependent step of splicing. Initial recognition of the bp by the U2 snRNP is through nucleation by the presenting nucleotides of the branchpoint interacting stem-loop (BSL) of U2 snRNA before this ATP dependent step. The exact mechanism of this ATP dependence is unknown. In yeast, we know that ATP dependence can be bypassed by loss of CUS2 or disruption of CUS2’s binding to HSH155. Here I use HeLa cells as a model system to determine if the human homolog of CUS2 also enforced the ATP-dependent step through its binding to SF3B1. I find that Tat-SF1 depletion or disruption of its interaction with SF3B1 is not enough to bypass the ATP requirement of early complex formation. Furthermore, I investigated the extended duplex that forms between the U2 snRNA and the eight-nucleotide region upstream of the bp sequence of the intron. We did not expect to see this duplex forming in cryo-EM structures especially without Watson-Crick base pairing. Using both HeLa and Saccharomyces cerevisiae as model organisms I found that perfect base pairing in the duplex may not be beneficial. In yeast, extra base pairing in the duplex can make a bp the preferred bp despite another identical bp in the intron. Hyperstabilized bp mutants trigger a decay pathway that would not normally be activated with canonical substrates. In mammals, extended duplex hyperstablization disrupts the stability of the A-complex spliceosomes but does not inhibit splicing. Taken together, these results uncover an evolutionary reason for the lack of Watson-Crick base pairing in the extended duplex upstream of the bp sequence.