Single molecule analysis reveals reversible and irreversible steps during spliceosome activation

Thank you for submitting your article "Single Molecule Analysis Reveals Reversible and Irreversible Steps During Spliceosome Activation" for consideration by eLife. Your article has been favorably evaluated by Jean Beggs and Jonathan Staley, who is a member of our Board of Reviewing Editors, and the evaluation was overseen by Michael Marletta as the Senior editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

In this manuscript, Hoskins et al. exploit the unique and innovative capabilities of CoSMoS to study at unprecedented resolution the activation step of the spliceosome, arguably the most dramatic and catalytically-critical stage of spliceosome assembly. They investigate formation of the B complex and its conversion to the Bact complex by observing splicing substrate association and dissociation of U1, the tri-snRNP components U4 and U5, and/or the NTC, the complex characteristic of catalytically-activated spliceosomes. At low ATP, which stalls spliceosome assembly, the authors observed increased lifetimes of U1, U4, and U5 but decreased lifetimes of the NTC, consistent with previous observations. By simultaneously monitoring U4 and NTC binding and release from the substrate by three-color CoSMoS, the authors demonstrate that the NTC only transiently samples the substrate before U4 release and generally binds stably only after release, providing the first direct evidence for the dependence of stable NTC binding on U4 release. By co-localization of U4 and U5 by three-color CoSMoS, the authors also observed a substantial frequency of events in which both dissociate simultaneously. Surprisingly, at high ATP, there is a very fast phase of this event that predominates, implying a novel discard pathway that dominates the spliceosome activation phase. Importantly, the authors show that such discard events can be followed by rebinding of the tri-snRNP components. Further, while some studies on the U4/U6 annealing factor Prp24 have suggested action on the spliceosome, the authors provide the first evidence that U4 release from U6 on the spliceosome is irreversible and, strikingly, that subsequent U5 release can be followed by simultaneous reassociation of U4 and U5, revealing for the first time spliceosome re-assembly after the presumed discard of the splicing substrate and concomitant disassembly of the spliceosome.

Overall, the presentation is clear and enjoyable to read, the quality of the work is high, and the results are convincing. The analysis has not only provided direct evidence for previously suspected features of spliceosome assembly (e.g., the irreversibility of U4 release, the dependence of NTC binding on U4 release) but also revealed unexpected evidence for an apparently ATP-dependent discard pathway for the tri-snRNP and provided the first evidence for repeated cycles of tri-snRNP sampling of a substrate – both with and without U4 release.

Additionally, while the approach of CoSMoS has been previously introduced by the authors, the successful and revealing application here to the activation step in splicing will underscore the power of this approach in investigating factor dynamics to the broader community – well beyond the splicing field.

Essential revisions:

1) A main question here is what underlies the apparent ATP-dependence of U4/U5(U6) dissociation. Does the fast phase at high ATP reflect a direct role for ATP in discard, as the authors imply, or do differences in assembly at high ATP lead to a spliceosome in which the tri-snRNP dissociates more rapidly but without an ATP-dependence? To test directly whether the simultaneous release of U4 and U5 is ATP dependent, could the authors easily wash away (low) ATP in the middle of an experiment, after assembly on substrate, and assay for a decrease in the rate of U4/U5 dissociation? If not, minimally, this ambiguity should be acknowledged in the manuscript.

2) It seems that the data shown in Figure 6 represent a subset of the data from the experiment shown in Figure 3B. What proportion of the RNAs undergo multiple tri-snRNP binding events?

3) Related to the above, can the authors provide any data in support of molecules that undergo multiple tri-snRNP binding events eventually proceeding to a functional splicing reaction, rather than being eventually discarded? In other words, how likely is it that such multiple association/dissociation events reflect a normal spliceosome assembly pathway? For example, is it possible to know which molecules become spliced? Could the authors simultaneously monitor release of the excised intron? Minimally, does the proportion of molecules that display multiple tri-snRNP binding events match an estimate of the number of molecules that undergo splicing in this setup? As is, the last line of the Abstract is not justified.

4) In the subsection “Spliceosomes remain dynamic even when stalled by limiting ATP” the authors discuss the need for binding of U1 and U2 prior to tri-snRNP binding. It should also be discussed, either here or later, that when tri-snRNP dissociates, these data do not appear to allow distinguishing between dissociation only of the tri-snRNP to leave a stable pre-spliceosome, or if U2 and/or U1 dissociate also, i.e. discard of the entire complex, and then assembly of a new spliceosome from scratch. If tri-snRNP discard leads back to the prespliceosome, then one would expect that tri-snRNP rebinding a second time would be kinetically faster than initial binding events. Can this be evaluated? (Similarly, in cases where U4 release is followed by U5 release and re-binding of the tri-snRNP, is rebinding faster than initial binding?)