Competition between the ATPase Prp5 and branch region-U2 snRNA pairing modulates the fidelity of spliceosome assembly

Abstract

ATPase-facilitated steps during spliceosome function have been postulated to afford opportunities for kinetic proofreading. Spliceosome assembly requires the ATPase Prp5p, whose activity might thus impact fidelity during initial intron recognition. Using alanine mutations in S. cerevisiae Prp5p, we identified a suboptimal intron whose splicing could be improved by altered Prp5p activity and then, using this intron, screened for potent prp5 mutants. These prp5 alleles specifically alter branch region selectivity, with improved splicing in vivo of suboptimal substrates correlating with reduced ATPase activity in vitro for a series of mutants in ATPase motif III (SAT). Because these effects are abrogated by compensatory U2 snRNA mutations or other changes that increase branch region-U2 pairing, these results explicitly link a fidelity event with a defined physical structure, the branch region-U2 snRNA duplex, and provide strong evidence that progression of the splicing pathway requires branch region-U2 snRNA pairing prior to Prp5p-facilitated conformational change.

Figure 1. Genetic screen for prp5 alleles that alter splicing substrate selectivity

A. ATPase-associated steps during splicing could offer opportunities for kinetic discrimination of suboptimal pre-mRNA substrates (Burgess and Guthrie, 1993). B. Schematic of S. cerevisiae Prp5p, indicating conserved motifs within its ATPase/helicase domain. Regions mutated in this study and the prp5-1 mutation are indicated. C. (Upper) Schematic of ACT1-CUP1 pre-mRNA reporter used for monitoring splicing effects in vivo (Lesser and Guthrie, 1993). (Lower) Growth on copper of strains carrying one reporter mutant, U257C, is slightly improved by alanine mutations in the SAT motif of Prp5p. D. Strategy of genetic screen for more-potent prp5 alleles that alter splicing of the U257C reporter, and for alleles conferring conditional and lethal growth defects. The selected prp5 alleles are listed in Table 1. E. Analysis of seven prp5 alleles that alter splicing of U257C. (Upper) Graph of maximum copper concentration that allows cell growth. (Middle) Cell growth on selected copper plates, as indicated. (Lower) Analysis of in vivo RNA levels for ACT1-CUP1 reporter by primer extension. Primer complementary to the 3′ exon was used to monitor levels of pre-mRNA, mRNA, and lariat intermediate (indicated by icons). F. Glycine at position T448 (SAG) is the critical change within motif III that confers improved splicing upon the U257C reporter.

Figure 2. prp5 alleles modulate substrate selectivity primarily at two positions in the branch-flanking region

A. Growth in the presence of copper (upper) and analysis of mRNA levels in vivo (lower) indicate that prp5 alleles improve splicing of U257A and U257G reporters, as well as the previously-tested U257C mutant. A longer exposure is provided to more clearly show mRNA levels. Primer complementary to U6 snRNA was used in the same reaction as a loading control. Quantitations: the mRNA level in each lane as determined by phosphorimaging was normalized to the U6 snRNA loading control, then divided by the normalized mRNA level in the strain containing wt-PRP5 and wt ACT1-CUP1 reporter to give relative mRNA levels. B. Schematic of base-pairing interactions between U2 snRNA and the branch region of ACT1-CUP1 pre-mRNA, indicating branch region mutations used in this figure and Fig. S1. C. prp5-TAG and -N399D alleles improve growth on copper and in vivo mRNA level of reporters mutated at position A258 in the branch region. D. prp5 alleles do not improve growth on copper for all single mutations that are outside of branch region U257 and A258. One mutant at each position is shown; for other mutants at these positions, see Fig. S1. E. Reporters carrying double mutations at branch region positions other than U257 and A258 also are improved by the prp5-N399D allele. F. prp5 alleles increase the steady-state levels of suboptimal branch region pre-mRNA present in splicing complexes in vivo. (Left) Quantitative-RT-PCR analysis of anti-Flag-Prp5p-purified complexes indicates an increased association of U257C pre-mRNAs with splicing complexes by prp5-TAG and prp5-N399D alleles. The level of pre-mRNA was normalized to U6 snRNA. Each bar is from a triplet assay; error bars represent standard deviations. (Right) Visualization of total Flag-Prp5p and U2A’ (Lea1p) by western blot indicates that levels of Prp5-TAG and Prp5-N399D proteins are similar to wt-Prp5p.

Figure 3. Suboptimal branch region–U2 snRNA base-pairing is the critical feature suppressed by prp5 alleles

A. prp5 alleles improve splicing of a wt intron in the presence of U2 snRNA mutants. B. prp5 alleles provide little or no additional improvement to the partial suppression by compensatory changes in the U2 branch-pairing region of defects due to U257C and A258C mutations. C. Additional potential base-pairs between U2 snRNA and the intron branch region partially suppress the U257C defect; prp5 alleles provide no additional improvement. A–C: (upper) Schematic of base-pairing interactions between U2 snRNA and the intron branch region, indicating U2 snRNA and ACT1-CUP1 reporter mutations used below. (middle) Graphical representation of maximum copper concentration that allows cell growth. (lower) Analysis by primer extension of in vivo RNA levels for ACT1-CUP1 reporters as in Fig. 1E. For panels A and B, all tested strains carried both wt-U2 and mutant U2 snRNA alleles on CEN plasmids; all strains were otherwise isogenic and contained disrupted chromosomal loci of U2 snRNA and PRP5. mRNA quantitations are as in Fig. 2A except divided by the normalized mRNA level in the strain containing wt-PRP5 and wt-U2 snRNA.

Figure 4. ATP hydrolysis by prp5-SAT alleles correlates with their phenotypes

A. In vitro ATPase assay of Prp5p mutants using His₆-tagged proteins purified from E. coli (Coomassie blue-stained gel, upper insert). All values, with standard errors, are averaged from at least two parallel reactions. B. Relative phenotypes and ATPase activities of Prp5p mutants. Calculation of relative ATPase activities was based on the least squares fitted lines in panel A. C. CUS2 deletion suppresses the cs phenotype of prp5-GAR (Left) and rescues the otherwise-lethal phenotype of prp5-PAE (Right). D. CUS2 deletion does not alter splicing fidelity. In the presence of wt-PRP5, growth on copper plates of strains carrying a variety of intron mutants did not change when CUS2 was deleted. E. In the absence of CUS2, the ATPase activity of prp5-SAT mutants correlates with ability to improve splicing of suboptimal intron substrates. Deletion of CUS2 improves growth of strains carrying prp5 alleles with low ATPase activity, but not prp5 alleles with relatively high ATPase activity, indicated by wt and U257C reporters (upper). In the absence of CUS2, lowered ATPase activity of prp5-SAT mutants correlated with stronger improved splicing of the U257C mutant. Relative improvement for each prp5 allele was calculated by normalizing to the growth of wt reporter (lower left) and plotted against the ATPase activity given in Fig. 4B (lower right).

Figure 5. Schematic of outcomes of Prp5p-mediated spliceosomal transition, depending on relative rates of branch region–U2 snRNA pairing and conformational change by Prp5p

A. Productive pathway of splicing, where k_on is fast and k_off is slow with wild-type substrate, and therefore duplex formation is faster than the rate of conformational change by Prp5p. B. Alternative outcomes for suboptimal substrates (X’s indicate a suboptimal intron branch-site-flanking region). k_on is slow and k_off is fast in spliceosomes formed on suboptimal substrates. In the case of wild-type Prp5p, conformational change occurs prior to duplex formation, resulting in Prp5p-mediated transition prior to branch region–U2 snRNA pairing (upper path). C. In the presence of mutations of the Prp5p ATPase domain, the rate of conformational change is also slowed, resulting in longer opportunity to achieve branch region–U2 snRNA pairing and allowing progression of the productive assembly pathway for suboptimal substrates (lower path).