Volume 21, no. 6, p. 1942-1952, 2001. Our paper described the in vivo and in vitro splicing phenotypes of a number of mutants with alterations in the 3′ splice site of a U12-dependent intron. In further work with these mutants, we have discovered that two mutants were not as described in the paper. The mutant named P120 +18 AG was, in fact, identical to the mutant named P120 +18 AC, while the mutant named P120 +27 AC contained an additional mutation at the 5′ splice site that inactivated U12-dependent splicing of this intron. We have reconstructed and analyzed these two mutants using the same protocols as described in the paper.
Our new results show that the phenotypes of the authentic P120 +18 AG mutant are very similar to those shown in the paper and to those of the P120 +18 AC mutant. In contrast, the phenotypes of the authentic P120 +27 AC mutant differed from those reported in our paper in several respects. First, a low but detectable level of in vivo and in vitro splicing occurred at the +27 3′ splice site in addition to activation of the cryptic UU 3′ splice site at +12 as seen in the +18 mutants. This is in contrast to the reported phenotype, where no detectable splicing was seen. Second, the authentic P120 +27 AC mutant was able to support formation of the spliceosomal A and B complexes in vitro. This differs from the mutant shown in the paper, which was unable to form either spliceosomal complex in vitro.
These results contradict our original conclusion that spliceosome formation in the U12-dependent splicing system is blocked if the 3′ splice site is located too far away from the branch site. Instead, it appears that spliceosome formation can occur in the absence of a properly positioned 3′ splice site sequence. The new data remove the apparent conflict between the result that a truncated RNA without a 3′ splice site could form spliceosomes while a mispositioned 3′ splice site appeared to block spliceosome formation. Our current understanding is that spliceosomal A and B complexes require only an active 5′ splice site and a branch site to form. These results do not contradict our primary conclusions in the paper that U12-dependent 3′ splice sites are functionally constrained to a small window downstream of the branch site with a strong preference for a spacing of 12 nucleotides and that the branch site is recognized in the absence of an active 3′ splice site.