T he spliceosome is a massive assembly of 5 RNAs and many proteins that, together, catalyze precursor-mRNA (pre-mRNA) splicing. The splicing mechanism involves 2 steps: (i) cleavage of the 5Ј exon-intron phosphodiester bond and formation of a new 2Ј-5Ј bond resulting in an intron ''lariat,'' and (ii) exon ligation (Fig. 1A). This 2-step phosphoryl transfer mechanism is suspiciously identical to the reaction catalyzed by the group II self-splicing introns, which are ribozymes. The precedent for a ribozymecatalyzed splicing mechanism, combined with the fact that the spliceosome has an RNA core that has been highly conserved for Ͼ1 billion years, led researchers to hypothesize that the spliceosome may use an RNA-based catalytic mechanism (1). However, the spliceosome as ribozyme hypothesis has been exceedingly difficult to prove, for 2 major reasons. First, the spliceosome contains many proteins that are essential for splicing (2). Second, the uncatalyzed rate of RNA ligation in vitro is significant when measured over many hours (3). This latter point makes it difficult to establish whether an inefficient reaction is actually caused by a catalytic rate enhancement or occurs spontaneously because of template-driven proximity. This is a particular concern for 1-step phosphoryl transfer reactions. Furthermore, it is possible to serendipitously obtain unrelated activities from RNAs, perhaps because a significant region of RNA conformational space retains some degree of catalytic activity (4). Therefore, it is important to be very cautious when interpreting an inefficient RNA-based reaction. The relevance of using engineered, protein-free RNA systems to study splicing has been recently argued (5, 6). In a recent issue of PNAS, Valadkhan et al. (7) demonstrated that an RNA complex derived from the spliceosome can perform a reaction that resembles splicing (Fig. 1B). This finding is significant because the reaction appears to be identical to the second step of splicing and is Ϸ10-fold more efficient than previous studies that used similar ).An interesting aspect of this new reaction is that the first step seems to be generated via hydrolysis rather than by a 2Ј-5Ј branching reaction (Fig. 1B). This finding is different from spliceosomecatalyzed pre-mRNA splicing (Fig. 1 A), but has precedence among the group II ribozmes (11). The second step of the reaction, however, results in exon ligation and formation of a new 5Ј-3Ј bond, just as in splicing. This reaction also has sequence requirements that closely parallel those of the spliceosome and requires magnesium. These data support, but do not prove, the idea that the reaction is related to pre-mRNA splicing. In future experiments, it will be important to investigate this reaction further for its mechanistic parallels to splicing. Currently, it is not clear how the substrate RNAs are coordinated in the reaction, and whether or not the system faithfully recapitulates the interactions in the spliceosome remains to be determined. Another important question concerns how the cata...