Some group I introns have been shown to be selfsplicing in vitro, but perhaps all require proteins for splicing in vivo. Sequence differences affect the stability of secondary structures and may explain why some group I introns function efficiently without protein cofactors while others require them. The terminal intron of the cytochrome b pre-mRNA from yeast mitochondria needs a nucleus-encoded protein for splicing, even though it splices autocatalytically in high salt in vitro. This system

more »

... the advantage that the protein is specific for this intron, and yet the structure of the catalytically active RNA can be studied in its absence. We have modified the intron by chemical and enzymatic treatment in the presence and absence of the protein to determine the impact of the protein on the secondary and tertiary structures of the intron. We found protein-induced formation of secondary and tertiary structures within the intron, and the same structures also form in high salt autocatalytic conditions. We have also studied UV cross-links to determine those bases of the intron that interact directly with the protein and found that the protein contacts the intron most intimately at the structures denoted P1, L2, P4, and P6a. Group I introns have been shown to be self-splicing in vitro. The reaction requires Mg 2ϩ , which has at least a structural role and perhaps a catalytic role in splicing (1, 2). Guanosine is required to initiate the first of two transesterfication reactions that yield the ligated exons (3). The group I intron of the large subunit rRNA of Tetrahymena requires no protein for splicing, but many group I introns do require protein cofactors for splicing in vivo (4). Consequently, a comprehensive view of group I introns must explain the role of ancillary proteins. In addition, protein-assisted splicing of group I introns may serve as a model for the function of small ribonucleoproteins, which are thought to have a catalytic function in the splicing of introns in nuclear pre-mRNA. Simple systems exist for the study of a few group I introns that require at least one protein for splicing. The best studied of these proteins is CYT-18 from Neurospora crassa, which can function in the splicing of several group I introns (5). CYT-18, which is a mitochondrial tRNA synthetase, cannot enhance the splicing of all group I introns, but it recognizes a subset of structural elements common to a class of group I introns (6). CYT-18 is believed to stabilize the correct tertiary folding of the catalytic core of the introns it stimulates (7, 8). Proteins from Escherichia coli, especially ribosomal protein S12, have been shown to stimulate splicing of group I introns from bacteriophage T 4 . Coetzee et al. (9) have shown that these proteins facilitate splicing by relatively weak, nonspecific binding to intron RNA. They suggest that S12 acts as an RNA chaperone, preventing formation of inhibitory structures and facilitating helix formation and perhaps tertiary interactions. Group I introns share common secondary and tertiary structures that have been analyzed based on phylogenetic and mutagenic analysis (10 -12). The catalytic core of the intron contains the guanosine binding site and the sites of primary sequence homology common among group I introns. Other conserved structures are the P1/P10 helices, which consist of the internal guide sequence base paired with the 5Ј-and 3Ј-splice junctions, respectively. It is not clear why a group of such structurally similar introns have a diverse requirement for protein cofactors when the basic chemical reaction mechanism of splicing is identical for all group I introns. Despite conservation of secondary structure, these introns vary considerably in sequence (13). Sequence differences that affect the stability of RNA helices may explain why some group I ribozymes function efficiently without protein cofactors while others require them. We have studied an intron that needs a single protein for activity in vivo. Splicing of the fifth intron of the cytochrome b pre-mRNA (bI5) from yeast mitochondria requires the nucleusencoded Cbp2 protein (14, 15), but in high salt this intron is self splicing (16). This system has the advantage that the protein is specific for this intron and yet the structure of the catalytically active RNA can be studied in its absence. We have modified the intron by chemical and enzymatic treatment in the presence and absence of the protein to determine what impact Cbp2 has on the secondary and tertiary structures of bI5. We have also performed UV cross-linking experiments to help determine those bases that interact directly with Cbp2. This is an effort to probe the RNA-protein interactions required for group I intron splicing. Our results suggest that Cbp2 stabilizes AU-rich helices and also forces the intron to assume the same tertiary structure under physiological conditions as it does in the high salt, autocatalytic conditions. EXPERIMENTAL PROCEDURES Expression of Cbp2 Protein-Plasmid pET3a-CBP2 was constructed by converting the ATG of the CBP2 gene to an NdeI site by directed mutagenesis and cloning the NdeI-SnaBI fragment between the NdeI and BamHI sites of pET3a (17). An overnight culture of BL21(DE3) containing this plasmid was used to inoculate 500 ml of LB/ampicillin. Cells were cultured at 37°C until A 550 reached 0.35. Expression of the