A Tyrosyl-tRNA Synthetase Protein Induces Tertiary Folding of the Group I Intron Catalytic Core

J. Mol. Biol. (1996) 257, 512–531

A Tyrosyl-tRNA Synthetase Protein Induces Tertiary Folding of the Group I Intron Catalytic Core Mark G. Caprara, Georg Mohr and Alan M. Lambowitz* Departments of Molecular Genetics, Biochemistry and Medical Biochemistry The Ohio State University 484 West Twelfth Avenue Columbus, Ohio 43210-1292 USA

The Neurospora crassa mitochondrial tyrosyl-tRNA synthetase (CYT-18 protein) functions in splicing group I introns. We have used chemical-structure mapping and footprinting to investigate the interaction of the CYT-18 protein with the N. crassa mitochondrial large subunit ribosomal RNA (mt LSU) and ND1 introns, which are not detectably self-splicing in vitro. Our results show that both these non-self-splicing introns form most of the short range pairings of the conserved group I intron secondary structure in the absence of CYT-18, but otherwise remain largely unfolded, even at high Mg2 + concentrations. The binding of CYT-18 promotes the formation of the extended helical domains P6a-P6-P4-P5 (P4-P6 domain) and P8-P3-P7-P9 (P3-P9 domain) and their interaction to form the catalytic core. In iodine-footprinting experiments, CYT-18 binding results in the protection of regions of the phosphodiester backbone expected for tertiary folding of the catalytic core, as well as additional protections that may reflect proximity of the protein. In both introns, most of the putative CYT-18 protection sites are in the P4-P6 domain, the region of the SU intron previously shown to bind CYT-18 as a separate RNA molecule, but additional sites are found in the other major helical domain in P8 and P9 in both introns and in L9 and P7.1/P7.1a in the mt LSU intron. Protease digestion of the CYT-18/intron RNA complexes results in the loss of CYT-18-induced RNA tertiary structure and splicing activity. Considered together with previous studies, our results suggest that CYT-18 binds initially to the P4-P6 region of group I introns to form a scaffold for the assembly of the P3-P9 domain, which may contain additional binding sites for the protein. A three-dimensional model structure of the CYT-18-binding site in group I introns indicates that CYT-18 interacts with the surface of the catalytic core on the side opposite the active-site cleft and may primarily recognize a specific three-dimensional geometry of the phosphodiester backbone of group I introns. 7 1996 Academic Press Limited

*Corresponding author

Keywords: catalytic RNA; Neurospora crassa; ribozyme; RNA splicing; RNA structure

Introduction The splicing of group I introns in vivo requires protein factors to promote formation of the catalytically-active RNA structure (Mannella et al., 1979; Lazowska et al., 1980; reviewed by Lambowitz Abbreviations used: CIA, chloroform:isoamyl alcohol; CMCT, 1-cyclohexyl-3-(2-morpholine-ethyl) carbodiimide metho-p-toluenesulfonate; DEPC, diethyl pyrocarbonate; DMS, dimethyl sulfate; LSU, large subunit ribosomal RNA; mt, mitochondria; ORF, open reading frame; PEI, polyethyleneimine; TyrRS, tyrosyl-tRNA synthetase. 0022–2836/96/130512–20 $18.00/0

& Perlman, 1990). Biochemical-genetic analyses of mitochondrial (mt) RNA splicing in Neurospora crassa and Saccharomyces cerevisiae have shown that some of these protein factors are encoded by host chromosomal genes, whereas others, maturases, are encoded by the introns themselves. Notably, the splicing factors encoded by host chromosomal genes include aminoacyl-tRNA synthetases and other cellular RNA binding proteins that bind specifically to group I intron RNAs. We suggested that these proteins may have adapted to function in splicing by recognizing structural features in group I introns that resemble those in their normal RNA substrates (Lambowitz & Perlman, 1990). Recently, 7 1996 Academic Press Limited

Protein-induced Folding of Group I Introns

other proteins with a more general affinity for RNA have been shown to function as ‘‘RNA chaperones’’ to promote folding of helical regions of catalytic RNAs (Herschlag et al., 1994). The Escherichia coli ribosomal protein S-12, which does not bind specifically to group I introns, apparently functions in this manner to facilitate folding of the phage T4 td intron in vitro (Coetzee et al., 1994). The RNA chaperones act during the folding process and their continued association with the intron RNA is not required to maintain the active RNA structure. Among the most extensively characterized group I intron splicing factors is the N. crassa mt tyrosyl-tRNA synthetase (mt TyrRS) or CYT-18 protein. This protein functions in splicing the mt large subunit ribosomal RNA (LSU) intron and other group I introns in N. crassa mitochondria (Collins & Lambowitz, 1985; Akins & Lambowitz, 1987; Wallweber & Lambowitz, unpublished results) While additional components are required for efficient splicing in vivo, purified CYT-18 protein isolated from mitochondria or synthesized via an expression plasmid in E. coli is by itself sufficient to splice the N. crassa mt LSU and other group I introns in vitro (Majumder et al., 1989; Kittle et al., 1991; Mohr et al., 1992, 1994). The mt TyrRSs, including the CYT-18 protein, are class I synthetases homologous to well-studied bacterial TyrRSs (Lambowitz & Perlman, 1990; Giege et al., 1993). Like the bacterial TyrRSs, the CYT-18 protein functions as a homodimer, with each CYT-18 homodimer having one binding site for tRNATyr or the group I intron RNA (Saldanha et al., 1995). The regions of the CYT-18 protein required for splicing overlap those required for binding tRNATyr and include a small, idiosyncratic N-terminal domain not found in the homologous bacterial TyrRSs, possibly parts of the nucleotide-binding region, and the C-terminal tRNA-binding domain (Cherniack et al., 1990; Kittle et al., 1991). Furthermore, a group I intron RNA was found to be a competitive inhibitor of aminoacylation, providing direct evidence that the group I intron and tRNA compete for the same or an overlapping binding site (Guo & Lambowitz, 1992). According to structure modelling, the group I intron catalytic core consists of two extended helical domains, one consisting of coaxially-stacked helices P6a, P6, P4 and P5 (the P4-P6 domain), and the other consisting of P8, P3, P7 and P9 (the P3-P9 domain; Michel & Westhof, 1990; Jaeger et al., 1991). The two helical domains create a cleft that constitutes the active site of the intron and contains binding sites for the guanosine cofactor and the P1 and P10 helices, which contain the 5'- and 3'-splice sites, respectively. The tertiary folding of group I introns is Mg2+-dependent, and one or two Mg2+ ions bound at the active site are believed to participate directly in catalysis (Cech et al., 1992). In addition to N. crassa mt group I introns, CYT-18 binds specifically to a variety of group I introns from other organisms, including introns of different structural subclasses and introns that are

513 self-splicing in vitro (Guo & Lambowitz, 1992). The dissociation constants (Kd s) at 25°C for the complexes between CYT-18 and the mini-derivatives of the N. crassa mt LSU and ND1 introns used in the present work are <0.3 pM, and the Kd s for other group I introns range from <0.3 to 420 pM (Saldanha et al., 1995; unpublished results). In general, the group I introns that bind tightly to CYT-18 have relatively little sequence similarity, suggesting that the protein recognizes conserved three-dimensional features of these introns (Guo & Lambowitz, 1992). The CYT-18-binding site in the N. crassa mt LSU intron was localized by testing in vitro transcripts containing different segments of the intron for binding to CYT-18 in a nitrocellulose-filter-binding assay (Guo & Lambowitz, 1992; Saldanha et al., unpublished results). It was found that CYT-18 binds strongly to a small RNA encompassing the P4-P6 domain, excluding P6b (Kd = 160 pM), but that sequences in the P7-P9 region of the other major helical domain increase the binding affinity by 0sevenfold (Kd = 22 pM). An E. coli genetic assay with the phage T4 td intron showed that CYT-18 could suppress structural mutations in different parts of the catalytic core, but not outside of the catalytic core region (e.g. in P1 and P2; Mohr et al., 1992). The structural defects suppressed by CYT-18 included mutations in both major helical domains of the catalytic core, as well as in J6/7, a junction region believed to play a key role in establishing the correct relative orientation of the two helical domains (Myers et al., unpublished results). Together, these findings suggest that CYT-18 functions in splicing by binding to each of the two major helical domains of the catalytic core and stabilizing them in the correct relative orientation to form the active site of the intron. Recent studies extended these findings by showing that CYT-18 functions similarly to a subgroup-specific peripheral RNA structure, P5abc, which interacts with the P4-P6 domain to stabilize the catalytic core of the Tetrahymena LSU intron (Mohr et al., 1994). A derivative of the Tetrahymena intron lacking P5abc is unable to self-splice at low Mg2+, but splicing is restored either by adding P5abc RNA in trans (van der Horst et al., 1991) or by adding CYT-18 protein (Mohr et al., 1994). Chemical-structure mapping showed directly that the binding of CYT-18 could substitute for the P5abc structure to induce the tertiary folding required for catalytic activity (Mohr et al., 1994). To understand precisely how the CYT-18 protein functions in splicing, we have under-taken a detailed investigation of the interaction of CYT-18 with group I introns using a variety of experimental approaches. In this work we report chemical-structure mapping and footprinting experiments investigating the interaction of CYT-18 with the N. crassa mt LSU and ND1 introns, which are not detectably self-splicing in vitro. Our results show that both introns form much of the conserved group I intron secondary structure in the absence of CYT-18, but

514

Protein-induced Folding of Group I Introns

Figure 1. Modification of the N. crassa mt LSU and ND1 introns with DMS, CMCT and DEPC. Modification reactions with the mt LSU (A) and ND1 (B) introns were carried out using in vitro transcripts pBD5A/BanI and pND1m/NdeI, respectively. RNAs were modified with DMS (lanes 1 to 4), CMCT (lanes 5 to 8) or DEPC (lanes 9 to 12) in reaction medium containing 0, 5, 15 or 25 mM Mg2+ at 37°C or 55°C. To detect sites of modification, the RNAs were reverse transcribed, using 5' end-labeled primers LS2 and N2 for the mt LSU and ND1 introns, respectively, and the products were analyzed by electrophoresis in 6% polyacrylamide/7 M urea gels, followed by autoradiography. RNA, RNA incubated in parallel in reaction medium containing 5 mM Mg2+ at 37°C in the absence of modifying reagent; lanes 1, 5 and 9, RNAs modified under denaturing conditions (0 mM Mg2+ ) at 55°C; lanes 2 to 4; 6 to 8 and 10 to 12, RNAs modified in reaction medium containing the indicated Mg2+ concentrations (mM) at 37°C. Dideoxy-sequencing ladders (G, U, C, A) obtained from pBD5A or pND1m using the same primers were run in parallel lanes.

that CYT-18 binding is required to induce tertiary structure required for splicing activity. Together with previous findings, our results suggest that CYT-18 binds initially to the P4-P6 region of group I introns to form a scaffold for the assembly of the P3-P9 domain. A three-dimensional model structure of the CYT-18-binding site indicates that CYT-18 interacts with each of the two major helical domains on the surface of the catalytic core opposite the active-site cleft and may primarily recognize a specific three-dimensional geometry of the phosphodiester backbone of group I introns.

Results Experimental strategy The N. crassa mt LSU and ND1 introns, chosen for detailed analysis, are not detectably self-splicing and are dependent on the CYT-18 protein for splicing in vivo and in vitro (Garriga & Lambowitz, 1986; Guo et al., 1991; Wallweber & Lambowitz, unpublished results). Because the mt LSU and ND1 introns belong to different structural subclasses (IA1 and IB2, respectively; Michel & Westhof, 1990), we anticipated that comparative analysis would aid

in identifying conserved structural features required for interaction with the CYT-18 protein. For both introns, the structural analysis was facilitated by the use of mini-derivatives that are still spliced in a CYT-18-dependent manner (Guo et al., 1991; Wallweber & Lambowitz, unpublished results). The 388 nt mini-derivative of the mt LSU intron is the standard substrate used in previous studies (Guo et al., 1991). The 197 nt mini-derivative of the ND1 intron lacks peripheral structures and is among the smallest group I introns described thus far (Wallweber & Lambowitz, unpublished results). Structure-mapping and footprinting experiments focused on the catalytic core region (P3 to P9), which was previously shown to contain the CYT-18-binding site in the mt LSU intron (Guo & Lambowitz, 1992). To obtain a detailed overview of the interaction, the experiments were carried out using a combination of base-specific reagents (dimethyl sulfate (DMS), 1-cyclohexyl-3-(2-morpholine-ethyl) carbodiimide metho-p-toluenesulfonate (CMCT) and diethyl pyrocarbonate (DEPC)) and reagents that cleave the phosphodiester backbone (Fe(II)-EDTA, iodine and Pb2+ ) at several different Mg2+ concentrations. Both the mt LSU and ND1 introns show optimal CYT-18-dependent splicing at

515

Protein-induced Folding of Group I Introns

5 mM Mg2+, and they are not detectably selfsplicing in vitro at any Mg2+ concentration (Guo et al., 1991; Wallweber & Lambowitz, unpublished results). Chemical-structure mapping with DMS, CMCT and DEPC In the initial experiments, chemical-structure mapping was performed with DMS, CMCT and DEPC. These reagents assess secondary structure formation and also provide tertiary structure information in cases where single-stranded junction or loop regions are internalized or participate directly in tertiary interactions (e.g. Pyle et al., 1992; Banerjee et al., 1993). Modified nucleotides were detected by reverse transcription using 5' endlabeled primers and quantitated by b-scanning (see Materials and Methods). Reverse transcription detects DMS modifications at the N1 position of unpaired adenines and the N3 position of unpaired cytosines. DEPC modifies primarily the N7 position of adenines, but also modifies other bases to some

extent, while CMCT modifies unpaired uridines and guanosines at the N3 and N1 positions, respectively (Ehresmann et al., 1987). Figure 1 shows representative DMS, CMCT- and DEPCmodification patterns for each N. crassa intron in the absence of CYT-18, and Figure 2 summarizes the modifications detected in reaction medium containing 5 mM Mg2+. For both introns, most of the bases comprising the short range pairings P4, P5, P6, P6a, P8 and P9 showed some protection relative to denaturing conditions in splicing buffer containing 5 to 25 mM Mg2+, indicating that most of the conserved group I intron secondary structure can form in the absence of CYT-18. However, the extent of modification within the secondary structure appears to be greater than in self-splicing group I introns under comparable conditions (cf Banerjee et al., 1993; Jaeger et al., 1993; von Ahsen & Noller, 1993). Thus, both N. crassa introns showed moderate modifications throughout the secondary structure, as well as some strong modifications within helices and at the terminal base-pairs of helices (see the legend to

Figure 2A legend opposite

516

Protein-induced Folding of Group I Introns

Figure 2. Summary of DMS, CMCT and DEPC-modification data for the N. crassa mt LSU and ND1 introns. Secondary structure models of the mt LSU (A) and ND1 (B) introns showing sites of modification in reaction medium containing 5 mM Mg2+. The inset shows modification data for the mt LSU intron P6b region. Circle, modification with DMS; square, modification with CMCT; triangle, modification with DEPC. A filled symbol indicates strong modification (band intensity 50% or more of the strongest modification in the denaturing lane) and an open symbol indicates moderate modification (band intensity 10% to 49% of the strongest modification in the denaturing lane). Strongly modified bases within predicted helices were: mt LSU intron–A111 in P4, U193 and U203 in P6b, U249 in P7 and U278 in P7.1; ND1 intron–A31 and A116 in P3, U79 and A80 in P6a and A104 in P7. Strongly modified bases at the ends of predicted helices were: mt LSU intron–A112 in P4, U185 and A194 in P6b and A244 in P7; ND1 intron–C118 in P3, U50 and A62 in P5, G103 in P7, and A119 in P8. Strongly modified bases predicted to be bulged out of helices Figure 2B were: mt LSU intron–A177, U183 and U189 in P6b; ND1 intron–A104 in P7. Modified bases at positions protected by tertiary interactions in self-splicing group I introns were: mt LSU intron–A105 (J3/4-P6 base-triple), U241 and C242 (J6/7-P4 base-triple), A113, A114, A331, A332, G333 and A334 (J4/5 and J8/7 interactions with P1), A157, G158, A160, U161, A262, U263, U266 (P11 pairing) and G355 and A356 (L9-P5 interaction); ND1 intron–A35 and U36 (J3/4-P6 base-triple), U100 and C101 (J6/7-P4 base-triple), A44, A45, A150 and A151 (J4/5 and J8/7 interactions with P1), and A174 and A175 (L9-P5 interaction). Bases that showed increased protection at 15 or 25 mM Mg2+ were: mt LSU intron–A111, A112, A141 and A142 in P4, A114 in J4/5, A123 in L5, A130, A131 and A134 in P5, A147 in J6a/6b, C242 and A243 in J6/7, A244 in P7, A261, A262 and U266 in L7.1a, A274 in J7.1a/7.1, U280 and C286 in J7.1/3, A287 in P3, A334 in J8/7 and A345 in J7/9; ND1 intron–A33, A34, A35 and U36 in J3/4, U37, G41 and G42 in P4, A43 and A44 in J4/5, U50 in P5, U51, U52, G53, A54, G55 and U58 in L5, G158, U159 and C160 in P7, and A162, A163 and A164 in J7/9. Bases that showed increased modification at 15 or 25 mM Mg2+ were: mt LSU intron–A181 in P6b, C245 and A250 in P7, A252 and A253 in P7.1, A258 in J7.1/7.1a, A261 and A267 in L7.1a, U272 and A273 in J7.1a/7.1, and U329 and G333 in J8/7; ND1 intron–A116 and C118 in P3, C136 in L8, A151 in J8/7, A162, A163 and A164 in J7/9, and C179 in P9. Bases whose modification could not be assessed because of strong stops in the reverse transcription reaction were; mt LSU intron–C109 in P4, C159 in J6a/6b, C238 and G239 in P6, C265 in L7.1a, C277 in P7.1 and A360 in P9; ND1 intron–C30 and C32 in P3, C48 and C49 in P5, C70 and A71 in P4, A75 in J6/6a, A146 in P8, U169 and A170 in P9, and G172 and U173 in L9. Regions not included in the analysis because of difficulties in assigning bands were: mt LSU intron–U217-A227 in P6b[3']-P6a[3'] and C290-G314 in P3[3']-P8[3']; ND1 intron–A81-A93 in P6a and A128-G135 in P8. Abbreviations: 5' and 3' SS, 5'- and 3'-splice sites, respectively. Nucleotides in intron and exons are shown in capital and lower case letters, respectively.

Figure 2). In addition, several strong modifications occurred within P3 of the ND1 intron (A31, A116, C118) and P7 of both introns (mt LSU: A244, U249; ND1: G103, A104), suggesting that these long-range pairings do not form properly in the absence of the protein. The same patterns of strong modifications were found with RNAs subjected to the renaturation protocols of Celander & Cech (1991) and Jaeger et al. (1991; data not shown). Most of the modifications within predicted helical regions were

not suppressed at the higher Mg2+ concentrations, and some residues showed enhanced modification at higher Mg2+ (see the legend to Figure 2). Beyond the secondary structure, the modification patterns also show that both N. crassa introns are deficient in forming the conserved group I intron tertiary structure. In self-splicing group I introns, the junction and loop regions J3/4, J4/5, J6/7, J8/7 and L9 are largely protected from base modification by tertiary structure under splicing conditions (Pyle

517

Protein-induced Folding of Group I Introns

Figure 3. Fe(II)-EDTA-cleavage patterns for the N. crassa mt LSU and ND1 introns. Fe(II)-EDTA cleavage of the mt LSU (A) and ND1 (B) introns was carried out using 5' end-labeled in vitro transcripts pBD5A/BanI and pND1m/NdeI, respectively, in reaction medium containing 0, 5, 15 or 25 mM Mg2+ at 37°C. The cleavage products were analyzed by electrophoresis in a 6% polyacrylamide/7 M urea gel, followed by autoradiography. Lanes OH−, G and A, 5' end-labeled RNA subjected to partial alkaline hydrolysis or partial digestion with RNase T1 or U2, respectively; RNA, RNAs incubated in parallel in reaction medium containing 5 mM Mg2+ at 37°C in the absence of Fe(II)-EDTA; lanes 1, RNAs treated in reaction medium without Mg2+ at 37°C (denaturing conditions); lanes 2 to 4, RNAs treated in reaction medium containing the indicated Mg2+ concentrations (mM) at 37°C.

et al., 1992; Banerjee et al., 1993; Jaeger et al., 1993; von Ahsen & Noller, 1993). By contrast, all these regions were strongly modified in the mt LSU intron, as were J3/4, J4/5 and J6/7 in the ND1 intron. J8/7 and L9 showed somewhat less modification in the ND1 intron, but modification in J8/7 of this intron increased at higher Mg2+ (see the legend to Figure 2). From the nucleotides that remain modified in the N. crassa introns, we infer that specific tertiary interactions do not form properly in the absence of CYT-18. These include the J3/4-P6 and J6/7-P4 base-triple interactions in both introns (Michel & Westhof, 1990; Michel et al., 1990; Green & Szostak, 1994), the L9-P5 interaction in both introns (Jaeger et al., 1994), and the P11 pairing between J6a/6b and L7.1a in the mt LSU intron (Jaeger et al., 1991; see Figure 2). In addition, bases in J4/5 and J8/7 that should be protected by interaction with the P1 helix (Michel & Westhof, 1990; Pyle et al., 1992) were modified in both introns, suggesting that docking of the 5'-splice site to the catalytic core does not occur efficiently in the

absence of CYT-18. Again, most of these modifications were not suppressed at higher Mg2+ concentrations, with a few exceptions noted in the legend to Figure 2. Thus, by contrast with self-splicing introns, Mg2+ is not by itself sufficient to promote formation of the active tertiary structure in the non-self-splicing N. crassa introns. Structure mapping with Fe(II)-EDTA The ability of the N. crassa introns to fold into the conserved tertiary structure was assessed further by cleavage-protection experiments with Fe(II)-EDTA, a reagent used extensively to monitor tertiary folding of self-splicing group I introns (e.g. Celander & Cech, 1991; Heuer et al., 1991; Weeks & Cech, 1995b). Fe(II)-EDTA in the presence of a reducing agent generates hydroxyl radicals that indiscriminantly cleave regions of the phosphodiester backbone exposed to solvent. With increasing Mg2+, self-splicing group I introns fold progressively into the conserved three-dimensional structure, and this

518

Protein-induced Folding of Group I Introns

Figure 4. DMS and DEPC modification of the N. crassa mt LSU and ND1 introns in the presence or absence of CYT-18 protein. Modification reactions for the mt LSU (A) and ND1 (B) introns were carried out using in vitro transcripts pBD5A/BanI or pND1m/NdeI, respectively. Complexes were formed by incubating 20 nM in vitro transcript with the indicated amounts of CYT-18 dimer (nM), as described in Materials and Methods. Free RNAs or complexes were modified with DMS (lanes 1 to 3) or DEPC (lanes 4 to 6) in reaction medium containing 5 mM Mg2+ at 37°C. Sites of modification were detected by reverse transcription as in Figure 1, using 5' end-labeled primers LS2 and N2 for the mt LSU and ND1 intron, respectively. RNA, control incubations of RNAs in the absence of modifying reagent as in Figure 1; lanes 1 and 4, RNAs modified in reaction medium containing 5 mM Mg2+ at 37°C in the absence of CYT-18; lanes 2, 3, 5 and 6, CYT-18/intron RNA complexes modified in reaction medium containing 5 mM Mg2+ at 37°C. Dideoxy-sequencing ladders (G, U, C, A) obtained from pBD5A or pND1m using the same primers were run in parallel lanes.

folding results in the internalization of the catalytic core, which becomes inaccessible to hydroxyl radical cleavage. Fe(II)-EDTA-cleavage patterns for the N. crassa introns are shown in Figure 3. By contrast with self-splicing introns, the cleavage patterns for the N. crassa introns are largely unchanged as the Mg2+ concentration is increased. Both introns remain almost completely unfolded in the absence of CYT-18, with their catalytic cores accessible to cleavage, even at 25 mM Mg2+. In the experiment shown in Figure 3, the mt LSU intron showed limited areas of increased protection in P6b, P8[3'] and J8/7, and both introns showed some enhanced cleavages at high Mg2+ (in P6 in the mt LSU intron and L6a in the ND1 intron), but these features were not observed reproducibly. The Fe(II)-EDTA cleavage patterns show again that Mg2+ is not by itself

sufficient to promote tertiary folding of the non-self-splicing N. crassa introns. DMS and DEPC modification in the presence of the CYT-18 protein As expected from the restoration of splicing activity, the binding of CYT-18 results in striking conformational changes in both N. crassa introns. These changes were assessed first by the effect of CYT-18 on the DMS and DEPC-modification patterns. The experiments were carried out in reaction medium containing 5 mM Mg2+, which is optimal for CYT-18 dependent splicing. (CMCT was not used in these experiments because the potassium borate buffer required for this reagent inhibited the protein-dependent splicing of the N. crassa introns; data not shown). Figure 4 shows

519

Protein-induced Folding of Group I Introns

the effect of CYT-18 on the DMS and DEPCmodification patterns in representative regions of the mt LSU and ND1 introns. The complete data for P3 to P9 are summarized in Figure 5, with bases showing increased protection in the presence of CYT-18 indicated by gray shading. The binding of CYT-18 resulted in increased DMS or DEPC protection at 66 sites in the mt LSU intron and 32 sites in the ND1 intron. The CYT-18-induced protections were found both within the conserved group I secondary structure, which appears to be stabilized by the protein, and in single-stranded junction and loop regions, expected for the formation of tertiary structure. The latter protections include those expected for the formation of the J3/4-P6 and J6/7-P4 base-triples in both introns, the L9-P5 interaction in both introns, and the P11 pairing in the mt LSU intron. In addition, nucleotides in J4/5 and J8/7 that have been predicted to interact with P1 were protected, providing evidence that the 5'-splice site can now dock with the catalytic core (see the legend to Figure 5 for details). After binding CYT-18, the DMS

and DEPC-protection patterns for the N. crassa introns were essentially the same as those resulting from secondary and tertiary structure in self-splicing group I introns. Thus, all of the observed base protections could be due to RNA structure, and there was no clear indication of additional protections that might reflect contact sites with the protein. Iodine cleavage in the presence of CYT-18 We initially attempted to investigate the effect of CYT-18 on the protection of the phosphodiester backbone by using Fe(II)-EDTA, but this was not possible because the CYT-18/intron RNA complexes were unstable in the presence of this reagent (data not shown). A satisfactory method proved to be iodine cleavage of phosphorothioate-substituted RNAs (Schatz et al., 1991; Rudinger et al., 1992). To monitor every phosphate position in each intron, we synthesized four separate transcripts, each substituted with 4% of a different phosphorothioate. We confirmed by nitrocellulose-filter-binding assays

Figure 5A legend opposite

520

Protein-induced Folding of Group I Introns

Figure 5. Summary of DMS and DEPC-modification data for CYT18/intron RNA complexes. Secondary structure models of the mt LSU (A) and ND1 (B) introns showing sites of modification in complexes formed between the 20 nM intron RNA and 40 nM CYT-18 dimer under the conditions of Figure 4. Shading indicates bases showing increased protection in the presence of the CYT-18 protein (50% or more reduction in band intensity). The inset shows modification data for the mt LSU intron P6b region. Circle, modification with DMS; triangle, modification with DEPC. A filled symbol indicates strong modification, and an open symbol indicates moderate modification, as defined in Materials and Methods. Bases within or at the ends of helices that were strongly modified in 5 mM Mg2+ and showed increased protection in the presence of CYT-18 were: mt LSU intron– A111 and A112 in P4, A244 in P7 and A194 in P6b; ND1 intron–A31, A116 and C118 in P3, A62 in P5, A80 in P6a, and G103 in P7. Bases bulged out of helices that were strongly modified in 5 mM Mg2+ and showed increased protection in the presence of CYT-18 were: mt LSU intron– A177 in P6b; ND1 intron–A104 in Figure 5B P7. Bases in junction and loop regions whose protection suggests tertiary structure formation in the presence of CYT-18 were: mt LSU intron–A105 (J3/4-P6 base-triple), U241 (J6/7-P4 base-triple), A113, A331, A332, G333 and A334 (J4/5 and J8/7 interactions with P1), A157, G158, A160 and A262 (P11 pairing), and G355 and A356 (L9-P5 interaction); ND1 intron–A35 (J3/4-P6 base-triple), C101 (J6/7-P4 base-triple), A44 and A45 (J4/5 interaction with P1) and A174 and A175 (L9-P5 interaction). Bases showing enhanced modification in the presence of CYT-18 were: mt LSU intron–A252, A253, U278 and U279 in P7.1, A267 in L7.1a, A268 in P7.1a and A274 and U275 in J7.1a/7.1; ND1 intron–none. Positions whose modification could not be clearly assessed were the same as in Figure 2. Abbreviations: 5' and 3' SS, 5'- and 3'-splice sites, respectively. Nucleotides in introns and exons are shown in capital and lower case letters, respectively.

that the 4% phosphorothioate-substitutions did not alter the affinity of the intron RNA for CYT-18 protein (data not shown). In addition, we confirmed that iodine and Fe(II)-EDTA gave similar cleavageprotection patterns with the L-21 ScaI version of the Tetrahymena ribozyme (data not shown), indicating that the data from the two approaches should be directly comparable. Figure 6 shows representative iodine-cleavage patterns for the two N. crassa introns in reaction medium containing 5 mM Mg2+. The complete data for P3 to P9 are summarized in Figure 7, with nucleotides whose 5' phosphate shows increased protection in the presence of CYT-18 indicated by black or gray shading. There are several points. First, as expected from the Fe(II)-EDTA experiments, the iodine-cleavage patterns show that both intron RNAs remain largely unfolded in the absence of CYT-18, with most of their phosphodiester backbones accessible to cleavage (Figure 6, lanes 2,

6, 10 and 14). Although a few Mg2+-dependent protections are seen in Figure 6, the only ones observed consistently were in P8[3'] in the mt LSU intron (see the legend to Figure 7). Protections in this region of the mt LSU intron were also observed in the Fe(II)-EDTA experiment shown in Figure 3 at 15 and 25 mM Mg2+ and could reflect limited RNA tertiary structure formation in the absence of the protein. Strikingly, the binding of CYT-18 resulted in protection of the phosphodiester backbone throughout the catalytic cores (Figure 7). Areas showing increased protection in both introns included P3, P4, P5, P6, P7, P8 and P9, as well as the junction and loop regions J3/4, J4/5, J5/4, J6/6a, J6/7, J8/7, L5 and L9. In the mt LSU intron, increased protection was also found in P6a, P7[3'] and P9[3'], and in the peripheral structures P6b and P7.1/P7.1a. Some of these additional protections in the mt LSU intron could reflect the presence of peripheral structures

Protein-induced Folding of Group I Introns

521

Figure 6. Iodine-cleavage patterns for the N. crassa mt LSU and ND1 introns in the presence or absence of CYT-18. Iodine-cleavage reactions for the mt LSU (A) and ND1 (B) introns were carried out using in vitro transcripts pBD5A/BanI or pND1m/NdeI, respectively, substituted with 4% of one phosphorothioate (A, C, U or G). Complexes were formed by mixing 20 nM in vitro transcript with the indicated amounts of CYT-18 dimer (nM), as described in Materials and Methods, and the free RNAs or complexes were subjected to iodine cleavage in reaction medium at 37°C. Sites of cleavage were detected by reverse transcription, using 5' end-labeled primers LS2 and N2 for the mt LSU and ND1 intron, respectively. RNA, control incubation of RNAs in reaction medium containing 5 mM Mg2+ at 37°C in the absence of iodine; lanes 1, 5, 9 and 13, RNAs cleaved in reaction medium without Mg2+ at 37°C (denaturing conditions); lanes 2, 6, 10 and 14, RNAs cleaved in reaction medium containing 5 mM Mg2+ at 37°C; lanes 3, 4, 7, 8, 11, 12, 15 and 16, CYT-18/intron RNA complexes cleaved in reaction medium containing 5 mM Mg2+ at 37°C. Dideoxy-sequencing ladders (G, U, C, A) obtained from pBD5A or pND1m using the same primers were run in parallel lanes. The sequencing ladder in panel B is taken from a longer exposure of the autoradiogram than the iodine-cleavage reactions.

not present in the ND1 intron. (Note: P8[5'], which has phosphate protections in the ND1 intron, could not be clearly surveyed in the mt LSU intron due to smeared primer extension products; see below and the legend to Figure 7.) Based on comparison with the published Fe(II)EDTA-cleavage patterns for self-splicing group I introns, many of the phosphate protections seen in the N. crassa introns can be accounted for by the internalization of the catalytic core accompanying the formation of the conserved group I intron tertiary structure (cf. Celander & Cech, 1991; Heuer et al., 1991; Weeks & Cech, 1995b). However, additional protections that may reflect proximity of the CYT-18 protein were found in P3[5'], P4, P5, L5, P6[5'], P8[3'] and P9[5'] in both introns, in P6a, P6b, P7.1/P7.1a and L9 in the mt LSU intron, and in P8[5'] in the ND1 intron (black shading in Figure 7; the P8[5'] sites could not be clearly surveyed in the mt LSU intron). These additional protections are at positions that should be accessible to solvent based on Fe(II)-EDTA analysis of self-splicing introns that

lack a large P5abc structure. Other protections at J3/4-2,3 P9[5']-2,7, L9-1,3 and P9[3']-4,5 in one or both introns could also reflect CYT-18-binding sites, but their assignment is more ambiguous since they are present in some self-splicing introns, but absent in others (Heuer et al., 1991; Weeks & Cech, 1995b). In both introns, most of the putative CYT-18protection sites are clustered in the P4-P6 domain, which was previously shown to contain a high affinity CYT-18-binding site in experiments testing different segments of the mt LSU intron (Guo & Lambowitz, 1992). However, additional sites of protection specific for the protein complexes are found in the other major helical domain of the catalytic core in P3, P8 and P9 in both introns and in the P7.1/P7.1a structure and L9 in the mt LSU intron. Evaluation using a three-dimensional model structure suggests that the protections in P3 could be due to docking of the P1 helix to the catalytic core, whereas protections in P8, P9, L9 and the P7.1/P7.1a structure could reflect sites of interaction with the protein (see Discussion).

522

Protein-induced Folding of Group I Introns

Figure 7. Summary of iodinecleavage data for CYT-18/intron RNA complexes. Secondary structure models of the mt LSU (A) and ND1 (B) introns showing sites of iodine cleavage or protection in complexes formed with 20 nM intron RNA and 40 nM CYT-18 dimer under the conditions of Figure 6. Black or gray shading indicates CYT-18-induced protection at the 5' phosphate of a nucleotide residue (50% or more reduction in band intensity relative to the absence of CYT-18). Black shading indicates a potential CYT-18 protection site identified by protection at a position accessible to Fe(II)-EDTA cleavage in self-splicing introns that lack a large P5abc structure, i.e. the phage T4 td and sunY introns (Heuer et al., 1991) and yeast cob-I5 (Weeks & Cech, 1995b). (Recent data of Shaw & Lewin (1995), which show additional protections in a full-length version cob-I5 that contains the intron ORF, are not included). The inset shows cleavage/protection data for the mt LSU intron P6b region. Arrows indicate iodine cleavage at the 5' phosphate of a nucleotide residue. 5' phosphates that consistently showed protection in splicing buffer containing 5 mM Mg2+ in the absence of CYT-18 were: mt LSU intron–A322, A323, A324, G325, A326 and A327 in P8[3']; ND1 intron, none. Unmarked residues in the analyzed region could not be quantitated either because of strong reverse transcription stops or because of multiple primer extension bands occurring at runs of the same nucleotide (see Materials and Methods). Abbreviations: 5' and 3' SS, 5'- and 3'-splice sites, respectively. Nucleotides in intron and exons are shown in capital and lower case letters, respectively.

Protein-induced Folding of Group I Introns

523

Figure 8. Lead cleavage of the N. crassa mt LSU and ND1 introns. Pb2+ cleavage of the mt LSU (A) and ND1 (B) introns was carried out with 32P-labeled in vitro transcripts pBD5A/BanI and pND1m/NdeI, respectively. Complexes were formed by incubating 20 nM in vitro transcript with 40 nM CYT-18 dimer, as described in Materials and Methods. The free RNAs or complexes were incubated in splicing buffer with or without Mg2+ for 15 minutes on ice followed by five minutes at 37°C, then placed at room temperature and incubated for three minutes in the presence or absence of 20 mM Pb2+. Cleavage products were analyzed by electrophoresis in a 6% polyacrylamide/7 M urea gel, followed by autoradiography. Lanes 1 and 11, in vitro splicing reactions of the 32P-labeled transcripts in the presence of 40 nM CYT-18 protein dimer. Splicing reactions were initiated by addition of 250 mM GTP. Lanes 2 and 12, RNAs incubated as for in vitro splicing reactions, but in the absence of CYT-18; lanes 3, 4, 13 and 14, control incubation of RNAs in splicing buffer containing 5 mM Mg2+ in the presence or absence of CYT-18; lanes 5, 6, 15 and 16, Pb2+ cleavage of RNAs or CYT-18/RNA complexes in reaction medium lacking Mg2+; lanes 7, 8, 17 and 18, Pb2+ cleavage of RNAs or CYT-18/RNA complexes in reaction medium containing 5 mM Mg2+; lanes 9, 10, 19 and 20, Pb2+ cleavage of RNAs or CYT-18/RNA complexes in reaction medium containing 30 mM Mg2+. The bottom portion of the autoradiogram in panel A is taken from a longer exposure to better show the 106 nt Pb2+-cleavage product. The numbers to the right indicate positions of major Pb2+-cleavage fragments. Asterisks to the right in panel B indicate minor Pb2+-cleavage fragments of the ND1 intron. Numbers to the left indicate RNA size markers ( 32P-labeled in vitro transcripts). The schematic at the bottom shows the sizes of fragments expected for Pb2+ cleavage at a single site in J8/7 of each intron.

524

Figure 9. Effect of protease K digestion on the in vitro splicing of CYT-18/intron RNA complexes. In vitro splicing reactions with the mt LSU (A) and ND1 (B) introns were carried out with 32P-labeled in vitro transcripts pBD5A/BanI or pND1m/NdeI, respectively. Complexes were formed by incubating 150 nM 32P-labeled RNA with 500 nM CYT-18 dimer, as described in the Materials and Methods. The free RNAs or complexes were then incubated in splicing buffer containing 5 mM Mg2+ with or without 0.5 mg/ml protease K for 60 minutes at 37°C, prior to initiating splicing reactions by the addition of 1 mM GTP. Products were analyzed by electrophoresis in a 6% polyacrylamide/7 M urea gel, followed by autoradiography of the dried gel. Lane 1, splicing reactions with RNAs preincubated in splicing buffer without protease K; lanes 2, splicing reactions with RNAs preincubated in splicing buffer with protease K. Lane 3, splicing reactions with CYT-18/RNA complexes preincubated in splicing buffer without protease K; lane 4, splicing reactions with CYT18/RNA complexes preincubated in splicing buffer with protease K. Abbreviations: P, precursor RNA; E1-I, product resulting from hydrolysis at the 3'-splice site; I-E2, product resulting from GTP-dependent cleavage at the 5'-splice site; I, excised intron; E1-E2, ligated exons.

Probing CYT-18/RNA complexes by Pb2+ cleavage Pb2+ cleavage provides an additional probe for formation of the catalytically-active group I intron tertiary structure (Streicher et al., 1993). Self-splicing group I introns are cleaved by Pb2+ in two regions of the catalytic core, J8/7 and P7, that have been implicated in binding functionally important Mg2+ ions. The formation of these metal ion-binding sites requires the presence of sufficient Mg2+ to stabilize the active RNA structure, but the Pb2+ cleavages are suppressed by higher Mg2+, which presumably competes for the same metal ion-binding site (Streicher et al., 1993). Pb2+ cleavage of the N. crassa mt LSU and ND1 introns is shown in Figure 8. In reaction medium containing 5 mM Mg2+, both N. crassa introns showed specific Pb2+ cleavage only in the presence of the CYT-18 protein (cf. lanes 7 and 17). This specific cleavage occurred predominantly at a single site in each intron, yielding two major fragments (397 and 106 nts for the mt LSU intron and 160 and 53 nts for the ND1 intron). The sizes of these fragments indicate that the Pb2+ cleavage occurred

Protein-induced Folding of Group I Introns

in the J8/7 region in both introns, and primer extension mapping showed that the predominant cleavage site was between J8/7-4 and 5 in the ND1 intron and J8/7-3 and 4 in the mt LSU intron (data not shown). These sites are at or within one nt of the cleavage site found in self-splicing group I introns (between J8/7-4 and 5; Streicher et al., 1993). The gel for the ND1 intron also shows minor cleavage products (approx. 150, 162 and 170 nts, indicated by asterisks), but these were not observed reproducibly and were not detected by the primer extension analysis (data not shown). Pb2+ cleavage in P7, which is relatively inefficient in self-splicing group I introns (Streicher et al., 1993), was not detected in either the mt LSU or ND1 introns. As with self-splicing introns, the Pb2+ cleavage in J8/7 of the N. crassa introns did not occur in the absence of Mg2+, which is required for tertiary structure formation, and it was suppressed by higher concentrations of Mg2+, which presumably compete for the same or an overlapping metal ion-binding site. Together, these findings show that the tertiary structure changes induced by the CYT-18 protein are required for the formation of a metal ion-binding site that is correlated with the catalytic activity of group I introns. CYT-18 must remain bound to the intron RNA to maintain the catalytically-active structure Protease-digestion experiments were carried out to determine whether CYT-18 must remain bound to the intron RNA to maintain splicing activity and the catalytically-active RNA structure. Figure 9 shows an experiment in which CYT-18 was bound to 32P-labeled precursor RNAs containing the mt LSU and ND1 introns, then incubated in the presence or absence of protease K and assayed for splicing activity. The results show that the protease treatment completely abolished the splicing of both introns (Figure 9, lanes 4). Parallel experiments showed that protease digestion of CYT-18/intron RNA complexes also resulted in the loss of the Pb2+-cleavage site in J8/7 in both introns and caused the DMS-modification patterns for both introns to revert largely to those prior to CYT-18 binding (data not shown). Clearly, CYT-18 differs from RNA chaperones in that its continued binding is required to maintain splicing activity and the catalyticallyactive RNA structure.

Discussion Our results show that the N. crassa mt LSU and ND1 introns, which are not detectably self-splicing, remain largely unfolded in the absence of the CYT-18 protein, irrespective of Mg2+ concentration. Although the introns by themselves form most of the expected group I intron secondary structure, CYT-18 binding is required to stabilize the secondary structure, including the long-range pairings P3 and P7, and to induce the tertiary structure required for splicing activity. Analogous

Protein-induced Folding of Group I Introns

525

protein-induced tertiary folding of the group I intron catalytic core has been observed for CYT-18 binding to the DP5abc-derivative of the Tetrahymena LSU intron (Mohr et al., 1994) and for the interaction of the yeast CBP2 protein with the mtDNA intron, cob-I5 (Shaw & Lewin, 1995; Weeks & Cech, 1995b). By contrast with RNA chaperones, which act transiently to facilitate the formation of helices, the CYT-18 protein is required for tertiary structure formation, and its continued binding is necessary to maintain this structure and splicing activity. Identification of the CYT-18-binding site Potential sites of interaction of the CYT-18 protein with the N. crassa group I intron RNAs were identified in iodine-cleavage experiments as phosphate protections at positions accessible to solvent in structurally-related self-splicing group I introns. Many of these potential CYT-18-interaction sites (34/49 and 10/18 in the mt LSU and ND1 introns, respectively) fall in the P4-P6 domain of the catalytic core, as expected from previous studies testing the binding of CYT-18 to separate domains of the mt LSU intron (Guo & Lambowitz, 1992). However, additional CYT-18-protection sites are found in the P3-P9 domain in both introns. Satisfyingly, the CYT-18-protection sites in the ND1 intron are largely a subset of those in the mt LSU intron, even though the ND1 intron belongs to a different structural subclass and lacks peripheral structures present in the mt LSU intron. Furthermore, the distribution of CYT-18-protection sites in the mt LSU intron is consistent with previous studies, showing that most of the binding energy can be accounted for by the interaction of CYT-18 with the isolated P4-P6 domain excluding P6b (Kd = 160 pM, −13.3 kcal/mole) and that addition of P7-P9 results in only a small additional increment in binding energy (−1.2 kcal/mole, decreasing Kd to 22 pM; Guo & Lambowitz, 1992; Saldanha et al., unpublished results). Notably, of 22 nucleotide residues flanking the 11 CYT-18-protection sites common to the mt LSU and ND1 introns, only seven are identical in both introns (the A-residue upstream of P3, P4 bp-3, P4[3']-1, J4/5-2 and P9-5, 6; Figure 7). Thus, it seems likely that many of the CYT-18/RNA protections identified by iodine footprinting reflect interactions between the protein and the phosphodiester backbone. Base-specific contacts remain possible, but were not detected in the DMS or DEPC-modification experiments. In the absence of detailed three-dimensional models of the N. crassa introns, the potential CYT-18-binding sites detected by iodine-footprinting in the N. crassa mt LSU intron were evaluated by using the three-dimensional model structure of a closely related group IA1 intron, the yeast mt LSU (v+ ) intron (Figure 10; Jaeger et al., 1991). In this model, the putative CYT-18-protection sites in the P4-P6 domain, P7.1/P7.la structure, P8, P9 and L9 are largely contiguous along one face of the intron on the side opposite the catalytic cleft, and all but

Figure 10. Potential CYT-18-protection sites mapped on a three-dimensional model structure of a group IA1 intron. The potential CYT-18-protection sites in the N. crassa mt LSU intron are shown as gray spheres on the three-dimensional model structure of the related yeast mt LSU intron (denoted v+; Jaeger et al., 1991). The potential contact sites were identified in iodine-cleavage experiments by protection at positions that should be accessible to cleavage based on Fe(II)-EDTA analysis of self-splicing introns that lack a large P5abc structure (see the legend to Figure 7). The phosphate protections in L5 and P8, which are not shown, fall in regions that have not been modeled in the yeast mt LSU intron, and the phosphate protection at P5[5']-1 is shown in the J4/5 segment, which is extended in the yeast compared to the N. crassa intron. The phosphate protections in P3[5'] are on the opposite side in this view. The P4-P6 domain is shown in green, the P3-P9 domain and P7.1/7.1a structure are shown in purple, P2 is shown in yellow, P1[3'] is shown in red and the 5'- and 3'-exons are shown in black. The Figure represents the precursor RNA just after 5'-splice site cleavage. The Figure was made using the DRAWNA software (Massire et al., 1994).

three of these sites (P6a-3 and 4 and L9-2) are clearly on the outer surface of the RNA where they could interact with the CYT-18 protein. By contrast, the protein-induced phosphate protections in P3 are on the opposite face of the catalytic core and are not contiguous with the remainder of the CYT-18protection sites. Although we cannot exclude that these P3 protections reflect an additional proteininteraction site, their location is such that they can be accounted for by more stable docking of the P1 helix to the catalytic core in the presence of CYT-18. The more stable docking of P1 could also account for

526 some of the protein-specific protections seen in the J4/5 region of both introns (see Wang et al., 1993). In the three-dimensional model structure, the P4-P6 domain, which comprises most of the potential CYT-18-binding site, is a continuous helix in which the coaxial stacking of P4 and P6 is stabilized by base-triples with the incoming and outgoing single-stranded regions, J3/4 and J6/7, respectively. Most of the putative CYT-18-protection sites fall on the outer surface of this helical domain, extending from P5 to P6b. The protected phosphates in P9[5'] on the other helical domain are adjacent to those in P5, and the protections in the P7.1/P7.1a structure are adjacent to those in P6a. The potential CYT-18-interaction site in P8 falls just outside the region represented in the yeast mt LSU intron model, but is predicted to be adjacent to the protected phosphate at P6a-5. Notably, many of the potential CYT-18-contact sites are clustered around the junction of the P4-P6 stacked helix, with CYT-18 facing the shallow groove of P4 and the deep groove of P6, the sides opposite those engaged in base-triple interactions. Footprinting and modification-interference experiments with the isolated P4-P6 domain have confirmed that the identified nucleotide positions in P4 and P6 are critical for CYT-18 binding (M. G. Caprara & A. M. Lambowitz, unpublished results). The possible CYT-18 interaction with P9/L9 is supported independently by experiments showing that 3' truncations of the mt LSU intron extending to within four nts of P9 do not affect CYT-18 binding or 5'-splice-site cleavage, whereas truncations extending five nts into P9 increase the Kd for CYT-18 from <0.3 pM to 7 pM and abolish 5'-splice-site cleavage, even at high Mg2+ (Guo & Lambowitz, 1992; Saldanha et al., unpublished results; M. G. Caprara & A. M. Lambowitz, unpublished results). Analogous 3' truncations extending into P9 of the ND1 intron abolish CYT-18 binding (M. G. Caprara & A. M. Lambowitz, unpublished results). The essential function of P9 in the CYT-18-dependent splicing reactions contrasts with results for the Tetrahymena LSU intron, where the deletion of P9 does not abolish ribozyme activity at high Mg2+ (Szostak, 1986; Beaudry & Joyce, 1990; Caprara & Waring, 1994). Other deletions in the N. crassa mt LSU intron that remove the protected phosphates in P3 or the protected phosphate in P8 had no detectable effect on CYT-18 binding (Saldanha et al., unpublished results). Thus, if these phosphate protections reflect CYT-18 contact sites, they must make a relatively small contribution to binding energy. The limited sequence similarity at the potential CYT-18-contact sites in the mt LSU and ND1 introns suggests that CYT-18 may recognize a specific three-dimensional geometry of the phosphodiester backbone of group I introns. In both introns, the protein appears to interact primarily with the P4-P6 domain (P5-P4-P6), with some contribution from P8 and P9 in the other major helical domain. In addition, CYT-18 may make intron specific contacts

Protein-induced Folding of Group I Introns

with P6a, P6b, and the P7.1/P7.la structure in the mt LSU intron. Assuming that the minimum CYT-18-binding site includes the core region of the P4-P6 domain, P8 and P9, and that the P4-P6 domain is a continuous A form helix with a ˚ (Michel & Westhof, 1990; base-pair step of 2.8 A Saenger, 1984), the CYT-18 footprints on the mt LSU ˚, and ND1 introns extend approximately 73 and 67 A respectively, comparable to the distances spanned by aminoacyl-tRNA synthetases interacting with their cognate tRNAs (Saenger, 1984; Giege et al., 1993). The footprint in the mt LSU intron would be larger if the CYT-18-binding site is extended to include P6b. It should be emphasized that our results serve only for an initial picture of the CYT-18-binding site. It remains possible that some protections attributed to CYT-18 are due to RNA structural changes, and that some protections attributed to RNA structure by comparison with self-splicing introns are due to the proximity of CYT-18 in the N. crassa introns. Mechanism of action of CYT-18 in RNA splicing The large binding energy for the interaction of CYT-18 with the isolated P4-P6 domain and the clustering of potential CYT-18 contact sites near the junction of the P4-P6 stacked helix suggest that a major function of CYT-18 might be to stabilize the correct geometry in this region. This hypothesis is supported by the structure mapping of the mt LSU and ND1 introns, which provides evidence that CYT-18 binding is required for correct folding of the P4-P6 domain, including stabilization of helical regions and the formation of the J3/4 and J6/7 base-triples. Likewise, in the DP5abc-derivative of the Tetrahymena intron, CYT-18 binding stabilized the P4 helix and also appeared to be required for the formation of the J3/4 and J6/7 base-triples (Mohr et al., 1994). The protease-digestion experiments for the mt LSU and ND1 introns show that the continued binding of CYT-18 is required to maintain the P4-P6 domain structure and that the J3/4 and J6/7 base-triples are not sufficient for this purpose. The Mg2+-induced folding of the Tetrahymena ribozyme appears to occur via an ordered pathway in which the initial formation of the P4-P6 domain nucleated by the P5abc extension provides a scaffold for the folding of the P3-P9 domain (Zarrinkar & Williamson, 1994). Since the structuremapping experiments show that both N. crassa introns remain largely unfolded in the absence of CYT-18, it seems likely that CYT-18 initially recognizes primary or secondary structural features in the unfolded intron and then makes additional contacts coincident with formation of tertiary structure. Such two-step binding is consistent with previous kinetic analysis, which indicated that the binding of CYT-18 to the mt LSU intron involves an initial bimolecular step that is close to diffusion limited (kon = 3.24 × 107 M−1 s−1 ), followed by a

527

Protein-induced Folding of Group I Introns

slower step (0.54 s−1 ), putatively an RNA conformational change (Saldanha et al., 1995). The findings that most of the potential CYT-18contact sites are in the P4-P6 domain (see above) and that this domain of the mt LSU intron can bind CYT-18 independently (Guo & Lambowitz, 1992) suggest that CYT-18 may bind initially to this region to promote the correct assembly of the P4-P6 domain. The formation of the P4-P6 domain may in turn be essential for the induction of one or more CYT-18-binding sites in the P3-P9 domain (Guo & Lambowitz, 1992; Saldanha et al., 1995). The P3-P9 domain of the N. crassa mt LSU intron does not bind CYT-18 independently, consistent with previous results for the Tetrahymena intron showing that the folding of P3-P9 is dependent on P4-P6 (Zarrinkar & Williams, 1994; Saldanha et al., unpublished results). The chemical-modification data provide evidence that a number of long range base-pairing and tertiary interactions that do not form in the absence of CYT-18 are established in the presence of the protein (e.g., P3, J3/4-P6, J6/7-P4, J4/5-P1, P7, J8/7-P1, P11 and L9-P5). The formation of these interactions is presumably dependent on an earlier step in the folding pathway that is defective in the N. crassa introns. At present, we cannot distinguish whether the binding of CYT-18 to the P4-P6 domain is itself sufficient to promote tertiary folding of the remainder of the catalytic core or whether this folding requires additional binding energy from contacts in the P3-P9 domain. As indicated above, most of the binding energy for the interaction of CYT-18 with the mt LSU intron can be accounted for by its interaction with the core region of the P4-P6 domain, and other regions of the intron contribute only a small increment to binding energy (Guo & Lambowitz, 1992; Saldanha et al., unpublished results). It remains possible, however, that the relatively small contribution from the P3-P9 domain reflects that binding to this region is coupled to an energetically unfavorable conformational change. Comparison of the mode of action of CYT-18 with that of the P5abc RNA structure and the yeast CBP2 protein The present results provide additional insight into how CYT-18 can functionally replace the peripheral RNA structure, P5abc, to promote splicing of the Tetrahymena LSU intron (Mohr et al., 1994). The P5abc structure in the Tetrahymena intron has been shown to make a sharp bend at the top of the P5 stem and fold down to interact with the outer surface of the P4-P6 stacked helix (Murphy & Cech, 1994), the same region of group I introns that interacts with the CYT-18 protein. Furthermore, the Tetrahymena P5abc structure folds independently at low Mg2+ concentration, and this folding is believed to provide a nucleation site for the subsequent folding of P4-P6 (Celander & Cech, 1991; Laggerbauer et al., 1994; Murphy and Cech, 1994; Zarrinkar

& Williamson, 1994). Thus, both CYT-18 and the P5abc structure interact with the P4-P6 region and may act early in the folding pathway to promote the assembly of this domain. In one case this is accomplished by an RNA–RNA interaction and in the other case by an RNA–protein interaction. The yeast CBP2 protein, which functions specifically in splicing the mtDNA intron cob-I5, also promotes formation of the catalytically-active RNA structure (Shaw & Lewin, 1995; Weeks & Cech, 1995b). However, the CBP2 protein appears to bind primarily to the opposite face of the intron, with the potential binding site including parts of P1, P2, P7.1/P7.1a and P8. Unlike CYT-18, CBP2 does not bind specifically to other group I introns, and it presumably recognizes structural features that are unique to cob-I5 (Gampel & Cech, 1991; Weeks & Cech, 1995a). Interestingly, CBP2 not only stabilizes the tertiary structure of the catalytic core at low Mg2+, but also appears to bind directly to the 5'-splice-site domain to promote its association with the catalytic core. Possible structural similarities between group I introns and tRNAs Finally, the initial picture of the protein-binding site obtained here permits some assessment of the possibility that CYT-18 recognizes structural features in group I introns that resemble those in tRNATyr (Guo & Lambowitz, 1992). Modeling studies with the Bacillus stearothermophilus TyrRS, for which a partial X-ray crystal structure is available (Brick et al., 1988), suggest that the synthetase interacts with the tRNATyr on the variable loop side. For the E. coli TyrRS/tRNATyr interaction, potential contact sites identified by UV-cross-linking are found in the acceptor stem, anticodon stem and loop, D-stem and the variable loop, and identity elements include the last base-pair of the acceptor stem, the discriminator base, the first and second bases of the anticodon loop, and the orientation of the variable arm (Bedouelle, 1990; Himeno et al., 1990; Bedouelle et al., 1993). The P4-P6 stacked helix of group I intron RNAs, which appears to be primarily responsible for the interaction with CYT-18, is analogous to the stacked helix formed by the D-arm and anticodon arm of tRNAs in being stabilized by base-triples with the incoming and outgoing single-stranded regions (Michel & Westhof, 1990). The interaction of CYT-18 with the phosphodiester backbone of the P4/P6 stacked helix and the possible contacts with the shallow groove of P4 may be analogous to the interaction of TyrRS with the stacked D and anticodon-arms of tRNATyr (Bedouelle et al., 1993; Putz et al., 1991). Notably, our results provide no indication that CYT-18 interacts with the P7 stem, which was previously noted to be similar in sequence and secondary structure to the variable stem of the N. crassa mt tRNATyr (Guo & Lambowitz, 1992). Further assessment of the structural similarities between group I introns and tRNAs will be based on work in progress

528 comparing the CYT-18-binding site in group I introns with that in tRNATyr.

Materials and Methods Recombinant plasmids for synthesis of intron-containing RNAs pBD5A and pND1m were used to synthesize in vitro transcripts containing the N. crassa mt LSU and ND1 introns, respectively. pBD5A contains a 388 nt miniderivative of the N. crassa mt LSU intron cloned behind the phage T3 promoter in Bluescribe vector (pBS+; Stratagene, La Jolla, CA; Guo et al., 1991). Digestion of pBD5A with BanI and transcription with phage T3 RNA polymerase (see below) yielded a 503 nt RNA containing a 65 nt 5'-exon, the 388 nt intron, and a 50 nt 3'-exon. pND1m contains a 197 nt-derivative of the N. crassa ND1 intron cloned behind the phage T7 promoter between the HindIII and NdeI sites of pUC18 (Wallweber & Lambowitz, unpublished results). The cloned segment of the ND1 intron includes the entire catalytic core and extends six nts beyond a naturally used, alternative 3'-splice site near the N-terminus of the intron ORF. The remainder of the intron ORF and the distal 3'-splice site have been deleted. For most experiments, pND1m was digested with NdeI and transcribed with phage T7 RNA polymerase to yield a 212 nt RNA containing a seven nt 5'-exon, the 197 nt intron, and an eight nt 3'-exon. Some experiments used a 460 nt transcript obtained from pND1m digested with AatII. This transcript contains a longer (256 nt) 3'-exon, enabling use of the down-stream primer N3 for better mapping of modification and cleavage sites in P9. Purification of the CYT-18 protein The CYT-18 protein was synthesized in E. coli from the expression plasmid pEX560 and purified by PEIprecipitation, (NH4 )2 SO4-fractionation and CM-Sepharose chromatography (Saldanha et al., 1995). pEX560 is a derivative of pET3a (Rosenberg et al., 1987), which contains the 638 amino acid cyt-18 ORF inserted behind the phage T7 promoter. The ORF lacks the N-terminal mitochondrial targeting sequence, but has an additional N-terminal AUG codon for expression in E. coli. The CYT-18 preparations used here were 90 to 95% pure and contained 95 to 100% active protein, as judged by determination of tyrosyl-adenylation active sites (Saldanha et al., 1995). Protein concentrations were determined by the Bradford assay (New England Biolabs, Beverly, MA) and expressed as amount of CYT-18 dimer. DNA oligonucleotides The following DNA oligonucleotide primers were used for mapping of modification or cleavage sites by reverse transcription: LS1, 5'-GGGGGGTACCAACTTGTTATCATATTGAT, complementary to a sequence at the intron/3'-exon junction of the mt LSU intron; LS2, 5'-TTATATCCCATGGGTAAAAAAAAG, complementary to a sequence in P8 of the mt LSU intron; LS3, 5' - AAAAAGAAAGTTGGGTTGTTTATAGTCGTTGAA CG, complementary to a sequence in P6-P7 of the mt LSU intron; N1, 5'-CCCATATGATTTCATTTTTTTT, complementary to a sequence spanning the intron/3'-exon junction of the ND1 intron; N2, 5'-TAGTCTCTAGGGTTTTAACA, complementary to a sequence in P6a-P7 of the ND1 intron and N3, 5'-TTAACTATGCGGCATCAGAGCA-

Protein-induced Folding of Group I Introns

GATTGTAC, complementary to a sequence in the 3'-exon of the ND1 intron RNA transcribed from pND1m/AatII. DNA oligonucleotides were obtained from Macromolecular Resources (Fort Collins, CO) and purified by electrophoresis in 15% (w/v) polyacrylamide (acrylamide:bisacrylamide, 19:1, (w/w))/7 M urea gels. Primers were visualized by UV-shadowing and eluted from gel slices by soaking in 0.3 M NaOAc, 10 mM Tris-HCl (pH 7.5), 1 mM EDTA for 12 to 24 hours at 37°C. The primers were then ethanol precipitated, centrifuged through a Sephadex G-25 spun column, and reprecipitated with ethanol. Synthesis of in vitro transcripts Transcription reactions were performed for two hours at 37°C in 100 ml of reaction medium containing 5 mg of DNA template, 50 to 100 units of phage T3 RNA polymerase (GIBCO BRL, Gaithersburg, MD) or phage T7 RNA polymerase (New England Biolabs), 40 mM Tris-HCl (pH 7.5), 6 mM MgCl2 , 2 mM spermidine, 100 mg/ml bovine serum albumin, 10 mM DTT, 1 mM of each rNTP and 40 units ribonuclease inhibitor (U.S. Biochemicals, Cleveland, OH). After the reaction, the DNA template was digested with DNase I (20 units; Pharmacia, Piscataway, NJ) for 20 minutes at 37°C. Transcripts were then extracted with phenol-CIA (phenol:chloroform:isoamyl alcohol; 25:24:1, by vol.), desalted through a Sephadex G-50 spun column, and precipitated with ethanol. Phosophorothioate-containing transcripts for iodine-cleavage experiments were synthesized as above, except that 0.04 mM of the specified NTPaS (Amersham, Arlington Heights, IL) was added to the transcription reactions. 32P-labeled transcripts used in the Pb2+-cleavage experiments were synthesized as above with 25 mCi of [a-32P]UTP (3000 Ci/mmole; NEN-DuPont, Boston, MA) and 0.25 mM NTPs, and those used in in vitro splicing experiments were synthesized with 50 mCi of [a-32P]UTP and 1 mM NTPs. RNAs to be 5' end-labeled were transcribed in the presence of 1.3 mM guanosine to give a 5' OH that could be phosphorylated directly with [g-32P] ATP (3000 Ci/mmole; NEN-DuPont) and phage T4 polynucleotide kinase (New England Biolabs; Sambrook et al., 1989). The 5' end-labeled RNAs were purified by electrophoresis in a 6% (w/v) polyacrylamide/7 M urea gel. Chemical modification RNA structure mapping and footprinting were carried out with in vitro transcripts pBD5A/BanI, which contains the mt LSU intron, and pND1m/NdeI or pND1m/AatII, which contain the ND1 intron (see above). Chemical modification was performed with dimethyl sulfate (DMS; Sigma, St Louis, MO), diethyl pyrocarbonate (DEPC; Sigma), or 1-cyclohexyl-3-(2-morpholine-ethyl) carbodiimide metho-p-toluenesulfonate (CMCT; Sigma). DMS and DEPC reactions were in 100 ml of 10 mM N-[2-hydroxyethyl]piperazine-N'-[2-ethanesulfonic acid] (HepesKOH; pH 7.5), 100 mM KCl and 5 to 25 mM MgCl2 (native conditions) or 10 mM Hepes-KOH (pH 7.5), 1 mM EDTA (denaturing conditions). The reaction medium used for CMCT modification contained 10 mM potassium borate (pH 7.9), instead of Hepes-KOH. For modification reactions, 20 nM RNA was preincubated in the reaction medium with or without 20 or 40 nM CYT-18 dimer for ten minutes at 37°C. The reactions were initiated by the addition of the modifying reagent (1 ml of DEPC, 1 ml of 25 mg/ml CMCT, or 1 ml of a 1/6 dilution of DMS in

529

Protein-induced Folding of Group I Introns

ethanol), incubated for four minutes at 37°C (native conditions) or one minute at 55°C (denaturing conditions), and then terminated by adding 2 ml 200 mM EDTA plus 10 mg of E. coli tRNA carrier and extraction with phenolCIA (25:24:1 by vol.) followed by ethanol precipitation. The reaction conditions result in less than one modification per RNA, as judged by primer extension analysis. The proportion of intron RNA complexed with protein was measured separately by nitrocellulose-filter binding (Guo & Lambowitz, 1992). With 20 nM CYT-18 dimer and 20 nM intron RNA, 75 to 85% of the input RNA was complexed with protein, and the amount of complex increased by 5 to 10% with 40 nM CYT-18 dimer. After binding CYT-18, 80 to 100% of the mt LSU and ND1 intron RNAs reacted upon the addition of GTP. For the pND1m/NdeI transcript (ND1 intron), a plot of the fraction of spliced RNA versus time followed a single exponential (kobs = 0.015 s−1 ), consistent with a uniform population of RNA/protein complexes. For the pBD5A/ BanI transcript (mt LSU intron), 080% of the active complexes spliced with kinetics following a single exponential (kobs = 0.0020 s−1 ), while the remainder were subject to rapid GTP-dependent cleavage at the 5'- or 3'-splice sites, giving rise to products that did not appear to be processed further. Additional controls for the mt LSU intron showed that subjecting the RNA to the renaturation protocol of Jaeger et al. (1991) or preincubating it with CYT-18 for a longer time (60 minutes) had no significant effect on the DMS and DEPC-modification patterns in the presence of CYT-18 (data not shown). Together, these findings suggest that both the mt LSU and ND1 intron RNAs are predominantly in a single active conformation following binding of CYT-18. Mapping of modification sites by reverse transcription Sites of modification were mapped by reverse transcription, using DNA oligonucleotide primers spanning the P3 to P9 regions of the introns (see above). A 5'end-labeled primer (5 pmoles) was annealed to 0.2 to 0.4 pmoles of RNA in 5 ml of 16 mM Tris-HCl (pH 8.3), 40 mM KCl for 30 seconds at 95°C, followed by 15 minutes at 55°C. The reaction mixture was then placed on ice and supplemented to 2.5 mM MgCl2 , 20 mM DTT, and 1 mM dNTPs. Reverse transcription reactions were initiated by the addition of M-MLV reverse transcriptase (200 units; GIBCO BRL), incubated for ten minutes at 50°C, and then terminated by adding EDTA to a final concentration of 50 mM. After heating to 95°C for seven minutes, the products were analyzed by electrophoresis in a 6% (w/v) polyacrylamide/7 M urea gel. Dideoxysequencing ladders obtained from pBD5A or pND1m by using Sequenase (U.S. Biochemicals) with the same 5' end-labeled primers were run in parallel lanes. After electrophoresis, the gels were dried and autoradiographed, and sites of modification were detected as blocks to reverse transcription, one base before the modified residue. Band intensities were measured with a Betascope 603 Blot analyzer (Betagen, Waltham, MA) and normalized for the amount of primer extension product in each lane. Positions were considered strongly modified if the band intensity was within 50% of the strongest modification in the denatured lane and moderately modified if the band intensity was 10 to 49% of the strongest modification in the denatured lane. Protections were assigned based on the change in band intensity relative to the reference sample, which was either the RNA under denaturing conditions or the RNA at 5 mM

Mg2+ in the absence of CYT-18 protein. Positions were considered protected if the band intensity was E50% of that in the reference sample and enhanced if the band intensity was >50% of that in the reference sample. Fe(II)-EDTA cleavage For Fe(II)-EDTA experiments, 5 ml of 5' 32P-labeled RNA (0.1 to 0.5 pmoles; approximately 50,000 cpm) was added to 2 ml of 5× reaction medium (50 mM Hepes-KOH (pH 7.5), 500 mM KCl and 0 to 125 mM MgCl2 ) in a 0.65 ml microfuge tube, and preincubated at 50°C for 20 minutes; followed by 37°C for ten minutes. To initiate the reactions, 1 ml each of 10 mM ammonium iron(II) sulfate (Aldrich, Milwaukee, WI), 20 mM EDTA, 100 mM DTT were added to the side of the tube and then rapidly mixed with the RNA by centrifugation for ten seconds. The reactions were incubated for two hours at 37°C, then terminated by adding thiourea (Aldrich) to a final concentration of 10 mM. Fe(II)-EDTA-cleavage products were analyzed by electrophoresis in a 6% (w/v) polyacrylamide/7 M urea gel, against sequencing ladders obtained from the same 5' end-labeled RNA by partial digestion with RNase U2 or T1 (Pharmacia; Donis-Keller et al., 1977) or NaOH (Caprara & Waring, 1993). The gel was dried and autoradiographed as above. Iodine cleavage of phosphorothioatecontaining RNAs Iodine cleavage reactions were carried out with RNAs or RNA/protein complexes (20 nM RNA plus 20 or 40 nM CYT-18 dimer). The RNAs or mixtures of RNA plus CYT-18 protein were preincubated in 100 ml of splicing buffer for ten minutes at 37°C. Cleavage reactions were initiated by the addition of 1 ml of a freshly prepared iodine-stock solution (1 mM in 100% ethanol; Sigma) and incubated for 45 seconds at 37°C, conditions that give less than one iodine cleavage per molecule, as judged by primer extension analysis. The reactions were terminated by placing them on ice and adding 2 ml of quench solution (500 ng of an unrelated 10% phosphorothioate-substituted RNA, 5 mg E. coli tRNA and 133 mM EDTA), followed by extraction with phenol-CIA and ethanol precipitation. We found that inclusion of the unrelated phosphorothioate-substituted RNA in the quench solution is important for obtaining reproducible and well-defined cleavage-protection patterns. Cleavage sites were mapped by reverse transcription using M-MLV reverse transcriptase (GIBCO BRL) essentially as described above for the analysis of base modifications, except that the reverse transcription reaction was shortened to 45 seconds to minimize the addition of noncoded nucleotides to the 3' ends of the cDNAs terminated at cleavage sites (Christian & Yarus, 1992). Under the conditions used, most of the primer extension bands are doublets reflecting the addition of one noncoded nucleotide to some proportion of the product. In cases where there are runs of the same nucleotide, the extent of protection was determined from the first and last bands in the series (assigned to the first and last positions, respectively), and other positions were considered ambiguous. Lead-cleavage reactions Lead-cleavage reactions were in 20 ml of reaction medium containing 20 nM 32P-labeled RNA (0.2 mCi/

530 pmole), 100 mM KCl, 20 mM Tris-HCl (pH 7.5), 5 mM DTT, 20 units of ribonuclease inhibitor (U.S. Biochemicals), different concentrations of MgCl2 , and 40 nM CYT-18 dimer where indicated (Mohr et al., 1994). The reaction mixtures were preincubated on ice for 15 minutes and then at 37°C for five minutes. After preincubation, the samples were shifted to room temperature and incubated for three minutes with 20 mM lead acetate (Jenneile Chemical Corp, Cincinnati, OH). The reactions were terminated by adding EDTA to 50 mM, followed by phenol-CIA (25:24:1 by vol.) extraction. Products were analyzed by electrophoresis in a 6% polyacrylamide (w/w)/7 M urea gel, followed by autoradiography of the dried gel.

In vitro splicing reactions Splicing reactions were carried out in 100 ml of reaction medium containing 150 nM 32P-labeled precursor RNA (0.12 mCi/pmole), 500 nM CYT-18 dimer, 100 mM KCl, 5 mM MgCl2 , 10 mM Tris-HCl (pH 7.5; Guo et al., 1991). Complexes were formed by incubating the 32P-labeled RNA with 500 nM CYT-18 dimer for ten minutes at 37°C, and then incubated in the presence or absence of protease K (0.5 mg/ml; Sigma) for 60 minutes at 37°C, prior to the splicing reactions. Splicing reactions were initiated by the addition of GTP to 1 mM, incubated for 15 minutes at 37°C, and terminated by addition of EDTA to 10 mM, followed by phenol-CIA (25:24:1 by vol.) extraction. Products were analyzed by electrophoresis in a 6% polyacrylamide (w/w)/7 M urea gel, followed by autoradiography of the dried gel.

Acknowledgements We thank Dr Eric Westhof (Institut de Biologie Mole´culaire et Cellulaire du CNRS, Strasbourg) for comments on the manuscript and for help in preparing Figure 10 and Ms Janet Gianelos for preparation of CYT-18 protein. M.G.C. was supported by NIH postdoctoral fellowship F32 GM16468. This work was supported by NIH grant GM37951.

References Akins, R. A. & Lambowitz, A. M. (1987). A protein required for splicing group I introns in Neurospora mitochondria is mitochondrial tyrosyl-tRNA synthetase or a derivative thereof. Cell, 50, 331–345. Banerjee, A. R., Jaeger, J. A. & Turner, D. H. (1993). Thermal unfolding of a group I ribozyme: the low-temperature transition is primarily disruption of tertiary structure. Biochemistry, 32, 153–163. Beaudry, A. A. & Joyce, G. F. (1990). Minimum secondary structure requirements for catalytic activity of a selfsplicing group I intron. Biochemistry, 29, 6534–6539. Bedouelle, H. (1990). Recognition of tRNATyr by tyrosyltRNA synthetase. Biochimie, 72, 589–598. Bedouelle, H. & Winter, G. (1986). A model of synthetase/transfer RNA interaction as deduced by protein engineering. Nature, 320, 371–373. Bedouelle, H., Guez-Ivanier, V. & Nageotte, R. (1993). Discrimination between transfer-RNAs by tyrosyltRNA synthetase. Biochimie, 75, 1099–1108. Brick, P., Bhat, T. N. & Blow, D. M. (1988). Structure of ˚ resolution. tyrosyl-tRNA synthetase refined at 2.3 A

Protein-induced Folding of Group I Introns

Interaction of the enzyme with the tyrosyl adenylate intermediate. J. Mol. Biol. 208, 83–98. Caprara, M. G. & Waring, R. B. (1993). Important 2'-hydroxyl groups within the core of a group I intron. Biochemistry, 32, 3604–3610. Caprara, M. G. & Waring, R. B. (1994). Deletion of P9 and stem-loop structures downstream from the catalytic core affects both 5' and 3' splicing activities in a group I intron. Gene, 143, 29–37. Cech, T. R., Herschlag, D., Piccirilli, J. A. & Pyle, A. M. (1992). RNA catalysis by a group I ribozyme. Developing a model for transition state stabilization. J. Biol. Chem. 267, 17479–17482. Celander, D. W. & Cech, T. R. (1991). Visualizing the higher order folding of a catalytic RNA molecule. Science, 251, 401–407. Cherniack, A. D., Garriga, G., Kittle, J. D., Jr, Akins, R. A. & Lambowitz, A. M. (1990). Function of Neurospora mitochondrial tyrosyl-tRNA synthetase in RNA splicing requires an idiosyncratic domain not found in other synthetases. Cell, 62, 745–755. Christian, E. L. & Yarus, M. (1992). Analysis of the role of phosphate oxygens in the group I intron from Tetrahymena. J. Mol. Biol. 228, 743–758. Coetzee, T., Herschlag, D. & Belfort, M. (1994). Escherichia coli proteins, including ribosomal protein S12, facilitate in vitro splicing of phage T4 introns by acting as RNA chaperones. Genes Dev. 8, 1575–1588. Collins, R. A. & Lambowitz, A. M. (1985). RNA splicing in Neurospora mitochondria. Defective splicing of mitochondrial mRNA precursors in the nuclear mutant cyt18-1. J. Mol. Biol. 184, 413–428. Donis-Keller, H., Maxam, A. M. & Gilbert, W. (1977). Mapping adenines, guanines and pyrimidines in RNA. Nucl. Acids Res. 4, 2527–2538. Ehresmann, C., Baudin, F., Mougel, M., Romby, P., Ebel, J.-P. & Ehresmann, B. (1987). Probing the structure of RNAs in solution. Nucl. Acids Res. 15, 9109–9128. Gampel, A. & Cech, T. R. (1991). Binding of the CBP2 protein to a yeast mitochondrial group I intron requires the catalytic core of the RNA. Genes Dev. 5, 1870–1880. Garriga, G. & Lambowitz, A. M. (1986). Protein-dependent splicing of a group I intron in ribonucleoprotein particles and soluble fractions. Cell, 46, 669–680. Giege, R., Puglisi, J. D. & Florentz, C. (1993). tRNA structure and aminoacylation efficiency. Prog. Nucl. Acid Res. 45, 129–205. Green, R., & Szostak, J. W. (1994). In vitro genetic analysis of the hinge region between helical elements P5-P4-P6 and P7-P3-P8 in the sunY group I self-splicing intron. J. Mol. Biol. 235, 140–155. Guo, Q. & Lambowitz, A. M. (1992). A tyrosyl-tRNA synthetase binds speciflcally to the group I intron catalytic core. Genes Dev. 6, 1357–1372. Guo, Q., Akins, R. A., Garriga, G. & Lambowitz, A. M. (1991). Structural analysis of the Neurospora mitochondrial large rRNA intron and construction of a mini-intron that shows protein-dependent splicing. J. Biol. Chem. 266, 1809–1819. Herschlag, D., Khosla, M., Tsuchihashi, Z. & Karpel, R. L. (1994). An RNA chaperone activity of non-speciflc RNA binding proteins in hammerhead ribozyme catalysis. EMBO J. 13, 2913–2924. Heuer, T. S., Chandry, P. S., Belfort, M., Celander, D. W. & Cech, T. R. (1991). Folding of group I introns from bacteriophage T4 involves internalization of the catalytic core. Proc. Natl Acad. Sci. USA, 88, 11105–11109.

Protein-induced Folding of Group I Introns

Himeno, H., Hasegawa, T., Ueda, T., Watanabe, K. & Shimizu, M. (1990). Conversion of aminoacylation specificity from tRNATyr to tRNASer in vitro. Nucl. Acids Res. 18, 6815–6819. Jaeger, L., Westhof, E. & Michel, F. (1991). Function of P11, a tertiary base-pairing in self-splicing introns of subgroup IA. J. Mol. Biol. 221, 1153–1164. Jaeger, L., Westhof, E. & Michel, F. (1993). Monitoring of the cooperative unfolding of the sunY group I intron of bacteriophage T4. The active form of the sunY ribozyme is stabilized by multiple interactions with 3' terminal intron components. J. Mol. Biol. 234, 331–346. Jaeger, L., Michel, F. & Westhof, E. (1994). Involvement of a GNRA tetraloop in long-range RNA tertiary interactions. J. Mol. Biol. 236, 1271–1276. Kittle, J. D. Jr, Mohr, G., Gianelos, J. A., Wang, H. & Lambowitz, A. M. (1991). The Neurospora mitochondrial tyrosyl-tRNA synthetase is sufficient for group I intron splicing in vitro and uses the carboxyterminal tRNA-binding domain along with other regions. Genes Dev. 5, 1009–1021. Laggerbauer, B., Murphy, F. L. & Cech, T. R. (1994). Two major tertiary folding transitions of the Tetrahymena catalytic RNA. EMBO J. 13, 2669–2676. Lambowitz, A. M. & Perlman, P. S. (1990). Involvement of aminoacyl-tRNA synthetases and other proteins in group I and group II intron splicing. Trends Biochem. Sci. 15, 440–444. Lazowska, J., Jacq, C. & Slonimski, P. P. (1980) Sequence of introns and flanking exons in wild-type and box3 mutants of cytochrome b reveals an interlaced splicing protein coded by an intron. Cell, 22, 328–333. Majumder, A. L., Akins, R. A., Wilkinson, J. G., Kelley, R. L., Snook, A. J. & Lambowitz, A, M. (1989). Involvement of tyrosyl-tRNA synthetase in splicing of group I introns in Neurospora crassa mitochondria: biochemical and immunochemical analysis of splicing activity. Mol. Cell. Biol. 9, 2089–2104. Mannella, C. A., Collins, R. A., Green, M. R. & Lambowitz, A. M. (1979). Defective splicing of mitochondrial rRNA in cytochrome-deficient nuclear mutants of Neurospora crassa. Proc. Natl Acad. Sci. USA, 76, 2635–2639. Massire, C., Gaspin, C. & Westhof, E. (1994). DRAWNA: a program for drawing schematic views of nucleic acids. J. Mol. Graphics, 12, 201–206. Michel, F. & Westhof, E. (1990). Modelling of the three-dimensional architecture of group I catalytic introns based on comparative sequence analysis. J. Mol. Biol. 216, 585–610. Michel, F., Ellington, A. D., Couture, S. & Szostak, J. W. (1990). Phylogenetic and genetic evidence for base-triples in the catalytic domain of group I introns. Nature, 347, 578–580. Mohr, G., Zhang, A., Gianelos, J. A., Belfort, M. & Lambowitz, A. M. (1992). The Neurospora CYT-18 protein suppresses defects in the phage T4 td intron by stabilizing the catalytically active structure of the intron core. Cell, 69, 483–494. Mohr, G., Caprara, M. G., Guo, Q. & Lambowitz, A. M. (1994). A tyrosyl-tRNA synthetase can function similarly to an RNA structure in the Tetrahymena ribozyme. Nature, 370, 147–150.

531 Murphy, F. L. & Cech, T. R. (1994). GAAA tetraloop and conserved bulge stabilize tertiary structure of a group I intron domain. J. Mol. Biol. 236, 49–63. Putz, J., Puglisi, J. D., Florentz, C. & Giege, R.(1991). Identity elements for specific aminoacylation of yeast tRNAAsp by cognate aspartyl-tRNA synthetase. Science, 252, 1696–1699. Pyle, A. M., Murphy, F. L. & Cech, T. R. (1992). RNA substrate binding site in the catalytic core of the Tetrahymena ribozyme. Nature, 358, 123–128. Rosenberg, A. H., Lade, B. N., Chui, D. S., Lin, S. W., Dunn, J. J. & Studier, F. W. (1987). Vectors for selective expression of cloned DNAs by T7 RNA polymerase. Gene, 56, 125–135. Rudinger, J., Puglisi, J. D., Putz, J., Schatz, D., Eckstein, F., Florentz, C. & Giege, R. (1992). Determinant nucleotides of yeast tRNAAsp interact directly with aspartyl-tRNA synthetase. Proc. Natl Acad. Sci. USA, 89, 5882–5886. Saenger, W. (1984). Principles of Nucleic Acid Structure. Springer-Verlag, NY, USA. Saldanha, R. J., Patel, S. S., Surendran, R., Lee, J. C. & Lambowitz, A. M. (1995). Involvement of Neurospora mitochondrial tyrosyl-tRNA synthetase in RNA splicing. A new method for purifying the protein and characterization of physical and enzymatic properties pertinent to splicing. Biochemistry, 34, 1275–1287. Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning: A Laboratory Manual, 2nd edit., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, NY, USA. Schatz, D., Leberman, R. & Eckstein, F. (1991). Interaction of Escherichia coli tRNASer with its cognate aminoacyltRNA synthetase as determined by footprinting with phosphorothioate-containing tRNA transcripts. Proc. Natl Acad. Sci. USA, 88, 6132–6136. Shaw, L. C. & Lewin, A. S. (1995). Protein-induced folding of a group I intron in cytochrome b pre-mRNA. J. Biol. Chem. 270, 21552–21562. Streicher, B., von Ahsen, U. & Schroeder, R. (1993). Lead cleavage sites in the core structure of group I intron-RNA. Nucl. Acids Res. 21, 311–317. Szostak, J. W. (1986). Enzymatic activity of the conserved core of a group I self-splicing intron. Nature, 322, 83–86. van der Horst, G., Christian, A. & Inoue, T. (1991). Reconstitution of a group I intron self-splicing reaction with an activator RNA. Proc. Natl Acad. Sci. USA, 88, 184–188. von Ahsen, U. & Noller, H. F. (1993). Methylation interference experiments identify bases that are essential for distinct catalytic functions of a group I ribozyme. EMBO J. 12, 4747–4754. Wang, J. F, Downs, W. D. & Cech, T. R. (1993). Movement of the guide sequence during RNA catalysis by a group I ribozyme. Science, 260, 504–508. Weeks, K. M. & Cech, T. R. (1995a). Efficient proteinfacilitated splicing of the yeast mitochondrial bI5 intron. Biochemistry, 34, 7728–7738. Weeks, K. M. & Cech, T. R. (1995b). Protein facilitation of group I intron splicing by assembly of the catalytic core and the 5' splice site domain. Cell, 82, 221–230. Zarrinkar, P. P. & Williamson, J. R. (1994). Kinetic intermediates in RNA folding. Science, 265, 918–924.

Edited by A. Klug (Received 9 October 1995; received in revised form 29 December 1995; accepted 8 January 1996)