Do DEAD-Box Proteins Promote Group II Intron Splicing without Unwinding RNA?

Molecular Cell

Short Article Do DEAD-Box Proteins Promote Group II Intron Splicing without Unwinding RNA? Mark Del Campo,1 Pilar Tijerina,1 Hari Bhaskaran,1 Sabine Mohr,1 Quansheng Yang,2,3 Eckhard Jankowsky,2,3 Rick Russell,1 and Alan M. Lambowitz1,* 1Institute

for Cellular and Molecular Biology, University of Texas at Austin, Austin, TX 78712, USA of Biochemistry 3Center for RNA Molecular Biology School of Medicine, Case Western Reserve University, Cleveland, OH 44106, USA *Correspondence: [email protected] DOI 10.1016/j.molcel.2007.07.028 2Department

SUMMARY

The DEAD-box protein Mss116p promotes group II intron splicing in vivo and in vitro. Here we explore two hypotheses for how Mss116p promotes group II intron splicing: by using its RNA unwinding activity to act as an RNA chaperone or by stabilizing RNA folding intermediates. We show that an Mss116p mutant in helicase motif III (SAT/AAA), which was reported to stimulate splicing without unwinding RNA, retains ATP-dependent unwinding activity and promotes unfolding of a structured RNA. Its unwinding activity increases sharply with decreasing duplex length and correlates with group II intron splicing activity in quantitative assays. Additionally, we show that Mss116p can promote ATP-independent RNA unwinding, presumably via single-strand capture, also potentially contributing to DEAD-box protein RNA chaperone activity. Our findings favor the hypothesis that DEAD-box proteins function in group II intron splicing as in other processes by using their unwinding activity to act as RNA chaperones. INTRODUCTION Group I and group II introns use RNA-catalyzed splicing reactions but require proteins for efficient splicing in vivo (Lambowitz et al., 1999). Such proteins may function by stabilizing the active RNA structure, by acting as RNA chaperones to resolve stable inactive structures that are ‘‘kinetic traps,’’ or by a combination of the two mechanisms. Proteins that bind specifically to group I or II intron RNAs and stabilize their active structures include the Neurospora crassa mitochondrial (mt) tyrosyl-tRNA synthetase (CYT-18 protein), the Saccharomyces cerevisiae CBP2 protein, and both group I and group II intronencoded maturases (Lambowitz et al., 1999). DEAD-box

proteins also function in splicing both group I and group II introns, typically acting in conjunction with splicing factors required for structural stabilization (Se´raphin et al., 1989; Mohr et al., 2002; Huang et al., 2005). DEAD-box proteins are found in all organisms and function in a wide variety of processes that require RNA structural rearrangements, including protein synthesis, ribosome assembly, and the splicing of nuclear pre-mRNAs (Tanner and Linder, 2001). Their involvement in group I and II intron splicing is important because the relatively simple in vitro splicing reactions of these introns allow detailed mechanistic studies that may shed light on how DEAD-box proteins function in more complex processes. DEAD-box proteins were first implicated in group I and II intron splicing by genetic analysis. In N. crassa, a mutation in the DEAD-box protein CYT-19 inhibits the splicing of several mt group I introns that require the CYT-18 protein for structural stabilization (Mohr et al., 2002). In S. cerevisiae, a related DEAD-box protein, Mss116p, was implicated similarly in splicing mt group I and II introns, including those that require intron-encoded maturases and other splicing factors for structural stabilization (Se´raphin et al., 1989; Huang et al., 2005). In addition to stimulating splicing, CYT-19 and Mss116p function in other mt RNA processing reactions and in translation. Further, the two proteins appear to act by similar mechanisms, because expression of CYT-19 suppresses all of the defects in mss116D strains (Huang et al., 2005). Biochemical studies showed that CYT-19 functions in group I intron splicing as an RNA chaperone that resolves inactive structures in an ATP-dependent manner (Mohr et al., 2002). At 25 C, the standard Neurospora growth temperature, CYT-19 strongly stimulates CYT-18-dependent splicing of an N. crassa mt group I intron, but the requirement for CYT-19 decreases at higher temperatures (e.g., 37 C), which allow more facile escape from kinetic traps. RNA structure mapping of the Tetrahymena thermophila LSU-DP5abc intron, whose splicing at low Mg2+ concentrations requires CYT-18, demonstrated that CYT-19 + ATP facilitates rearrangement from a misfolded to a functional structure (Mohr et al., 2002). Studies using a ribozyme derivative of the T. thermophila intron reinforced the conclusion that CYT-19 functions by resolving

Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc. 159

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

inactive RNA structures and provided additional mechanistic insight into how this is accomplished (Tijerina et al., 2006; Grohman et al., 2007). Despite the substantial evidence that DEAD-box proteins function in group I intron splicing as RNA chaperones, controversy has arisen about their role in group II intron splicing. The yeast genetic studies above showed that Mss116p functions in splicing all four mt group II introns: two group IIA introns, aI1 and aI2, which encode maturases for structural stabilization, and two small group IIB introns, aI5g and bI1, which do not encode maturases. We found that under near-physiological conditions CYT19 or Mss116p alone promotes the splicing of the aI5g and bI1 introns, as well as RNA cleavage by D135 RNA, a ribozyme derivative of aI5g (Mohr et al., 2006; Halls et al., 2007). In all three cases, binding of the DEAD-box protein alone is insufficient to promote the reaction, and ATP is absolutely required. For bI1, protease-digestion experiments using a reverse splicing assay showed that CYT-19 functions as an archetypical RNA chaperone, whose presence is no longer required after RNA folding (Mohr et al., 2006). For the D135 RNA, however, protease digestion or ATP depletion prior to the addition of substrate RNA abolished the reaction, suggesting that the DEAD-box protein accelerates a folding transition that follows substrate binding (Halls et al., 2007). Based on these findings, we proposed that DEAD-box proteins promote the splicing of group II introns primarily by using their ATP-dependent RNA unwinding activity to act as RNA chaperones, although they may also exert ATP-independent effects, such as structural stabilization, in some reactions (Halls et al., 2007). A different role for Mss116p in the splicing of aI5g was suggested by Solem et al. (2006). Although they also showed that Mss116p promotes splicing of aI5g under near-physiological conditions in an ATP-hydrolysis dependent manner, they proposed that Mss116p acts on aI5g and possibly other group I and II introns by stabilizing on-pathway folding intermediates. Although two other arguments were presented and are discussed below, this conclusion was based primarily on in vitro experiments with an Mss116p mutant in conserved RNA helicase motif III (SAT/AAA), which they suggested completely uncouples ATP hydrolysis from RNA unwinding. This mutant displayed no detectable RNA unwinding activity on a model 12 bp duplex but appeared to be only 2-fold less active than the wild-type (WT) protein in aI5g splicing. However, prior studies of the same Mss116p SAT/AAA mutant in vivo had shown residual activity not only in group II intron splicing but also in group I intron splicing and translational activation (Huang, 2004). As these other activities are expected to require disruption of nonfunctional structures, the results raised the possibility that Mss116p SAT/AAA retains some RNA unwinding activity. Here, we demonstrate that this is the case and that the residual unwinding activity with short duplexes correlates with group II intron splicing activity in quantitative assays. We also show that the SAT/AAA mutant can unfold a struc-

tured RNA, the T. thermophila group I ribozyme, in an ATP-dependent manner and that both WT and mutant Mss116p accelerate the unwinding of short duplexes in an ATP-independent manner, albeit less efficiently than with ATP. The simplest model for the action of DEADbox proteins in group II intron splicing remains that they function primarily as RNA chaperones, accelerating structural rearrangements by unwinding short RNA helices or disrupting other structural elements. RESULTS Purification and Biochemical Characterization of Mss116p SAT/AAA We expressed and purified the Mss116p WT and SAT/ AAA proteins by using a maltose-binding protein (MBP) fusion method (Halls et al., 2007). Both the WT and mutant proteins displayed RNA-dependent ATPase activity (kcat = 168–180 and 49–82 min1, respectively, in the presence of saturating concentrations of group I or II intron RNAs and ATP), and no detectable ATPase activity without RNA (Table S1). By comparison, RNA-stimulated ATPase activities reported by Solem et al. (2006) for proteins purified by a different method were lower (56.7 ± 13 and 9.1 ± 1 min1 for Mss116p WT and SAT/AAA, respectively), with significant background rates in the absence of RNA. Effect of the SAT/AAA Mutation on Unwinding a Short Duplex The conclusion of Solem et al. (2006) that the SAT/AAA mutation completely abolishes duplex unwinding was based on assays with a 12 bp duplex. These experiments left open the possibility that the mutant protein could unwind shorter duplexes, which are more common in the structured RNAs on which Mss116p normally acts. To investigate this possibility, we measured unwinding of the 6 bp P1 substrate duplex of the T. thermophila group I ribozyme, with P1 either free in solution or in its natural context, attached at the 50 end of the ribozyme. The latter provides a particularly useful model for DEADbox action on natural substrates (Tijerina et al., 2006). Figures 1A and 1B show that WT Mss116p + ATP efficiently unwound the free P1 duplex with a rate linearly dependent on protein concentration (%100 nM; kcat/Km = 6.8 3 106 M1 min1). Strikingly, Mss116p SAT/AAA + ATP also unwound this duplex and was only 7-fold less efficient under standard conditions (kcat/Km = 0.97 3 106 M1 min1). Similar results were obtained under group II intron splicing conditions (100 mM KCl, 8 mM Mg2+) at 25 C, with the mutant unwinding the duplex only 4-fold less efficiently than WT (Figure 1C). With the P1 duplex in its natural context, linked to the ribozyme, WT Mss116p was 16-fold more efficient than the mutant (Figure 1D). Both the WT and mutant proteins also unwound the P1 duplex in the absence of ATP (kcat/Km values 1.4 3 105 M1 min1 for WT and SAT/AAA with ribozyme-linked or free P1; Figure 1D and data not shown) or with the nonhydrolyzable analog AMP-PNP (kcat/Km = 2.5 3 105 M1

160 Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc.

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

min1 and 3.4 3 105 M1 min1 with ribozyme-linked P1 for WT and SAT/AAA, respectively; data not shown). Although less efficient than ATP hydrolysis-driven unwinding for both the WT and mutant proteins, such ATP-independent unwinding of short duplexes could contribute to the RNA chaperone activity of DEAD-box proteins on natural substrates (see Discussion).

Figure 1. Unwinding of the 6 bp P1 Duplex (A) Gels showing unwinding of P1 in solution under standard conditions. Cartoons show the duplex and separated strand, with the asterisk indicating the radiolabel. (B and C) Unwinding of P1 in solution at 25 C by Mss116p WT (circles) or SAT/AAA (triangles) under standard (B) or splicing (C) conditions. Dissociation rates without protein are also shown (squares), and kcat/ Km values are indicated. Both proteins also displayed P1 unwinding activity under splicing conditions at 30 C, but spontaneous helix dissociation was too fast to quantitatively measure enhancement by Mss116p (data not shown). (D) Unwinding of the P1 duplex linked to the ribozyme by Mss116p WT (circles) or SAT/AAA (triangles). Open and closed symbols are for reactions with or without saturating ATP-Mg2+ (2 mM), respectively.

Effect of the SAT/AAA Mutation on Unwinding Longer Duplexes To investigate the ability of the SAT/AAA mutant to unwind longer duplexes of the type used by Solem et al. (2006), we assayed unwinding of RNA oligonucleotide substrates that form a 13 bp duplex with a 25 nt 30 overhang. The assays were done both under standard conditions in the absence of free Mg2+ and under group II intron splicing conditions (see above). Again, Mss116p SAT/AAA had detectable activity, but for the longer duplex, the difference between the WT and mutant proteins was much larger. Under splicing conditions, both proteins displayed saturation and similar concentration dependences (i.e., similar K1/2 values), with WT 1300-fold more active (Figure 2). The observation of saturation for the 13 bp duplex, but not the 6 bp P1 duplex (Figure 1), suggests that functional binding is weaker to P1 or that binding to P1 is rate limiting for unwinding. Under low-salt conditions, unwinding activity of the 13 bp duplex increased such that only a lower limit could be obtained for the WT protein activity, which was at least 150-fold greater than that of the mutant. Under both conditions, WT Mss116p gave sigmoidal protein concentration dependences, suggesting the involvement of multiple protomers, whereas the mutant gave hyperbolic dependences, consistent with a single functional unit. With 13 bp duplexes, no activity was observed for the WT or mutant protein without ATP or with AMP-PNP under any of the above conditions (data not shown). Notably, when the duplex was shortened to 10 bp, the ATP-dependent unwinding activity of Mss116p SAT/AAA again increased dramatically, with kunwind = 0.87 and 4.5 min1 under splicing and standard conditions, respectively, the latter approaching the maximum rate measurable by this assay (Figure S1). Collectively, these findings show that the unwinding activity of the mutant relative to WT scales strongly with duplex length, with the mutant having relatively high activity on shorter duplexes common in natural structured RNAs. Effect of the SAT/AAA Mutation on Group II Intron Splicing Solem et al. (2006) reported that the SAT/AAA mutation caused only a ‘‘small reduction’’ in the efficiency of aI5g splicing relative to WT Mss116p. This conclusion was based on an experiment done at saturating protein concentration with a single reaction time point (120 min), which suggested that the mutant protein was 2-fold less active than WT. However, the true activity of the mutant protein may be substantially lower, because the use of

Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc. 161

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

Figure 2. Unwinding of a 13 bp Duplex (A and B) Reactions with Mss116p WT and SAT/AAA, respectively. Gels show reactions with 400 nM Mss116p under splicing (left) and standard lowsalt (right) conditions. Cartoons depict the duplex and separated strand as in Figure 1. Plots show the unwinding rate constant, calculated by including strand-annealing activity (Yang and Jankowsky, 2005), against protein concentration. Differences in the amount of RNA unwound in the steady state reflect that Mss116p’s annealing activity is higher under the low-salt conditions (Halls et al., 2007). Data points are averages of multiple measurements, error bars indicate standard deviations, and solid lines are best fits by Hill binding curves. Kinetic parameters are shown below.

a saturating protein concentration could obscure RNA binding defects of the mutant and because splicing by WT Mss116p presumably reached an endpoint well before the 120 min time point (6 half-lives; Halls et al., 2007), while the slower reaction with the mutant was likely still proceeding. To quantitatively assess the effect of the SAT/AAA mutation, we measured the splicing rates of the aI5g and bI1 group II introns in the presence of various concentrations of the WT and mutant proteins (Figure 3). As found previously, WT Mss116p gave ATP-dependent stimulation of aI5g splicing, saturating in the nM range and inhibiting splicing at higher concentrations (>40 nM; Halls et al., 2007). In the lower range, before saturation, the dependence of splicing rate on protein concentration was linear and gave a kcat/Km of 8.8 3 105 M1 min1 (Figure 3D). By contrast, the mutant protein required higher concentrations for saturation and gave a hint of upward curvature (Figure 3D). For comparison with the WT Mss116p, we estimated lower and upper limits of kcat/Km for the mutant by excluding or including the upward curvature and obtained values of 0.31 and 1.1 3 105 M1 min1, respec-

tively (dashed lines). Thus, the SAT/AAA mutant is 8- to 28fold less efficient than WT in promoting aI5g splicing, a considerably larger difference than reported by Solem et al. (2006). For the bI1 intron, the splicing efficiency was 25- to 52-fold lower for the mutant (Figures 3F–3H). Thus, the deleterious effects of the SAT/AAA mutation in group II intron splicing are comparable to its effects on unwinding (Figure 1), most simply suggesting that the defect in splicing results from diminished unwinding activity. Disruption of RNA Tertiary Structure by the SAT/AAA Mutant The ability to unwind a model helix does not necessarily indicate that a DEAD-box protein is capable of unfolding a structured RNA to an extent sufficient to function as an RNA chaperone. Thus, we explored whether Mss116p WT and SAT/AAA could mediate partial unfolding of the native T. thermophila group I intron ribozyme, with misfolding detected afterward by monitoring the ribozyme’s RNA cleavage activity (Russell et al., 2006; Tijerina et al., 2006). Both the WT and mutant proteins were active, giving ATP-dependent decreases in the fraction of native

162 Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc.

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

Figure 3. Group II Intron Splicing (A and E) Splicing reactions with group II introns aI5g and bI1, respectively, at 30 C for 120 min. Lane 1, RNA without incubation; lanes 2–10, reactions with indicated additions. P, precursor RNA; I-lar, intron lariat; I-lin, mix of linear intron and broken lariat; and E1-E2, ligated exons. (B, C, F, and G) Time courses of precursor RNA disappearance at 30 C in the presence or absence of the indicated concentrations of Mss116p WT or SAT/AAA plus 1 mM ATP-Mg2+. Data were fit with a single exponential. (D and H) Rate constant for precursor RNA disappearance as a function of protein concentration for aI5g and bI1, respectively. kcat/Km values are shown. Rate constants represented by closed symbols are from the time courses in (B), (C), (F), and (G), whereas open symbols are from an independent experiment.

ribozyme (Figure 4A). Similar to results in splicing assays, the mutant protein was 15-fold less efficient than the WT (Figure 4B). Further, both proteins showed low activity without ATP or with AMP-PNP (<0.8 3 105 M1 min1; data not shown). Thus, our results show that the SAT/ AAA mutant retains the ability to disrupt RNA tertiary structure in both an ATP-dependent and an ATP-independent manner. DISCUSSION Our results address whether DEAD-box proteins promote group II intron splicing by using their unwinding activity to act as RNA chaperones (Mohr et al., 2006; Halls et al., 2007) or by structural stabilization without RNA unwinding (Solem et al., 2006). The latter conclusion was based primarily on results seeming to show that an Mss116p SAT/AAA mutation completely abolished RNA-unwinding activity, with only a ‘‘small reduction’’ in aI5g splicing activity. By contrast, we find that, although the SAT/AAA mutant may less efficiently couple ATP hydrolysis to RNA unwinding, it in fact retains substantial ATP-dependent RNA unwinding activity. This activity scales strongly with duplex length, being >1000-fold lower than WT for a 13 bp duplex but only 4- to 16-fold lower for a 6 bp duplex, comparable to most helices in group I and II in-

trons. Moreover, our results show that the SAT/AAA mutation decreases group II intron splicing efficiency by 8to 52-fold, considerably more than estimated by Solem et al. (2006) and in the range expected from the decrease in unwinding activity with shorter duplexes. Additionally, the SAT/AAA mutant can partially unfold a structured RNA, the T. thermophila group I ribozyme, with an efficiency 15-fold lower than WT. Together, these findings indicate that the SAT/AAA mutant retains the ability to mediate ATP hydrolysis-dependent disruptions of RNA structure, negating the central argument that led Solem et al. (2006) to conclude that Mss116p activates group II intron splicing without unwinding RNA. The proposal of Solem et al. (2006) was also supported by two other arguments. The first was that the unwinding activity of DEAD-box proteins decreases at the higher Mg2+ concentration used in the splicing assays and is thus unlikely to be relevant. However, our results in Figure 1 as well as prior Mg2+-titration experiments (Halls et al., 2007) show that Mss116p in fact retains substantial unwinding activity under splicing conditions (100 mM KCl, 8 mM Mg2+). Second, previous studies by Pyle and coworkers were interpreted to indicate that folding of the aI5g intron and its ribozyme derivatives are not rate limited by kinetic traps (reviewed in Pyle et al., 2007), so that acceleration of structure disruption by a chaperone would

Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc. 163

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

Figure 4. Group I Ribozyme Unfolding (A) Progress curves of unfolding with Mss116p WT or SAT/AAA at 800 or 1500 nM. (B) Observed rate constants for loss of native ribozyme as a function of Mss116p WT (circles) and SAT/AAA (triangles) concentration. kcat/Km values are shown.

not be expected to give a rate enhancement. Although there is good evidence for lack of rate-limiting kinetic traps for aI5g ribozyme derivatives under the nonphysiological conditions used historically by Pyle and coworkers (500 mM KCl, 60–100 mM Mg2+ at 42 C), the conclusion that kinetic traps do not limit folding under more physiological conditions rests on the observation that subdenaturing urea concentrations do not accelerate folding (Fedorova et al., 2007). As noted by the authors, however, the lack of acceleration by urea leaves open the possibility that a kinetic trap is rate limiting if the necessary unfolding does not result in a substantial increase in solvent-exposed surface (Shelton et al., 1999; Fedorova et al., 2007). Additionally, the physiological relevance of the aI5g ribozyme constructs is unclear because they lack exons, which were shown previously to contribute to kinetic traps in aI5g precursor RNAs (Nolte et al., 1998). In the absence of contrary evidence from the SAT/AAA mutant, the finding that stimulation of group II intron splicing by DEAD-box proteins is strictly ATP dependent (Mohr et al., 2006; Solem et al., 2006; Halls et al., 2007) strongly implicates RNA unwinding activity in this process. Although ATP-independent activities of DEAD-box proteins, including RNA binding, strand annealing, and strand exchange, may contribute, they are not by themselves

sufficient to stimulate group II intron splicing to measurable levels. Notably, we find that DEAD-box proteins can accelerate the strand separation of short duplexes even without ATP hydrolysis. This activity may reflect ‘‘strand capture’’ that results from preferential binding to single-stranded RNA, and it may enable DEAD-box proteins to function as ATPindependent RNA chaperones, similar to many nonspecific single-stranded RNA binding proteins (Herschlag, 1995; Schroeder et al., 2004). Indeed, there may be a continuum of ATP-dependent and -independent activity, as ATP could participate in such ‘‘passive’’ RNA chaperoning by tightening binding to RNA and/or promoting successive cycles of RNA binding and release. The ATP-independent unwinding activity is higher for Mss116p than CYT-19 (P.T., M.D., A.M.L., and R.R., unpublished data), possibly explaining the ability of Mss116p, but not CYT-19, to partially stimulate in vitro splicing of a group I intron in the absence of ATP (Halls et al., 2007). However, as ATP is always present in the cell, ATP-driven duplex unwinding is likely to be the favored mode of operation of DEAD-box proteins in vivo. As general RNA chaperones, CYT-19 and Mss116p are hypothesized to resolve diverse kinetic traps. Both CYT19 and Mss116p accelerate partial unfolding of the T. thermophila group I ribozyme by disrupting tertiary structure stabilized by long-range tertiary contacts (Russell et al., 2006; Tijerina et al., 2006). In the case of group II introns, DEAD-box proteins may act by resolving nonnative structures, such as those resulting from exon-dependent misfolding in aI5g splicing (Nolte et al., 1998), or by accelerating the slow transition from the initial unfolded state to a state with a folded DI, which serves as a scaffold for assembly of the other RNA domains (Fedorova et al., 2007). This slow transition may involve bending of an internal loop (Waldsich and Pyle, 2007) and may require transient unstacking of coaxially stacked helices, a process that could be intrinsically slow and accelerated by a chaperone (Esteban et al., 1998; Zhuang et al., 2002). Supplementing its primary ATP-dependent role, Mss116p binding could help stabilize the bend or the interactions of other RNA domains with the folded DI at low Mg2+ concentrations (cf., Fedorova et al., 2007; Waldsich and Pyle, 2007). The latter interactions may be stabilized by maturases in other group II introns (Lambowitz et al., 1999). We note that splicing of aI5g and bI1 by DEADbox proteins alone is much slower than maturase-promoted splicing of the Ll.LtrB group II intron (Saldanha et al., 1999). Thus, it is possible that other as yet unidentified proteins stabilize the structures of aI5g and bI1 in vivo. Finally, our results favor a single unified mechanism by which Mss116p functions in group I and II intron splicing and in translational activation: by using its RNA-unwinding activity to act as an RNA chaperone (Herschlag, 1995). The ability of Mss116p to act as an RNA chaperone on such a wide range of structurally diverse RNAs presumably reflects that it binds these RNAs with minimal specificity, allowing its unwinding activity to resolve

164 Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc.

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

different kinetic traps. This mechanism is supported by the interchangeability of Mss116p with the DEAD-box proteins CYT-19 and Ded1p, which also function in diverse cellular processes (Solem et al., 2006; Tijerina et al., 2006; Halls et al., 2007). Further, the similar activities suggest that these other proteins use analogous mechanisms to perform their natural functions. Indeed, DEAD-box proteins that function as general RNA chaperones may play important roles in the RNA metabolism of all organisms.

labeled substrate (CCCUCUA5) that was rapidly cleaved by the ribozyme. Supplemental Data Supplemental Data include one figure and one table and can be found with this article online at http://www.molecule.org/cgi/content/full/ 28/1/159/DC1/. ACKNOWLEDGMENTS This work was supported by NIH grants GM067700 to E.J., GM070456 to R.R., and GM037951 to A.M.L. M.D. is the recipient of NIH postdoctoral fellowship F32-GM76961.

EXPERIMENTAL PROCEDURES Protein Purification Mss116p WT and SAT/AAA were expressed as fusions to MBP using pMAL-derived expression plasmids, with the Mss116p SAT/AAA plasmid constructed by PCR amplifying the Mss116p SAT/AAA ORF (codons 37–664) of pHRH182 (Huang, 2004) and cloning it between the BamHI and HindIII sites of pMAL-c2t (Halls et al., 2007). The proteins were purified as described (Halls et al., 2007), except that prior to TEV protease digestion, glycerol was added (10% v/v), and after protease digestion, proteins were purified by gel filtration (Superdex 200; GE Healthcare Biosciences Corp., Piscataway, NJ) in 20 mM Tris-HCl (pH 7.5), 500 mM KCl, 1 mM EDTA, 1 mM DTT, and 10% glycerol. Peak fractions were pooled, mixed with 1 ml amylose resin to remove residual MBP, filtered through a 0.45 mm polyethersulfone membrane, and dialyzed overnight at 4 C against 20 mM Tris-HCl (pH 7.5), 500 mM KCl, 1 mM EDTA, 1 mM DTT, and 50% glycerol. Proteins were flash frozen and stored at 80 C. RNA Unwinding Unwinding of the P1 duplex was assayed by native PAGE (Tijerina et al., 2006). Reactions were at 25 C with 50 mM Na-MOPS (pH 7.0), 50 mM KCl, 10 mM MgCl2, 5% glycerol, and 2 mM ATP-Mg2+. Trace 32 P-labeled CCCUCUA5 was incubated with the complementary oligonucleotide GGAGGGA or with the T. thermophila LSU ribozyme for 5 min to form the P1 helix. The RNA was then diluted to 10 nM and excess unlabeled CCCUCUA5 was added (2.5 mM). Reactions were initiated by adding protein. Unwinding assays with 13 bp duplex substrates RNA were done at 30 C with 0.1 nM [32P]-RNA substrate and 2 mM ATP-Mg2+ in 30 ml of reaction medium containing 50 mM KCl (standard conditions) or 100 mM KCl, 8 mM MgCl2 (splicing conditions), 20 mM Tris-HCl (pH 7.5), 2 mM DTT, 0.5 U/ml RNasin, and 0.01% (v/v) NP-40 (Halls et al., 2007). RNA oligonucleotides were 50 -AGCACCGUAAAGA, 30 -(A4C)4AAAAUUCGUGGCAUUUCU. RNA Splicing Group II intron splicing reactions were at 30 C with 10 nM 32P-labeled RNA and indicated amounts of Mss116p in 100 mM KCl, 8 mM MgCl2, 10 mM Tris-HCl (pH 7.5), 10% glycerol, and 1 mM ATP-Mg2+, ADPMg2+, or AMP-PNP-Mg2+. Ten millimolar stocks of ADP and AMPPNP were treated for 60 min at 30 C with 1.5 U/ml hexokinase (Roche Diagnostics) plus 25 mM glucose to eliminate residual ATP. Reactions were initiated by adding RNA. RNA Unfolding with the T. thermophila LSU Ribozyme Partial unfolding of the ribozyme (A-rich bulge mutant, 183-188U used to facilitate quantitation; Russell et al., 2006) was monitored as described (Tijerina et al., 2006), except that the RNA was preincubated (50 C for 30 min in the presence of 10 mM Mg2+) to generate the native ribozyme. Unfolding reactions were at 25 C in 50 mM Na-MOPS (pH 7.0), 50 mM KCl, 5 mM MgCl2, 5% glycerol, and 2 mM ATP-Mg2+. Reactions were initiated by adding protein and stopped by adding 50 mM Mg2+ and 1 mg/ml Proteinase K. The fraction of native ribozyme was determined after quenching by measuring the fraction of input 32P-

Received: May 10, 2007 Revised: July 12, 2007 Accepted: July 26, 2007 Published: October 11, 2007 REFERENCES Esteban, J.A., Walter, N.G., Kotzorek, G., Heckman, J.E., and Burke, J.M. (1998). Structural basis for heterogeneous kinetics: reengineering the hairpin ribozyme. Proc. Natl. Acad. Sci. USA 95, 6091–6096. Fedorova, O., Waldsich, C., and Pyle, A.M. (2007). Group II intron folding under near-physiological conditions: collapsing to the near-native state. J. Mol. Biol. 366, 1099–1114. Grohman, J.K., Del Campo, M., Bhaskaran, H., Tijerina, P., Lambowitz, A.M., and Russell, R. (2007). Probing the mechanisms of DEAD-box proteins as general RNA chaperones: the C-terminal domain of CYT19 mediates general recognition of RNA. Biochemistry 46, 3013–3022. Halls, C., Mohr, S., Del Campo, M., Yang, Q., Jankowsky, E., and Lambowitz, A.M. (2007). Involvement of DEAD-box proteins in group I and group II intron splicing. Biochemical characterization of Mss116p, ATP hydrolysis-dependent and -independent mechanisms, and general RNA chaperone activity. J. Mol. Biol. 365, 835–855. Herschlag, D. (1995). RNA chaperones and the RNA folding problem. J. Biol. Chem. 270, 20871–20874. Huang, H.R. (2004) Functional studies of intron- and nuclear-encoded splicing factors in the mitochondria of Saccharomyces cerevisiae, PhD thesis, The University of Texas Southwestern Medical Center, Dallas, Texas. Huang, H.R., Rowe, C.E., Mohr, S., Jiang, Y., Lambowitz, A.M., and Perlman, P.S. (2005). The splicing of yeast mitochondrial group I and group II introns requires a DEAD-box protein with RNA chaperone function. Proc. Natl. Acad. Sci. USA 102, 163–168. Lambowitz, A.M., Caprara, M.G., Zimmerly, S., and Perlman, P.S. (1999). Group I and group II ribozymes as RNPs: clues to the past and guides to the future. In The RNA World, R.F. Gesteland, T.R. Cech, and J.F. Atkins, eds. (Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press), pp. 451–485. Mohr, S., Stryker, J.M., and Lambowitz, A.M. (2002). A DEAD-box protein functions as an ATP-dependent RNA chaperone in group I intron splicing. Cell 109, 769–779. Mohr, S., Matsuura, M., Perlman, P.S., and Lambowitz, A.M. (2006). A DEAD-box protein alone promotes group II intron splicing and reverse splicing by acting as an RNA chaperone. Proc. Natl. Acad. Sci. USA 103, 3569–3574. Nolte, A., Chanfreau, G., and Jacquier, A. (1998). Influence of substrate structure on in vitro ribozyme activity of a group II intron. RNA 4, 694–708. Pyle, A.M., Fedorova, O., and Waldsich, C. (2007). Folding of group II introns: a model system for large, multidomain RNAs? Trends Biochem. Sci. 32, 138–145.

Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc. 165

Molecular Cell DEAD-Box Proteins in Group II Intron Splicing

Russell, R., Das, R., Suh, H., Travers, K., Laederach, A., Engelhardt, M., and Herschlag, D. (2006). The paradoxical behavior of a highly structured misfolded intermediate in RNA folding. J. Mol. Biol. 363, 531–544. Saldanha, R., Chen, B., Wank, H., Matsuura, M., Edwards, J., and Lambowitz, A.M. (1999). RNA and protein catalysis in group II intron splicing and mobility reactions using purified components. Biochemistry 38, 9069–9083. Schroeder, R., Barta, A., and Semrad, K. (2004). Strategies for RNA folding and assembly. Nat. Rev. Mol. Cell Biol. 5, 908–919. Se´raphin, B., Simon, M., Boulet, A., and Faye, G. (1989). Mitochondrial splicing requires a protein from a novel helicase family. Nature 337, 84–87. Shelton, V.M., Sosnick, T.R., and Pan, T. (1999). Applicability of urea in the thermodynamic analysis of secondary and tertiary RNA folding. Biochemistry 38, 16831–16839.

Solem, A., Zingler, N., and Pyle, A.M. (2006). A DEAD protein that activates intron self-splicing without unwinding RNA. Mol. Cell 24, 611–617. Tanner, N.K., and Linder, P. (2001). DExD/H box RNA helicases: from generic motors to specific dissociation functions. Mol. Cell 8, 251–262. Tijerina, P., Bhaskaran, H., and Russell, R. (2006). Non-specific binding to structured RNA and preferential unwinding of an exposed helix by the CYT-19 protein, a DEAD-box RNA chaperone. Proc. Natl. Acad. Sci. USA 103, 16698–16703. Yang, Q., and Jankowsky, E. (2005). ATP- and ADP-dependent modulation of RNA unwinding and strand annealing activities by the the DEAD-box protein Ded1. Biochemistry 44, 13591–13601. Waldsich, C., and Pyle, A.M. (2007). A folding control element for tertiary collapse of a group II intron ribozyme. Nat. Struct. Mol. Biol. 14, 37–44. Zhuang, X., Kim, H., Pereira, M.J., Babcock, H.P., Walter, N.G., and Chu, S. (2002). Correlating structural dynamics and function in single ribozyme molecules. Science 296, 1473–1476.

166 Molecular Cell 28, 159–166, October 12, 2007 ª2007 Elsevier Inc.