RNA Substrate Specificity and Structure-guided Mutational Analysis of Bacteriophage T4 RNA Ligase 2
2004; Elsevier BV; Volume: 279; Issue: 30 Linguagem: Inglês
10.1074/jbc.m402394200
ISSN1083-351X
AutoresJayakrishnan Nandakumar, C. Kiong Ho, Christopher D. Lima, Stewart Shuman,
Tópico(s)Bacterial Genetics and Biotechnology
ResumoHere we report that bacteriophage T4 RNA ligase 2 (Rnl2) is an efficient catalyst of RNA ligation at a 3′-OH/5′-PO4 nick in a double-stranded RNA or an RNA·DNA hybrid. The critical role of the template strand in approximating the reactive 3′-OH and 5′-PO4 termini is underscored by the drastic reductions in the RNA-sealing activity of Rnl2 when the duplex substrates contain gaps or flaps instead of nicks. RNA nick joining requires ATP and a divalent cation cofactor (either Mg or Mn). Neither dATP, GTP, CTP, nor UTP can substitute for ATP. We identify by alanine scanning seven functionally important amino acids (Tyr-5, Arg-33, Lys-54, Gln-106, Asp-135, Arg-155, and Ser-170) within the N-terminal nucleotidyl-transferase domain of Rnl2 and impute specific roles for these residues based on the crystal structure of the AMP-bound enzyme. Mutational analysis of 14 conserved residues in the C-terminal domain of Rnl2 identifies 3 amino acids (Arg-266, Asp-292, and Glu-296) as essential for ligase activity. Our findings consolidate the evolutionary connections between bacteriophage Rnl2 and the RNA-editing ligases of kinetoplastid protozoa. Here we report that bacteriophage T4 RNA ligase 2 (Rnl2) is an efficient catalyst of RNA ligation at a 3′-OH/5′-PO4 nick in a double-stranded RNA or an RNA·DNA hybrid. The critical role of the template strand in approximating the reactive 3′-OH and 5′-PO4 termini is underscored by the drastic reductions in the RNA-sealing activity of Rnl2 when the duplex substrates contain gaps or flaps instead of nicks. RNA nick joining requires ATP and a divalent cation cofactor (either Mg or Mn). Neither dATP, GTP, CTP, nor UTP can substitute for ATP. We identify by alanine scanning seven functionally important amino acids (Tyr-5, Arg-33, Lys-54, Gln-106, Asp-135, Arg-155, and Ser-170) within the N-terminal nucleotidyl-transferase domain of Rnl2 and impute specific roles for these residues based on the crystal structure of the AMP-bound enzyme. Mutational analysis of 14 conserved residues in the C-terminal domain of Rnl2 identifies 3 amino acids (Arg-266, Asp-292, and Glu-296) as essential for ligase activity. Our findings consolidate the evolutionary connections between bacteriophage Rnl2 and the RNA-editing ligases of kinetoplastid protozoa. Bacteriophage T4 encodes two RNA strand-joining enzymes, RNA ligase 1 (Rnl1) 1The abbreviations used are: Rnl, RNA ligase; REL, RNA-editing ligase; ds, double-stranded; DTT, dithiothreitol. and RNA ligase 2 (Rnl2), that exemplify different branches of the RNA ligase family (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar). The function of Rnl1 in vivo is to repair a break in the anticodon loop of Escherichia coli tRNALys triggered by phage activation of a host-encoded anticodon nuclease (2Amitsur M. Levitz R. Kaufman G. EMBO J. 1987; 6: 2499-2503Google Scholar). Rnl1-like ligases are few in number, and they have a relatively narrow phylogenetic distribution that is limited, as far as we know, to bacteriophages, fungi, and baculoviruses (3Uhlenbeck O.C. Gumport R.I. Enzymes. 1982; 15: 31-58Google Scholar, 4Baymiller J. Jennings S. Kienzle B.K. Gorman J.A. Kelly R. McCullough J.E. Gene. 1994; 142: 129-134Google Scholar, 5Phizicky E.M. Schwartz R.C. Abelson J. J. Biol. Chem. 1986; 261: 2978-2986Google Scholar, 6Durantel D. Croizier L. Ayres M.D. Croizier G. Possee R.D. Lopez-Ferber M. J. Gen. Virol. 1998; 79: 629-637Google Scholar, 7Sawaya R. Schwer B. Shuman S. J. Biol. Chem. 2003; 278: 43298-43398Google Scholar, 8Blondal T. Hjorleifdottir S.H. Fridjonsson O.F. Evarsson A. Skirnissdottir S. Hermannsdottir A.G. Hreggvidsson G.O. Smith A.V. Kristjansson J.K. Nucleic Acids Res. 2003; 31: 7247-7252Google Scholar, 9Martins A. Shuman S. J. Biol. Chem. 2004; (in press)Google Scholar). T4 Rnl2 typifies a separate branch (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar) that includes vibriophage KVP40 Rnl2 (10Yin S. Ho C.K. Miller E.S. Shuman S. Virology. 2004; 319: 141-151Google Scholar), the RNA-editing ligases (RELs) of Trypanosoma and Leishmania (11McManus M.T. Shimamura M. Grams J. Hajduk S.L. RNA. 2001; 7: 167-175Google Scholar, 12Rusche L.N. Huang C.E. Piller K.J. Hemann M. Wirtz E. Sollner-Webb B. Mol. Cell. Biol. 2001; 21: 979-989Google Scholar, 13Schnaufer A. Panigrahi A.K. Panicucci B. Igo R.P. Salavati R. Stuart K. Science. 2001; 291: 2159-2162Google Scholar) (Fig. 1), putative RNA ligases encoded by certain eukaryotic viruses, and putative RNA ligases encoded by many species of archaea (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar). Thus, the Rnl2-like ligases are present in all three phylogenetic domains. The function of T4 Rnl2 during phage infection is unknown. RNA ligases join 3′-OH and 5′-PO4 RNA termini through a series of three nucleotidyl transfer steps (3Uhlenbeck O.C. Gumport R.I. Enzymes. 1982; 15: 31-58Google Scholar, 14Silber R. Malathi V.G. Hurwitz J. Proc. Natl. Acad. Sci. U. S. A. 1972; 69: 3009-3013Google Scholar, 15Cranston J.W. Silber R. Malathi V.G. Hurwitz J. J. Biol. Chem. 1974; 249: 7447-7456Google Scholar, 16Sugino A. Snopek T.J. Cozzarelli N.R. J. Biol. Chem. 1977; 252: 1732-1738Google Scholar). Step 1 is the reaction of ligase with ATP to form a covalent ligase-(lysyl-N)-AMP intermediate and pyrophosphate. In Step 2, the AMP is transferred from ligase-adenylate to the 5′-PO4 RNA end to form an RNA-adenylate intermediate (AppRNA). In Step 3, attack by an RNA 3′-OH on the RNA-adenylate seals the two ends via a phosphodiester bond and releases AMP. Biochemical characterization of T4 and KVP40 Rnl2 revealed an interesting effect of ATP whereby reaction of Rnl2 with a 5′-PO4 single-stranded 18-mer RNA in the presence of ATP resulted in the accumulation of high levels of AppRNA and scant RNA end sealing (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar, 17Yin S. Ho C.K. Shuman S. J. Biol. Chem. 2003; 278: 17601-17608Google Scholar). This ATP-trapping effect is likely caused by dissociation of Rnl2 from newly formed AppRNA, followed immediately by re-adenylylation of Rnl2, which precludes the third step of the strand-joining pathway. The biochemical properties of the phage Rnl2 enzymes raised questions about the kinds of reactions they might catalyze in a cellular milieu where ATP is present at millimolar concentrations (10Yin S. Ho C.K. Miller E.S. Shuman S. Virology. 2004; 319: 141-151Google Scholar). One possibility is that Rnl2 functions to adenylate pRNA ends in vivo. The resulting 5′ AppRNA terminus could potentially influence the stability of host or phage-derived RNAs and, in the case of mRNAs, affect their efficiency of translation. Note that ATP-dependent synthesis of AppRNA by Rnl2 is reminiscent of the capping of eukaryotic mRNA by GTP-RNA guanylyltransferase. The analogy between Rnl2 and capping enzymes is underscored by: (i) the presence of shared primary structure motifs I, III, IIIa, IV, and V that define the covalent nucleotidyltransferase superfamily (18Shuman S. Liu Y. Schwer B. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 12046-12050Google Scholar, 19Shuman S. Schwer B. Mol. Microbiol. 1995; 17: 405-410Google Scholar) (Fig. 1); (ii) concordance of site-directed mutational analyses at homologous residues of capping enzymes and T4 Rnl2, consistent with a shared constellation of catalytic residues (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar, 17Yin S. Ho C.K. Shuman S. J. Biol. Chem. 2003; 278: 17601-17608Google Scholar, 20Sawaya R. Shuman S. Biochemistry. 2003; 42: 8240-8249Google Scholar); and (iii) similarity in the tertiary structures of the N-terminal nucleotidyltransferase domains of eukaryotic capping enzymes and T4 Rnl2 (21Håkansson K. Doherty A.J. Shuman S. Wigley D.B. Cell. 1997; 89: 545-553Google Scholar, 22Fabrega C. Shen V. Shuman S. Lima C.D. Mol. Cell. 2003; 11: 1549-1561Google Scholar, 23Ho C.K. Wang L.K. Lima C.D. Shuman S. Structure. 2004; 12: 327-339Google Scholar). In a second scenario, we speculated that the biological function of Rnl2 is indeed ATP-dependent RNA strand joining and that Rnl2 either recognizes a specific RNA structure, from which it would not dissociate after the RNA adenylation step, or it requires a partner protein that anchors it to RNA adenylate to promote completion of the sealing step (10Yin S. Ho C.K. Miller E.S. Shuman S. Virology. 2004; 319: 141-151Google Scholar). Studies of the kinetoplastid RNA-editing ligases (which we classify as Rnl2-like enzymes) do indicate that the RELs prefer to join RNA termini that are splinted together by a bridging RNA template strand (24Blanc V. Alfonso J.D. Aphasizhev R. Simpson L. J. Biol. Chem. 1999; 274: 24289-24296Google Scholar, 25Palazzo S.S. Panigrahi A.K. Igo R.P. Salavati R. Stuart K. Mol. Biochem. Parasitol. 2003; 127: 161-167Google Scholar). It is notable that the specificity for RNA versus DNA as the template bridge differed between recombinant REL and the REL present in the native “editosome” complex, suggesting that editosome components associated with REL can alter its substrate preference (25Palazzo S.S. Panigrahi A.K. Igo R.P. Salavati R. Stuart K. Mol. Biochem. Parasitol. 2003; 127: 161-167Google Scholar). We report here that T4 Rnl2 displays a vigorous ATP-stimulated RNA-sealing activity when the reactive 3′-OH and 5′-PO4 RNA termini are opposed at a nick in a doubled-stranded (ds) RNA or an RNA·DNA hybrid. Rnl2 activity is reduced sharply, and then abolished, when the reactive RNA ends are separated incrementally by 1-, 2-, or 3-nucleotide gaps. Activity is also inhibited when the RNA ends protrude as flaps from the template strand. We exploited the newly defined optimal nicked substrate to conduct a mutational analysis of Rnl2, focusing on amino acids that are conserved in vibriophage Rnl2 and the protozoan RELs (Fig. 1). We have identified seven functionally important amino acids within the N-terminal nucleotidyltransferase domain of Rnl2 and suggest specific roles for these residues based on the available crystal structure of the AMP-bound protein. In addition, mutational analysis of 14 conserved residues in the C-terminal domain identifies 3 amino acids as essential for ligase activity. Our findings underscore the evolutionary links between Rnl2 and the RNA-editing ligases. We suggest that templated RNA repair is an ancient phenomenon that is probably not limited presently to kinetoplastid protozoa. T4 Rnl2—Missense mutations were introduced into the RNL2 gene by PCR using the two-stage overlap extension method as described previously (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar, 17Yin S. Ho C.K. Shuman S. J. Biol. Chem. 2003; 278: 17601-17608Google Scholar). The PCR products were digested with NdeI and BamHI and inserted into pET16b (Novagen). The inserts of the mutant pET-RNL2 plasmids were sequenced completely to exclude the acquisition of unwanted changes during amplification and cloning. The pET-RNL2 plasmids were transformed into E. coli BL21(DE3). Wild-type Rnl2 and mutated versions thereof were produced as follows (except as noted below). Cultures (200 ml) of E. coli BL21(DE3)/pET-RNL2 were grown at 37 °C in Luria Bertani medium containing 0.1 mg/ml ampicillin until the A600 reached ∼0.45. The cultures were adjusted to 0.4 mm isopropyl-β-d-thiogalactoside, and incubation was continued at 37 °C for 4 h. Cells were harvested by centrifugation, and the pellet was stored at –80 °C. All subsequent procedures were performed at 4 °C. Thawed bacteria were resuspended in 10 ml of buffer A (50 mm Tris-HCl (pH 7.5), 0.25 m NaCl, 10% sucrose). Lysozyme and Triton X-100 were added to final concentrations of 50 μg/ml and 0.1%, respectively. The lysates were sonicated to reduce their viscosity, and insoluble material was removed by centrifugation. The soluble extracts were applied to 1-ml columns of nickel-nitrilotriacetic acid-agarose (Qiagen, Chatsworth, CA) that had been equilibrated with buffer A containing 0.1% Triton X-100. The columns were washed with 5 ml of the same buffer and then eluted stepwise with 2-ml aliquots of 50, 100, 200, and 500 mm imidazole in buffer B (50 mm Tris-HCl (pH 8.0), 0.25 m NaCl, 10% glycerol, 0.05% Triton X-100). The polypeptide compositions of the column fractions were monitored by SDS-PAGE. Wild-type and mutant Rnl2 were recovered predominantly in the 200-mm imidazole eluates, which typically contained 2–4 mg of protein. The Rnl2 preparations were stored at –80 °C. Protein concentrations were determined with the BioRad dye reagent using bovine serum albumin as the standard. Rnl2 mutants R33A, D135A, and R155A were insoluble when produced by isopropyl-β-d-thiogalactoside induction at 37 °C as described above. Solubility was improved by varying the induction method as follows. The cultures were grown at 37 °C until the A600 reached ∼0.45. Then they were placed on ice for 30 min, after which they were adjusted to 0.4 mm isopropyl-β-d-thiogalactoside and 2% (v/v) ethanol and then incubated at 17 °C for 20 h with continuous shaking. Wild-type Rnl2 was also produced using this protocol. The wild-type and mutant proteins were purified from soluble extracts by nickel-agarose chromatography as described above. Adenylyltransferase Assay—Reaction mixtures (20 μl) containing 50 mm Tris acetate (pH 6.5) or 50 mm Tris-HCl (pH 9.0), 12 mm NaCl, 5 mm DTT, 1 mm MgCl2, 20 μm [α-32P]ATP, and 25 pmol of wild-type or mutant Rnl2 were incubated for 5 min at 22 °C. The reactions were quenched with SDS, and the products were analyzed by SDS-PAGE. The Rnl2·[32P]AMP adduct was visualized by autoradiography of the dried gel and quantified by scanning the gel with a Fujix BAS2500 imager. RNA Ligase Substrates—Oligoribonucleotides were purchased from Dharmacon (Lafayette, CO) and deprotected as instructed by the vendor. Oligodeoxyribonucleotides were purchased from BIOSOURCE International (Camarillo, CA). The RNA or DNA strands were 5′ 32P-labeled using T4 polynucleotide kinase and [γ-32P]ATP and then purified by electrophoresis through a nondenaturing 18% polyacrylamide gel. To form the nicked dsRNA and dsDNA substrates, complementary RNA or DNA strands were annealed at equimolar concentrations (1 nm each) in 150 mm NaCl, 10 mm Tris-HCl (pH 8.0), 1 mm EDTA by incubation for 10 min at 65 °C, followed by incubation for 15 min at 37 °C and then 30 min at 22 °C. To form the RNA-labeled nicked RNA·DNA hybrid, the 32P-labeled RNA strand was annealed to a 2-fold molar excess of an unlabeled 5-OH-terminated DNA strand. The ligase substrates were stored at –20 °C and thawed on ice immediately prior to use. RNA Ligase Assay—Reaction mixtures (10 μl) containing 50 mm Tris acetate (pH 6.5), 40 mm NaCl, 5 mm DTT, 1 mm MgCl2, 1 pmol of 5′ 32P-labeled ligase substrate, and ATP and Rnl2 as specified were incubated for 10 min at 22 °C. The reactions were quenched by adding 10 μl of 90% formamide, 20 mm EDTA. The samples were analyzed by electrophoresis through a 14-cm 6% polyacrylamide gel containing 7 m urea in 45 mm Tris borate, 1 mm EDTA for 75 min at 4 watts of constant power. The ligation products were visualized by autoradiography and quantified by scanning the gel with a Fujix BAS2500 imager. ATP-stimulated Sealing of a Nicked dsRNA Substrate by T4 Rnl2—A nicked dsRNA substrate was prepared by annealing two 5′ 32P-labeled synthetic 24-mer RNA oligonucleotides with overlapping complementarity. The strands were designed to form a 12-bp duplex with complementary 12-nucleotide 5′-tails (Fig. 2C) that can self-assemble into an extended duplex containing staggered 3′-OH/5′-PO4 nicks on both strands. Rnl2 was incubated with 100 nm of the annealed 5′ 32P-labeled RNA strands in the presence of Mg and ATP. The reaction products were analyzed by gel electrophoresis under denaturing conditions. As shown in Fig. 2A, Rnl2 generated a mixed ladder of multimers of the 24-mer strands. The yield of ligated products was saturated in the range of Rnl2 concentrations employed in this titration experiment (6–100 nm). Equivalent concentrations of Rnl2 were incapable of joining a nicked dsDNA substrate of identical sequence and complementarity (Fig. 2, A and C). A control reaction confirmed that the nicked dsDNA substrate was reactive with T4 DNA ligase (Fig. 2A). These results show that T4 Rnl2 specifically seals RNA strands at a duplex nick. A back-titration experiment established that the extent of ligation of 100 nm dsRNA nicks increased with increasing enzyme concentration in the range of 0.2–1.6 nm Rnl2 and attained saturation at ≥1.6 nm Rnl2 (Fig. 2B). The instructive finding was that dsRNA ligation by Rnl2 was dependent on added ATP (Fig. 2B). Simple comparison of the Rnl2 titration profiles with and without ATP indicated that the nicked RNA ligase activity was stimulated at least 16-fold by ATP (Fig. 2B). The trace levels of ligated product in the absence of ATP are attributable to the presence of preadenylated Rnl2·AMP in the recombinant enzyme preparation (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar). The stimulation of dsRNA nick ligation by ATP contrasts sharply with the ATP inhibition of ligation of single-stranded 18-mer RNA substrates reported previously for T4 and KVP40 Rnl2 (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar, 10Yin S. Ho C.K. Miller E.S. Shuman S. Virology. 2004; 319: 141-151Google Scholar, 17Yin S. Ho C.K. Shuman S. J. Biol. Chem. 2003; 278: 17601-17608Google Scholar). DNA-templated Ligation of Nicked RNA—A nicked dsRNA·DNAsubstrate was prepared by annealing a 5′ 32P-labeled 24-mer RNA oligonucleotide to an unlabeled 24-mer DNA oligonucleotide with overlapping complementarity (Fig. 3B). The annealed strands form a 12-bp duplex with complementary 12-nucleotide 5′-tails that promote self-assembly into an RNA·DNA hybrid duplex containing staggered 3′-OH/5′-PO4 RNA nicks on one strand. (Note that the DNA strand of the RNA·DNA hybrid contains staggered 3′-OH/5′-OH nicks, which cannot be sealed by polynucleotide ligases.) Increasing concentrations of Rnl2 were reacted with 100 nm of the annealed 5′ 32P RNA-labeled RNA·DNA hybrid in parallel with the dsRNA substrate. Rnl2 displayed similar activity in sealing 3′-OH/5′-PO4 RNA ends that were splinted by either an RNA or DNA template strand (Fig. 3A). The salient difference was that the ladder of sealed 32P-labeled RNAs generated with the RNA·DNA hybrid substrate comprised a perfectly spaced series of n-mers, with n values ranging from 2–12. In contrast, the ladder of sealed 32P-labeled RNAs generated with the dsRNA substrate was spaced irregularly (Fig. 3A). The irregular mobility of the dsRNA ligation products may reflect the fact that both component RNA strands contain reactive 5′-PO4 termini that, when situated at the ends of the substrate, can be joined by Rnl2 to the 3′-OH terminus of the complementary RNA strand to produce a hairpin product that migrates aberrantly during electrophoresis. In the case of the RNA·DNA hybrid substrate, the 5′-OH of the DNA strand cannot be ligated to the 3′-OH of the labeled RNA strand and, although it is conceivable that the 3′-OH of the DNA strand could be joined to the 5′-PO4 of the RNA strand, in practice T4 Rnl2 is not proficient in ligating a substrate containing a deoxynucleotide at the 3′-OH position (10Yin S. Ho C.K. Miller E.S. Shuman S. Virology. 2004; 319: 141-151Google Scholar). To evaluate whether Rnl2 might be better able to ligate DNA strands if they were splinted by an RNA template, we prepared a nicked DNA·RNA duplex by annealing a 5′ 32P-labeled 24-mer DNA oligonucleotide to an unlabeled 24-mer RNA oligonucleotide with overlapping complementarity. The resulting DNA·RNA hybrid duplex contains potentially ligatable staggered 3′-OH/5′-PO4 DNA nicks on one strand and nonligatable 3′-OH/5′-OH RNA nicks on the other strand. We were unable to detect any DNA strand joining with this nicked DNA·RNA substrate, even at Rnl2 concentrations that were saturating for sealing of the nicked RNA·DNA hybrid (data not shown). A kinetic analysis of the reaction of 100 nm nicked RNA·DNA hybrid with 1.6 nm Rnl2 is shown in Fig. 3B. The evolution of the n-mer ladder as a function of time was consistent with a distributive mechanism of sealing of a substrate containing multiple ligatable nicks, i.e. a dimer 48-mer product of a single sealing event predominated at early times, prior to the appearance of higher order ligated RNAs. The extent of ligation was quantified by scanning the gel with a Fujix imaging apparatus. The radioactivity signal was determined for each n-mer (e.g. n = 1 for the input substrate strand, n = 2 for linear dimer, etc.) that was visualized. The radioactivity for individual n-mers in each reaction was summed, and each n-mer species was expressed as the fraction of the total. The amount (in fmol) of 5′ 32P-labeled 24-mer comprising each labeled n-mer was determined by multiplying this fraction by the known amount of input 24-mer substrate strands. Each linear n-mer is necessarily generated as a consequence of “n minus 1” ligation events. Thus, the amount (in fmol) of ligation required to form each n-mer product was calculated for each n-mer by the equation: fmol(ligation) = fmol(24-mer) × [(n-1)/n]. The total amount of strand ligation (fmol of 5′ ends joined) for each reaction was then determined by summing the ligation events for each n-mer product in the ladder. A plot of ligation versus reaction time is shown in Fig. 6A. The initial rate of reaction was 11 fmol of ends ligated/fmol of enzyme/min. 67% of the available 5′ 32P-labeled RNA ends were sealed after 15 min. Thus, Rnl2 acts catalytically and efficiently in sealing RNA nicks. Rnl2 sealing of the nicked RNA·DNA hybrid substrate was optimum at pH 6.0–7.0 in 50 mm Tris acetate buffer (Fig. 3C). Activity was virtually nil at pH ≤5.0 but persisted at alkaline pH up to 9.0. Nicked RNA ligation required a divalent cation cofactor, which could be magnesium or manganese, but not calcium, copper, or zinc (Fig. 4A). Nicked RNA ligation in the absence of added ATP was 2% of the activity in the presence of 10 μm ATP; neither GTP, CTP, UTP, nor dATP could substitute for ATP in stimulating the ligation reaction (Fig 4B). Joining of nicked RNA by Rnl2 was remarkably efficient with respect to ATP utilization insofar as the yield of ligated product was saturated at 0.1 μm ATP, which was equivalent to the concentration of 32P-labeled RNA strands included in the reaction (Fig. 4C). Reducing ATP to 0.01 μm elicited a proportional reduction in the extent of ligation (Fig. 4C). The standard ligation reaction mixtures contained 40 mm NaCl contributed by the enzyme and RNA substrate solutions. The effect of increasing ionic strength on RNA ligation efficiency was gauged by supplementing the reactions with NaCl to final concentrations of 65, 90, 140, 240, 340, or 540 mm NaCl. Activity was unaffected up to 65 mm NaCl but was reduced by 36% and 97% at 90 and 240 mm NaCl, respectively. Rnl2 activity was undetectable at ≥340 mm NaCl (data not shown). Requirement for Alignment of the RNA Ends at a Nick versus a Flap—The DNA strand of the RNA·DNA hybrid was altered by removing 1 nucleotide in the center such that, upon annealing to the 24-mer 32P-labeled RNA strand, either the 3′-OH RNA end or the 5′-PO4 end protruded from the bridging template strand as a 1-nucleotide flap (Fig. 5). Rnl2 was much less effective in sealing a 1-nucleotide flap on the 5′-PO4 side than it was in joining at a perfectly aligned nick and was nearly inert in sealing a 1-nuclotide flap on the 3′-OH side (Fig. 5). We then further altered the DNA strand of the RNA·DNA hybrid by removing 2 central nucleotides so that annealing to the 24-mer 32P-labeled RNA strand yielded 1-nt flaps on both the 3′-OH RNA and the 5′-PO4 RNA ends (Fig. 5). Rnl2 displayed minimal activity on the double-flap substrate, comparable with that seen with the 3′ flap alone. Quantitation of the titration data revealed that Rnl2-specific activities on the 5′-flap, 3-flap, and double-flap substrates were 11, 0.3, and 1% of the specific activity on the nicked substrate, respectively (Fig. 6B). We conclude that Rnl2 ligase activity is impeded by protruding nucleotides and is especially sensitive to a flap at the 3′-OH terminus. Requirement for Alignment of the RNA Ends at a Nick versus a Gap—The DNA strand of the RNA·DNA hybrid was altered by adding 1, 2, or 3 deoxyadenosine nucleotides in the center of the oligonucleotide such that, upon annealing to the 24-mer 32P-labeled RNA strand, the 3′-OH RNA end was separated from the 5′-PO4 end by a 1, 2-, or 3-nt gap (Fig. 7B). Rnl2 was much less effective in sealing across the 1-nucleotide gap than it was in joining at a perfectly aligned nick (Fig. 7A, top panel). Activity was minimal at a 2-nt gap and undetectable at a 3-nt gap. Rnl2-specific activities on the 1-, 2-, and 3-nt gapped substrates were 10, 1.6, and <0.1% of the specific activity on the nicked substrate, respectively (not shown). We surmise that Rnl2 requires exact approximation of the reactive termini at a nick stabilized by the template strand. Structure-guided Mutational Analysis of the N-terminal Domain of T4 Rnl2—T4 Rnl2 provides an excellent model for structural and mechanistic analysis of the Rnl2/REL class of RNA-joining enzymes. There is primary structure similarity between Rnl2 and the kinetoplastid RELs through the entire length of the Rnl2 protein (Fig. 1). Many of the conserved positions localize to the N-terminal adenylyltransferase domain (spanning amino acids 1–249 and demarcated by the arrowhead in Fig. 1), which includes nucleotidyltransferase motifs I, III, IIIa, IV, and V that are conserved in DNA ligases and mRNA capping enzymes. Previous mutational analysis of Rnl2 pinpointed 12 individual amino acids that are essential for Rnl2-catalyzed circularization of an 18-nucleotide single-stranded pRNA substrate (indicated by vertical bars in Fig. 1) (1Ho C.K. Shuman S. Proc. Natl. Acad. Sci. U. S. A. 2002; 99: 12709-12714Google Scholar, 17Yin S. Ho C.K. Shuman S. J. Biol. Chem. 2003; 278: 17601-17608Google Scholar); these include at least one conserved residue in each of the nucleotidyltransferase motifs. A crystal structure of the Rnl2 adenylyltransferase domain with AMP bound at the active site revealed a shared fold and catalytic mechanism for RNA ligases, DNA ligases, and mRNA capping enzymes and provided plausible atomic explanations for the observed mutational effects on Rnl2 activity (23Ho C.K. Wang L.K. Lima C.D. Shuman S. Structure. 2004; 12: 327-339Google Scholar). For example, essential residues Lys-35, Asn-40, Arg-55, Glu-99, Phe-119, Glu-204, Lys-225, and Lys-227 are implicated directly in the nucleotidyl transfer reaction via direct or indirect atomic contacts to AMP (23Ho C.K. Wang L.K. Lima C.D. Shuman S. Structure. 2004; 12: 327-339Google Scholar). Other essential residues appeared to play noncatalytic structural roles. Here we conducted an alanine scan of 13 positions located in the N-terminal domain of Rnl2 that were chosen on the basis of their conservation in other Rnl2/REL family members, their hydrophilic character, and/or their positions in the Rnl2 fold suggestive of either internal structural roles or contributions to a putative surface RNA-binding site (see below). The mutated residues are indicated by question marks over the alignment in Fig. 1. The Y5A, E29A, R33A, K54A, E63A, Q106A, K107A, D135A, Y136A, E139A, R155A, S170A, and R221A proteins were produced in E. coli as His10-tagged fusions and purified from soluble bacterial extracts by Ni-agarose chromatography. The 42-kDa Rnl2 polypeptide was the predominant species detected by SDS-PAGE, and the extents of purification were comparable for mutant and wild-type Rnl2 (Fig. 8). The adenylyltransferase activity of the Rnl2·Ala proteins was assayed by label transfer from [α-32P]ATP to the Rnl2 polypeptide to form the covalent Rnl2·AMP intermediate (Fig. 8). The E29A, K54A, E63A, Q106A, K107A, Y136A, E139A, and R221A mutants displayed near wild-type adenylyltransferase activity, i.e. the yield of mutant Rnl2·AMP was within a factor of two of the extent of adenylylation of wild-type Rnl2 produced under the same conditions. Five mutants displaying significant defects in adenylylation were: R33A (3% of the wild-type control), D135A (4%), R155A (16%), and S170A (8%). The Y5A mutation had a modest effect on adenylylation (39% of the wild-type control). Each of the mutants was assayed for RNA ligase activity with the nicked RNA·DNA hybrid substrate. The E29A, E63A, K107A, Y136A, E139A, and R221A mutants displayed wild-type or near wild-type strand-sealing activity, whereas mutants Y5A, R33A, K54A, Q106A, D135A, R155A, and S170A were either unreactive or severely impaired in RNA sealing (Fig. 9). The relative activities of the Rnl2·Ala mutants are quantified in Fig. 10A. The retention of activity of the E29A, E63A, K107A, Y136A, E139A, and R221A mutants in RNA sealing correlated with their retention of adenylyltransferase activity. The defects of the R33A, D135A, R155A, and S170A mutants in overall RNA ligation were in keeping with their defects in forming the ligase-adenylate intermediate, whereas the Y5A mutation exerted a relatively greater effect on RNA sealing than on ligase adenylylation. The most instructive findings pertained to the K54A and Q106A mutants, which were active in ligase adenylylation but defective in overall RNA ligation. These results suggest that Lys-54 and Gln-106 are specifically required for one or more steps of the ligation pathway subsequent to Rnl2·AMP formation.Fig. 10Mutational effects on RNA ligation. A, extents of
Referência(s)