Title: The ATPase, RNA Unwinding, and RNA Binding Activities of Recombinant p68 RNA Helicase
Abstract: p68 RNA helicase, a nuclear RNA helicase, was identified 2 decades ago. The protein plays very important roles in cell development and organ maturation. However, the biological functions and enzymology of p68 RNA helicase are not well characterized. We report the expression and purification of recombinant p68 RNA helicase in a bacterial system. The recombinant p68 is an ATP-dependent RNA helicase. ATPase assays demonstrated that double-stranded RNA (dsRNA) is much more effective than single-stranded RNA in stimulating ATP hydrolysis by the recombinant protein. Consistently, RNA-binding assays showed that p68 RNA helicase binds single-stranded RNA weakly in an ATP-dependent manner. On the other hand, the recombinant protein has very high affinity for dsRNA. Binding of the protein to dsRNA is ATP-independent. The data indicate that p68 may directly target dsRNA as its natural substrate. Interestingly, the recombinant p68 RNA helicase unwinds dsRNA in both 3′ → 5′ and 5′ → 3′ directions. This is the second example of a Asp-Glu-Ala-Asp (DEAD) box RNA helicase that unwinds RNA duplexes in a bi-directional manner. p68 RNA helicase, a nuclear RNA helicase, was identified 2 decades ago. The protein plays very important roles in cell development and organ maturation. However, the biological functions and enzymology of p68 RNA helicase are not well characterized. We report the expression and purification of recombinant p68 RNA helicase in a bacterial system. The recombinant p68 is an ATP-dependent RNA helicase. ATPase assays demonstrated that double-stranded RNA (dsRNA) is much more effective than single-stranded RNA in stimulating ATP hydrolysis by the recombinant protein. Consistently, RNA-binding assays showed that p68 RNA helicase binds single-stranded RNA weakly in an ATP-dependent manner. On the other hand, the recombinant protein has very high affinity for dsRNA. Binding of the protein to dsRNA is ATP-independent. The data indicate that p68 may directly target dsRNA as its natural substrate. Interestingly, the recombinant p68 RNA helicase unwinds dsRNA in both 3′ → 5′ and 5′ → 3′ directions. This is the second example of a Asp-Glu-Ala-Asp (DEAD) box RNA helicase that unwinds RNA duplexes in a bi-directional manner. Asp-Glu-Ala-Asp double-stranded RNA single-stranded RNA dithiothreitol nickel-nitrilotriacetic acid isopropyl-1-thio-β-d-galactopyranoside open reading frame nucleotide methylene blue bovine serum albumin hepatitis c virus The nuclear p68 RNA helicase was first identified by cross-reaction with a monoclonal antibody PAb204 that was originally raised against SV40 large T antigen 2 decades ago (1.Crawford L. Leppard K. Lane D. Harlow E. J. Virol. 1982; 42: 612-620Crossref PubMed Google Scholar, 2.Lane D.P. Hoeffler W.K. Nature. 1980; 288: 167-170Crossref PubMed Scopus (138) Google Scholar). The helicase plays very important roles in cell proliferation and organ maturation (2.Lane D.P. Hoeffler W.K. Nature. 1980; 288: 167-170Crossref PubMed Scopus (138) Google Scholar, 3.Stevenson R.J. Hamilton S.J. MacCallum D.E. Hall P.A. Fuller-Pace F.V. J. Pathol. 1998; 184: 351-359Crossref PubMed Scopus (83) Google Scholar, 4.Mahal B. Nellen W. Biol. Chem. Hoppe Seyler. 1994; 375: 759-763Crossref PubMed Scopus (7) Google Scholar). p68 is highly conserved throughout evolution. The human protein shows 98% sequence identity with mouse p68 (5.Kitajima Y. Yatsuki H. Zhang R. Matsuhashi S. Hori K.A. Biochem. Biophys. Res. Commun. 1994; 199: 748-754Crossref PubMed Scopus (11) Google Scholar). The Saccharomyces cerevisiae and Schizosaccharomyces pombe “p68”-Dbp2p share 55% identity with the human protein (6.Iggo R.D. Jamieson D.J. MacNeill S.A. Southgate J. McPheat J. Lane D.P. Mol. Cell. Biol. 1991; 11: 1326-1333Crossref PubMed Scopus (124) Google Scholar). The growth defect in a DBP2 yeast mutant strain can be rescued by human p68 (7.Barta I. Iggo R. EMBO J. 1995; 14: 3800-3808Crossref PubMed Scopus (65) Google Scholar), suggesting that the biological functions of the p68 RNA helicase are conserved among different species. The p68 RNA helicase belongs to a large family of highly evolutionarily conserved proteins, the so-called Asp-Glu-Ala-Asp (DEAD)1 box family of putative ATPases and helicases (for reviews see Refs. 8.Luking A. Stahl U. Schmidt U. Crit. Rev. Biochem. Mol. Biol. 1998; 33: 259-296Crossref PubMed Scopus (220) Google Scholar and 9.Schmid S.R. Linder P. Mol. Microbiol. 1992; 6: 283-291Crossref PubMed Scopus (449) Google Scholar). The RNA helicases are found in almost all organisms, from bacteria to human. This family of proteins is involved in almost every process of RNA metabolism in cells such as pre-mRNA splicing, translation, RNA degradation, and ribosome biogenesis (10.de la Cruz J. Kressler D. Linder P. Trends Biochem. Sci. 1999; 24: 192-198Abstract Full Text Full Text PDF PubMed Scopus (433) Google Scholar). RNA helicases are enzymes that catalyze the separation of strands of duplex RNA by utilization of the energy derived from hydrolysis of NTP (generally ATP). Unlike the previously defined DNA helicase, which processively unwinds long stretches of dsDNA, most RNA helicases probably modulate only a short duplex region in a long RNA molecule. In addition, recent studies also suggest that members of DEAD/DE xH superfamily of RNA helicase may also be involved in dissociation of protein-RNA interactions (11.Jankowsky E. Gross C.H. Shuman S. Pyle A.M. Science. 2001; 291: 121-125Crossref PubMed Scopus (253) Google Scholar, 12.Chen J.Y. Stands L. Staley J.P. Jackups Jr., R.R. Latus L.J. Chang T.H. Mol. Cell. 2001; 7: 227-232Abstract Full Text Full Text PDF PubMed Scopus (162) Google Scholar) and/or modulation of other RNA secondary structures (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar). Despite a rapidly growing number of identified putative RNA helicases, only a handful of these putative RNA helicases, including eIF-4A, U5–200K, HCV-NS3, Brr2p, Prp16p, and vaccinia virus NPH-II, have demonstrated RNA unwinding activity in vitro (14.Rozen F. Edery I. Meerovitch K. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1990; 10: 1134-1144Crossref PubMed Scopus (498) Google Scholar, 15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar, 16.Raghunathan P.L. Guthrie C. Curr. Biol. 1998; 8: 847-855Abstract Full Text Full Text PDF PubMed Scopus (244) Google Scholar, 17.Wang Y. Wagner J.D. Guthrie C. Curr. Biol. 1998; 8: 441-451Abstract Full Text Full Text PDF PubMed Scopus (83) Google Scholar, 18.Warrener P. Collett M.S. J. Virol. 1995; 69: 1720-1726Crossref PubMed Google Scholar, 19.Shuman S. Proc. Natl. Acad. Sci. U. S. A. 1992; 89: 10935-10939Crossref PubMed Scopus (101) Google Scholar). The ATPase and the RNA unwinding activities of p68 RNA helicase were documented (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar, 20.Iggo R.D. Lane D.P. EMBO J. 1989; 8: 1827-1831Crossref PubMed Scopus (124) Google Scholar, 21.Ford M.J. Anton I.A. Lane D.P. Nature. 1988; 332: 736-738Crossref PubMed Scopus (160) Google Scholar) with the protein that was purified from human 239 cells. ATPase activity of the purified p68 RNA helicase is stimulated by RNA polynucleotides and to a lesser extent by DNAs. The RNA unwinding assay showed that the p68 RNA helicase was able to unwind a partial dsRNA with a long 162-bp duplex (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar). This indicated that p68 RNA helicase unwinds RNA with high processivity. In addition to unwinding RNA, p68 RNA helicase was also suggested to catalyze RNA annealing (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar). Due to the difficulty in expression and purification of recombinant p68 RNA helicase in an appropriate expression system, the biological functions and enzymatic activities of p68 RNA helicase are not well characterized (10.de la Cruz J. Kressler D. Linder P. Trends Biochem. Sci. 1999; 24: 192-198Abstract Full Text Full Text PDF PubMed Scopus (433) Google Scholar, 22.Heinlein U.A. J. Pathol. 1998; 184: 345-347Crossref PubMed Google Scholar). By using a unique methylene blue-mediated cross-linking technique (MB cross-linking) (23.Liu Z.R. Wilkie A.M. Clemens M.J. Smith C.W. RNA (New York). 1996; 2: 611-621PubMed Google Scholar), p68 RNA helicase was detected interacting with the U1:5′-ss duplex during the splicing process (24.Liu Z.R. Sargueil B. Smith C.W. Mol. Cell. Biol. 1998; 18: 6910-6920Crossref PubMed Scopus (22) Google Scholar). The protein is essential for pre-mRNA splicing. 2Z.-R. Liu, unpublished results. In this article, we report the expression and purification of recombinant His tag p68 RNA helicase from Escherichia coli. We analyzed the ATPase and RNA unwinding activities of the recombinant His tag p68. Our data show that the ATPase activity of the protein was polynucleotide-dependent. dsRNA is much more efficient than ssRNA in stimulating the ATPase activity of the protein. We demonstrate here that p68 RNA helicase unwound RNA in both 3′ → 5′ and 5′ → 3′ directions. In addition, our data demonstrate that p68 RNA helicase bound strongly to dsRNA, and the binding was ATP-independent. RNAs were synthesized by run-off transcriptions of the linearized transcription vectors that carry appropriate DNA inserts in the transcription region using T7 or SP6 RNA polymerase. The RNAs were uniformed labeled with [α-32P]UTP. The DNA vectors for transcribing each RNA substrate are listed in Table I. The partial dsRNAs were prepared by annealing each pair of complementary transcripts at a 3-fold excess of unlabeled strand over labeled strand. The dsRNA substrate for both ATPase assays and RNA binding assays is the hybridization of equal molar amounts of two complementary strands. Annealing solution contained 30 mm Tris-HCl, pH 7.5, 100 mm NaCl, and 80% formamide. The RNA annealing mixture was heated to 85 °C for 10 min and was then slowly cooled down to room temperature. The RNA hybrids were used without further treatments.Table ITranscripts used in ATPase, RNA unwinding, and RNA-binding assaysRNALengthRNA polymeraseVectorModificationsDigestionnt1167SP6pGEM-3ZDelete the sequence between EcoRI/HindIIIPvuII2222T7pGEM-3ZDelete the sequence between EcoRI/HindIIIPvuII3214SP6pGEM-3ZNOPvuII4269T7pGEM-3ZNOPvuII528SP6pSP73NOEcoRI693T7pSP73NOBglII758SP6pGEM-3ZDelete the sequence GAGTATT between the SP6 starting site and HindIII siteEcoRI861T7pGEM-3ZNOHindIII9297SP6pGEM-3ZNONdeI Open table in a new tab According to the GenBankTM sequence (GenBankTM accession number AF015812), a pair of primers, 5′GCGGATCCTCGAGTGACCGAGACCGC3′ and 5′ATTGGGAATATCCTGTTG3′, was used. The open reading frame (ORF) that encodes p68 RNA helicase was amplified from a cDNA library (Stratagene). The PCR products were cloned into pBluescript SK(+) vector. The obtained DNA clones were examined by auto-DNA sequencing, and the sequences of resultant DNA completely match the DNA sequences of p68 RNA helicase retrieved from GenBankTM. The ORF of p68 RNA helicase was subcloned into an expression vector pET-30a by BamHI/HindIII sites. The expression clones were transformed to E. coli. Bacteria were grown in LB medium to A600 nm = 0.6. The expression of recombinant protein was induced by IPTG for 5 h. To purify the recombinant protein, the harvested bacterial cells were subjected to one freeze/thaw cycle at −80 °C. After digestion with lysozyme in lysis buffer (50 mm Tris-HCl, pH 7.5, 100 mm NaCl, 1 mm DTT, 0.5 mmEDTA) at 30 °C for 40 min, the bacterial cells were further disrupted by ultrasonication. After centrifugation, the soluble bacterial lysates were passed through a Ni-NTA column, and the recombinant proteins were eluted with 200 mm imidazole in protein buffer (50 mm Tris-HCl, pH 7.5, 200 mmNaCl, 0.5 mm DTT, 10% glycerol). The protein was further dialyzed against the same buffer with only 100 mmimidazole. ATPase activities were determined by measuring the released inorganic phosphate during ATP hydrolysis using a direct colorimetric assay (25.Chan K.-M. Delfert D. Junger K.D. Anal. Biochem. 1986; 157: 375-380Crossref PubMed Scopus (827) Google Scholar, 26.Pugh G.E. Nicol S.M. Fuller-Pace F.V. J. Mol. Biol. 1999; 292: 771-778Crossref PubMed Scopus (42) Google Scholar). The method is based on the change in absorbance (A623 nm) of malachite green-molybdenum complex in the presence and absence of inorganic phosphate. A typical ATPase assay was carried out in 50-μl reaction volumes, containing 20 mm Tris-HCl, pH 7.5, 200 mm NaCl, 1 mm MgCl2, 5 mm DTT, ∼1–2 μg of appropriate RNA, 40 units of RNasin, 4 mm NTP, and 10 μl of helicase. The ATPase reactions were incubated at 37 °C for 30 min. After incubation, 1 ml of malachite green-molybdenum reagent was added to the reaction mixture, and reactions were further incubated at room temperature for exactly 5 min. The absorption (A) at 630 nm was then measured. The concentrations of inorganic phosphate were determined by matching the A630 nm in a standard curve of A630 nmversus known phosphate concentrations. RNA unwinding activities were determined by the method similar to that described by Rozen and co-workers (14.Rozen F. Edery I. Meerovitch K. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1990; 10: 1134-1144Crossref PubMed Scopus (498) Google Scholar). Briefly, the RNA unwinding reactions were carried out in a 20-μl reaction volume containing 70 mm Tris-HCl, pH 7.5, 200 mm NaCl, 1 mm MgCl2, 5 mm DTT, 2.5 fmol of partial dsRNA, 40 units of RNasin, 16 mm ATP, and 2–4 μl of helicase. Reactions were incubated at 37 °C for 60 min. The reaction mixtures were directly loaded onto the appropriate percentage of SDS-PAGE, and the gel was subjected to autoradiography. RNA bindings were analyzed by gel-mobility shift assays and MB-mediated dsRNA-protein cross-linking (23.Liu Z.R. Wilkie A.M. Clemens M.J. Smith C.W. RNA (New York). 1996; 2: 611-621PubMed Google Scholar, 27.Liu Z.R. Laggerbauer B. Luhrmann R. Smith C.W. RNA (New York). 1997; 3: 1207-1219PubMed Google Scholar). In a typical gel mobility shift assay, 100 ng of recombinant proteins were mixed with 5 fmol of appropriate RNA in buffer solution containing 30 mmTris-HCl, pH 7.5, 100 mm NaCl, 2 mmMgCl2, 1 mm DTT, and 20 units of RNasin with or without ATP as indicated. After 15 min of incubation at room temperature, the reaction mixtures were loaded on to 6% native-PAGE (acrylamide:bis = 40:1). The methylene blue mediated cross-linkings were carried out as described previously (23.Liu Z.R. Wilkie A.M. Clemens M.J. Smith C.W. RNA (New York). 1996; 2: 611-621PubMed Google Scholar, 24.Liu Z.R. Sargueil B. Smith C.W. Mol. Cell. Biol. 1998; 18: 6910-6920Crossref PubMed Scopus (22) Google Scholar). The same protein:RNA reaction mixtures used in the gel mobility shift assays were used in the cross-linking experiments. After RNase mixture (RNase A, T1, and V1) digestion, the cross-linking mixture was separated by the appropriate percentage of SDS-PAGE and subjected to autoradiography. The Western blot analyses were performed with various sources of bacterial lysates, the purified recombinant p68, and the recombinant HCV-NS3 using the commercial ECL Western blot and detection kits. The antibody pAb204 was used in the blot as a 5:1 dilution and pAbN1 was used as a 3000:1 dilution. Previous data demonstrated that p68 RNA helicase is an essential splicing factor acting at the U1 small nuclear RNA and 5′-splice site duplex (24.Liu Z.R. Sargueil B. Smith C.W. Mol. Cell. Biol. 1998; 18: 6910-6920Crossref PubMed Scopus (22) Google Scholar).2 According to the cDNA sequence of human p68 RNA helicase from GenBankTM(accession number AF015812), the open reading frame (ORF) of p68 RNA helicase was cloned, and the recombinant His tag protein was overexpressed in E. coli (BL21) and purified over a Ni-NTA column (Fig. 1A, lane 5). The concentration of the recombinant protein was estimated to be about 250 ng/μl (estimated by Bradford method, Bio-Rad). We noted that to keep the recombinant protein in a soluble form, it is necessary to maintain the protein in a buffer solution containing at least 100 mmimidazole (data not shown). To verify whether the obtained recombinant protein is p68 RNA helicase, we performed Western blot analyses using a monoclonal antibody pAb204 and a polyclonal antibody pAbN1 (raised against a peptide that spans 15 amino acids residues at the N-terminal region of human p68). It is evident that the purified recombinant protein was recognized by both antibodies in the Western blot analyses (Fig. 1, B, lane 1, and C,lane 1). No bacterial proteins were recognized by these two antibodies (Fig. 1, B, lane 3, and C, lane 3). The data suggested that a soluble recombinant His tag p68 RNA helicase was expressed and purified in a bacterial expression system. To determine enzymatic activities of the bacterially expressed recombinant p68 RNA helicase, we first examined the ATPase activity of the His-p68 by measuring the released inorganic phosphate during ATP hydrolysis using a direct colorimetric assay (25.Chan K.-M. Delfert D. Junger K.D. Anal. Biochem. 1986; 157: 375-380Crossref PubMed Scopus (827) Google Scholar, 26.Pugh G.E. Nicol S.M. Fuller-Pace F.V. J. Mol. Biol. 1999; 292: 771-778Crossref PubMed Scopus (42) Google Scholar). The ATPase assays demonstrated that the recombinant protein hydrolyzed ATP in a polynucleotide-dependent manner (Fig. 2A). As a control, no ATP hydrolysis was observed in the same assay using BSA (Fig. 2A). We also tested the acceptance of other nucleotide triphosphates by the recombinant p68. Our experiments showed that the protein only hydrolyzed ATP and dATP but not CTP, UTP, and GTP (data not shown). To examine the nucleic acid-dependent ATPase activity of the bacterially expressed recombinant p68 RNA helicase, we carried out ATPase assays in the presence of ssRNA (RNA 9, Table I), ssDNA (DNA oligonucleotides, 24 nt), dsRNA (RNA 7 versus RNA 8, Table I), or dsDNA (pGEM-3Z, linearized with EcoRI). Consistent with previous analyses (20.Iggo R.D. Lane D.P. EMBO J. 1989; 8: 1827-1831Crossref PubMed Scopus (124) Google Scholar) carried out with p68 purified from human 239 cells, the ATPase activity of the bacterially expressed p68 RNA helicase was activated by both the ssRNA and the dsDNA and to much less extent by the ssDNA (Fig. 2B). Interestingly, the ATPase activity of the recombinant protein was greatly enhanced by the dsRNA. The ATPase activity of the protein was more than doubled in the presence of the dsRNA compared with that in the presence of the same molar amounts of the ssRNA (Fig. 2B). Because doubling the amounts of each individual ssRNA that formed the duplex RNA did not have a similar effect on the ATPase activity of p68 (data not shown), it was less likely that the enhancement of the ATPase activity by the dsRNA was the result of the residues of ssRNAs that did not form duplex in our annealing conditions. The dsRNA used in our test contains a 67-bp duplex, a 4-nt 5′-overhang on one side, and an 1-nt 5′-overhang on the other side (Table I). It is unlikely that the very short ssRNA tails on both sides of the RNA duplex are responsible for the ATPase activity enhancement. Thus our conclusion is that the ATPase activity of the recombinant p68 RNA helicase is strongly stimulated by dsRNA. The preceding data showed that the ATPase activity of the recombinant p68 was strongly enhanced by dsRNA. We reasoned that the recombinant p68 RNA helicase may bind dsRNA, and the affinity of the protein for dsRNA may be higher than that for ssRNA. To measure the RNA-binding properties of the recombinant p68 RNA helicase, we employed the gel mobility shift assay. The same ssRNA and dsRNA substrates used in the ATPase assays were utilized here to examine the ssRNA and dsRNA binding activities of the recombinant p68 RNA helicase. A weak mobility-shifted band was observed with the recombinant p68 RNA helicase and the ssRNA (Fig. 3A, lane 2), and the shifted band was ATP-dependent (Fig. 3A, compare lane 2 to lane 4). A similar ssRNA binding property was also observed with another RNA helicase derived from NS3 of HCV (Fig. 3A, lane 3). However, a very strong mobility-shifted band was observed with the recombinant p68 RNA helicase and the dsRNA substrate, and the formation of this slow mobility RNA-protein complex band is ATP-independent (Fig. 3A, lanes 8 and 9). In contrast, no mobility shift band was observed with the HCV-NS3 RNA helicase and the same dsRNA substrate (Fig. 3A, lane 6). Quantitative analyses of both the free RNA bands and RNA-protein complex bands indicate that less than 10% of the ssRNA was associated with the ssRNA-p68 complex. On the other hand, more than 90% of the labeled dsRNA was associated with the shifted dsRNA-protein complex band. The data suggested that the recombinant p68 has higher affinity for dsRNA compared with that for ssRNA. To verify further the dsRNA binding by the recombinant p68 RNA helicase, we utilized the MB cross-linking technique, an RNA-protein photocross-linking method that has high preference for mediating cross-links between dsRNA and dsRNA-binding proteins (23.Liu Z.R. Wilkie A.M. Clemens M.J. Smith C.W. RNA (New York). 1996; 2: 611-621PubMed Google Scholar). The same ssRNA and dsRNA substrates used in gel mobility shift assays were used in the MB cross-linking tests. It is evident that the recombinant p68 RNA helicase was cross-linked to the dsRNA target in an ATP-independent manner (Fig. 3B, lanes 1 and 2). The cross-linking signal was not observed with the ssRNA:p68 or with HCV-NS3:dsRNA (Fig. 3B, lanes 3 and 4). The results from the cross-linking experiments were completely consistent with the data obtained from gel mobility shift assays. Our experimental data strongly suggested that the recombinant p68 RNA helicase has high affinity for dsRNA compared with that for ssRNA. This dsRNA binding property is not a common characteristic to all members of DEAD box family of RNA helicases. To examine the RNA unwinding activities of the recombinant His tag p68, we carried out an RNA unwinding assay using a procedure similar to that described by Rozen and co-workers (14.Rozen F. Edery I. Meerovitch K. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1990; 10: 1134-1144Crossref PubMed Scopus (498) Google Scholar). The RNA substrate for the RNA unwinding assay was a partial dsRNA containing a short RNA duplex (∼22 bp in length) and long 186- and 88-nt 3′-overhangs on both sides (RNA 1 versusRNA 2, Table I and Fig. 4). Consistent with previous studies (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar, 15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar), the partial dsRNA-containing 3′-overhang was unwound in the presence of 25 ng/μl of the His tag p68 (Fig. 4A, lane 4). In the control reactions, the unwinding was not observed in the absence of ATP and in the absence of the recombinant protein (lanes 3 and 5). The RNA duplex was also not unwound by a nonspecific protein, BSA (lane 6). The p68 RNA helicase purified from human 239 cells was able to unwind a partial dsRNA with a long 162-bp RNA duplex (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar). This feature is unique to very few members of DEAD/DE xH RNA helicases. It is generally believed that, unlike DNA helicases, the majority of so-called RNA helicases work only on a short stretch of RNA duplex or local RNA structures (8.Luking A. Stahl U. Schmidt U. Crit. Rev. Biochem. Mol. Biol. 1998; 33: 259-296Crossref PubMed Scopus (220) Google Scholar). NPH-II from vaccinia virus is another example of unwinding long RNA duplex (28.Jankowsky E. Gross C.H. Shuman S. Pyle A.M. Nature. 2000; 403: 447-451Crossref PubMed Scopus (197) Google Scholar). To test whether the bacterially expressed p68 RNA helicase is able to unwind the long RNA duplex, we employed another partial dsRNA substrate in the unwinding assays. The partial dsRNA contains a 69-bp duplex and a 186-nt 3′-overhang on one side and a 88-nt 3′-overhang on the other side (RNA 3 versus RNA 4, Table I and Fig. 4). In contrast to the observations made by Hirling and co-workers (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar), our unwinding experiments showed that very small fractions of the partial dsRNA substrate were unwound by the recombinant p68 (Fig. 4B, lane 8). Our observations are, however, consistent with observations made by Rossler and co-workers (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar). The above unwinding data indicated that the recombinant p68 is a weak processive RNA helicase. A member of the DEAD box family, translational initiation factor eIF-4A has been shown to unwind RNA duplex in both 3′ → 5′ and 5′ → 3′ directions. It was suggested that this unwinding property of the protein helps the melting of the mRNA structure during the translational initiation process (14.Rozen F. Edery I. Meerovitch K. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1990; 10: 1134-1144Crossref PubMed Scopus (498) Google Scholar, 29.Jaramillo M. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1991; 11: 5992-5997Crossref PubMed Scopus (86) Google Scholar, 30.Pause A. Sonenberg N. Curr. Opin. Struct. Biol. 1993; 3: 953-959Crossref Scopus (53) Google Scholar). To investigate the polarity of RNA unwinding by the p68 RNA helicase, we synthesized a partial dsRNA containing a 22-bp duplex and a 71-nt 5′-overhang on the one side and a 6-nt 5′-overhang on the other side (RNA 5versus RNA 6, Table I and Fig. 4). The RNA unwinding assay was carried out with this RNA substrate. In contrast to the observations made by Rossler and co-workers (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar), we have repeatedly observed that this partial dsRNA with the 5′-overhang was completely unwound by the bacterially expressed p68 RNA helicase (Fig. 4C, lane 4). Thus our RNA unwinding assays have demonstrated that the bacterially expressed recombinant p68 RNA helicase unwound RNA duplex in both 3′ → 5′ and 5′ → 3′ directions. We presented in this report the expression and purification of recombinant p68 RNA helicase from an E. coli expression system, and we demonstrated that the protein is an ATP-dependent RNA helicase. We have shown that the ATPase activity of the recombinant protein is greatly enhanced by dsRNA. The RNA-binding assays demonstrated that p68 RNA helicase binds ssRNA weakly in an ATP-dependent manner. On the other hand, the recombinant protein has very strong affinity for dsRNA. The binding of the protein to dsRNA is ATP-independent. Interestingly, the bacterially expressed recombinant p68 helicase unwinds dsRNA in both 5′ → 3′ and 3′ → 5′ directions. This is the second example of the DEAD box family of RNA helicases that unwinds RNA in both directions. It has been observed previously (31.Linder P. Daugeron M.C. Nat. Struct. Biol. 2000; 7: 97-99Crossref PubMed Scopus (44) Google Scholar) that the ATPase activity of so-called DEAD/DE xH box of ATPases is polynucleotide-dependent. More detailed characterization carried out with eIF-4A suggested that ATP binding and hydrolysis are tightly coupled to polynucleotide binding by the protein (32.Lorsch J.R. Herschlag D. Biochemistry. 1998; 37: 2194-2206Crossref PubMed Scopus (135) Google Scholar, 33.Lorsch J.R. Herschlag D. Biochemistry. 1998; 37: 2180-2193Crossref PubMed Scopus (173) Google Scholar). Thus, it is believed that binding and hydrolysis of ATP, in return, has great effects on the polynucleotides binding by a DEAD/DE xH box protein. Consistent with the previous observations (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar, 20.Iggo R.D. Lane D.P. EMBO J. 1989; 8: 1827-1831Crossref PubMed Scopus (124) Google Scholar), the ATPase activity of the recombinant p68 is strictly polynucleotide-dependent. We also observed that the ssRNA binding affinities of the recombinant p68 RNA helicase in the presence and absence of ATP are different. In the absence of ATP, the ssRNA binding by the recombinant protein is not detectable by our gel mobility shift assay. Thus, the data from the experiments support the notion that ATP binding and hydrolysis change the RNA-binding properties of an RNA helicase. The dsRNA-binding property of p68 is unique in the DEAD/DE xH box family. It was assumed that RNA helicases, especially those that act as monomer, should recognize dsRNA substrate (9.Schmid S.R. Linder P. Mol. Microbiol. 1992; 6: 283-291Crossref PubMed Scopus (449) Google Scholar, 34.Gibson T.J. Thompson J.D. Nucleic Acids Res. 1994; 22: 2552-2556Crossref PubMed Scopus (89) Google Scholar). Although sequence alignments revealed that a number of other RNA helicases, including human helicase A, Drosophilamaleless, and Caenorhabditis elegans ORF T20G5, contained sequence motifs that closely resemble the dsRNA-binding motif (34.Gibson T.J. Thompson J.D. Nucleic Acids Res. 1994; 22: 2552-2556Crossref PubMed Scopus (89) Google Scholar), none of the RNA helicases have demonstrated dsRNA binding. What is the functional role of this dsRNA-binding property of p68? Because the dsRNA substrate is much more effective than an ssRNA substrate in stimulating the ATPase activity, it is likely that the dsRNA binding is functional relevant to the enzymatic activity of the protein. The best explanation for the functional role of the dsRNA binding is that, unlike many other DEAD/DE xH box proteins that load on the RNA substrate at the single-stranded 5′- or 3′-overhangs (35.Tanner N.K. Linder P. Mol. Cell. 2001; 8: 251-262Abstract Full Text Full Text PDF PubMed Scopus (619) Google Scholar), p68 RNA helicase may directly load on a dsRNA substrate. To test this possibility, we reasoned that the 3′ or 5′ single-stranded overhangs of RNA substrate may not be required for RNA unwinding. Thus, we carried out unwinding experiments using a dsRNA substrate with only a 1-nt 3′-overhang on the one side and a 2-nt 3′-overhang on the other side. However, due to quick annealing of the separated ssRNA, we were unable to reach unambiguous conclusions (data not shown). Because p68 RNA helicase was shown to potentially have RNA annealing activities (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar), an alternative possibility is that the dsRNA-stimulated ATPase activity is required for the RNA annealing activity of the protein. Our data showed the strong dsRNA-binding properties by p68 in the presence and absence of ATP. The effects of ATP binding and hydrolysis on the dsRNA binding are, however, not determined under our current experiments. It will be interesting to perform further experiments to determine what are the effects of the ATP binding and hydrolysis on the dsRNA binding. It is conceivable that ATP binding and hydrolysis would have opposite effects on the dsRNA binding in the above two alternative possibilities. The modes and mechanisms of RNA binding by the DEAD/DE xH box RNA helicases are not well understood. The three-dimensional structure of the HCV-NS3:oligonucleotide complex revealed that the nucleic acid is bound in the second cleft between the domain 1–2 and domain 3 (36.Kim J.L. Morgenstern K.A. Griffith J.P. Dwyer M.D. Thomson J.A. Murcko M.A. Lin C. Caron P.R. Structure. 1998; 6: 89-100Abstract Full Text Full Text PDF PubMed Scopus (581) Google Scholar,37.Yao N. Hesson T. Cable M. Hong Z. Kwong A.D. Le H.V. Weber P.C. Nat. Struct. Biol. 1997; 4: 463-467Crossref PubMed Scopus (421) Google Scholar). However, this nucleic acid binding cleft of HCV-NS3 is not clearly defined in the assembled three-dimensional structure of translation initiation factor, eIF-4A (38.Johnson E.R. McKay D.B. RNA (New York). 1999; 5: 1526-1534Crossref PubMed Scopus (73) Google Scholar, 39.Benz J. Trachsel H. Baumann U. Struct. Fold. Des. 1999; 7: 671-679Abstract Full Text Full Text PDF Scopus (110) Google Scholar, 40.Caruthers J.M. Johnson E.R. McKay D.B. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 13080-13085Crossref PubMed Scopus (244) Google Scholar). Mutational analyses and RNA binding studies with a number of DEAD/DE xH boxes of RNA helicases revealed weak RNA binding by the common sequence motif VI in the DEAD/DE xH box. The weak RNA-binding property of this motif is believed to play an important catalytic role in RNA unwinding (41.Pause A. Methot N. Sonenberg N. Mol. Cell. Biol. 1993; 13: 6789-6798Crossref PubMed Scopus (259) Google Scholar). In addition to the RNA binding by motif VI, many DEAD/DE xH boxes of RNA helicases also bind RNA with additional RNA-binding motifs (13.Rossler O.G. Straka A. Stahl H. Nucleic Acids Res. 2001; 29: 2088-2096Crossref PubMed Scopus (80) Google Scholar, 42.Lorkovic Z.J. Herrmann R.G. Oelmuller R. Mol. Cell. Biol. 1997; 17: 2257-2265Crossref PubMed Scopus (36) Google Scholar, 43.Fuller-Pace F.V. Nicol S.M. Reid A.D. Lane D.P. EMBO J. 1993; 12: 3619-3626Crossref PubMed Scopus (138) Google Scholar). Nevertheless, the relationship between the enzymatic activities and the RNA-binding by these additional RNA binding motifs has not yet been demonstrated. An interesting question regarding the strong dsRNA binding by the recombinant p68 is whether the dsRNA substrate is bound by the sequence motif VI in the DEAD box or by any other sequence motifs outside of the DEAD box. If the dsRNA substrate is bound outside of the DEAD box, how are the enhancement effects of dsRNA binding on the ATPase activity explained? It was noted previously (15.Hirling H. Scheffner M. Restle T. Stahl H. Nature. 1989; 339: 562-564Crossref PubMed Scopus (240) Google Scholar) that p68 RNA helicase contained sequence motifs at its C terminus that resemble the RGG repeats. Is it possible that this sequence motif contributes to dsRNA binding? The amino acid sequence of p68 RNA helicase does not contain any recognizable double-stranded RNA-binding motif. Thus, the dsRNA binding by p68 RNA helicase may represent a new dsRNA recognition mechanism. The bacterially expressed recombinant p68 RNA helicase unwinds dsRNA in both 3′ → 5′ and 5′ → 3′ directions. To date, the bi-directional RNA unwinding activities are peculiar to two DEAD box helicases. The DEAD box of RNA helicase eIF-4A in complex with eIF-4B unwound dsRNA in a bi-directional manner (14.Rozen F. Edery I. Meerovitch K. Dever T.E. Merrick W.C. Sonenberg N. Mol. Cell. Biol. 1990; 10: 1134-1144Crossref PubMed Scopus (498) Google Scholar). The exact functional relevance of this bi-directional RNA unwinding is not clear. It was suggested that this bi-directional RNA unwinding property of eIF-4A played an important role in helping the ribosome land on the 5′-untranslated region of mRNA for translation initiation. p68 RNA helicase was suspected to be involved in a number of biological processes (8.Luking A. Stahl U. Schmidt U. Crit. Rev. Biochem. Mol. Biol. 1998; 33: 259-296Crossref PubMed Scopus (220) Google Scholar), including pre-mRNA splicing,2 RNA degradation (44.He F. Jacobson A. Genes Dev. 1995; 9: 437-454Crossref PubMed Scopus (219) Google Scholar), ribosome biogenesis (45.Nicol S.M. Causevic M. Prescott A.R. Fuller-Pace F.V. Exp. Cell Res. 2000; 257: 272-280Crossref PubMed Scopus (58) Google Scholar), and transcriptional regulation (46.Watanabe M. Yanagisawa J. Kitagawa H. Takeyama K. Ogawa S. Arao Y. Suzawa M. Kobayashi Y. Yano T. Yoshikawa H. Masuhiro Y. Kato S. EMBO J. 2001; 20: 1341-1352Crossref PubMed Scopus (246) Google Scholar,47.Endoh H. Maruyama K. Masuhiro Y. Kobayashi Y. Goto M. Tai H. Yanagisawa J. Metzger D. Hashimoto S. Kato S. Mol. Cell. Biol. 1999; 19: 5363-5372Crossref PubMed Google Scholar). It would be expected that p68 must be able to work on wide spectra of nucleic acids targets. The wide variety of putative RNA substrates may require the protein to be able to unwind RNA duplex in both directions. An alternative possibility is that, as discussed above, p68 may directly target dsRNA for its unwinding substrate. The single-stranded overhangs are really not required. Therefore, the dsRNAs with 3′ or 5′ or without single-stranded overhangs are the equal unwinding substrates for the protein. Unlike eIF-4A, which mainly unwinds dsRNA by complex with eIF-4B, p68 RNA helicase does not need a helper to unwind RNA. Although the functional relevance of the bi-directional RNA unwinding by p68 is not clear, the RNA unwinding by the protein will potentially provide an excellent model system to characterize the mechanism and functional relevance of this bi-directional RNA unwinding. We thank Frances Fuller-Pace for providing hybridoma cells for the antibody PAb204 and Roger Bridgeman for antibody PAb204 production. We are also grateful to Dr. D. L. Peterson for providing the vector for expression of HCV-NS3. We thank Jenny Yang, Liuqing Yang, R. L. Rill, Mariano A. Garcia-Blanco, Christopher W. J. Smith, and Becky Tarleton for detailed critical comments on the manuscript.