INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Members of the DExH family of nucleic acid-dependent nucleoside triphosphatases (NTPases) contain six conserved motifs arrayed in a collinear fashion (7). The number of DExH proteins is increasing rapidly as the signature elements are encountered during genome sequencing. The DExH box proteins are often designated as putative helicases, yet only a subset of these nucleic acid-dependent NTPases are actually able to unwind duplex DNA or RNA in vitro.

The vaccinia virus enzymes nucleoside triphosphate phosphohydrolase I (NPH-I) and NPH-II were the first DExH box NTPases to be purified and characterized (20, 21). NPH-II, a 676-amino-acid monomer, is a nucleoside triphosphate (NTP)-dependent 3'-to-5' RNA helicase (10, 26, 27). NPH-II hydrolyzes any NTP or dNTP, and its NTPase activity is stimulated 10- to 40-fold by either single-stranded DNA or RNA (8, 21). NPH-II exemplifies a subgroup of structurally related DExH proteins that includes human RNA helicase A; the yeast pre-mRNA splicing factors Prp2, Prp16, and Prp22; Drosophila maleless; and hepatitis C virus RNA helicase NS3 (9).

Vaccinia virus NPH-I, a 631-amino-acid monomeric protein, catalyzes the hydrolysis of ATP and dATP only, and its activity is absolutely dependent on a single-strand DNA cofactor (1, 5, 20, 21, 25). NPH-I is a component of the vaccinia virus RNA polymerase transcription elongation complex and serves as a transcription termination factor during the synthesis of viral early mRNAs (4, 5). Release of the transcript from the elongation complex requires ATP hydrolysis by NPH-I (3, 5). Attempts to demonstrate a helicase function for NPH-I have been unsuccessful. NPH-I is the prototype member of a distinct subgroup of DExH proteins that includes Snf2 and its numerous homologues (13). Snf2-like proteins are implicated in transcriptional activation and repression and in chromatin remodeling (24). As with NPH-I, most Snf2-like proteins do not score positively when assayed for helicase activity in vitro.

Our aim was to delineate structure-function relationships for DExH box proteins that would provide insights into the mechanism of nucleic acid-dependent NTP hydrolysis and, where applicable, the coupling of phosphohydrolase activity to duplex unwinding. We previously reported the effects of alanine substitution mutations in NPH-II (8, 9, 11). This analysis revealed that amino acids (underlined) in motifs I (GxGKT), II (DExHE), and VI (QRxGRxGR) are important for NTP hydrolysis. That the NTPase-defective mutants were also defective in RNA unwinding confirms the dependence of strand displacement on NTP hydrolysis. A distinct class of mutations, involving the threonines in motif III (TAT) and the histidine moiety of the DExH box, elicited defects in RNA unwinding but spared the NTPase function (8, 11). Thus, certain moieties are required to couple energy generation to strand displacement.

In the present study, we initiated a mutational analysis of vaccinia virus NPH-I with the intent of establishing structure-function relationships for the Snf2-like ATPase subfamily. An alanine scan of 12 residues in NPH-I motifs I, II, III, and VI identified 9 amino acids that are important for ATP hydrolysis. By introducing conservative substitutions at these positions, we defined the essential functional groups. Comparison of our mutagenesis results for NPH-I with those reported for other DExH or DEAD box proteins underscores the fact that structure-function relationships at conserved residues in the motifs are context dependent; i.e., mutational effects on one nucleic acid-dependent NTPase cannot be extrapolated to other proteins.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

Missense mutants of NPH-I. Plasmid pET-His-NPH-I encodes the 631-amino-acid NPH-I polypeptide fused to an N-terminal leader peptide containing 10 tandem histidines (5). Expression of this protein is under the control of a T7 RNA polymerase promoter. Missense mutations were introduced into the NPH-I gene by PCR, using the two-stage overlap extension method. Residues targeted for amino acid substitutions were Lys-61 and Thr-62 (motif I); Asp-141, Glu-142, His-144, and Asn-145 (motif II); Ser-183 and Thr-185 (motif III); and Gln-472, Gly-475, Arg-476, and Arg-479 (motif VI). The mutated DNA products of the second-stage amplification were either digested with NdeI and BglII and inserted into pET16b or else restricted with NdeI and BamHI and inserted into pET-His-NPH-I in lieu of the corresponding wild-type restriction fragment. The presence of the desired mutations was confirmed by DNA sequencing; the segment corresponding to the inserted restriction fragment was sequenced completely in order to exclude acquisition of unwanted mutations during amplification and cloning.

Carboxyl-terminal truncations of NPH-I. Carboxyl-terminal truncation mutants were constructed by PCR amplification with antisense primers that introduced translation stop codons in lieu of the codons for Tyr-604, Ile-563, Asp-545, Lys-532, Lys-522, and Glu-511 and BglII sites immediately 3' of the new stop codons. The PCR products were digested with NdeI and BglII and then inserted into pET16b. The truncated NPH-I genes and proteins were named according to the number of amino acids deleted from the C terminus, i.e., 28, 69, 87, 100, 110, and 121.

NPH-I expression and purification. The NPH-I expression plasmids were transformed into Escherichia coli BL21(DE3). Transformants were inoculated into Luria-Bertani medium containing 0.1 mg of ampicillin/ml and grown at 37°C until the A₆₀₀ reached approximately 0.6. Cultures (1 liter each) were then placed on ice for 30 min, ethanol was added to a final concentration of 2%, and the cultures were subsequently incubated at 18°C for 24 h with continuous shaking. Cells were harvested by centrifugation, and the pellets were stored at 80°C. All subsequent procedures were performed at 4°C. The cells were thawed and resuspended in 80 ml of lysis buffer (50 mM Tris HCl [pH 7.5], 0.15 M NaCl, 10% sucrose) containing 0.2 mg of lysozyme/ml. After 30 min on ice, Triton X-100 was added to the suspensions to a final concentration of 0.1%. The lysates were sonicated to reduce their viscosity. Insoluble material was removed by centrifugation for 40 min at 18,000 rpm in a Sorvall SS-34 rotor. The supernatants were mixed with 1 ml of Ni-nitrilotriacetic acid-agarose resin (Qiagen) for 1 h. The slurries were poured into columns, which were eluted serially with IMAC buffer (20 mM Tris-HCl [pH 8.0], 0.5 M NaCl, 1 mM phenylmethylsulfonyl fluoride, 10% glycerol) containing 5, 25, 50, and 200 mM imidazole. The polypeptide composition of the column fractions was monitored by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). Fractions enriched for the expressed NPH-I polypeptide (which eluted at 200 mM imidazole) were pooled and dialyzed against 0.5× buffer A (1× buffer A contains 50 mM Tris-HCl [pH 8.0], 2 mM EDTA, 2 mM dithiothreitol (DTT), 0.1% Triton X-100, and 10% glycerol). The dialysates were applied to 1-ml columns of phosphocellulose that had been equilibrated with 0.5× buffer A. The columns were eluted stepwise with 1× buffer A containing 50, 100, and 500 mM NaCl. The NPH-I proteins were recovered in the 500 mM NaCl eluate fractions.

Determination of NPH-I protein concentration. Aliquots (1× and 2×) of each phosphocellulose NPH-I preparation were analyzed by SDS-PAGE in parallel with a twofold dilution series of bovine serum albumin (BSA; from 0.125 to 4 µg per lane). The gels were stained with Coomassie blue dye. The staining intensities of the polypeptides were quantitated with a digital imaging and analysis system (Alpha Innotech Corporation). NPH-I concentrations were calculated by extrapolation to the BSA standard curve.

ATPase assay. ATPase activity was determined by the release of ³²P_i from [-³²P]ATP in the presence of a DNA cofactor. Reaction mixtures (10 µl) containing 40 mM Tris-HCl (pH 8.0), 1 mM MgCl₂, 2 mM DTT, 1 mM [-³²P]ATP, 1 µg of heat-denatured herring sperm DNA, and NPH-I were incubated for 30 min at 37°C. An aliquot (2 µl) of the reaction mixture was applied to a polyethyleneimine-cellulose thin-layer chromatography plate that was developed in 0.5 M LiCl-1 M formic acid. [-³²P]ATP and ³²P_i were quantitated by scanning the thin-layer chromatography plate with a FUJIX model BAS2500 Bio-Imaging Analyzer.

Western blot analysis. Aliquots (20 ng) of purified wild-type NPH-I and C-terminal truncation mutants of NPH-I were electrophoresed through a 10% polyacrylamide gel containing 0.1% SDS, and the polypeptides were transferred electrophoretically to a nitrocellulose membrane. After preincubation with a solution consisting of 10 mM Tris-HCl (pH 8.0), 150 mM NaCl, 0.05% Tween 20, 5% (wt/vol) BSA, and 1% (wt/vol) powdered milk, the membrane was incubated for 2 h at 22°C with rabbit anti-NPH-I serum diluted 1:1,500 in 10 mM Tris-HCl (pH 8.0)-150 mM NaCl. The membrane was washed with a solution containing 10 mM Tris-HCl (pH 8.0), 150 mM NaCl, and 0.4% Tween 20 and then incubated for 1 h with anti-rabbit immunoglobulin conjugated to horseradish peroxidase. The immunoreactive polypeptides were visualized with an enhanced chemiluminescence system (ECL; Amersham) in accordance with the instructions of the vendor. The anti-NPH-I serum was a gift of E. G. Niles, State University of New York at Buffalo.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Mutagenesis strategy. Alanine mutations were introduced in lieu of 12 residues (underlined) within motifs I (58-GVGKT-62), II (141-DECHN-145), III (183-SAT-185), and VI (472-QIVGRAIR-497) of vaccinia virus NPH-I. Wild-type NPH-I and the NPH-I Ala mutants were expressed in bacteria as N-terminal His-tagged fusion proteins and then purified from soluble bacterial lysates by Ni-agarose and phosphocellulose column chromatography (5). Initial assessment of the DNA-dependent ATPase activity of the NPH-I Ala mutant proteins revealed significant activity decrements for nine of the mutants. (We define a 1-order-of-magnitude decrement in ATPase specific activity as the threshold of significance.) Conservative amino acid substitutions were then introduced for these nine residues, and the mutants enzymes were expressed in bacteria and purified. SDS-PAGE analysis of wild-type NPH-I and 30 different NPH-I mutants showed that the phosphocellulose preparations were highly enriched with respect to NPH-I and that the extents of purification were similar (Fig. 1).

View larger version (32K): [in this window] [in a new window]   FIG. 1.   NPH-I purification. Aliquots (1 µg) of the phosphocellulose preparations of wild-type (WT) NPH-I and the indicated NPH-I mutants were electrophoresed through 10% polyacrylamide gels containing 0.1% SDS. Polypeptides were visualized by staining with Coomassie blue dye. The positions and sizes (in kilodaltons) of marker proteins are indicated at the left of each gel.

DNA-dependent ATPase activity. Recombinant wild-type NPH-I catalyzed the release of ³²P_i from [-³²P]ATP in the presence of magnesium and single-stranded DNA. The extent of ATP hydrolysis was linear with respect to the level of input enzyme at limiting concentrations, and the reaction proceeded to completion at saturating enzyme levels (5). The specific activity of wild-type NPH-I in the presence of 1 mM ATP, 1 mM MgCl₂, and 100 µg of denatured DNA/ml was 7.8 nmol of ATP hydrolyzed per ng of protein in 30 min. Assuming that all NPH-I molecules are active, this value translates into a turnover number of ~320 s¹. ATP hydrolysis was undetectable in the absence of either magnesium or single-stranded DNA (18). Recombinant wild-type NPH-I was unable to hydrolyze GTP in place of ATP (18).

View larger version (19K): [in this window] [in a new window]   FIG. 2.   DNA-dependent ATPase activities of NPH-I mutants. The sequences of motifs I, II, III, and VI are shown. (Top) Motif amino acids residues subjected to mutational analysis are underlined and in boldface. (Bottom) ATPase activity was assayed as described in Materials and Methods. Aliquots (2 µl) of serial twofold dilutions of each phosphocellulose enzyme preparation were included in the reaction mixtures. Between two and four titration experiments were performed for each mutant protein, and the specific activities were calculated from the averages of the slopes of the titration curves. The ATPase specific activities shown are normalized to the specific activity of wild-type (WT) NPH-I.

Structure-function relationships for motif I. Motif I corresponds to the Walker A box, GxGK(T/S), which is common to many proteins that bind and hydrolyze NTPs. We found that replacement of the NPH-I motif I lysine (Lys-61) by alanine elicited a >10⁴ decrement in ATPase activity. Glutamine at this position reduced activity to the same extent (Fig. 2). The ATPase function was restored partially when Lys-61 was replaced by arginine (K61R). Although the K61R mutant enzyme was only 1.4% as active as wild-type NPH-I, it was nonetheless >100-fold more active than the K61A or K61Q mutant. These data establish a clear requirement for a positively charged residue, specifically lysine, at this critical position. To gain further insights into the function of the motif I lysine, we determined the K_m for ATP of wild-type NPH-I and the K61R mutant. ATP hydrolysis was measured as a function of ATP concentration in the presence of single-stranded DNA and 5 mM MgCl₂. The affinity of the K61R mutant for ATP (K_m = 0.39 mM) was similar to that of wild-type NPH-I (K_m = 0.43 mM), but the V_max of K61R was 1.4% of the wild-type value (Table 1).

Structure-function relationships for motif II. Replacement of Asp-141 of motif II by alanine reduced the ATPase activity by 3 orders of magnitude. Similar effects were seen when Asp-141 was replaced by asparagine. Activity was restored to about half the wild-type value when glutamate was introduced (Fig. 2). The D141E mutation did not affect the K_m for ATP, and the V_max was 60% of the wild-type value (Table 1). An acidic side chain at this position is clearly essential for catalysis, and there is no significant steric constraint that precludes replacement of Asp-141 by the larger glutamate side chain.

3

Motif III side chains are not essential for ATP hydrolysis by NPH-I. Alanine substitutions at Ser-183 and Thr-185 of the SAT motif in NPH-I reduced the ATPase activity to 21 and 49% of the wild-type level, respectively (Fig. 2). These modest effects did not meet the criterion of a 10-fold activity decrement that we had set for designating a given side chain as being important for ATP hydrolysis.

Structure-function relationships for motif VI. Replacement of Gln-472 by alanine lowered the ATPase activity to 0.5% of the wild-type value. Conservative changes to asparagine and histidine restored the specific activity to 6 and 17%, respectively, of that of the wild type. The Q472N and Q472H mutations reduced the V_max values to 7 and 29% of that of the wild type, respectively, with little effect on the K_m for ATP (Table 1). These findings suggest that the essential function of Gln-472 entails hydrogen bonding interactions with the amide moiety that are sensitive to its distance from the main chain.

Carboxyl truncations of NPH-I define a minimal ATPase domain. A set of six carboxyl-terminal truncation mutants of NPH-I (named 28, 69, 87, 100, 110, and 121, according to the number of C-terminal amino acids deleted) were expressed in E. coli with N-terminal His tags and then purified from soluble bacterial lysates by Ni-agarose and phosphocellulose column chromatography. SDS-PAGE analysis indicated that the phosphocellulose preparations were highly enriched with respect to the NPH-I polypeptides (Fig. 3A). Western blotting with anti-NPH-I antiserum confirmed that the predominant polypeptide in the enzyme preparations corresponded to NPH-I (Fig. 4B). The specific ATPase activities of the 28, 69, and 87 proteins were essentially identical to that of full-length wild-type NPH-I, whereas 100, 110, and 121 were catalytically inactive (Fig. 4C). This experiment defines a catalytically active ATPase domain (87) from amino acids 1 to 544 that includes all of the conserved motifs. Our findings confirm and refine the C-terminal deletion analysis of Christen et al. (3), who noted that NPH-I residues 1 to 563 were sufficient for ATP hydrolysis whereas a derivative containing residues 1 to 524 was inactive. We suspect that mutants with C-terminal deletions beyond 87 may be defective by virtue of improper protein folding, because whereas the catalytically active 87 enzyme sedimented as a discrete monomer in a glycerol gradient, the inactive 100 polypeptide sedimented diffusely in fractions of larger than monomer size (data not shown). This suggested that the recombinant 100 protein formed nonspecific aggregates. Similar results were obtained for 121. We eschewed N-terminal deletion analysis of NPH-I in light of the close proximity of essential motif I to the amino terminus.

View larger version (33K): [in this window] [in a new window]   FIG. 3.   Carboxyl-terminal truncation mutants of NPH-I. (A) Polypeptide composition. Aliquots (0.3 µg) of the phosphocellulose preparations of wild-type (WT) NPH-I and the indicated C-terminal truncation mutants were electrophoresed through a 10% polyacrylamide gel containing 0.1% SDS. Polypeptides were visualized by staining with Coomassie blue dye. The positions and sizes (in kilodaltons) of marker proteins are indicated at the left of each gel. (B) Western blotting was performed as described in Materials and Methods. (C) DNA-dependent ATPase activities. Reaction mixtures (50 µl) containing 40 mM Tris-HCl (pH 8.0), 2 mM MgCl2, 2 mM DTT, 1 mM [-32P]ATP, 1 µg of heat-denatured DNA, and enzyme as specified were incubated for 30 min at 37°C.

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

Here we have presented an analysis of the role of conserved motifs I, II, III, and VI in DNA-dependent ATP hydrolysis by vaccinia virus NPH-I. Our characterization of 30 different mutant enzymes is the most extensive in vitro structure-function study of a DExH protein to date and the first to address in detail the requirements for ATP hydrolysis by an Snf2-like enzyme. Mutational analyses of motifs I, II, III, and VI have also been performed for vaccinia virus NPH-II (8, 9, 11) and hepatitis C virus NS3 (12, 15) (both of which are DExH box NTPases/helicases), for the yeast DExH box splicing factor Prp16 (this analysis, described in reference 14, was confined to in vivo mutational effects), and for the mammalian DEAD box ATPase eIF4A (22, 23). In the following sections, we review and compare the mutagenesis results for these DEAD or DExH box proteins and interpret them in light of the recently reported crystal structures of several nucleic acid-dependent NTPases.

Motif I. High-resolution crystallographic studies of H-ras p21 showed that the invariant lysine of the Walker A motif makes direct contact with the - and -phosphates of the bound nucleotide (19). Hence, this lysine has been mutated in numerous NTPases. Our finding that the K61A mutation of vaccinia virus NPH-I abrogates NTP hydrolysis agrees with the effects of mutations at the corresponding lysines in eIF-4A (23), NPH-II (11), Prp16 (14), RAD3 (30), PriA (32), and UvrD (6). The K61Q and K61R mutants were also catalytically defective. Kinetic analysis of the K61R mutant showed a severe decrement in the V_max with no significant effect on the K_m for ATP. This argues for an essential role of the motif I lysine in catalysis but not in ATP binding. Similar conclusions were drawn from studies of motif I mutants of Rad3 (a Lys-to-Arg mutant) and UvrD (a Lys-to-Met mutant) (6, 30). In contrast, in the case of eIF4A, mutation of the motif I lysine to asparagine abolished ATP binding as well as ATP hydrolysis (23). Remarkably different effects were reported by Kim et al. (15) for a mutant of hepatitis C virus NS3 in which the lysine was replaced by glutamic acid; they found that the Lys-to-Glu mutant retained NTPase activity (at 29 to 39% of the specific activity of wild-type NS3) and actually had a sixfold-lower K_m for ATP than the wild-type enzyme. Lys-to-Glu and Lys-to-Ala mutations of yeast Prp16 are lethal in vivo: however, an arginine at this position supports cell growth (14).

Motif II. The aspartic acid residue in the Walker B motif of the p21 structure (equivalent to the Asp in the DExH box) interacts via a water molecule with the magnesium ion (19). The aspartate residue of the PcrA DExx box is located near the bound ADP (29). As noted above, this aspartate side chain of NS3 forms a salt bridge with the motif I lysine side chain in the unliganded enzyme. The putative metal-binding and lysine-interacting functions of this residue can be mediated by either Glu or Asp in NPH-I. An asparagine side chain, which would not be expected to fulfill either role, is unable to support the ATPase activity of NPH-I. An Asp-to-Glu change in eIF4A has little effect on the K_m or V_max of the ATPase (but, interestingly, this mutation abolishes RNA unwinding by eIF4A). An NEAD mutation of the DEAD box of eIF4A abolishes ATP hydrolysis without affecting ATP binding (23). In Prp16, the EEAH mutant is viable whereas AEAH and NEAH mutations are lethal (14). The Asp side chain of NPH-II is also important for ATP hydrolysis (8). Thus, the mutational effects at this position are concordant among NPH-I, eIF4A, Prp16, and NPH-II.

Motif III. Motif III of the hepatitis C virus NS3 protein is a flexible linker connecting domains 1 and 2 (31). In NPH-I, neither the serine nor the threonine of motif III is essential for ATP hydrolysis; i.e., replacement with either of these amino acids elicited only a modest decrement in specific activity. Analogous mutations to alanine in the TAT motif of NPH-II also caused a modest (twofold) decrement in DNA-dependent ATPase activity (11). Changing the SAT motif of eIF4A to AAA actually increased the ATPase activity twofold relative to that of the wild-type enzyme (23). Motif III mutants of NPH-II and eIF4A inhibit helicase function even though ATP hydrolysis is preserved. Distinctive effects were reported for the AAT mutation of NS3, which had no effect on the ATPase activity in the absence of nucleic acid but prevented an increase in ATP hydrolysis in response to an RNA cofactor, even though the AAT mutant bound RNA with wild-type affinity (15). The NS3 AAT mutant had 50% of the helicase activity of the wild-type enzyme. It has been suggested that motif III mediates conformational changes coordinated with the ATPase reaction cycle. Helicases like NPH-II and eIF4A may couple the NTP-driven conformational changes to strand separation, whereas NPH-I- and Snf2-like proteins may use NTP-derived energy to effect other macromolecular rearrangements. NPH-I's function in transcription termination depends on its ability to hydrolyze ATP (3-5). The T185A mutation of motif III, which preserves ATP hydrolysis, also preserves the termination factor activity (3). Hence, motif III may not be critical for the biological function of NPH-I. This appears to be the case for Prp16 also, because mutations to alanine in the SAT motif result in functional proteins in vivo (14).

Motif VI. Substitutions with alanine showed that the one glutamine and two arginine side chains of motif VI (QxxGRxxR) are essential for ATP hydrolysis by NPH-I. These three amino acids are conserved throughout the DExH family. In the DEAD box proteins, the first position of motif VI is a histidine rather than a glutamine. It is interesting that the Q472H mutant of NPH-I retains substantial ATPase activity (29% of the wild-type V_max), as does the reciprocal His-to-Gln mutation in motif VI of eIF4A (23% of the wild-type V_max) (23). Similarly, the Gln-to-His mutant of Prp16 is functional in vivo, although the alanine substitution is lethal (14). In contrast, the Gln-to-His mutation in motif VI of NS3 completely abolishes NTPase activity (15).

Conclusions. We have identified by alanine scanning 9 amino acids in motifs I, II, III, and VI of vaccinia virus NPH-I that are essential for ATP hydrolysis. Structure-activity relationships were established for each position by performing conservative amino acid substitutions. Comparison of our findings with mutational studies of other DExH box and DEAD box proteins underscores similarities as well as numerous disparities. We conclude that the functions of the conserved amino acids of the NTPase motifs are context dependent and must therefore be established empirically for any given member of the DExH and DEAD families.