PROTEINS: Structure, Function, and Bioinformatics 67:1128–1137 (2007)
Human DEAD-Box ATPase DDX3 Shows a Relaxed Nucleoside Substrate Specificity Raffaella Franca, Amalia Belfiore, Silvio Spadari, and Giovanni Maga* DNA Enzymology and Molecular Virology Unit, Istituto di Genetica Molecolare IGM-CNR, via Abbiategrasso 207, 27100 Pavia, Italy
ABSTRACT Human DDX3 (hDDX3) is a DEADbox protein shown to possess RNA-unwinding and adenosine triphosphatase (ATPase) activities. The hDDX3 protein has been implicated in nuclear mRNA export, cell growth control, and cancer progression. In addition, a role of this protein in the replication of human immunodeficiency virus Type 1 and in the pathogenesis of hepatitis C virus has been recently proposed. Its enzymological properties, however, are largely unknown. In this work, we characterized its ATPase activity. We show that hDDX3 ATPase activity is stimulated by various ribo- and deoxynucleic acids. Comparative analysis with different nucleoside triphosphate analogs showed that the hDDX3 ATPase couples high catalytic efficiency to a rather relaxed substrate specificity, both in terms of base selection and sugar selection. In addition, its ability to recognize the Lstereoisomers of both 30 deoxy- and 20 ,30 dideoxyribose, points to a relaxed stereoselectivity. On the basis of these results, we hypothesize the presence of structural determinants on both the base and the sugar moieties, critical for nucleoside binding to the enzyme. Our results expand the knowledge about the DEAD-box RNA helicases in general and can be used for rational design of selective inhibitors of hDDX3, to be tested as potential antitumor and antiviral agents. Proteins 2007;67:1128– 1137. VC 2007 Wiley-Liss, Inc. Key words: RNA helicase; ATPase; kinetics; substrate specificity; nucleotides INTRODUCTION RNA helicases are enzymes that participate in numerous biochemical pathways including transcription, splicing, transport, translation, RNA decay and ribosome biogenesis.1,2 The amino acid sequence of RNA helicases is characterized by the presence of eight conserved motifs. On the basis of variations of the sequence in the motifs of the helicase domain, RNA helicases are further classified into families. The two largest human RNA helicase gene families, DDX and DHX, are named after the DEAD-box and DEAH box, respectively, present in the motif II (Walker B) of the helicase domain.3 Biochemically, RNA helicases are capable of binding and hydrolyzing nucleoside triphosphate (NTP), mainly adenosine triphosphate C 2007 WILEY-LISS, INC. V
(ATP). The ATPase activity is an essential function of RNA helicases, since ATP hydrolysis is used as a source of energy for translocation along the nucleic acid lattice and unwinding.4 Contrary to their yeast counterparts,5 the biochemical activities and biological functions of the majority of human RNA helicases are largely unknown. However, two important biological functions for RNA helicases can be recognized, namely dysregulation in cancer and involvement in differentiation.6 Very recently, another important functions has been ascribed to the human DDX3 (hDDX3) RNA helicase, namely to facilitate HIV-1 replication in infected cells, by promoting Revdependent nuclear export of unspliced viral RNAs.7,8 The hDDX3 protein is a DEAD box helicase, whose homologous have been identified in S. cerevisiae (DED1p, DBP1),9,10 X. laevis (AN3)11,12 and M. musculus (PL10).13 Beside its roles in RNA nuclear export and in HIV-1 replication, hDDX3 has been implicated in regulation of ribosome biogenesis and translation during mitosis, as well as in cell growth control.14 Interestingly, hDDX3 has been shown to interact with the hepatitis C virus (HCV) core protein15–17 and to be dysregulated in HCV-induced hepatocellular carcinoma,18 suggesting a role both in cancer progression and in HCV pathogenesis. The recombinant hDDX3 protein has been shown to possess ATPase and RNA helicase activities,8 but no investigation on its biochemical properties has been performed yet. Due to the relevant interest of this protein as a possible target for anticancer and antiviral therapy,19 we have started to characterize its NTPase activity. We show that hDDX3 as an NTPase has rather relaxed substrate specificity, both in terms of base discrimination and sugar stereoselectivity. Comparative analysis with different NTP analogs allowed to hypothesize the presence of structural determinants on both the base and the sugar moieties, critical for binding to the enzyme.
Grant sponsor: European Project; Grant number: FP6 LSHB-CT503480-TRIoH; Grant sponsor: ISS-AIDS National Program Contract; Grant numbers: 40F.48, 40F.78; Grant sponsor: CARIPLO Foundation project ‘‘Oncogenetica e Proteomica della Replicazione’’; Grant number: 2003.1663/10.8441. *Correspondence to: Giovanni Maga, DNA Enzymology and Molecular Virology Unit, Istituto di Genetica Molecolare IGM-CNR, via Abbiategrasso 207, 27100 Pavia, Italy. E-mail:
[email protected] Received 12 September 2006; Revised 2 November 2006; Accepted 11 January 2007 Published online 13 March 2007 in Wiley InterScience (www. interscience.wiley.com). DOI: 10.1002/prot.21433
CHARACTERIZATION OF hDDX3 ATPase ACTIVITY
MATERIAL AND METHODS Chemicals Unlabelled dNTPs and ddNTPs were from Roche Molecular Biochemicals. All other reagents were of analytical grade and purchased from Merck or Fluka. The thymidine analog d4TTP was a generous gift of Dr. S. Victorova. L-(b)deoxy- and L-(b)-dideoxynucleoside triphosphates were a generous gift of Prof. G. Gosselin and were prepared and purified as described.20 Total RNA Extraction Approximately 3 3 107 frozen HeLa cells were lysed in 1 mL TRI REAGENTTM (Sigma) by repeated pipetting and let stand for 5 min. The homogenized sample was vigorously vortexed with 0.2 mL chloroform. After 15 min at room temperature, phases were separated by centrifugation (12,000g, 15 min, 48C). The upper aqueous phase was mixed with 0.5 mL isopropanol and let stand for 10 min. RNA was pelleted by centrifugation (12,000g, 10 min, 48C), washed twice with 75% ethanol, air-dried, resuspend in RNase free water, and heated at 558C for 10 min. DNA contamination was removed by DNase treatment (Ambion). RNA integrity was checked by electrophoresis. Reverse Transcription Reverse transcription was carried out according to the protocol suggested for the Moloney Murine Leukemia Virus Reverse Transcriptase (Promega). Briefly, HeLa total RNA was heated in the presence of 1 lg oligodT (Pharmacia)/lg RNA at 728C for 5 min (15 lL final volume) and then cooled to allow the annealing. After addition of the enzyme, the single strand cDNA synthesis reaction was performed at 428C for 1 h in the manufacturer’s buffer supplemented enzyme, dNTPs (300 lM each) and RNase inhibitor (25 units), according to standard protocol. PCR Amplification and Cloning The human DDX3 coding sequence was amplified from HeLa cells total RNA. The full-length cDNA was first obtained by PCR amplification, using the Pfu DNA polymerase (Promega) and the reverse transcribed single strand cDNAs as template. The product was then further amplified by a subsequent round of PCR. Cycling parameters were: 5 min at 948C followed by 30 cycles as 1 min at 948C/1 min at 638C/5 min at 728C and a final elongation of 10 min at 728C. The BamHI and HindIII restriction sites were introduced at the 50 - and 30 -end of the DDX3 coding sequence respectively by the primers designed (50 -ATACGGATCCTGCATGAGTCATGTGGCAGTG-30 (sense), 50 -ATCGAAGCTTGTTACCCCACCAGTCAACC-30 (antisense), restriction sites sequences in italic). The insert was cloned into the recipient vector pTrcHisA (Invitrogen) and the resulting plasmid was used to transform DH5a E. coli cells according to standard protocols. All purification steps were performed using the Wizard1 SV gel and PCR Clean-up System (Promega).
1129
Correct ligation was checked by control digestion. The resulting pTrcHisA-DDX3 vector encodes the helicase DDX3 fused to an N-terminal peptide of 3 kDa (comprising the 6xHis tag) and seven additional amino acids (KQGCFGG) at the COOH-terminal part. Protein Expression and Purification DH5a cells transformed with the expression vector pTrcHisA-DDX3 were grown at 378C in 0.5 L of Luria Broth to an OD600 of 0.4. Isopropyl b-D-thiogalactoside (IPTG) was added to a final concentration of 1 mM and growth continued for 4 h. Cells were lysed under non denaturating conditions (Buffer A: Sodium phosphate buffer 0.1M pH 8, Tris-HCl 0.01M pH 8, 0.01% NP-40), in the presence of lysozime (1 mg/mL). The lysate was sonicated and suspension was centrifugated at 9000 rpm for 10 min in a Beckman JA20 rotor. DDX3 in the supernatant was purified by chromatography through a FPLC-NiNTA-column, which was first washed with 10 mL of the Buffer B (50 mM NaH2PO4 (pH 8.0), 300 mM NaCl, 20 mM imidazole). hDDX3 protein bound to the column was eluted with Buffer B containing 250 mM imidazole. SDS-PAGE of samples taken after the purification step was performed and the presence of the recombinant protein in the collected fractions was checked by staining (GelCode1 BlueStain Reagent-Pierce) and Western blot with anti-His tag antibodies. Purified hDDX3 was dialyzed against Buffer C (Tris-HCl 0.01M pH 8.0, 0.5 mM DTT, 100 mM NaCl, 30% glycerol) and stored in aliquots at 808C. ATPase Activity Assays The ATPase activity was determined as previously described,20 by directly monitoring [g-32P] ATP hydrolysis by thin-layer chromatography (TLC). A final volume of 5 lL contained: hDDX3 protein, 10 lM [g-32P] ATP as tracer, plus different concentrations of cold ATP and nucleic acids as indicated. Samples were incubated for 15 min at 258C and dotted on to TLC sheets of polyethylenimine cellulose. The products were separated by ascending chromatography with 0.5M KH2PO4 (pH 3.4).The intensities of the radioactive bands corresponding to ATP and Pi were quantified by densitometric scanning with PhosphoImager. Helicase Assays The hDDX3 helicase assay was performed as previously described.21 Briefly, increase in the absorbance at 260 nm was measured with a Ultrospec 2001pro spectrophotometer (GE Healthcare) in a 300 lL quartz cuvette. A final volume of 300 lL contained hDDX3 protein, 50 mM Tris-HCl (pH 7.5), 5 mM MgCl2, 100 lM ATP or NTPs and NA as indicated in the figure legends. Samples containing the NA substrate without enzyme were measured and subtracted to give the increase in AU (FAU). Helicase reactions were started by addition of the enzyme. Samples were equilibrated at room temperature (258C) before measurement. Each measurement was
PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
1130
R. FRANCA ET AL.
repeated at least three times and the mean values used for the interpolation.
50% of inhibition) were calculated by fitting the experimental data to the equation: k0 ¼ k0max ð½I=ID50 Þn =1 þ ð½I=ID50 Þn
Inhibition Assays and Kinetic Constants Determination The apparent affinity for the ATP substrate and the apparent rate of ATP hydrolysis were calculated by measuring the variation of the initial velocities of the ATPase reaction, as a function of ATP concentration. Data were analyzed according to the Michelis–Menten equation: v ¼ ðk E0 Sn Þ=ðKm þ Sn Þ
app kcat =KmATP ¼ ðkcat =KmATP ½NAÞ=ðKmOLIGO þ ½NAÞ
ð1Þ
where kcat/KmATP is the second-order rate constant for ATP hydrolysis, [NA] is the nucleic acid concentration and KmOLIGO is the apparent affinity constant for the nucleic acid substrate. The ratio (kcat/KmATP)/KmOLIGO represents the third-order rate constant (k3) for the interaction of enzyme with free ATP and free nucleic acid as described by the equation: k3
E þ NA þ ATP ! E þ NA þ ADP þ Pi
ð2Þ
For the determination of the kinetic parameters of the helicase reaction, time-course experiments were performed as described in the Figure legends. Data were analyzed according to the mixed exponential equation: 0
FAU ¼ A ð1 ek t Þ þ k00 t
ð5Þ
where S is the concentration of the competing substrate (ATP). The ATPase reaction in the absence of nucleic acid can be described as a one-substrate one-product reaction by the equation: Kd
k2
E þ ATP ! E : ATP ! E þ ADP þ Pi
ð6Þ
where Kd ¼ k1/k1 is the true equilibrium dissociation constant for ATP binding and k2 is the overall rate of product formation. According to a simple Briggs–Haldane mechanism, the apparent affinity for ATP (KATP m ) is given by: ATP ¼ KdATP þ k2 =k1 Km
ð7Þ
Thus, Km Kd only when k1 k2, i.e., in kinetic terms, when the rate limiting step of the reaction is not substrate binding (expressed by the second-order rate constant k1) but a subsequent step leading to product formation (represented by the overall rate k2). Even though we did not determine the rate limiting step of the hDDX3 ATPase reaction, for the vast majority of ATPases studied so far, this condition has been verified. In particular, it has been shown to hold for the prototype DEAD-box protein eIF4,23 so that KATP KATP . m d Similarly, the competitive inhibition of the ATPase reaction by any NTP, can be summarized by the following equation: E þ ATP
KdATP
k2
! E : ATP ! E þ ADP þ Pi
þ NTP
ð3Þ
where A, is the burst amplitude; k0 is the apparent unwinding rate (burst rate), k@ is the rate-limiting steady-state dissociation rate of the enzyme from the NA and t is time. Inhibition assays for ATPase activity were performed under the conditions described for the ATPase assay, in the presence of increasing amounts of the inhibitors. Dose-response curves were generated and ID50 values (representing the concentration of the inhibitor giving PROTEINS: Structure, Function, and Bioinformatics
where k0 is the reaction rate in the presence of the inhibitor concentration [I], k0max is the reaction rate in the absence of the inhibitor and n is the cooperativity index. According to a fully competitive mechanism of action, the ID50 values were converted to the respective inhibition constants (Ki) through the relationship: Ki ¼ ID50 =ð1 þ S=Km Þ
where k is the catalytic rate, E0 is the input enzyme concentration, S is the ATP concentration and n is the cooperativity index. For the analysis of ATPase stimulation by nucleic acid (NA), the method developed by Peck and Herschalg was used.22 Briefly, ATPase stimulation was measured in the presence of different NA concentrations and subsaturating ATP (10 lM, the KmATP is comprised between 25 and 30 lM in the presence or absence of NA, respectively). The apparent second-order rate constant for ATP hydrolysis (kcat/Kapp mATP) was obtained from the observed steady-state hydrolysis rate (kobs), by subtracting the rate constant measured in the absence of NA and then dividing the difference by the enzyme concentration. A concentration of 100 nM hDDX3 was used in all the assays. The dependence of the kcat/Kapp mATP from the NA concentration was analyzed according to the equation:
ð4Þ
#
ð8Þ
KdNTP k02
E : NTPð! E þ NDP þ PiÞ ¼ k01/k10 Where Kd and k2 are as defined in Eq. (8), KNTP d is the equilibrium dissociation constant for NTP binding, and k0 2 is the overall rate for NTP hydrolysis, in case of a substrate inhibitor. Thus, in case of a NTP, which is not hydrolyzed (nonsubstrate inhibitor) the inhibitory constant KiNTP ¼ KNTPd, whereas in the case of a sub-
DOI 10.1002/prot
CHARACTERIZATION OF hDDX3 ATPase ACTIVITY
Fig. 1. Purification of recombinant his-tagged human DDX3 protein. (A) ATPase activity profile of the eluted fractions from the NiNTA chromatographic step. Reactions were carried out as described in Materials and Methods. The numbers refer to the fractions eluted from the column. Each fraction was 0.25 mL. Three microliter of each fraction was used in the assay. L, loading; FT, flow-through; W, wash. Mix: control reaction without added enzyme/fraction. (B) SDS-PAGE elution profile of the NiNTA column. Numbers refer to the same fractions as in Figure 1(A). Forty microliter of each fraction was loaded on the gel. L, loading; FT, flow-through; W, wash. MWM, molecular weight markers. Gel was stained, dried and scanned to obtain the digital image. (C) Dose-dependency of the ATPase reaction was measured by titrating different amounts of the main peak fractions in an ATPase assay. Numbers refer to the fractions of the NiNTA column, as indicated in panels A and B.
strate inhibitor, KiNTP ¼ KdNTP 1 k20 /k10 . Then, the same considerations exposed above for the relationship between ATP Km and KATP , apply also to KNTP and KNTP . d i d Thus, based on the assumption that binding of ATP or NTPs was not the rate-limiting step of the reaction, we directly compared the KATP with the KNTP values for the m i different NTPs, to derive the corresponding specificity indexes (S.I.).
1131
Fig. 2. Dependence of human DDX3 ATPase activity from Mg11 and Mn11. (A) Increasing concentrations of either Mg11 (lanes 1–4) or Mn11 (lanes 6–10) were titrated into an ATPase assay in the presence of 0.75 pmol of recombinant hDDX3. Reactions were carried out as described in Materials and Methods. Lane 5, control reaction in the absence of metal cofactor. (B) The radioactive spots corresponding to ATP and Pi were quantified and the relative values of ATP hydrolysis (%) were plotted as a function of the metal cofactor concentration. Error bars represent S.D. values of three independent experiments.
pTrCHis vector. The his-tagged recombinant protein was expressed in E. coli cells and purified by Ni11-affinity chromatography. Figure 1(A) shows the elution profile of the ATPase activity through the Ni-NTA column. One single peak of ATPase was detected in fractions 15–21. Figure 1(B) shows the SDS-PAGE analysis of the same fractions. As can be seen, a single polypetide of the expected molecular weight could be seen in the fractions containing the ATPase activity. Analysis of the fractions by SDS-PAGE allowed to estimate a purity of the recombinant protein >95% (data not shown). Western Blot analysis with antihis tag antibodies confirmed that the polypeptide contained a his-tag (data not shown). The main peak was dialyzed to eliminate imidiazole and the specific activity was determined, as shown in Figure 1(C), and found to be 350 pmol ATP 3 lg1 3 min1. Thus, hDDX3 could be successfully expressed in bacterial cells as a his-tagged protein and purified to near homogeneity in a single step. The hDDX3 Protein has Mg11/Mn11-Dependent ATPase Activity
RESULTS Expression and Purification of Recombinant Human DDX3 The human (h) DDX3 full length cDNA was amplified by RT-PCR from total HeLa cells RNA and cloned into a
Next, the activating ion specificity for the ATPase activity of hDDX3 was investigated. Increasing concentrations of either Mg11 or Mn11 were titrated into an ATPase assay. As shown in Figure 2(A), no ATP hydroly-
PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
1132
R. FRANCA ET AL.
sis was observed in the absence of metal ions (lane 5), whereas stimulation of ATPase could be measured in the presence of both Mg11 (lanes 1–4) and Mn11 (lanes 6– 10). Quantification of the products [Fig. 2(B)], allowed to identify Mg11 as the optimal metal cofactor at a concentration of 5 mM. The optimal concentration of Mn11 was 0.5 mM, but the overall activity was two-fold lower than with Mg11. Additional metals tested were Zn11, Be11, Ca21 and Fe11. None of them supported the
Fig. 3. The ATPase activity of human DDX3 is partially stimulated by various nucleic acids. (A) Increasing concentrations of activated DNA (lanes 2–4); ss DNA oligonucleoside dA200 (lanes 5–7); ss RNA oligonucleoside rA200 (lanes 8–10) or ss RNA oligonucleoside rA40 (lanes 11–13), were titrated into an ATPase assay in the presence of 0.35 pmol of recombinant hDDX3 protein. Lane 1, control reaction without added nucleic acid. (B) Increasing concentrations of partially ds RNA/DNA hybrids dT20/rA200 (lanes 3–6) or rU20/dA200 (lanes 7–10) were titrated into an ATPase assay in the presence of 0.35 pmol of recombinant hDDX3 protein. Lane 1, control reaction without enzyme; lane 2 control reaction without added nucleic acid. The relative fold stimulation with respect to the reaction in the absence of NA is indicated on bottom of each panel.
ATPase activity of hDDX3, indicating a strict requirement for either Mg11 or Mn11 (data not shown). The ATPase Activity of hDDX3 is Only Partially Stimulated by Nucleic Acids Next, the effects of various nucleic acids on the ATPase activity of hDDX3 were investigated. In a first set of experiments, both DNAs and RNAs were titrated into the reaction. Limiting amounts of enzyme were used, in order to amplify the stimulatory effects. As shown in Figure 3(A), a stimulatory effect was observed in the presence of genomic calf thymus DNA, which has been partially digested with DNaseI to create singestrand (ss) gaps (lanes 2–4), or with the ss DNA oligonucleoside dA200 (lanes 5–7), as well as with ss RNA oligonucleosides rA200 (lanes 8–10) and rA40 (lanes 11–13). Figure 3(B) shows similar experiments performed in the presence of increasing amounts of partially doublestrand (ds) RNA/DNA hybrids dT20/rA200 (lanes 3–6) and rU20/dA200 (lanes 7–10). As can be seen from the quantifications, the ATPase activity of hDDX3 was stimulated only two-fold by both deoxy- and ribonucleic acid substrates, either ss or partially ds, suggesting that the presence of a NA is not essential for the ATPase reaction. By performing nucleic acid (NA) titrations, over a broad range of concentrations, into the ATPase assay, the apparent affinity of hDDX3 for the different NAs (KmOLIGO) and the corresponding second-order rate constant for ATP hydrolysis (kcat/KmATP) and the third-order rate constant for the interaction of the enzyme with free NA and free ATP (k3) were calculated (see also Methods section) and the values are summarized in Table I. In absolute terms, the differences among the different NAs tested were small (two-fold max.). However, a clear correlation was noted between the affinity for the NA and the stimulation of ATP hydrolysis, so that the NA with the highest affinity showed the highest kcat/KmATP values. The order of preference of hDDX3 for ssNA substrates was: (dA)200 (rA)200 (rA)40, whereas for partially dsNAs it was (rU)20/(dA)200 (dT)20/(rA)200. The Kinetic of ATP Hydrolysis by hDDX3 is Influenced by the Presence of a Nucleic Acid The experiments shown above demonstrated that the ATPase activity of hDDX3 is partially stimulated by NA. Next, the apparent rate of ATP hydrolysis was measured
TABLE I. Kinetic Parameters for Stimulation of the ATPase Activity of hDDX3 by Various DNA and RNA Substrates Nucleic acid (dA)200 (rA)200 (rA)40 (dT)20/(rA)200 (rU)20/dA(200)
kcat/KmATPa (3 106) (M1 min1) 1.2 0.7 0.9 0.98 1.2
0.2 0.1 0.1 0.07 0.2
a
KmOLIGO (lM) 0.010 0.020 0.14 0.020 0.015
0.002 0.003 0.03 0.002 0.003
k3 (M2 min1) (1.2 0.1) (3.5 0.5) (6 1) (4.9 0.5) (8 1)
3 3 3 3 3
1014 1013 1012 1013 1013
Kinetic parameters kcat/KmATP, KmOLIGO, and k3 were calculated according to Eqs. (1) and (2), for details see Materials and Methods.
PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
1133
CHARACTERIZATION OF hDDX3 ATPase ACTIVITY
as function of the ATP concentration and in the absence or in the presence of NA. Half-saturating amounts of NA were used in the reaction, in order to be in the linear range of stimulation. As shown in Figure 4, in the absence of NA the reaction followed an hyperbolic relationship with the substrate concentration, whereas addition of NA induced a sigmoidal dose-response curve. Analysis with the Hill equation gave a cooperativity index (nH) of 0.9 in the absence of NA, whereas in the presence of NA the nH value increased to 2.03, indicating an apparent positive cooperativity. The kinetic parameters are summarized in Table II. Addition of NA increased the apparent rate of hydrolysis (kcat) as well as the catalytic efficiency (kcat/Km) of the ATPase activity of hDDX3. The hDDX3 Protein ATPase is Sensitive to Various NTPs Next we tested the substrate binding specificity of hDDX3 NTPase activity by testing various NTPs, whose structures are shown in Figure 5(A), for their ability to compete with ATP as the substrate for hydrolysis by hDDX3. In a first set of experiments, all four ribo- and deoxy-NTPs were tested in the ATPase assay. As shown in Figure 5(B), dATP (lane 4), dGTP (lane 7), rCTP (lane 10), and rGTP (lane 11) were able to inhibit ATP hydrolysis by hDDX3. Unmodified dideoxynucleosides such as ddA [Fig. 5(C), lane 3], ddT (lane 4) and ddC (lane 5), but not ddG, potently inhibited the ATPase reaction.
Fig. 4. Nucleic acid induces cooperative ATP binding by human DDX3. The ATPase activity of 0.75 pmol of recombinant hDDX3 was measured as a function of ATP concentration, in the absence (circles) or in the presence (triangles) of 0.05 lM. DNA. Both data sets were fitted to the Hill’s equation, resulting in cooperativity coefficients (nH) of 0.9 and 2.03, respectively. Error bars represent S.D of three independent measurements.
Among the modified dideoxynucleosides analogs, only the 20 ,30 -didehydro TTP analog (d4T, lane 10) showed an inhibitory effect. We also tested some L-stereoisomers of the natural D-nucleosides. As shown in Figure 5(C), LddCTP (lane 7) and L-dATP (lane 8), were able to efficiently suppress ATP hydrolysis. Quantitative Analysis of hDDX3 ATPase Inhibition by Different Nucleosides Triphosphates Figure 6A shows the results of experiments in which different concentrations of dATP, ddATP, and L-dATP were titrated into the reaction. Their apparent inhibition constants (Ki) were then determined [Fig. 6(B)]. The Ki values of other NTPs were similarly determined and listed in Table III. As explained in Materials and Methods, the Ki values derived reflect the apparent binding affinities of the enzyme for the tested triphosphates and were thus compared with the apparent binding affinity (Km) of ATP (Fig. 4(A) and Table III), allowing the calculation of the respective selectivity indexes (S.I. defined as KdNTP/KmATP, see also Materials and Methods). As summarized in Table III, hDDX3 shows S.I. values of 0.53, 0.3, and 1.2 for dATP, ddATP and L-dATP, respectively. Similar experiments were performed with the other triphosphates and the calculated Ki and S.I. values are reported in Table III. We also tested whether additional NTPs other than ATP were able to support the RNA helicase activity of hDDX3. Saturating amounts of each nucleoside (100 lM) were used, so that the observed differences in unwinding rates could be directly ascribed to the catalytic step and not to differences in binding affinities. The ability of recombinant hDDX3 to unwind an RNA/RNA duplex was measured by continuously monitoring the increase in absorbance at 260 nm (AU260) resulting from doubleto single-strand transition. The variation of AU260 of control reactions in the presence of the various NTPs but in the absence of NA was determined and subtracted as background. As shown in Figure 6(C), CTP, dATP, and LdATP were able to be utilized as phosphate donors for the RNA helicase reaction in the presence of rU20/rA40, albeit with different efficiencies. The apparent unwinding rates (k0 , see Material and Methods), were 7.9, 7.5, 2.5, and 0.18 min1 for ATP, CTP, dATP, and L-dATP, respectively. From this analysis, however, it was not possible to determine whether these different rates reflected the relative efficiencies of hydrolysis for the different nucleosides, or whether other steps of the reaction were slowed down.
TABLE II. Kinetic Parameters for ATP Hydrolysis by hDDX3 in the Absence or in the Presence of Nucleic Acid Nucleic acid None (dA)200 a b
ATP Km
a
(mM)
0.040 0.003 0.030 0.004b
kcat (min1)
kcat/KATP (3 106) (M1 min1) m
nH
14 1 36 3
0.35 0.06 1.2 0.3
0.9 2.03
Values are the means of three independent experiments. Number in brackets are S.D. units. K0.5 value obtained from the Hil equation. PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
Fig. 5. The ATPase activity of human DDX3 is inhibited by various NTPs. (A) Structures of the NTPs analogs used. (B) All four deoxy- and ribonucleosides were tested as inhibitors of the ATPase reaction catalyzed by 0.75 pmol of hDDX3. Lane 1, control reaction without enzyme; lanes 2 and 3, control reactions without added inhibitors. Lanes 4–7: dATP, dTTP, dCTP, and dGTP, respectively at 100 lM each; lanes 8– 11: rATP, rUTP, rCTP, and rGTP, respectively at 100 lM each. (C) Various nucleoside analogs were tested as inhibitors of the ATPase reaction catalyzed by 0.75 pmol hDDX3. Lane 1, control reaction without enzyme; lane 2, control reaction without added inhibitors. Lanes 3–6: ddATP, ddTTP, ddCTP, and ddGTP, respectively at 100 lM each. Lane 7: L-ddCTP at 10 lM final concentration; lanes 8 and 9: L-dATP and LdTTP, respectively at 100 lM each. Lanes 10–12: d4TTP, AZTTP, and 30 FTTP, respectively at 100 lM each.
CHARACTERIZATION OF hDDX3 ATPase ACTIVITY
Fig. 6. Effects of ATP analogs on the ATPase activity of human DDX3. (A) Increasing concentrations of dATP (lanes 4–8); L-dATP9–13 or ddATP (lanes 14–18), were titrated in the ATPase reaction catalyzed by 0.75 pmol of human DDX3. Lane 1, control reaction without enzyme, Lanes 2 and 3: control reaction without inhibitors. (B) Radioactive spots corresponding to ATP and Pi were quantified from experiments similar to those shown in panel A, and the corresponding apparent velocity values were plotted as a function of the different NTPs concentrations. Dose-response curves were generated and the corresponding Ki values were calculated according to a competitive mechanism of action. (C) Saturating amounts (100 lM) of ATP, CTP, dATP or L-dATP were used in the RNA helicase reaction catalyzed by 0.75 pmols of hDDX3, in the presence of 300 nM rU20/rA40. The reaction was monitored continuously at 260 nm and the absorbance values (AU) plotted as a function of time. The apparent unwinding rate (burst rate, k0 ) was determined as described in Material and Methods. Absorbance values for control reactions in the absence of enzyme or in the absence of NTP were AU260 ¼ 0.045 0.004 and 0.05 0.006, respectively. The absorbance of the denatured template, corresponding to 100% of unwound substrate was AU260 ¼ 0.13 0.004 as indicated by the dashed line on the plot.
DISCUSSION The best characterized DEAD RNA helicase to date is the eIF4A protein.22–24 Under comparable reaction conditions, eIF4A displayed a lower catalytic efficiency for ATP hydrolysis than hDDX3 (Tables I and II), with Km and kcat values of 330 lM and 4 min1, respectively, giving a 0.012 lM1, min1 efficiency value. The ATPase ac-
1135
tivity of eIF4 was greatly stimulated by different RNA substrates, contrary to the slight stimulation observed for hDDX3. Analysis of the stimulation of eIF4 ATPase activity by oligo(U) substrates of increasing lengths, showed maximal kcat/Km values for ATP hydrolysis in the 105M1, min1 range, which are comparable to the values listed in Table I for hDDX3 with oligo(A) substrates. However, the apparent affinity values (KmOLIGO) displayed by hDDX3 for all the NAs tested were significantly lower than the reported values for eIF4 with oligo(U) substrates, resulting in third-order rate constant k3 values of about two order of magnitude higher. These differences are unlikely due to the different sequence of the NAs (A vs. U homopolymers), since it has been shown that eIF4 ATPase stimulation is not dependent on the particular RNA sequence. Another difference was that DNA substrates such as poly(dA) stimulated the ATPase activity of hDDX3, whereas no stimulation was observed by DNA for eIF4. Thus, the ATPase activity of hDDX3 appears to have different properties than the one of eIF4. The data shown in Figure 4 demonstrate that NA causes a sigmoidal kinetics for ATP hydrolysis by hDDX3. This can be due to a conformational change in the hDDX3 monomer converting its ATP binding site from a low-affinity to an high-affinity state, or to a NA induced multimerization of hDDX3, resulting in a different binding affinity for ATP of the adjacent subunits. ATP-induced conformational changes have been shown to play a role in the unwinding mechanism by various cellular and viral DEAD-box RNA helicases.4,25–33 To understand the basis of this effect on hDDX3, it would important to determine whether its active form is a monomer or a multimer. The native quaternary structure of hDDX3 is currently under investigation in our laboratory. The different binding affinities shown by hDDX3 with respect to different NTPs can be explained both in terms of base and sugar selection. As summarized in Table III, for both ribo- and deoxynucleosides, the order of binding preference to the enzyme is ATP > CTP TTP UTP GTP. Comparison of the base rings of adenine and cytosine shows that the NH2 substituent at position N-6 and N-4 of the purine and pyrimidine rings, respectively, occupies a spatially equivalent position. By comparison, the guanine, uracil, and thymine rings bear an oxygen group at the same positions (O-6 and O-4, respectively). Thus, the NH2 group in the base ring appears to be a critical determinant for nucleoside binding. Interestingly, in the crystal structure of the DECD family human RNA-dependent ATPase UAP5634,35 in complex with ADP, the N-6 of the adenine ring was shown to be hydrogen bonded with a Glutamine residue of the conserved Q-motif. This residue (Q72 in UAP56) is conserved also in the Q-motif of DEAD-family helicases, including hDDX3 (Q207),14 suggesting that a similar interaction might be important also for hDDX3 and providing additional support for a crucial role of the N-6 substituent in nucleoside binding.
PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
1136
R. FRANCA ET AL.
TABLE III. Apparent Binding Affinities and Selectivity Indexes of hDDX3 NTPase Activity for Different Nucleoside Triphosphatesa Km (lM)
Triphosphate ATP dATP ddATP L-dATP CTP dCTP ddCTP L-ddCTP UTP dTTP L-dTTP ddTTP d4TTP AZTTP 30 FTTP GTP dGTP ddGTP
20 1
Ki (lM)
S.I. (Ki/KmATP)
10.7 0.3 6.1 0.8 24.1 0.5 35.2 0.9 25 10 2.4 0.7 29.7 0.8 >100 >100 >100 17.7 0.7 9.6 0.6 >100 >100 60 10 78 10 80 10
1 0.53 0.3 1.2 1.75 1.25 0.12 1.5 >5 >5 >5 0.88 0.48 >5 >5 3 3.9 4
Stereoselectivity (Ki(L)/Kd(D))
2.25
12.3
a
Assays were performed as described in Material and Methods. Values represent the mean of three independent experiments. Numbers in brackets are S.D.
As for the nature of the sugar ring, it appears that the affinity of hDDX3 decreases in the order ribo- < (i.e. lower affinity) < deoxy- < dideoxyribose, for both ATP and CTP, suggesting that the lack of substituents at positions 20 and 30 of the sugar facilitates nucleoside binding. This is more clear in the case of the TTP analogs series. Whereas UTP and dTTP do not show any measurable affinity to the enzyme, the dideoxy- and didehydro-derivatives (ddTTP and d4TTP) bind to hDDX3 with high affinity. When the 30 position of ddTTP is substituted with an azido (AZTTP) or a fluorine (30 FTTP) group, the affinity drops dramatically, indicating a role of the 20 and 30 positions in determining the binding of nucleosides to the active site of the enzyme. Overall, these data allow to hypothesize that specificity of nucleoside binding by hDDX3 is attained by two mechanisms: base selection at the level of the substituent at position N-6 of purines and N-4 of pyrimidines and a tight steric control at the positions 20 and 30 of the sugar ring, whereby decreasing the steric hindrance of the sugar substituents decreases binding specificity. When the optical L-enantiomers of natural D- nucleosides were tested, as in the cases of dATP/ L-dATP and ddCTP/L-ddCTP, a moderate stereoselectivity was noted, ranging from 2.25- to 12.3-fold. These results indicated that, even though the preferred configuration of the sugar is still the naturally occurring D-, this is not an absolute requirement, since also L-nucleosides can be recognized with significant affinities. CONCLUSIONS In summary, our results suggest that the NTPase activity of hDDX3 has a very broad and relaxed substrate binding specificity, both in terms of base selection (since it was inhibited by purine and pyrimidine nucleosides) PROTEINS: Structure, Function, and Bioinformatics
and sugar selection (being able to recognize also 30 deoxy-, 20 ,30 dideoxy- and 20 ,30 didehydro-ribose moieties). In addition, its ability to recognize the L-stereoisomers of both 30 deoxy- and 20 ,30 dideoxy-ribose, points to a relaxed stereoselectivity. We have hypothesized the presence of some key structural determinants in the nucleoside scaffold, which are important for binding to hDDX3. These data expand our knowledge of the DEADbox RNA helicases in general and can be used for rational design of selective inhibitors of hDDX3, to be tested as potential antitumor and antiviral agents. Indeed, L-stereoisomers of nucleoside analogs have been already developed as potent anti-HIV and anti-hepatitis B virus agents. The rationale for their use is based on the general high steroselectivity (i.e. preference of the Disomer) of cellular enzymes with respect to the viral ones.36 The discovery that hDDX3, albeit being a cellular enzyme, can be targeted by L-nucleosides, allows to hypothesize the possibility of designing specific inhibitors of its activity, minimizing their unspecific interaction with other nucleotide metabolizing enzymes. REFERENCES 1. Rocak S, Linder P. DEAD-box proteins: the driving forces behind RNA metabolism. Nat Rev Mol Cell Biol 2004;5:232–241. 2. Tanner NK, Linder P. DExD/H box RNA helicases: from generic motors to specific dissociation functions. Mol Cell 2001;8:251– 262. 3. Abdelhaleem M, Maltais L, Wain H. The human DDX and DHX gene families of putative RNA helicases. Genomics 2003;81:618– 622. 4. Delagoutte E, von Hippel PH. Helicase mechanisms and the coupling of helicases within macromolecular machines. I. Structures and properties of isolated helicases. Q Rev Biophys 2002;35:431– 478. 5. Linder P. Yeast RNA helicases of the DEAD-box family involved in translation initiation. Biol Cell 2003;95:157–167.
DOI 10.1002/prot
CHARACTERIZATION OF hDDX3 ATPase ACTIVITY 6. Abdelhaleem M. RNA helicases: regulators of differentiation. Clin Biochem 2005;38:499–503. 7. Krishnan V, Zeichner SL. Alterations in the expression of DEAD-box and other RNA binding proteins during HIV-1 replication. Retrovirology 2004;1:42. 8. Yedavalli VS, Neuveut C, Chi YH, Kleiman L, Jeang KT. Requirement of DDX3 DEAD box RNA helicase for HIV-1 RevRRE export function. Cell 2004;119:381–392. 9. Jamieson DJ, Beggs JD. A suppressor of yeast spp81/ded1 mutations encodes a very similar putative ATP-dependent RNA helicase. Mol Microbiol 1991;5:805–812. 10. Iost I, Dreyfus M, Linder P. Ded1p, a DEAD-box protein required for translation initiation in Saccharomyces cerevisiae, is an RNA helicase. J Biol Chem 1999;274:17677–17683. 11. Gururajan R, Weeks DL. An3 protein encoded by a localized maternal mRNA in Xenopus laevis is an ATPase with substratespecific RNA helicase activity. Biochim Biophys Acta 1997; 1350:169–182. 12. Askjaer P, Rosendahl R, Kjems J. Nuclear export of the DEAD box An3 protein by CRM1 is coupled to An3 helicase activity. J Biol Chem 2000;275:11561–11568. 13. Leroy P, Alzari P, Sassoon D, Wolgemuth D, Fellous M. The protein encoded by a murine male germ cell-specific transcript is a putative ATP-dependent RNA helicase. Cell 1989;57:549–559. 14. Sekiguchi ,T, Fukumura J. Phosphorylation of dead-box RNA helicase DDX3 by mitotic cyclin B/CDC2, but not cyclin A/CDK2. J Biol Chem, in press. 15. Mamiya N, Worman HJ. Hepatitis C virus core protein binds to a DEAD box RNA helicase. J Biol Chem 1999;274:15751–15756. 16. Owsianka AM, Patel AH. Hepatitis C virus core protein interacts with a human DEAD box protein DDX3. Virology 1999;257:330–340. 17. You LR, Chen CM, Yeh TS, Tsai TY, Mai RT, Lin CH, Lee YH. Hepatitis C virus core protein interacts with cellular putative RNA helicase. J Virol 1999;73:2841–2853. 18. Chang PC, Chi CW, Chau GY, Li FY, Tsai YH, Wu JC, Wu Lee YH. DDX3, a DEAD box RNA helicase, is deregulated in hepatitis virus-associated hepatocellular carcinoma and is involved in cell growth control. Oncogene 2005;25:1991–2003. 19. Kwong AD, Rao BG, Jeang KT. Viral and cellular RNA helicases as antiviral targets. Nat Rev Drug Discov 2005;4:845–853. 20. Locatelli GA, Gosselin G, Spadari S Maga G. Hepatitis C virus NS3 NTPase/helicase: different stereoselectivity in nucleoside triphosphate utilisation suggests that NTPase and helicase activities are coupled by a nucleoside-dependent rate limiting step. J Mol Biol 2001;313:683–694. 21. Maga G, Gemma S, Fattorusso C, Locatelli GA, Butini S, Persico M, Kukreja G, Romano MP, Chiasserini L, Savini L, Novellino E, Nacci V, Spadari S, Campiani G. Specific targeting of hepatitis C virus NS3 RNA helicase. Discovery of the potent and selective competitive nucleoside-mimicking inhibitor QU663. Biochemistry 2005;44:9637–9644.
1137
22. Peck ML, Herschlag D. Effects of oligonucleotide length and atomic composition on stimulation of the ATPase activity of translation initiation factor eIF4A. RNA 1999;5:1210–1221. 23. Lorsch JR, Herschlag D. The DEAD Box Protein eIF4A. I. A minimal kinetic and thermodynamic framework reveals coupled binding of RNA and nucleotide. Biochemistry 1998;37:2180– 2193. 24. Rogers GW, Jr, Richter NJ, Merrick WC. Biochemical and kinetic characterization of the RNA helicase activity of eukaryotic initiation factor 4A. J Biol Chem 1999;274:12236–12244. 25. Dumont S, Cheng W, Serebrov V, Beran RK, Tinoco I, Jr, Pyle AM Bustamante C. RNA translocation and unwinding mechanism of HCV NS3 helicase and its coordination by ATP. Nature 2006;439:105–108. 26. Hsieh J, Moore KJ, Lohman TM. A two-site kinetic mechanism for ATP binding and hydrolysis by E. coli Rep helicase dimer bound to a single-stranded oligodeoxynucleoside. J Mol Biol 1999;288:255–274. 27. Levin MK, Gurjar M, Patel SS. A Brownian motor mechanism of translocation and strand separation by hepatitis C virus helicase. Nat Struct Mol Biol 2005;12:429–435. 28. Locatelli GA, Spadari S, Maga G. Hepatitis C virus NS3 ATPase/helicase: an ATP switch regulates the cooperativity among the different substrate binding sites. Biochemistry 2002;41:10332–10342. 29. Moolenaar GF, Herron MF, Monaco V, van der Marel GA, van Boom JH, Visse R, Goosen N. The role of ATP binding and hydrolysis by UvrB during nucleoside excision repair. J Biol Chem 2000;275:8044–8050. 30. Myong S, Rasnik I, Joo C, Lohman TM, Ha T. Repetitive shuttling of a motor protein on DNA. Nature 2005;437:1321–1325. 31. Serebrov V, Pyle AM. Periodic cycles of RNA unwinding and pausing by hepatitis C virus NS3 helicase. Nature 2004;430: 476–480. 32. Sharma S, Sommers JA, Choudhary S, Faulkner JK, Cui S, Andreoli L, Muzzolini L, Vindigni A Brosh RM, Jr. Biochemical analysis of the DNA unwinding and strand annealing activities catalyzed by human RECQ1. J Biol Chem 2005;280:28072–28084. 33. Talavera MA, De La Cruz EM. Equilibrium and kinetic analysis of nucleoside binding to the DEAD-box RNA helicase DbpA. Biochemistry 2005;44:959–970. 34. Shi H, Cordin O, Minder CM, Linder P, Xu RM. Crystal structure of the human ATP-dependent splicing and export factor UAP56. Proc Natl Acad Sci USA 2004;101:17628–17633. 35. Zhao R, Shen J, Green MR, MacMorris M, Blumenthal T. Crystal structure of UAP56, a DExD/H-box protein involved in premRNA splicing and mRNA export. Structure 2004;12:1373– 1381. 36. Focher F, Spadari S, Maga G. Antivirals at the mirror: the lack of stereospecificity of some viral and human enzymes offers novel opportunities in antiviral drug development. Curr Drug Targets Infect Disord 2003;3:41–53.
PROTEINS: Structure, Function, and Bioinformatics
DOI 10.1002/prot
