The Rep52 Gene Product of Adeno-Associated Virus Is a DNA Helicase with 3′-to-5′ Polarity

Abstract

The rep gene of adeno-associated virus type 2 encodes four overlapping proteins from two separate promoters, termed P₅ and P₁₉. The P₅-promoted Rep proteins, Rep78 and Rep68, are essential for viral DNA replication, and a wealth of data concerning the biochemical activities of these proteins has been reported. In contrast, data concerning the biochemical functions of the P₁₉-promoted Rep proteins, Rep52 and Rep40, are lacking. Here, we describe enzymatic activities associated with a bacterially expressed maltose-binding protein (MBP)-Rep52 fusion protein. Purified MBP-Rep52 possesses 3′-to-5′ DNA helicase activity that is strictly dependent upon the presence of nucleoside triphosphate and divalent cation cofactors. In addition, MBP-Rep52 demonstrates a constitutive ATPase activity that is active in the absence of DNA effector molecules. An MBP-Rep52 chimera bearing a lysine-to-histidine substitution at position 116 (K116H) within a consensus helicase- and ATPase-associated motif (motif I or Walker A site) was deficient for both DNA helicase and ATPase activities. In contrast to a Rep78 A-site mutant protein bearing a corresponding amino acid substitution at position 340 (K340H), the MBP-Rep52 A-site mutant protein failed to exhibit a trans-dominant negative effect when it was mixed with wild-type MBP-Rep52 or MBP-Rep78 in vitro. This lack of trans dominance, coupled with the results of coimmunoprecipitation and gel filtration chromatography experiments reported here, suggests that the ability of Rep52 to engage in multimeric interactions may differ from that of Rep78 or -68.

Adeno-associated virus type 2 (AAV-2) is a nonpathogenic parvovirus of humans that requires coinfection with a helper virus for efficient productive replication (1, 2, 14, 35). In the absence of helper virus coinfection, AAV establishes a latent infection via integration into a specific locus (termed AAVS1) within the long arm of human chromosome 19 (9, 20, 21, 38, 50). The AAV virion consists of a single-stranded, linear DNA molecule packaged within a nonenveloped, 20- to 25-nm icosahedral capsid composed of three antigenically related capsid proteins (VP1, VP2, and VP3) (1, 14, 36). The viral genome is approximately 4.7 kb in length and encodes two genes, rep and cap (1, 42). The rep and cap genes are flanked by 145-nucleotide (nt) inverted terminal repeats that serve as the viral origins of replication and encapsidation signals (15, 37, 42, 47). The rep open reading frame (ORF) is expressed from two promoters positioned at map units 5 (P₅ promoter) and 19 (P₁₉ promoter) (26). The cap gene is expressed from a third promoter positioned at map unit 40 (P₄₀ promoter). The P₅ promoter directs the synthesis of the two largest rep gene products, Rep78 and Rep68 (26, 30, 42). P₁₉, a promoter internal to the full-length rep ORF, directs the synthesis of Rep52 and Rep40, which lack the 224-amino-acid N-terminal domain common to Rep78 and Rep68 (26, 30, 42). In addition, a splicing event results in the replacement of 92 C-terminal amino acid residues common to Rep78 and Rep52 with a 7-amino-acid tail in Rep68 and Rep40. Of the four Rep proteins, Rep78 and Rep52 are the most abundantly expressed during helper virus-induced AAV replication (32, 34).

Rep78 and Rep68 serve as viral replication initiator proteins (4, 15, 31, 48). These proteins recognize cognate binding sites within the viral origin of replication and nick the origin at a specific site (known as the terminal resolution site) (6, 7, 17, 29, 40, 41). The nicking event provides a free 3′-hydroxyl group that primes viral DNA synthesis. In addition to DNA-binding and site-specific endonuclease activities, Rep78 and Rep68 have been shown to possess helicase and ATPase activities in vitro (5, 8, 17, 18, 45, 46, 51). In contrast to the relatively large amount of data concerning the biochemical activities of the P₅-promoted Rep proteins, information concerning the biochemical functions of the P₁₉-promoted Rep proteins is lacking. Rep52 and Rep40 fail to bind AAV inverted-terminal-repeat DNA in electrophoretic mobility shift assays and also fail to demonstrate detectable endonuclease activities in vitro (18, 40, 49). These results are consistent with the modular domain structure of the Rep proteins. The site-specific DNA-binding domain, as well as conserved amino acid motifs associated with DNA cleavage and covalent attachment, occurs within the first 200 residues of Rep78 and -68 and is therefore absent from Rep52 and -40 (16, 33, 45). However, amino acid sequences common to all four Rep proteins display motifs associated with numerous prokaryotic and eukaryotic helicase and ATPase proteins (10–13, 44). This observation has prompted us to investigate the ability of a representative P₁₉-promoted Rep protein, Rep52, to function as a helicase and/or ATPase. To this end, we have cloned and expressed Rep52 as a maltose-binding protein (MBP)-Rep52 fusion protein in Escherichia coli. We report here that MBP-Rep52 possesses a 3′-to-5′ DNA helicase activity as judged by its ability to unwind various partial duplex DNA substrates. MBP-Rep52-mediated helicase activity is strictly dependent upon the presence of both nucleoside 5′-triphosphate (NTP) and divalent cation cofactors. In addition, MBP-Rep52 possesses an ATPase activity that is not dependent upon the presence of DNA effector molecules. Coimmunoprecipitation and gel filtration chromatography experiments with in vitro-translated Rep52 suggest that the ability of Rep52 to multimerize may differ from that of the P₅-promoted Rep proteins.

MATERIALS AND METHODS

Plasmids and recombinant proteins.

Chimeric MBP-Rep fusion proteins were expressed from plasmids pMBP-Rep52 and pMBP-Rep78, which contain AAV sequences 993 to 2187 and 321 to 2187, respectively, cloned in frame with the malE gene of pMAL-c2 (New England Biolabs, Inc. [NEB], Beverly, Mass.). To generate plasmids pMBP-Rep52(NTP) and pMBP-Rep78(NTP), the 0.9-kb XmnI-XbaI fragments of pMBP-Rep52 and pMBP-Rep78, respectively, were replaced with the 0.9-kb XmnI-XbaI fragment of pRep(1-621)-K340H (39). MBP fusion proteins were expressed in E. coli DH5α and purified by amylose affinity chromatography as described previously (5, 6). In vitro-translated Rep52 was expressed from plasmid pSR378, which contains AAV sequence 988 to 2187 under the control of a bacteriophage T7 promoter in a pGEM-3Z background (Promega, Madison, Wis.). In vitro-translated Rep78 was expressed from plasmid pRep(1-621), which has been described previously (39).

Helicase assays.

To generate partial duplex DNA substrate, a bacteriophage M13 gene III-specific oligodeoxyribonucleotide (5′-CAACAAATCGTTTTAGGGTAT-3′) was 5′ end labeled with [γ-³²P]ATP, hybridized to single-stranded M13mp18 DNA, and purified by Sephadex G-50 chromatography. A radiolabeled partial duplex substrate (approximately 5 fmol) was added to standard helicase reaction mixtures (18 μl) containing 30 mM Tris-HCl (pH 7.5), 4 mM MgCl₂, 25 mM NaCl, 1 mM dithiothreitol (DTT), 1 mM ATP, 0.1 mg of bovine serum albumin (BSA) per ml, and various amounts of MBP fusion protein. The reaction mixtures were incubated at 30°C for 30 min, and then reactions were terminated with 4 μl of gel loading buffer (100 mM EDTA, 1% sodium dodecyl sulfate, 50% glycerol, 0.1% bromophenol blue). Reaction products were resolved by electrophoresis through a nondenaturing 6% polyacrylamide gel in 1× Tris-borate-EDTA (TBE) buffer. The reaction products were visualized by autoradiography and quantified by PhosphorImager analysis (Molecular Dynamics) (except for experiments generating the results shown in Fig. 3A and C, which were quantified by scanning laser densitometry [Molecular Dynamics] of the exposed autoradiographs).

FIG. 3.

Properties of MBP-Rep52 helicase activity. Helicase assays were performed for 30 min at 30°C with 0.8 μg of MBP-Rep52 fusion protein, 5 fmol of ³²P-labeled partial duplex M13 substrate, and various concentrations of either ATP (A), NaCl (B), or MgCl₂ (C). All other components of the assay were as described in Materials and Methods.

To generate substrates used to determine the polarity of MBP-Rep52 helicase activity, a 19-nt oligodeoxyribonucleotide (5′-GCTGACAGGGACCCGAGTG-3′) was 5′ end labeled with [γ-³²P]ATP and hybridized to either 5′-TTTCACTCTAATTCTTGTCACTCGGGTCCCTGTCAGC-3′ (to generate the 5′-overhang substrate) or 5′-CACTCGGGTCCCTGTCAGCTTTCACTCTAATTCTTGT-3′ (to generate the 3′-overhang substrate). Standard helicase assays were performed with approximately 0.2 pmol of substrate DNA and the amounts of MBP-Rep52 fusion protein indicated in Fig. 5. Helicase products were resolved by electrophoresis on a 10% polyacrylamide gel in 1× TBE buffer.

FIG. 5.

Polarity of MBP-Rep52 helicase activity. (A) A 5′-end-labeled oligodeoxyribonucleotide (19-mer) was hybridized to each of two unlabeled oligodeoxyribonucleotides (37-mers) to generate helicase substrates bearing either 3′ or 5′ overhangs (diagramed at the bottom of each panel). Approximately 0.2 pmol of the labeled helicase substrate was incubated for 30 min at 30°C in the presence (0.1, 0.2, and 0.4 μg; lanes 3 to 5 and 8 to 10, respectively) or absence (lanes 1, 2, 6, and 7) of MBP-Rep52 fusion protein. Lanes marked “Heat” received reaction mixtures that had been heated to 95°C for 5 min to denature the helicase substrate prior to loading. The positions of the partial duplex substrate and displaced 19-mer are indicated. Reaction products were resolved on a 10% polyacrylamide gel. (B) The relative amount of ³²P-labeled fragment displaced was quantified by PhosphorImager analysis (Molecular Dynamics) and is graphed for each polarity substrate as a function of the amount of fusion protein added.

ATPase assays.

The production of inorganic phosphate from ATP was measured by a colorimetric assay (23). ATPase assays were performed in a 20-μl reaction volume containing 30 mM Tris-HCl (pH 7.5), 4 mM MgCl₂, 25 mM NaCl, 1 mM DTT, and 1 mM ATP. MBP fusion proteins were added in the amounts indicated in Fig. 4, and the reaction mixtures were incubated at 30°C for 30 min. Following incubation, 10-μl samples of each reaction mixture were transferred to the individual wells of a 96-well microtiter plate containing (per well) 200 μl of a 9:1 (vol/vol) mix of acid molybdate solution (Sigma Chemical Co., St. Louis, Mo.) and 0.1125% aqueous malachite green. Following a 1-min incubation, 25 μl of 34% (wt/vol) sodium citrate was added to each well. Color development was allowed to proceed for 20 min, after which time the absorbance at 660 nm was determined. The amount of inorganic phosphate generated was determined by comparison to a standard KH₂PO₄ curve.

FIG. 4.

ATPase activity of MBP-Rep52. The ability of MBP-Rep52 to hydrolyze ATP was determined by measuring the release of inorganic phosphate via a colorimetric assay (23). ATPase assays were performed in helicase assay buffer (without BSA) for 30 min at 30°C. Samples of each reaction mixture were transferred to a 96-well microtiter plate containing (per well) 200 μl of acid molybdate solution (Sigma)–0.1125% malachite green (9:1, vol/vol). Each well was quenched with 25 μl of 34% sodium citrate, and following a 20-min incubation period, the optical density at 660 nm (OD 600 nm) was determined. The amount of inorganic phosphate released was determined by comparison to a standard curve. Each assay was performed in duplicate.

In vitro translation and coimmunoprecipitation.

In vitro-translated Rep proteins were synthesized from DNA templates with a TNT-coupled reticulocyte lysate system (Promega) supplemented with MgCl₂ (to 1 mM) and 40 μCi of l-[³⁵S]methionine (>1,000 Ci/mmol; Amersham Life Science, Inc., Arlington Heights, Ill.). Coimmunoprecipitation experiments were performed as described previously (39). Briefly, in vitro translation products (10 μl) were diluted in 250 μl of binding buffer (phosphate-buffered saline, 0.5% Nonidet P-40, 1 mg of BSA per ml) and incubated with 1 μg of MBP-Rep or MBP-LacZα fusion protein for 30 min at 4°C. The samples then received 2 μl of an MBP-specific polyclonal rabbit antiserum (NEB) and were incubated an additional 30 min. Forty microliters of a buffer-equilibrated protein G-agarose bead slurry (Gibco/BRL; Grand Island, N.Y.) was then added to each tube. After an additional 30 min at 4°C, the agarose beads were washed extensively with binding buffer. Bound proteins were eluted from the pelleted beads by heating them to 95°C for 10 min in 30 μl of NuPAGE sample buffer (NOVEX, San Diego, Calif.). Eluted proteins were electrophoresed in a NuPAGE 4 to 12% Bis-Tris gel with 1× MES [2-(N-morpholino)ethanesulfonic acid]-sodium dodecyl sulfate running buffer (NOVEX). Radiolabeled proteins were visualized by fluorography.

Gel filtration chromatography.

In vitro-translation reaction mixtures programmed with a Rep52-encoding plasmid were diluted twofold with buffer A (30 mM Tris-HCl [pH 7.5], 100 mM NaCl) and supplemented with ATP and MgCl₂ to a 2 mM final concentration of each. The translation reaction mixtures were then incubated for 1 h at 4°C in the presence or absence of DNA effector molecules (single-stranded oligodeoxythymidylic acid [dT₅₀] or a 105-bp double-stranded oligonucleotide derived by PCR amplification of the pGEM-3Z polylinker with SP6 and T7 promoter-specific primers). The reaction mixtures were resolved by fast protein liquid chromatography on a prepacked HR 10/30 Superose 12 column (Pharmacia) equilibrated with buffer A (flow rate, 0.4 ml/min). Fractions were collected, and aliquots of each fraction were counted in a liquid scintillation counter. Fast protein liquid chromatography calibration standards were thyroglobulin (669 kDa), ferritin (440 kDa), catalase (232 kDa), BSA (67 kDa), and carbonic anhydrase (29 kDa).

RESULTS

MBP-Rep52 DNA helicase activity.

The P₁₉-promoted rep gene products, Rep52 and Rep40, contain amino acid sequence motifs associated with helicase and ATPase activities (10–13). To determine whether a representative P₁₉-promoted rep gene product can function as a helicase, Rep52 was expressed in E. coli as an MBP-Rep52 fusion protein, purified by amylose affinity chromatography, and tested for the ability to displace a 22-nt ³²P-labeled oligodeoxyribonucleotide hybridized to single-stranded M13 DNA. To ensure that enzymatic activity associated with purified MBP-Rep52 was not due to contamination with a copurifying bacterial enzyme, an MBP-Rep52 null mutant [termed MBP-Rep52(NTP)] bearing a histidine residue in place of lysine at position 116 (within a consensus NTP-binding motif [10]) was constructed and purified from E. coli for use as a negative control. The Rep52 K116H mutation directly corresponds to the K340H mutation of Rep68, which has been shown to abolish the helicase and ATPase activities of this P₅-promoted protein (5, 45, 46, 51). An additional negative control was provided by MBP-LacZα, which contains the MBP domain fused to a partial β-galactosidase polypeptide. An MBP-Rep78 chimera was employed as a positive control, since this protein has been shown to possess helicase activity in vitro (5). As shown in Fig. 1A, MBP-LacZα was devoid of helicase activity, yielding background levels of signal equivalent to that of a reaction mixture receiving no fusion protein (compare lanes 2 and 3). In contrast, incubation of MBP-Rep52 fusion protein with the partial duplex substrate resulted in a substantial amount of detectable helicase activity, unwinding approximately 60% of the substrate in this experiment (lane 4). As expected, the MBP-Rep52(NTP) mutant protein lacked helicase activity (lane 5), yielding levels of strand displacement equivalent to those of the MBP-LacZα and no-added-protein controls. Consistent with previous reports, MBP-Rep78 demonstrated a substantial level of helicase activity that is abolished by mutation of the conserved NTP-binding motif (compare lanes 6 and 7). Taken together, these results demonstrate that MBP-Rep52 possesses DNA helicase activity. The limit of detectable MBP-Rep52-mediated helicase activity was observed with between 60 and 120 ng of fusion protein added (Fig. 1B, lanes 4 and 5), and a linear dose-response relationship between fusion protein concentration and helicase activity was observed (Fig. 1B, lanes 4 to 8).

FIG. 1.

MBP-Rep52 has DNA helicase activity. Standard helicase assays were performed as described in Materials and Methods. Each reaction mixture received approximately 5 fmol of ³²P-labeled partial duplex M13 substrate. Reaction mixtures were incubated for 30 min at 30°C and resolved by electrophoresis on a nondenaturing 6% polyacrylamide gel. The positions of the partial duplex substrate and displaced oligonucleotide product are indicated. The heat-denatured substrate was heated to 95°C for 5 min prior to being loaded on the gel. The percentage of ³²P-labeled fragment displaced in each reaction was quantified by PhosphorImager analysis (Molecular Dynamics) and is indicated at the bottom of each lane. (A) Standard reaction mixtures received either no fusion protein (lanes 1 and 2) or 0.8 μg of the indicated affinity-purified fusion protein (lanes 3 to 7). MBP-Rep52 (NTP) and MBP-Rep78 (NTP) contain a substitution of histidine for lysine within an evolutionarily conserved NTP-binding motif (Walker A site [44]). (B) Effects of increasing amounts of MBP-Rep52 on strand displacement. Lane 1 received a heat-denatured control reaction mixture. Standard reaction mixtures that received increasing amounts of MBP-Rep52 fusion protein (0.0, 0.03, 0.06, 0.12, 0.25, 0.5, and 1.0 μg) were resolved in lanes 2 to 8, respectively.

Characterization of MBP-Rep52 DNA helicase activity.

Helicase-mediated DNA unwinding is typically coupled to the hydrolysis of a NTP cofactor in the presence of Mg²⁺ (or other) cations (24, 25, 28). To confirm the requirement of MBP-Rep52 for an NTP cofactor and to determine which NTPs serve the cofactor function most efficiently, the ability of MBP-Rep52 to unwind substrate DNA in the presence or absence of ATP, as well as other ribo- and deoxyribonucleoside 5′-triphosphates, was examined (Fig. 2A). In the absence of an NTP cofactor, MBP-Rep52-mediated helicase activity was not observed (lane 2). Addition of ATP to the assay, however, resulted in the induction of substantial levels of helicase activity (compare lanes 2 and 5). In addition to ATP, CTP and GTP could also function as helicase cofactors. UTP, however, was used poorly by MBP-Rep52. Certain deoxyribonucleotides were also effective cofactors of helicase activity. dATP served as efficiently as ATP in this experiment (although additional experimentation showed it to be slightly less active than rATP at inducing helicase activity). A relatively low level of helicase activity was sustained by dGTP, and no significant helicase activity was observed with a dCTP or dTTP cofactor.

FIG. 2.

MBP-Rep52 helicase activity is dependent upon NTP and divalent cation cofactors. (A) MBP-Rep52 can use a variety of NTP cofactors. Standard helicase assays (see Materials and Methods) were performed with 0.8 μg of MBP-Rep52 fusion protein (lanes 2 to 10) and a 1 mM final concentration of one of various NTP cofactors (indicated above each lane). Reaction products were resolved on a nondenaturing 6% polyacrylamide gel in 1× TBE running buffer. (B) MBP-Rep52 helicase activity requires metal cations and NTP hydrolysis. Helicase assay buffer (30 mM Tris-HCl [pH 7.5], 25 mM NaCl, 1 mM DTT, 0.1 mg of BSA per ml) was supplemented with 5 fmol of radiolabeled substrate DNA and either no fusion protein (lanes 1 and 2) or MBP-Rep52 fusion protein (0.8 μg; lanes 3 to 9) and the indicated NTP and/or metal chloride cofactor (at final concentrations of 1 and 4 mM, respectively). ATPγS, adenosine 5′-O-(3-thiotriphosphate), a nonhydrolyzable ATP analog.

To determine if MBP-Rep52 requires a divalent cation cofactor for helicase activity, helicase assays were performed in the presence or absence of Mg²⁺ or Mn²⁺ cations. As seen in Fig. 2B, incubation of MBP-Rep52 with helicase substrate in the presence of ATP without Mg²⁺ resulted in no detectable helicase activity (lane 3). Similarly, the addition of Mg²⁺ as the sole cofactor was not sufficient to support helicase activity. When both ATP and Mg²⁺ cofactors were present, however, helicase activity was readily observed (lane 5). Mn²⁺ was also found to serve as a cofactor, albeit less efficiently than Mg²⁺ (compare lanes 5 and 7). To determine whether NTP hydrolysis was necessary for helicase activity, the ability of MBP-Rep52 to function as a helicase in the presence of the nonhydrolyzable ATP analog adenosine 5′-O-(3-thiotriphosphate) (ATPγS) was determined. As shown in Fig. 2B (lanes 8 and 9), MBP-Rep52 functioned poorly as a helicase when adenosine 5′-O-(3-thiotriphosphate) was provided as the energy source, indicating that ATP hydrolysis is necessary for MBP-Rep52 helicase activity.

To characterize MBP-Rep52-mediated helicase activity further, helicase assays were performed under a variety of experimental conditions (Fig. 3). Levels of MBP-Rep52 helicase activity rose rapidly in response to increasing concentrations of ATP (Fig. 3A). A plateau of activity was reached at an ATP concentration of approximately 2.5 mM, above which levels of helicase activity gradually declined. The optimal NaCl concentration was between approximately 50 and 100 mM. Higher concentrations of NaCl significantly inhibited helicase activity (Fig. 3B). Helicase activity was found to rise rapidly in response to magnesium chloride, achieving maximum levels at a concentration of approximately 4 mM (Fig. 3C).

MBP-Rep52 ATPase activity.

All helicases examined in detail contain an intrinsic NTPase activity (25, 27, 28); therefore, we have examined directly the ability of MBP-Rep52 to function as an NTPase using a colorimetric assay to detect inorganic phosphate released from the NTP substrate during hydrolysis (23) (Fig. 4). MBP-LacZα and the NTP-binding site mutant MBP-Rep52(NTP) were used as negative controls. MBP-Rep52 demonstrated substantial levels of ATPase activity that increased with increasing amounts of fusion protein added, whereas MBP-Rep52(NTP) and MBP-LacZα were devoid of ATPase activity. The rate of ATP hydrolysis associated with MBP-Rep52 was approximately 0.2 nmol of ATP hydrolyzed per min per μg of fusion protein. Notably, MBP-Rep52 ATPase activity occurred in the absence of DNA effector molecules. The addition of single-stranded M13 DNA failed to stimulate MBP-Rep52 ATPase activity (data not shown), in contrast to what has been reported for MBP-Rep68Δ (51). Recently, Costello et al. (8) reported that a glutathione S-transferase–Rep52 fusion protein possesses an ATPase activity that is independent of the addition of single-stranded DNA; however, glutathione S-transferase–Rep52 ATPase activity was stimulated to some degree by the addition of the nonhistone chromosomal protein HMG1.

Polarity of MBP-Rep52 helicase activity.

The majority of DNA helicases prefer to initiate unwinding at the junction of single-stranded and duplex DNAs (25). The preference for a 5′ single-stranded flanking region or a 3′ single-stranded flanking region defines the helicases polarity and is believed to reflect the direction of translocation of the helicase along single-stranded DNA. We sought to determine the polarity of MBP-Rep52 helicase activity via the use of specific substrate DNAs bearing either 5′ or 3′ overhangs (Fig. 5A). The results of helicase assays in which each of the polarity substrates was incubated with increasing amounts of MBP-Rep52 fusion protein are shown in Fig. 5A and quantitatively graphed in Fig. 5B. These results indicate that MBP-Rep52 helicase activity possesses 3′-to-5′ polarity. Preliminary results with an MBP-Rep78 fusion protein indicate that the P₅-promoted Rep gene products also possess 3′-to-5′ polarity (data not shown).

trans dominance.

Rep78 and Rep68 mutants bearing a substitution of histidine for lysine within the conserved NTP-binding domain (Walker A site) have been reported to exhibit a trans-dominant inhibition of wild-type Rep78 and -68 function both in vitro and in vivo (4, 22, 33, 45, 51). One possible explanation for this phenotype is a physical interaction between wild-type and mutant proteins that results in the functional inactivation of a multimeric Rep protein complex. Indeed, Rep78 has been shown to self-associate both in vitro and in vivo and to form hexameric complexes in vitro (39). Therefore, it was of interest to determine whether an MBP-Rep52 NTP-binding-site mutant [MBP-Rep52(NTP)] demonstrated a trans-dominant negative effect, either in combination with MBP-Rep78 or with wild-type MBP-Rep52 (Fig. 6). As observed previously, MBP-Rep52 and MBP-Rep78 mutants bearing an NTP-binding-site mutation lacked helicase activity (Fig. 6A, lanes 2 and 3, respectively). In contrast, wild-type MBP-Rep78 was fully active, displacing 70% of labeled oligonucleotide from the partial duplex substrate (lane 4). Although introducing MBP-Rep78(NTP) fusion protein into the reaction mix in a 1:4 ratio (by weight) with wild-type MBP-Rep78 had little effect on strand displacement, the addition of MBP-Rep78(NTP) in a 1:2 ratio with wild-type MBP-Rep78 resulted in an approximately 40% decrease in MBP-Rep78 helicase activity (Fig. 6A, lane 6). Incubation of the two proteins in a 1:1 ratio resulted in a 64% decrease in activity (lane 7). In contrast, incubation of MBP-Rep78 with MBP-Rep52(NTP) at similar-weight ratios failed to inhibit MBP-Rep78 helicase activity (lanes 8 to 10).

FIG. 6.

trans-dominant inhibition. (A) Inhibition of MBP-Rep78-mediated helicase activity by NTP-binding-site mutant proteins. Standard helicase assays were performed with the amounts of fusion protein indicated below. Lane 1 received heat-denatured substrate DNA. Reaction mixtures with 0.8 μg of either MBP-Rep52(NTP), MBP-Rep78(NTP), or MBP-Rep78 fusion protein were loaded in lanes 2, 3, and 4, respectively. MBP-Rep78 (0.8 μg) was mixed with increasing concentrations of either MBP-Rep78(NTP) (0.2, 0.4, or 0.8 μg; lanes 5 to 7, respectively) or MBP-Rep52(NTP) (0.2, 0.4, or 0.8 μg; lanes 8 to 10, respectively) and incubated for 30 min at 30°C. Reaction products were separated on a 6% nondenaturing polyacrylamide gel. (B) Ability of MBP-Rep52(NTP) and MBP-Rep78(NTP) to inhibit MBP-Rep52 helicase activity. Heat-denatured partial duplex substrate is shown in lane 1. Reaction mixtures with 0.8 μg of either MBP-Rep52(NTP), MBP-Rep78(NTP), or MBP-Rep52 fusion protein were loaded in lanes 2, 3, and 4, respectively. Reaction mixtures with 0.8 μg of MBP-Rep52 and various concentrations of either MBP-Rep78(NTP) (0.2, 0.4, or 0.8 μg) or MBP-Rep52(NTP) (0.2, 0.4, or 0.8 μg) were loaded in lanes 5 to 7 and lanes 8 to 10, respectively.

The ability of an MBP-Rep52(NTP) mutant to inhibit the helicase activity of MBP-Rep52 was also tested (Fig. 6B). No trans-dominant inhibition of helicase activity was observed, even at a 1:1 wild-type/mutant weight ratio. Similarly, MBP-Rep78(NTP) exhibited no inhibition of MBP-Rep52 helicase activity (lanes 5 to 7). These results suggest that Rep52 does not participate in homo- or heterotypic Rep-Rep interactions. However, it is possible that the MBP-Rep52(NTP) mutant protein employed in these assays may be inactivated in some way that prevents interaction with wild-type Rep polypeptides. Therefore, as described below, the ability of Rep52 to oligomerize was addressed by other biochemical assays with other forms of Rep52.

Oligomerization.

To investigate further the potential interaction of Rep52 with itself or with Rep78, coimmunoprecipitation experiments were performed with MBP-Rep52 and MBP-Rep78 in conjunction with in vitro-translated, ³⁵S-labeled Rep52 and Rep78 (Fig. 7). In this assay, bacterially expressed MBP fusion proteins were incubated with radiolabeled, in vitro-translated Rep proteins and precipitated with an MBP-specific polyclonal rabbit antiserum bound to protein G-agarose beads. Samples of ³⁵S-labeled, in vitro-translated Rep78 and Rep52 are shown in lanes 1 and 2, respectively. It should be noted that in vitro-synthesized Rep78 and Rep52 migrate with apparent molecular weights that reflect their sequence-predicted molecular masses of 70.6 and 44.6 kDa, respectively. As previously reported (39), in vitro-translated Rep78 readily coprecipitated with bacterially expressed MBP-Rep78 (Fig. 7, lane 5). In contrast, no protein-protein interaction was detected between MBP-Rep78 and in vitro-translated Rep52 (lane 6). In a reciprocal experiment, in vitro-translated Rep78 and Rep52 proteins failed to coprecipitate with MBP-Rep52 (lanes 7 and 8, respectively). One possible explanation for the lack of detectable oligomerization of Rep52 is that the protein forms highly stable multimers shortly after protein synthesis, thereby precluding the subunit exchange necessary to detect protein-protein interactions by a functional trans-dominance assay or by coprecipitation. We therefore sought to determine the oligomerization state of native Rep52 by direct analysis using gel filtration chromatography. Cloned copies of the Rep52 gene were translated in vitro in the presence of [³⁵S]methionine, and the in vitro translation products were subjected to separation on a Superose 12 column. As shown in Fig. 8, Rep52 presented an elution volume between that of the BSA (67-kDa) and carbonic anhydrase (29-kDa) molecular mass standards, consistent with the monomeric form of the protein. It should be noted, however, that due to column resolution in this molecular mass range, the leading edge of the elution profile may potentially include some level of Rep52 dimers. Since multimerization of certain helicases is induced by interaction with single- or double-stranded DNA (25), we tested the ability of DNA substrates to induce Rep52 oligomerization. Incubation of in vitro-translated Rep52 with a 50-nt single-stranded oligodeoxyribonucleotide (dT₅₀) or with a 105-bp double-stranded DNA oligomer did not alter the elution profile of Rep52 (Fig. 8). It should be noted, however, that posttranslational modifications lacking in in vitro- or bacterially synthesized Rep52 may be required for efficient Rep52 multimerization.

FIG. 7.

Coimmunoprecipitation of bacterially expressed MBP-Rep fusion protein and in vitro-translated Rep polypeptides. Rep78 and Rep52 were synthesized in vitro in the presence of [³⁵S]methionine with a coupled in vitro transcription-translation system. Samples (1 μl) of the completed in vitro translation reaction mixtures were loaded in lanes 1 and 2. Ten microliters of each in vitro translation reaction mixture was diluted 25-fold in binding buffer (phosphate-buffered saline containing 0.5% Nonidet P-40 and 1 mg of BSA per ml) and mixed with 1 μg of the indicated MBP fusion protein. Following incubation, protein complexes were precipitated with anti-MBP polyclonal rabbit serum and protein G-agarose beads. Coprecipitating radiolabeled proteins were separated on a 4 to 12% NuPAGE gel (NOVEX) and visualized by fluorography. IVT, in vitro translation products; IP, immunoprecipitates.

FIG. 8.

Rep52 is a monomer in solution. Rep52 was translated in vitro in the presence of [³⁵S]methionine. In vitro translation reaction mixtures were diluted twofold with buffer A (30 mM Tris-HCl [pH 7.5], 100 mM NaCl). Following addition of MgCl₂ and ATP to a 2 mM final concentration of each, the diluted translation products were incubated for 1 h at 4°C in the absence (filled circles) or presence of DNA effector molecules: a single-stranded DNA oligonucleotide (dT₅₀; open circles) or a 105-bp double-stranded DNA oligonucleotide (filled triangles). After incubation, the in vitro translation products were subjected to gel filtration chromatography on a Superose 12 column equilibrated in buffer A. Samples of each fraction were counted in a liquid scintillation counter. The elution profiles of various molecular mass standards are indicated at the top. V, void volume; T, thyroglobulin (669 kDa); F, ferritin (440 kDa); Ct, catalase (232 kDa); B, BSA (67 kDa); C, carbonic anhydrase (29 kDa).

DISCUSSION

In this article, we reported that a bacterially expressed MBP-Rep52 fusion protein possesses DNA helicase and ATPase activities. The ability of MBP-Rep52 to function as a DNA helicase was strictly dependent upon the presence of NTP and divalent cation cofactors. However, a variety of ribo- and deoxyribonucleoside triphosphates, and at least two different divalent cations, could fulfill the cofactor requirement. The dependence of MBP-Rep52 upon NTP and divalent cation cofactors directly correlates with helicase-associated structural motifs identified within the amino acid sequence of Rep52. Of the seven helicase-associated consensus motifs deduced from sequence alignments of numerous prokaryotic and eukaryotic helicase proteins (10–12), the Rep52 gene product, as well as the remaining overlapping AAV Rep proteins, clearly possesses motif I (GXXXXGKT/S, where X represents any amino acid) and motif II (UUUU[D/E][D/E], where U represents any hydrophobic amino acid). These helicase-associated motifs occur within Rep52 between residues 110 and 117 and 150 and 155, respectively. The remaining helicase-associated consensus motifs are more difficult to identify; however, a probable motif VI sequence (HXXGKXXK, compared to the H/QXXGRXXR consensus sequence) occurs between Rep52 residues 232 and 239. This placement suggests that the Rep helicase domain is confined to an approximately 130-amino-acid region of the protein. Recent crystallographic analyses of the PcrA helicase of Bacillus stearothermophilus and the Rep helicase of E. coli (19, 43) have confirmed that motif I, also known as the Walker A site or P-loop, participates in NTP binding by interacting with the triphosphate moiety of the nucleotide cofactor. A second contribution to NTP binding is provided by a stacking interaction between the NTP base and a tyrosine residue within motif IV. With PcrA, it was suggested that the limited number of base- or ribose-specific interactions observed may explain the ability of PcrA to use a wide selection of NTPs as cofactors of helicase activity (43). A similar situation may apply to Rep52, which also uses a variety of NTP cofactors. Motif II, the Walker B site, is believed to coordinate a metal ion involved in NTP hydrolysis. NTP hydrolysis is believed to provide the energy needed to disrupt the hydrogen bonds of the DNA duplex during strand displacement and to promote conformational changes of the helicase molecule that drive its translocation along the DNA substrate.

The majority of helicases examined in detail are multimeric structures, typically dimers or hexamers (24). Recently, this observation has been extended to include Rep78 (39). Three lines of evidence, however, suggest that Rep52 fails to form stable homotypic interactions. These are (i) lack of detectable trans-dominant inhibition in vitro, (ii) the inability to coprecipitate in vitro-translated Rep52 with an MBP-Rep52 binding partner, and (iii) the elution of in vitro-translated Rep52 as a monomeric species during gel filtration chromatography (despite incubation with potential DNA effectors). Lack of Rep52 self-association is also consistent with results of studies mapping amino acid domains associated with the oligomerization of the larger P₅-promoted Rep78 protein, in that a short coiled-coil domain vital to efficient Rep-Rep interaction mapped within the amino-terminal region of Rep78 and is therefore lacking in Rep52 and -40 (39). These data also suggest that there is little cross-talk between the P₅- and P₁₉-promoted Rep proteins. It is possible, however, that Rep52 is not monomeric under all conditions and that some level of oligomerization may exist that is below the detection limits of the assays employed in this study.

With the identification of Rep52 as a DNA helicase, it becomes apparent that the AAV rep gene encodes two families of overlapping helicase proteins derived from the same ORF. An obvious question concerns the purpose of expressing multiple forms of the rep-encoded helicase. Presumably the P₅- and P₁₉-promoted helicases perform distinct functions during viral replication. Indeed, genetic analyses of an AAV-2 mutant defective in the expression of the P₁₉-promoted rep gene products, Rep52 and Rep40, indicated that these proteins play a role in the accumulation of viral single-stranded genomic DNA and the subsequent production of infectious AAV virions (3). In contrast, the P₅-promoted rep gene products, Rep78 and Rep68, were found to be involved in the production of double-stranded replicative-form AAV DNA intermediates (3). Thus, one function of the P₁₉-promoted Rep52 and -40 proteins may be to maintain single-stranded, unit-length viral genomes for encapsidation. Alternatively, the P₁₉-promoted Rep proteins may actively unwind duplex viral genomes concomitantly with packaging of the progeny strand into preformed, or possibly nucleating, capsid structures. Further investigation will be required to define the precise role of the P₁₉-promoted Rep proteins in AAV replication.