Biochemical and genetic analysis of the vaccinia virus d5 protein: Multimerization-dependent ATPase activity is required to support viral DNA replication

Abstract

The vaccinia virus-encoded D5 protein is an essential ATPase involved in viral DNA replication. We have expanded the genotypic and phenotypic analysis of six temperature-sensitive (ts) D5 mutants (Cts17, Cts24, Ets69, Dts6389 [also referred to as Dts38], Dts12, and Dts56) and shown that at nonpermissive temperature all of the tsD5 viruses exhibit a dramatic reduction in DNA synthesis and virus production. For Cts17 and Cts24, this restriction reflects the thermolability of the D5 proteins. The Dts6389, Dts12, and Dts56 D5 proteins become insoluble at 39.7 degrees C, while the Ets69 D5 protein remains stable and soluble and retains the ability to oligomerize and hydrolyze ATP when synthesized at 39.7 degrees C. To investigate which structural features of D5 are important for its biological and biochemical activities, we generated targeted mutations in invariant residues positioned within conserved domains found within D5. Using a transient complementation assay that assessed the ability of D5 variants to sustain ongoing DNA synthesis during nonpermissive Cts24 infections, only a wtD5 allele supported DNA synthesis. Alleles of D5 containing targeted mutations within the Walker A or B domains, the superfamily III helicase motif C, or the AAA+ motif lacked biological competency. Furthermore, purified preparations of these variant proteins revealed that they all were defective in ATP hydrolysis. Multimerization of D5 appeared to be a prerequisite for enzymatic activity and required the Walker B domain, the AAA+ motif, and a region located upstream of the catalytic core. Finally, although multimerization and enzymatic activity are necessary for the biological competence of D5, they are not sufficient.

FIG. 1.

Predicted sequence of the D5 protein and identification of conserved domains and sites affected in temperature-sensitive and site-directed mutants. The predicted amino acid sequence of the vaccinia virus D5 protein (Western Reserve strain; protein database entry P04305) is shown. The residues altered in the various tsD5 strains are circled; those shown in gray circles have been previously mapped (T₁₄₃I in Cts24, S₁₆₁F in Cts17, and P₆₈₂S in Ets69) (11). The residues shown in white type within black circles indicate lesions that were mapped in this study: M₁₁₆I in Dts12, V₂₁₂A in Dts56, and S₁₆₁F and A₂₈₃T in Dts6389. The four conserved motifs identifying D5 as a member of the SFIII helicase subgroup of the AAA+ family of proteins (17) are boxed. These motifs include a Walker A box/P-loop (white box, black type), a Walker B box (light gray box, black type), the motif C of SFIII proteins (dark gray box, white italic type), and the AAA+ family motif (black box, white type). The invariant residues targeted for site-directed mutagenesis within these motifs are underlined (K₅₀₉A, E₅₅₇Q, N₆₀₅D, and R₆₁₉A,R₆₂₀A). The inverted triangles mark the beginning of the 301-D5 (filled) or 413-D5 (empty) fragment. While the majority of the ts lesions map to the N-terminal portion of the protein, the C-terminal portion of D5 contains the putative enzymatic/functional domains.

FIG. 2.

The tsD5 viruses exhibit a temperature-sensitive reduction in the production of infectious virus during a single round of infection. Confluent 35-mm dishes of BSC40 (A) or L929 (B) cells were infected with the WR (wt, Cts17, Cts24, and Ets69) (left graphs) and IHD-W (wt, Dts12, Dts56, and Dts6389) (right graphs) strains at an MOI of 2 and incubated for 24 h at either 31.5°C or 39.7°C. The yield of cell-associated virus was quantitated by titration on BSC40 cells at 31.5°C. The dashed line denotes the amount of infectious virus made during a wt infection at 31.5°C. Each infection was performed in duplicate, and each assay was titrated in duplicate. Error bars denote standard deviations.

FIG. 3.

tsD5 viruses show a defect in viral DNA synthesis at high temperature. Confluent 35-mm dishes of BSC40 (A) or L929 (B) cells were infected with the WR (left plots) or IHD (right plots) strain at an MOI of 5 and maintained at 31.5°C or 39.7°C. Individual cultures were harvested at 3, 6, 9, 12, or 24 hpi, and the levels of accumulated viral DNA were determined by Southern dot blot hybridization. The average of the data obtained from a representative experiment, which was analyzed in quadruplicate, is shown with the standard deviation. The legends are shown at the bottom of the figure; note that the scales for the y axes differ among the panels.

FIG. 4.

Analysis of accumulation and solubility of tsD5 proteins during infections performed at permissive and nonpermissive temperatures. BSC40 (A) or L929 (B) cells were either mock infected or infected with the various WR and IHD strains at an MOI of 10 and maintained at either 31.5°C or 39.7°C for 8 h prior to harvesting. Whole-cell lysates were subjected to immunoblot analysis with an anti-D5 antiserum; the level of D5 in each sample was normalized to the amount of D5 that accumulated during infection with the appropriate wt strain at 31.5°C. Infections were performed in duplicate, and each infected cell lysate was analyzed in duplicate. A representative blot is shown. (C) BSC40 cells were infected with the various IHD viruses at an MOI of 10 and maintained at either the permissive (odd lanes) or nonpermissive (even lanes) temperature for 8 h. Postnuclear supernatants were prepared and analyzed by immunoblotting with the anti-D5 antiserum; the boxed numbers below the lanes represent the relative levels of the tsD5 proteins compared to the amount of wtD5 that accumulated at the respective temperature.

FIG. 5.

Comparative analysis of the ATP binding and hydrolysis activities of wt and mutant 3XFLAG-D5 proteins. (A) 3XFLAG-D5 proteins were purified to near homogeneity. Cells were infected so as to overexpress the various 3XFLAG-D5 proteins at permissive temperature. The D5 proteins were affinity purified from lysates prepared at 24 hpi; a silver-stained gel of the purified proteins is shown in panel A (for each preparation, 1/10 and 1/20 of the yield from 1 × 10⁷ cells are shown in adjacent lanes). The black arrowhead indicates the 3XFLAG-D5 proteins; protein standards are indicated on the left, with their masses shown in kilodaltons. (B) tsD5 proteins retain near-wt activity for ATP binding and ATP hydrolysis. ATP binding and ATP hydrolysis assays were carried out as described in Materials and Methods. The average of the data obtained from several replicates is shown with the standard deviation.

FIG. 6.

Transient expression of wtD5 enables sustained viral DNA synthesis when Cts24 infections are shifted to 39.7°C in the midst of DNA replication. (A) Transient expression of wtD5, but not Cts24 D5, rescues viral DNA synthesis. BSC40 cells were infected with Cts24 at an MOI of 5 at 31.5°C; at 3 hpi, cells were mock transfected (no DNA) or transfected with supercoiled empty vector (pInt) or pInt:D5 constructs. At 5 hpi, cells were shifted to 39.7°C for the remainder of the experiment. At 5, 8, and 24 hpi, cells were harvested, and the levels of accumulated viral DNA were quantitated by Southern dot blot hybridization; the analysis was performed in quadruplicate, and the average value is shown with error bars. The inset shows an immunoblot that confirms the accumulation of the D5 proteins. (B) D5 alleles containing targeting mutations within conserved motifs are unable to rescue DNA replication during Cts24 infections. The transient complementation assay described for panel A was applied to the analysis of pInt:D5 constructs encoding D5 proteins containing the indicated mutations.

FIG. 7.

Biochemical analysis of site-directed D5 mutants. (A) Expression and purification of 3XFLAG-D5 proteins containing site-directed mutations in conserved motifs. The 3XFLAG-D5 proteins with site-directed mutations were expressed, purified, and visualized as described for Fig. 5A. The black arrowhead indicates the 3XFLAG-D5 proteins; protein standards are indicated on the left. (B) ATP binding activity of the site-directed mutated D5 proteins. An ∼150-ng aliquot of the 3XFLAG-D5 proteins, or 1.5 μg of the Klenow fragment of E. coli DNA polymerase I, was incubated with [α-³²P]dATP on ice and UV irradiated (or left on ice without UV exposure [lane 2]). The samples were resolved by electrophoresis on an SDS-10% acrylamide gel, and bound [α-³²P]dATP was visualized by autoradiography. (C) Mutagenesis of the invariant residues within the conserved D5 motifs ablates ATPase activity. Reaction mixtures containing increasing concentrations of 3XFLAG-D5 proteins (wt at 30, 75, 150, and 300 ng/reaction mixture [lanes 2 to 5]; all others were at 150 and 300 ng/reaction mixture [lanes 6 to 13]) or no protein (lane 1) and 1 mM [α-³²P]dATP were analyzed by ascending chromatography. ADP and ATP were visualized by autoradiography. (D) Quantitative summary of the ATP binding and hydrolysis activity of the wt and mutant D5 proteins. The average activity obtained for each protein preparation was calculated after phosphorimager analysis of replicate experiments and is shown with the standard deviation.

FIG. 8.

Participation of conserved family domains in D5 oligomerization. (A) The site-directed D5 mutants retained the ability to interact with wt D5. BSC40 cells were coinfected with the vTF7.5 and vTMD5 (vD5) viruses (lane 2 received only vTF7.5 infection) each at an MOI of 2. At 3 hpi, cells were left untransfected (lane 1, no DNA) or transfected with supercoiled pTM1-3XFLAG:D5 constructs. The 3XFLAG proteins and any tightly associated untagged D5 were affinity purified on anti-FLAG resin. The eluates were analyzed by fractionation on an SDS-8% acrylamide gel; only the relevant portion of the silver-stained gel is shown. (B) Some of the site-directed mutants showed a decreased ability to self-associate. BSC40 cells were infected with vTF7.5 and, at 3 hpi, cotransfected with supercoiled pTM-1 containing the untagged and 3XFLAG-tagged versions of the same (lanes 8 to 12) or different (lanes 13 and 14) D5 alleles. A portion of the clarified cytoplasmic lysates was removed prior to affinity purification (input, top panel); the remainder was used for affinity purification of the 3XFLAG-D5 and any associated untagged D5 (eluate, bottom panel). Input and eluate samples were resolved electrophoretically; the input samples were subjected to immunoblot analysis with an anti-D5 antiserum, and the eluates were visualized by silver staining. (C) Graphic presentation of the relative efficiency of copurification of untagged D5 with 3XFLAG-D5 protein. The data shown in panel B were quantitated and used to obtain a ratio of the relative efficiency of pull down. For each sample, the ratio within the eluates of the untagged D5 to the 3XFLAG-tagged D5 was compared to the ratio of untagged D5 to tagged D5 present in the input sample. The average of replicate experiments is shown with the standard deviation.

FIG. 9.

Preliminary mapping of the oligomerization/enzymatic domain. (A) A domain located between amino acids 301 and 413 is required for stoichiometric association with full-length wtD5. Cells were programmed to overexpress a full-length untagged version of D5 in the context of full-length 3XFLAG-tagged D5 (lane 1) or N-terminal truncation D5 mutants 3XFLAG-301-D5 (lane 2) or 3XFLAG-413-D5 (lane 3). Affinity purification of the tagged protein and any associated, untagged D5 was performed as previously described; the eluates were resolved electrophoretically and visualized by silver staining. While both the full-length 3XFLAG-wtD5 (black arrowhead) and the 3XFLAG-301-D5 (white oval) were able to retrieve full-length untagged D5 (gray arrow), the 3XFLAG-413-D5 (white arrowhead) was deficient in this activity. (B) Expression and purification of wt and N-terminal-truncated 3XFLAG-tagged D5 proteins. The 3XFLAG-wt, 3XFLAG-301-D5, and 3XFLAG-413-D5 proteins were expressed and purified, and increasing amounts of the purified preparations were resolved electrophoretically and visualized by silver staining. These affinity-purified proteins were used in biochemical analyses. (C) Both N-terminal truncation mutants retain ATP binding activity. ATP binding studies were conducted as described previously. The black arrowhead marks the migration of full-length 3XFLAG-D5, the white oval marks the migration of the 3XFLAG-301-D5 protein, and the white arrowhead marks the migration of the 3XFLAG-413-D5 protein. (D) Only the 301-D5 protein, which is capable of multimerization, retains enzymatic activity. Reaction mixtures containing either no protein (lane 1) or 150 to 300 ng of the 3XFLAG-wt (lanes 2 and 3), 3XFLAG-301-D5 (lanes 4 and 5), or 3XFLAG-413-D5 (lanes 6 and 7) proteins were incubated with [α-³²P]dATP, and the conversion of ATP to ADP was monitored by ascending chromatography. (E) The 413-D5 protein lacks intrinsic ATPase activity which cannot be restored by the addition of exogenous nucleic acid. Reaction mixtures containing either no protein (lanes 1) or ∼150 ng of the 3XFLAG-wt or 3XFLAG-413-D5 proteins were incubated with [α-³²P]dATP in the absence of DNA (lanes 2) or in the presence of 250 (lanes 3, 5, and 7) or 500 (lanes 4, 6, and 8) ng of DNA. ssM13mp10 was the source of single-stranded DNA (ss-DNA; lanes 3 and 4), salmon sperm DNA was the source of double-stranded DNA (ds-DNA; lanes 5 and 6), and activated salmon sperm was the source of nicked double-stranded DNA (lanes 7 and 8). (F) Neither the 301- nor the 413-D5 protein can rescue viral DNA synthesis during nonpermissive Cts24 infection. The transient complementation assay previously described was applied to pInt constructs expressing either 301-D5 or 413-D5; neither protein retained the ability to restore DNA synthesis.