Simian virus 40 Vp1 DNA-binding domain is functionally separable from the overlapping nuclear localization signal and is required for effective virion formation and full viability

Abstract

A DNA-binding domain (DBD) was identified on simian virus 40 (SV40) major capsid protein Vp1, and the domain's function in the SV40 life cycle was examined. The DBD was mapped by assaying various recombinant Vp1 proteins for DNA binding in vitro. The carboxy-terminal 58-residue truncated Vp1DeltaC58 pentamer bound DNA with a K(d) of 1.8 x 10(-9) M in terms of the protein pentamer, while full-length Vp1 and carboxy-terminal-17-truncated Vp1DeltaC17 had comparable apparent K(d)s of 5.3 x 10(-9) to 7.3 x 10(-9) M in terms of the protein monomers. Previously identified on Vp1 was a nuclear localization signal (NLS) consisting of two N-terminal basic clusters, NLS1 (4-KRK-6) and NLS2 (15-KKPK-18). Vp1DeltaC58 pentamers harboring multiple-point mutations in NLS1 (NLSm1), NLS2 (NLSm2), or both basic clusters (NLSm1. 2) had progressively decreased DNA-binding activity, down to 0.7% of the Vp1DeltaC58 level for NLSm1. 2 Vp1. These data, along with those of N-terminally truncated proteins, placed the DBD in overlap with the bipartite NLS. The role of the Vp1 DBD during infection was investigated by taking advantage of NLS phenotypic complementation (N. Ishii, A. Nakanishi, M. Yamada, M. H. Macalalad, and H. Kasamatsu, J. Virol. 68:8209-8216, 1994), in which an NLS-defective Vp1 could localize to the nucleus in the presence of wild-type minor capsid proteins Vp2 and Vp3. This approach made it possible to dissect the role of the bifunctional Vp1 NLS-DBD in virion assembly in the nucleus. Mutants of the viable nonoverlapping SV40 (NO-SV40) DNA NLSm1, NLSm2, and NLSm1. 2 replicated normally following transfection into host cells and produced capsid proteins at normal levels. All mutant Vp1s were able to interact with Vp3 in vitro. The mutants NLSm1 and NLSm1. 2 were nonviable, and the mutant Vp1s unexpectedly failed to localize to the nucleus though Vp2 and Vp3 did, suggesting that the mutated NLS1 acted as a dominant signal for the cytoplasmic localization of Vp1. Mutant NLSm2, for which the mutant Vp1's nuclear localization defect was complemented by Vp2 and Vp3, displayed a 5,000-fold reduced viability. Analysis of NLSm2 DNA-transfected cell lysate revealed a 10-fold reduction in the level of DNase I-protected viral DNA, and yet virion-like particles were found among the DNase I-resistant material. Collective results support a role for Vp1 NLS2-DBD2 in the assembly of virion particles. The results also suggest that this determinant can function in the infection of new cells.

FIG. 1

DNA binding by recombinant Vp1 proteins. (A and B) Southwestern blots. Protein samples were resolved on SDS–10% polyacrylamide gels and either stained with Coomassie blue (A, lanes 1 and 2; B, lanes 1 to 5) or electrotransferred to nitrocellulose and probed with nick-translated ³²P-labeled SV40 DNA (10 to 12 ng/ml) (A, lanes 3 and 4; B, lanes 6 to 10). In panel A, 0.5 μg of Vp1 (Vp1∗, lanes 1 and 3) as a cleavage product of GST-Vp1, or 0.5 μg of Vp1ΔC17 (lanes 2 and 4), were used. Intact GST-Vp1 (GST-Vp1) and the GST moiety (GST∗) were also present in the samples of lanes 1 and 3. In panel B, 2 μg of Vp1ΔC17 (lanes 1 and 6) or of the following N-terminal mutant ΔC17 proteins were used: NLSm2 (lanes 2 and 7), NLSm1 · 2 (lanes 3 and 8), ΔN(2–21) (lanes 4 and 9), and ΔN(2–51) (lanes 5 and 10). Four bars to the left of each Coomassie gel mark the positions (from top to bottom) for molecular mass standards of 110, 74, 45, and 26 kilodaltons. (C) Isolation of pentameric Vp1ΔC58 and ΔN(2–21)-ΔC58. Sedimentation through sucrose gradients was performed as described in Materials and Methods, and an aliquot from each of the 15 fractions was analyzed by SDS-PAGE and Coomassie blue staining. In the profiles shown, twice as much protein was sedimented for ΔN(2–21)-ΔC58 (lower panel) than for Vp1ΔC58 (upper panel). Six bars to the left of each gel mark the positions for six molecular mass standards of 100, 71, 44, 28, 19, and 14 kilodaltons. Pentamers were found in fractions 8 and 9 for Vp1ΔC58 and in fractions 8 through 10 for ΔN(2–21)-ΔC58. NLSm1ΔC58, NLSm2ΔC58, and NLSm1 · 2ΔC58 gave sedimentation profiles similar to that of Vp1ΔC58. (D) Solution-phase DNA binding. Filter-binding assays were performed by incubating various concentrations of each protein with ³²P-labeled DNA, and the percentages of the input radiolabel that was retained on nitrocellulose membrane upon filtration were determined. Average values from three experiments are shown with error bars for the binding of a 326-bp PCR-derived SV40 fragment by pentameric Vp1ΔC58 or its N-terminal mutant derivatives. Values from one experiment are shown for the binding of nick-translated SV40 DNA by ΔN(2–51)-ΔC17 whose monomeric concentrations are given in parentheses. Dissociation constants (K_ds) were determined as molar protein concentrations at 50% DNA retention. (E) Summary of DNA-binding activities for recombinant Vp1s. For GST-Vp1-derived Vp1 and for Vp1ΔC17, the apparent K_d was given in protein monomer concentration for the binding of nick-translated SV40 (SV) or pBR322 (pBR) DNA. For ΔC58 proteins, an average K_d along with the standard deviation was given in the protein pentamer concentration for the binding of a 326-bp SV40 fragment. The relative activity is the reciprocal of K_d made relative to that of the Vp1ΔC58 K_d, which was taken to be 100%. An “X” on the schematic protein diagram represents the mutation of an N-terminal basic cluster; a dot beneath residues 104 and 254 indicates their mutation from cysteines into alanines, n.d., not done.

FIG. 2

Subcellular localization of NLS-mutant Vp1ΔC58-GFP proteins. Cells transfected with wild-type (a), NLSm1 (b), NLSm2 (c), and NLSm1.2 mutant pSG5-Vp1ΔC58-GFP were fixed and stained with guinea pig anti-Vp1, followed by rhodamine-labeled anti-guinea pig antibody. Photographs of the rhodamine fluorescence are shown. The GFP autofluorescence in each case gave an essentially identical pattern, although the intensity was less than the corresponding Vp1 immunofluorescence.

FIG. 3

DNA replication and capsid protein production by N-terminal Vp1 mutants. (A) Time course of viral DNA replication. Cells transfected with each NO-SV40 DNA were harvested at the indicated time points, and the total intracellular viral DNA was extracted and probed with nick-translated SV40 DNA in a Southern slot blot. The radioactivities of the individual slots were counted and plotted against time. (B) Levels of capsid proteins. Cells transfected with individual NO-SV40 DNAs were analyzed by immunoblotting with anti-Vp1 (upper panel) or anti-Vp3 (lower panel) antibodies. The amount of cells analyzed was adjusted for transfection efficiency as measured by the activity of β-galactosidase expressed from pmiwZ, which was cotransfected with each NO-SV40 DNA. Bands corresponding to Vp1, Vp2, and Vp3 are indicated at the right.

FIG. 4

Subcellular localization of capsid proteins expressed from Vp1 N-terminal-mutant NO-SV40 DNAs. Cells were nuclearly microinjected with individual NO-SV40 DNAs, cultured for 24 h, fixed, and doubly stained with guinea pig anti-Vp1 (a, c, e, and g) and rabbit anti-Vp3 (b, d, f, and h), followed by rhodamine-labeled (a, c, e, and g) or fluorescein-labeled (b, d, f, and h) secondary antibodies. The same cells are shown in panels a and b, c and d, e and f, and g and h.

FIG. 5

Interaction of N-terminal-mutant Vp1s with Vp3. In vitro-transcribed and -translated, ³⁵S-labeled Vp1 from each pBS-Vp1 DNA was mixed with the resin to which either GST-Vp3 (first panel) or GST (second panel) had been immobilized. The resin-bound proteins were analyzed by SDS-PAGE and fluorography. The amounts of input ³⁵S-labeled Vp1s used for the pull-down experiments are also shown (third panel).

FIG. 6

Virion particle formation by NO-SV40-NLSm2. (A and B) Sonicated lysate prepared from wild-type (Wt, A) or mutant NLSm2 (NLSm2, B) NO-SV40 DNA transfected cells was treated with DNase I, and aliquots containing equivalent amounts of DNase I-resistant viral DNAs (60 μl for wild-type raised to 600 μl with buffer and 600 μl for NLSm2) were sedimented through 5 to 32% sucrose gradients and fractionated from the bottom into 17 fractions. Five-sixths of each fraction was analyzed for viral DNA by Southern blot (upper panels). One-sixth of each wild-type fraction or one-thirtieth of each NLSm2 fraction was also analyzed for Vp1 by Western blot (lower panels). An arrowhead points to fraction 4, the peak fraction for viral DNA and Vp1 from purified virions sedimented in a parallel gradient. (C) Distribution of post-DNase I viral DNA and Vp1 in sucrose fractions. Wild-type (upper plot) and NLSm2 (lower plot). For each Southern or Western profile, the radioactivity per lane of the DNase I-resistant viral DNA or Vp1, obtained by phosphorimager quantitation of the observed band(s), was made relative to the total radioactivity of the DNA or Vp1 for all 17 fractions, which was taken to be 100%. The percentages were plotted against the fraction number.