The herpes simplex virus type 1 cleavage/packaging protein, UL32, is involved in efficient localization of capsids to replication compartments

Abstract

Six genes, including UL32, have been implicated in the cleavage and packaging of herpesvirus DNA into preassembled capsids. We have isolated a UL32 insertion mutant which is capable of near-wild-type levels of viral DNA synthesis; however, the mutant virus is unable to cleave and package viral DNA, consistent with the phenotype of a previously isolated temperature-sensitive herpes simplex virus type 1 mutant, tsN20 (P. A. Schaffer, G. M. Aron, N. Biswal, and M. Benyesh-Melnick, Virology 52:57-71, 1973). A polyclonal antibody which recognizes UL32 was previously used by Chang et al. (Y. E. Chang, A. P. Poon, and B. Roizman, J. Virol. 70:3938-3946, 1996) to demonstrate that UL32 accumulates predominantly in the cytoplasm of infected cells. In this report, a functional epitope-tagged version of UL32 showed that while UL32 is predominantly cytoplasmic, some nuclear staining which colocalizes with the major DNA binding protein (ICP8, UL29) in replication compartments can be detected. We have also used a monoclonal antibody (5C) specific for the hexon form of major capsid protein VP5 to study the distribution of capsids during infection. In cells infected with wild-type KOS (6 and 8 h postinfection), 5C staining patterns indicate that capsids are present in nuclei within replication compartments. These results suggest that cleavage and packaging occur in replication compartments at least at 6 and 8 h postinfection. Cells infected with the UL32 mutant exhibit a hexon staining pattern which is more diffusely distributed throughout the nucleus and which is not restricted to replication compartments. We propose that UL32 may play a role in "bringing" preassembled capsids to the sites of DNA packaging and that the failure to localize to replication compartments may explain the cleavage/packaging defect exhibited by this mutant. These results suggest that the UL32 protein is required at a step distinct from those at which other cleavage and packaging proteins are required and may be involved in the correct localization of capsids within infected cells.

FIG. 1

Physical map of the region of the HSV genome containing the UL32 gene. The region of HSV-1 DNA spanning the UL32 and UL33 open reading frames has been expanded to show the cleavage sites for BamHI (B), SalI (S), XbaI (X), NcoI (N), HpaI (H), and BstEII (E) and the start (nt 69162) and the stop (nts 67374 and 69625) sites for UL32 and UL33, respectively. On the last line is a diagram of the insertion of the ICP6::lacZ cassette in UL32 at the BamHI site (nt 68338). ▨, ICP6 promoter. U_L, unique long region; U_S, unique short region.

FIG. 2

Immunoblot of extracts from KOS- and hr64-infected cells. Monolayers of Vero and 158 cells were infected with either KOS, hr74, or hr64 (5 PFU/cell) or were mock infected. Extracts were prepared at 18 h postinfection as described under Materials and Methods and resolved by SDS-PAGE. Proteins were revealed by Western blotting with a UL32 polyclonal antibody and alkaline phosphatase-conjugated secondary antibodies. Lane 1, mock-infected Vero cell extracts; lane 2, KOS-infected Vero cell extracts; lane 3, hr64-infected Vero cell extracts; lane 4, hr74-infected Vero cell extracts; lane 5, mock-infected 158 cell extracts; lane 6, hr64-infected 158 cell extracts.

FIG. 3

Southern blot analysis of total DNA from KOS-, hr74-, and hr64-infected Vero cells. (A) Vero cells were infected with the indicated virus at 5 PFU per cell. At 18 h postinfection, DNA was isolated as described under Materials and Methods and digested with BamHI. The blotted DNA was hybridized with a ³²P-labeled probe containing the BamHI SQ fragment. SQ represents viral DNA junctions, and S and Q represent viral DNA termini. Lanes 1, 3, and 5, total DNA from KOS-, hr74-, and hr64-infected Vero cells, respectively; lanes 2, 4, and 6, total DNA from KOS-, hr74-, and hr64-infected 158 cells, respectively. (B) Vero cells were infected at 5 PFU per cell. At 18 h postinfection the cells were collected in low-melting-point agarose. Agarose blocks were introduced into the well of a 1.3% pulsed-field gel. The complex high-molecular-weight DNA which does not enter the gel is labeled “well,” and unit-length monomeric DNA is labeled “152Kb.” Lane 1, total DNA from KOS-infected Vero cells; lane 2, total DNA from hr64-infected Vero cells; lane 3, total DNA from hr64-infected 158 cells.

FIG. 4

Southern blot analysis of DNase-treated DNA from KOS- and hr64-infected Vero cells. Total DNA extracts from KOS or hr64 were treated with DNase I and subsequently digested with BamHI as described under Materials and Methods. The samples were collected at the end of a 1-h incubation time (time zero) and at 2, 4, 6, 8, and 16 h postinfection.

FIG. 5

UL32 can be detected in KOS-infected cell extracts but not in B capsid or virions. B capsids from KOS-infected Vero cells were collected from a sucrose gradient and purified on a second gradient as described under Materials and Methods. Monolayers of Vero cells were infected with either KOS or hr64 (5 PFU/cell) or were mock infected. Extracts were prepared at 18 h postinfection as described under Materials and Methods. Virions were prepared as described under Materials and Methods. Proteins were revealed by ECL Western blotting with UL32 polyclonal antibodies (13) and horseradish peroxidase-labeled secondary antibodies. Lane 1, B capsids from KOS-infected Vero cells; lane 2, mock-infected Vero cell extracts; lane 3, KOS-infected Vero cell extracts; lane 4, hr64-infected Vero cell extracts; lane 5, virion from KOS-infected Vero cells.

FIG. 6

UL6 and UL25 are present in B capsids from wild-type- and hr64-infected cells. The presence or absence of UL6 and UL25 in KOS- and hr64-infected Vero cell extracts, in B capsids from KOS- and hr64-infected cells, and in virions from KOS-infected Vero cells was determined as described under Materials and Methods. Proteins were revealed by Western blotting with α-UL6 antibodies and alkaline phosphatase-conjugated secondary antibodies (A) or by ECL Western blotting with a UL25 polyclonal antibody and horseradish peroxidase-labeled secondary antibodies (B). Panel A lanes: 1, B capsids from KOS-infected Vero cells; 2, B capsids from hr64-infected Vero cells; 3, mock-infected Vero cell extracts; 4, KOS-infected Vero cell extracts; 5, hr74-infected Vero cell extracts; 6, virions from KOS-infected Vero cells (this lane was developed for a longer time to visualize the band). Panel B lanes: 1, mock-infected Vero cell extracts; 2, KOS-infected Vero cell extracts; 3, B capsids from KOS-infected Vero cells; 4, B capsids from hr64-infected Vero cells; 5, extracts from Vero cells infected with virus lacking UL25; 6, virions from KOS-infected Vero cells.

FIG. 7

Localization of UL32 in cells transfected with pAPVEEUL32. Cells shown in panels A to C were transfected with pAPVUL32 and pVP16 as described under Materials and Methods. Cells shown in panels D to F were transfected with pAPVEEUL32 and superinfected with hr64. In panels A and B, green represents staining with the EE monoclonal antibody and red represents staining with the UL32 polyclonal antibody, respectively. In panel D green represents staining with the EE monoclonal antibody, and in panel E red represents staining for ICP8. Panels C and F show the merged images of the staining patterns. Bars = 10 μm.

FIG. 8

Replication compartments and capsid localization. Vero cells were infected with KOS (A to C and J), hr74 (D to F), or hr64 (G to I and K) for 6 h and stained with anti-ICP8 polyclonal antibodies (A, D, and G) and 5C monoclonal antibodies (B, E, and H). Panels C, F, and I represent the merged images of staining patterns with ICP8 and 5C antibodies. Panels J and K represent three-dimensional reconstructions from Z series images of KOS- and hr64-infected Vero cells obtained with the Voxel View program as described under Materials and Methods. In a control experiment, mock-infected cells were treated with primary and secondary antibodies, and infected cells were treated with the secondary antibodies alone; in neither case was any cross-reactivity for anti-ICP8 or monoclonal antibody 5C observed (data not shown). Bars = 10 μm.