Induction of Cell-Cell Fusion from Without by Human Herpesvirus 6B

Abstract

Human herpesvirus (HHV) 6A induce fusion from without (FFWO), whereas HHV-6B is believed to be ineffective in this process. Here, we demonstrate that HHV-6B induces rapid fusion in both epithelial cells and lymphocytes. The fusion was identified 1 h postinfection, could be inhibited by antibodies to HHV-6B gH and to the cellular receptor CD46, and was dependent on virus titer but independent of de novo protein synthesis and UV inactivation of the virus. Comparisons indicate that HHV-6A is only 10-fold more effective in inducing FFWO than HHV-6B. These data demonstrate that HHV-6B can induce FFWO in epithelial cells and lymphocytes.

Two closely related human herpesviruses, human herpesvirus 6A (HHV-6A) and HHV-6B (22), both utilize the cellular receptor CD46 and possibly a yet-unknown coreceptor for entry into host cells (20, 21). Both viruses display primary tropism for T lymphocytes, but abortive infection has been shown in a variety of cell types (4, 6).

Elucidating biological differences and similarities are important to further understand the characteristics of HHV-6A and HHV-6B infections. Akkapaiboon et al. reported that HHV-6A, but not HHV-6B, induced syncytia formation in a range of different human cells (1). Fusion of infected cells may be caused by the de novo production of viral glycoproteins days after infection. However, virus may also cause cell-cell fusion from without (FFWO), a process that occurs early and is not dependent on de novo synthesis of viral glycoproteins (3). CD46 and the viral glycoproteins gH, gL, and gQ were essential for HHV-6A-induced FFWO (1, 15, 16). Although HHV-6B could induce multinucleated cells in MT-4 cultures, the ability of HHV-6B to induce syncytia formation was poor in, for example, SupT1, a human T-cell line, and 293T, a human epithelial kidney-cell line derived from HEK 293 cells (15). However, our findings suggest that HHV-6B is also capable of inducing FFWO.

We investigated whether the PL-1 strain of HHV-6B (12) was able to induce fusion in HEK 293 and SupT1. MOLT 3 cells were used to propagate the PL-1 strain of HHV-6B. When the cytopathic effect was seen in more than 80% of the cells, the supernatant and infected cells were collected. Freeze-thaw cycles were applied to the infected cells to release additional virus, which along with the supernatants were cleared from cellular debris by centrifugation at 5,000 rpm. Further concentration of the virus was performed by ultracentrifugation at 30,000 rpm for 1 h, followed by resuspension of the pellet in Iscove modified Dulbecco medium (Invitrogen, Taastrup, Denmark) supplemented with 10% fetal calf serum. The virus titer was determined by microscopically defined cytopathic effects on MOLT 3 cells as 50% tissue infective culture doses (TCID₅₀) by the Reed-Muench method (19). Prior to infection, HEK 293 cells were seeded in 24-well plates and left to adhere for 24 h. SupT1 (5 × 10⁵) and HEK 293 (2 × 10⁵) cells were infected with HHV-6B at various TCID₅₀s. At 4 hours postinfection (hpi) progress of syncytium formation was observed (Fig. 1A). The percentage of nuclei in syncytia were obtained for HEK 293 cells by scoring polykaryocytes containing more than three nuclei per syncytia in infected cultures stained with crystal violet (Fig. 1B). Nuclei were scored in a blinded fashion by three independent observers. At 182 TCID₅₀ HEK 293 cells showed widespread formation of syncytia, including up to 90% of the cells. Likewise, SupT1 cells formed large syncytia (Fig. 1C) with the presence of multinucleated cells in crystal violet staining (data not shown). With decreasing virus titer, a correlating decrease in the formation of syncytia was observed, with the absence of detectable syncytium formation at 4 hpi using 23 TCID₅₀.

FIG. 1.

HHV-6B-induced fusion in HEK 293 and SupT1 cells. (A) HEK 293 and SupT1 cells were infected with HHV-6B, strain PL-1, at virus titers of 182, 91, 46, 23, or 12 TCID₅₀. At 4 hpi, fusion was documented using an Olympus IX71 microscope fitted with a Leica DFC350F camera. Arrows indicate fused cells. (B) HEK 293 cells were infected with HHV-6B, strain PL-1, at 364, 182, 91, 46, 23, 12, 6, or 0 TCID₅₀. Cultures were stained with 0.5% crystal violet prior to being photographed, as described in panel A. Nuclei present in polykaryocytes containing more than three nuclei and total nuclei were scored in a blinded fashion by three individual observers. The percentage of nuclei in syncytia were obtained for the indicated virus titers. The standard deviations are shown on the top of the bars. (C) HEK 293 and SupT1 cells were infected with HHV-6B, strain PL-1, at 182 TCID₅₀, and fusion was documented as described in panel A at 0, 2, and 4 hpi. Arrows indicate fused cells. (D) HEK 293 cells were infected with HHV-6B, strain PL-1, at 132 TCID₅₀. The progress of syncytia formation was documented as described in panel A at 0, 1, 2, 3, and 4 hpi. Arrows indicate lammelipodia movement.

To further demonstrate that the syncytia were generated by FFWO, kinetic analysis of their formation was investigated. HEK 293 and SupT1 cells were infected with HHV-6B at 182 TCID₅₀, and the formation of syncytia was studied at 0, 2, and 4 hpi (Fig. 1B and C). Fusion was detectable in both cell lines at 2 hpi, and widespread cell-cell fusion was observed at 4 hpi (Fig. 1B and C, arrows). The fusion between cells that formed contact occurred within an hour (Fig. 1D). The rapid formation of syncytia was further documented by flow cytometry using forward-scatter and side-scatter (FS/SS) analysis (Fig. 2A). The percentage of enlarged cells provides a measure of fused SupT1 cells (Fig. 2A). HHV-6B-infected cells displayed an increased percentage of enlarged cells at 1 hpi, and this fraction increased during the following hours (Fig. 2A and B). Notably, the enlarged cells contained several nuclei and, furthermore, after 4 h the formation of large syncytia that were out of range on the flow cytometer contributed to an underestimation of the percentage of fused cells at these time points. The infected cells also showed a higher amount of granules so, to investigate whether the infected cells were viable, a propidium iodide (PI; Sigma-Aldrich, Brondby, Denmark) analysis was performed (Fig. 2C). The fraction of PI-positive cells at 4 hpi did not differ markedly from the fraction at 0 hpi. This indicated that the enlarged cells were indeed live cells. Thus, these experiments clearly indicated the rapid formation of syncytia induced by HHV-6B.

FIG. 2.

HHV-6B-induced fusion is dependent on the human cellular receptor CD46 and the viral glycoprotein H (gH). (A) FS/SS analysis of HHV-6B-induced FFWO in SupT1 cells. SupT1 cells were infected with HHV-6B at 182 TCID₅₀ and FS/SS analyzed 0, 1, 2, and 4 hpi on an FC500 Coulter/Beckmann flow cytometer. A dot plot with gated events represents the enlarged cells in the culture. The data represent 10,000 events. (B) Percentage of enlarged cells in HHV-6B-infected SupT1 culture was evaluated by flow cytometry as in panel A. The presence of anti-CD46 or anti-gH antibodies prevents the development of an increased FS. (C) Viability analysis of HHV-6B-infected SupT1 cells. Infected cells were stained with PI at 4 hpi and analyzed on a BD Calibur flow cytometer. (D) One hour prior to infection, SupT1 and HEK 293 cells were incubated in the presence or absence of either anti-gH (1:5,000 diluted ascites), 10 μg of anti-CD46/ml, 10 μg of anti-HLA class I/ml, or 10 μg of IgG1 isotype immunoglobulins/ml. Cells were then mock infected or infected with HHV-6B at 182 TCID₅₀. FFWO was documented 4 hpi, as described in Fig. 1A. Arrows indicate fused cells.

Viral glycoproteins, in particular gL and gH, are important for the binding of human herpesvirus to their host cells (2, 8, 10, 11). For HHV-6A the heterotetrameric complex gH-gL-gQ1(80K)-gQ2(37K) (1) is the viral ligand for the cellular receptor CD46, which is essential for HHV-6A-mediated FFWO (5, 15). We therefore tested whether antibodies against the viral glycoprotein gH or human CD46 could prevent HHV-6B-induced fusion. HEK 293 and SupT1 cells were incubated in the presence or absence of either anti-gH 2E4 ascites diluted 1:5,000 (kindly provided by G. Campadelli-Fiume, University of Bologna, Bologna, Italy), 10 μg of anti-CD46 (Acris Antibodies GmbH, Hiddenhausen, Germany)/ml, 10 μg of anti-HLA class I (Dako A/S, Glostrup, Denmark)/ml, or 10 μg of immunoglobulin G1 (IgG1) isotype immunoglobulins (Sigma-Aldrich) 1 h prior to infection with HHV-6B (182 TCID₅₀). Fusion was documented 4 hpi by microscopy (Fig. 2D) or at different time points by flow cytometry for SupT1 cells (Fig. 2B). The presence of anti-gH or anti-CD46 antibodies virtually prevented fusion in infected HEK 293 cells (Fig. 2D) and SupT1 cells (Fig. 2B and D), whereas anti-HLA class I or IgG1 immunoglobulins had no influence on the induced fusion (Fig. 2B and D). These data indicate that the HHV-6B-mediated fusion is dependent on gH and the human cellular receptor CD46.

FFWO is independent of de novo viral protein synthesis (3). We therefore assessed whether HHV-6B-induced fusion in SupT1 and HEK 293 cells required de novo protein synthesis by blocking the translation of mRNA with 10 μg cycloheximide (Sigma-Aldrich)/ml 1 h prior to infecting the cells with HHV-6B (23). In addition, the syncytium-generating ability of 182 TCID₅₀ UV-inactivated HHV-6B (17) was examined. After 4 hpi, cells infected with HHV-6B in the presence or absence of cycloheximide (Fig. 3A) or treated with UV-inactivated HHV-6B (Fig. 3B) induced fusion in HEK 293 cells. Similar data were obtained with SupT1 cells (data not shown). These observations indicate that HHV-6B-induced fusion is independent of de novo protein synthesis and is therefore caused by FFWO.

FIG. 3.

Influence of de novo cellular and viral protein synthesis on syncytia formation. (A) Cellular protein synthesis in HEK 293 cells were blocked by incubation with 10 μg of cycloheximide (CHX)/ml at 1 h before infection. Cells were then either mock infected or infected with HHV-6B at 182 TCID₅₀ for 4 h. FFWO was documented as described in Fig. 1. Arrows indicate fused cells. (B) HHV-6B was UV inactivated for 20 min before infecting HEK 293 cells at 182 TCID₅₀. FFWO was documented 4 hpi as described in Fig. 1. Arrows indicate fused cells.

Studies of the fusion in HEK 293 cells suggested the involvement of cytoskeletal proteins in the process through protraction and retraction of lammelipodia, creating contact points between fusing cells (Fig. 1D). We therefore investigated the influence of inhibitors of the cytoskeletal proteins, actin and tubulin, on the fusion process by blocking their function with respectively cytochalasin D (Medinova Scientific AS, Glostrup, Denmark) and colchicine (Sigma-Aldrich) (data not shown). Even though fusion could be detected at 4 hpi in the presence of these inhibitors, their interference with cell morphology made microscopic evaluations more difficult. We can therefore not entirely rule out an involvement of tubulin or actin in the fusion process.

Together, these results disclose a novel functional role of HHV-6B by showing that HHV-6B is capable of inducing FFWO in both epithelial cells and lymphocytes. This conclusion is based on our findings that the fusion occurs early and is dependent on the amount of HHV-6B, on viral gH, and on CD46 cellular receptors. Furthermore, fusion occurs in the presence of cycloheximide or with UV-inactivated HHV-6B.

Rapid FFWO requires a high virus titer. We therefore sought to compare the fusogenic ability of HHV-6B with that of HHV-6A. Since FFWO can occur in the presence of inactivated virus, we decided to compare the amount of the viral glycoprotein gp60/110 (Chemicon International, Temecula, CA) in the preparations of HHV-6B and HHV-6A by Western blotting. Virus supernatant containing HHV-6A (strain U1102, kindly provided by S. Jacobson, National Institutes of Health, Bethesda, MD) was shown to be able to induce FFWO in HEK 293 cells (data not shown). Comparisons of HHV-6B dilutions and undiluted HHV-6A supernatant demonstrated that HHV-6B contained approximately 100-fold more gp60/110 (Fig. 4A). However, an endpoint titration of HHV-6B and HHV-6A based on the ability to induce FFWO in HEK 293 cells at 4 hpi (Fig. 4B), indicated that the HHV-6B virus stock was approximately 10-fold better at inducing FFWO. Based on these data, we conclude that HHV-6B is clearly capable of inducing FFWO but that HHV-6A has approximately 10-fold better ability to do so than HHV-6B.

FIG. 4.

Comparison of HHV-6A, strain U1102, and HHV-6B, strain PL-1. (A) HHV-6B was diluted twofold in RPMI plus 10% fetal calf serum. Western blotting was performed with the diluted HHV-6B, undiluted HHV-6A, and RPMI plus 10% fetal calf serum. The blot was probed with an antibody against the HHV-6A/B viral glycoprotein gp60/110. Ponceau-S staining was performed as a loading control. (B) Endpoint titration of HHV-6B and HHV-6A based on induced FFWO at 4 hpi. HEK 293 cells were infected with different virus titers. At 4 hpi, cultures were scored for the presence of syncytia. Scoring was performed in quadruplets.

The discrepancy between Mori et al. (15) and our data in the ability of HHV-6B to form FFWO can be explained in several ways. First, the titer of HHV-6B during infection may be different. However, it is not possible to directly compare the titers of HHV-6B in the two experiments. Second, an interesting possibility is the potential differences between the strains of HHV-6B. We used the PL-1 strain of HHV-6B, which appears to be very similar to Z29 (18), whereas Mori et al. used HST. Our PL-1 strain was propagated in MOLT 3 cells, whereas Mori et al. used cord blood mononuclear cells for propagation. Since the viral tetrameric complex contains gQ, there is a possibility for cell type-specific splicing patterns in the U100 protein product (1, 7, 13, 14). Indeed, it has been shown that the cell type for herpes simplex virus type 1 propagation may not only influence the titer of virus in the supernatant but also subsequent phenotypic results (9). In any case, the formation of syncytia by FFWO can be achieved by the PL-1 strain of HHV-6B.