Human Cytomegalovirus gH/gL/gO Promotes the Fusion Step of Entry into All Cell Types, whereas gH/gL/UL128-131 Broadens Virus Tropism through a Distinct Mechanism

Cell lines.

Primary human foreskin fibroblasts (HFF) (Life Technologies), MRC-5 fibroblasts (ATCC CCL-171), and HFFtet (which express the tetracycline [Tet] repressor protein [35]; provided by Richard Stanton [Cardiff University, Cardiff, United Kingdom]) were grown in Dulbecco's modified Eagle's medium (DMEM) (Life Technologies) supplemented with 6% heat-inactivated fetal bovine serum (FBS) (Rocky Mountain Biologicals, Inc., Missoula, MT, USA) and 6% bovine growth serum (BGS) (Rocky Mountain Biologicals, Inc., Missoula, MT, USA). The retinal pigment epithelial cell line ARPE-19 (American Type Culture Collection, Manassas, VA, USA) was grown in a 1:1 dilution mixture of DMEM and Ham's F-12 medium (DMEM–F-12) (Life Technologies) supplemented with 10% FBS.

HCMV.

All human cytomegalovirus (HCMV) strains were derived from bacterial artificial chromosome (BAC) clones. BAC clone TR was provided by Jay Nelson (Oregon Health and Sciences University, Portland, OR, USA) (36). BAC clone TB40/e (BAC-4) (TB) was provided by Christian Sinzger (University of Ulm, Ulm, Germany) (37). BAC clone Merlin (ME) (pAL1393), which contains tetracycline operator sequences within the transcriptional promoter of UL130 and UL131, was provided by Richard Stanton (Cardiff University, Cardiff, United Kingdom) (35). Infectious HCMV was recovered by electroporation of BAC DNA into MRC-5 fibroblasts, as described by Wille et al. (34). Cell-free HCMV stocks were produced by infecting HFF or HFFtet at 2 PFU per cell. At 8 to 10 days postinfection (p.i.) (when cells were still visually intact), culture supernatants were harvested, and cellular contaminants were removed by centrifugation at 1,000 × g for 10 min and again at 6,000 × g for 10 min. Stocks were judged cell free by the lack of calnexin and actin in Western blot analyses and stored at −80°C. The number of PFU was determined by plaque assay on triplicate HFF or ARPE-19 cultures. Freeze/thaw cycles were avoided.

Antibodies.

Monoclonal antibodies (MAbs) specific for HCMV major capsid protein (MCP) 28-4 and gB 27-156 were provided by Bill Britt (University of Alabama, Birmingham, AL, USA) (38, 39). Anti-UL128 MAb 4B10 was provided by Tom Shenk (Princeton University, Princeton, NJ, USA) (24). Rabbit polyclonal antipeptide antibodies directed against HCMV gH/gL, UL130, and UL131 were provided by David Johnson (Oregon Health and Sciences University, Portland, OR, USA) (40). Rabbit polyclonal antipeptide antibodies directed against MEgO were described previously (25).

Western blotting.

Cell-free virions from culture supernatants (as described above) were concentrated by centrifugation at 50,000 × g for 1 h and resuspended in 2% SDS in 20 mM Tris-buffered saline (TBS) (pH 6.8). Insoluble material was removed by centrifugation at 16,000 × g for 30 min, and the cleared extracts were heated to 95°C for 10 min. For reducing blots, extracts were adjusted to 25 mM dithiothreitol (DTT). Proteins were separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Whatman) in a buffer containing 10 mM NaHCO₃ and 3 mM Na₂CO₃ (pH 9.9) plus 10% methanol. The transferred proteins were probed with MAbs or rabbit polyclonal antibodies specific for HCMV proteins, followed by horseradish peroxidase-conjugated secondary antibodies; chemiluminescence was detected using a Bio-Rad ChemiDoc MP imaging system. Band densities were quantified using Bio-Rad Image Lab v5.1 or ImageJ v1.48 software.

qPCR.

Cell-free HCMV stock (200 μl) was first treated with DNase I to remove DNA not protected within viral capsids. Capsids were disrupted by using the viral lysis buffer and proteinase K provided in the PureLink Viral RNA/DNA minikit (Life Technologies). Viral DNA was eluted in 50 μl RNase-free water. A region within UL83 conserved among ME, TR, and TB was chosen as the amplicon, and genomes were quantified by real-time quantitative PCR (qPCR) using SYBR green dye (Bio-Rad) and the forward primer TGGTCACCTATCACCTGCAT and the reverse primer GAAAGAGCCCGACGTCTACT. PCR samples were made to a total well volume of 20 μl with 1× SYBR green master mix (Bio-Rad), a 1:100 dilution of viral DNA, and 500 nM forward and reverse primers. PCR products were detected using the MyiQ real-time PCR detection system (Bio-Rad). The qPCR results were compared to a plasmid pTRUL83 standard and then calculated as genomes per milliliter. The efficiency of DNase I treatment was tested in independent experiments by spiking plasmid DNA. For each experiment, two independent viral DNA extractions and three independent qPCRs were performed.

Determination of PFU.

Cell monolayers were infected with cell-free HCMV or HCMV-infected cells (infectious centers), and the cell culture medium was replaced with a 1:1 mixture of 2× DMEM and 1.2% SeaPlaque agarose to prevent the spread of progeny virus by diffusion. Plaques were counted by light microscopy 3 weeks later.

Antibody neutralization of HCMV.

Anti-UL130 and anti-UL131 rabbit sera diluted 1:100 (vol/vol) were added to culture medium containing cell-free HCMV. The mixture was incubated at room temperature for 1 h and then added to HFF or ARPE-19 cultures.

PEG treatment.

Cell-free HCMV was adsorbed to cells for 2 h at 10°C under 800 × g centrifugation or at 37°C without centrifugation. Unbound virus was removed by two cold phosphate-buffered saline (PBS) washes, and cells were subsequently treated with 44% polyethylene glycol (PEG) (prewarmed to 37°C) for 30 s, followed by 10 washes with warm culture medium to remove the PEG. The cultures were incubated at 37°C until immunofluorescence (IF) analysis was performed. PEG 6000 (Fluka) was prepared as a 60% (wt/wt) solution in PBS and diluted with PBS to 44%.

Immunofluorescence.

HCMV-infected cells were fixed with 2% formaldehyde in PBS for 15 min, followed by three PBS washes. The fixed cells were incubated for 1 h in IF buffer (0.5% Triton X-100, 0.5% deoxycholate, 1% bovine serum albumin [BSA], 0.05% sodium azide in PBS). To visualize HCMV-infected cells, anti-IE1/IE2 MAb P63-27 was used at 1:50 in IF buffer, and Alexa 488-conjugated goat anti-mouse antibodies were diluted 1:1,000 in IF buffer. DAPI (4′,6′-diamidino-2-phenylindole dihydrochloride) was used to visualize total cells.

HCMV strains vary in both the ratio of gH/gL/gO to gH/gL/UL128-131 and the total amount of gH/gL complexes.

Zhou et al. (25) analyzed gH/gL complexes of different HCMV strains by nonreducing Western blotting. Since the main goal of those experiments was to compare the ratios of gH/gL/gO to gH/gL/UL128-131 among strains, the gel loads were adjusted to equalize the total amounts of gH/gL. TR and TB virions were found to contain gH/gL mostly in the form of gH/gL/gO, whereas ME had mostly gH/gL/UL128-131 (25). The experiments made use of a recombinant ME in which the assembly of gH/gL/UL128-131 could be reduced by the tetracycline-repressor protein (TetR), which resulted in ME virions containing mostly gH/gL/gO (denoted Merlin-trimer [ME-T]) (25, 35). To further characterize the amounts of gH/gL complexes among HCMV strains, extracellular virions of TB, TR, ME, and ME-T (ME virions produced in TetR-expressing cells) were analyzed by Western blotting under reducing conditions, and the gel loads were normalized to major capsid protein to allow comparison of the amounts of the different glycoproteins per virion (Fig. 1). All strains contained comparable amounts of gB, but the amounts of gH/gL varied dramatically. TB contained the most total gH/gL, followed by TR, ME, and ME-T (Fig. 1A). Despite the small amount of gH/gL, ME virions contained dramatically more UL131 (a marker of the gH/gL/UL128-131 complex) (Fig. 1B). Note that in Fig. 1B, the gel loads were adjusted to allow visualization of the smaller amounts of UL131 in TB and TR virions, suggesting that the difference in the level of UL131 between ME and the other viruses was greater than 10-fold.

UL128, UL130, and gO were not included in the above-mentioned reducing-gel analyses because amino acid polymorphisms between strains affect the cross-reactivities of available antibodies (25). Thus, HCMV gH/gL complexes were analyzed by nonreducing Western blotting, probing for the antigen common to both complexes, gL (Fig. 2). Consistent with analyses reported by Zhou et al., the vast majority of gH/gL in TB, TR, and ME-T virions migrated at a size indicative of disulfide-linked gH/gL/gO, whereas most of the gH/gL in ME virions was at the size of disulfide-linked gH/gL/UL128 (Fig. 2A). Note that UL130 and UL131 are not disulfide linked to gH/gL/UL128-131 and are thus removed by SDS-PAGE (24, 40). Unbound, or free, gH/gL comprised a very small fraction of the total gH/gL for all strains. Whether this represents spontaneous reduction of disulfide bonds in the extract samples or is indicative of small amounts of free gH/gL in virions is not clear. Band density analyses of the three gH/gL species on nonreducing blots corresponded well with the quantitation of total gH and gL on reducing blots (compare Fig. 1A and 2).

The small amounts of gH/gL complexes in ME and ME-T virions made quantitation in these viruses difficult to perform on the same blots as TB and TR. Thus, ME and ME-T virions were further analyzed on separate nonreducing gels with increased gel loads to allow detection in a quantifiable range (Fig. 3). Furthermore, since ME and ME-T are genetically identical viruses, the virions could also be directly compared using anti-MEgO and anti-UL128 antibodies as Western blot probes (25). As in the previous analyses, ME virions contained more total gH/gL than ME-T virions, although at these increased loads, the difference was less pronounced. This likely reflected the limitations of band density quantitation on the lower end of the spectrum in Fig. 1 and 2. Comparison of the amounts of gH/gL/gO and gH/gL/UL128-131 in ME and ME-T virions by probing blots for either the common antigen (gL) or the unique antigen (gO or UL128) confirmed that ME-T virions contained more gH/gL/gO than ME virions, whereas ME virions had more gH/gL/UL128-131 than ME-T virions (compare analogous bands in Fig. 3A, B, and C). Together, these analyses indicated that strains of HCMV vary widely in the total amount of gH/gL complexes in the virion envelope and also vary in whether that gH/gL is in the form of gH/gL/gO or gH/gL/UL128-131.

Particle-to-PFU ratio analysis of strains of HCMV that contain different amounts of gH/gL complexes.

One model of HCMV tropism suggests that gH/gL/gO and gH/gL/UL128-131 serve mechanistically analogous roles for entry into fibroblasts and epithelial cells, respectively. However, that model is inconsistent with observations that a gO-null mutant TR was impaired for entry into fibroblasts and epithelial cells (34). To help resolve this issue, we reasoned that the particle-to-PFU ratio of HCMV on either fibroblasts or epithelial cells should be influenced by the amounts of the gH/gL complexes in the virion envelope. Thus, cell-free virions of TB, TR, ME, and ME-T were quantitated by real-time qPCR for viral genomes, the titers were determined by plaque assay on both fibroblasts and epithelial cells, and the particle-to-PFU ratio was calculated (Fig. 4).

Particle-to-PFU ratio measurements on fibroblasts varied widely between strains and were statistically significant in all comparisons (Fig. 4A). The extremes were TB and ME at 11 and 5,825, respectively, which correlated with the large and small amounts, respectively, of gH/gL/gO detected in these strains (Fig. 2). The values for TR and ME-T were intermediate (217 and 83, respectively), which generally correlated with their intermediate amounts of gH/gL/gO relative to TB and ME. Notably, however, TR contained considerably more gH/gL/gO than ME-T but was 2.6-fold less infectious than ME-T. It is possible that this relatively modest difference might indicate that the level of gH/gL/gO in ME-T represents a minimum threshold. Additionally, it might reflect other genetic differences between the strains that affect plaque formation. Notably, ME-T was 70-fold more infectious than ME on fibroblasts (Fig. 4A). Since ME-T and ME are genetically identical viruses, this difference likely reflects the direct effect of the different amounts of gH/gL/gO and gH/gL/UL128-131 on entry into these cells.

Particle-to-PFU ratios were generally higher on epithelial cells, consistent with previous observations that these cells are more difficult to infect with HCMV than are fibroblasts, perhaps reflecting the more complicated route of entry, which involves internalization followed by fusion from within endosomes (27). TB was the most infectious on epithelial cells (419 particles/PFU), followed by ME-T (1,250 particles/PFU), ME (10,529 particles/PFU), and TR (325,330 particles/PFU) (Fig. 4B). All comparisons were statistically significant. Given that ME virions contained dramatically more gH/gL/UL128-131 than the other viruses, the observation that ME was nearly 10-fold less infectious than ME-T and 25-fold less infectious than TB on epithelial cells was inconsistent with the model that gH/gL/UL128-131 performs the conserved herpesvirus gH/gL function of promoting gB-mediated fusion on these cells. Rather, the comparatively poor infectivity of ME on both fibroblasts and epithelial cells is more consistent with the hypothesis that gH/gL/gO provides the core fusion function for entry into all cell types and that gH/gL/UL128-131 provides a mechanistically distinct additional function necessary for entry into epithelial cells. However, the particle-to-PFU ratio of TR on epithelial cells was several log-fold higher than any of the other viruses, despite being relatively rich in gH/gL/gO. This discrepancy was likely due in part to the fact that plaque formation is a function not only of initial entry of cell-free virions, but also of replication and the spread of progeny.

To address postentry differences in plaque formation among HCMV strains, an infectious-center experiment was performed. Fibroblasts or epithelial cells were infected with TB, TR, or ME, and at 2 days p.i., the cells were detached from the culture dishes and replated onto monolayers of the homologous cell type, and the resultant plaques were counted (Fig. 5). Note that since ME-T virions were produced by propagating ME in cells expressing TetR to repress transcription from the UL128 locus (25, 35), they are genetically identical to ME virions. Thus, there was no logical distinction to be made between ME and ME-T in these infectious-center experiments. On fibroblasts, TB, TR, and ME were all comparably efficient at forming plaques from infectious centers (40 to 60 plaques per 100 infectious centers seeded) (Fig. 5A). On epithelial cells, TB and ME formed plaques from infectious centers with efficiencies similar to those on fibroblasts, but TR was notably less efficient, yielding only 10 plaques per 100 infectious centers (Fig. 5B). In interpreting these results, it is important to note that gO-null TR was severely impaired for entry but more efficient at cell-cell spread in these epithelial cells than the parental wild-type TR (34), suggesting a mechanistic distinction between entry and spread on these cells. Thus, the poor ability of TR to form plaques on epithelial cells from infectious centers helps to explain the apparently high particle-to-PFU ratio of TR on these cells and is also consistent with reports indicating that HCMV tropism is not solely determined at the level of entry (41, 42).

Perhaps the most surprising result of the particle-to-PFU analysis presented in Fig. 4 was that ME-T was approximately 10-fold more efficient than ME on epithelial cells (Fig. 4B). Western blot analysis clearly demonstrated that ME-T virions contain mostly gH/gL/gO and very little gH/gL/UL128-131, whereas ME contains abundant gH/gL/UL128-131 and smaller amounts of gH/gL/gO (Fig. 1 to 3). It is possible that these differences in gH/gL complexes affected the fundamental mechanism or route of entry for ME and ME-T into these cells. To test this, antibody neutralization experiments were performed using anti-UL130 and anti-UL131, which have been shown previously to block HCMV infection of epithelial cells, but not fibroblasts (43) (Fig. 6). Anti-UL130/131 antibodies blocked infection of both ME and ME-T on epithelial cells, but not on fibroblasts, indicating that for both viruses, infection of epithelial cells, but not fibroblasts, was gH/gL/UL128-131 dependent. These results indicate that (i) ME-T virions were not completely devoid of gH/gL/UL128-131 and (ii) the enhanced infectivity of ME-T on epithelial cells was not due to an aberrant, gH/gL/UL128-131-independent process. Taken together, these analyses strongly contradict the model that gH/gL/gO and gH/gL/UL128-131 serve mechanistically analogous roles for entry of HCMV into fibroblasts and epithelial cells, respectively. Instead, it seems more likely that the two complexes act through distinct mechanisms to facilitate entry into these cell types.

The particle-to-PFU ratio difference between ME and ME-T reflects the fusion step of entry.

Since ME and ME-T are genetically identical viruses, their postentry replication and spread characteristics are identical in cells that do not express TetR (25, 35). Thus, the observed differences in particle-to-PFU ratios (Fig. 4) were likely indicative of differences in the entry process as a direct result of the amounts of gH/gL/gO and gH/gL/UL128-131 in the virion envelope, likely acting on the ability of these viruses to accomplish the fusion step of entry. To test this hypothesis, equivalent numbers of ME or ME-T virion particles were adsorbed onto fibroblasts or epithelial cells, and then both viruses were provided with an equivalent surrogate fusion mechanism, PEG treatment (Fig. 7) (27).

As expected from the particle-to-PFU analyses, ME-T efficiently infected fibroblasts, but ME did not (66% and 2.5%, respectively) (Fig. 7A). PEG treatment had no significant effect on the number of cells infected with ME-T, suggesting that the endogenous fusion mechanism of ME-T on these cells was similar to the efficiency of PEG-induced fusion. In contrast, PEG resulted in an approximately 20-fold increase in the numbers of cells infected with ME, bringing the percentage of cells infected with ME very near the values for ME-T. These results suggested that ME virions were poorly infectious on fibroblasts, largely due to an inefficient fusion mechanism.

At the dose of virus used in these experiments (150 particles per cell), neither ME nor ME-T infected quantifiable numbers of epithelial cells (Fig. 7B). This was consistent with the much higher particle-to-PFU ratio measured on these cells than on fibroblasts (Fig. 4). PEG treatment resulted in similar numbers of infected cells for both ME and ME-T, consistent with the notion that the high particle-to-PFU ratio of ME compared to ME-T reflected a difference in the abilities of these viruses to accomplish the fusion step of entry, not later stages of the replication cycle. However, because of the relatively high particle-to-PFU ratios for both ME and ME-T on these cells (1,250 and 10,529, respectively) (Fig. 4B), it was possible that the 150 particles per cell used in these experiments did not reflect the replication-competent virions in each population. To address this caveat, another set of experiments were performed in which 10 or 30 PFU were plated on epithelial cells, virus entry was induced with PEG, and the number of resultant plaques was determined (Fig. 8). PEG had no significant effect on the numbers of ME-T plaques, suggesting that the endogenous entry machinery of replication-competent ME-T virions was as efficient as PEG-induced entry. In contrast, PEG treatment resulted in plaque numbers 4- to 5-fold greater than the inputs of ME (Fig. 8). This represents a recovery of approximately half of the particle-to-PFU difference between ME and ME-T on these cells (Fig. 4B). The remaining half may reflect the limits of PEG efficiency. These results strongly suggest that the comparatively poor infectivity of ME on epithelial cells is due to an inability to accomplish the fusion stage of entry, despite their elevated levels of gH/gL/UL128-131.