HSV-1 infection of human corneal epithelial cells: Receptor-mediated entry and trends of re-infection

Cells, viruses, and antibodies

The human corneal epithelial (HCE) cell line (RCB1834 HCE-T) was obtained from Dr. Kozaburo Hayashi (National Eye Institute, Bethesda, MD) [23]. The cells were grown in Minimal Essential Media (MEM; Gibco-Invitrogen Corp., Rockville, MD) and supplemented with penicillin, streptomycin, and 10% fetal bovine serum (FBS).

The African green monkey kidney (Vero) cell line as well as the wild-type Chinese hamster ovary (CHO-K1) cell lines were obtained (Patricia G. Spear, Northwestern University, Chicago, IL) and cultured as previously decribed [19]. The viruses used in the study – HSV-1(KOS), HSV-1(KOS)804, HSV-1(KOS)gL86, and HSV-1(KOS)RFP – were also provided by Patricia G. Spear [19]. All virus stocks were kept at a low multiplicity of infection (MOI), propagated in Vero cell lines, and stored at −80 °C. Titers for all three of the viruses were determined using Vero cells as well.

Anti-nectin-1 antibodies used were monoclonal mouse anti-nectin-1 (37–5900; Zymed Laboratories, San Francisco, CA). The monoclonal mouse anti-HVEM antibody was obtained from Santa Cruz Biotechnology (sc-74089; Santa Cruz, CA) and the monoclonal rat anti-PILR- α was obtained from Dendritics (ddx0230; Lyon, France). The secondary antibodies FITC-conjugated anti-mouse IgG and FITC-conjugated anti-rat IgG were obtained from Sigma-Aldrich (St. Louis, MO).

Viral entry assays

Viral entry assays were performed using both HCE cells and naturally-resistant CHO-K1 cells, the latter of which do not possess the known viral entry receptors. Cells were plated on a 96-well tissue culture dish at full confluency (approximately 2×10⁴ cells per well). Cells were then infected with serial dilutions of β-galactosidase expressing HSV-1(KOS)gL86 virus at a maximum MOI of 0.01 in a solution of 3% BSA (BSA) in phosphate buffered saline (PBS). After 6 h of infection, cells were washed with 1× PBS and then incubated with o-nitrophenyl- β-D-galactopyranoside (ImmunoPure ONPG; Pierce, Rockford, IL), thus allowing quantification of β-galactosidase expressing HSV-1(KOS)gl86 viral entry into each of the cell types. Measurements of activity were then obtained at an optical density of 410nm by a spectrophotometer (Molecular Devices, Sunnyvale, CA).

HSV-1 entry into HCE and CHO-K1 cells was then confirmed qualitatively by 5-bromo-4-chloro-3-indolyl-β-D-galactopyranoside (X-gal) assays. Monolayers of cells were plated in a 6-well tissue culture dish (approximately 1.5×10⁶ cells per well). Cells were again infected for 6 h using the HSV-1(KOS)gL86 virus. After the incubation period, cells were washed, fixed, permeabilized, and incubated with the X-gal substrate. The X-gal substrate forms a blue color with cells that contain β-galactosidase. Microscopy images were collected using a 20× objective from the inverted microscope (Aciovert 100M; Zeiss, Jena, Germany) with the software Slidebook, version 3.0.

Time point plaque assay

Viral entry and replication were again confirmed using a time point plaque assay. Monolayers of HCE, CHO-K1, and Vero cells were plated on a 6-well tissue culture dish (approximately 1.5×10⁶ cells per well). Cells were infected at 0.01 MOI with HSV-1(KOS)804 virus or mock infected with 1× PBS for 2 h. After primary incubation, the viral or mock solution were replaced with the proper media for each cell type – MEM with 10% fetal bovine serum (FBS) and penicillin/streptomycin additives for HCE cells, Ham's F-12 medium with 10% FBS and penicillin/streptomycin additives for CHO-K1 cells, and DMEM with 10% FBS and penicillin/streptomycin additives for Vero cells. Cells were then incubated at 37 °C for a series of time points. At the time points of 0, 12, 18, 24, and 36 h, incubated cells were fixed using 100% methanol and subsequently stained with a solution of 1% crystal violet in 50% ethanol at room temperature. Cells were then washed three times with nanopure water and the number and size of plaques were assessed as a measure of viral replication. Plaque size was measured and plotted as an average of 100 plaques formed at each time interval. Plaques were not significantly formed by the naturally-resistant CHO-K1 cells and thus plaque sizes were not measured or plotted. Images of the plaques were taken using the 40× objective from the inverted microscope (Aciovert 100M; Zeiss) using the software Slidebook version 3.0.

pH dependence of entry

HCE cells were plated on a 96-well tissue culture dish (approximately 2×10⁴ cells per well). Cells were then treated with serial dilutions of BFLA-1 (1 μM) or chloroquine (1 mM) and incubated for 1 hat 37 °C. Concentrations of BFLA-1 and chloroquine followed those of previous studies [24,25]. Cells were then treated with equal concentrations of HSV-1(KOS)gL86 virus. The viral entry assay was then performed and β-galactosidase activity was determined using the same protocol as described above.

Isolation of RNA

RNA was isolated from the Human Corneal Epithelial cells and CHO-K1 cells with TRIzol® reagent (Invitrogen, Carlsbad, CA). Total RNA samples were purified by RNeasy Mini kit (Qiagen, Valencia, CA) according to the manufacturer’s manual and checked the quantity of RNAs using NanoDrop (Thermo Scientific, Wilmington, DE). The quality of RNA was evaluated by NanoDrop (Thermo Scientific) and warranted that the ratio (28S/18S) of high RNA quality were >1.5. Densitometry analysis was performed using the NIH ImageJ software (version 1.43) and data for relative intensity was plotted as a histogram.

Gene expression analysis and verification by qRT–PCR

To examine the gene expression by quantitative RT–PCR, 10 µg of the RNA samples from HCE and CHO-K_i cells were converted to first strand cDNAs using the High-Capacity cDNA Reverse Transcription Kits (Applied Biosystems, Foster City, CA) following the manufacturer’s protocol. The cDNA was added into the Platinum SYBR Green qPCR SuperMix-UDG Master Mixes (Invitrogen, Carlsbad, CA) containing SYBR Green I fluorescent dye and Taq DNA polymerase. The mixtures were aliquoted into 96-well plate containing gene specific primer sets. The sequence of the target gene primers included HVEM primers: 5′-TCT CTG CTG CCA GAC A-3′ and 5′-GCC ACA GCA GAA CAG A-3′, Nectin-1 primers: 5′-TCC TTC ACC GAT GGC ACT ATC C-3′ and 5′-TCA ACA CCA GCA GGA TGC TC-3′, PILR-α primers: 5′-AAG GTC AGC AGC GGA CTA AA-3′ and 5′-CAG TCT TGA GAG GGC TGT CC-3′, beta-actin (ACTB) primers: 5′-GTG ATG GTG GGA ATG GGT CAG-3′ and 5′-TTT GAT GTC ACG CAC GAT TTC C-3′. The qRT–PCR reactions were performed using ABI Prism 7500 system (Applied Biosystems) with the following parameters; repeat 1 (1 cycle), 2 min at 50 °C; 2 min at 95 °C and repeat 2 (40 cycles), 15 s at 95 °C and followed by 30 s at 60 °C, and the data were collected at each 60 °C. PCR products were ran on a 1% agarose gel, stained with ethidium bromide and illuminated with UV light. Expected PCR gene product sizes were 1,270 bp for HVEM, 738 bp for nectin-1 and 250 bp for PILR-α.

Immunocytochemistry

To visualize various receptors on the cell surface, monolayers of HCE cells were plated on chamber slides (Lab-Tek; Nunc, Rochester, NY) at 100% confluency. After a 1× PBS wash, cells were incubated with anti-nectin-1 (1:50 dilution), anti-HVEM (1:50 dilution), or anti-PILR-α (1:25 dilution) primary antibody for 90 min at 4 °C. After the primary incubation period, cells were washed three times with a cold solution of PBS containing 3% BSA for 10 min. Next, cells were incubated with secondary antibody. The secondary antibody used for the anti-nectin-1 and anti-HVEM primary antibody was FITC-conjugated anti-mouse IgG (Sigma) while the secondary used for anti-PILR-α antibody was FITC-conjugated anti-rat IgG (Sigma); all secondary antibodies were administered at a concentration of 1:1,000 and had an incubation period of 45 min at 4 °C. Cells were then washed three times with a cold solution of PBS containing 3% BSA for 10 min. Final products were imaged using laser-scanning spectrum confocal microscopy (TCS SP2; Leica, Wetzlar, Germany). HCE cells treated with only secondary antibody as well as primary antibodies replaced with isotypes served as negative controls. A monoclonal anti c-Myc IgG in generated in mouse and a rat anti c-Myc IgG (Abcam) provided the isotype controls.

Flow cytometry

Flow cytometry was used to confirm the presence of viral entry receptors nectin-1, HVEM, and PILR-α on the cell surface visualized by immunocytochemistry. Monolayers of HCE cells plated on 6-well tissue culture dishes were transferred to 1.5 ml Eppendorf tubes using Hank's based enzyme-free cell dissociation buffer (Invitrogen) and were then washed three times for 10 min with a cold solution of 1× PBS with 10% glucose and 10% calf serum. Cells were then incubated at 4 °C for 90 min with either anti-nectin-1 (1:100 dilution), anti-HVEM (1:100 dilution), or anti-PILR-α (1:100 dilution) primary antibody. Cells were again washed three times for 10 min with a cold solution of 1× PBS with 10% glucose and 10% calf serum and then incubated for 45 min at 4 °C with the appropriate FITC-conjugated secondary antibody as noted above at a concentration of 1:1,000. After incubation of secondary antibody, the cells had a final wash for 10 min with a cold solution of 1× PBS with 10% glucose and 10% calf serum. Flourescence-activated cell sorter (FACS) analysis was then performed.

Antibody blocking of receptors

HCE cells plated on a 96-well tissue culture dish (approximately 2×10⁴ cells per well) were treated with serial dilutions of the appropriate antibody (anti-nectin-1, anti-HVEM, or anti-PILR-α) to block their corresponding receptor. The dilutions of the antibodies used ranged from 1:25 to 1:1600. Following incubation at room temperature for 2 h, viral entry assays were performed using HSV-1(KOS)gL86 virus as per the protocol stated above. Cells not blocked with primary antibody were used as controls.

Re-infection

HCE cells plated on a 96-well tissue culture dish (approximately 2×10⁴ cells per well) were treated with a primary viral dose of HSV-1(KOS) at a concentration of 0.01 PFUs/cell. After 48 h of incubation at 37 °C, cells were treated with a serial dilution of a secondary viral dose of HSV-1(KOS)gL86. Viral entry assays were subsequently performed as per the protocol above. Cells treated with only the secondary dose of virus served as controls. To determine the proportion of cells infected during the primary infection, cell staining was performed. HCE cells plated in a 6-well culture dish at full confluency were treated with a viral dose of HSV-1(KOS) at a concentration of 0.01 PFUs/cell. After 48 h of incubation at 37 °C, cells were rinsed with 1× PBS wash. Primary HSV-1 polyclonal antibody (Dako, Carpinteria, CA) was added at a 1:50 dilution and incubated for 1 h at room temperature. After a 1× PBS wash, endogenous peroxidase activity was blocked by adding 3% hydrogen peroxide for 10 min at room temperature. After another 1× PBS wash, secondary anti-HSV-1 antibody conjugated to horseradish peroxidase (1:200 dilution) was added and incubated for 30 min at room temperature. After a final wash, DAB was added. Cells infected with virus were stained brown. The proportion of infected cells in 100 high power fields were determined and averaged.

Flow cytometry post-primary infection

HCE cells plated on a 6-well tissue culture dish were treated with a viral dose of HSV-1(KOS)RFP at a concentration of 0.01 PFUs/cell. After 48 h of incubation at 37 °C, the monolayers of cells were transferred to 1.5 ml Eppendorf tubes. Before FACS analysis was performed, cells were treated with propidium iodide to distinguish live from dead cells. Flow cytometry was then selectively performed to test levels of nectin-1 on living cells tagged with red flourescent protein as per the method described above. The concentration of anti-nectin-1 primary antibody was 1:100 and the concentration of anti-mouse IgG secondary antibody was 1:1,000. HCE cells untreated with a viral dose were used as a control.

Statistical analysis

All of the experiments were conducted a minimum of 3 times with results at a steady level. Quantitative data was expressed as a mean±standard deviation.

HSV-1 entry and replication in HCE cells

As shown in Figure 1A, viral entry into HCE cells compared to that of CHO-K1 cells followed a dose-dependent pattern, with maximum entry occurring before maximum viral concentration. Upon establishing initial data on viral entry using these results, viral entry was verified using the insoluble substrate X-gal. In Figure 1B, β-galactosidase activity was not observed in the naturally resistant CHO-K1 cells, whereas β-galactosidase activity was observed in HCE cells (Figure 1C).

After establishing HSV-1 viral entry into HCE cells, we examined the ability of HSV-1 to successfully replicate within the cells using a time point plaque assay. As seen in Figure 2A, the number of plaques followed a time-dependent pattern for HCE cells and Vero cells (positive control). Image confirmation of plaques after 18 h post-infection in Vero and HCE cells are demonstrated in Figure 2B,C, respectively. Figure 2D confirms that plaques were not formed in the naturally resistant CHO-K1 cells at 18 h post-infection. Plaque size was also measured at the time intervals of 12, 18, and 24 h and plotted in Figure 2E. An increase in average plaque size at successive time points was observed indicating the ability of virus to spread to adjacent uninfected cells, which is also observed during a primary or recurrent infection in the cornea [26].

pH dependency of viral entry

HSV-1 entry can be pH dependent or independent depending on cell-type [18]. Bafilomycin and chloroquine are both lysosomotropic agents that increase pH and therefore influence cellular uptake in pH dependent mechanisms (e.g., endocytosis). As shown in Figure 3, both bafilomycin and chloroquine strongly inhibited HSV-1 entry in a dose-dependent manner; when concentrations of these lysosomotropic agents were increased, measurements of viral entry decreased. This suggests that HSV-1 entry into HCE cells requires acidification.

Identification and expression of entry receptors

To explore the molecular basis for entry, we decided to examine entry receptor expression by quantitative RT–PCR. It was evident that genes for HVEM, Nectin-1 and PILRalpha were clearly expressed by HCE cells (Figure 4A). Densitometry confirmed gene expression of all three receptors as well (Figure 4B). The control, CHO-K1 cells, showed possible expression of PILRalpha but not HVEM or nectin-1. As expected, both cell-types expressed β-actin. The results from CHO-K1 cells are consistent with previous findings [21]. Next, we decided to verify the gene expression results by protein expression analysis using immunofluorescence and flow-cytometry. As shown in Figure 5A,C,E, respectively, HCE cells express nectin-1, HVEM, and PILR-α along the cell membrane. Cells treated with only FITC-conjugated secondary antibody as negative controls (Figure 5B,D,F, respectively) did not show fluorescence and likewise, isotype control experiments also did not result in any signals (data not shown). The expression of all three receptors was also confirmed by flow cytometry and is demonstrated in Figure 6. While the intensity of positive cells in the x-axis is somewhat low, these results were reproducible and confirm our findings from immunofluorescence (Figure 5). The expression of nectin-1 was most abundant, followed by HVEM and PILR-α.

Antibody blocking and down-regulation of receptors

Reduction of viral entry with blocking by receptor-specific antibodies would indicate the role of those receptors in pathogenesis. The results shown in Figure 7 indicate that there is remarkably diminished entry when any of the receptors (nectin-1, HVEM, and PILR-α) are blocked. Data was recorded as percent blocking, meaning the percentage that entry was reduced compared to controls. Maximum percent blocking of approximately 45%, 30%, and 20% were seen when the cells were treated with anti-nectin-1 (Figure 7A), anti-HVEM (Figure 7B), and anti-PILR-α (Figure 7C) antibodies, respectively. Further reduction was seen when multiple receptors were blocked – there was approximately 50% blocking when the anti-nectin-1 and anti-HVEM antibodies were used in coordination (Figure 7D) and nearly 70% blocking when all three antibodies were used (Figure 7E). Absolute numbers for blocking are shown in Table 1.

Nectin-1 expression may reduce re-infection

The re-infection experiment was performed to indicate whether HSV-1 entry was affected by a previous low-level infection. The results shown in Figure 8 demonstrate a decrease in HSV-1 (gL-86) entry into cells that have already undergone a primary infection with wild-type HSV-1 (KOS) (superinfected) compared to control, which were not exposed to HSV-1(KOS) before HSV-1(gL86) expression. To analyze the percentage of cells that were infected at this time point (48 h after primary infection), cells were stained for β-galactosidase expression from the secondary viral genome [27]. The number of cells infected in by secondary reporter virus averaged at 27±4% (data not shown) compared to about 100% infected control cells. To gain insight into a possible mechanism for the decrease in infection, flow cytometry was performed post-primary infection to determine nectin-1 expression compared to control. The results in Figure 9 show a decrease in the level of nectin-1 expression on living cells 48 h post-primary infection compared to living control cells that underwent no primary infection and instead were simply incubated for 48 h. This result, however, has to be seen in the light of the observation that overall expression of nectin-1 goes up significantly from about 24 h of incubation (Figure 6) to after 48 h of incubation. Thus, a prior infection by a HSV-1 strain can reduce the nectin-1 expression relative to uninfected cells.