Herpes simplex virus 1 infection induces the expression of proinflammatory cytokines, interferons and TLR7 in human corneal epithelial cells

Abstract

Herpetic epithelial and stromal keratitis is a sight-threatening ocular infection. To study the role of the epithelium in the innate response to herpes simplex virus 1 (HSV-1) infection of the cornea, we used a telomerase-immortalized human corneal epithelial cell (HCEC) line, HUCL, and primary HCECs as a model and infected the cells with HSV-1 (KOS strain). HSV-1 infection of HCECs resulted in a two-phase activation of nuclear factor-kappaB (NF-κB), JNK and p38, with the first peak at 1–4 hr and a second peak at 8 hr. Concomitant with the first peak of activation, transcriptional expression of interleukin (IL)-6, IL-8, tumour necrosis factor (TNF)-α and interferon (IFN)-β was rapidly induced in HSV-1-infected cells. HSV-1 infection also induced the production of IL-6, IL-8, and TNF-α in both HUCL cells and primary HCECs. Coincident with the second phase of NF-κB activation in HSV-1-infected HCECs, the expression of Toll-like receptor 7 (TLR7) was induced, whereas the level of TLR3 was greatly down-regulated. Thus, in response to HSV-1 infection, HCECs produce proinflammatory cytokines, leading to infiltration, and IFNs to enhance the antiviral activity in the cornea, probably through sequential activation of TLRs.

Introduction

Herpes simplex keratitis is a disease initiated by infection of the epithelial and stromal layers of the cornea with herpes simplex virus 1 (HSV-1), resulting in infiltration of neutrophils and mononuclear lymphocytes into the stromal layer. Although the inflammatory response is necessary to clear the organism from the infected tissue, it can also be destructive to the host cornea, leading to scar formation and vision loss. ^1–6 Corneal epithelial cells constitute the first line of defence against microbial pathogens, including viruses, and therefore should possess the ability to respond to their presence. It is reasonable to propose that, in response to HSV-1 infection, the epithelial cells send initial signals, such as the release of proinflammatory cytokines, to recruit neutrophils and mononuclear lymphocytes into the cornea. Epithelial cells could also produce factors such as type 1 interferons (IFNs) to enhance the antiviral activity of residential cells of the cornea.

Recent studies have indicated that epithelial recognition of pathogens is largely attributable to an evolutionarily conserved family of receptors, the Toll-like receptors (TLRs), that function in innate immunity via the recognition of pathogen-associated molecular patterns.^7,8 Pattern recognition by TLRs then leads to cytokine production and the expression of costimulatory molecules, which can constitute protective innate and adaptive immunity.⁹ To date, 11 TLRs have been identified, and agonists have been identified for most, but not all, of them.^10,11 Of interest in terms of viral infection, TLR3 has been shown to be activated by polyinosine-polycytidylic acid (poly I:C), which structurally mimics the double-stranded RNA associated with viral infection and has long been used to induce IFN-α/β expression.¹² TLR3 transmits signals through the Toll/interleukin (IL)-1 receptor homology (TIR)-containing adaptor to activate nuclear factor (NF)-κB and induce proinflammatory cytokine and IFN-β expression.¹³ Similar to bacterial DNA, HSV-1 DNA contains abundant deoxycytidylate phosphatedeoxyguanylate (CpG) motifs. These motifs stimulate multiple cellular components of the immune system through TLR9. ^11–14 TLR9 transduces intracellular signals through myeloid differentiation factor 88 (MyD88), which initiates a signalling cascade leading to NF-κB activation and cytokine production. Thus, both TLR3 and TLR9 might provide pathways that link HSV-1 recognition in mammalian cells, although the role of these TLRs in viral infection in vivo remains to be determined.^14,15 Imidazoquinoline compounds are therapeutics that have antiviral properties (by inducing the expression of IFN-α). These compounds were recently shown to signal through TLR7.¹⁶ Unlike other TLRs such as TLR3, 4 and 9, TLR7 expression is restricted to the IFN-producing plasmacytoid dendritic cells in humans and is induced in macrophages upon viral infection.¹⁷ Recently, Heil et al.¹⁸ and Diebold et al.¹⁹ demonstrated that TLR7 and TLR8 recognize the single-stranded RNAs (ssRNAs) found in many viruses, leading to IFN-α production in virus-infected macrophages and dendritic cells. Furthermore, TLR7 recognizes the ssRNA viruses in vivo and mice deficient in TLR7 have reduced responses to in vivo infection with vesicular stomatitis virus (VSV).²⁰ Expression of TLR7 at the protein level in non-myeloid cells has not been reported.

To understand the role of epithelial cells in the innate response to HSV-1 infection of the cornea and the involvement of TLRs, we used an in vitro infection model and observed that the infection of human corneal epithelial cells (HCECs) resulted in the activation of NF-κB, p38, and JNK. Associated with NF-κB activation, expression and secretion of proinflammatory cytokines, IL-6, IL-8 and tumour necrosis factor (TNF)-α were also up-regulated. Interestingly, at a later stage, when TLR3 expression is down-regulated, the expression of TLR7 is up-regulated, suggesting a sequential activation of TLRs during HSV-1 infection of HCECs.

Materials and methods

HCEC culture

Human telomerase-immortalized corneal epithelial (HUCL) cells, kindly provided by Dr Rheinwald and Dr Gipson (Harvard Medical School, Boston, MA),^21,22 were maintained in defined keratinocyte-serum free media (SFM) (Invitrogen Life Technologies, Carlsbad, CA) in a humidified 5% CO₂ incubator at 37°. For viral infection, after cells were grown to 90% confluence, the medium was replaced with Keratinocyte Basic Medium (KBM; BioWhittaker, Walkersville, MD) for 12 hr (growth factor starvation overnight).

To confirm the results obtained with HUCL cells, primary HCECs were isolated from human donor corneas obtained from the Georgia Eye Bank (Atlanta, GA). The epithelial sheet was separated from the underlying stroma after overnight treatment with dispase (2·5 U/ml; Sigma-Aldrich, St. Louis, MO) at 4°. The dissected epithelial sheets were trypsinized and cells collected by centrifugation (500 g for 5 min). HCECs were cultured in T25 flasks coated with fibronectin-collagen (FNC; 1 : 3 mixture) coating mix (Athena Environmental Service, Inc., Baltimore, MD) and used at passage 4.

Virus infection

Stocks of HSV-1 (KOS strain) used in this study were propagated on Vero cells (American Type Culture Collection, Manassas, VA) grown in complete Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS) and 1% penicillin and streptomycin. The titre of virus stocks was determined by standard plaque assay on Vero cells and titres were expressed as plaque-forming units (PFU)/ml. Stocks were stored at −70° in 1-ml aliquots, and a fresh aliquot of stock virus was thawed and used for each experiment. To infect cells, epithelial cultures were infected at a multiplicity of infection of 5 or 0·5. After 1 hr of adsorption at 37°, the inoculum was removed and cells were replenished with KBM. At the indicated time, cells were processed for RNA preparation and immunoblotting, and conditioned media were collected for cytokine determination.

In the experiment with inhibitors, cells were pretreated with the inhibitor for 30 min, and then infected with HSV-1. The inhibitors (Calbiochem, San Diego, CA) used included the NF-κB inhibitor MG-132 (5 µm), the p38 inhibitor SB203580, and/or the JNK inhibitor SP600125. Several concentrations (up to 10 µm) for the latter two inhibitors were tested for their effects on TLR7 expression (see Fig. 5 below).

Figure 5.

Toll-like receptor (TLR) expression in human telomerase-immortalized corneal epithelial (HUCL) cells during herpes simplex virus 1 (HSV-1) infection. (a) HUCL cells and (b)primary human corneal epithelial cells (HCECs) were infected with HSV-1 at 5 plaque-forming units (PFU) per cell and were cultured for the indicated times post-infection (p.i.). (c) HUCL cells were infected with HSV-1 (HSV-1) with no infection as control (control) at 5 PFU per cell in the absence (–) or the presence of the p38 inhibitor SB203580 (SB, 5 µm), the JNK inhibitor SP600125 (SP, 10 µm), or the nuclear factor (NF)-κB inhibitor MG-132 (MG, 10 µm) for 10 hr. At each time-point, cells were lysed or the cell membrane was prepared by ultracentrifugation (for TLR3) and subjected to sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) followed by immunoblotting with antibodies against TLR7, TLR3, TLR9, TLR2, and/or smooth muscle actin (control), using a chemical luminescent technique. These results are representative of three independent experiments.

RNA isolation and reverse transcriptase–polymerase chain reaction (RT-PCR) analysis

Total cellular RNA was isolated with TRIzol reagent (Invitrogen, Carlsbad, CA) and treated with RNase-free DNase I. Total RNA (2 µg) was reverse-transcribed with a first-strand synthesis system for RT-PCR (Invitrogen). cDNA was amplified by PCR using the specific primers listed in Table 1. IL-6 was amplified in 25 cycles with an annealing temperature of 55°, IL-8 was amplified in 22 cycles with an annealing temperature of 55°, TNF-α was amplified in 25 cycles with an annealing temperature of 60°, IFN-α was amplified in 32 cycles with an annealing temperature of 55°, and IFN-β was amplified in 35 cycles with an annealing temperature of 55°. The PCR products for IL-6, TNF-α, ΙFN-α and IFN-β were subjected to electrophoresis on 1·5% agarose gels containing ethidium bromide. The products for IL-8 and its internal control 18S RNA were separated on 12% polyacrylamide gel electrophoresis (PAGE) and stained with SYBR green (Molecular Probes, Inc., Eugene, OR). DNA gel images were captured using the Kodak EDAS 290 system.

Table 1.

Primer sequences and length of amplicons

Gene	Primer sequences	Product size (bp)
IL-6	Forward CTCCTTCTCCACAAGCGCCTTC	583
	Reverse GCGCAGAATGAGATGAGTTGTC
IL-8	Forward CACCGGAAGGAACCATCTCA	72
	Reverse GGAAGGCTGCCAAGAGAGC
TNF-α	Forward GAAAGCATGATCCGGGACGTG	510
	Reverse GATGGCAGAGAGGAGGTTGAC
IFN-α1	Forward TTTCTCCTGCCTGAAGGACAG	373
	Reverse TCTCATGATTTCTGCTCTGACA
IFN-β	Forward CGCAGTGACCATCTATGAGA	332
	Reverse TTCGGAGGTAACCTGTAAGT
18S	Forward ACATCCAAGGAAGGCAGCAG	65
	Reverse TTTTCGTCACTACCTCCCCG
GAPDH	Forward CACCACCAACTGCTTAGCAC	515
	Reverse CCCTGTTGCTGTAGCCAAAT

¹Amplifying isoforms 5 and 8.

IL, interleukin; TNF, tumour necrosis factor; IFN, interferon.

Determination of IL-6, IL-8, and TNF-α secretion from HUCL cells

Secretion of IL-6, IL-8 and TNF-α was determined by enzyme-linked immunosorbent assay (ELISA). After growth factor starvation, HUCL cells were infected with HSV-1. At the indicated time, supernatants were harvested for ELISA analysis. Measurement of IL-6, IL-8 and TNF-α by ELISA was performed according to the manufacturer's instructions (R&D Systems, Minneapolis, MN). The amount of cytokines in the culture media was normalized to the total amount of cellular protein lysed with radioimmune precipitation assay (RIPA) buffer [150 mm NaCl, 100 mm Tris-HCl, pH 7·5, 1% deoxycholate, 0·1% sodium dodecyl sulphate (SDS), 1% Triton X-100, 50 mm NaF, 100 mm sodium pyrophosphate, 3·5 mm sodium orthovanadate, proteinase inhibitor cocktails, and 0·1 mm phenylmethylsulphonyl fluoride (PMSF)]. The protein concentration of cell lysates was determined with the Micro BCA kit (Pierce Biotechnology, Rockford. IL). Results were expressed as pg cytokine per mg cell lysate (mean ± standard error; n = 3). P-values were determined by analysis of variance (ANOVA).

Western blot analysis

HSV-1-infected HUCL cells were lysed with RIPA buffer, and the protein concentration was determined with the Micro BCA kit (Pierce). NF-κB activation was determined by IκB-α phosphorylation and degradation using anti-IκB-α or anti-phospho-IκB-α antibody, respectively (Cell Signaling Technology, Beverly, MA). Phospho-JNK1/2, phospho-p38, TLR7, TLR2, TLR3 and TLR9 expression was also assessed by Western blotting using anti-phospho-JNK1/2 and phospho-p38 (Cell Signaling Technology), and anti-TLR7 (Santa Cruz Biotechnology, Santa Cruz, CA), anti-TLR2 (Imgenex, San Diego, CA), anti-TLR3 and anti-TLR9 (Calbiochem, San Diego, CA) antibodies.

Immunohistochemical staining

HUCL cells were cultured on 8-well chamber slides and infected with HSV-1. At the indicated time-points, cells were photographed for morphological changes under an inverted microscope, and then fixed with 4% freshly made formaldehyde. Cells were stained with fluorescein isothiocyanate (FITC)-labelled anti-human HSV-1 antibody (DakoCytomation Denmark A/S, Glostrup, Denmark). For double immunofluorescent staining, rabbit anti-human HSV-1 and goat anti-TLR7 antibodies (Santa Cruz Biotechnology) were used. Mounting medium (Vectorshield; Vector Laboratories, Burlingame, CA) with 4′,6-diamidino-2-phenylindole (DAPI) was used to stain nuclei.

Results

HSV-1 infection induces NF-κB activation in HUCL cells

HUCL cells were infected with HSV-1 (KOS strain) at a multiplicity of infection of 5 PFU/ml. While there were no noticeable morphological changes 4 hr post-infection (p.i.) compared with the control, uninfected cells, by 8 hr p.i. a few round cells were seen, and by 24 hr p.i. most cells exhibited rounding (Fig. 1); however, the cells stayed in that state for up to 3 days (the longest time tested) without significant loss of cell numbers. The virus-infected cells were identified using anti-HSV-1 antibody. Most HUCL cells stained positively, but the staining intensity varied between cells at 8 h p.i. By 24 hr p.i., most cells were intensely stained.

Figure 1.

Immunostaining of herpes simplex virus 1 (HSV-1)-infected human telomerase-immortalized corneal epithelial (HUCL) cells. HUCL cells (control, uninfected cells, a–c) were infected with HSV-1 (KOS strain) at 5 plaque-forming units (PFU) per cell and were cultured for 4 hr (d–f), 8 hr (g–i), and 24 hr (j–l) post-infection (p.i.). At each time-point, cells were photographed (a, d, g, j) and then fixed in 1% paraformaldehyde/phosphate-buffered saline, followed by staining with an HSV-1 antibody (b, e, h, k) or DAPI for nuclei (c, f, i, l).

Activation of NF-κB, a transcription factor and master regulator of inflammation,²³ in response to HSV-1 infection was assessed in HCECs by immunodetection of IκB-α phosphorylation and degradation (Fig. 2). HSV-1 infection resulted in IκB-α phosphorylation and degradation in a time-dependent manner in both HUCL cells and primary HCECs. An elevated level of phospho-IκB-α was first detected 1 hr p.i. Phosphorylation appeared to be down-regulated by 4 hr p.i., followed by a slight increase at 8 hr. At 12 hr p.i., phospho-IκB-α was undetectable and there was a decrease in total IκB-α as well as p65, a subunit of NF-κB. By 24 hr p.i., p65 was barely detectable and no IκB-α immunoreactivity was found (Fig. 2). A similar pattern of NF-κB activation was also observed in primary HCECs in response to HSV-1 infection (Fig. 2b).

Figure 2.

Herpes simplex virus 1 (HSV-1) infection stimulates nuclear factor (NF)-κB, p38 and JNK activation in human telomerase-immortalized corneal epithelial (HUCL) cells. HUCL cells (a) or primary human corneal epithelial cells (HCECs) (b) were infected with HSV-1 at 5 plaque-forming units (PFU) per cell and were cultured for the indicated times post-infection (p.i.). At each time-point, cells were lysed and subjected to sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE), followed by immunoblotting with paired antibodies using a chemiluminescent technique to determine IκB-α phosphorylation and degradation with antibodies against IκB-α (IκB) and phospho-IκB-α (pIκB); the levels of p65 of NF-κB with anti-p65; JNK phosphorylation with antibodies against JNK (JNK) and phospho-JNK (pJNK); and p38 phosphorylation with antibodies against p38 (p38) and phospho-p38 (pP38). Results are representative of two independent experiments.

Interestingly, unlike p65 of NF-κB, the levels of JNK and p38, two other signalling effectors associated with TLR activation, remained unchanged in HSV-1-infected cells up to 24 hr p.i. (Fig. 2). However, HSV-1 infection resulted in two phases of activation of JNK as assessed by JNK tyrosine phosphorylation. Relatively low but clearly detectable levels of JNK phosphorylation were observed within 8 hr p.i., while much higher levels were detected from 8 to 24 hr p.i. Low levels of phospho-p38 were observed in control, uninfected cells, which increased at 4 hr p.i. Similar to JNK, much higher levels of phospho-p38 were observed 8 hr p.i and thereafter. Taken together, these results suggest that there are two phases of response of HUCL cells during HSV-1 infection, corresponding to exposure of epithelial cells to the virus (early) and replication (later) of virus in cells.

HSV-1 infection-associated cytokine expression and secretion in HUCL cells

To assess the biological relevance of induced NF-κB activation, we measured the effect of HSV-1 infection on HCEC proinflammatory and antiviral cytokine expression and production (secretion). Infection-induced expression of cytokines was determined first by RT-PCR (Fig. 3). The expression of IL-6, IL-8 and TNF-α was up-regulated in infected cells, peaking at 2 hr p.i. and declining thereafter. By 12 hr p.i., levels of IL-6 and IL-8 transcripts had returned to control values. However, an elevated level of TNF-α mRNA was still observed at 24 hr p.i.

Figure 3.

Herpes simplex virus 1 (HSV-1)-induced cytokine and interferon (IFN) mRNA expression in human corneal epithelial cells. Human telomerase-immortalized corneal epithelial (HUCL) cells were infected with HSV-1 at 5 plaque-forming units (PFU) per cell and were cultured for the indicated times. Total RNA was extracted, reverse-transcribed, and amplified using the primers listed in Table 1 with GAPDH as control. PCR products were separated and stained as described in Materials and methods. Results are representative of three independent experiments.

We also assessed the expression of the type I interferons IFN-α/β in response to HSV-1 infection (Fig. 3). The expression profile of IFN-β was similar to that of TNF-α; the expression was induced rapidly and persisted for up to 24 hr p.i. For IFN-α, there was a low level of mRNA in uninfected cells with a slight down-regulation of basal expression consistently observed at 1 hr p.i. Expression was then elevated at 2 and 4 hr p.i and further up-regulated to a level much higher than the control at 8 hr p.i and thereafter. The level of GAPDH remained unchanged in the control during the course of the study.

The effects of HSV-1 infection on IL-6, IL-8 and TNF-α secretion in HCECs cells were assessed by ELISA (Fig. 4). Very low levels of IL-6 (Figs 4a and b), IL-8 (Figs 4c and d) or TNF-α (Figs 4e and f) accumulated in the culture medium at 8 and 24 hr for control, uninfected HUCL cells (Figs 4a, c and e) and primary HCECs (Figs b, d and f). IL-6 secretion gradually increased in HSV-1-infected HCECs (Figs 4a and b). While the pattern of HSV-1-induced IL-6 expression was similar in HUCL cells and in primary HCECs, the level of IL-6 accumulated in the culture medium of primary cells infected with HSV-1 was much lower than that in HUCL cells. An earlier study showed that primary HCECs failed to produce IL-8 in response to HSV-1 infection.²⁴ As shown in Fig. 4, HSV-1 infection induced primary HCECs to secrete comparable amounts of IL-8 in a time–course similar to that of the cultured HUCL cells (Figs 4c and d). Furthermore, a significant increase in TNF-α secretion was observed 4 hr p.i and thereafter (up to 24 hr p.i., the longest time studied) in both cell types.

Figure 4.

Time–course of herpes simplex virus 1 (HSV-1)-induced interleukin (IL)-6 and IL-8 secretion in human corneal epithelial cells (HCECs). Human telomerase-immortalized corneal epithelial (HUCL) cells (a, c, e) and primary HCECs (b, d, f) were infected with HSV-1 at 5 plaque-forming units (PFU) per cell and were cultured for the indicated times. IL-6 (a, b), IL-8 (c, d) and tumour necrosis factor (TNF)-α (e, f) secretion was measured in cell culture supernatants by enzyme-linked immunosorbent assay (ELISA). All values are expressed as mean ± standard deviation. Statistically significant differences in secreted IL-6 (ng/mg cell lysate), IL-8 (ng/mg cell lysate) and TNF-α (pg/mg cell lysate) in infected HUCL cells were determined by analysis of variance with probabilities shown for the overall significance when compared with medium collected after 8 and 24 hr culture of uninfected cells. Data shown are representative of three experiments.

Induction of TLR7 expression in HSV-1-infected HCECs

Several TLRs, including TLR2, 3, 7 and 9, have been suggested to be involved in the recognition of infection with HSV-1^25,26 or other viruses,^15,20,27 and thus we investigated the expression of TLRs in cultured HCECs (Fig. 5). In normal, uninfected HUCL cells, both TLR9 and 2 were readily detected in cell lysate, whereas TLR3 was detected only in the enriched membrane fraction. During the course of HSV-1 infection, the levels of TLR9 and 2 remained unchanged, whereas TLR3 levels decreased 8 hr p.i. and were undetectable at 12 hr p.i. As expected, cultured HUCL cells did not express TLR7, which is classically found in type I IFN-producing plasmacytoid dendritic cells^28,29 as well as eosinophils.³⁰ Eight hours post-HSV-1 infection, TLR7 was detected at the protein level, and abundant expression was observed at 12 and 24 hr p.i., nearly paralleling the disappearance of TLR3 immunoreactivity. Similarly, no TLR7 was detected in uninfected primary HCECs and infection of the cells with HSV-1 resulted in induction of TLR7 at the protein level, although the TLR7 expression was induced at an earlier time in primary HCECs when compared with HUCL cells (Fig. 5b).

The presence of the NF-κB inhibitor MG-132, but not the p38 inhibitor SB203580 or the JNK inhibitor SP600125, in the culture medium blocked HSV-1-induced TLR7 expression in HUCL cells (Fig. 5c). Two other NF-κB-specific inhibitors, isohelenin and kamebakaurin, exhibited similar effects on HSV-1-induced TLR7 expression (data not shown).

HSV-1-induced TLR7 expression was further confirmed by immunocytochemistry (Fig. 6). Most cells were stained intensely for HSV-1 proteins at 8 hr p.i., while some cells exhibited light or no HSV-1 staining (arrows, Fig. 6b). The cells with intense HSV-1 staining were also TLR7 positive. Furthermore, TLR7 was located intracellularly, consistent with its role as an endosomal membrane-associated receptor for viral nucleic acids.³¹ The cells lacking HSV-1 staining, however, were also TLR7-negative, suggesting a close association of TLR7 expression with HSV-1 infection or replication.

Figure 6.

Expression of Toll-like receptor 7 (TLR7) in herpes simplex virus 1 (HSV-1)-positive human telomerase-immortalized corneal epithelial (HUCL) cells. HUCL cells in chamber slides were infected with HSV-1 at 5 plaque-forming units (PFU) per cell for 8 hr and were then incubated with anti-HSV-1 visualized with rhodamine (b), anti-TLR7 visualized with fluorescein isothiocyanate (FITC)-conjugated anti-mouse immunoglobulin G (IgG) (c), or DAPI for nuclear staining (a). Cells lacking strong HSV-1 staining (arrows) were also negative for TLR7 staining.

Discussion

We investigated the response of cultured primary HCECs and HUCL cells to HSV-1 infection in vitro. We demonstrated that HSV-1 infection activates the NF-κB, p38 and JNK signalling pathways in HCECs and observed two phases of activation of these signalling pathways in response to infection. The infection-induced expression of the proinflammatory cytokines IL-6, IL-8 and TNF-α at the mRNA level was associated with the first peak of NF-κB and JNK activation and declined 8 hr p.i. or thereafter. Production of these cytokines was also observed in HSV-1-infected HCECs. Furthermore, mRNAs of the type I IFNs were also induced in HUCL cells in response to HSV-1 infection. The up-regulation of IFN-β expression continued throughout the course of the study (24 hr p.i.), whereas the expression of IFN-α appeared to be up-regulated 8 hr p.i. HUCL cells express multiple TLRs that are known to be involved in cell responses to viruses, including TLR2, 3 and 9, but not TLR7. HSV-1 infection induced the expression of TLR7, an innate immunity receptor of virus that recognizes viral ssRNA (mouse but not human TLR7)^18,19 in both HUCL cells and primary HCECs. Thus, our data suggest that corneal epithelial cells possess the ability to recognize HSV-1 infection and play a role in innate immune responses in the cornea and subsequent inflammation in herpes simplex keratitis.

Of 11 TLRs, roles for TLR2, 3, 7 and 9 in response to virus infection have been established either in functional TLR-deleted mice or in immune cells.^{12,25,32–37} Epithelial cells are the primary site for viruses such as HSV. We and others have observed that multiple TLRs are expressed in corneal epithelial cell lines and in primary HCECs including TLR2, 3 and 9.^38–41 The presence of these virus-recognizing TLRs would render the epithelial cells able to respond to viral infection in vitro and in vivo. Using the phosphorylation of IκB-α, JNK1/2 and p38 as markers for the activation of TLR-mediated signalling pathways, we documented that epithelial cells are able to respond to HSV-1 challenge. Thus, we suggest that one or more TLRs are involved in corneal epithelial recognition of and response to HSV-1 infection.

The activation and the translocation of NF-κB to the nuclei induced by TLR-mediated signalling pathways result in the transcription of proinflammatory cytokine genes. Indeed, in our cultured epithelial cell model, the transcriptional expression of IL-6, IL-8 and TNF-α was rapidly induced, coincident with the activation of NF-κB, in response to HSV-1 infection. These cytokines were also secreted from the infected HUCL and primary cultured HCE cells, as measured by ELISA. Studies by Lausch et al. using excised mouse corneas revealed that the corneal epithelial layer is the most abundant producer of IL-6.⁴² Kanangat et al. reported that HSV-1 infection selectively up-regulates IL-6 gene expression in the murine epithelial-like cell line EMT-6.⁴³ The secretion of IL-6 has been linked to the recruitment of neutrophils to the virus infection site.^42,44 Our study confirmed the results of these earlier studies on HSV-1-induced IL-6 expression in corneal epithelial cells. Like IL-6, TNF-α is a potent proinflammatory cytokine and functions in the mediation of the acute-phase response, chemotaxis, and activation of inflammatory and antigen-presenting cells.^45,46 TNF-α is produced mainly by activated macrophages and T cells and has been reported to be released by the epithelial cells of the mouse central cornea in response to lipopolysaccharide.⁴⁶ It has been shown to be an important factor in the pathogenesis of murine recurrent herpetic stromal keratitis.⁴⁷ The demonstration that HSV-1 infection induces TNF-α expression and secretion suggests that the epithelium is an important source for the cytokine in vivo. IL-8 is a potent chemoattractant for both neutrophils and lymphocytes. An earlier study failed to detect any HSV-1-induced IL-8 expression at both the mRNA and protein levels in primary cultures of HCECs.²⁴ However, a recent study using a HCEC line revealed that, at the early phase of HSV-1 infection, the cells produced neutrophil chemoattractants that were neutralized by anti-IL-8 antibodies, suggesting that corneal epithelial cells play a role in inducing neutrophil chemotaxis through the production of IL-8.⁴⁸ Our results showed that HSV-1 infection induces the expression and secretion of IL-8, along with IL-6 and TNF-α, both in primary culture and in telomerase-immortalized HCECs. The reason for the discrepancy among the studies of IL-8 synthesis in HSV-1-infected HCECs is not clear. One possible explanation is that the strains used by the three groups are different; strain RE was used by Oakes et al.,²⁴ strain MP by Miyazaki et al.⁴⁸ and strain KOS in this study. Nevertheless, the ability of the epithelium to produce multiple proinflammatory cytokines in response to HSV-1 infection indicates that the epithelium is an important part of the innate response system for the cornea and plays a role in eliciting the infiltration of inflammatory cells into the tissue.

We also demonstrated that HSV-1 infection induces the expression of the type I interferons IFN-β and IFN-α. Type I interferon (IFN-α/β) induction is a primary defence system for virus infection.^49,50 In vitro studies suggested that TLR3 recognizes dsRNA, a signature pattern specific to viruses, and induces the phosphorylation of IRF-3, leading to IFN-β induction via the TLR adapter Toll interleukin-1 (TIR)-domain-containing adaptor molecule (TICAM-1) or TIR-domain-containing adaptor molecule (TRIF).^51,52 IFN-β production has been suggested to contribute to the induction of IFN-α. Using ELISA, we were unable to detect either IFN-β or IFN-α in the culture media of infected HUCL cells, perhaps because the levels of the cytokines released into the medium in response to HSV-1 infection were below the measurable range of the method. The produced IFN-α/β may function in an autocrine or paracrine fashion to enhance the antiviral activity of the epithelial cells. Indeed, we recently showed that exogenously added IFN-α or IFN-β induces the expression of the antiviral genes myxovirus resistance gene 1 and 2′,5′ oligoadenylate synthetase in HUCL cells.³⁸

The most striking observation of the study is the induced expression of TLR7. Recently, TLR7 was reported to recognize ssRNA and induce IFN-α expression in vitro^18,19 and to sense the presence of ssRNA viruses (vesicular stomatitis virus and influenza virus) in vivo.²⁰ TLR7 exhibits restricted expression and is abundantly expressed in macrophages infected with virus and in natural IFN-α/β-producing cells (IPCs). IPCs play a central role in innate immunity against microbial infections. Thus, TLR7 has been suggested to function as a sensor in immune cells for viral infection.^31,32,53 The expression of TLR7 by non-myeloid cells is highly contentious; its expression was detected in human uterine epithelial cells⁵⁴ and corneal epithelial cells³⁹ but not in human epithelial A549 cells or human umbilical vein endothelial cells (HUVECs)⁵⁵ by RT-PCR. No TLR7 expression at the protein level in non-myeloid cells was reported. Indeed, we detected TLR7 expression by Western blotting in uninfected HCECs, both in HUCL cells and in primary HCECs. However, HSV-1 infection induced abundant TLR7 expression at the protein level in these cells. The primary HCEC cultures we used were at passage 4, maintained in serum-free KGM for over 1 month, and should be largely free from myeloid cell contamination. The virus-induced TLR7 expression was further confirmed by double staining of HSV-1 proteins and TLR7 in HUCL cells. Thus we conclude that HSV-1 infection induces TLR7 expression in HCECs. To our knowledge, this is the first report of TLR7 expression in non-myeloid cells at the protein level in response to viral infection.

The induced TLR7 expression was coincident with the down-regulation of TLR3, a receptor for dsRNA, suggesting that TLR3 and TLR7 may function sequentially in these cells to regulate epithelial response to viral infection and replication. To date, the mechanisms underlying TLR expression in epithelial cells have proved elusive. It would therefore be of great interest to determine how the expression of TLR7 and TLR3 is regulated in HCECs in response to viral infection. Furthermore, the time–course of TLR7 expression suggests a role for the protein in the induction of the second peak of the activation of signalling pathways. In our in vitro infection model, the up-regulation of IFN-α appears to be correlated to the induction of TLR expression in HSV-1-infected cells. Interestingly, the activation of TLR7 by ssRNA has been shown to induce IFN-α expression and secretion in dendritic cells and macrophages. Thus, TLR7 may function similarly in epithelial cells in the induction of IFN-α expression in response to viral infection. It is, however, intriguing that HSV-1 is a dsDNA virus; how this pathogen might induce and activate TLR7 is a question that warrants further investigation.

In summary, our study suggests that corneal epithelial cells recognize HSV-1, probably through the sequential action of multiple TLRs. These cells can also initiate innate immune responses in the cornea by releasing proinflammatory cytokines and chemokines that recruit neutrophils and enhance their antiviral activity by the autocrine/paracrine action of IFNs. Understanding the molecular events of HSV-1–epithelial interactions and the inflammatory consequences of TLR activation may permit the development of novel, specific therapies that can promote innate defence and prevent some of the destructive consequences of herpes simplex keratitis.