Roles Played by Toll-like Receptor-9 in Corneal Endothelial Cells after Herpes Simplex Virus Type 1 Infection

Abstract

Purpose.: To determine the roles played by toll-like receptor 9 (TLR9) in cultured human corneal endothelial (HCEn) cells after herpes simplex virus type 1 (HSV-1) infection and to characterize the TLR9-mediated antiviral responses.

Method.: Immortalized HCEn cells were examined for TLR expression. The upregulation of inflammatory cytokines after HSV-1 infection was determined by real-time RT-PCR or protein array analyses. The TLR9-mediated HSV-1 replication was determined by real-time PCR and plaque assay. To determine whether there was an activation of the signal transduction pathway, HCEn cells that were transfected with pathway-focused transcription factor reporters were examined for promoter activity.

Results.: TLR9 was abundantly expressed intracellularly in HCEn cells. The CpG oligonucleotide, a TLR9 ligand, stimulated the NF-κB activity in HCEn cells. HSV-1 infection also stimulated NF-κB and induced NF-κB -related inflammatory cytokines, including RANTES, IP-10, MCP-2, MIF, MCP-4, MDC, MIP-3α, IL-5, TARC, MCP-1, and IL-6. The induction of these cytokines was significantly reduced by blocking the activity of TLR9. In addition, viral replication in HCEn cells was significantly reduced by the inhibition of TLR9, but was preserved by a concomitant activation of the NF-κB cascade. Of the different HSV-1–induced inflammatory cascade–related transcription factors, TLR9 was found to activate NF-κB, cyclic AMP response element (CRE), and the CCAAT-enhancer-binding proteins (C/EBP) the most.

Conclusions.: Corneal endothelial cells transcriptionally initiate inflammatory programs in response to HSV-1 infection related to NF-κB, CRE, and C/EBP and express arrays of inflammatory cytokine induction by TLR9. On the other hand, HSV-1 exploits TLR9-mediated NF-κB activation for its own replication.

The tissues of the ocular surface help maintain the clarity of the cornea and protect the eye from numerous environmental pathogens or dead cell constituents. The endothelial cells lining the inner surface of the cornea are also responsible for maintaining the optical clarity of the cornea. Because endothelial cells do not replicate in vivo, a decrease in their density can lead to blinding bullous keratopathy. ^1,2  

Generally, mucosal surfaces are armed with pattern-recognition receptors (PRRs) for sensing foreign materials. The PRRs recognize various types of ligands, such as bacterial and fungal cell wall components, bacterial lipoproteins, and nucleic acids derived from bacteria, virus, and self. ³ An invasion by viruses is recognized by the toll-like receptor (TLR) family, and the recently recognized categories of intracellular PRRs that detect nucleic acids in the cytoplasm. ⁴ The retinoic acid-inducible gene I (RIG-I) and melanoma differentiation-associated gene 5 (MDA5) detect the RNA of pathogens. DNA-dependent activator of interferon-regulatory factors (DAI), absent in melanoma 2 (AIM2), RNA polymerase III, leucine-rich repeat flightless-interacting protein 1 (LRRFIP1), and interferon γ–inducible protein 16 (IFI16) detect intracellular DNA. 

The TLRs were the first discovered and major category of PRRs. The recognition of pathogens by TLRs leads to the induction of innate immunity, inflammation, and adaptive immunity. The TLRs on the mucosal body surface function not only to keep bacteria from invading the body but also to form mutually beneficial relationships. ⁵ TLRs recognize commensal or pathogen-associated molecular patterns to control the function of the mucosal surface cells. For example, the TLRs regulate the proliferation of epithelial cells after intestinal injury. An absence of TLRs significantly impairs the repair of the epithelial barrier. ⁵ Signaling by the TLRs leads to increased inflammation and promotes the development of inflammation-associated neoplasia. ⁶ Thus, intricate interactions operate for the host and microbes by the many functions of the TLRs. 

Corneal endothelial cells have been recently found to act as immune modulators that suppress T cell receptor-mediated CD4⁺ T cell proliferation. They also stimulate the conversion of CD8⁺ T cells into regulatory T cells. ^7,8 These functions may contribute to the immune privilege of the eye. TLRs are especially recognized as important modulators of innate and acquired immunity. Thus, understanding how the endothelial cells behave after TLR stimulation may provide important clues on how to control immune-mediated diseases. 

TLR9 is a well-known sensor of the nucleic acids of viruses and microbes. HSV-1 is the most common viral pathogen permissive to the corneal endothelial cells, and an infection by HSV-1 is manifested as herpetic keratitis. To recognize herpesvirus, the host uses a distinct repertoire of TLRs. First, the surface glycoproteins ligate to TLR2. ^9,10 Second, the DNAs of herpesvirus which are rich in CpG sequences, stimulate TLR9. ^11,12 And third, double-strand RNAs, generated through self-hybridization of viral genes, activate TLR3. ^13,14 TLR9 has been reported to be a crucial component in corneal epithelial cells that recognize the HSV-1 infection. ^15,16 However, the roles played by TLRs in corneal endothelial cells have not been determined. 

The activation of TLR9 can also cause collateral damage or exacerbation of immune-mediated diseases. For example, when self nucleic acids activate TLR9 chaperoned by anti-DNA autoantibody, ³ the TLR9 activation initiates or exacerbates autoimmune diseases. ^{17 –19} Thus, understanding the roles played by TLR9 may help develop effective strategies to prevent unwanted inflammatory responses in the anterior chamber or corneal endothelium. 

The purpose of this study was to determine the response of the TLR9 in cultured human corneal endothelial (HCEn) cells after herpes simplex virus type 1 (HSV-1) infection and to characterize the TLR9-mediated anti-viral responses. We shall show that HCEn cells express TLR9 intracellularly, and HSV-1 infection leads to the upregulation of arrays of inflammatory cytokines mediated by TLR9. Especially important was that the NF-κB cascade downstream of TLR9 can be hijacked by HSV-1 and diverted for its own replication. 

Materials and Methods

Cells

An HCEn cell line was established as described in detail. ⁷ The HCEn cells were propagated to confluence on 6- or 96-well plates in Dulbecco's modified Eagle's medium (DMEM; Gibco, Grand Island, NY) supplemented with 10% fetal bovine serum. Primary corneal endothelial cells were obtained from the corneoscleral rims of donor eyes after the central cornea was used for keratoplasty. 

The procedures used conformed to the tenets of the Declaration of Helsinki. 

Viruses

Confluent monolayers of Vero cells were infected with the KOS strain of HSV-1 (generous gift from Kozaburo Hayashi, National Institutes of Health [NIH], Bethesda, MD) After 1 hour of adsorption, the medium containing the virus was aspirated, and the monolayers of cells were refed with fresh HSV-1–free medium. After attaining the maximum cytopathic effect (48–72 hours after infection), the medium was discarded, and cells with the small amount of remaining medium were frozen, thawed, and sonicated. The supernatant was collected after centrifugation at 3000 rpm for 10 minutes and overlaid onto a sucrose density gradient (10–60% wt/vol). The solution was centrifuged with a swing rotor (SW28; Beckman Instruments, Fullerton, CA) for 1 hour at 11,500 rpm. The resultant visible band at the lower part of the gradient containing the HSV-1 was washed using centrifugation at 14,000 rpm for 90 minutes and resuspended in a small volume of serum-free DMEM. The virus was then divided into aliquots and stored at −80°C until use. To infect the HCEn cells with HSV-1, the HCEn cells were adsorbed with the sucrose-density gradient purified virus stock for 1 hour at a multiplicity of infection (MOI) of 0.01 to 1, and refed with fresh medium. 

Flow Cytometry

Flow cytometry was used to determine the degree of TLR expression using the following monoclonal antibodies (mAbs): TLR2 (Alexis, Plymouth Meeting, PA), TLR4 (Monosan, Uden, Netherlands), TLR3 (abCam, Cambridge, UK), and TLR9 (Oncogene, San Diego, CA). Mouse isotype IgG was used as the control. FITC-conjugated anti-mouse IgG₁ or IgG₂ (BD Pharmingen, Franklin Lakes, NJ) was used as the secondary antibody. 

For flow cytometric analysis of the surface expression of TLRs on the HCEn cells, a suspension of subconfluent cells was obtained by adding 0.5% trypsin/EDTA to the HCEn cells and incubated with anti-TLR antibodies. This was followed by incubation with FITC-conjugated anti-mouse IgG (BD PharMingen). For intracellular staining of the TLRs, HCEn cell suspensions were permeabilized (Cytofix/Cytoperm; BD Biosciences) before staining. After they were washed twice in PBS, the stained cells (live-gated on the basis of the forward and side scatter profile and propidium iodide exclusion) were analyzed by flow cytometry. 

Luciferase Reporter Assays

HCEn cells were transfected with luciferase reporter plasmids for AP-1, C/EBP, CRE, Elk-1, ISRE, NFAT, or NF-κB (Agilent, Santa Clara, CA). For the, internal control, HCEn cells were co-transfected with pRL-CMV (Promega, Madison, WI; using Geneporter 3000 transfection reagent; Genlantis, San Diego, CA). 

For inhibition of TLR-9, TLR-9 inhibitory oligonucleotide (forward 5′-TCCTGGCGGGGAAGT-3′) (Alexis, San Diego, CA) or TLR-9 siRNA (Qiagen, Hilden, Germany) was used. For activation of the NF-κB cascade, the IκBα on the HCEn cells was inhibited by IκBα siRNA (Invitrogen, Carlsbad, CA). For transfection of siRNA, HCEn cells were transfected (RNAifect; Qiagen) 2 days after transfection of the reporter plasmids, according to the manufacturer's protocol. HCEn cells were infected with HSV-1 48 hours after siRNA transfection. The luciferase activity was measured using the dual-luciferase reporter assay system (Promega). 

The target sequences of the siRNA were TLR-9 siRNA: forward 5′-CGGCAACTGTTATTACAAGAA-3′, and IκBα siRNA: forward 5′-GAGCTCCGAGACTTTCGAGGAAATA-3′. 

Pharmacologic Inhibition of NF-κB Cascade

An IKK inhibitor peptide or control peptide (Merck, Darmstadt, Germany) was used to block the IκB kinase activity. The IKK inhibitor peptide contained a sequence corresponding to the active IκB phosphorylation recognition sequence. For inhibition of NF-κB p65, NF-κB p65 (Ser276) inhibitor peptide or control peptide (Imgenex, San Diego, CA) was used. 

Real-Time RT-PCR

Total RNA was isolated from HSV-1–infected HCEn cells and reverse transcribed (QuantiTect Reverse Transcription Kit; Qiagen). The cDNAs were amplified and quantified on a thermal cycler (LightCycler; Roche, Mannheim, Germany) using a PCR kit (QuantiTect SYBR Green; Qiagen). The sequences of the real-time PCR primer pairs were VEGF: forward 5′-GCAGCTTGAGTTAAACGAACG-3′, reverse 5′-GGTTCCCGAAACCCTGAG-3′; IL-6: forward 5′-GATGAGTACAAAAGTCCTGATCCA-3′, reverse 5′-CTGCAGCCACTGGTTCTGT-3′; HSV-1 DNA polymerase: forward 5′-CATCACCGACCCGGAGAGGGAC-3′, reverse 5′-GGGCCAGGCGCTTGTTGGTGTA-3′; and glyceraldehyde-3-phosphate dehydrogenase (GAPDH); forward 5′-AGCCACATCGCTCAGACAC-3′, reverse 5′-GCCCAATACGACCAAATCC-3′. 

To ensure equivalent loading and amplification, all products were normalized to GAPDH transcript as an internal control. 

Enzyme-Linked Immunosorbent Assay

To determine the levels of secreted IL-6, supernatants collected from HSV-1–infected HCEn cells were assayed with a commercial ELISA kit (Peprotech, Rocky Hill, NJ). 

For inflammatory cytokine and chemokine profiling after HSV-1 infection, supernatants were collected from HCEn cells 12 hours post infection (pi) and assayed with human cytokine antibody arrays (RayBiotech, Norcross, GA). This analysis determined the level of expression of 80 cytokines and chemokines. The intensity of the chemiluminescence signals was digitized (LAS-1000plus; Fujifilm, Tokyo, Japan, and MultiGauge software ver.2.0; Fujifilm) and normalized by using the positive control signals in each membrane. 

Pathways Analysis

The set of extracted genes was analyzed for transcriptional networks of molecular events using computerized pathway analysis (Pathways Analysis 7.0; Ingenuity Systems, Redwood, CA; based on the Ingenuity Pathways Knowledge Base). The resulting networks were evaluated by the significance scores, which were expressed as the negative logarithm of the P value. The obtained score indicate the likelihood that the assembly of a set of focus genes in a network could be explained by random chance alone. 

Statistical Analyses

Data are presented as the mean ± SEM. Statistical analyses were performed using t-tests or ANOVA, as appropriate. 

Results

TLR9 Expression in HCEn Cells

We used flow cytometry to determine whether TLRs are expressed on HCEn cells grown in culture or primary HCEn cells, because TLRs can be expressed on the cell surface in nonhematopoietic lineage cells. No significant cell surface expression was observed (data not shown). 

Next, we assessed the intracellular expression of TLRs by staining permeabilized HCEn cells and primary corneal endothelial cells. TLR9 was found to be significantly expressed intracellularly, whereas expression of TLR2, TLR3, and TLR4 was barely detectable (Fig. 1). 

Figure 1.

 

Intracellular expression of TLRs in HCEn cells. TLR 9 was significantly expressed in HCEn cells and primary human corneal endothelial cells. Expression of TLR2, -3, and -4 was barely detectable. Solid line: unstained; dotted line: control IgG stained, gray line: anti-TLR antibody stained.

TLR9-Mediated NF-κB Promoter Activation in HCEn Cells after HSV-1 Infection

To determine whether the input from TLR9 is functional, we examined whether TLR9 ligand activates the NF-κB cascade, since NF-κB is the representative signaling cascade of TLR-mediated signaling. When TLR9 was stimulated with B class CpG oligonucleotide, a TLR9 ligand, there was a significant upregulation of NF-κB promoter activity, indicating that TLR9 is functional in HCEn cells (Fig. 2). 

Figure 2.

 

NF-κB promoter activation in HCEn cells by TLR9. HCEn cells transfected with NF-κB reporter plasmids were stimulated with CpG oligonucleotide for 24 hours and measured for luciferase activity. CpG transfection significantly elevates the NF-κB promoter activity (n = 6; *P < 0.05).

We next evaluated whether HSV-1 infection would activate the NF-κB cascade in HCEn cells and whether TLR9 plays a role in this activation. We found that HSV-1 infection significantly stimulated the promoter of NF-κB as early as 6 hours pi, and the level of expression continued to increase up to 12 hours pi (Fig. 3) The elevated NF-κB promoter activity was significantly reduced by an inhibition of TLR9. Thus, the TLR9 cascade that stimulates NF-κB is activated after HSV-1 infection. 

Figure 3.

 

Inhibition of HSV-1 infection–induced NF-κB promoter by TLR9 inhibition. HCEn cells transfected with reporter plasmids were stimulated with HSV-1 infection for 6 (A) and 12 (B) hours, and measured for luciferase activity. Treatment by TLR9 inhibitory oligonucleotide significantly inhibited NF-κB promoter activation (n = 6; *P < 0.05).

TLR9-Mediated Inflammatory Cytokine and Chemokine Induction in HCEn Cells after HSV-1 Infection

Next, we examined whether TLR9 is involved in the induction of cytokines and chemokines in HCEn cells after HSV-1 infection. After HCEn cells were infected with HSV-1, the level of IL-6 transcript was significantly increased at 12 hours pi (i.e., the IL-6 expression relative to GAPDH was 1.4 × 10⁻⁴ ± 1.0 × 10⁻⁵ at an HSV-1 MOI of 0.1 and 2.0 × 10⁻⁶ ± 3.1 × 10⁻⁷ for mock infection (P < 0.01). The level was lower at 24 hours pi: 1.1 × 10⁻⁴ ± 4.0 × 10⁻⁶ IL-6/GAPDH at an HSV-1 MOI of 0.1 and 2.3 × 10⁻⁶ ± 3.3 × 10⁻⁷ for mock infection. 

To examine the contribution of TLR9 to the IL-6 induction, HCEn cells treated with TLR9 inhibitory oligonucleotide were infected with HSV-1 and assessed for IL-6 induction by real-time PCR. The HSV-1 infection significantly elevated IL-6 induction at 12 hours pi (Figs. 4A, 4B). The level of IL-6 after HSV-1 infection was significantly reduced by blocking TLR9 with a TLR9 inhibitory oligonucleotide (Fig. 4A). Inhibition of TLR9 by siRNA transfection also had a similar inhibitory effect on the IL-6 induction (Fig. 4B). 

Figure 4.

 

Inhibition of HSV-1 infection–induced IL-6 activation. Effect of inhibiting HSV-1 infection–induced IL-6 mRNA induction by TLR9 inhibitory oligonucleotide treatment (A) and by transfection of siRNA of TLR9 (B) at 12 hours. TLR9 blockade by inhibitory oligonucleotide or siRNA significantly reduced the HSV-1 infection–induced IL-6 mRNA activation. n = 4; *P < 0.01. (C, D) Reduction of HSV-1 infection–induced IL-6 secretion by TLR9 blockade. TLR9 inhibitory oligonucleotide significantly reduced IL-6 secretion at 12 (C) and 24 (D) hours pi in a dose-dependent manner (n = 6; *P < 0.01).

We then determined whether HSV-1 infection can stimulate IL-6 secretion through TLR9. When supernatants of HSV-1–infected HCEn cells were assessed for IL-6 by ELISA, we found that HSV-1 infection significantly stimulated IL-6 secretion (Figs. 4C, 4D). This HSV-1 infection–induced IL-6 secretion was significantly suppressed by a TLR9 inhibitory oligonucleotide in a dose-dependent manner (Figs. 4C, 4D). 

Next, we assessed how HSV-1 infection modulated the cytokine and chemokine milieu of HCEn cells through TLR9. Supernatants from HSV-1–infected HCEn cells were assayed for 80 cytokines and chemokines using protein array analysis and were tested for their sensitivity to TLR9 inhibition. Twenty cytokines and chemokines were significantly upregulated after HSV-1 infection, and of them,TLR9 inhibition significantly reduced the upregulation of RANTES (CCL5), IP-10 (CXCL10), MCP-2 (CCL8), MIF, MCP-4 (CCL13), MDC (CCL22), MIP-3α (CCL20), IL-5, TARC (CCL17), and MCP-1 (CCL2) (Fig. 5). 

Figure 5.

 

TLR9-mediated inflammatory cytokine and chemokine induction by HSV-1–infected HCEn cells. HCEn cells were adsorbed with HSV-1 at an MOI of 0.1 for 1 hour and refed with the DMEM. After 12 hours of incubation, the supernatant of HSV-1–infected HCEn cells was assayed for cytokines. TLR9-induced inflammatory cytokines and chemokines were significantly reduced by exposure to TLR9 inhibitory oligonucleotide (n = 4/group; *P < 0.05, **P < 0.01).

Role of TLR9-Mediated NF-κB Activation and HSV-1 Replication

TLR9 participates in the primary defense systems against viral infection and functions to induce inflammatory cytokines after infection of HCEn cells by HSV-1. We examined whether TLR9 affects the entry and replication of HSV-1 into HCEn cells. To accomplish this, HSV-1 was adsorbed on HCEn cells, and the number of HSV-1 copies was determined by real-time PCR of HSV-1 DNA polymerase. After TLR9 was inhibited by pretreatment with TLR9 inhibitory oligonucleotide, a significant reduction in the copy number was not observed (Fig. 6A). 

Figure 6.

 

Inhibition of HSV-1 proliferation by blockade of TLR9 and NF-κB signaling cascade. (A) Unperturbed entry of HSV-1 into the HCEn cells by TLR9 inhibition. HSV-1 was adsorbed on HCEn cells for 1 hour. HCEn cells were washed and assessed for HSV-1 DNA polymerase copy number by using real-time PCR (n = 8). TLR9 inhibitory oligonucleotide did not appreciably affect HSV-1 absorption. (B) TLR9 inhibitory oligonucleotide impaired HSV-1 replication, shown by the reduction in copy number of HSV-1 DNA polymerase mRNA (n = 4, 24 hours pi). (C) Reduced proliferation of HSV-1 by IKK inhibitor. HCEn cells were infected with HSV-1 at the indicated MOI and assessed at 24 hours pi for copy number of HSV-1 DNA polymerase mRNA, with reverse transcription real-time PCR (n = 4; P < 0.01). (D) Restoration of TLR9 inhibition–mediated reduction of HSV-1 proliferation by IκBα inhibition. Treatment of TLR9 inhibitory oligonucleotide significantly reduced the copy number of HSV-1 DNA polymerase mRNA at 24 hours pi. This reduction was restored by NF-κB activation using transfection of siRNA of IκBα (n = 4; *P < 0.05, **P < 0.01).

The contribution of TLR9 to viral replication was determined by real-time RT-PCR, and the results showed a significant reduction in the expression of the mRNA of HSV-1 DNA polymerase in HCEn cells after exposure to TLR9 inhibitory oligonucleotide (Fig. 6B). This reduction was also confirmed by titration (control oligo-treated at an MOI of 0.1: 9.3 × 10⁸ ± 0.3 × 10⁸; TLR9 inhibitory oligonucleotide-treated at an MOI of 0.1: 9.8 × 10⁷ ± 0.9 × 10⁷, P < 0.01). 

We then assessed whether the TLR9-mediated viral replication in HCEn cells was related to NF-κB activation. The classic NF-κB cascade is regulated by IκB kinase (IKK), leading to the nuclear translocation of p65, a component of the NF-κB pathway. When HCEn cells were treated with IKK inhibitory peptide, which contained sequences corresponding to the active IκBα phosphorylation recognition sequence, the induction of HSV-1 DNA polymerase was significantly inhibited in HSV-1–infected HCEn cells (Fig. 6C). Treatment with a p65 inhibitor also significantly reduced the number HSV-1 copies (data not shown). These findings indicate that the classic NF-κB cascade is involved in HSV-1 replication in HCEn cells. 

We next examined whether TLR9-inhibition–mediated suppression of HSV-1 replication can be restored by NF-κB activation. Because the activation of the classic NF-κB cascade is regulated by the degradation of IκBα, the NF-κB cascade is activated by siRNA-mediated inhibition of IκBα. Exposure to TLR9 inhibitory oligonucleotide reduced the copy number of HSV-1 DNA polymerase mRNA, and the transfection of IκBα siRNA reduced the effect of TLR9 inhibition (Fig. 6D). Collectively, these findings indicate that HSV-1 used the TLR9-mediated NF-κB activation for its own replication in HCEn cells. 

Alternative Transcription Factor Activation by TLR9 in HSV-1–Infected HCEn Cells

HSV-1 infection induces an array of inflammatory cascades. This can be summarized by the transcriptional induction profiles of representative transcriptional factors. To characterize the profiles of the signaling cascades activated by HSV-1 infection and show the possible involvement of TLR9, we determined whether HSV-1 infection can activate representative transcriptional factors related to the TLR9 cascades by using transfection of reporter plasmids. The activities of transcriptional factors of cascades of NF-κB, MAPK/ERK, cAMP/PKA, MAPK/JNK, C/EBP, interferon response, and PKC/calcium were measured using reporter plasmids for NF-κB, Elk-1, cyclic AMP response element (CRE), AP-1, C/EBP, ISRE, and NFAT, respectively. HSV-1 infection significantly stimulated the transcription factors of NF-κB, Elk-1, CRE, AP-1, C/EBP, and NFAT at 24 hours pi (Fig. 7). 

Figure 7.

 

Signaling cascade–focused promoter activation in HCEn cells by HSV-1 infection. HCEn cells transfected with reporter plasmids were stimulated with HSV-1 infection for 24 hours at an MOI of 0.1 and measured for luciferase activity for (A) NF-κB, (B) Elk-1, (C) CRE, (D) AP-1, (E) C/EBP, (F) ISRE, and (G) NFAT. HSV-1 infection significantly elevated promoter activities of NF-κB, ELK-1, CRE, AP-1, C/EBP, and NFAT (n = 6; *P < 0.05).

We then tested whether TLR9 contributes to the induction of transcription factor activities of the inflammatory cascades. When TLR9 was inhibited by siRNA transfection, the HSV-1–induced activation of CRE and C/EBP reporters was significantly reduced (Fig. 8). The other transcription factor activities, including Elk-1, AP-1, and NFAT were not appreciably affected (data not shown). Thus, HCEn cells used TLR9 leading to various types of promoter activation, including NF-κB, after HSV-1 infection. 

Figure 8.

 

Inhibition of HSV-1 infection–induced CRE and C/EBP promoter activation by TLR9 inhibition. HCEn cells transfected with reporter plasmids were infected with HSV-1 at an MOI of 0.1 and measured for luciferase activity at 12 hours pi. Transfection of siRNA of TLR9 significantly inhibited the reporter activities of CRE (A) and C/EBP (B). (n = 6; *P < 0.05).

TLR9-Mediated Inflammatory Network after HSV-1 Infection

To summarize how HCEn cells used TLR9-mediated signals after HSV-1 infection, the TLR9-dependent cytokines induced after HSV-1 infection were analyzed for signaling interactions using a systems biological approach. Using a database of known signaling networks (Ingenuity Pathways Knowledge Base; Ingenuity Systems), we successfully generated two major biological networks with high significance scores (network 1, P < 10⁻³¹; network 2, P < 10⁻¹⁵). The most significant network was network 1, which was annotated as cell-to-cell signaling and interaction, hematologic system development and function, and immune cell trafficking, where NF-κB was centrally positioned (data not shown). 

Discussion

Our results showed that TLR9 was abundantly expressed in HCEn cells and was used to initiate inflammatory responses after HSV-1 infection. HSV-1 exploited the TLR9-mediated NF-κB activation for its own replication. To resist the assault, HCEn cells transcriptionally initiate an array of inflammatory programs related to the cascades of NF-κB, ERK, MAPK (P38), JNK, cAMP/PKA, PKC, and interferon responses. Of these, TLR9 activation was especially used for the signal transduction cascades of NF-κB, CRE, C/EBP, and arrays of inflammatory cytokines, including IL-6. 

In sensing microbial pathogens, conserved structural moieties are recognized by germline encoded PRRs, including the TLRs, NOD-like receptors, and C-type lectin receptors. ²⁰ Apoptotic or necrotic cells or degradation products of the extracellular matrix, damage-associated molecules or cytokines, such as dsDNA, RNA, high-mobility group box 1 (HMGB1), ATP, hyaluronan, versican, heparin sulfate, and heat shock proteins, are abundantly present. These damage-associated molecular patterns (DAMPs) are also recognized by PRRs. Of the different PRRs, the TLRs are the most important class of receptors that are able to sense pathogen-associated molecular patterns (PAMPs). Nucleic acids, especially DNAs, are a major class of molecules that stimulate TLRs. Previously, the DNAs derived from bacteria had been considered the exclusive ligand of TLR9. However, viral genomes and self DNAs derived from necrotic or apoptotic cells have also been shown to activate TLR9. Physiologically, ligands of TLRs, including TLR9, are ubiquitous, and the corneal endothelium is continuously exposed to various components of PRR ligands. Thus, the cornea and the host are exposed to and sense the environment using combinations of PRRs. In this setting, cascades initiated from such PRRs generally converge to NF-κB or inflammasomes, where the converged signal inputs can elicit robust immune responses in synergy. ²¹  

For entry of HSV-1 into the host, glycoproteins, gB, gD, gH, and gL are required. For example, gB binds to paired immunoglobulin-like type 2 receptor α (PILRα) on the host. ²² gD binds to herpesvirus entry mediator (HVEM), nectin-1, or 3-O sulfated heparan sulfate, after which the host recognizes the viral invasion by the PRRs. In the TLR-mediated recognition cascade, three major molecular components—TLR2, TLR9, and TLR3—are engaged to activate innate immune responses. ²³ However, the TLR-mediated interaction does not appear to affect viral entry (Fig. 6A). The sequential recognition of TLR2 and -9 that occurs after HSV-1 infection leads to a robust NF-κB activation which then induces a wide array of cytokines, chemokines, and interferons, where NF-κB plays a central role in regulating numerous cellular metabolic events. Concomitantly, HSV-1 redirects the host transcriptional machinery to express its own genes in a tightly regulated temporal cascade. ²⁴ The three classes of genes, the immediate-early (IE) genes, including ICP-0, -4, -22, -27, and -47, followed by the early and the late genes are sequentially expressed. 

To initiate productive replication of HSV-1, ICP0 plays a crucial role as a strong activator of all classes of HSV-1 genes and a propagator of lytic infections. ICP0 possesses NF-κB-binding elements on its promoter. The transcription of ICP0 is dependent on activation of NF-κB of the host, which is triggered by the recruitment of p65/RelA. ²⁴ Inhibition of the NF-κB cascade, including the inhibition of IKK or dominant negative IκBα, significantly suppresses viral replication (Fig. 6). ^{24 –26} Very recently, the UL31 of HSV-1 was also shown to be necessary for optimal NF-κB activation and expression of ICP4, ICP8, and glycoprotein C. ²⁷  

The use of host NF-κB for viral replication is not limited to HSV-1 because NF-κB-binding sites are also located in the genome of different members of the herpes virus family. ^28,29 Moreover, HSV-1 is equipped with the ability to effectively block multiple innate signaling for its survival. For example, virion host shutoff protein (VHS) degrades the host mRNA by its RNase function, US11 or γ34.5 inhibits PKR (RNA-activated protein kinase), and ICP47 inhibits MHC class I loading. ^{30 –37} After the viral replication is completed, ICP-0 directs the inhibition of inflammatory responses by ubiquitin-specific peptidase 7 (USP7) translocation, which leads to the inhibition of NF-κB and JNK. ³⁰ Thus, HSV-1 hijacks and exploits the crucial components of the host immune system, TLR9 and NF-κB, for its own use. 

There are two major signaling pathways for TLR: NF-κB and MAPKs. In the MAPK cascade, the JNK, p38, and ERK pathways are conventionally activated, leading to the activation of AP-1, CRE, Elk-1, and C/EBP elements in the promoters. In addition, the C/EBP family of transcription factors is involved in many biological functions, including regulating cytokine expression, proliferation, and tumor progression. ^{38 –40} We found that the reporter activity of NF-κB, CRE, and C/EBP are activated by TLR9 after HSV-1 infection. Analysis of the HSV-1 infection–induced transcriptome of HCEn cells showed strong inductions of CREBBP and C/EBPα, which are representative transcription factors related to CRE and C/EBP. 

Generally, transcriptional activation is regulated by different levels of transcriptional factor activation and interactions. On infection by Helicobacter pylori, the AP-1 and CRE elements in the cyclooxygenase promoter are activated by TLR2 and -9. ⁴¹ In TLR-mediated activation of IL-6 and TNF-α, both NF-κB and C/EBP binding elements in the promoter are critical for their transcriptional activation. ^{42 –44} In HSV-1–infected HCEn cells, TLR9 input activated the NF-κB signal transduction cascade (Fig. 3), and our bioinformatic analysis of the induced cytokines and chemokines which are sensitive to TLR9 inhibition, were summarized as NF-κB-dependent inflammatory cascade. However, the NF-κB cascade may not be sufficient to fully explain the transcriptional activation of inflammatory cytokines. In HCEn cells, the activations of CRE, C/EBP, and NF-κB were involved in the TLR9-mediated signaling cascade (Fig. 8) and presumably in the TLR9-mediated induction of inflammatory cytokines and chemokines. At least two of the recognition sequences of these transcription factors exist in the promoters of TLR responsive cytokines and chemokines (data not shown). This may explain the unexpectedly wide array of inflammatory cytokines that was inhibited by TLR9 suppression. 

We used immortalized HCEn cells as models of corneal endothelial cells in situ. The HCEn cells have similar capabilities as primary corneal endothelial cells and organ cultured corneal endothelial cell in inducing representative cytokines including MCP-1, IL-6, IL-8, CXCL2, TGFβ2, and thrombospondin 1. ^8,45 However, there is still some question of whether immortalized HCEn cells can truly reflect the in vivo properties of corneal endothelial cells such as HSV-1 infection–induced endotheliitis. For this, in vivo analysis may help in gaining a better understanding of the physiological roles of the endothelial cells during a viral infection. 

At present, a murine model of HSV-1–induced corneal endotheliitis is not available. We used the KOS strain for this study because our initial hypothesis was based on the findings of our earlier studies. ^46,47 Very recently, the KOS strain has been reported to have a mutation of the US8A gene and defective US9 gene. ⁴⁸ US9 is especially involved in neuronal virulence. However, a defective US9 does not appear to affect the cell-to-cell spread in permissive epithelial cells. ⁴⁹ In addition, no apparent dysfunction was reported for the elongated US8A by mutation. 

To summarize, corneal endothelial cells express TLR9 intracellularly to recognize dsDNAs and HSV-1 infection. HSV-1 usurps this TLR-mediated NF-κB activation for its own replication. 

Footnotes

 Supported by Grant-in-Aid 20592076 and 21592258 for Scientific Research from the Japanese Ministry of Education, Science, and Culture.

Footnotes

 Disclosure: S. Takeda, None; D. Miyazaki, None; S. Sasaki, None; Y. Yamamoto, None; Y. Terasaka, None; K. Yakura, None; S. Yamagami, None; N. Ebihara, None; Y. Inoue, None

The authors thank Duco Hamasaki for editing this article.