November 2006
Volume 47, Issue 11
Free
Cornea  |   November 2006
AsialoGM1-Mediated IL-8 Release by Human Corneal Epithelial Cells Requires Coexpression of TLR5
Author Affiliations
  • Jojo W. Du
    From the Department of Biological Sciences, College of Optometry, State University of New York, New York City, New York; the
  • Fan Zhang
    From the Department of Biological Sciences, College of Optometry, State University of New York, New York City, New York; the
  • José E. Capó-Aponte
    From the Department of Biological Sciences, College of Optometry, State University of New York, New York City, New York; the
  • Souvenir D. Tachado
    Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts; and the
  • Jing Zhang
    Department of Ophthalmology, Wayne State University of Medicine, Detroit, Michigan.
  • Fu-Shin X. Yu
    Department of Ophthalmology, Wayne State University of Medicine, Detroit, Michigan.
  • Robert A. Sack
    From the Department of Biological Sciences, College of Optometry, State University of New York, New York City, New York; the
  • Henry Koziel
    Division of Pulmonary, Critical Care and Sleep Medicine, Department of Medicine, Beth Israel Deaconess Medical Center and Harvard Medical School, Boston, Massachusetts; and the
  • Peter S. Reinach
    From the Department of Biological Sciences, College of Optometry, State University of New York, New York City, New York; the
Investigative Ophthalmology & Visual Science November 2006, Vol.47, 4810-4818. doi:https://doi.org/10.1167/iovs.06-0250
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Jojo W. Du, Fan Zhang, José E. Capó-Aponte, Souvenir D. Tachado, Jing Zhang, Fu-Shin X. Yu, Robert A. Sack, Henry Koziel, Peter S. Reinach; AsialoGM1-Mediated IL-8 Release by Human Corneal Epithelial Cells Requires Coexpression of TLR5. Invest. Ophthalmol. Vis. Sci. 2006;47(11):4810-4818. https://doi.org/10.1167/iovs.06-0250.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

purpose. In this study, it was determined that human corneal epithelial cells (HCECs) express asialoganglioside ganliotetraosylceramide (asialoGM1) and toll-like receptor (TLR)-5, and their interaction induces interleukin (IL)-8 release through Ca2+ transient activation and mitogen-activated protein kinase (MAPK) stimulation.

methods. Expression of asialoGM1 and TLR5 was detected in SV40 HCECs by Western blot and flow cytometry analyses and their association by coimmunoprecipitation. Single-cell fluorescence imaging was used to measure intracellular free Ca2+ transients in fura-2–loaded cells. The enzyme-linked immunosorbent assay (ELISA) was used to quantify IL-8 production in both cultured and primary HCECs.

results. The HCECs expressed both asialoGM1 and TLR5 receptors. Ligation of asialoGM1 resulted in protein–protein interaction with TLR5, followed by transient increases in Ca2+ influx through L-type voltage-dependent Ca2+ channels. This led to P2Y receptor stimulation along with membrane depolarization, resulting from increases in ATP release into the medium. Intracellular Ca2+ transients led to time-dependent extracellular signal-regulated kinase (ERK) MAPK pathway stimulation, followed by a 9.5-fold increase in IL-8 release. Similarly, in primary HCECs, asialoGM1 receptor stimulation resulted in an 8.1-fold increase. With a TLR5 neutralizing antibody, no asialoGM1-induced increases in IL-8 release occurred, and this response was not suppressed in the presence of a TLR2 neutralizing antibody.

conclusions. IL-8 release by HCECs is mediated through ligand-induced asialoGM1 protein–protein interactions with TLR5. This response is dependent on ATP efflux into the medium, followed by P2Y receptor stimulation. Such activation, in turn, results in increases in Ca2+ influx through L-type voltage-dependent Ca2+ channels, as well as stimulation of the ERK pathway.

The pathogen Pseudomonas aeruginosa (PA) is an opportunistic bacterium that can cause bacterial keratitis in contact lens patients, especially with the increased use of extended-wear contact lenses. 1 Corneal epithelial cells (CECs) constitute the first line of defense against microbial pathogens and have the ability to sense the presence of pathogen-associated molecular patterns (PAMPs) present in pathogenic bacteria such as PA. 2 3 Recognition is largely due to the expression of asialoganglioside gangliotetraosylceramide (asialoGM1), which is composed of Galβ1, 2GalNAcβ1, 4Galβ1, 4-Glcβ1, and 1Cer (a glycolipid receptor), 4 5 and in association with toll-like receptors (TLRs), a family of pattern recognition receptors. 6 7 8 9 10 Ligation of these receptors leads to activation of mitogen-activated protein kinases (MAPKs) and nuclear factor (NF)-κB, and production of proinflammatory cytokines. 3 4 11 12 13 Release of these cytokines can augment the infection and contribute to corneal destruction. 14 15  
Asialylated glycolipids, such as asialoGM1, act as coreceptors for many pathogens, including Haemophilus influenzae, Staphylococcus aureus, Streptococcus pneumoniae, Klebsiella pneumoniae, Escherichia coli, and PA, 5 16 as well as PAMPs such as pilin and flagellin. 5 Recognition of PAMPs by asialoGM1 can be mimicked by ligation with asialoGM1 antibody. This interaction stimulates (1) nuclear translocation of NF-κB, (2) phosphorylation of extracellular signal-related kinase (ERK)1/2, 4 and (3) expression of the mucin (MUC-2) 9 10 and interleukin (IL)-8 genes. 17 Because asialoGM1 lacks transmembrane and intracellular domains, it is incapable of direct contact with cytoplasmic signaling molecules. A possible link between asialoGM1 and downstream signaling components is an assemblage of different TLRs. 
TLRs are an evolutionarily conserved family of pattern-recognition receptors that function in innate immunity through recognition of PAMPs. 6 7 8 To date, 11 TLRs types have been identified. 18 19 20 TLR2 agonists include a variety of bacterial cell wall products, such as peptidoglycan (PGN), lipoteichoic acid (LTA), and lipoproteins, 18 19 20 whereas TLR3 agonists are viral double-stranded (ds)RNAs. 21 TLR4 agonists include Gram-negative bacterial lipopolysaccharide (LPS), respiratory syncytial virus protein F, and the plant product taxol. 22 23 Bacterial flagellin has been identified as a TLR5 agonist, 24 and unmethylated CpG-containing DNA as a TLR9 agonist. 25 In the cornea, the expression of TLR2, -4, -5, and -9 have been documented. 3 26 27 Furthermore, colocalization of TLRs and gangliosides has been described in many cell types to function in mediating signal transduction. 28 TLRs are possible candidates for interaction with asialoGM1 as coreceptors. 
CECs are an important component of mucosal defenses, expressing chemokines and cytokines to direct the influx and activation of phagocytic cells in response to bacterial pathogens. Hence, part of the corneal epithelial barrier function depends on its ability to sense and respond to danger signals. Although asialoGM1 serves as a receptor for detecting many pathogens in some tissues, 5 its presence in the corneal epithelium is controversial. 29 30 Earlier studies by Hazlett et al. 29 provided evidence that asialoGM1 is present in the mouse and bovine 31 corneas, where it serves as a receptor for PA and is spatially localized in the wound site. However, immunohistochemical and thin-layer chromatogram immunostaining suggest that the isolated rabbit corneal epithelium contains no detectable levels of asialoGM1, even after attempts to expose potential cryptic sites with trypsin, PA, or PA exoproducts. 30 Furthermore, preliminary immunohistochemical analyses indicate that no asialoGM1 is present in human corneas. 30 Recently, Yamamoto et al. 32 showed that asialoGM1, as a constituent of lipid rafts, participates in PA internalization. Of significance, the more proximal components of the epithelial cell signaling cascade, the receptor, and the kinases involved in signal transduction in the cornea have never been characterized. 
In this study, we determined the effect of asialoGM1 ligation in SV40-immortalized human corneal epithelial cells (HCECs) on Ca2+ mobilization, ERK1/2 phosphorylation, and proinflammatory IL-8 production. We demonstrated that ligation of asialoGM1 induces its heterodimerization with TLR5, inducing adenosine 5′-triphosphate (ATP) release and in turn P2Y receptor stimulation, which leads to L-type voltage-dependent Ca2+ channel activation followed by transient increases in intracellular [Ca2+]. We also showed that subsequent activation of the ERK limb of MAPK cascade leads to very similar increases in release of IL-8 in both SV40-immortalized and primary HCECs. This interaction between asialoGM1 and TLR5 is specific, since it was abrogated during exposure to TLR5 blocking antibody, and TLR2 neutralization did not suppress IL-8 release. 
Methods
Cell Culture
SV40-immortalized HCECs, kindly provided by Kaoru Araki-Sasaki et al. 33 were maintained in Dulbecco’s modified Eagle’s medium (DMEM/F12) supplemented with 10% fetal bovine serum (FBS), 5 ng/mL epidermal growth factor (EGF), 5 μg/mL insulin, and 40 μg/mL gentamicin in a humidified 5% CO2 incubator at 37°C. Primary cultured HCECs, purchased from Cascade Biologics (Portland, OR) were grown in medium containing human corneal growth supplement (HCGS; EpiLife; Cascade Biologics), bovine pituitary extract, bovine insulin, hydrocortisone, bovine transferrin, and mouse EGF. First- and second-passage cells were used for cytokine secretion experiments. Cells were grown to 80% to 90% confluence, serum starved, and deprived of growth factors for 16 to 24 hours before experimentation. 
Western Blot Analysis
HCECs were pretreated with U0126 (Cell Signaling Technology, Beverly, MA), EGTA, nifedipine, or phosphate-buffered saline (PBS) and then treated with anti-asialoGM1 antibody. After treatment, cells were washed twice with ice-cold PBS (pH 7.4) and lysed for at least 5 minutes on ice in whole cell lysis buffer (20 mM Tris [pH 7.5], 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100, 2.5 mM sodium pyrophosphate, 1 mM β-glycerolphosphate, and 1 mM Na3VO4), with a protease inhibitor mixture (1 mM phenylmethylsulfonyl fluoride [PMSF], 1 mM benzamide, 10 μg/mL leupeptin, and 10 μg/mL aprotinin). Cells were harvested by scraping followed by sonication (three bursts at 40 mV for 10 seconds), and cell lysates were clarified by centrifugation at 4°C for 15 minutes at 13,000 rpm. Protein concentrations of the lysates were measured with a bicinchoninic acid assay (Micro BCA protein assay kit; Pierce Biotechnology, Rockford, IL), and equal amounts of total protein were fractionated on 10% sodium dodecyl sulfate (SDS)-polyacrylamide gels. After semidry transfer to polyvinylidene difluoride (PVDF) membranes, the membranes were blocked with PBS containing 0.1% Tween-20 (PBS-T) and 5% nonfat dry milk (Carnation, Nestlé USA, Glendale, CA) for 1 hour at room temperature and then incubated with specific antibodies for 2 hours at room temperature or overnight at 4°C. The antibodies used were: anti-asialoGM1 (1:50,000–1:100,000 dilution; Wako Chemical, Richmond, VA), anti-TLR5 (1:1000), anti-phospho-ERK1/2 (1:1000), and anti-ERK1/2 (1:2000; Santa Cruz Biotechnology, Santa Cruz, CA). After primary antibody incubation, PVDF membranes were washed three times with 5% nonfat dry milk in PBS-T and then incubated with an appropriate horseradish peroxidase (HRP)-conjugated secondary antibody (1:2,000–1:10,000 dilution; Santa Cruz Biotechnology) for 1 hour at room temperature. After successive washings, bound antibodies were visualized with an enhanced chemiluminescence system (ECL Plus Western Blot detection system; GE Healthcare, Piscataway, NJ). 
Flow Cytometry Analysis
Cell surface expression of asialoGM1 and TLR5 was determined by flow cytometry (Epics XL; Beckman-Coulter, Fullerton, CA) with laser power of 5.76 mW. The instrument was calibrated before each measurement with standardized fluorescent particles (Immunocheck; AMAC, Inc. Westbrook, ME). Fluorescence signals of cells were measured simultaneously by three photomultiplier tubes and optical filters and are expressed as the mean log fluorescence intensity of the cell population within the gate. The HCECs were incubated with an anti-asialoGM1 and TLR5 antibody on ice for 60 minutes, washed three times, incubated with a phycoerythrin (PE)-conjugated secondary antibody for 30 minutes on ice protected from light, fixed in lysing solution (Optilyse; Beckman-Coulter) at room temperature for 5 to 10 minutes, and analyzed by flow cytometry. The cells were initially identified by the characteristic forward- and side-scatter parameters on unstained cells and confirmed by staining with PE-conjugated primary anti-human HLA-DR (Beckman-Coulter). Data are expressed as the mean relative fluorescence units (RFU) and percentage of cells staining positively. Isotype primary conjugated antibodies served as a negative control. Samples were prepared and analyzed in duplicate, and a minimum of 5000 cells were counted from each sample. 
Fluorescence Microscopy
Cells were subcultured on 22-mm diameter glass coverslips (Fisher Scientific, Pittsburgh, PA), loaded with 2 μM Fura-2/AM (Invitrogen-Molecular Probes, Eugene, OR), with or without inhibitors (nifedipine, apyrase [Grade III], ARL67156, suramin, or PPADS) at 37°C for 30 minutes, then washed four times with Ringer’s solution (141 mM NaCl, 4.2 mM KCl, 2 mM KH2PO4, 1 mM MgCl2.6H2O, 0.8 mM CaCl2, 5.5 mM glucose, and 10 mM HEPES, 300 mOsmol/L [pH 7.4]), Ca2+-free counterpart solution buffered by 0.5 mM EGTA or Cl-free solution (141.6 mM d-gluconic acid sodium salt, 2.5 mM K2SO4, 2 mM MgSO4.7 H2O, 5 mM glucose, 5.4 mM CaSO4.2H2O, and 5.3 mM HEPES [pH 7.4], 300 mOsmol/L). Cells were alternately illuminated with 340 and 380 nm and their emission at 510 nm was monitored every 5 seconds with an inverted microscope (Diaphot 200; Nikon, Tokyo, Japan) and a digital charge coupled device (CCD) camera (model 1400 microimager; Xillix, Vancouver, British Columbia, Canada). Image-analysis software (Ratio Tool; Inovision, Durham, NC) was used to record and analyze fluorescence ratios. In a given microscopic field, 25 individual cells were chosen for monitoring, only if they had a comparatively large cytoplasmic/nuclear ratio and homogeneous brightness. At least five coverslips were examined for each condition. Results were plotted as ΔF340/F380 ± SEM from at least five independent experiments. 
Determination of IL-8 Secretion
Secretion of proinflammatory cytokines was determined with protein cytokine arrays, and the amount of IL-8 was measured using an enzyme-linked immunosorbent assay (ELISA). Immortalized HCECs and primary HCECs were plated in 35-mm cell culture plates. After 24 hours of growth factor starvation, cells were washed twice with PBS and then exposed to the indicated condition for 30 minutes before the cells were stimulated with anti-asialoGM1 antibody (1:80) for 8 hours. Supernatants were harvested and centrifuged at 4000 rpm for 3 minutes, to remove cell debris. Each supernatant was stored at −80°C until analysis. Protein concentration of each cell lysate was determined using the BCA protein assay kit. Protein microwell TH1/TH2 cytokine arrays were immobilized with nine different capture antibodies: IL-2, -4, -5, -8, -10, -12, and -13; interferon (IFN)-γ; and tumor necrosis factor (TNF)-α (Pierce Biotechnology). IL-8 ELISA (R&D Systems, Minneapolis, MN) was performed according to the manufacturer’s instructions. The amount of IL-8 in the culture medium was normalized to the total amount of cellular protein lysed with 5% SDS and 0.5 N NaOH. Results are expressed as mean picograms of IL-8 per milligram cell lysate ± SEM (n = 3). 
Coimmunoprecipitation
Cells grown to 85% to 90% confluence in 100-mm plates (Fisher Scientific) were washed twice with ice-cold PBS and then lysed with 1 mL of radioimmunoprecipitation assay (RIPA) buffer (150 mM NaCl, 20 mM Tris [pH 7.5], 1% Triton X-100, 0.5% sodium deoxycholate, 1 mM EDTA, and 0.1% SDS) with a protease and phosphatase inhibitor mixture (1 mM PMSF, 10 μg/mL aprotinin, 10 μg/mL leupeptin, 1 μg/mL pepstatin, 1 mM sodium fluoride, and 1 mM sodium orthovanadate) for at least 20 minutes on ice at 4°C. After the cells were harvested by scraping and sonication, the sample was centrifuged at 13,000 rpm at 4°C for 15 minutes to remove debris. Nonspecific proteins in the supernatant were removed by preclearing the sample with 50 μL of 50% vol/vol of Proteins A and G (Santa Cruz Biotechnology) with 1 μg of the appropriate control IgG, for 30 minutes at 4°C, followed by centrifugation (2500 rpm for 30 seconds). The protein concentration of the supernatant was determined with the BCA protein assay kit. Equal amounts of protein (1300 μg) were subjected to immunoprecipitation with 6 μg of anti-TLR5, 15 μg of anti-asialoGM1, or 15 μg nonspecific immunoglobulin (normal mouse control IgG; Santa Cruz Biotechnology) and 20 μL of Protein A/G mixture (1:1) or with the beads alone at 4°C overnight. Immunoprecipitates were washed with RIPA buffer three times and PBS once, dissolved in 60 μL of 2× sample buffer (12.5 mM Tris [pH 6.8], 4% SDS, 20% glycerol, 2.5% β-mercaptoethanol, 0.1% bromphenol blue), boiled for 10 minutes, and centrifuged at 10,000 rpm for 2 minutes at 4°C to remove the beads. The samples were then subjected to SDS-PAGE and Western blot analysis. 
Statistical Analysis
Data are expressed as the mean ± SEM of experimental sets, each of which were repeated at least three times, unless otherwise indicated. All statistical analyses were performed by one-way ANOVA (Statistica 6.0; StatSoft, Tulsa, OK) with the Fisher least-significant difference (LSD) post hoc probability matrix, unless otherwise specified, and P < 0.05 was considered statistically significant. 
Results
AsialoGM1 and TLR5 Expression in HCECs
To determine whether HCECs express asialoGM1 and TLR5, we used SDS-PAGE and Western blot analysis with flow cytometry. The Western blot analyses shown in Figures 1A and 1Bdocument the presence of TLR5 and asialoGM1 based on close correspondence with their described apparent molecular weights. Unlike in the isolated rabbit corneal epithelium, flow cytometry of the HCECs clearly demonstrated surface expression of both asialoGM1 and TLR5. 
Ligation of AsialoGM1 in Calcium Mobilization
To probe for asialoGM1-mediated Ca2+ signaling, asialoGM1 was ligated with anti-asialoGM1 antibody and analyzed for intracellular calcium concentration, [Ca2+]i. There were steep increases at 45 seconds in the F340/F380 ratio induced by this antibody at dilutions from 1:320 to 1:80 (Fig. 2A) . At a dilution of 1:80, the maximum transient increase occurred. Subsequent to the peak increase, there was a gradual waning of the response that resulted in near recovery to the baseline level after approximately 15 minutes. The interaction between asialoGM1 and TLR5 in eliciting these responses is indicated by the finding that with the function-blocking anti-TLR5 antibody, the asialoGM1 agonist dose dependently suppressed the maximum transient. The largest suppression obtained with 20 μg/mL was approximately 84%. To determine the source of calcium underlying the aforementioned transients, the effect of anti-asialoGM1 antibody on this response was determined in Ca2+-free Ringer’s solution, buffered with 0.5 mM EGTA. Under this condition, this agonist failed to elicit calcium mobilization (Fig. 2B) , indicating that anti-asialoGM1 antibody–induced [Ca2+]i increases depend on calcium influx from the extracellular space. 
The dependency of increases in L-type Ca2+ channel activity on mediating [Ca2+]i transients was evaluated by determining whether the L-type Ca2+ channel blocker nifedipine affects the response. Cells were preincubated with nifedipine (0.5 and 1 μM) for 30 minutes, and anti-asialoGM1 antibody and nifedipine were applied simultaneously. With nifedipine, there were dose-dependent reductions in anti-asialoGM1 antibody–induced [Ca2+]i transients (F4,30 = 39.9758, P < 0.00001), with maximum inhibition of 72% with 1 μM nifedipine (Fig. 2C) . A reduction in anti-asialoGM1 antibody–induced [Ca2+]i increase by nifedipine indicates a role for L-type Ca2+ channels in mediating agonist-induced Ca2+ transients. 
To confirm the involvement of L-type voltage-dependent Ca2+ channel activation in mediating anti-asialoGM1 antibody–induced Ca2+ transients, we incubated the cells in Cl-free Ringer’s solution. Such activation is dependent on membrane depolarization to less than −30 mV, resulting from the cAMP-mediated stimulation of net Cl transport. When net Cl transport activity is eliminated in Cl-free Ringer’s, TLR5 receptor activation cannot induce L-type Ca2+ channel activation. Such stimulation is obviated because despite adenylate cyclase–mediated cAMP formation, the membrane voltage cannot be depolarized to a level sufficient to activate L-type Ca2+ channel activity. The results shown in Figure 2Care consistent with this formulation, since there was a near equivalence between the diminution of the anti-asialoGM1 antibody–induced Ca2+ transients in Cl-free Ringer’s and those measured during exposure to nifedipine. This agreement provides further support that TLR5-induced Ca2+ transients stem from stimulation of L-type Ca2+ activity. 
Involvement of ATP on Transduction of Ca2+ Signals from AsialoGM1
To determine whether asialoGM1 ligation induces downstream Ca2+ transients through ATP release into the medium, the effects of apyrase (an ATP dephosphatase) and ARL67156 (an ectonucleotidase inhibitor) on Ca2+ mobilization after asialoGM1 ligation were studied. ATPγ-S elicited dose-dependent increases in [Ca2+]i (Figs. 3A 3B) . Apyrase suppressed in a dose-dependent manner the anti-asialoGM1 antibody–induced [Ca2+]i increase (Fig. 3C ; F5,24 = 11.7417, P < 0.00001), thus indicating a role for ATP. To confirm further a role for ATP, the cells were preincubated with 100 μM of the 5′-ectonucleotidase inhibitor ARL67156 for 30 minutes before the simultaneous applications of 1:160 anti-asialoGM1 antibody and 100 μM ARL67156. This inhibitor increased the efficacy of anti-asialoGM1 antibody–induced Ca2+ mobilization, presumably due to suppression of ATP hydrolysis (Fig. 3D ; F4,20 = 25.6559, P < 0.00001; Fisher LSD P < 0.005). Furthermore, the presence of the P2 receptor antagonists—100 μM suramin and 100 μM PPADS—significantly reduced anti-asialoGM1 antibody–induced Ca2+ transients (Fig. 3D ; F4,20 = 25.6559, P < 0.00001; Fisher LSD P < 0.00001). These results are consistent with those described in human lung epithelial cells in which asialoGM1 ligation promotes ATP release and induces P2 receptor stimulation followed by transients increases in [Ca2+]i. 10  
Activation of ERK1/2
Antibody ligation of asialoGM1 induced ERK1/2 phosphorylation in a time-dependent manner (Fig. 4A) . Control cells showed a low, but detectable, level of phosphorylated ERK1/2. On asialoGM1 stimulation, the levels increased by 5 minutes and peaked at 30 minutes, followed by a slow decline for up to 2 hours, at which time phospho-ERK1/2 was still detectable. 
Dependence of ERK Pathway Activation on AsialoGM1-Induced Ca2+ Transients
To determine the dependency of asialoGM1 antibody–induced increases in p-ERK1/2 levels on Ca2+ influx through L-type voltage-dependent Ca2+ channels, we examined the effects of EGTA and nifedipine on ERK1/2 phosphorylation. Either EGTA in Ca2+-free medium or nifedipine in Ca2+-containing medium blocked anti-asialoGM1 antibody–induced ERK1/2 phosphorylation (Fig. 4B) . U0126 (10 μM), an MAPK and ERK kinase (MEK) inhibitor, abolished anti-asialoGM1 antibody–induced ERK1/2 phosphorylation. Taken together, these data suggest that anti-asialoGM1 antibody–induced Ca2+ mobilization and ERK1/2 phosphorylation are dependent on Ca2+ influx through L-type voltage-dependent Ca2+ channels. 
Effect of AsialoGM1 Ligation on IL8 Secretion
To assess the biological relevance of anti-asialoGM1 antibody–induced Ca2+ mobilization, we measured the effect of anti-asialoGM1 antibody on proinflammatory cytokine production (secretion) with a protein microwell array. Of the nine cytokines examined, only IL-8 secretion was detected (data not shown). ELISA was performed to quantify the secretion of IL-8 in immortalized HCECs and primary HCECs. Both cell lines constitutively secreted IL-8. Increased amounts of IL-8 were detected after 8 hours of exposure to a 1:80 dilution of the antibody agonist (F11,24 = 54.8384, P < 0.0001; Fisher LSD P < 0.0001). After 8 hours, the amounts of IL-8 accumulated in the culture medium of immortalized HCECs and primary HCECs stimulated with asialoGM1 were 9.5 and 8.1 times, respectively, more than that in their control cell medium. 
AsialoGM1-Mediated IL-8 Release
To determine the dependency of IL-8 secretion on Ca2+ mobilization after asialoGM1 ligation, the effects of Ca2+ chelators on IL-8 secretion were determined. In the absence of extracellular Ca2+ or presence of the anti-TLR5 antibody, anti-asialoGM1 antibody–induced secretion of IL-8 was inhibited (Fig. 5) . Pretreatment with anti-TLR2 antibody did not prevent asialoGM1-induced secretion of IL-8, suggesting that the release of AGM1-mediated release of this chemokine is solely dependent on its interaction with TLR5. Moreover, U0126 (10 μM) reduced anti-asialoGM1 antibody–induced IL-8 accumulation by 86%. In parallel experiments, similar results were found for IL-8 secretion in primary HCECs. These data suggest that asialoGM1 ligation stimulates IL-8 secretion, which is mediated by TLR5 and not TLR2. Furthermore, asialoGM1-mediated Ca2+ influx from the extracellular space stimulates the ERK pathway. 
Colocalization of AsialoGM1 with TLR5
AsialoGM1 lacks transmembrane and cytoplasmic domains and is therefore not capable of contact with downstream cytoplasmic signaling molecules. To determine whether asialoGM1 forms a coreceptor association with TLR5, coimmunoprecipitation was performed. Immunoprecipitates were probed by SDS-PAGE along with immunoblot analysis with anti-asialoGM1 antibodies (Fig. 6) . AsialoGM1 colocalization with TLR5 was evident and is highly suggestive of a coreceptor association with this receptor. 
Discussion
CECs constitute the first line of ocular defense against microbial pathogens. In addition, glycolipid receptors, such as asialoGM1 and TLRs, play crucial roles in sensing the presence of pathogenic bacteria. In this study, ligation of the asialoGM1 receptor caused Ca2+ influx through L-type voltage-dependent Ca2+ channels along with subsequent activation of the ERK pathway of the MAPK cascade, leading to IL-8 production (Fig. 7) . Furthermore, the HCECs expressed TLR5, an innate immunity receptor of flagellin 3 and that TLR5 was a coreceptor for the asialoGM1-mediated inflammatory response—that is, IL-8 release. Thus, asialoGM1, in association with TLR5 and through Ca2+ mobilization and ERK activation, plays a role in eliciting innate immune responses in the corneal epithelium and in subsequent inflammatory processes in corneal diseases (e.g., bacterial keratitis). 
CECs are important components of the mucosal defense system and express chemokines and cytokines to direct the influx and activation of phagocytic cells in response to bacterial pathogens. Owing to its superficial location, the cornea is constantly exposed to diverse organisms, although it usually remains immunosilent, as the eye is relatively impermeable to microorganisms. 34 35 36 However, if epithelial integrity is breached by trauma, such as that imposed by contact lens–induced insult, sight-threatening bacterial infection may occur. 37 38 Hence, the corneal epithelium must possess the ability to sense and respond to the presence of pathogenic bacteria. AsialoGM1 serves as a receptor for a large number of pathogens, 5 but its presence in the corneal epithelium has been controversial. 29 30 Our studies revealed for the first time in a physiologically relevant HCEC line, that these cells express asialoGM1. Although the inflammatory response is necessary to clear offending organisms from infected tissue, activation of epithelial signaling and recruitment of polymorphonuclear leukocytes is not innocuous to the host. Persistent inflammation is clearly detrimental to the function of the corneal epithelium in maintaining corneal transparency and may itself compromise the cornea by leading to corneal scarring, melting, and blindness. 15 35 39 Thus, blocking the asialoGM1 receptor or TLR5 on CECs should be part of any therapeutic strategy aimed at preventing the destructive sequelae of ocular bacterial infections in response to inflammation. 
AsialoGM1 has an exposed GalNAcβ1-4Gal moiety that serves as a receptor for a large number of pathogens. 16 Owing to the corneal epithelium’s lack of transmembrane and cytoplasmic components, the link between ligand recognition and downstream kinase activation in the corneal epithelium remains to be clarified. In airway epithelial cells, asialoGM1 ligation promotes extracellular ATP release, followed by autocrine activation of nucleotide receptors, possibly P2Y, resulting in activation of phospholipase C (PLC), Ca2+ mobilization, phosphorylation of ERK1/2 and activation of mucin transcription. 9 The mechanism of ATP release, whether by exocytosis or by a pumping out by members of the ATP-binding cassette (ABC) transporter family, remains unknown. Moreover, in parainfluenza viral infections, activation of Cl secretion and inhibition of Na+ absorption across the tracheal epithelium have been reported. 40 These changes in ion transport activity result in inhibition of volume regulation and are a consequence of binding to a neuraminidase-insensitive glycolipid, possibly asialoGM1, triggering the release of ATP, which then acts in an autocrine fashion on apical P2Y receptors to produce the aforementioned changes in ion transport. 40 Changes in Ca2+ transient patterns of activation in the presence of apyrase, ARL67156, and P2 receptor antagonists confirm that asialoGM1 ligation in HCECs cause extracellular ATP release and binding to P2 receptors. 
Purinergic receptors represent a potential link between asialoGM1 ligation and downstream signaling components. 9 10 40 On binding of ATP to purinergic receptors, large Ca2+ oscillations are induced. 41 42 These increases in cytosolic Ca2+ are generated by P2X or P2Y receptor activation, which occur via distinct mechanisms. 43 P2X receptors, which play an important role in excitable cell types, are ligand-gated ion channels that allow Ca2+ influx from the extracellular space after nucleotide binding. 44 On the other hand, P2Y receptors are G-protein coupled receptors (GPCRs) that stimulate release of intracellular Ca2+ stores through PLC-mediated PIP2 hydrolysis and activation of the IP3 pathway. 44 These receptors produce a biphasic Ca2+ response in which the initial depletion of intracellular Ca2+ stores leads to a sustained Ca2+ release due to the opening of specific store-operated voltage-independent Ca2+ channels in the plasma membrane that allow Ca2+ entry from the extracellular space. 45  
The innate immune response to microbes consists of an assortment of soluble components including cytokines, chemokines, and antimicrobial peptides. 46 Epithelial cells produce IL-8 in response to PA challenge through activation of NF-κB and transcription of the IL-8 gene. 17 47 In airway epithelial cells, intracellular Ca2+ transients mediate IL-8 production; cells pretreated with BAPTA/AM to chelate intracellular Ca2+ had significantly reduced PA-induced IL-8 responses, whereas EGTA, an extracellular Ca2+ chelator, and NiCl2, an external Ca2+ channel blocker, did not inhibit IL-8 secretion. 4 Moreover, BAPTA/AM and thapsigargin, which deplete Ca2+ stores by blocking Ca2+ ATPases in the endoplasmic reticulum and clamping intracellular Ca2+ concentration, respectively, significantly reduce anti-asialoGM1 antibody–induced ERK1/2 phosphorylation as well as MUC-2 production in airway epithelial cells. 9 Our results indicate that, in immortalized and primary HCECs, IL-8 production—induced by asialoGM1 ligation—is mediated by Ca2+ influx, suggesting a tissue-specific signaling cascade. 
In HCECs, anti-asialoGM1 antibody–induced Ca2+ transients are a result of Ca2+ influx from the extracellular space after cAMP-mediated increases in net transepithelial Cl transport activity. This response was suppressed in Cl-free medium, since the membrane voltage remained too negative for L-type voltage-dependent Ca2+ channel stimulation. In this condition, despite ATP release after TLR5 stimulation, its hydrolysis by adenylate cyclase to produce cAMP had no depolarizing effect on membrane voltage in a Cl-free medium. Alternatively, blockage of volume regulation clamped the membrane voltage at levels that were too hyperpolarized in another system, which prevented asialoGM1 coreceptor–induced increases in Ca2+ influx through stimulation of L-type voltage-dependent Ca2+ channel activity. 40 Taken together, the evidence shows that, although asialoGM1 participates in innate immunity, its effects are mediated by different pathways in different tissues. 
Another potential transmembrane link between asialoGM1 ligation and downstream signaling events are the TLRs. In airway epithelial cells, the signaling capabilities of TLR2 are amplified, through its association with asialoGM1 within the context of lipid rafts, to provide a broadly responsive signaling complex that initiates the host response to potential bacterial infection. 13 Our characterization of Ca2+ transients indicate that both TLR2 (PGN, LTA Gram-positive bacteria cell wall components) and TLR4 (LPS and lipid A Gram-negative bacteria cell wall components) agonists did not elicit Ca2+ mobilization and that anti-asialoGM1 antibody–induced increases in [Ca2+]i are not potentiated by TLR2 agonists (data not shown). Similarly, inhibition of TLR2 did not prevent anti-asialoGM1 antibody–induced increase in IL-8 secretion. This suggests that TLRs2 and -4 are unlikely candidates for asialoGM1 coreceptors, and that a tissue-specific relationship may exist to account for differences in pathogen recognition. 
Another possible candidate for a receptor complex is TLR5. Recent studies have shown that flagella—surface structures that confer motility—enhance the pathogenicity of certain organisms, either by promoting adhesion to the host tissues or by directly activating host inflammatory signaling pathways. 34 48 49 50 PA elaborates a multitude of factors, including flagella. 14 It has been demonstrated that flagella bind specifically to host cell glycosphingolipid asialoGM1 and that binding is essential for epithelial cell invasion and cytotoxicity. 5 14 In airway epithelial infection by PA, flagella initially localized with asialoGM1, inducing TLR5 mobilization to the apical cell surface and its subsequent colocalization with superficial flagella. 12 Flagella-induced Ca2+ fluxes induce activation of Src, Ras, ERK1/2, and NF-κB through asialoGM1, TLR2, and TLR5, which is dependent on the availability of exposed receptors on the apical surface of polarized airway epithelial cells. 12  
Finally, within the specific context of IL-8 production induced by pathogenic bacteria or their PAMPs, significant common themes are notable. For example, the signaling pathway by which asialoGM1 ligands stimulate IL-8 production 51 is similar to that stimulated by flagellin, 12 52 53 LPS, 54 and PGN/LTA, 12 55 in that it requires activation of NF-κB and MAPKs, especially ERK1/2. Our results suggest a role for the ERK limb of the MAPK cascade, consistent with findings in previous studies. In agreement with results in previous studies, ours suggest a role for the ERK pathway of the MAPK cascade. Convergence of these bacterial recognition pathways on these kinases represents a coordinated innate immune response. 
In summary, our studies suggest that HCECs can initiate immune responses via asialoGM1 ligation, thus resulting in the clearing of pathogens and, potentially, the elimination of the excessive inflammatory response that result in corneal scarring and loss of vision. Understanding the molecular events of bacterial–epithelial interactions and the inflammatory consequences of asialoGM1 ligation may permit the development of novel, specific therapies that can promote innate defenses and prevent some of the destructive sequelae of ocular bacterial infections. 
 
Figure 1.
 
HCECs expressed both asialoGM1 and TLR5. Cell lysates were resolved with Western blot analysis and probed for asialoGM1 and TLR5 receptor expression with specific antibodies (A, B). (A) Sixty, 40, and 20 μg of total proteins were loaded into lanes 1, 2, and 3, respectively. (B) NAWALMA (positive control) and HCEC lysates were loaded into lanes 1 and 2, respectively. (C) AsialoGM1 and TLR5 surface expression on HCECs was analyzed by flow cytometry. Left traces: the isotype control (which was similar to background autofluorescence); right traces: PE-conjugated anti-asialoGM1 and TLR5 labeling.
Figure 1.
 
HCECs expressed both asialoGM1 and TLR5. Cell lysates were resolved with Western blot analysis and probed for asialoGM1 and TLR5 receptor expression with specific antibodies (A, B). (A) Sixty, 40, and 20 μg of total proteins were loaded into lanes 1, 2, and 3, respectively. (B) NAWALMA (positive control) and HCEC lysates were loaded into lanes 1 and 2, respectively. (C) AsialoGM1 and TLR5 surface expression on HCECs was analyzed by flow cytometry. Left traces: the isotype control (which was similar to background autofluorescence); right traces: PE-conjugated anti-asialoGM1 and TLR5 labeling.
Figure 2.
 
Anti-asialoGM1 antibody induced dose-dependent increases in [Ca2+]i. (A) Cells were initially imaged in 100 μL of Ringer’s solution, and after 100 seconds, 100 μL of Ringer’s solution containing a 1:40 (final 1:80) dilution of anti-asialoGM1 antibody was added (arrow). (B) Dose response of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were challenged with Ringer’s solution and anti-rabbit IgG as the vehicle and the negative control, respectively. After 100 seconds of baseline measurements, 100 μL of Ringer’s solution containing a 1:160 to 1:20 (final 1:320–1:40) dilution of anti-asialoGM1 antibody was added. The maximum change in fura-2/AM ratios was elicited by 1:40 (final 1:80) dilution, with EC50 at approximately 1:90 at 45 seconds. Preincubation with 20 μg/mL anti-TLR5 antibody or absence of extracellular Ca2+ (Ca2+-free solution buffered by 0.5 mM EGTA) completely abolished the anti-asialoGM1 antibody–induced calcium transient. One-way ANOVA F8,36 = 43.2815, P < 0.00001; Fisher LSD *P < 0.05, **P < 0.0005, and ***P < 0.0001. (C) Nifedipine caused a dose-dependent reduction of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were preincubated with the indicated concentrations of nifedipine for 30 minutes and then 1:80 dilution of anti-asialoGM1 antibody with the indicated concentrations of nifedipine were added simultaneously. The absence of extracellular Cl, in the presence or absence of nifedipine, significantly reduced the anti-asialoGM1 antibody–induced increase in [Ca2+]i. F4,30 = 39.9758, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 2.
 
Anti-asialoGM1 antibody induced dose-dependent increases in [Ca2+]i. (A) Cells were initially imaged in 100 μL of Ringer’s solution, and after 100 seconds, 100 μL of Ringer’s solution containing a 1:40 (final 1:80) dilution of anti-asialoGM1 antibody was added (arrow). (B) Dose response of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were challenged with Ringer’s solution and anti-rabbit IgG as the vehicle and the negative control, respectively. After 100 seconds of baseline measurements, 100 μL of Ringer’s solution containing a 1:160 to 1:20 (final 1:320–1:40) dilution of anti-asialoGM1 antibody was added. The maximum change in fura-2/AM ratios was elicited by 1:40 (final 1:80) dilution, with EC50 at approximately 1:90 at 45 seconds. Preincubation with 20 μg/mL anti-TLR5 antibody or absence of extracellular Ca2+ (Ca2+-free solution buffered by 0.5 mM EGTA) completely abolished the anti-asialoGM1 antibody–induced calcium transient. One-way ANOVA F8,36 = 43.2815, P < 0.00001; Fisher LSD *P < 0.05, **P < 0.0005, and ***P < 0.0001. (C) Nifedipine caused a dose-dependent reduction of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were preincubated with the indicated concentrations of nifedipine for 30 minutes and then 1:80 dilution of anti-asialoGM1 antibody with the indicated concentrations of nifedipine were added simultaneously. The absence of extracellular Cl, in the presence or absence of nifedipine, significantly reduced the anti-asialoGM1 antibody–induced increase in [Ca2+]i. F4,30 = 39.9758, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 3.
 
ATP transduced Ca2+ transients via P2 receptors induced by anti-asialoGM1 antibody. (A) ATP induced increases in [Ca2+]i. Cells were initially imaged in 100 μL of Ringer’s solution, and, after 100 seconds, 100 μL of Ringer’s solution containing 1000 μM (final 500 μM) of ATPγ-S was added (arrow). (B) Dose–response of ATP-induced increases in [Ca2+]i. F3,16 = 130.6576, P < 0.00001; Fisher LSD *P < 0.0001. (C) The presence of apyrase significantly reduced anti-asialoGM1 antibody–induced increases in [Ca2+]i. F5,24 = 11.7417, P < 0.00001; Fisher LSD *P < 0.005, **P < 0.001 and ***P < 0.0001. (D) ARL67156 significantly enhanced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. The P2 receptor antagonists suramin and PPADS significantly reduced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. F4,20 = 25.6559, P < 0.00001; Fisher LSD *P < 0.005 and **P < 0.00001.
Figure 3.
 
ATP transduced Ca2+ transients via P2 receptors induced by anti-asialoGM1 antibody. (A) ATP induced increases in [Ca2+]i. Cells were initially imaged in 100 μL of Ringer’s solution, and, after 100 seconds, 100 μL of Ringer’s solution containing 1000 μM (final 500 μM) of ATPγ-S was added (arrow). (B) Dose–response of ATP-induced increases in [Ca2+]i. F3,16 = 130.6576, P < 0.00001; Fisher LSD *P < 0.0001. (C) The presence of apyrase significantly reduced anti-asialoGM1 antibody–induced increases in [Ca2+]i. F5,24 = 11.7417, P < 0.00001; Fisher LSD *P < 0.005, **P < 0.001 and ***P < 0.0001. (D) ARL67156 significantly enhanced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. The P2 receptor antagonists suramin and PPADS significantly reduced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. F4,20 = 25.6559, P < 0.00001; Fisher LSD *P < 0.005 and **P < 0.00001.
Figure 4.
 
Anti-asialoGM1 antibody induced ERK1/2 activation. (A) HCECs were grown in 35-mm plates and serum deprived for 24 hours before stimulation with anti-asialoGM1 antibody (1:80 dilution) for the indicated times. Cell lysates were fractionated by SDS-PAGE and then analyzed by Western blot analysis with antibodies directed against phospho-ERK1/2. Membranes were then stripped and reprobed using an antibody against total ERK1/2 as the control for equal protein loading. (B) The effect of 30 minutes’ preincubation of 10 μM U0126, 1 mM EGTA, and 1 μM nifedipine (NF) on anti-asialoGM1 antibody (AGM; 1:80 dilution for 30 minutes)–induced ERK activation, compared with the unstimulated control (CTL).
Figure 4.
 
Anti-asialoGM1 antibody induced ERK1/2 activation. (A) HCECs were grown in 35-mm plates and serum deprived for 24 hours before stimulation with anti-asialoGM1 antibody (1:80 dilution) for the indicated times. Cell lysates were fractionated by SDS-PAGE and then analyzed by Western blot analysis with antibodies directed against phospho-ERK1/2. Membranes were then stripped and reprobed using an antibody against total ERK1/2 as the control for equal protein loading. (B) The effect of 30 minutes’ preincubation of 10 μM U0126, 1 mM EGTA, and 1 μM nifedipine (NF) on anti-asialoGM1 antibody (AGM; 1:80 dilution for 30 minutes)–induced ERK activation, compared with the unstimulated control (CTL).
Figure 5.
 
IL-8 secretion in HCECs stimulated by anti-asialoGM1 antibody: effects of Ca2+ chelators, MEK1 inhibitor, and anti-TLR5 antibody. Confluent monolayers of SV40 and primary HCECs grown in 35-mm plates were pretreated with the compounds indicated for 30 minutes before 8 hours of exposure to anti-asialoGM1 antibody (1:80 dilution), maintained in the presence of the indicated compounds. The secreted IL-8 in cell supernatants was measured by ELISA. (▪) SV40 HCECs; (□) primary HCECs. F11,24 = 54.8384, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 5.
 
IL-8 secretion in HCECs stimulated by anti-asialoGM1 antibody: effects of Ca2+ chelators, MEK1 inhibitor, and anti-TLR5 antibody. Confluent monolayers of SV40 and primary HCECs grown in 35-mm plates were pretreated with the compounds indicated for 30 minutes before 8 hours of exposure to anti-asialoGM1 antibody (1:80 dilution), maintained in the presence of the indicated compounds. The secreted IL-8 in cell supernatants was measured by ELISA. (▪) SV40 HCECs; (□) primary HCECs. F11,24 = 54.8384, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 6.
 
Colocalization of asialoGM1 and TLR5 in HCECs. For immunoprecipitation, 1.3 mg of protein was immunoprecipitated (IP) with 6 μg of anti-TLR5 (TLR5), 15 μg of anti-asialoGM1 (AGM), and 15 μg of nonspecific immunoglobulins (IgG) with 20 μL of protein A/G mixture, or with the agarose beads (Beads) alone. The samples were subjected to SDS-PAGE and probed (WB) with rabbit anti-asialoGM1.
Figure 6.
 
Colocalization of asialoGM1 and TLR5 in HCECs. For immunoprecipitation, 1.3 mg of protein was immunoprecipitated (IP) with 6 μg of anti-TLR5 (TLR5), 15 μg of anti-asialoGM1 (AGM), and 15 μg of nonspecific immunoglobulins (IgG) with 20 μL of protein A/G mixture, or with the agarose beads (Beads) alone. The samples were subjected to SDS-PAGE and probed (WB) with rabbit anti-asialoGM1.
Figure 7.
 
Events in host cell signaling after asialoGM1 ligation. Anti-asialoGM1 antibody binds to its receptor, a membrane glycolipid, which is associated with TLR5. This causes the extracellular release of ATP. Membrane surface adenylate cyclase (AC) catalyzes the conversion of ATP to cAMP, which then acts on cAMP gCl-sensitive channels, leading to Cl secretion. The secretion results in depolarization of the plasma membrane, thus activating L-type voltage-dependent Ca2+ channels, leading to Ca2+ influx. Ca2+ mobilization, as a result of Ca2+ influx, leads to phosphorylation of MEK1/2 and ERK1/2 and to IL-8 production.
Figure 7.
 
Events in host cell signaling after asialoGM1 ligation. Anti-asialoGM1 antibody binds to its receptor, a membrane glycolipid, which is associated with TLR5. This causes the extracellular release of ATP. Membrane surface adenylate cyclase (AC) catalyzes the conversion of ATP to cAMP, which then acts on cAMP gCl-sensitive channels, leading to Cl secretion. The secretion results in depolarization of the plasma membrane, thus activating L-type voltage-dependent Ca2+ channels, leading to Ca2+ influx. Ca2+ mobilization, as a result of Ca2+ influx, leads to phosphorylation of MEK1/2 and ERK1/2 and to IL-8 production.
The authors thank Kathryn Pokorny for editorial suggestions and Jinping Zhu for flow cytometry analysis. 
WoodsCA, MorganPB. Use of silicone hydrogel contact lenses by Australian optometrists. Clin Exp Optom. 2004;87:19–23. [CrossRef] [PubMed]
SongPI, AbrahamTA, ParkY, et al. The expression of functional LPS receptor proteins CD14 and toll-like receptor 4 in human corneal cells. Invest Ophthalmol Vis Sci. 2001;42:2867–2877. [PubMed]
ZhangJ, XuK, AmbatiB, YuFS. Toll-like receptor 5-mediated corneal epithelial inflammatory responses to Pseudomonas aeruginosa flagellin. Invest Ophthalmol Vis Sci. 2003;44:4247–4254. [CrossRef] [PubMed]
RatnerAJ, BryanR, WeberA, et al. Cystic fibrosis pathogens activate Ca2+-dependent mitogen-activated protein kinase signaling pathways in airway epithelial cells. J Biol Chem. 2001;276:19267–19275. [CrossRef] [PubMed]
FeldmanM, BryanR, RajanS, et al. Role of flagella in pathogenesis of Pseudomonas aeruginosa pulmonary infection. Infect Immun. 1998;66:43–51. [PubMed]
MedzhitovR. Toll-like receptors and innate immunity. Nat Rev Immunol. 2001;1:135–145. [CrossRef] [PubMed]
MedzhitovR, JanewayC, Jr. Innate immunity. N Engl J Med. 2000;343:338–344. [CrossRef] [PubMed]
MedzhitovR, Preston-HurlburtP, JanewayCA, Jr. A human homologue of the Drosophila Toll protein signals activation of adaptive immunity. Nature. 1997;388:394–397. [CrossRef] [PubMed]
McNamaraN, KhongA, McKemyD, et al. ATP transduces signals from ASGM1, a glycolipid that functions as a bacterial receptor. Proc Natl Acad Sci USA. 2001;98:9086–9091. [CrossRef] [PubMed]
McNamaraN, GallupM, SucherA, MaltsevaI, McKemyD, BasbaumC. AsialoGM1 and TLR5 cooperate in flagellin-induced nucleotide signaling to activate ERK1/2. Am J Respir Cell Mol Biol. 2006;34:653–660. [CrossRef] [PubMed]
TachadoSD, ZhangJ, ZhuJ, PatelN, KozielH. HIV impairs TNF-alpha release in response to Toll-like receptor 4 stimulation in human macrophages in vitro. Am J Respir Cell Mol Biol. 2005;33:610–621. [CrossRef] [PubMed]
AdamoR, SokolS, SoongG, GomezMI, PrinceA. Pseudomonas aeruginosa flagella activate airway epithelial cells through asialoGM1 and toll-like receptor 2 as well as toll-like receptor 5. Am J Respir Cell Mol Biol. 2004;30:627–634. [CrossRef] [PubMed]
SoongG, ReddyB, SokolS, AdamoR, PrinceA. TLR2 is mobilized into an apical lipid raft receptor complex to signal infection in airway epithelial cells. J Clin Invest. 2004;113:1482–1489. [CrossRef] [PubMed]
LyczakJB, CannonCL, PierGB. Establishment of Pseudomonas aeruginosa infection: lessons from a versatile opportunist. Microbes Infect. 2000;2:1051–1060. [CrossRef] [PubMed]
Kurpakus-WheaterM, KernackiKA, HazlettLD. Maintaining corneal integrity how the “window” stays clear. Prog Histochem Cytochem. 2001;36:185–259. [PubMed]
KrivanHC, RobertsDD, GinsburgV. Many pulmonary pathogenic bacteria bind specifically to the carbohydrate sequence GalNAc beta 1–4Gal found in some glycolipids. Proc Natl Acad Sci USA. 1988;85:6157–6161. [CrossRef] [PubMed]
DiMangoE, RatnerAJ, BryanR, TabibiS, PrinceA. Activation of NF-kappaB by adherent Pseudomonas aeruginosa in normal and cystic fibrosis respiratory epithelial cells. J Clin Invest. 1998;101:2598–2605. [CrossRef] [PubMed]
MedvedevAE, KopydlowskiKM, VogelSN. Inhibition of lipopolysaccharide-induced signal transduction in endotoxin-tolerized mouse macrophages: dysregulation of cytokine, chemokine, and toll-like receptor 2 and 4 gene expression. J Immunol. 2000;164:5564–5574. [CrossRef] [PubMed]
AkiraS, TakedaK, KaishoT. Toll-like receptors: critical proteins linking innate and acquired immunity. Nat Immunol. 2001;2:675–680. [CrossRef] [PubMed]
ZhangD, ZhangG, HaydenMS, GreenblattMB, BusseyC, FlavellRA. A toll-like receptor that prevents infection by uropathogenic bacteria. Science. 2004;303:1522–1526. [CrossRef] [PubMed]
AlexopoulouL, HoltAC, MedzhitovR, FlavellRA. Recognition of double-stranded RNA and activation of NF-kappaB by Toll-like receptor 3. Nature. 2001;413:732–738. [CrossRef] [PubMed]
ChowJC, YoungDW, GolenbockDT, ChristWJ, GusovskyF. Toll-like receptor-4 mediates lipopolysaccharide-induced signal transduction. J Biol Chem. 1999;274:10689–10692. [CrossRef] [PubMed]
KawasakiK, AkashiS, ShimazuR, YoshidaT, MiyakeK, NishijimaM. Mouse toll-like receptor 4.MD-2 complex mediates lipopolysaccharide-mimetic signal transduction by Taxol. J Biol Chem. 2000;275:2251–2254. [CrossRef] [PubMed]
HayashiF, SmithKD, OzinskyA, et al. The innate immune response to bacterial flagellin is mediated by Toll-like receptor 5. Nature. 2001;410:1099–1103. [CrossRef] [PubMed]
HemmiH, TakeuchiO, KawaiT, et al. A Toll-like receptor recognizes bacterial DNA. Nature. 2000;408:740–745. [CrossRef] [PubMed]
JohnsonAC, HeinzelFP, DiaconuE, et al. Activation of toll-like receptor (TLR)2, TLR4, and TLR9 in the mammalian cornea induces MyD88-dependent corneal inflammation. Invest Ophthalmol Vis Sci. 2005;46:589–595. [CrossRef] [PubMed]
UetaM, NochiT, JangMH, et al. Intracellularly expressed TLR2s and TLR4s contribution to an immunosilent environment at the ocular mucosal epithelium. J Immunol. 2004;173:3337–3347. [CrossRef] [PubMed]
OkamotoT, SchlegelA, SchererPE, LisantiMP. Caveolins, a family of scaffolding proteins for organizing “preassembled signaling complexes” at the plasma membrane. J Biol Chem. 1998;273:5419–5922. [CrossRef] [PubMed]
HazlettLD, MasinickS, BarrettR, RosolK. Evidence for asialo GM1 as a corneal glycolipid receptor for Pseudomonas aeruginosa adhesion. Infect Immun. 1993;61:5164–5173. [PubMed]
ZhaoZ, PanjwaniN. Pseudomonas aeruginosa infection of the cornea and asialo GM1. Infect Immun. 1995;63:353–355. [PubMed]
GuptaSK, BerkRS, MasinickS, HazlettLD. Pili and lipopolysaccharide of Pseudomonas aeruginosa bind to the glycolipid asialo GM1. Infect Immun. 1994;62:4572–4579. [PubMed]
YamamotoN, PetrollMW, CavanaghHD, JesterJV. Internalization of Pseudomonas aeruginosa is mediated by lipid rafts in contact lens-wearing rabbit and cultured human corneal epithelial cells. Invest Ophthalmol Vis Sci. 2005;46:1348–1355. [CrossRef] [PubMed]
Araki-SasakiK, OhashiY, SasabeT, et al. An SV40-immortalized human corneal epithelial cell line and its characterization. Invest Ophthalmol Vis Sci. 1995;36:614–621. [PubMed]
FleiszigSM, AroraSK, VanR, RamphalR. FlhA, a component of the flagellum assembly apparatus of Pseudomonas aeruginosa, plays a role in internalization by corneal epithelial cells. Infect Immun. 2001;69:4931–4937. [CrossRef] [PubMed]
WilliamsJM, FiniME, CousinsSW, ReposeJS. Corneal responses to infection.KrachmerJH MannisMJ HollandEJ eds. Cornea. 1997;128–162.Mosby St. Louis.
GipsonI, GrillSM. Cell biology of the corneal epithelium.AlbertDM JacobiecFA eds. Principles and Practice of Ophthalmology. 1994;3–16.WB Saunders Philadelphia.
ScheinOD, GlynnRJ, PoggioEC, SeddonJM, KenyonKR. The relative risk of ulcerative keratitis among users of daily-wear and extended-wear soft contact lenses: a case-control study. Microbial Keratitis Study Group. N Engl J Med. 1989;321:773–778. [CrossRef] [PubMed]
StapletonF, DartJK, SealDV, MathesonM. Epidemiology of Pseudomonas aeruginosa keratitis in contact lens wearers. Epidemiol Infect. 1995;114:395–402. [CrossRef] [PubMed]
ColeN, KrockenbergerM, BaoS, BeagleyKW, HusbandAJ, WillcoxM. Effects of exogenous interleukin-6 during Pseudomonas aeruginosa corneal infection. Infect Immun. 2001;69:4116–4119. [CrossRef] [PubMed]
KunzelmannK, KonigJ, SunJ, et al. Acute effects of parainfluenza virus on epithelial electrolyte transport. J Biol Chem. 2004;279:48760–48766. [CrossRef] [PubMed]
Kubo-WatanabeS, GotoS, SainoT, TazawaY, SatohYI. ATP-induced [Ca2+]i changes in the human corneal epithelium. Arch Histol Cytol. 2003;66:63–72. [CrossRef] [PubMed]
SrinivasSP, YehJC, OngA, BonannoJA. Ca2+ mobilization in bovine corneal endothelial cells by P2 purinergic receptors. Curr Eye Res. 1998;17:994–1004. [CrossRef] [PubMed]
RalevicV, BurnstockG. Receptors for purines and pyrimidines. Pharmacol Rev. 1998;50:413–492. [PubMed]
KlepeisVE, WeingerI, KaczmarekE, Trinkaus-RandallV. P2Y receptors play a critical role in epithelial cell communication and migration. J Cell Biochem. 2004;93:1115–1133. [CrossRef] [PubMed]
LewisRS. Store-operated calcium channels. Adv Second Messenger Phosphoprotein Res. 1999;33:279–307. [PubMed]
TosiMF. Innate immune responses to infection. J Allergy Clin Immunol. 2005;116:241–249.quiz 250 [CrossRef] [PubMed]
DiMangoE, ZarHJ, BryanR, PrinceA. Diverse Pseudomonas aeruginosa gene products stimulate respiratory epithelial cells to produce interleukin-8. J Clin Invest. 1995;96:2204–2210. [CrossRef] [PubMed]
PhilpottDJ, GirardinSE, SansonettiPJ. Innate immune responses of epithelial cells following infection with bacterial pathogens. Curr Opin Immunol. 2001;13:410–416. [CrossRef] [PubMed]
Eaves-PylesT, MurthyK, LiaudetL, ViragL, RossG, SorianoFG. Flagellin, a novel mediator of Salmonella-induced epithelial activation and systemic inflammation: I kappa B alpha degradation, induction of nitric oxide synthase, induction of proinflammatory mediators, and cardiovascular dysfunction. J Immunol. 2001;166:1248–1260. [CrossRef] [PubMed]
GewirtzAT, NavasTA, LyonsS, GodowskiPJ, MadaraJL. Cutting edge: bacterial flagellin activates basolaterally expressed TLR5 to induce epithelial proinflammatory gene expression. J Immunol. 2001;167:1882–1885. [CrossRef] [PubMed]
LiJ, JohnsonXD, IazvovskaiaS, TanA, LinA, HershensonMB. Signaling intermediates required for NF-kappa B activation and IL-8 expression in CF bronchial epithelial cells. Am J Physiol Lung Cell Mol Physiol. 2003;284:L307–L315. [CrossRef] [PubMed]
RheeSH, KeatesAC, MoyerMP, PothoulakisC. MEK is a key modulator for TLR5-induced interleukin-8 and MIP3alpha gene expression in non-transformed human colonic epithelial cells. J Biol Chem. 2004;279:25179–25188. [CrossRef] [PubMed]
TallantT, DebA, KarN, LupicaJ, de VeerMJ, DiDonatoJA. Flagellin acting via TLR5 is the major activator of key signaling pathways leading to NF-kappa B and proinflammatory gene program activation in intestinal epithelial cells. BMC Microbiol. 2004;4:33. [CrossRef] [PubMed]
PaikYH, SchwabeRF, BatallerR, RussoMP, JobinC, BrennerDA. Toll-like receptor 4 mediates inflammatory signaling by bacterial lipopolysaccharide in human hepatic stellate cells. Hepatology. 2003;37:1043–1055. [CrossRef] [PubMed]
KumarA, ZhangJ, YuFS. Innate immune response of corneal epithelial cells to Staphylococcus aureus infection: role of peptidoglycan in stimulating proinflammatory cytokine secretion. Invest Ophthalmol Vis Sci. 2004;45:3513–3522. [CrossRef] [PubMed]
Figure 1.
 
HCECs expressed both asialoGM1 and TLR5. Cell lysates were resolved with Western blot analysis and probed for asialoGM1 and TLR5 receptor expression with specific antibodies (A, B). (A) Sixty, 40, and 20 μg of total proteins were loaded into lanes 1, 2, and 3, respectively. (B) NAWALMA (positive control) and HCEC lysates were loaded into lanes 1 and 2, respectively. (C) AsialoGM1 and TLR5 surface expression on HCECs was analyzed by flow cytometry. Left traces: the isotype control (which was similar to background autofluorescence); right traces: PE-conjugated anti-asialoGM1 and TLR5 labeling.
Figure 1.
 
HCECs expressed both asialoGM1 and TLR5. Cell lysates were resolved with Western blot analysis and probed for asialoGM1 and TLR5 receptor expression with specific antibodies (A, B). (A) Sixty, 40, and 20 μg of total proteins were loaded into lanes 1, 2, and 3, respectively. (B) NAWALMA (positive control) and HCEC lysates were loaded into lanes 1 and 2, respectively. (C) AsialoGM1 and TLR5 surface expression on HCECs was analyzed by flow cytometry. Left traces: the isotype control (which was similar to background autofluorescence); right traces: PE-conjugated anti-asialoGM1 and TLR5 labeling.
Figure 2.
 
Anti-asialoGM1 antibody induced dose-dependent increases in [Ca2+]i. (A) Cells were initially imaged in 100 μL of Ringer’s solution, and after 100 seconds, 100 μL of Ringer’s solution containing a 1:40 (final 1:80) dilution of anti-asialoGM1 antibody was added (arrow). (B) Dose response of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were challenged with Ringer’s solution and anti-rabbit IgG as the vehicle and the negative control, respectively. After 100 seconds of baseline measurements, 100 μL of Ringer’s solution containing a 1:160 to 1:20 (final 1:320–1:40) dilution of anti-asialoGM1 antibody was added. The maximum change in fura-2/AM ratios was elicited by 1:40 (final 1:80) dilution, with EC50 at approximately 1:90 at 45 seconds. Preincubation with 20 μg/mL anti-TLR5 antibody or absence of extracellular Ca2+ (Ca2+-free solution buffered by 0.5 mM EGTA) completely abolished the anti-asialoGM1 antibody–induced calcium transient. One-way ANOVA F8,36 = 43.2815, P < 0.00001; Fisher LSD *P < 0.05, **P < 0.0005, and ***P < 0.0001. (C) Nifedipine caused a dose-dependent reduction of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were preincubated with the indicated concentrations of nifedipine for 30 minutes and then 1:80 dilution of anti-asialoGM1 antibody with the indicated concentrations of nifedipine were added simultaneously. The absence of extracellular Cl, in the presence or absence of nifedipine, significantly reduced the anti-asialoGM1 antibody–induced increase in [Ca2+]i. F4,30 = 39.9758, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 2.
 
Anti-asialoGM1 antibody induced dose-dependent increases in [Ca2+]i. (A) Cells were initially imaged in 100 μL of Ringer’s solution, and after 100 seconds, 100 μL of Ringer’s solution containing a 1:40 (final 1:80) dilution of anti-asialoGM1 antibody was added (arrow). (B) Dose response of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were challenged with Ringer’s solution and anti-rabbit IgG as the vehicle and the negative control, respectively. After 100 seconds of baseline measurements, 100 μL of Ringer’s solution containing a 1:160 to 1:20 (final 1:320–1:40) dilution of anti-asialoGM1 antibody was added. The maximum change in fura-2/AM ratios was elicited by 1:40 (final 1:80) dilution, with EC50 at approximately 1:90 at 45 seconds. Preincubation with 20 μg/mL anti-TLR5 antibody or absence of extracellular Ca2+ (Ca2+-free solution buffered by 0.5 mM EGTA) completely abolished the anti-asialoGM1 antibody–induced calcium transient. One-way ANOVA F8,36 = 43.2815, P < 0.00001; Fisher LSD *P < 0.05, **P < 0.0005, and ***P < 0.0001. (C) Nifedipine caused a dose-dependent reduction of anti-asialoGM1 antibody–induced increases in [Ca2+]i. Cells were preincubated with the indicated concentrations of nifedipine for 30 minutes and then 1:80 dilution of anti-asialoGM1 antibody with the indicated concentrations of nifedipine were added simultaneously. The absence of extracellular Cl, in the presence or absence of nifedipine, significantly reduced the anti-asialoGM1 antibody–induced increase in [Ca2+]i. F4,30 = 39.9758, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 3.
 
ATP transduced Ca2+ transients via P2 receptors induced by anti-asialoGM1 antibody. (A) ATP induced increases in [Ca2+]i. Cells were initially imaged in 100 μL of Ringer’s solution, and, after 100 seconds, 100 μL of Ringer’s solution containing 1000 μM (final 500 μM) of ATPγ-S was added (arrow). (B) Dose–response of ATP-induced increases in [Ca2+]i. F3,16 = 130.6576, P < 0.00001; Fisher LSD *P < 0.0001. (C) The presence of apyrase significantly reduced anti-asialoGM1 antibody–induced increases in [Ca2+]i. F5,24 = 11.7417, P < 0.00001; Fisher LSD *P < 0.005, **P < 0.001 and ***P < 0.0001. (D) ARL67156 significantly enhanced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. The P2 receptor antagonists suramin and PPADS significantly reduced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. F4,20 = 25.6559, P < 0.00001; Fisher LSD *P < 0.005 and **P < 0.00001.
Figure 3.
 
ATP transduced Ca2+ transients via P2 receptors induced by anti-asialoGM1 antibody. (A) ATP induced increases in [Ca2+]i. Cells were initially imaged in 100 μL of Ringer’s solution, and, after 100 seconds, 100 μL of Ringer’s solution containing 1000 μM (final 500 μM) of ATPγ-S was added (arrow). (B) Dose–response of ATP-induced increases in [Ca2+]i. F3,16 = 130.6576, P < 0.00001; Fisher LSD *P < 0.0001. (C) The presence of apyrase significantly reduced anti-asialoGM1 antibody–induced increases in [Ca2+]i. F5,24 = 11.7417, P < 0.00001; Fisher LSD *P < 0.005, **P < 0.001 and ***P < 0.0001. (D) ARL67156 significantly enhanced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. The P2 receptor antagonists suramin and PPADS significantly reduced the anti-asialoGM1 antibody–induced increases in [Ca2+]i. F4,20 = 25.6559, P < 0.00001; Fisher LSD *P < 0.005 and **P < 0.00001.
Figure 4.
 
Anti-asialoGM1 antibody induced ERK1/2 activation. (A) HCECs were grown in 35-mm plates and serum deprived for 24 hours before stimulation with anti-asialoGM1 antibody (1:80 dilution) for the indicated times. Cell lysates were fractionated by SDS-PAGE and then analyzed by Western blot analysis with antibodies directed against phospho-ERK1/2. Membranes were then stripped and reprobed using an antibody against total ERK1/2 as the control for equal protein loading. (B) The effect of 30 minutes’ preincubation of 10 μM U0126, 1 mM EGTA, and 1 μM nifedipine (NF) on anti-asialoGM1 antibody (AGM; 1:80 dilution for 30 minutes)–induced ERK activation, compared with the unstimulated control (CTL).
Figure 4.
 
Anti-asialoGM1 antibody induced ERK1/2 activation. (A) HCECs were grown in 35-mm plates and serum deprived for 24 hours before stimulation with anti-asialoGM1 antibody (1:80 dilution) for the indicated times. Cell lysates were fractionated by SDS-PAGE and then analyzed by Western blot analysis with antibodies directed against phospho-ERK1/2. Membranes were then stripped and reprobed using an antibody against total ERK1/2 as the control for equal protein loading. (B) The effect of 30 minutes’ preincubation of 10 μM U0126, 1 mM EGTA, and 1 μM nifedipine (NF) on anti-asialoGM1 antibody (AGM; 1:80 dilution for 30 minutes)–induced ERK activation, compared with the unstimulated control (CTL).
Figure 5.
 
IL-8 secretion in HCECs stimulated by anti-asialoGM1 antibody: effects of Ca2+ chelators, MEK1 inhibitor, and anti-TLR5 antibody. Confluent monolayers of SV40 and primary HCECs grown in 35-mm plates were pretreated with the compounds indicated for 30 minutes before 8 hours of exposure to anti-asialoGM1 antibody (1:80 dilution), maintained in the presence of the indicated compounds. The secreted IL-8 in cell supernatants was measured by ELISA. (▪) SV40 HCECs; (□) primary HCECs. F11,24 = 54.8384, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 5.
 
IL-8 secretion in HCECs stimulated by anti-asialoGM1 antibody: effects of Ca2+ chelators, MEK1 inhibitor, and anti-TLR5 antibody. Confluent monolayers of SV40 and primary HCECs grown in 35-mm plates were pretreated with the compounds indicated for 30 minutes before 8 hours of exposure to anti-asialoGM1 antibody (1:80 dilution), maintained in the presence of the indicated compounds. The secreted IL-8 in cell supernatants was measured by ELISA. (▪) SV40 HCECs; (□) primary HCECs. F11,24 = 54.8384, P < 0.0001; Fisher LSD *P < 0.0001.
Figure 6.
 
Colocalization of asialoGM1 and TLR5 in HCECs. For immunoprecipitation, 1.3 mg of protein was immunoprecipitated (IP) with 6 μg of anti-TLR5 (TLR5), 15 μg of anti-asialoGM1 (AGM), and 15 μg of nonspecific immunoglobulins (IgG) with 20 μL of protein A/G mixture, or with the agarose beads (Beads) alone. The samples were subjected to SDS-PAGE and probed (WB) with rabbit anti-asialoGM1.
Figure 6.
 
Colocalization of asialoGM1 and TLR5 in HCECs. For immunoprecipitation, 1.3 mg of protein was immunoprecipitated (IP) with 6 μg of anti-TLR5 (TLR5), 15 μg of anti-asialoGM1 (AGM), and 15 μg of nonspecific immunoglobulins (IgG) with 20 μL of protein A/G mixture, or with the agarose beads (Beads) alone. The samples were subjected to SDS-PAGE and probed (WB) with rabbit anti-asialoGM1.
Figure 7.
 
Events in host cell signaling after asialoGM1 ligation. Anti-asialoGM1 antibody binds to its receptor, a membrane glycolipid, which is associated with TLR5. This causes the extracellular release of ATP. Membrane surface adenylate cyclase (AC) catalyzes the conversion of ATP to cAMP, which then acts on cAMP gCl-sensitive channels, leading to Cl secretion. The secretion results in depolarization of the plasma membrane, thus activating L-type voltage-dependent Ca2+ channels, leading to Ca2+ influx. Ca2+ mobilization, as a result of Ca2+ influx, leads to phosphorylation of MEK1/2 and ERK1/2 and to IL-8 production.
Figure 7.
 
Events in host cell signaling after asialoGM1 ligation. Anti-asialoGM1 antibody binds to its receptor, a membrane glycolipid, which is associated with TLR5. This causes the extracellular release of ATP. Membrane surface adenylate cyclase (AC) catalyzes the conversion of ATP to cAMP, which then acts on cAMP gCl-sensitive channels, leading to Cl secretion. The secretion results in depolarization of the plasma membrane, thus activating L-type voltage-dependent Ca2+ channels, leading to Ca2+ influx. Ca2+ mobilization, as a result of Ca2+ influx, leads to phosphorylation of MEK1/2 and ERK1/2 and to IL-8 production.
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×