Activation of Focal Adhesion Kinase in Adenovirus-Infected Human Corneal Fibroblasts

Abstract

purpose. Focal adhesion kinase (FAK), a nonreceptor protein tyrosine kinase with protean downstream influences on cell cycle regulation, cytoskeletal dynamics, and cell attachment, is activated by integrin binding and aggregation. Adenoviruses, including those associated with human keratitis, enter permissive cells by an integrin-mediated mechanism. Hence, a possible relationship between adenovirus infection and tyrosine phosphorylation of FAK in human corneal cells was explored.

methods. Human corneal fibroblasts (HCFs) derived from donor corneas were infected for various periods with adenovirus type 19 (Ad19) or were mock infected with virus-free medium. Parallel experiments included echistatin, which is a snake venom disintegrin and a partial inhibitor of FAK. For immunoblot analysis, Triton-X–soluble and Triton-X–insoluble fractions were analyzed by SDS-PAGE and immunoblotted with phosphospecific antibodies. Expression of the interleukin (IL)-8 gene was analyzed by RT-PCR and ELISA.

results. Ad19 infection of HCFs induced tyrosine phosphorylation of a protein at 125 kDa, which was evident within 15 minutes after infection and was established by immunoprecipitation to be FAK. Immunoblot with antibody to FAK tyrosine-397 confirmed phosphorylation of this key binding site for downstream signaling proteins. Immunoblot analysis further suggested a shift in the intracellular location of phosphorylated FAK from the cytosol (Triton-X–soluble cell lysate fraction) to the cytoskeleton (Triton-X–insoluble cell lysate pellet) on infection. Finally, treatment of HCFs with echistatin reduced virus-induced expression of the neutrophil chemotaxin IL-8, previously implicated in adenoviral pathogenesis.

conclusions. Ad19 infection of HCFs induces rapid phosphorylation of FAK, and a dramatic change in its intracellular distribution. Activation of FAK may contribute to the inflammatory response to adenovirus infection of the human cornea.

Keratocytes, the resident cells within the corneal stroma, maintain the stroma in a precisely organized and transparent state. Keratocytes also capably amplify acute stromal inflammation in vitro by the secretion of chemokines ^{1 2 3 4} and appear to contribute to necrotizing stromal inflammation in vivo due to herpes simplex virus ⁵ and Gram-negative bacteria. ^{6 7} Thus, keratocytes play an important role in inflammatory responses to pathogens in the corneal stroma. 

Adenoviruses are significant human pathogens. ⁸ Forty-nine serotypes subdivide into six distinct subgroups (A–F) on the basis of genetic sequencing. Adenovirus subgroups associate broadly with specific clinical syndromes. Epidemic keratoconjunctivitis (EKC), the only adenoviral syndrome that significantly affects the cornea, occurs most commonly with infection by subgroup D adenoviruses of serotype 8, 19, or 37. The keratitis of EKC is distinctive and is clinically defined by delayed-onset, multifocal, subepithelial (stromal) corneal infiltrates. Subepithelial infiltrates can cause long-lasting ocular morbidity, including reduced vision, foreign-body sensation, and photophobia. The infiltrating cells within the subepithelial corneal stroma in experimental animal models of adenovirus infection are primarily polymorphonuclear neutrophils in the early stages, ⁹ followed later by lymphocytes. ¹⁰ We recently demonstrated that infection of cultured human keratocytes with adenovirus type 19 (Ad19) induces secretion of neutrophil chemotaxins including interleukin (IL)-8, ⁴ consistent with the neutrophil infiltration observed in animal models of corneal adenovirus infection. Therefore, expression of IL-8 by infected keratocytes may cause the formation of subepithelial infiltrate in human EKC. 

In immortalized (nonocular) cell lines, the interaction of adenoviruses with the cell surface integrins αvβ5 and αvβ3 ^{11 12} mediates internalization of the virus by a phosphatidylinositol 3-kinase (PI3K)–dependent mechanism. ^{13 14} Other signal transduction events lead to chemokine upregulation and may induce host inflammatory responses to adenoviral gene vectors. ¹⁵ In the work presented herein, we examine the relationship between adenovirus infection of human corneal cells and the activation of focal adhesion kinase (FAK), a 125-kDa intracellular signaling molecule activated by binding and aggregation of αvβ5 and αvβ3 ¹⁶ and critical in a broad array of cellular functions (reviewed in Ref. ¹⁷ ). 

Materials and Methods

Immunologic Reagents

Alkaline phosphatase–conjugated goat anti-rabbit and goat anti-mouse IgG were obtained from Bio-Rad (Richmond, CA), horseradish peroxidase-conjugated donkey anti-rabbit IgG from Amersham (Arlington Heights, IL), and the chemiluminescent reagents from Amersham. The mAb-P-Y and N-terminus–specific polyclonal antibody to FAK were obtained from Santa Cruz Biotechnology (Santa Cruz, CA). The polyclonal antibody to phosphorylated tyrosine residue 397 (FAK tyrosine-397) was obtained from Biosource (Camarillo, CA). Echistatin and other chemical reagents were purchased from Sigma (St. Louis, MO) and Pierce (Rockford, IL). 

Cells and Viruses

Primary keratocytes were derived from donor corneas (North Florida Lions Eye Bank, Jacksonville, FL), as previously described. ² Briefly, after mechanical debridement of the corneal epithelium and endothelium, corneas were cut into 2-mm diameter sections, and each section placed in individual wells of six-well tissue culture plates (Falcon; Fisher Scientific, Pittsburgh, PA) with Dulbecco’s modified Eagle’s medium (DMEM), containing 10% heat inactivated fetal bovine serum (FBS), penicillin G sodium, and streptomycin sulfate. Corneal fragments were removed before confluence of each cell culture. Cells were grown and maintained at 37°C in 5% CO₂. Cells from multiple donors were pooled and the cell monolayers used at passage 3. After serial passages in serum-containing medium, keratocytes maintain a fibroblast phenotype ¹⁸ and are referred to in the remainder of this article as human corneal fibroblasts (HCFs). A fibroblast phenotype was confirmed by indirect immunofluorescence staining with polyclonal anti-vimentin (positive reactivity) and anti-cytokeratin (no reactivity) antibodies, by methods described previously. ⁴ All experiments presented were performed with primary HCFs at passage 3. 

Ad19 was cultured directly from the cornea of a patient with EKC and grown in human lung carcinoma cells (A549 cells, CCL 185; American Type Culture Collection, Manassas, VA) in minimum essential medium (MEM) with 2% FBS, penicillin G sodium, streptomycin sulfate, and amphotericin B. The State of Oklahoma Department of Health confirmed the viral serotype. Typical adenoviral cytopathic effect, positive immunofluorescence staining for adenovirus hexon proteins, and increasing titers of virus within 1 week after infection of HCFs with this virus have been described. ⁴ Virus was purified from infected A549 cells on CsCl gradients. Purified virus was dialyzed against a 10 mM Tris (pH 8.0) buffer with 80 mM NaCl, 2 mM MgCl₂ and 10% glycerol. The tissue culture infectious dose (TCID) of this Ad19 preparation was determined to be 2 × 10⁸ TCID/mL, and stock preparations were stored at −80°C. 

Adenoviral Infection for Cellular Protein Recovery

Because of prior evidence that adenovirus-induced intracellular signaling occurs within several minutes of exposure to virus, ¹⁵ we chilled the cells before infection and maintained them at 4°C for 1 hour after infection to allow synchronous onset of intracellular signaling in the presence of adenovirus bound to its host cell receptor. HCFs grown to 95% confluence in six-well plates were first washed gently four times with chilled MEM containing 2% FBS and then kept at 4°C for 35 minutes before infection in duplicate with cold Ad19 in MEM 2% FBS at a multiplicity of infection of 50 or application of cold MEM+2% FBS without virus as a control. After a 1-hour viral adsorption at 4°C, HCFs were allowed to incubate at 37°C for various times before harvest. A multiplicity of infection of 50 was chosen to ensure complete and synchronous activation of adenoviral receptors. Cells were harvested by adding 200 μL of chilled lysis buffer consisting of phosphate-buffered saline (PBS), 1% Triton X-100, and 2 mM EDTA, along with protease inhibitors, including phenylmethylsulfonyl fluoride (1 mM), pepstatin A (5 μg/mL), leupeptin (10 μg/mL), and aprotinin (10 μg/mL), and incubated at 4°C for 5 minutes. Cells were scraped from the plates with a cell scraper, and the cell lysates collected in microfuge tubes and centrifuged at 13,000g for 30 minutes at 4°C. The supernatants were separated from the lysate pellets, and both were frozen for later use. 

SDS-PAGE and Immunoblot Analysis

SDS-PAGE on solubilized proteins was performed on 8% acrylamide gels, after spectrophotometric bicinchoninic acid (BCA) analysis to normalize protein loading. To quantify the protein concentration of Triton-X–insoluble cell lysate pellets, the pellets were resuspended in 3× sample buffer (60 mM Tris-HCl [pH 6.8], 10% glycerol, 0.02% (wt/vol) SDS, 0.05% bromophenol blue (wt/vol), and 5% β-mercaptoethanol), and quantified by amidoschwarz protein assay. ¹⁹ For SDS-PAGE on Triton-X–insoluble lysate pellets containing cytoskeletal and associated proteins, pellets were resuspended in 3× sample buffer, boiled for 4 minutes, and immediately loaded on paired separating gels (10% resolving gel, 4% stacking gel). After SDS-PAGE, one gel was stained with Coomassie blue as additional confirmation of equal protein loading of the pellet samples, and the other gel was transferred to nitrocellulose by means of a transblot apparatus (Mini-Protean II, Bio-Rad). SDS-PAGE gels containing solubilized proteins from lysate supernatants were similarly transferred. Nitrocellulose sheets were blocked overnight at 4°C with Tris-buffered-saline (TBS: 10 mM Tris-HCl [pH 7.4] and 300 mM NaCl containing 0.1% Tween-20) containing 5% crystalline grade bovine serum albumin. Incubations with primary antisera (diluted in blocking buffer) were performed for 2 hours at room temperature. Immunoblots were washed three times with TBS after both the primary and secondary antibody incubations. Antibody reactivity was determined with enhanced chemiluminescent reagents with peroxidase-conjugated sheep anti-mouse antibody or donkey anti rabbit-secondary antibodies. Before reprobe of developed blots, blots were incubated in 62.5 mM Tris solution (pH 6.5) containing 2% SDS and 0.1 M β-mercaptoethanol at 50°C for 30 minutes. Each Western blot experiment was performed at least three times. 

Densitometric scans of immunoblots were analyzed by computer (One-DScan software; Scanalytics, Billerica, MA) in the linear range of detection and absolute values were then normalized. Densitometry parameters included threshold and bandwidth, and were set identically for tested bands. 

Immunoprecipitation

Cells were solubilized by suspension in 200 μL ice-cold PBS (pH 7.4) containing 1% Triton-X-100, 2 mM EDTA, and 1 mM orthovanadate (lysis buffer) for 5 minutes, followed by centrifugation at 13,000g for 30 minutes at 4°C. The supernatants were removed and precleared by mixing with 50 μL protein-A agarose (5 mg/mL) for 30 minutes with gentle mixing at 4°C, followed by centrifugation at 6000g. The cleared supernatants were then mixed with 5 μg of mAb-P-Y or anti-FAK and 50 μL protein A agarose in lysis buffer containing 1 mM orthovanadate and maintained overnight at 4°C. The following morning, the mixture was centrifuged at 6000g for 5 minutes, the supernatant removed, and the immunoprecipitates washed three times with 300 μL lysis buffer. The immune complex obtained was solubilized in SDS-PAGE sample buffer, and boiled for 3 minutes. The proteins were resolved using 8% acrylamide gels and immunoblot analysis performed as described earlier. 

Viral Infection for RNA isolation

Monolayer HCF cultures grown to 95% confluence in flasks were washed in MEM (GibcoBRL, Life Technologies, Rockville, MD), and infected with purified Ad19 in MEM+2% FBS (GibcoBRL) at a multiplicity of infection of 50 or MEM+2% FBS with dialysis buffer as a control. Virus was adsorbed for 1 hour at 37°C and then incubated for another hour before addition of reagents for isolation of RNA from the cells. 

Total RNA was isolated by a single-step RNA isolation method with extraction reagent (TRIzol; GibcoBRL), according to the instructions provided by the manufacturer. Briefly, cells were lysed with the reagent at room temperature. Proteins were removed by chloroform extraction of the lysate. RNA was precipitated from the supernatant with ethanol, and the pellet was resuspended in Tris-EDTA (pH 8.0). Ribonuclease inhibitor (RNAsin; Promega, Madison, WI) was added to the RNA solution to prevent RNase action. Contaminating DNA was removed by DNase I (Promega) treatment followed by a phenol-chloroform extraction and subsequent ethanol precipitation of the RNA. The RNA was resuspended at a concentration of 1 mg/mL in diethyl pyrocarbonate (DEPC)-treated water. A spectrophotometric reading at a wavelength of 260 nm was used to determine the concentration of RNA. The quality of each RNA sample was determined by calculating the ratio of optical density of each RNA sample at 260 to 280 nm. A ratio of approximately 1.8 indicated that samples contained only nondegraded RNA. 

Reverse Transcription—Polymerase Chain Reaction

For synthesis of cDNAs, 5 μg total RNA was reverse transcribed with Moloney murine leukemia virus reverse transcriptase (M-MLV; Promega), with an oligo-dT 15mer (Promega) used as the primer. The reaction mixture for the reverse transcription reaction was composed of 1.5 U/μL of ribonuclease inhibitor (RNasin; Promega), 50 mM Tris-HCl (pH 8.3), 75 mM KCl, 3 mM Mgcl2, 10 mM dithiothreitol, 500 μM dNTPs, and 10 U/μL of M-MLV reverse transcriptase. A reaction without reverse transcription, which comprised all these reactants except for the reverse transcriptase enzyme, was run with each experiment to rule out the possibility of contaminating genomic DNA amplification in the PCR reaction step. 

For PCR amplification, interleukin-8 primers were designed based on the GenBank human IL-8 complete cDNA sequence (Unigene Hs.624; provided in the public domain by the National Center for Biotechnology Information, Bethesda, MD) and amplified a 481 bp-product (forward, nucleotide [nt] 489–508: 5′-GTGTGGGTCTGTTGTAGGGT-3′, and reverse, nt 951–970: 5′-CTGTGAGGTAAGATGGTGGC-3′). GAPDH primers were generated from the human GAPDH sequence (GenBank accession No. X01677) and amplified a 165-bp product (forward, nt 76–96: 5′-GTCGGAGTCAACGGATTTGGTCGT-3′, and reverse, nt 226–240: 5′-GACGGTGCCATGGAATTTGCCATG-3′). Two microliters of the cDNA obtained by reverse transcription was used with the PCR reaction mixture composed of 50 mM Tris-HCl (pH 9.0), 50 mM NaCl, 10 mM MgCl₂, 200 μM dNTPs, and 20 μg/μL primers. Thin-walled PCR reaction tubes were used for the reactions and the assay was performed on a programmable thermominicycler (MJ Research, Watertown, MA), by using one cycle each that comprised two successive denaturation steps at 94°C for 2 minutes and 1 minute, 50°C for 2 minutes, 72°C for 3 minutes with a 2-second increase in extension temperature with every cycle, followed by another 72°C for 5 minutes. This cycle was then repeated 30 times. The amplification products were analyzed by gel electrophoresis in 1% agarose gels and the sizes of the amplicons were verified by comparing them with a 100-bp DNA marker (GibcoBRL). In separate experiments, multiplex RT-PCR for the presence of transcripts IL-1β, TNF-α, IL-6, granulocyte-macrophage–colony-stimulating factor (GM-CSF), TGF-α, GAPDH, and IL-8 was performed using a kit (CytoXpress; Biosource), according to the manufacturer’s instructions. Each PCR experiment was performed at least five times. 

IL-8 ELISA

HCFs were infected in 48-well tissue culture plates with purified Ad19 or were mock infected reduced serum medium (OptiMem; GibcoBRL) as a control. Cell supernatants were harvested 4 hours after infection, and the amount of IL-8 protein quantified with calorimetric sandwich enzyme–linked immunosorbent assay kit (Quantikine; R&D Systems, Minneapolis, MN) with a linear sensitivity range from 31.25 to 2000 pg/mL. Plates were read on a microplate reader (Emax; Molecular Devices, Sunnyvale, CA) and analyzed by computer (SOFTmax; Molecular Devices). The ELISA experiment was performed three times. For each ELISA experiment, the means of triplicate ELISA results for each of the virus- or mock-infected wells were compared by Student’s t-test. 

Results

Tyrosine Phosphorylation of FAK after Ad19 Infection of HCFs

In preliminary experiments, HCFs demonstrated increased tyrosine phosphorylation of a 125-kDa protein within 5 to 15 minutes after Ad19 infection, compared with mock-infected control cells (data not shown). To confirm an increase in FAK tyrosine phosphorylation after Ad19 infection, we immunoprecipitated cell lysates from 15-minute mock- and Ad19-infected cells with either mAb-P-Y or anti-FAK. The immune complexes were resolved by SDS-PAGE and transferred to nitrocellulose and the respective blots probed with anti-FAK (Fig. 1A) or mAb-P-Y (Fig. 1B) . The amount of FAK protein recovered from mAb-P-Y immunoprecipitates was 13 times greater after Ad19 infection than with mock infection (Fig. 1A) , whereas the amount of tyrosine phosphorylated protein recovered from anti-FAK immunoprecipitates was almost three times greater with infection (Fig. 1B) . We then compared mock- and Ad19-infected cells for phosphorylation at tyr-397 (Fig. 1C) , an autophosphorylation site essential to activation of signaling proteins with an Src homology 2 domain ²⁰ (see Ref. ²¹ for a review) By densitometry, we observed a ninefold increase in phosphorylation of FAK at this key activation site upon infection. 

FAK Moves from the Triton-X–Soluble to the Triton-X–Insoluble Protein Fraction after Ad19 Infection of HCFs

To test whether the apparent increase in tyrosine phosphorylation of HCF FAK after Ad19 infection is due to changes in the cellular location of FAK, a mAb-P-Y immunoblot from HCFs infected 15 minutes before harvest (Fig. 2 , anti-P-Y) was stripped in 0.1 M β-mercaptoethanol and 2% SDS, washed, and reprobed with anti-FAK. Infection appeared to reduce the total amount of FAK in the Triton-X–soluble protein fraction, suggesting the possibility that infection with Ad19 increases the association of FAK with Triton-X–insoluble cytoskeletal proteins. Taking into account an overall decrease (2.6-fold) in the total amount of soluble FAK after infection, as quantified by densitometric analysis on the anti-FAK blot (Fig. 2) , FAK phosphorylation, as measured by densitometry on the blot probed with mAb-P-Y, was increased more than 16-fold in HCFs after 15 minutes of infection with Ad19. 

To test whether the apparent reduction of total FAK in the Triton-X–soluble protein fraction after infection might be due to an increased association of FAK with the Triton-X–insoluble protein fraction, equal amounts of the Triton-X–insoluble pellet (as quantitated by the amidoschwarz assay) were subjected to SDS-PAGE and immunoblotted with anti-FAK antibody. More than twice as much FAK was apparent in the Triton-X–insoluble pellet of Ad19-infected cells compared with the mock-infected control (Fig. 3A ; increase of 2.2-fold after 15 minutes of infection). Coomassie blue staining of a gel run in parallel showed equal loading of total protein from infected versus uninfected cells (Fig. 3B) and confirmed the veracity of the amidoschwarz assay. In a separate experiment, the Triton-X–insoluble protein pellets from Ad19- and mock-infected HCFs were subjected to SDS-PAGE, immunoblotted with mAb-P-Y, and then stripped, washed, and reprobed with anti-FAK. Densitometric analysis, even taking into account the approximate twofold increase in FAK in the pellet after Ad19 infection, showed a 37-fold greater degree of phosphorylation of Triton-X–insoluble FAK after infection. Taken together, these data suggest that Ad19 infection of HCFs induces both activation of FAK as a tyrosine kinase and increased association of activated FAK with cytoskeletal (Triton-X–insoluble) cell components. 

Echistatin Reduces FAK Tyrosine Phosphorylation in HCFs after Ad19 Infection

To examine a possible downstream effect of FAK phosphorylation in Ad19-infected HCFs, we first tested the impact of the disintegrin and FAK inhibitor, echistatin, ^{22 23 24} on FAK phosphorylation in cultured HCFs (Fig 4) . HCFs were incubated in echistatin (0.2 μM) or its solvent (ddH₂O) for 18 hours before mock infection or infection with Ad19. Solubilized proteins were resolved by SDS-PAGE and immunoblotted with mAb-P-Y (Fig. 4 , top). The blot was then stripped, washed, and reprobed with anti-FAK tyr-397 (Fig. 4 , middle). Finally, the same blot was again stripped, washed, and reprobed with anti-FAK (Fig. 4 , bottom). Densitometric analysis normalized on the basis of the lower blot readings showed that, in HCF, echistatin blocked approximately 25% of infection-induced FAK tyrosine phosphorylation and approximately 75% of infection-induced phosphorylation at the key autophosphorylation site of FAK, tyr-397. 

Inhibition of FAK with the Disintegrin Echistatin Reduces IL-8 Expression by Ad19-Infected HCFs

Based on the apparent inhibition of FAK phosphorylation by echistatin, we sought to use this inhibitor to examine a putative downstream effect of phosphorylation and activation of FAK. We have shown that Ad19 infection of cultured human keratocytes induces expression of the neutrophil chemotaxin IL-8, ⁴ implicating IL-8 in the pathogenesis of corneal stromal inflammation after adenovirus infection. Therefore, we infected HCFs in the presence or absence of echistatin (Fig. 5) . RT-PCR performed on samples harvested 1 hour after Ad19 infection showed that echistatin noticeably reduced IL-8 mRNA levels (Fig. 5A) . Multiplex RT-PCR for other transcripts, including IL-1β, TNF-α, and IL-6, revealed no increases after viral infection for 1 hour and no compensatory increases due to the presence of echistatin (data not shown). Supernatants from echistatin-treated HCFs 4 hours after Ad19 infection demonstrated a significant reduction by ELISA in Ad19-induced IL-8 protein secretion (3406 ± 1146 pg/mL in ddH₂O-pretreated vs 1286 ± 189 pg/mL in echistatin-pretreated cell supernatants, Fig. 5B , P < 0.02). 

Discussion

To gain insight into the early host cell responses to adenovirus infection of the cornea, we infected HCFs with a human corneal isolate of Ad19 and studied the tyrosine phosphorylation of FAK, a cellular tyrosine kinase with protean downstream effects. Bruder and Kovesdi ¹⁵ have demonstrated that adenovirus-induced intracellular signaling in immortalized human cervical carcinoma cells occurs within several minutes of exposure to virus. In our studies, infection of HCFs with a corneal adenoviral isolate reproducibly induced tyrosine phosphorylation of FAK within 15 minutes of warming, as confirmed by immunoblot with an anti-phosphotyrosine antibody, an anti-tyr-397 antibody, and two immunoprecipitation strategies. In addition, Ad19 infection of HCFs induced a shift of FAK from the Triton-X–soluble cellular proteins to the Triton-X–insoluble cellular compartment. Tyrosine phosphorylation of FAK in vitro has been demonstrated after exposure of HCFs to fibronectin ²⁵ and mechanical injury. ²⁶ Our findings represent the first demonstration of virus-induced FAK phosphorylation in human corneal cells and the first indication of signal transduction events in any cell type after infection with a subgroup D adenovirus. 

A nonreceptor protein tyrosine kinase localized at focal adhesions, FAK participates in diverse cellular functions including survival, attachment, proliferation and cell cycle regulation. ^{17 27 28} Binding of cell surface integrins to extracellular matrix proteins can induce the tyrosine phosphorylation of FAK, ^{16 29} and activated FAK has been implicated in cytoskeletal reorganization ³⁰ and the internalization of pathogenic bacteria. ³¹ Upon integrin clustering through the binding of a multivalent ligand, FAK is autophosphorylated at tyr-397, creating a high-affinity binding site for the SH2 domains of Src family kinases. ³² The interaction between FAK and Src kinases leads to phosphorylation of FAK on additional tyrosine residues and subsequent activation of downstream mitogen activated protein kinases (MAPKs) that in turn mediate cellular cytokine gene transcription. ³³ In this context, it is interesting that infection of human lung carcinoma cells by Ad5 (subgroup C)-derived gene vectors induce MAPK-mediated chemokine synthesis in vitro, ¹⁵ providing a possible mechanism for adenoviral gene vector-induced inflammation in vivo. 

The nonenveloped protein capsid of the adenovirus forms a regular icosahedron with 20 triangular surfaces composed of 240 hexons and 12 vertices. A single penton with a base and projecting fiber makes up each vertex. For most adenoviral subgroups, ³⁴ the projecting penton fiber serves as the ligand for the cellular adenovirus receptor. ³⁵ However, a requisite secondary interaction must also occur between the reiterated Arg-Gly-Asp (RGD) motif in the penton base and host cell α_v integrin transmembrane proteins ¹¹ for successful internalization of the adenovirus. For Ad2, virus internalization rapidly follows αv integrin–mediated activation of PI3K, which in turn activates downstream members of the Rho family of small guanosine triphosphate (GTP)–binding proteins, inducing polymerization of cytoskeletal actin, ¹⁴ and receptor-mediated endocytosis through clathrin-coated pits. ³⁶ After adenovirus internalization, the virus capsid hexon interacts with intracytoplasmic microtubules ³⁷ that assist viral transport to the nucleus for eventual viral gene transcription. Interactions between Ad19 and HCF cell membrane integrins probably mediate early intracellular events, as suggested by the rapidity (within 15 minutes of infection) of change in the phosphorylation state and cellular location of FAK seen in our studies. 

Recent reports suggest that Ad2 binding to immortalized human colon carcinoma cells initiates tyrosine phosphorylation of FAK. ¹³ However, although PI3K can be a downstream target of FAK, ²⁰ FAK phosphorylation does not appear essential for viral internalization. ¹³ In our studies, use of echistatin, a snake venom disintegrin and an inhibitor of FAK phosphorylation, blocked downstream chemokine expression by Ad19-infected HCFs. Echistatin also inhibits other focal adhesion-associated signaling events, including phosphorylation of paxillin and syk. ^{23 24} Verification of a specific relationship between activation of FAK and expression of IL-8 in Ad19-infected keratocytes awaits identification of a truly FAK-specific inhibitor. Nonetheless, our data suggest that the signaling cascade necessary for adenovirus internalization may be distinct or divergent from that leading to activation of MAPK and synthesis of proinflammatory cytokines. ^{13 14} Our prior published work ⁴ taken in context with the results reported herein suggest that signaling pathways are activated in human keratocytes after infection with tropic subgroup D adenoviruses, and that these pathways may play a role in the ocular pathogenesis of Ad19 infection of the cornea. 

 Supported by Grants EY00357, EY13124, and EY12190 from the National Institutes of Health; a Career Development Award (JC) from Research to Prevent Blindness; and an Honors Research Program grant (RSS) from the University of Oklahoma Health Sciences Center College of Medicine.

 Submitted for publication December 12, 2001; revised April 3, 2002; accepted April 26, 2002.

 Commercial relationships policy: N.

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

 

Figure 1.

 

Determination of FAK tyrosine phosphorylation in HCFs in response to Ad19 infection. Cell cultures were infected with Ad19 or were mock infected (M) with virus-free medium, and after 1 hour of adsorption at 4°C, were incubated at 37°C for 15 minutes. The cells were solubilized and (A) immunoprecipitated (IP) with mAb-P-Y, with the immunoprecipitates subjected to SDS-PAGE and immunoblot analysis with anti-FAK, or (B) immunoprecipitated with anti-FAK, with immunoprecipitates subjected to SDS-PAGE and immunoblotted with mAb-P-Y, or (C) subjected to SDS-PAGE and immunoblotted with antibody specific for phosphorylated FAK tyr-397. Normalized densitometric values are shown above each band. In each case, Ad19 infection of HCFs increased the degree of tyrosine phosphorylation of FAK.

Figure 2.

 

Immunoblot comparing solubilized protein from virus- and mock-infected HCFs for amount of FAK and degree of tyrosine phosphorylation. HCFs were infected with Ad19 or were mock infected (M) with virus-free medium at 4°C. After a 1-hour adsorption at 4°C, HCFs were incubated at 37°C for 15 minutes before harvest. Solubilized proteins were loaded on separating gels for SDS-PAGE, transferred to a nitrocellulose membrane, and probed for phosphorylated tyrosine residues with mAb-P-Y. The nitrocellulose membrane was then stripped, washed, and reprobed with anti-FAK. Total Triton-X–soluble FAK appeared to decrease after infection with Ad19. Densitometric values are shown, normalized for the decrease in total FAK in the solubilized protein component.

Figure 3.

 

Effect of Ad19 infection on Triton-X–insoluble HCF proteins. Cell cultures were infected with Ad19 or were mock infected (M) with virus-free medium, and after 1 hour of adsorption at 4°C, were incubated at 37°C for 15 minutes. The cells were solubilized, centrifuged, and the supernatants removed. The remaining Triton-X–insoluble pellets were resuspended in sample buffer, quantified by amidoschwarz protein assay, boiled, divided, and loaded on two separating gels for SDS-PAGE. Paired gels were run in parallel and then transferred to nitrocellulose and immunoblotted with anti-FAK (A), stained with Coomassie blue to assure equal loading of the gels (B), or immunoblotted with mAb-P-Y and then stripped, washed, and reprobed with anti-FAK (C). Normalized densitometric values given in (A) and (C) demonstrate that Ad19 infection induced a shift in the location of FAK from the Triton-X–soluble to the Triton-X–sinsoluble protein compartment.

Figure 4.

 

Effect of the FAK inhibitor echistatin on tyrosine phosphorylation of soluble FAK in Ad19-infected HCFs. Cell cultures were infected with Ad19 or were mock infected (M) with virus-free medium in the presence or absence of echistatin (0.2 μM). After 1 hour of adsorption at 4°C, cells were incubated at 37°C for 15 minutes. Solubilized proteins were separated on gels for SDS-PAGE, transferred to nitrocellulose, and probed for phosphorylated tyrosine residues with mAb-P-Y. The nitrocellulose membrane was stripped, washed, and reprobed with anti-FAK-tyr-397, again stripped and washed, and probed with anti-FAK. Normalized densitometric values adjusted to reflect the total quantity of FAK in each lane based on densitometry of the bottom panel reflect partial inhibition of the effect of Ad19 infection on FAK phosphorylation in cultured HCFs.

Figure 5.

 

Effect of echistatin on expression of IL-8 by Ad19-infected HCFs. (A) Cell cultures were infected with Ad19 or were mock infected (M) with virus-free medium in the presence or absence of echistatin (0.2 μM). After 1 hour of adsorption at 4°C, cells were incubated for an additional hour at 37°C. Total RNA was isolated, reverse transcribed, and subjected to PCR with primers specific for IL-8 and the control gene GAPDH. In a separate experiment (B), mock- and Ad19-infected HCFs were incubated for 4 hours at 37°C before collection of cell supernatants for quantification of expression of IL-8 by ELISA. The means of results in triplicate ELISAs for each of the virus- or mock-infected wells were compared by Student’s t-test. Ad19 infection significantly increased the quantity of secreted IL-8 protein (P < 0.01) compared with mock infection, and echistatin significantly reduced Ad19 infection-driven IL-8 expression (P < 0.02).

The authors thank Mary Butler and Roger Astley for technical assistance.