Syk Associates with Clathrin and Mediates Phosphatidylinositol 3-Kinase Activation during Human Rhinovirus Internalization¹

Human rhinovirus (HRV),³a cause of the common cold, is the most common acute infection in humans. Although HRV infections are self-limited in healthy individuals, they cause significant morbidity in patients with underlying lung diseases such as asthma and chronic obstruction pulmonary disease where HRV infections are associated with up to 80 and 50% of exacerbations, respectively (1, 2). The underlying mechanism is believed to be related to the ability of HRV to induce airway epithelial expression of inflammatory mediators (3) upon binding to its major receptor, ICAM-1. HRV-ICAM-1 binding is sufficient to initiate a signaling cascade that leads to the expression of inflammatory mediators such as IL-8, IL-1, IL-6, and GM-CSF (4, 5), with perpetuation of the inflammatory response occurring following viral cell entry and replication (6, 7). Studies with human airway epithelial cell lines (5, 6, 7, 8, 9) and primary cells (10) have indicated that the epithelium is a primary source of the inflammatory mediators induced by HRV. In addition, HRV infection also preferentially up-regulates the membranous form of ICAM-1 while concomitantly decreasing soluble ICAM-1, thus providing a positive feedback loop for increasing infectivity and airway epithelial cell activation (11).

We have recently identified the protein kinase Syk as being an early signaling molecule that regulates replication-independent, HRV-induced p38 MAPK activation and IL-8 expression (4). We also demonstrated that HRV induces the recruitment of Syk to ICAM-1 in conjunction with ezrin, a cytoskeletal linker protein that binds to ICAM-1 in vivo (12) and possesses an ITAM, the canonical Syk binding motif (13). The time course of these signaling events (10–90 min) and the observation that an irradiated replication-deficient form of the virus activated the same signaling events (4) identified the Syk signaling pathway as being independent of viral replication. HRV has also been found to induce a second signaling pathway in a replication-independent manner involving the activation of PI3K and recruitment of the PI3K substrate Akt to HRV-containing endosomes (14). Pharmacologic inhibition of PI3K impaired not only viral internalization but also IL-8 expression in an NFκB-dependent manner. A role for Syk in the regulation of PI3K in airway epithelial cells is not known.

In leukocytes, Syk is a known upstream regulator of PI3K in the BCR and FcγR signaling pathways (15, 16) and plays an important role in mediating the internalization of extracellular material upon receptor binding (reviewed in Ref. 17). Although this has been most extensively studied in the context of the phagocytosis of IgG-opsonized particles via the FcγR (18, 19, 20, 21, 22, 23, 24), Syk has also been implicated in phagocytosis mediated by CD44 (25) and by the complement receptor (26). Syk has also been shown to participate in the regulation of endocytosis, an internalization pathway that is distinct from phagocytosis in ways that include the smaller size of the internalized particle and the requirement for clathrin. In leukocytes, Syk plays a role in the later stages of the endocytic pathway by mediating endosome-lysosome fusion following internalization of the BCR (27) and the Fc receptor associated γ-chain (28, 29) as well as the delivery of IgG-opsonized particles to the lysosome following FcγR-mediated internalization (22). In HeLa epithelial cells, Syk has been shown to regulate CD77-mediated Shiga toxin internalization by inducing the phosphorylation of clathrin (30). Down-regulation of Syk activity by small interfering RNA (siRNA) or expression of a dominant negative Syk mutant abrogated clathrin phosphorylation and Shiga toxin internalization (30). A role for airway epithelial Syk in the regulation of endocytosis is not known, although both minor (31, 32) and major (33) HRV serotypes have been shown to be internalized, at least in part, by clathrin-mediated endocytosis in HeLa cells.

In the current report, we investigated the role of Syk in HRV internalization and HRV-induced PI3K activation. We found Syk to be associated with clathrin and p85, the catalytic subunit of PI3K, and with the activation of PI3K upon HRV-16-ICAM-1 binding. Confocal microscopy revealed the redistribution of the normally cytosolic Syk initially to the plasma membrane upon HRV16-ICAM-1 binding and subsequently to punctate structures resembling endosomes at temperatures permissive for internalization. Overexpression of wild-type (WT) Syk enhanced HRV16 internalization, whereas expression of the dominant negative Syk^K396R decreased HRV16 cell entry. Together, our observations indicate that Syk mediates HRV-induced PI3K activation and clathrin-mediated internalization of the virus.

BEAS-2B, a human bronchial epithelial cell line, was provided by Dr. C. Harris of the National Cancer Institute (Bethesda, MD). Primary normal human bronchial epithelial cells (NHBE; formerly available as Cambrex NHBE cells) were purchased from Lonza. BEAS-2B and NHBE cells were cultured in Clonetics bronchial epithelial growth medium (BEGM; Cambrex BioScience) at 37°C and in a humidified environment containing 5% CO₂. BEGM is composed of Clonetics bronchial epithelial basal medium (BEBM; catalog no. cc-3171) supplemented with Clonetics BEGM SingleQuots (catalog no. cc-4175) containing 2 ml of bovine pituitary extract, 0.5 ml of insulin, 0.5 ml of hydrocortisone, 0.5 ml of gentamicin sulfate amphotericin (GA-1000), 0.5 ml of retinoic acid, 0.5 ml of transferrin, 0.5 ml of tri-iodothyronine, 0.5 ml of epinephrine, and 0.5 ml of recombinant human epidermal growth factor at proprietary concentrations.

The following Abs were purchased from the indicated sources: rabbit polyclonal anti-ezrin from Upstate Cell Signaling; mouse monoclonal anti-Syk (clone 4D10), rabbit polyclonal anti-Syk (clone C20), and mouse monoclonal anti-clathrin from Santa Cruz Biotechnology; rabbit polyclonal anti-phospho-Src (Y416), mouse monoclonal anti-Src, rabbit polyclonal anti-phospho-PI3K p85 (Y458), rabbit polyclonal anti-PI3K p85, rabbit polyclonal anti-phospho-Akt (T308), and rabbit polyclonal anti-Akt from Cell Signaling Technology. The mouse mAb R16.7 directed against the VP2 capsid protein of HRV-16 (34) was provided by Dr. W.-M. Lee (University of Wisconsin, Madison, WI). The HRP-labeled anti-mouse and anti-rabbit, the FITC-labeled goat anti-rabbit, the Texas Red-labeled goat anti-mouse, Cy3-labeled goat, and the donkey anti-mouse secondary Abs were from Jackson ImmunoResearch Laboratories. The isotype control Abs for the anti-Syk4D10 and anti-clathrin used in the immunoprecipitation experiments were mouse IgG2a-κ and mouse IgG1, respectively; both were purchased from Sigma-Aldrich. The rabbit IgG isotype control Ab used as a negative control in immunofluorescence experiments was purchased from Lab Vision. The mouse monoclonal anti-human-ICAM-1 domain D1 Ab for ICAM-1 receptor blocking experiments was obtained from Fitzgerald Industries.

BEAS-2B and NHBE cells were transfected with the Amaxa Nucleofector system according to the manufacturer’s instructions using 4 × 10⁶ cells and 4 μg of plasmid DNA or 2 × 10⁶ cells with 0.75 μg of Cy3-luciferase control SMARTpool siRNA reagent or Syk SMARTpool siRNA (Upstate). The cells were plated in normal culture medium following transfection and cultured at 37°C in 5% CO₂ for 36–48 h before being used for the experiments.

The plasmids expressing human Syk mutants were generated from the pcDNA3 plasmid containing amino-terminal, hemagglutinin-tagged, WT human Syk using the QuikChange site-directed mutagenesis kit (catalog no.200518; Stratagene). The Syk cDNAs were excised from pcDNA3 as a 1.553-kb BamHI fragment and subcloned into the BglII and BamHI sites of pEGFP-N2 (Clontech). All mutants were verified by sequencing and Western blot analysis to ensure the expression of a protein of the appropriate size and immunoreactivity before being used for experimentation.

BEAS-2B cells were grown to ∼90% subconfluence. Cells were deprived of growth factors overnight prior to virus inoculation. The HRV16 viral stocks used for experiments were generated by propagation in WI-38 cells and purified by centrifugation through sucrose to remove ribosomes and soluble factors as previously described (35). The purified viral preparations, stored in BEBM, contained a 50% tissue culture infective dose of ∼10^4.5 virus/ml assessed using a microtiter plate assay as previously described (36) and were diluted 1/10 in BEBM for inoculation. Cells were incubated at 4°C for 1 h and then washed once with prewarmed PBS. Prewarmed cell culture medium was added to the cells and samples were returned to incubation at 34°C and 37°C for 0 to 30 min and harvested for immunoprecipitation and immunoblotting experiments or confocal microscopy as described below. Nonstimulated cells were treated in the same manner in BEBM without HRV16.

Cells were washed with PBS and harvested with lysis buffer (50 mM Tris · HCl (pH 8.0), 120 mM NaCl, and 1% Triton X-100 supplemented with 1 tablet per 10 ml of Complete-Mini Protease Inhibitor Cocktail (Roche)). The protein assay was performed using the Bradford method. For Western analysis of whole cell lysates, 25–30 μg of protein was loaded per lane and separated by SDS-PAGE using a 7.5–10% polyacrylamide gel as previously described (4). For immunoprecipitation, the cell lysates were precleared with 20 μl of protein A/G-Sepharose (Santa Cruz Biotechnology) for 60 min then incubated with 1 μg of anti-Syk4D10 Ab or the isotype control IgG2a-κ for 60 min before the addition of 30 μl of protein A/G-Sepharose for another 60 min in an end-over-end rotor. The reaction was washed three times with radioimmunoprecipitation assay buffer (120 mM NaCl, 50 mM Tris · HCl (pH 8.0), 1% Nonidet P-40, 0.5% deoxycholate, and 0.1% SDS). The entire procedure was done at 4°C. After the final wash, the pellet was resuspended in 40 μl of 2× Laemmli sample buffer and boiled for 10 min. The samples were separated by SDS-PAGE and Western blot analysis was performed as described (4). Densitometry was analyzed using GelEval 1.07 (FrogDance Software for Mac OS X Version) and expressed as mean ± SEM. ANOVA was used for factorial analysis. Post hoc tests were conducted when ANOVA reached p < 0.05 using Tukey’s adjustment method. The statistical analysis program Prism 4.0a was used for analysis (GraphPad Software).

BEAS-2B and NHBE cells were cultured to subconfluence on 18-mm glass coverslips (Thomas Scientific) in 12-well plates and stimulated with 1/10 dilution of the purified HRV16 described above. For the ICAM-1 receptor blocking experiments, subconfluent cells were incubated with anti-ICAM-1 domain D1 Ab diluted 15 μg/ml in BEGM for 1 h at 4°C before virus inoculation. To synchronize the binding of the virus to ICAM-1, the initial incubation period (1 h) was performed at 4°C. Excess unbound virus was washed twice with prewarmed PBS. The cells were then shifted to a 37°C incubator with prewarmed normal culture medium for the periods indicated. The cells were fixed with 4% paraformaldehyde for 15 min, quenched with 100 mM glycine for 15 min, permeabilized with 0.1% Triton X-100 for 15 min, blocked with 10% goat serum plus 1% BSA in PBS for 2 h, and then incubated with 1/100 mouse IgG2a-κ isotype control, 1/100 rabbit IgG isotype control, 1/100 mouse IgG1 isotype control, 1/100 mouse anti-Syk 4D10, 1/100 rabbit anti-ezrin, 1/300 mouse anti-clathrin, 1/100 rabbit anti-Syk C-20, or 1/1250 mouse anti-VP2 overnight at 4°C. The cells were then washed with PBS and counterstained with 1/300 FITC-, Texas Red-, or Cy3-labeled goat anti-rabbit or anti-mouse IgG (Jackson ImmunoResearch Laboratories) for 1 h. The coverslips were mounted using DakoCytomation fluorescent mounting medium onto a glass slide. Images were obtained using a laser-scanning Zeiss LSM510 confocal microscope (Plan-Apochromat ×63/1.4 oil differential interference contrast objective) and the LSM510 Image software. Z-stack images were obtained with 0.5–1.0 μM collimations through the cell monolayer. Single xy-planes (parallel to cell monolayer) and reconstructed xz/yz-planes (orthogonal to cell monolayer) were processed using the LSM 510 Image software and exported as TIFF images into Canvas X. Quantification of HRV16 staining (primary Ab, mouse anti-HRV16 VP2; secondary Ab, Cy-3-labeled anti-mouse IgG) was performed on images obtained using the same confocal microscope setting, comparing BEAS-2B cells transfected with WT-Syk-GFP and Syk^K396R-GFP that were stimulated with HRV16 and immunostained on the same day.

We had previously shown by immunoprecipitation and Western analysis that Syk associates with ezrin in airway epithelial cells at basal conditions and that this coassociation is enhanced following HRV stimulation (4). In the present study we evaluate the intracellular localization of Syk and ezrin in response to HRV inoculation. To minimize variations in HRV-ICAM-1 binding among the different conditions and to synchronize internalization, we incubated BEAS-2B cells with HRV16 for 60 min at 4°C to allow binding but not internalization to occur. Excess virus was washed off and internalization was allowed to proceed at 37°C. At defined times following incubation at 37°C, the cells were fixed, permeabilized, and dually stained for Syk (red) and ezrin (green) as described in Materials and Methods. Intracellular localization of the two proteins was then assessed by immunofluorescence and confocal microscopy. As shown in Fig. 1,A, both Syk and ezrin exhibit a primarily cytosolic distribution under basal conditions. Upon the binding of HRV to ICAM-1 and before internalization (<1 min at 37°C; Fig. 1,B) there is a recruitment of both proteins to the plasma membrane with evidence of Syk-ezrin colocalization as indicated by the yellow color in the overlay image (Fig. 1,B, far right panel). As internalization was permitted to proceed at 37°C, we observed redistribution of Syk and ezrin to punctate intracellular structures that resemble early endosomes at 15 min (Fig. 1,C) and perinuclear structures resembling late endosomes at 30 min (Fig. 1,D). Incubation of the cells at 4°C for 1 h in the absence of HRV (Fig. 1,E) did not affect the basal distribution of Syk and ezrin at basal conditions (Fig. 1,A), indicating that the changes in the intracellular localization of the two proteins at <1 min (Fig. 1,B) are a result of HRV-ICAM-1 binding rather than the lowering of the temperature to 4°C. As controls, we performed immunostaining using the isotype control Abs for the mouse monoclonal Syk and rabbit polyclonal ezrin Abs, mouse IgG2a-κ, and rabbit IgG, respectively, followed by counterstaining with appropriate Cy3-labeled anti-mouse IgG and FITC-labeled anti-rabbit IgG. As depicted in Fig. 1, F and G, respectively, staining was absent in both the red channel and the green channel, confirming the binding specificities of both Syk and ezrin Abs for the proteins of interest.

As an additional control, we pretreated BEAS-2B cells with an anti-ICAM-1 Ab that recognizes ICAM-1 at the D1 domain, a known binding site for the major group rhinoviruses such as HRV16 (37), to serve as a blocking Ab to HRV16-ICAM-1 interaction. As shown in Fig. 1 H, blocking the interaction of HRV16 to ICAM-1 resulted in no change in the cytosolic distribution of Syk at <1 and 15 min after incubation at 37°C (middle and far right panel) when compared with no HRV (far left panel). Taken together, this suggests that the intracellular trafficking of Syk and ezrin was, in fact, dependent on HRV binding to its receptor.

These observations reveal colocalization of Syk and ezrin with enhancement following HRV16 exposure and confirm our biochemical data showing enhanced Syk-ezrin coassociation in response to HRV16 (4). The confocal analyses revealed the recruitment of Syk and ezrin to the plasma membrane with subsequent trafficking to endosomal compartments during HRV16 internalization, suggesting a role for Syk in the endocytosis of HRV. Clathrin-mediated endocytosis has been found to be a major route of cell entry for both minor (31, 32) and major (33) group HRV serotypes. Therefore, we set out to determine whether Syk and ezrin associated with clathrin during HRV uptake by BEAS-2B cells.

We used immunoprecipitation and Western blot analysis to evaluate the coassociation of Syk and clathrin following HRV infection. As shown in Fig. 2 A, Western blot analysis of the Syk immunoprecipitates revealed the coassociation of the H chain of clathrin with Syk upon HRV16-ICAM-1 binding (<1 min) with enhancement upon shift to 37°C, a permissive temperature for endocytosis. Syk-clathrin association was sustained to 30 min. Densitometry of the clathrin band normalized to Syk, expressed as a mean fold change ± SE when compared with control (No HRV was defined as 1), is as follows: <1 min, 3.37 ± 2.18; 10 min, 1.73 ± 0.91; 20 min, 4.38 ± 3.43; and 30 min, 5.38 ± 3.14; n = 4. There is a trend to increased Syk-clathrin coassociation over time, but this did not reach statistical significance. The converse experiment, using an Ab to H chain clathrin as the immunoprecipitating Ab, also revealed enhanced Syk-clathrin association upon HRV binding and incubation at 37°C up to 30 min (data not shown).

We have previously established that HRV16 enhances the formation of a protein complex composed of ICAM-1, ezrin, and Syk (4) and postulated ezrin to be a linker protein that mediates the interaction of Syk to ICAM-1. Therefore, we assessed the association of ezrin with clathrin during HRV16 internalization using confocal microscopy following dual staining for ezrin (primary Ab, rabbit polyclonal anti-ezrin; secondary Ab, FITC anti-rabbit IgG) followed by an anti-clathrin H chain Ab (secondary Ab, Cy3 anti-mouse IgG). When compared with basal conditions in the absence of HRV (Fig. 2,C), we observed the recruitment of ezrin (green) and clathrin (red) to the plasma membrane upon HRV16-ICAM-1 binding (time, <1 min; Fig. 2,D) with evidence of significant colocalization as indicated by the yellow color in the overlay images (Fig. 2,D, far right panel). Incubation at 37°C to permit HRV internalization revealed the trafficking of both ezrin and clathrin to punctate structures resembling early (15 min) and late (30 min) endosomes-lysosomes with significant colocalization of ezrin and clathrin as indicated by the yellow color in the overlay images (Fig. 2, E and F, right panel) at both time points. Immunostaining with the isotype control IgG1 Ab did not reveal staining in either the red or the green channel (Fig. 2 B), confirming the specificity of the anti-clathrin H chain Ab.

Taken together, these observations indicate that Syk and ezrin are recruited to the plasma membrane in conjunction with clathrin upon the binding of HRV16 to ICAM-1. Furthermore, in response to permissive temperatures for HRV internalization, the intracellular redistribution of Syk and ezrin paralleled that of clathrin with evidence of coassociation biochemically and by confocal microscopy, suggesting a role for Syk in clathrin-mediated endocytosis of HRV.

Next, we evaluated the intracellular localization of HRV16 following incubation at temperatures that are permissive for internalization by using a mAb directed against the HRV16 capsid protein VP2. To facilitate the assessment of Syk colocalization with HRV16, we transfected BEAS-2B cells with GFP-tagged Syk. Twenty-four to 36 h posttransfection we synchronized HRV16 binding to ICAM-1 with an initial 1-h incubation at 4°C, washed off unbound HRV16, and then shifted the cells to 37°C to allow internalization to occur. The cells were then fixed at defined time points and immunostained with anti-VP2 Ab followed by a Cy-3-labeled anti-mouse secondary Ab. As shown in Fig. 3,A, there is no background staining and surface-bound extracellular HRV-16 is poorly detected (Fig. 3,B). However, after 15 and 30 min of incubation at 37°C, the internalized HRV16 is visualized (Fig. 3, C and D, left panels) with evidence of colocalization with Syk as indicated by the yellow color in the overlay images (Fig. 3, C and D, right panels). Orthogonal sections obtained from z-stack images acquired using 0.5–1.0 μm collimations are shown in both the xz and yz axes (Fig. 3, B–D) and show colocalization of Syk and HRV16 with trafficking to a more perinuclear localization at 30 min. The lack of HRV16 staining at the cell surface at <1 min (Fig. 3 B) may be a result of the poor binding affinity of the anti-VP2 Ab to the extracellular virus. In our experimental protocol, excess unbound virus is washed off before the temperature shift to 37°C to permit internalization and viral replication is not possible at the 15- and 30-min time points. Therefore, HRV16 staining at 15 and 30 min cannot be a result of increased viral particles. Uncoating of the viral capsid proteins occurs following the internalization of HRV as the virus is delivered to progressively more acidic endoluminal pH environments along the endocytic pathway (38, 39, 40). Therefore, it is possible that modifications of the capsid proteins following HRV16 internalization alter the binding affinity of the anti-VP2 Ab to allow visualization of the virus at 15 and 30 min following internalization.

To determine the role of Syk kinase activity in HRV internalization, we compared BEAS-2B cells transfected with the kinase-inactive mutant Syk^K396R-GFP (Fig. 4,C) to those transfected with WT-Syk-GFP (Fig. 4,B). As control, BEAS-2B cells were also transfected with pEGFP alone (Fig. 4,A). HRV-16 binding to ICAM-1 was synchronized by incubation for 1 h at 4°C. Unbound virus was washed off and internalization was allowed to proceed for 30 min at 37°C. The cells were then fixed and immunostained with anti-VP2 Ab followed by Cy-3-labeled anti-mouse Ab. Confocal microscopy shows that the overexpression of WT Syk (Fig. 4,B) enhanced the staining for HRV16 when compared with cells expressing GFP alone (Fig. 4,A). Expression of the kinase-inactive Syk^K396R mutant (Fig. 4,C) significantly decreased HRV16 staining. Quantitation of the high intensity pixels in the red channel (i.e., HRV16 staining) of the confocal images, taken with the same settings, reveals substantially fewer pixels in the high intensity ranges in the Syk-K396R-expressing cells when compared with those expressing WT-Syk (Fig. 4 D). These observations suggest that Syk kinase activity plays a role in HRV internalization.

The PI3K/Akt signaling pathway has recently been identified as a downstream effector of HRV-induced cell signaling (14), mediating HRV internalization and IL-8 expression. We had previously shown that Syk mediates HRV-induced IL-8 via a p38-dependent pathway (4). Syk is a known upstream regulator of PI3K in leukocytes (15, 16) and interacts with the p85 regulatory subunit of PI3K directly (41) and indirectly via adaptor proteins such as CD19 (42) and Gab2 (43). Thus, we sought to examine the role of Syk as a potential upstream regulator of PI3K in HRV-induced signal transduction in airway epithelial cells.

In immunoprecipitation studies using anti-Syk-4D10 as the precipitating Ab, we demonstrated coprecipitation of Syk and p85 under basal conditions with no significant change following exposure to HRV16 (Fig. 5,A, second and third panels from the top). However, immunoblotting with an anti-phospho-specific p85 Ab revealed that HRV induced phosphorylation of the p85 subunit upon ICAM-1 binding and it appeared to peak by 20 min (Fig. 5,A, top panel). To determine whether p85 phosphorylation was accompanied by the induction of p85 activity, we performed Western blot analysis of the whole cell lysates from the same experiment and assessed phosphorylation of the p85 substrate Akt. As shown in Fig. 5 B, HRV16 induced Akt-phosphorylation within the same time frame as p85 phosphorylation. Activation occurred upon HRV-ICAM-1 binding (incubation at 4°C for 1 h and at 37°C for <1 min) with a peak at 10–30 min of incubation at 37°C. Densitometry values of the phospho-Akt band when normalized to total Akt, expressed as fold change ± SE normalized to no HRV stimulation (defined as 1) are as follows: <1 min, 20.43 ± 10.43; 10 min, 66.36 ± 33.27; 20 min, 67.37 ± 38.23; and 30 min, 61.2 ± 35.89 (p < 0.05 for all; n = 4).

To determine the role of Syk in HRV-induced PI3K activation, we down-regulated Syk expression using siRNA (Syk SMARTpool siRNA; Upstate Cell Signaling) and subsequently exposed the cells to HRV16 as described above. We used sham and control siRNA (control SMARTpool siRNA; Upstate Cell Signaling)-transfected BEAS-2B cells for control. We have previously shown that expression of ICAM-1 and ezrin are not affected under these transfection conditions (4). As shown in Fig. 6,A, transfection with control siRNA had no effect on Syk expression when compared with sham-transfected BEAS-2B cells, unlike those transfected with Syk-siRNA (top panel). Densitometry of the Syk bands revealed a 75% knockdown of Syk expression in Syk-siRNA-transfected cells when compared with sham-transfected cells and a 79% knockdown compared with control siRNA-transfected cells (n = 3; p < 0.001). Knockdown of Syk expression by siRNA significantly abrogated HRV-induced Akt phosphorylation at all time points assessed when compared with sham and control siRNA-transfected cells (Fig. 6,A, second panel from the top). Densitometry of the phospho-Akt bands (Fig. 6,B) normalized to total Akt (Fig. 6 A, bottom panel) and expressed as mean ± SE (bottom panel) are as follows; sham-transfected: 0.58 ± 0.11 (<1 min), 1.58 ± 0.26 (10 min), 1.42 ± 0.06 (20 min), and 1.37 ± 0.02 (30 min); control siRNA: 0.57 ± 0.08 (<1 min), 1.21 ± 0.13 (10 min), 1.05 ± 0.07 (20 min), and 0.83 ± 0.14 (30 min); Syk siRNA: 0.19 ± 0.01 (<1 min), 0.49 ± 0.21 (10 min), 0.35 ± 0.03 (20 min), and 0.35 ± 0.05 (30 min). Statistical analysis using ANOVA, followed by Tukey’s post hoc analysis, revealed that Akt phosphorylation in response to HRV16 in the sham and control siRNA-transfected cells was not significantly different from one another (p > 0.05). However, Akt phosphorylation in Syk siRNA-transfected cells was significantly different from that in sham (p < 0.001) and control siRNA-transfected (p < 0.01) cells at all time points, suggesting that Akt phosphorylation following HRV stimulation is Syk dependent. Together, these observations reveal basal association of Syk with p85 in BEAS-2B cells and Syk-dependent PI3K activation in response to HRV infection.

To further validate our observations, we repeated these experiments in primary human airway epithelial cells using the Clonetics normal human bronchial epithelial (NHBE) cells (Lonza). As shown in Fig. 6,C, transfection with Syk siRNA decreased Syk expression in NHBE cells when compared with sham and control siRNA-transfected cells (top panel) along with decreased phosphorylation of Akt (second and third panels from the top). We had previously shown that HRV16 induced a Syk-dependent activation of p38 (4) within the same time frame as our current observations of PI3K activation and Akt phosphorylation. Therefore, we also evaluated the phosphorylation of p38 at 20 min following HRV16 inoculation in NHBE cells. As shown in Fig. 6 D, knockdown of Syk by siRNA (top panel) also abrogated phosphorylation of p38 (middle panel) when compared with sham and control siRNA-transfected cells, confirming our earlier observations in BEAS-2B cells (4). Therefore, by using human airway epithelial cell lines and primary human airway epithelial cells we have observed that HRV16 induces activation of PI3K/Akt and p38 within 30 min of HRV inoculation. Moreover, these two HRV-induced signaling pathways are Syk dependent.

HRV-induced PI3K activation has recently been reported to be dependent on Src (44), an upstream regulator of Syk in the immunoreceptor signaling pathway in leukocytes (22, 45, 46, 47, 48). The relationship between Syk and Src in HRV-induced signaling in airway epithelial cells is not known and was investigated next. As shown in Fig. 7,A (control siRNA; first three lanes from the left), exposure of overnight growth factor-deprived BEAS-2B cells to HRV16 induced the phosphorylation of Src within 1 min when compared with nonexposed control cells (lanes labeled C). Following transfection with Syk siRNA (Fig. 7,A, last three lanes from the left), Syk expression was down-regulated (top panel) but HRV16-induced Src phosphorylation (middle panel) was not affected; HRV induced phosphorylation of Src within 1 min and it was sustained at 15 min. To evaluate the role of Syk kinase activity on HRV-induced Src phosphorylation, we overexpressed the kinase-inactive Syk^K396R mutant in BEAS-2B cells and compared the effects of HRV inoculation with cells transfected with WT-Syk. The Syk mutant constructs possessed a GFP tag to allow differentiation from endogenous Syk. As shown in Fig. 7 B, the response to HRV was similar in the cells expressing WT-Syk and Syk^K396R with evidence of enhanced Src phosphorylation within 1 min that was sustained at 15 min (middle panel). Taken together, these observations indicate that Syk was dispensable for HRV-induced Src activation and are congruent with other ITAM-mediated signaling pathways in which Syk was found to be downstream of Src (45, 49).

To further confirm our observations in BEAS-2B cells, we also assessed the intracellular localization of Syk and ezrin in NHBE cells in response to HRV inoculation by using confocal microscopy. NHBE cells grown to subconfluence on glass coverslips were incubated with HRV16 for 60 min at 4°C to synchronize ICAM-1 binding. Excess virus was washed off and internalization was allowed to proceed at 37°C for defined time periods. NHBE cells then were fixed, permeabilized, and immunostained with primary Abs to Syk and ezrin followed by the appropriate fluorophore-labeled secondary Abs as described in Materials and Methods. As shown in Fig. 8,A, Syk and ezrin display a cytosolic distribution under basal conditions but are recruited to the plasma membrane upon the binding of HRV to ICAM-1 (Fig. 8,B) with evidence of colocalization as indicated by the yellow color in the overlay image (Fig. 8,B, far right panel). Incubation at 37°C to permit internalization resulted in redistribution of both Syk and ezrin to punctate submembranous regions at 15 min (Fig. 8,C) with progression toward a more central perinuclear localization at 30 min (Fig. 8,D). Syk-ezrin colocalization, indicated by the yellow color in the overlay images (Fig. 8, B–D), is present up to 30 min. In Fig. 8,E we show again that incubation at 4°C for 60 min does not alter the basal distribution of Syk and ezrin. These observations of Syk/ezrin distribution and colocalization in NHBE cells in response to HRV16-ICAM-1 binding and HRV16 internalization mirror the observations in BEAS-2B cells (Fig. 1).

Taken together, our observations in the human BEAS-2B airway epithelial cell line and in primary NHBE cells indicate that Syk mediates the activation of PI3K independently of viral replication as well as the clathrin-mediated endocytosis of HRV16 following the engagement of ICAM-1.

In this report we used NHBE cells and the BEAS-2B human airway epithelial cell line and identified Syk to be an early signaling molecule that mediates PI3K/Akt activation and internalization of HRV via clathrin-mediated endocytosis. Our biochemical and confocal data indicate that the binding of HRV16 to ICAM-1 leads to the recruitment of Syk, ezrin, and clathrin to the plasma membrane with subsequent trafficking of this protein complex to endosomes during HRV internalization. These observations, in conjunction with our previous data demonstrating replication-independent HRV-induced p38 phosphorylation and IL-8 expression (4), suggest that Syk may have two roles in HRV-induced cell signaling: 1) one that is independent of viral replication and signals in response to ICAM-1 engagement by HRV binding, leading to activation of the p38 and PI3K signaling pathways and IL-8 expression; and 2) a second role that regulates viral cell entry and, in this way, regulates viral replication and replication-dependent signaling events. A model of the intracellular signaling pathways that have been identified in the airway epithelium following HRV inoculation and the role of Syk in these pathways is shown in Fig. 9.

Syk is known to play an important role in the receptor-mediated internalization of extracellular materials in leukocytes. Although the role of Syk has been particularly well characterized in FcγR-mediated phagocytosis (18, 50, 51, 52), recent studies suggest that Syk has a more generalized role in the receptor-mediated internalization. Syk has been shown to regulate complement-mediated phagocytosis (26) and CD44-mediated phagocytosis (25) as well as endocytosis. In HeLa epithelial cells Syk has been shown to mediate Shiga toxin internalization via Gb3 (also known as CD77), a globotriaosyl ceramide that serves as a receptor for Shiga toxin (53), by inducing the phosphorylation of clathrin (30). Down-regulation of Syk activity abrogated not only clathrin-mediated internalization of Shiga toxin but also intracellular transport from the endosome to the Golgi apparatus (30).

Our observations in BEAS-2B and NHBE cells indicate a role for Syk in the early stages of internalization with the recruitment of Syk upon HRV-ICAM-1 binding (Figs. 1,B and 8,B). Importantly, our studies in cells overexpressing the kinase-inactive Syk^K396R mutant (Fig. 4) indicate that the kinase activity of Syk is important for mediating HRV internalization. Curiously, in the HRV16 immunostaining experiments we failed to detect surface-bound HRV16 (Fig. 3,B) although we were clearly able to detect the presence of the virus at 15 and 30 min following internalization (Figs. 3, C and D, and 4,B). As our experimental protocol calls for the removal of unbound virus after the initial 1-h incubation step at 4°C and as viral replication is not possible in this short period of time, we speculate that changes in the viral capsid proteins, which occur as the virus is delivered to progressively more acidic environments during endocytosis (38, 39, 40), may unmask binding sites for the anti-VP2 Ab that could explain the detection of HRV16 after incubation for 15 and 30 min at 37°C. Indeed, our observations of persistent Syk-HRV colocalization at 15 and 30 min following the initiation of internalization at 37°C (Fig. 3, C and D) suggest that Syk may play a role in endosomal traffic to the lysosome, where the lower intracompartmental pH has been shown to facilitate capsid uncoating and thus viral replication (38, 54, 55). Alternatively, there may be inadequate viral clustering at 4°C, but upon warming the ICAM-1-virus complexes may cluster during internalization and increase the signal strength to a detectable level.

We observed increased retention of the kinase inactive Syk^K396R mutant in the nucleus of BEAS-2B cells when compared with WT-Syk (Fig. 4). Studies in mammary epithelial cells (56) and B lymphocytes (57) have identified a sequence at the junction of the kinase domain and the linker B region of Syk to be responsible for nuclear targeting. Although an intact kinase domain appears to be required for the export of Syk from the nucleus to the cytosol (57), the kinase domain is apparently dispensable for the translocation of cytosolic Syk to the plasma membrane; in the DT40 B lymphocytic cell line Syk^K396R is recruited to the plasma membrane and colocalizes with the BCR upon receptor engagement (58). This concurs with our observations in BEAS-2B cells using confocal microscopy and coprecipitation studies (data not shown). Thus, we conclude that the Syk-mediated signaling is responsible for regulating HRV internalization.

The specific mechanisms by which Syk mediates endosomal traffic have yet to be elucidated but may be related to Syk-mediated PI3K activation. In studies with 16HBE14o- human bronchial epithelial cells and the major group rhinovirus HRV39, Newcomb et al. have shown the activation of PI3K within 10 min of inoculation and the colocalization of HRV39 with Akt within 15 min of internalization (14). The inhibition of PI3K activity with LY294002 significantly decreased HRV39 internalization when assessed at 30 min despite the equivalent binding of the virus at the membrane surface at time 0 (14). Furthermore, in studies with the minor group rhinovirus HRV2, which also undergoes cell entry via clathrin-mediated endocytosis, the inhibition of PI3K with wortmannin significantly delayed transport of the virus from the early endosomes to late endosomes (59). Our previous finding of Syk activation upon HRV-ICAM-1 binding (4) and our recent observation of the constitutive association of Syk with the p85 catalytic subunit of PI3K (Fig. 5) and Syk-dependent PI3K activation (Fig. 6) are highly supportive of the hypothesis that Syk mediates HRV internalization and endosomal trafficking via PI3K.

The regulation of endosomal traffic by PI3K appears to be a signaling pathway conserved through evolution from yeast (60) to mammalian cells, where the endosome-to-lysosome transport of ligand-cell surface receptor complexes plays an important role in the delivery of essential nutrients and substrates as well as the regulation of receptor recycling and receptor-mediated signaling (61, 62). The regulation of endosome-lysosome transport by PI3K in leukocytes plays an important role in the regulation of innate and adaptive immunity. Complementary reports of Syk and PI3K activities in the regulation of MHC class II Ag internalization and presentation are highly suggestive of a link between these two signaling molecules, although direct evidence demonstrating Syk-mediated PI3K activation in endosomal traffic remains to be published. Studies in B cell lymphoma cell lines have revealed that the inhibition of PI3K impaired several steps in the endosomal traffic of BCR complexes, including the delivery of the complexes into late endosomes (63), the maturation of the late endosomes into the MHC class II-enriched compartments (64), and the presentation of MHC class II Ag (65). Syk has been implicated as playing a role in the same trafficking steps. In the B lymphoma IIA1.6 cell line, impaired Syk recruitment and activation due to mutation of the Fc receptor-associated γ-chain (which contains the binding site for Syk) or expression of a kinase-inactive Syk resulted in impaired γ-chain-mediated transport to the lysosome and presentation of MHC class II-restricted Ag (28). Similar observations have been made in dendritic cells where Syk deficiency in dendritic cells derived from Syk knockout mice (29) or dendritic cells treated with pharmacological inhibitors (66) exhibited impaired FcγR-mediated internalization and presentation of MHC class II-restricted Ags. Syk-mediated endosomal traffic regulates other aspects of immunoreceptor cell signaling as well; in DT40 B lymphocytes Syk plays a critical role in the fusion of BCR-containing endosomes with lysosomes and in this way regulate BCR-mediated apoptosis (27). Taken together, it is reasonable to conclude that Syk mediates endosome-lysosome traffic by regulating PI3K activation.

Syk can activate PI3K by two different mechanisms: 1) indirectly via adaptor proteins such as the B cell adaptor for PI3K (BCAP) (67), CD19 (42), and Grb2-associated binders 1 (Gab1) (68, 69) and 2 (Gab2) (43), which, following tyrosine phosphorylation, serve as binding sites for the p85 subunit of PI3K; and 2) directly by the binding of the carboxyl-terminal Src homology 2 (SH-2) domain of p85 to phosphotyrosine 317 of Syk (41). In BEAS-2B cells we observes the basal association of Syk with p85 and the phosphorylation of both p85 and Akt upon HRV binding (Fig. 5). These findings are highly suggestive of a direct role for Syk in PI3K activation. Current studies are underway in our laboratory to investigate the role of Syk-Y317 in p85 association and activation as well as the role of the Syk-SH2 domains in Syk recruitment to ICAM-1/ezrin following HRV infection.

Collectively, our observations have identified a role for Syk in the early signaling pathways that regulate replication-independent airway epithelial cell activation and IL-8 expression as well as HRV internalization, suggesting an important role for Syk in regulating the replication-dependent signaling events. These results should prompt further studies to evaluate the effect of Syk inhibition in rhinovirus infections and to decrease the detrimental effects of the dysregulated airway inflammatory response that ensues in susceptible populations such as patients with underlying asthma.

The authors have no financial conflict of interest.