VP1 of Foot-and-Mouth Disease Virus Induces Apoptosis via the Akt Signaling Pathway*

Foot-and-mouth disease virus (FMDV) binds to cellular integrins through an RGD motif in its capsid protein, VP1. It is unclear, however, what kind of cellular event(s) are triggered after the binding of VP1 to the cells. In this study, we show that aqueous soluble recombinant DNA-derived VP1 (rVP1) of FMDV induced apoptosis of BHK-21 cells after binding to integrins. In addition, treatment of BHK-21 cells with rVP1 resulted in deactivation of Akt and enhancement of several pro-apoptotic responses such as dephosphorylation of glycogen synthase kinase-3 (cid:1) and cleavage of procaspase-3, -7, and -9. Additional studies revealed that the rVP1 treatment caused apoptosis of cancer cells, including MCF-7 (a breast carcinoma cell line with a functional deletion of the caspase-3 gene) and PC-3 (a sphingosine 1-phosphate receptor subtype 3-deficient androgen-in-dependent prostate cancer cell line). These results suggest that rVP1 of FMDV may be used selectively as a potent apoptotic agent for human cancer by modulating the Akt signaling pathway and that its effect is not primarily dependent on either activation of procaspase-3 or deactivation of sphingosine 1-phosphate receptor subtype 3.

Foot-and-mouth disease virus (FMDV) 1 is the etiological agent of foot-and-mouth disease, a deadly epidemic that affects many economically important domestic livestock such as cattle, pigs, goats, and sheep (1). There are seven serotypes of FMDV that all belong to the Aphthovirus genus of the family Picornaviridae (2).
The capsid of FMDV is made up of 60 copies each of four proteins, VP1, VP2, VP3, and VP4 (2). Although, in some cases, FMDV may use heparin sulfate as an alternative receptor for internalization (3), it usually infects cells by attaching to integrin receptors through a long conformationally flexible loop (G-H loop) of VP1 (4,5). The sequence of this loop contains a conserved RGD tripeptide motif, which is characteristic of the ligands that bind to integrin receptors (3, 6 -9).
Regulation of apoptosis, or programmed cell death, allows the organism to replace aged cells, to control the cell number and tissue size, and to protect itself from rogue cells that may lead to lethality (13). Apoptosis requires integration and finetuning of multiple proteins that are either regulators or executors of the survival and death processes. Cancer, autoimmune diseases, immunodeficiency disease, reperfusion injury, and neurodegenerative disorders are characterized by disregulation of apoptosis. Integrin signaling can activate the Akt pathway (14), an anti-apoptotic mechanism utilized by many types of cells (15). It has been shown that Akt mediates cell survival in response to stimulation by growth factors (16,17). Akt is a serine/threonine kinase that binds phosphorylated lipids at the membrane in response to activation of phosphatidylinositol 3-kinase (PI3K) (18). Activation of PI3K through G proteincoupled receptors such as sphingosine 1-phosphate (S1P) receptors or use of PI3K inhibitors (e.g. LY294002) may thus modulate the activity of Akt (19). Phosphorylated Akt can activate anti-apoptotic or inhibit pro-apoptotic processes in the cell by phosphorylating Bad, Forkhead transcription factors, glycogen synthase kinase-3␤ (GSK-3␤), and caspase-9 (20 -25). On the other hand, deactivation of Akt has been shown to activate pro-apoptotic responses (26). Both GSK-3␤ and caspase-9 are pro-apoptotic factors (27,28). Cleavage of caspase-9 results in induction of a caspase cascade, including the processing of procaspase-3 and -7 (29). The initiation of the caspase cascade eventually leads to apoptosis (28).
Although integrins have been shown to mediate FMDV infection (6,11,12), whether the binding of FMDV and VP1 to integrins results in activation, deactivation, or any cellular response is unclear. Full-length and truncated forms of recombinant DNAderived VP1 (rVP1) of FMDV have been generated previously by several investigators (30). To solve the problem of poor water solubility, however, rVP1 has been routinely used together with denaturing agents such as urea (31). The presence of denaturing agents has made it quite difficult to evaluate the biological effects of rVP1. Recently, we have not only expressed and isolated two aqueous soluble forms of rVP1, but also showed that, like native VP1, they can elicit protective immunity against FMDV (31). In this study, we further investigated the binding of purified rVP1 to integrins and examined their cellular effects, mechanism of action, and potential applications.
Expression and Purification of rVP1-The mutant VP1 gene in the expression vector pET24a(ϩ) was expressed in Escherichia coli and purified according to our procedure described previously (31). Preliminary experiments showed that the apoptotic effect of the monomeric rVP1 mutant (rVP1(C187S)) was of the same magnitude as that of rVP1 without mutation at position 187. This monomeric rVP1 mutant was thus used in all experiments and is hereafter referred to as rVP1 in this study. After being processed according to our purification method, wildtype and mutant rVP1 become water-soluble (up to 5 mg/ml). Moreover, the lyophilized powder can be easily reconstituted in water without any precipitation. Because our purification process is quite simple (31), it can be easily used for large-scale production of rVP1.
Cell Lines and Treatments-BHK-21 cells are known to be permissive for FMDV to bind and replicate through the RGD motif (34). In this study, BHK-21 cells and several human cancer cell lines such as MCF-7, 22Rv1, and PC-3 were maintained at 37°C in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum, 2 mM L-glutamine, 100 units/ml penicillin, and 100 g/ml streptomycin. Cells were plated at 2 ϫ 10 5 cells/well in a 12-well plate (400 l/well) 1 day before the experiments. The cells were washed twice with phosphatebuffered saline (PBS) and treated with rVP1 in Dulbecco's modified Eagle's medium without fetal calf serum. Some of the cells were then lysed with 0.2 ml of boiling protein loading buffer (Invitrogen) at the indicated time points, and 20-l samples were analyzed for Akt and GSK-3␤ phosphorylation by Western blotting. The remaining cells were incubated overnight at 37°C for apoptosis experiments.
DNA Fragmentation Assay-In brief, BHK-21 cells were cultured in monolayer cultures overnight and treated with various amounts of rVP1. The treated cells were washed with ice-cold PBS and transferred to a centrifuge tube. The cells were recovered by centrifugation at 1500 ϫ g for 10 min at 4°C and resuspended in 100 l of solution containing 10 mM Tris-Cl and 1 mM EDTA (pH 8.0). One ml of extraction buffer consisting of 0.1 M EDTA, 0.5% (w/v) SDS, 20 g/ml pancre- atic RNase, and 10 mM Tris-Cl (pH 8.0) was added, followed by incubation at 37°C for 1 h. Proteinase K (Sigma) was added to a final concentration of 100 g/ml, and the suspension of lysed cells was kept at 50°C for 3 h with periodically swirling. An equal volume of phenol was added to the solution. The mixture was kept at room temperature for 10 min and then subjected to centrifugation at 5000 ϫ g for 15 min to separate into two phases. After extraction three times with phenol, the aqueous phases were pooled, transferred to a centrifuge tube, and mixed with 0.2 volume of 10 M ammonium acetate and 2 volumes of ethanol. The genomic DNA was collected by centrifugation at 5000 ϫ g for 5 min, washed with 70% ethanol, and separated by electrophoresis on a 0.8% agarose gel.
Assay for Cell Survival and Apoptosis-BHK-21 cells (5 ϫ 10 5 cells/ ml) were transfected with plasmid pEGFP-C3 (0.5 g/well), which encodes green fluorescent protein (GFP), in Effectene for 24 -48 h. After incubation with rVP1 at the indicated time points, the surviving and dead cells were distinguished by examination under a fluorescence microscope. To visualize DNA condensation in nuclei, the control and treated cells were stained with DAPI (0.5 g/ml in PBS) according to the manufacturer's instructions and then examined by light microscopy. The level of apoptosis was determined as the ratio of cells with morphologically changed nuclei to the total number of cells present.
Cell Transfection and Plasmid Expression-BHK-21 cells were transfected with pIBSY1 or pIBSY1-VP1 (32) using Lipofectamine Plus TM according to the protocol specified by the manufacturer. The cell extracts were assayed by Western blotting 48 h after the start of transfection.
Indirect Immunofluorescence-The binding of rVP1 and binary ethyleneimine (BEI)-inactivated FMDV to the cells was detected following an established procedure (35), with minor modifications. Briefly, BHK-21 cells were grown on a glass coverslip overnight and then treated with rVP1 (1 M) or BEI-inactivated FMDV (125 g/ml) at 4°C for 2 h. Cells on the coverslip were washed with PBS once and immersed in 100% methanol at Ϫ20°C for 10 min. Fixed cells were rinsed three times with PBS, blocked in 1% bovine serum albumin in PBS at room temperature for 40 min, washed with PBS, and then incubated with primary antibodies (rabbit anti-FMDV neutralizing antibodies) at 1:500 in 1% bovine serum albumin in PBS for 16 h at 4°C. After the primary antibody reaction, cells were washed twice with PBS, incubated with secondary antibodies (fluorescein isothiocyanate-labeled pig anti-rabbit IgG) at 1:1000 in 1% bovine serum albumin in PBS for 60 min at 30°C under low light conditions, and washed twice with PBS. Coverslips were mounted on slides using Gel/Mount (Biomeda, Foster City, CA). The slides were air-dried for 1 h and stored at room temperature in the dark. Cells were observed using an Olympus IX70 fluorescence microscope.

RESULTS
rVP1 Induces Apoptosis-VP1 of FMDV contains an RGD motif, which is characteristic of the ligands that bind to cellular integrins (4,8,36). Although we previously produced aqueous soluble rVP1, it was found to form monomers or dimers depending upon the redox conditions of the experiments (31). To avoid any uncertainty in this study, we thus generated a monomeric rVP1 mutant by changing Cys 187 to Ser by site-directed mutagenesis and investigated the cellular effect of this rVP1. Treating BHK-21 cells with 1 M rVP1 for 24 h resulted in DNA fragmentation (Fig. 1A), one of characteristic features of apoptosis (28). To evaluate whether DNA fragmentation was accompanied by cell death and DNA condensation, we transfected BHK-21 cells with plasmid pEGFP-C3 and observed them under a fluorescence microscope. The number of GFP-transfected cells (green color under fluorescence) was reduced when the cells were treated with 1 M rVP1 for 2 h. The number of GFP-transfected cells was severely reduced when the treatment lasted for 24 h (Fig. 1B, left panels). Staining of cell nuclei with DAPI confirmed that rVP1-treated cells had nuclear condensation indicative of apoptosis (Fig. 1B, right panels). The apoptotic effect of rVP1 was not due to site-directed mutagenesis at position 187, as wild-type rVP1 also showed a similar effect (Fig. 1C, open bars). However, both FMDV and FMDV inactivated by BEI, an agent known to inactivate FMDV without damaging viral proteins (37), did not show any apoptotic effect (Fig. 1D).
rVP1 Induces Apoptosis by Binding to Integrin-To compare the apoptotic effect of rVP1 with that of other RGD motifcontaining molecules, BHK-21 cells were treated with rVP1, cyclic RGD, or peptide P29 (containing residues 131-159 with the RGD motif at positions 145-147) (32). Treating BHK-21 cells with 1 M rVP1 resulted in apoptosis of almost all BHK-21 cells (Fig. 1C, closed bars), whereas both peptide P29 and cyclic RGD at the same concentration had little effect (data not shown). The 50% effective concentration of rVP1 in causing apoptosis was ϳ0.3 M, which was ϳ100and 10,000-fold more potent than those of peptide P29 (30 M) and cyclic RGD (3

FIG. 2. Binding of BEI-inactivated FMDV and rVP1 to BHK-21 cells and the effect of intracellularly expressed rVP1.
A, BHK-21 cells were pretreated with or without FMDV, rVP1, and anti-integrin ␣ 5 ␤ 1 antibodies at 4°C for 2 h as indicated. The cells were then treated with rabbit anti-FMDV neutralizing antibodies and subsequently subjected to indirect immunofluorescence staining using fluorescein isothiocyanatelabeled pig anti-rabbit IgG secondary antibodies (magnification ϫ400). B, the cells were transfected with 0.3 g of vector alone (pIBSY1) or vector encoding the VP1 gene (pIBSY1-VP1). After 3 days of culture, the cells were harvested, and the cell lysate was subjected to Western blot analysis. C, the viability of the transfected cells with or without 0.5 M extracellular rVP1 treatment was determined. The viability of the transfected cells was examined by counting the percentage of morphologically normal cells. Data represent the means of two experiments. mM), respectively. Pretreatment of the cells with antibodies before addition of rVP1 revealed that the apoptotic effect of rVP1 was blocked specifically not only by rabbit anti-FMDV neutralizing antibodies, but also by anti-integrin ␣ 5 ␤ 1 antibodies (Fig. 1D).
Because fibronectin is a natural ligand for the integrin ␣ 5 ␤ 1 receptor (9, 36), we investigated whether addition of fibronectin could reverse the apoptotic effect caused by rVP1. Our results show that fibronectin at 0.8 M abolished the apoptotic effect of 1 M rVP1 on BHK-21 cells (Fig. 1D). To further determine whether the apoptotic effect of rVP1 could be obtained only after its interaction with integrin receptors, we constructed a mutant VP1 protein containing GGD instead of RGD and found that this mutant (rVP1(R145G)) caused little, if any, apoptosis (Fig. 1D). In addition, we were able to show that rVP1, like BEI-inactivated FMDV, bound to BHK-21 cells and that the binding was blocked by anti-integrin ␣ 5 ␤ 1 antibody ( Fig. 2A). Moreover, we transfected BHK-21 cells with plasmid pIBSY1-VP1, an expression vector containing the VP1 gene, to express rVP1 intracellularly. Although rVP1 was significantly expressed in the transfected cells, little apoptosis was observed even after 3 days of culture (Fig. 2, B and C). On the other hand, addition of rVP1 to the culture medium resulted in the death of the transfected cells (Fig. 2C). These results suggest that induction of apoptosis by rVP1 is most likely due to its binding to integrin receptors.
rVP1 Treatment Deactivates Akt-Binding of growth factors such as PDGF to cells activates PI3K/Akt pathways and causes cell proliferation (17). On the other hand, deactivation of Akt causes apoptosis (28). To determine whether rVP1 has any effect on the Akt pathway after binding to integrin, we treated BHK-21 cells with PDGF and rVP1. As expected, PDGF activated Akt phosphorylation in a concentration-dependent manner, and this effect was inhibited by the PI3K inhibitor LY294002 (Fig. 3, A and B). When BHK-21 cells were treated with rVP1 for 30 min, Akt phosphorylation was inhibited, in contrast to the activation induced by PDGF (Fig. 3C). The effect was more pronounced when the incubation time was increased to 60 min (Fig. 3C). This inhibitory effect of rVP1 on Akt was reversed not only by increasing concentrations of PDGF (0.1-10 g/ml) (Fig. 3D), but also by pretreatment of cells with anti-integrin ␣ 5 ␤ 1 antibodies for 30 min (Fig. 3C). Because BEI-inactivated FMDV, unlike rVP1, did not cause apoptosis (Fig. 1D), we also investigated its effect on Akt. BEI-inacti- Cell lysates were analyzed by Western blotting using anti-Akt or anti-phospho-Akt (p-Akt) antibodies as the primary antibodies. B, the cells were treated with increasing concentrations of PDGF as indicated, and the cell lysates were analyzed as described for A. C, the cells were treated with PBS (control), 0.5 M rVP1, or both anti-integrin ␣ 5 ␤ 1 antibodies (Ab) and 0.5 M rVP1 for fixed periods of time as indicated. For cells treated with both anti-integrin ␣ 5 ␤ 1 antibodies and 0.5 M rVP1, the antibodies were added 30 min before rVP1. D, the cells were treated with PDGF and/or rVP1 as indicated. E, the cells were treated with rVP1 or BEI-inactivated FMDV for 1 h as indicated. Anti-integrin ␣ 5 ␤ 1 antibodies were added 30 min before the rVP1 or BEI-inactivated FMDV treatment. F, after rVP1 and PDGF treatment, cell lysates were analyzed by Western blotting using anti-phospho-GSK-3␤ (p-GSK-3␤) antibodies as the primary antibodies. G, BHK-21 cells were treated with PBS (control), PDGF, or increasing concentrations of rVP1 as indicated, and the cell lysates were analyzed as described for F.

VP1 of FMDV Induces Apoptosis
vated FMDV was found to increase rather than decrease the phosphorylation of Akt (Fig. 3E), substantiating a positive correlation between reduction of Akt phosphorylation and apoptosis.
We then determined whether downstream elements of the Akt pathway, such as GSK-3␤ phosphorylation (24), could also be inhibited by rVP1 treatment. We found that PDGF treatment of BHK-21 cells increased, whereas rVP1 treatment reduced, the phosphorylation of GSK-3␤ (Fig. 3F). The reduction in phosphorylation of Akt and GSK-3␤ upon treatment of BHK-21 cells with rVP1 was dose-dependent (Fig. 3G). To verify that the level of GSK-3␤ correlated positively with BHK-21 cell apoptosis, BHK-21 cells were treated with a non-ATP competitive GSK-3␤ inhibitor, GSK-3␤I, to inhibit the apoptotic effect of GSK-3␤ (38). Our results show that treatment of BHK-21 cells with both PDGF and GSK-3␤I attenuated the apoptotic effect of rVP1 in a dose-dependent manner (Fig. 4, A and B).
rVP1 Activates Cleavage of Procaspases-Caspases, a family of cysteine/aspartic acid proteases, are central regulators of apoptosis (28,29). Fragmentation of DNA has been shown to cause cleavage of procaspase-9, resulting in processes of other procaspases such as procaspase-3 and -7. The initiation of the caspase cascade eventually leads to apoptosis (28,29). The effects of rVP1 treatment on the activation of procaspase-9 and downstream procaspase-7 and -3 were therefore evaluated. We found that rVP1 treatment caused the cleavage of procaspase-9, -7, and -3 in a concentration-dependent manner (Fig. 5).
rVP1 Induces Cancer Cell Apoptosis Irrespective of the Absence of Caspase-3 and S1P Receptor Subtype 3 (S1P 3 )-The Akt pathway has been a major target for treatment of all four major human cancers, i.e. breast, prostate, lung, and colorectal (39 -42). To evaluate further the apoptotic effect and mechanism of action of rVP1, we concentrated on cells that might not be responsive to it. One of these cancer cells is MCF-7, a breast carcinoma cell line with a functional deletion of the caspase-3 gene (43). Buckley et al. (43) found that seven peptides containing an RGD motif were unable to induce apoptosis of MCF-7 cells and concluded that the apoptotic effect of these RGD motif-containing peptides is caspase-3-dependent. To evaluate whether the apoptotic effect of rVP1 would also be caspase-3dependent, we treated MCF-7 cells with rVP1 to evaluate their apoptosis. Our results show that rVP1 caused apoptosis of MCF-7 cells in a concentration-dependent manner and that this effect was reversed by addition of increasing concentrations of GSK-3␤I (Fig. 6, A and B), suggesting that the apoptotic effect of rVP1 is not necessarily dependent on caspase-3.
Other cancer cell lines that we selected for study were PC-3 and 22Rv1, both of which are prostate cancer cells. 22Rv1 cells express S1P 3 and responds to androgen stimulation, whereas PC-3 cells do not (44). Androgen-independent prostate cancer cells such as PC-3 are usually much more lethal and difficult to treat. As S1P binds to S1P 3 to mediate Akt activation and cross-talk with PDGF receptors (44), rVP1 could cause deactivation of Akt via S1P 3 . In this case, treating PC-3 with rVP1 would not result in apoptosis. Our results show that rVP1 not only caused apoptosis of both cell types, but also showed greater apoptotic effect on PC-3 cells, suggesting that the apoptotic effect of rVP1 is not via deactivation of S1P 3 (Fig. 6C). DISCUSSION The capsid protein(s) of FMDV (45) and a variety of pathogens, including adenovirus 2 (46), coxsackievirus A9 (47), and echovirus 22 (48), contain an RGD motif that can bind to cellular integrin receptors such as ␣ v ␤ 1 , ␣ v ␤ 3 , ␣ v ␤ 6 , and ␣ 5 ␤ 1 (6,11,12). Whether the binding of the RGD motif of these pathogens to the receptors elicits any cellular events has not been well documented. Our results in this study show clearly that rVP1 of FMDV bound to the integrin receptors and resulted in cellular apoptosis. Because integrin-mediated cell attachment has been shown to regulate both PI3K and Akt (49,50), our observation that rVP1 treatment resulted in apoptosis and inhibition of Akt within the same effective dose range (Figs. 1C and 3G) is consistent with the proposal that the effect of rVP1 is most likely via its deactivation of Akt after its binding to integrin. It must be noted, however, that FMDV and BEIinactivated FMDV caused activation of Akt and resulted in less cellular apoptosis (Figs. 1D and 3E). These results suggest that, although both FMDV and VP1 bind to integrin, the former acts as an agonist and the latter as an antagonist of integrin receptors. Which portion(s) of FMDV enables it to act  5. Cleavage of procaspase-3, -7, and -9 after rVP1 treatment of BHK-21 cells. The cells were treated with increasing amounts of rVP1 as indicated. After incubation for 16 h, the cells were lysed, and lysates were analyzed by Western blotting using antibodies against caspase-3, -7, and -9 as well as antibodies against procaspase-3 (pro-casp3), -7 (pro-casp7), and -9 (pro-casp9) as the primary antibodies.
as an agonist to activate Akt remains to be elucidated.
The Akt signaling pathway has been shown to regulate cell survival via both BAD and caspase-3 in Chang liver cells (51) and via only caspase-3 in cardiomyocytes (52). Recently, Zhang et al. (53) found that the Akt pathway in HeLa cells affect cell survival via regulation of not only caspase-3, but also caspase-9 and GSK-3␤. It thus appears that, depending on cell types and stimuli, the Akt pathway may regulate cell survival and apoptosis via modification of different downstream molecular targets. Although we have not tested all the potential downstream targets, our results so far reveal clearly that rVP1 treatment of BHK-21 fibroblast cells down-regulates the Akt pathway, resulting in apoptosis via dephosphorylation of GSK-3␤ and cleavage of procaspase-9, -7, and -3 (Figs. [3][4][5]. Whether these are the only major factors or there are other factors involved requires further elucidation. However, the observation that GSK-3␤I could effectively attenuate the apoptotic effect of rVP1 suggests that the effect of rVP1 involves primarily the Akt/GSK-3␤/caspase pathway. By observing that seven RGD motif-containing peptides could not cause apoptosis of MCF-7 cells (human breast carcinoma cells deficient in a functional caspase-3 gene), Buckley et al. (43) concluded that the apoptotic effect of these peptides was most likely due to their direct intracellular activation of caspase-3. However, our results show that the potency of rVP1 in causing apoptosis of MCF-7 cells was only slightly lower than that observed for BHK-21 cells (Fig. 6B), suggesting that the apoptotic effect of rVP1 is not necessarily dependent on the direct activation of caspase-3. Comparison of the amino acid sequences of rVP1 and the peptides used by Buckley et al. reveals that, unlike rVP1, none of the peptides used by Buckley et al. contained leucine at position ϩ1 after the RGD motif. Whether this difference or other amino acids of rVP1 contribute to the difference in the mechanism of action of rVP1 compared with that of the RGD motif-containing peptides used by Buckley et al. remains to be elucidated.
A number of tumor cells undergo apoptotic cell death when treated with chemotherapeutic agents such as camptothecin, cisplatin, adriamycin, and vincristine. Induction of apoptosis has thus been suggested as one of the major means to explore for cancer treatment (54,55). By identifying RGD as an active moiety of fibronectin (36) and by isolation of fibronectin and vitronectin receptors (8,9), Ruoslahti and co-workers have revealed the biological and pharmacological importance of RGD motif-containing molecules. In this study, we found that aqueous soluble rVP1, an RGD motif-containing protein, was extremely potent in causing apoptosis of MCF-7, PC-3, and 22Rv1 cells (ED 50 Ͻ 1 M). Its effect was not completely dependent on either caspase-3 or S1P 3 (Fig. 6). Moreover, as more integrin ␣ 5 ␤ 1 receptors on the cells have been linked to a higher metastatic potential of cancer cells (56), rVP1 may selectively cause the apoptosis of these cancer cells and prevent the occurrence of metastasis.
FMDV affects many domestic livestock and results in economic disaster. However, our finding that a recombinant form of VP1 of FMDV can kill cancer cells shows that the capsid protein of this pathogen may be beneficial for human beings.