HeLa cells purchased from American Type Culture Collection (Manassas, VA, USA) were cultured in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal bovine serum at 37°C in a humidified incubator with 5% CO₂. CV-B3 (Nancy strain) was maintained at −80°C freezers. Ultraviolet (UV)-irradiated virus was prepared and viral infection was performed as previously described (Luo et al., 2002).

Hemagglutinin (HA)-tagged wild-type GAB2 (GAB2^WT) was a generous gift from Dr. Roger Daly at the Monash University (Melbourne, Australia). The HA-tagged GAB2^G238E was established by replacing the glycine (G) residue at amino acid 238 of wild-type GAB2 with glutamic acid (E). The wild-type GAB2 was used as a template to generate two truncated forms of GAB2. The resulting fragments were cloned into a vector expressing 3 × Flag at its N-terminus (3 × Flag-GAB2-N₁₋₂₃₇ and 3 × Flag-GAB2-C_238−676). The small interfering RNAs (siRNAs) against human GAB1 and human GAB2 were purchased from Dharmacon (Ottawa, ON, Canada).

Transfection of plasmids and siRNAs was performed according to manufacturer's instructions using Lipofectamine® 2000 (#11668019, Invitrogen, Burlington, ON, Canada) for plasmid transfection or Oligofectamine® (#12252-011, Invitrogen, Burlington, ON, Canada) for siRNA transfection. For drug treatment, pan-caspase inhibitor zVAD (50 μM, #550377, BD Biosciences, San Jose, CA, USA) was applied for the blockage of general caspase activation. SB203580 (50 μM, #S8307, Sigma-Aldrich, St. Louis, MO, USA) was used for inhibition of p38 activity.

Cellular proteins were extracted using modified oncogene science lysis buffer (250 mM NaCl, pH 7.2, 50 mM Tris-HCl, 0.1% NP-40, 2 mM EDTA, and 10% glycerol) supplemented with protease inhibitors. The concentrations of the protein were determined by Bradford assay. A total of 20–40 μg of protein per sample was loaded for sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Western blotting was performed as described previously (Shi et al., 2014). Primary antibodies used in this study were: (1) anti-human GAB2 (#3239, Cell signaling, Beverly, MA, USA); (2) anti-Flag (sc-807, Santa Cruz, Dallas, TX, USA); (3) anti-HA (sc-805, Santa Cruz, Dallas, TX, USA); (4) anti-viral capsid protein VP1 (NCL-ENTERO, Leica biosystems, Concord, ON, Canada); (5) anti-cleaved caspase-3 (#9661, Cell signaling, Beverly, MA, USA); (6) anti-phospho-p44/42 MAPK (ERK1/2) (#4094, Cell signaling, Beverly, MA, USA); (7) anti-phospho-p38 MAPK (#4511, Cell signaling, Beverly, MA, USA); (8) anti-phospho-SAPK/JNK (#4668, Cell signaling, Beverly, MA, USA); (9) anti-β-actin (#2228, Sigma-Aldrich, St. Louis, MO, USA); and (10) anti-phospho-HSP27 (sc-81498, Santa Cruz, Dallas, TX, USA).

Purified viral proteinase 2A and catalytically inactive 2A were generous gifts from Dr. Eric Jan at the University of British Columbia. The assay was performed according to the protocol previously described (Deng et al., 2015). The cleaved fragments of GAB2 were examined by western blotting. CV-B3-infected HeLa cells were used as a positive control.

The virus titers in the supernatant of CV-B3-infected cells were measured by plaque assay as described previously (Deng et al., 2015). Briefly, the supernatant was serially diluted and overlaid on a monolayer of HeLa cells with 90% confluence for 1 h prior to the replacement of the medium with complete DMEM containing 0.75% agar. After 3-day incubation, the cells were fixed with Carnoy's fixative (75% ethanol and 25% acetic acid) and then stained with 1% crystal violet. The resulting plaques were counted and the virus titers were calculated as plaque forming unit (PFU) per milliliter.

All results presented are representative of at least three independent experiments. Results are expressed as means ± standard deviations (SD). Statistical analysis was performed with unpaired student's t-test. The p < 0.05 were considered to be statistically significant.

To understand the possible role of GAB2 in CV-B3 infection, we first examined the protein level of GAB2 upon viral infection. As shown in Figure 1A, protein level of GAB2 began to decrease at 5 h and disappeared at 7 h post-infection, accompanied by the generation of an additional band at around 60 kDa using an antibody against the C-terminus of GAB2, suggesting a possible cleavage event. To verify this, a plasmid expressing C-terminal HA-tagged GAB2 (GAB2-HA) was utilized. HeLa cells transfected with GAB2-HA were infected with CV-B3 for indicated hours and protein expression of GAB2 was examined. We found that, similar to the finding in Figure 1A, an extra band of exogenous GAB2 was detected at around 60 kDa using an anti-HA antibody (Figure 1B), indicating that GAB2 is also cleaved after CV-B3 infection. It is noted that the cleavage efficiency of exogenously transfected GAB2 is much lower than that of endogenous GAB2. We have previously reported that liposome-mediated cDNA transfection inhibits CV-B3 attachment to the cells by disrupting membrane cholesterol (Wong et al., 2006). Thus, the decreased cleavage efficiency is likely a result of reduced viral infectivity in transfected cells.

We then investigated the mechanism leading to GAB2 cleavage. We first tested whether viral replication is required for GAB2 cleavage. We utilized UV-irradiated viruses, which are capable of interacting with viral receptor and subsequent entering into cells, but unable to replicate. We showed that GAB2 was not cleaved in cells infected with UV-CV-B3, suggesting that GAB2 cleavage is dependent on viral replication (Figure 2A).

CV-B3 encodes two proteinases, 2A and 3C, which not only process viral polyprotein into individual structural and nonstructural protein, but also target cellular proteins to either facilitate infection or provoke host anti-viral machinery (Jurgens et al., 2006; Zaragoza et al., 2006; Feng et al., 2014). In vitro cleavage assay was performed to determine whether viral proteinases contribute to the cleavage of GAB2 upon CV-B3 infection. As shown in Figure 2B, incubation with wild-type 2A (2A^WT) led to the generation of a cleavage band at ~60 kDa, corresponding to what was detected during CV-B3 infection. However, the catalytic mutant of 2A (2A^mut) failed to cleave GAB2, suggesting that cleavage of GAB2 is mediated through the catalytic activation of 2A. In vitro cleavage assay was also conducted using purified 3C proteinase. However, no cleavage products were detected (data not shown).

Furthermore, to rule out the possible role of caspase activation in GAB2 cleavage, we treated cells with the general caspase inhibitor, z-VAD. We found that caspase inhibition did not prevent the cleavage of GAB2 (Figure 2C). The inhibition of caspase activity by z-VAD was confirmed by the blockage of caspase-3 cleavage (Figure 2D). Taken together, our results suggested that the cleavage of GAB2 detected during CV-B3 infection is mediated via the catalytic activity of viral proteinase 2A, independent of the activation of caspase.

Based on the reported cleavage consensus motif of proteinase 2A (i.e., L/I/MxT/S/Nx//G, // indicates the scissile bond, P4 position—L (leucine), I (isoleucine), or M (methionine), P2 position—T (threonine), S (serine), or N (asparagine), P1' position is commonly G (glycine), x indicates any amino acid residues) (Wong et al., 2012), we found one potential cleavage sequence (²³⁴LASHG²³⁸) on GAB2. To determine whether GAB2 is cleaved at this site, we performed site-directed mutagenesis to replace G at position 238 with glutamic acid (E). As shown in Figure 3A, GAB2^G238E mutant was uncleavable upon CV-B3 infection, suggesting that G238 is the cleavage site. Figure 3B illustrates the functional domains, the cleavage site, and the resulting cleavage fragments of GAB2 following CV-B3 infection.

We have previously shown that knockdown of GAB1 inhibits CV-B3 replication (Deng et al., 2015). Here, we questioned whether GAB2 plays a similar role in CV-B3 infection. Figure 4 revealed that gene-silencing of GAB2 resulted in significant decreases in viral protein production (Figures 4A,B) and virus titers (Figure 4C), accompanied by a large reduction of CV-B3-induced phosphorylation of JNK and p38 (Figure 4A), suggesting a pro-viral function for GAB2. The importance of p38 activation in viral replication was further demonstrated using a selective p38 inhibitor (SB203580). Figure 4D showed that treatment with SB203580 led to a marked reduction of viral protein expression. Inhibition of p38 activity by SB203580 was confirmed by reduced levels of phospho-HSP27, a downstream target of p38 (Figure 4D). No significant effects of JNK inhibitor on viral replication were observed (data not shown). Interestingly, unlike deletion of GAB1 that caused a decrease in phosphorylated ERK1/2, knockdown of GAB2 had no effect on virus-mediated ERK1/2 phosphorylation (Figure 4A). We further examined the impacts of forced expression of exogenous GAB2 on viral replication. We found that viral protein levels and viral loads failed to further increase in cells overexpressing GAB2 (data not shown), indicating that the levels of endogenous GAB2 are already saturated for viral replication.

We next sought to determine the consequence of GAB2 cleavage during viral infection. HeLa cells were either transfected with GAB2^WT or non-cleavable GAB2 (GAB2^G238E). We showed that the levels of viral protein (Figures 5A,B) and virus titers (Figure 5C) were comparable between cells expressing GAB2^WT and GAB2^G238E, suggesting that cleavage of GAB2 had no direct benefits to virus replication. Moreover, we examined the influence of overexpression of either GAB2-N or GAB-C on viral replication. As shown in Figures 5D–F, neither VP1 levels nor virus titers were significantly altered in cells expressing GAB2 cleavage fragments compared with empty vector control. Collectively, our results suggest that cleavage of GAB2 results in the loss of its function in promoting viral infection, rather than the gain of a pro-viral activity.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. The reviewer RPK and handling Editor declared their shared affiliation and the handling Editor states that the process nevertheless met the standards of a fair and objective review.