P58IPK inhibits coxsackievirus-induced apoptosis via the PI3K/Akt pathway requiring activation of ATF6a and subsequent upregulation of mitofusin 2

Authors

  • Huifang M. Zhang,

    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Ye Qiu,

    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Xin Ye,

    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Maged G. Hemida,

    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Paul Hanson,

    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author
  • Decheng Yang

    Corresponding author
    1. Department of Pathology and Laboratory Medicine, University of British Columbia, Center for Heart Lung Innovation, St. Paul's Hospital, Vancouver, British Columbia, Canada
    Search for more papers by this author

Summary

Previously we found that prolonged endoplasmic reticulum (ER) stress caused by coxsackievirus B3 (CVB3) infection led to p58IPK downregulation and subsequent cell apoptosis. This finding implies that p58IPK expression benefits cell survival and counteracts CVB3-induced apoptosis. In testing this hypothesis, we first found that PI3K/Akt survival pathway is more sensitive than ERK1/2 in response to p58IPK expression. This finding was further verified by silencing p58IPK with specific siRNAs, which led to the significant suppression of phosphorylation of Akt (p-Akt) but not ERK1/2. Further, using CVB3-infected cell line expressing dominant negative ATF6a (DN-ATF6a), we found that expression of p58IPK and p-Akt was significantly reduced, which led to the decreased cell viability. However, when the DN-ATF6a cells were transiently transfected with p58IPK, an opposite result was obtained. Finally, by CVB3 infection of cells stably expressing p58IPK, we found that CVB3-induced mitochondria-mediated apoptosis was suppressed, which was evidenced by the reduced cytochrome c release and upregulation of the mitochondrial membrane protein mitofusin 2. However, silencing p58IPK with either specific siRNAs or DN-ATF6a sensitized cells to CVB3-induced apoptosis. These results suggest that p58IPK suppresses CVB3-induced apoptosis through selective activation of PI3K/Akt pathway that requires activation of ATF6a and subsequently upregulates mitofusin 2.

Introduction

The endoplasmic reticulum (ER) contains a number of molecular chaperones and co-chaperones physiologically involved in protein synthesis, folding and transportation. ER also serves as a site for calcium storage (Paschen, 2001). Perturbing the protein synthetic machinery or ER Ca2+ homeostasis will induce a condition that is referred to as ER stress (Kaufman, 1999) or unfolded protein response (UPR) (Schroder and Kaufman, 2005). Initially, the UPR seeks to re-establish ER homeostasis via translation attenuation to decrease ER load of new proteins, transcriptional activation of chaperone genes to increase the folding capability, and activation of the ER-associated degradation machinery to clear misfolded proteins (Harding et al., 2000; Schroder and Kaufman, 2005). If ER stress is unresolved, apoptosis is triggered (Lucia et al., 2006). Three ER transmembrane proteins are instrumental in co-ordinating the UPR signal. In response to ER stress, (i) the activating transcription factor 6a (ATF6a) is translocated to the Golgi complex, where it is cleaved by proteases to yield an active transcription factor for expression of chaperone genes (e.g. GRP78) involved in protein folding (Haze et al., 1999); (ii) inositol-requiring enzyme 1(IRE1) catalyses X box binding protein 1 (Xbp1) mRNA splicing, leading to the generation of an active transcription factor for expression of chaperones and co-chaperones (e.g. p58IPK) involved in protein folding and degradation (Reimold et al., 2000; Lee et al., 2003); (iii) PKR-like ER protein kinase (PERK) phosphorylates eukaryotic initiation factor 2α (eIF2α), leading to rapidly attenuation of cellular mRNA translation (Harding et al., 2002).

p58IPK is an ER-luminal co-chaperone associated with GRP78. During ER stress, it plays an important role in protein processing efficiency of the ER by binding to misfolded proteins and acting as a co-chaperone (Oyadomari et al., 2006; Rutkowski et al., 2007; Petrova et al., 2008). More importantly, p58IPK is also a PERK cellular inhibitor that is activated during ER stress. Activation of PKR or PERK will result in eIF2α phosphorylation and global attenuation of eIF2α-dependent cellular mRNA translation. However, p58IPK activation inhibits eIF2α phosphorylation and helps to recover protein translation suppression caused by ER stress (Harding et al., 2002).

The ER stress response can be triggered by a variety of stimuli and is linked to pathophysiology of numerous diseases including viral infections (Trujillo-Alonso et al., 2011; Galindo et al., 2012). Coxsackievirus B3 (CVB3), a positive single-stranded RNA virus in the Picornaviridae family, is a major pathogen of viral myocarditis, particularly in children and young adults (Blauwet and Cooper, 2010). This heart disease often progresses into its chronic phase, dilated cardiomyopathy, which is a major cause of sudden unexpected death (Kuhl et al., 2005). Our previous study of host responses to CVB3 infection found that CVB3 triggers ER stress and prolonged ER stress results in cell apoptosis by activating proapoptotic gene expression (Zhang et al., 2010). Intriguingly, we further revealed that p58IPK, an ER co-chaperone and PERK inhibitor that is often upregulated during many other types of viral infections (Goodman et al., 2007), was downregulated at both transcription and translational levels during CVB3 infection (Zhang et al., 2010). This downregulation seems to be responsible for the transition of signal pathways from protective UPR to an apoptosis program. It has been documented that CVB3 infection induces mitochondria-mediated apoptosis (Chau et al., 2007), which implies that p58IPK expression can inhibit CVB3-induced mitochondrial injury and block the apoptosis pathway. This hypothesis was successfully verified in this study and we further found that p58IPK expression counteracted the CVB3-induced mitochondria-mediated apoptosis through activation of phosphatidylinositol 3-kinase (PI3K)/Akt survival pathway at the early stage of infection, which requires the activation of ATF6a and the subsequent upregulation of expression of mitochondrial membrane protein mitofusin 2.

Results

p58IPK expression activates PI3K/Akt but not ERK1/2 pathways

Our previous studies and others found that activation of ERK1/2 or PI3K/Akt pathway benefits cell survival and CVB3 replication (Luo et al., 2002; Esfandiarei et al., 2004). Recently, we also showed that CVB3 infection downregulated p58IPK and shifted the protective ER stress response to cell apoptosis at the late phase of infection (Zhang et al., 2010). To determine the relationship between p58IPK expression levels and the status of both ERK1/2 and PI3K/Akt survival pathways, p58IPK expressing cells and vector-transfected control cells were cultured in growth medium for 18 h and then starved in a serum-free Dulbecco's modified Eagle's medium (DMEM) for 2 h. The cells were treated with inhibitors of PI3K or DMSO (control) at different concentrations for 1 h. Cell lysates were used for Western blot analyses to detect expression of p58IPK and phosphorylation of both Akt (p-Akt, Ser473) and ERK1/2 (p-44/p-42 ERK1/2). Figure 1A demonstrated that LY294002 completely blocked p-Akt basal level expression in control cells (lanes 1–4) while in p58IPK cells the p-Akt basal level expression was still detectable (lanes 5–8) and the DMSO-treated cells showed a significant increase in p-Akt (lane 5) compared with control (lane 1). This increase is probably due to that p58IPK expression counteracts the inhibitory effect of LY294002 on phosphorylation of Akt. Nevertheless, the pERK1/2 expression could only be detectable at basal level in control cells, indicating that p58IPK expression reduces p-ERK1/2 basal level expression. These data suggest that p58IPK expression activates PI3K/Akt but not ERK1/2 pathways.

Figure 1.

p58IPK expression activates PI3K/Akt but not ERK1/2 pathway.

A. p58IPK expression counteracts the LY294002-caused suppression of Akt phosphorylation. HeLa cells stably expressing p58IPK were treated with DMSO or inhibitor LY294002 at the indicated concentrations for 1 h and then the cell lysates were harvested for Western blot analyses of p58IPK expression and phosphorylation of Akt and ERK1/2. Total Akt and ERK1/2 were also detected. Vector-transfected cells were used as a control. Levels of p-Akt and p-ERK1/2 were determined by densitometric analysis of data from 3 independent experiments using NIH ImageJ program, normalized to total Akt or total ERK1/2 and presented as mean ± SD (lower panel). *P < 0.05.

B. CVB3-induced downregulation of p58IPK suppresses PI3K/Akt but slightly enhances ERK1/2 activation. p58IPK cells were infected with CVB3 or sham-infected with PBS. At different time point pi, cell lysates were subject to Western blot analysis using indicating antibodies. β-actin was used as a loading control. Levels of p-Akt and p-ERK1/2 were quantified by densitometric analysis as described in (A) (lower panel).

C and D. (C) p58IPK cells were treated with either U0126 (20 μM) or LY294002 (50 μM) (D) for 1 h and then infected with CVB3. At indicated time point pi, Western blot analyses were conducted using indicated antibodies. Levels of p-Akt and p-ERK1/2 were quantified by densitometric analysis as described in (A) (lower panel).

The above experiment has demonstrated the positive regulatory role of p58IPK on PI3K/Akt activation but not on the ERK1/2 pathway in the absence of CVB3 infection. Next, we further confirmed this result using p58IPK cells infected with CVB3. Briefly, p58IPK cells were incubated in the same conditions as described above and then infected with CVB3 or sham-infected with phosphate-buffered saline (PBS). At the indicated time points post infection (pi), cell lysates were collected for Western blot analysis. As shown in Fig. 1B, both Akt and ERK1/2 were activated by CVB infection via phosphorylation. However, with the progression of CVB3-induced downregulation of p58IPK, p-Akt levels in both p58IPK and vector cells were sharply decreased at 6 h pi; while p-ERK1/2 production level was slightly increased at 6 h pi in both p58IPK and control cells, and particularly in p58IPK cells the p-ERK1/2 levels did not show a decrease until 10 h pi. This data implies that CVB3-induced downregulation of p58IPK expression dramatically reduced phosphorylation of Akt but not that of ERK1/2 at 6 h pi. In other words, this experiment indicates that p58IPK expression activates PI3K/Akt but not ERK1/2 pathways during CVB3 infection. It is worth mentioning that with the progression of CVB3-induced downregulation of p58IPK and subsequent transition of UPR to cell death, both p-Akt and p-ERK1/2 were decreased at 10 h pi. This is probably due to the virus-induced cell death and proteasome-mediated cellular protein degradation (Breusing and Grune, 2008).

To further solidify the finding, p58IPK expressing cells were treated with either LY294002 (50 μM) or ERK1/2 inhibitor U0126 (20 μM) for 1 h before CVB3 infection. Western blot analyses demonstrated that inhibitor LY294002 did not completely block CVB3-induced production of p-Akt at 6–8 h pi (lanes 7–8) (Fig. 1C); this may be due to that p58IPK-induced activation of Akt counteracted the inhibitory effect of LY294002 on Akt phosphorylation. Nevertheless, the p58IPK-induced counteractive effect against this inhibitor was not observed in the activation of ERK1/2, which was evidenced by the complete suppression of ERK1/2 phosphorylation. However, in inhibitor U0126-treated cells (Fig. 1D), although p-ERK1/2 production was dramatically reduced (lane 7–8) compared with the control (lanes 3–4), p-Akt production not only was not reduced but also had a significant increase (lane 6–8) compared with the controls (Lane 2–4). This increase may be attributed to both the induction by p58IPK and suppression of p-ERK1/2 signalling by U0126. This latter result supports the previous report on the cross-talk between PI3K/Akt and ERK1/2 signalling pathways during CVB3 infection (Luo et al., 2002; Esfandiarei et al., 2004). These results imply that p58IPK expression is a sensitive signal for Akt but not for ERK1/2 activation. These data further suggest that PI3K/Akt survival pathway can positively respond to p58IPK expression and play an important role in p58IPK-induced promotion of cell survival.

Downregulation of p58IPK decreased cell viability through the inactivation of Akt

To further determine the mechanism of p58IPK-mediated regulation of CVB3 replication, we silenced p58IPK expression in CVB3-infected p58IPK cells with specific siRNAs and then detect the suppression of p58IPK expression at mRNA and protein levels by RT-PCR and Western blot analysis respectively. We also examined ERK1/2, Akt and eIF2α phosphorylation by Western blot analyses. In addition, we further determined the effect of gene silencing on CVB3 replication by measuring the host cell viability, viral VP1 production and CVB3 particle formation. Figure 2A and B, respectively, show the attenuation of p58IPK expression at both the transcriptional and translational levels by siRNAs compared with the control cells treated with scrambled siRNAs. This downregulation of p58IPK at protein level was not dramatic at different time points pi, which may be due to the suppression caused partially by CVB3 infection (Zhang et al., 2010), which masked the changes caused by siRNA. However, this p58IPK attenuation is mirrored by the increased expression of p-eIF2α, a downstream target of PKR or PERK, which are negatively regulated by p58IPK (Harding et al., 2002). The siRNA treatment also markedly reduced p-Akt but only slightly decreased p-ERK1/2 expression compared with the control. However, the total Akt and ERK1/2 levels were not significantly affected. The attenuated p-Akt signals resulted in decreased cell viability, which was demonstrated by MTS assay (Fig. 2C). The reduced CVB3 replication was demonstrated by the decrease of both VP1 protein synthesis (Fig. 2B) and infectious viral particle formation (Fig. 2D). These data indicate that PI3K/Akt survival pathway plays an important role in p58IPK-mediated regulation of CVB3 replication and that downregulation of p58IPK causes inactivation of Akt and inhibition of CVB3 replication at early phase of infection.

Figure 2.

siRNA silencing of p58IPK decreases cell viability through the inactivation of Akt but not ERK1/2 during CVB3 infection.

A. HeLa cells were transfected with siRNAs targeting p58IPK or scrambled (Scr) siRNA and then infected with CVB3. Total RNAs isolated from the cells were used to detect p58IPK mRNA expression by RT-PCR. Quantification of the mRNAs was conducted by densitometric analysis using NIH ImageJ program, normalized to β-actin.

B. Western blot analysis was conducted using the cell lysates to detect expression of p58IPK, phosphorylation of signalling molecules and production of CVB3 VP1. Total Akt (T-Akt) and total ERK1/2 (T-ERK1/2) were also detected. Levels of p-Akt and p-ERK1/2 were quantified by densitometric analysis using NIH ImageJ program, normalized to total Akt and total ERK1/2 respectively, and presented as mean ± SD (lower panel).

C. Cell viability was measured by MTS assay at 10 h pi and converted to a percentage of the control HeLa cells receiving sham infection and no siRNA transfection (set as 100% survival).

D. Viral plaque assay was performed using supernatants (Sup) and cell pellets (Pel) from the culture at 10 h pi. A cell lysate treated with scrambled siRNAs was used as a control. Error bars represent mean ± SD. *P < 0.05.

Dominant negative (DN)-ATF6a expression suppresses p58IPK-mediated activation of survival signal Akt

To further verify that p58IPK-mediated activation of PI3K/Akt survival pathway requires activation of AFT6a, HeLa cell line stably expressing DN-ATF6a (pcDNA3.1-ATF6a (171–373) was established and used to detect the expression levels of p58IPK, p-Akt and VP1 after CVB3 infection by Western blot analysis. Figure 3A shows that p58IPK expression was dramatically downregulated and phosphorylation of Akt was also gradually diminished compared with the vector-transfected control cells at each corresponding time point pi. However, the total Akt levels were not dramatically affected. As a result, VP1 protein expression levels were significantly reduced, which is consistent to the viral plaque assay result (Fig. 3B). This experiment was repeated by using HL-1 cardiomyocytes transiently transfected with the DN-ATF6a plasmid and similar results were obtained (Fig. 3C). To further confirm that p58IPK is a downstream effector of ATF6a, we performed two additional experiments: first, we transiently transfected the DN-ATF6a cell line with either a p58IPK plasmid or an empty vector and then detected the same gene expression as described above in Fig. 3A after CVB3 infection. We demonstrated that the expression of p58IPK, p-Akt and VP1 was all increased compared with the control (Fig. 3D). Nevertheless, in our second experiment using established p58IPK cell line transiently transfected with a DN-ATF6a plasmid, an opposite result was observed, i.e., expression of p58IPK, p-Akt and VP1 was significantly suppressed (Fig. 3E). However, in these two experiments total Akt expression was not significantly changed. These results suggest that p58IPK is a downstream effector of ATF6a, regulating PI3K/Akt survival pathway.

Figure 3.

DN-ATF6a expression suppresses p58IPK-mediated activation of Akt.

A. HeLa cells stably expressing DN-ATF6a were infected with CVB3. Western blot analysis was conducted to detect p58IPK expression, Akt phosphorylation and CVB3 VP1 production. Total Akt and β-actin were used as loading controls. Levels of p-Akt were quantified by densitometric analysis using NIH Image J program, normalized to total Akt, and presented as mean ± SD.

B. CVB3 replication was further examined by plaque assay to determine the PFU ml−1. *P < 0.05.

C. HL-1 cardiomyocytes were transiently transfected with a DN-ATF6a plasmid and then infected with CVB3. Western blot analysis was performed to detect gene expression and the levels of p-Akt expression were quantified by densitometry as described in (A) above.

D. HeLa cells stably expressing DN-ATF6a were transiently transfected with a p58IPK plasmid and then infected with CVB3. Western blot was performed to detect downstream gene expression and the levels of p-Akt were quantified as described in (A).

E. HeLa cells stably expressing p58IPK were transiently transfected with a DN-ATF6a plasmid and then infected with CVB3. Western blot was performed to detect downstream gene expression and the levels of p-Akt were quantified as described in (A).

p58IPK is a mediator linking the ATF6a and PI3K/Akt survival pathway

To determine the position of co-chaperone p58IPK relative to PI3K/Akt and ATF6a in the signalling cascade, two experiments were conducted. First, p58IPK cell line transiently transfected with a Myc-His-tagged DN-Akt (K179M) plasmid (Myc-His-DN-Akt) was infected with CVB3. Western blot analysis showed that expression of DN-Akt induced downregulation of both p-Akt and VP1 expression in p58IPK cells at each time point pi compared with the corresponding control samples transfected with an empty vector (Fig. 4A). The second experiment was performed by transient transfection of DN-ATF6a cell line with a haemagglutinin (HA)-tagged constitutively active (CA)-Akt(T308D, S473D) plasmid (HA-CA-Akt). Figure 4B demonstrates the opposite results, which shows that DN-ATF6a expression did not block activation of Akt and the promotion of CVB3 VP1 production caused by CA-Akt. These data suggest that p58IPK is an ATF6a-regulated downstream effector and an upstream modulator of PI3K/Akt.

Figure 4.

DN-Akt attenuates VP1 expression in p58IPK cells while CA-Akt enhances VP1 production in DN-ATF6a cells during CVB3 infection.

A. HeLa cells stably expressing p58IPK were transiently transfected with a Myc-His-tagged DN-Akt plasmid and then infected with CVB3. At different time points pi, expression of p-Akt and CVB3 VP1 protein was detected by Western blot analysis and the levels of VP1 were quantified by densitometric analysis using NIH ImageJ program, normalized to total β-actin, and presented as mean ± SD. *P < 0.05.

B. HeLa cells stably expressing DN-ATF6a were transiently transfected with a HA-tagged CA-Akt plasmid and then infected with CVB3. At different time points pi, expression of p-Akt and CVB3 VP1 protein was detected by Western blot analysis and the levels of VP1 were quantified by densitometric analysis as described in (A).

p58IPK expression inhibits CVB3-induced mitochondria injury by upregulation of mitofusin 2 and reduction of cytochrome c release

It has been documented that CVB3 infection induces mitochondria-mediated apoptosis (Chau et al., 2007). Our recent study showed that p58IPK was gradually downregulated during CVB3-induced progression of cell apoptosis (Zhang et al., 2010). These previous studies imply that p58IPK is a negative regulator of CVB3-induced mitochondrial injury and subsequent cell death. To verify this speculation, p58IPK cells and vector cells were infected with CVB3 or sham-infected with PBS and then Western blot analysis was conducted using proteins prepared from cytosolic and mitochondrial fractions to detect cytochrome c release from mitochondria. Figure 5A demonstrates that p58IPK expression reduced cytochrome c release from mitochondria in p58IPK cells compared with the vector cells. By using whole protein lysates, we also detected the expression of mitofusin 2 protein after CVB3 infection and found that mitofusin 2, a mitochondrial outer membrane protein responsible for maintenance and operation of the mitochondrial network, was enhanced in p58IPK cells compared with the control cells at each corresponding time point pi (Fig. 5B). These data on p58IPK-mediated mitofusin 2 upregulation were further solidified by immunochemical staining of mitofusin 2 protein in p58IPK expressing cells (Fig. 5C).

Figure 5.

p58IPK expression inhibits CVB3-induced mitochondria-mediated injury by upregulation of mitofusin 2 and reduction of cytochrome c release.

A. HeLa cells stably expressing p58IPK or vector cells were infected with CVB3 or sham-infected with PBS. Cells were harvested at 10 h pi for isolation of cytosolic and mitochondrial proteins to detect cytochrome c release. Cox IV protein was used as a purity control for mitochondrial proteins. The levels of cytochrome c were quantified by densitometric analysis of data from three independent experiments using NIH ImageJ program, normalized to total α-tubulin and presented as mean ± SD. *P < 0.05.

B. HeLa cells stably expressing p58IPK or vector cells were infected (+) with CVB3 or sham-infected (−) with PBS. At different time points pi cells were harvested for isolation of total proteins to detect mitofusin 2 by Western blot analysis. The levels of mitofusin 2 were quantified by densitometry as described in (A).

C. Immunostaining p58IPK cells and vector cells culturing on glass coverslips were stained with an anti-mitofusin 2 primary antibody and then with a goat anti-rabbit IgG labelled with red fluorescent Alexa Fluor 594. The nuclei were counterstained with DAPI. Cells images were analysed by confocal microscopy.

Downregulation of p58IPK sensitizes the cells to mitochondria-mediated apoptosis during CVB3 infection

The data presented in Fig. 5 indicate that p58IPK counteracts CVB3-induced mitochondria-mediated apoptosis. This finding was further verified by downregulation of p58IPK using DN-ATF6a cells. We first demonstrated that in CVB3-infected DN-ATF6a cells downregulation of p58IPK by expression of the DN-ATF6a protein resulted in a decreased expression of mitofusin 2 protein in each time point pi compared with control cells (Fig. 6A). We further found that this downregulation of p58IPK by DN-ATF6a also coincidently decreased expression of prosurvival protein Bcl-2, enhanced activation of caspase-9 and caspase-3 and cleavage of poly ADP-ribose polymerase (PARP) (Fig. 6B). A similar result was also observed when treating the HeLa cells with specific siRNAs to silence p58IPK expression (Fig. 6C).

Figure 6.

Knockdown of p58IPK expression by DN-ATF6a or specific siRNAs sensitizes cells to CVB3-induced mitochondria-dependent apoptosis.

A. HeLa cells stably expressing DN-ATF6a or vector cells were infected with CVB3 or sham-infected with PBS. Western blot analysis was conducted to detect p58IPK and suppression of mitofusin-2 expression. The levels of mitofusin 2 were quantified by densitometry, normalized to β-actin and presented as mean ± SD. *P < 0.05.

B. Cell lysates obtained in (A) were also used to detect Bcl-2 expression, activation of caspase-9 and caspse-3 and cleavage of PARP. The levels of Bcl-2 expression were quantified by densitometry as described in (A).

C. HeLa cells were transfected with p58IPK siRNAs or scrambled siRNA and then infected with CVB3 or sham-infected with PBS. At indicated time points pi, cell lysates were used for Western blot analysis of p58IPK expression and the corresponding effects on gene expression as described in (B). The levels of Bcl-2 expression were quantified as described in (A).

Discussion

p58IPK was initially identified as a cellular inhibitor of dsRNA-activated eIF2α kinase, PKR, in influenza virus infection (Lee et al., 1990; 1992; Goodman et al., 2011). Since then, its role in viral replication has been studied only in a couple of other viruses, such as VSV (Goodman et al., 2009) and plant viruses (Bilgin et al., 2003). p58IPK upregulation is also induced by some non-infectious stimuli (Gupta et al., 2010). Further studies have demonstrated that p58IPK plays a protective role during ER stress and lack of p58IPK would cause a number of disorders such as diabetes and lung injury (Ladiges et al., 2005; Goodman et al., 2009; Datta et al., 2010). p58IPK upregulation in the ER lumen is a cellular response against the detrimental stress caused by some virus infections. Intriguingly, our previous study found that CVB3 infection, an oncolytic virus, significantly downregulated p58IPK at both the transcriptional and translational levels (Zhang et al., 2010). As a result of this downregulation, the protective UPR was shifted to cell apoptosis by activation of proapoptotic transcription factors and several caspases at the late stage of infection.

As mentioned above, p58IPK, a co-chaperone of GRP78 localized in ER lumen, plays a protective role in response to ER stress (Rutkowski et al., 2007). Lack or downregulation of this gene expression would weaken the host defence mechanism. Based on these previous findings, one can hypothesize that if the p58IPK expression is artificially enhanced, it may improve the pathophysiological conditions. To test this hypothesis we performed a series of experiments using either stably transfected HeLa cells expressing p58IPK or other approaches, such as gene silencing with siRNA or transfection with a DN-ATF6a plasmid, to alter p58IPK expression levels. Then we test the effect of the altered p58IPK expression on the downstream signal pathways leading to cell survival or death during CVB3 infection. We found that p58IPK expression promoted cell survival and counteracted the virally induced cell apoptosis.

The ER stress response consists of an early phase in which protein synthesis is inhibited by eIF2α phosphorylation and a later phase in which the genes that promote ER capability are induced. During the early phase of CVB3 infection, eIF2α is phosphorylated due to PKR activation, which results in inhibition of cellular protein translation and thus the decrease of cell viability. At this early stage when p58IPK has not been downregulated yet by CVB3, p58IPK expression can inhibit PKR and decrease the p-eIF2α-dependent suppression of global protein translation. This host defence process will enhance cell survival capability. To identify the signal pathways that p58IPK relies on to promote cell survival and counteract the shift from protective UPR to cell apoptosis caused by persistent CVB3 infection, we infected p58IPK expressing cells with CVB3 to gradually downregulate p58IPK expression and then determined which cellular survival pathway(s) was more sensitively affected in a timely manner. We found that this downregulation of p58IPK dramatically affected activation of PI3K/Akt pathway but not that of ERK pathway. This finding was further confirmed by using inhibitors of ERK and PI3K, demonstrating that p58IPK expression counteracted the inhibitory effect of LY294002 on phosphorylation of Akt but not the U0126 on phosphorylation of ERK1/2. These data indicate that p58IPK expression activates PI3K/Akt but not the ERK1/2 signalling.

On the basis of these results, we further demonstrated that the activation of PI3K/Akt was tightly controlled by the upstream ATF6a-p58IPK cascade. Silencing of p58IPK with specific siRNAs or transfection of cells with a DN-ATF6a plasmid strongly suppressed PI3K/Akt activation and the subsequent CVB3 replication. However, overexpression of p58IPK resulted in activation of PI3K/Akt survival pathway and promotion of viral protein synthesis at early phase of infection.

It has been reported that ER stress triggered by different stimuli activates PI3K/Akt and/or ERK1/2 pathway, conferring cellular resistance to ER stress-mediated cell death. However, the activation of specific survival pathway(s) is highly variable and depends on cell type or growth conditions. Hu et al. showed that both PI3K/Akt and ERK1/2 are activated in ER stress and subsequently prevent the apoptosis response in human cancer cells (Hu et al., 2004). Price et al. demonstrated PI3K-mediated signalling pathway promotes chondrocyte survival from ER stress but it does not depend on ERK1/2 activation (Price et al., 2010). Similar reports on the involvement of PI3K/Akt pathway in regulating cell's survival-death decisions in ER stress conditions have been documented (Hyoda et al., 2006; Hosoi et al., 2007). All above reports are studies on UPR induced by pharmacological agents and have not provided data on the relationship between activation of survival pathway and expression of p58IPK. Recently, an investigation using influenza viral infection, a negative single-stranded RNA virus, found that MK2 and MK3, the downstream targets of MAPK pathway, are activated and support viral replication by promoting cell survival through interactions with p58IPK (Luig et al., 2010). Although this report links the activation of cell survival signalling pathway to p58IPK expression, it identified a different survival pathway (comparing with ours) activated by p58IPK during virus-induced UPR. In our study, p58IPK expression activated PI3K/Akt but not ERK1/2 pathway during infection of CVB3, a positive stranded RNA virus.

To further explore whether p58IPK-mediated cell survival inhibits cell apoptosis and if so, what anti-apoptosis pathway it activates, we next examined the apoptosis pathways in CVB3-infected HeLa cells expressing different levels of p58IPK. As CVB3 infection has been documented to induce mitochondria-mediated apoptosis, we first determined if p58IPK expression could suppress cytochrome c release from mitochondria during CVB3 infection. This question was successfully addressed by Western blot analysis showing the decreased amount of cytochrome C in cytosolic fraction. Based on these findings we further determined the upregulation of another critical mitochondrial marker, mitofusin 2. Mitofusin 2 is an outer mitochondrial transmembrane protein responsible for mitochondria fusion. It is also located in ER where it regulates ER shape, tethers ER and mitochondria, and regulates mitochondrial uptake of Ca2+ released from the ER (de Brito and Scorrano, 2008). It has been reported that mitofusin 2 is upregulated during ER stress induced by pharmacological agents and that loss of mitofusin 2 sensitizes cells to ER stress-induced cell death (Ngoh et al., 2012). Based on this previous study, it is reasonable to speculate that mitofusin 2 likely has a similar function as p58IPK in protecting cell from injury. However, there is no such report thus far for this important notion. Here we show for the first time that mitofusin 2 is upregulated in p58IPK expressing cells at early phase of CVB3 infection. This upregulation of mitofusin 2 may benefit from the anti-PERK activity of p58IPK, which mediates the dephosphorylation of eIF2α to restore cellular protein translation. We also showed that this mitofusin 2 upregulation is in parallel with upregulation of Bcl-2, suppression of cytochrome c release from mitochondria and activation of caspases-9 and -3, indicating that the mitofusin 2 likely plays an important role in concert with p58IPK to counteract CVB3 induced cell death.

In summary, this study has found that p58IPK is a positive regulator of host defence mechanism during CVB3 infection. It enhances cell viability and inhibits CVB3-induced apoptosis at the early phase of infection through activation of PI3K/Akt survival pathway, which requires activation of upstream ER stress sensor ATF6a and subsequently upregulates expression of mitofusin 2 (Fig. 7). As persistent CVB3 infection causes downregulation of p58IPK and shifts the protective UPR to cell apoptosis, search for natural or chemically synthesized small molecule activators that can upregulate p58IPK may discover effective antiviral drugs to block CVB3-induced cell death during infections.

Figure 7.

Putative model of p58IPK action in inhibiting CVB3-induced apoptosis. Transient ER stress induced by CVB3 at early phase of infection activates PI3K/Akt survival pathway and inhibits virus-induced cell apoptosis, which requires activation of ATF6a-p58IPK signalling cascade and subsequent upregulation of mitochondrial membrane protein mitofusin 2. Prolonged ER stress caused by persistent CVB3 infection will downregulate p58IPK and lead to cell apoptosis.

Experimental procedures

Virus, cell culture, materials and transfection

CVB3 (CG strain) was obtained from Dr Charles Gauntt (University of Texas Health Center) routinely propagated in HeLa cells (ATCC). The virus stock was isolated from cells by three freeze-thaw cycles followed by centrifugation to remove cell debris and was stored at −80°C. Virus titres were determined by plaque assay as described previously (Zhang et al., 2002). Viral infection of cells was conducted at a multiplicity of infection (moi) of 10 or sham-infected with PBS. HeLa cells, an established cell line widely used to study CVB3 replication and cytopathic effect, were grown in DMEM supplemented with 100 μg ml−1 penicillin, 100 μg ml−1 streptomycin, 2 mM glutamine, and 10% fetal bovine serum (FBS) (Clontech). The HL-1 cell line, a cardiac muscle cell line established from a mouse atrial cardiomyocyte tumour lineage, was a gift from William C. Claycomb (Louisiana State University Health Science Center). The cells were maintained in Claycomb medium supplemented with 10% FBS (JRH Biosciences), 100 μg of penicillin-streptomycin ml−1, 0.1 mM norepinephrine (Sigma), and 2 mM l-glutamine (Invitrogen). LY294002 and U0126 inhibitors were purchased from Cell Signaling. siRNAs targeting human p58IPK were obtained from Qiagen. Transfection of siRNAs was performed under optimal conditions following the manufacturer's instructions (Gibco BRL). Briefly, 2 × 105 cells were grown at 37°C overnight. When the cells reached 60–70% confluence, they were washed with PBS and overlaid with transfection complexes containing siRNAs and Oligofectamine (Life technologies). After incubation for 6 h at 37°C, the mixture was replaced with DMEM supplemented with 10% FBS and incubation was continued for 2 days. The transfection of plasmid pcDNA1/neo-p58IPK (a gift from Dr Michael Katze, University of Washington), DN-ATF6a, pcDNA3.1-ATF6a (171–373) (a gift from Dr K. Moris, Kyoto University, Japan), Myc-His-tagged-DN-Akt(K179M) (Upstate) and HA-tagged constitutively activated Akt (HA-CA-Akt)(T308D S473D) (Addgene) was conducted by the same procedures as those for siRNAs described above except Lipofectamine (Invitrogen) was the transfection reagent.

Establishment of the stably transfected HeLa cell lines

The stable HeLa cell line expressing p58IPK or DN-ATF6a was established by plasmid transfection as per the manufacturer's instructions (Life Technologies). Briefly, 1 × 105 HeLa cells were transfected with pcDNA1/neo-p58IPK or pcDNA3.1-ATF6a (171–373) by the Lipofectamine method. After treatment with 8 μl of Lipofectamine and plasmid complexes for 6 h, the mixture was replaced with DMEM supplemented with 10% FBS and incubation was continued for 2 days before initiating selection. The stable cells were selected with G418 at a concentration of 600 μg ml−1. The resistant colonies were picked up 3 weeks later and further screened for p58IPK or ATF6a expression by Western blotting using a monoclonal anti-human p58IPK (Cell Signaling) and polyclonal anti-mouse ATF6a antibody (Santa Cruz) respectively. The colonies showing a high-level expression of p58IPK or ATF6a were selected. Cells transfected with the corresponding empty vectors were also screened by G418 in parallel for use as controls. Cell cultures prepared for detection of phosphorylation signals were serum-starved at least 2 h before performing the experiments.

Western blot analysis

Western blotting was performed by the standard protocol as previously described (Liu et al., 2008). Briefly, the cells were washed with cold PBS and resuspended in an appropriate volume of lysis buffer (0.025 M Tris-HCl, pH 8.0, 137 mM NaCl, 10% glycerol, 1 mM EDTA, 1 mM EGTA, 1% Triton X-100 and proteinase inhibitor cock tail). After incubation for 20 min on ice, the supernatant containing the proteins was collected by centrifugation at 13 000 g for 15 min at 4°C. For mitochondrial proteins isolation, the mitochondria isolation kit for cell culture (Thermo Science) was employed as per the manufacture's instructions. The isolated proteins were separated by 18% SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and transferred onto nitrocellulose membranes. The membranes were blocked with 5% skim milk in PBS and incubated with one of the following primary antibodies: monoclonal anti-phospho-Akt (Ser473), monoclonal anti-Akt, polyclonal anti-ERK1/2, polyclonal anti-phospho-ERK1/2, polyclonal anti-Bcl-2, polyclonal anti-phospho-eIF2α, polyclonal anti-cytochrome c, polyclonal anti-caspase-9, polyclonal anti-PARP and polyclonal anti-mitofusin-2 (Cell Signaling); monoclonal mouse anti-VP1 (Leica Microsystem); monoclonal anti-caspase-3, monoclonal anti-α-tubulin (Santa Cruz); polyclonal anti-Myc (Upstate); monoclonal anti-HA(Covance); polyclonal anti-COX IV (Abcam); monoclonal anti-β-actin (Sigma). After several washes with PBS, each blot was further incubated with an appropriate secondary antibody (goat anti-mouse or donkey anti-rabbit) conjugated to horseradish peroxidase (Amersham). Membranes were further developed by the enhanced chemiluminescence method as per the manufacturer's instructions (Amersham). Data were quantified by densitometric analysis of bands using NIH Image J software and normalized to the control.

Cell viability assay

Cell viability was analysed by using a 3-(4–5-dimethylthiazol-2-yl)-5-(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium salt (MTS) assay kit following the manufacturer's instructions (Promega). Briefly, after treatment with siRNAs and infected with CVB3 at a moi of 10 for 10 h, the cells were incubated with MTS solution for 2 h, and the absorbency of formazan was measured at 492 nm using an enzyme-linked immunosorbent assay (ELISA) plate reader. The absorbency of sham-infected/sham-treated cells was defined as the value of 100% survival (control), and the remaining data were converted to a percentage of the control.

Viral plaque assay

Virus titres were determined by plaque assay as described previously (Yuan et al., 2005). Briefly, HeLa cells were seeded into 6-well plates (8 × 105 cells/well) and incubated at 37°C for 20 h to a confluence of approximately 90% and then washed with PBS and overlaid with 1 ml of virus-containing samples serially diluted in cell culture medium. After a viral absorption period of 60 min, the supernatants were removed and the cells were overlaid with 2 ml of sterilized soft Bacto-agar-minimal essential medium, cultured at 37°C for 72 h, fixed with Carnoy's fixative for 30 min, and stained with 1% crystal violet. The plaques were counted and the amount of viral plaque-forming units (PFU) per ml was calculated.

Immunocytochemistry and confocal microscopy

p58IPK cells and vector cells proliferating on glass coverslips at ∼ 70% confluence were stained by following the method described previously (Sall et al., 2010). Briefly, the cells were fixed with 4% paraformaldehyde, permeabilized in 0.1% Triton X-100 for 10 min and stained with an anti-mitofusin 2 primary antibody for 1 h and then the slides were washed and treated with goat anti-rabbit IgG (H + L) labelled with red fluorescent Alexa Fluor 594 (Invitrogen, A11012). Nuclei were counterstained with 4',6'-diamidine-2'-phenylindole (DAPI) dihydrochloride. Cells were observed under a Leica SP2 AOBS confocal microscope.

Statistical analysis

The Student's t-test was employed to analyse data. The results are expressed as means ± standard deviations (SD) of three independent experiments. A P-value of less than 0.05 was considered statistically significant.

Acknowledgements

We would like to thank Dr Michael G. Katze (University of Washington, USA) and Dr Kazutoshi Mori, Kyoto University, Japan) for providing us the plasmid pcDNA1/neo-p58IPK and pcDNA3.1-AFT6 (171–373) respectively. This work was supported by grants from the Canadian Institutes of Health Research and Heart and Stroke Foundation of BC and Yukon. Dr Maged Hemida is a recipient of the CIHR-IMPACT and Heart and Stroke Foundation of Canada postdoctoral fellowship. Xin Ye is supported by a UGF Award from the University of British Columbia.

Ancillary