Porcine Circovirus 2 Activates the PERK-Reactive Oxygen Species Axis To Induce p53 Phosphorylation with Subsequent Cell Cycle Arrest at S Phase in Favor of Its Replication

ABSTRACT

Porcine circovirus type 2 (PCV2), the causative agent of porcine circovirus-associated diseases (PCVAD), is known to induce oxidative stress, activate p53 with induction of cell cycle arrest, and trigger the PERK (protein kinase R-like endoplasmic reticulum kinase) branch of the endoplasmic reticulum (ER) stress pathway. All these cellular responses could enhance PCV2 replication. However, it remains unknown whether PERK activation by PCV2 is involved in p53 signaling with subsequent changes of cell cycle. Here, we demonstrate that PCV2 infection induced cell cycle arrest at S phase to favor its replication via the PERK-reactive oxygen species (ROS)-p53 nexus. PCV2 infection promoted phosphorylation of p53 (p-p53) at Ser15 in porcine alveolar macrophages. Inhibition of PERK by RNA silencing downregulated total p53 (t-p53) and p-p53. Treatment with the MDM2 inhibitor nutlin-3 led to partial recovery of t-p53 in perk-silenced and PCV2-infected cells. perk silencing markedly downregulated ROS production. Scavenging of ROS with N-acetylcysteine (NAC) of PCV2-infected cells downregulated t-p53 and p-p53. Increased accumulation of p-p53 in the nuclei during PCV2 infection could be downregulated by silencing of perk or NAC treatment. Further studies showed that perk silencing or NAC treatment alleviated S phase accumulation and downregulated cyclins E1 and A2 in PCV2-infected cells. These findings indicate that the PCV2-activated PERK-ROS axis promotes p-p53 and contributes to cell cycle accumulation at S phase when more cellular enzymes are available to favor viral DNA synthesis. Overall, our study provides a novel insight into the mechanism how PCV2 manipulates the host PERK-ROS-p53 signaling nexus to benefit its own replication via cell cycle arrest.

IMPORTANCE Coinfections or noninfectious triggers have long been considered to potentiate PCV2 infection, leading to manifestation of PCVAD. The triggering mechanisms remain largely unknown. Recent studies have revealed that PERK-mediated ER stress, oxidative stress, and cell cycle arrest during PCV2 infection are conducive to viral replication. However, how PCV2 employs such host cell responses requires further research. Here, we provide a novel mechanism of PCV2-induced ER stress and enhanced viral replication: the PCV2-activated PERK-ROS-p53 nexus increases S phase cell population, a cell cycle period of DNA synthesis favorable for PCV2 replication. The fact that PCV2 deploys the simple ROS molecules to activate p53 to benefit its replication provides novel insights into the triggering factors, that is, certain stimuli or management measures that induce ER stress with subsequent generation of ROS would exacerbate PCVAD. Use of antioxidants is justified on farms where PCVAD is severe.

INTRODUCTION

Porcine circovirus type 2 (PCV2), a small nonenveloped DNA virus belonging to the genus Circovirus of the Circoviridae family (1), is considered the main pathogen of porcine circovirus-associated diseases (PCVAD), including postweaning multisystemic wasting syndrome (PMWS), dermatitis and nephropathy syndrome, respiratory disease syndrome, and congenital tremor in piglets (2, 3). There are two main open reading frames (ORFs) in the PCV2 genome: ORF1, encoding the replicase protein that is involved in viral replication (4), and ORF2, coding for the capsid protein (Cap) with high immunogenicity (5). Several studies have also indicated that there are ORF3, ORF4, and ORF5, which exert different functions during PCV2 infection (6–8). PCV2 infection can induce a variety of cellular responses, including endoplasmic reticulum (ER) stress, oxidative stress, autophagy, cell cycle arrest, and apoptosis (9–13). These cellular responses play important roles in the game between virus and host and might function as double-edged swords for the host cells because ER stress, oxidative stress, and autophagy are found to favor PCV2 proliferation (9–11).

Tumor suppressor protein 53 (p53) is an important transcription factor of diverse functions (14, 15). When activated by various stress signals, p53 translocates into the nucleus and regulates expression of a wide variety of genes involved in apoptosis, cell cycle progression, differentiation, and DNA repair (16–18). Recent studies have suggested that some viruses could hijack p53 functions for their replication. For example, the nucleocapsid protein of porcine epidemic diarrhea virus (PEDV) could interact with p53 to induce cell cycle arrest in S phase to promote viral replication (19). Transmissible gastroenteritis virus (TGEV) infection increased phosphorylation of p53 at serine 15 (Ser15) and Ser20 with subsequent cell cycle arrest, thereby enhancing its replication (20). Xu et al. showed that p53 signaling mediated cell cycle accumulation in S phase in favor of PCV2 replication that could be suppressed by deletion of the p53 gene (12). However, how PCV2 induces p53 signaling in favor of its replication remains an enigma.

As the center of protein synthesis and processing, the ER plays an important role in virus infection. Our previous study revealed that PCV2 infection selectively activates the stress sensor PERK (protein kinase R-like endoplasmic reticulum kinase) in the ER and deploys the PERK pathway to enhance its replication (9). Chen et al. reported that PCV2 infection induces oxidative stress as evidenced with substant accumulation of reactive oxygen species (ROS) in infected PK-15 cells, while N-acetylcysteine (NAC) treatment could attenuate ROS generation with decreased virus replication (10). Our recent report showed that ROS originating from PCV2-induced ER stress could promote translocation of nuclear HMGB1 into the cytoplasm, which derepresses HMGB1-mediated inhibition of viral DNA replication in the nuclei to benefit PCV2 proliferation (21). These studies indicate that PCV2 could activate ER stress with subsequent oxidative stress to manipulate the downstream targets in favor of its replication.

However, we wondered whether p53 signaling during PCV2 infection is mediated by PERK activation and oxidative stress to enhance virus replication. Here, we report that the PCV2-activated PERK-ROS axis was involved in enhanced p53 phosphorylation at Ser15 and its nuclear migration, leading to cell cycle arrest at S phase and increased viral replication.

RESULTS

PCV2 infection enhanced p53 phosphorylation.

PCV2 infection was reported to increase the expression of p53 in PK-15 cells (6) and induce cell cycle arrest via p53 signaling to facilitate its replication (12). Several studies have revealed that p53 phosphorylation at serine 15 (Ser15) is involved in the regulation of cell cycle (22–24). While immunosuppression is one of the main features of PCVAD, immune stimulation plays an important role in the development of this disease (25). The PAM cells not only are important immune cells but also are critical target cells of PCV2 infection (26). To examine whether PCV2 could affect p53 level and induce its phosphorylation at Ser15 in PAM cells, cells at different time points from 12 to 48 h postinfection (hpi) were used for Western blotting. Figure 1 shows that there was a time-dependent increase of p-p53 (phosphorylated form at Ser15) with the progression of PCV2 infection, while t-p53 (total p53/β-actin) remained constant from 24 hpi in both PCV2- and mock-infected cells (Fig. 1A).

FIG 1.

Porcine circovirus type 2 infection promoted p53 phosphorylation at serine 15 in both PAM and PK-15 cells. PAM or PK-15 cells were mock infected or infected with PCV2 or its ΔORF3 mutant at an MOI of 1. (A and B) PAM cells were infected with PCV2 for the indicated times (A) or with ΔORF3 for 36 h (B) for Western blotting with antibodies to t-p53, p-p53 (Ser15), and PCV2 Cap. β-Actin was used as a loading control. (C and D) Ratio of t-p53 to β-actin or p-p53 to t-p53 as shown in panel B. (E) Western blotting of t-p53, p-p53, Cap, and β-actin in the whole-cell lysates of PK-15 cells infected with PCV2 or its mutant ΔORF3. (F and G) Ratio of t-p53 to β-actin or p-p53 to t-p53 as shown in panel E. The images in panels B and E are representative of three independent experiments. The bar charts in panels C, D, F, and G show the means ± SD of three independent experiments. Statistical significance was analyzed by Student t test (*, P < 0.05; **, P < 0.01).

We thought that changes of p53 might be related to cell type. Also, an early study reported that the PCV2 ORF3 protein could interact with porcine ubiquitin E3 ligase Pirh2 and upregulate cellular p53 levels (27). Therefore, we analyzed the protein levels, by Western blotting, of both t-p53 and p-p53 in PAM and PK-15 cells infected for 36 h with either PCV2 or its ΔORF3 mutant strain. We found that infection with both viral strains decreased t-p53 in PAM cells but increased t-p53 in PK-15 cells compared with mock-infected control cells (Fig. 1B, C, E, and F). However, PCV2 or ΔORF3 infection significantly increased p-p53 in both PAM and PK-15 cells compared with control cells (Fig. 1B, D, E, and G). There were no obvious differences of t-p53 and p-p53 levels between the ΔORF3 mutant and its parental virus in either cell type (Fig. 1B to G). These results indicate that PCV2 infection enhanced p53 phosphorylation at Ser15 in both cell types, a phosphorylated form active in cell cycle regulation, and that the ORF3 protein did not affect phosphorylation at this particular serine residue. Because alveolar macrophages are one of the main cell types of PCV2 infection and are more immune responsive, PAM cells were used in the following experiments.

Downregulation of PERK in PCV2-infected cells reduced both total p53 and phosphorylated p53 at Ser15.

ER stress was reported to activate p53 signaling (28, 29). To examine whether ER stress induced by PCV2 infection is also involved in p53 signaling, we investigated the effect of perk-specific RNA interference (shPERK) on total p53 and its phosphorylation at Ser15. We found that shPERK was effective in downregulating PERK expression and resulted in reduced PCV2 replication shown as decreased Cap level and virus titers (Fig. 2A to C), consistent with our previous findings (9). Downregulation of PERK did not affect the ratio of p-p53 to t-p53 (Fig. 2D) but did decrease the ratio of t-p53 to β-actin, with corresponding reduction of the ratio of p-p53 to β-actin in PCV2-infected cells (Fig. 2E and F). To further explore if downregulation of total p53 by perk silencing is related to reduced transcription, relative quantitative reverse transcription-PCR (qRT-PCR) was performed to detect the p53 mRNA level. Figure 2G indicates that downregulation of PERK did not have a negative effect on p53 transcription.

FIG 2.

Silencing of perk by RNA interference in PCV2-infected cells downregulated both total and phosphorylated p53. PAM cells were first transfected with perk-specific RNA interference plasmid (shPERK) or control plasmid (shNC) for 12 h and then infected with PCV2 (MOI = 1) for 36 h. (A) Western blotting was used to analyze the protein levels of t-PERK, p-PERK, t-p53, p-p53 (Ser15), and Cap in PCV2-infected cells with β-actin as the loading control. The images are representative of three independent experiments. (B) Ratios of t-PERK and Cap to β-actin from panel A. (C) Virus titers between shPERK group and shNC group shown as log₁₀ TCID₅₀. (D to F) Ratios of p-p53 to t-p53 (D), t-p53 to β-actin (E), and p-p53 to β-actin (F) from panel A. (G) Total RNA was extracted from PCV2-infected PAM cells with or without perk silencing for qRT-PCR to examine the p53 mRNA level. The bar charts in panels B to G represent the means ± SD of three independent experiments. Statistical significance was analyzed by Student t test (ns, not significant; *, P < 0.05; **, P < 0.01).

Reduced total p53 in PCV2-infected and perk-silenced cells could be partially recovered by nutlin-3 treatment.

Because downregulation of PERK reduced total p53 without affecting its transcription, we speculated that reduced total p53 could be due to proteolytic degradation. MDM2 acts as E3 ligase that ubiquitinates p53 for proteasome degradation (30, 31). We used nutlin-3, an MDM2 inhibitor, to inhibit the interaction between p53 and MDM2 to see if it affects the p53 level. Western blotting revealed that nutlin-3 treatment led to a significant increase of total p53, but without affecting its phosphorylation at Ser15, in the PCV2-infected cells with perk silencing compared to corresponding cells without nutlin-3 (Fig. 3A and B). Apparently, this would result in significant reduction of the p-p53/t-p53 ratio (Fig. 3C). However, PCV2 replication, shown as the Cap expression level, was dependent on both PERK activation and p53 phosphorylation, but not on total p53 because the Cap protein level was still low in perk-silenced cells irrespective of increased total p53 as a result of nutlin-3 treatment (Fig. 3A). These findings, together with those in the preceding sections, suggest that activation of PERK signaling by PCV2 infection has double effects on p53, increased stability and upregulated phosphorylation at Ser15, and that activation of both PERK and p53 signaling is involved in enhanced PCV2 replication.

FIG 3.

Nutlin-3 treatment could reverse reduction of total p53 due to perk silencing in PCV2-infected PAM cells. After transfection with shPERK or shNC vector for 12 h, PAM cells were infected with PCV2 and then incubated in the medium containing nutlin-3 for 36 h. (A) Western blotting of t-PERK, p-PERK, t-p53, p-p53, and Cap in whole-cell lysates with β-actin as a loading control. The gels are representative of three independent experiments. (B and C) Ratios of t-p53 to β-actin (B) and p-p53 to t-p53 (C) from panel A. The bar charts in panels B and C are the means ± SD of three independent experiments. Statistical significance was analyzed by Student t test (*, P < 0.05; **, P < 0.01).

PCV2 induced p53 phosphorylation at Ser15 and its nuclear translocation via the PERK-ROS axis.

Our recent study has demonstrated that PCV2 infection increases expression of ERO1α via PERK activation, with subsequent ROS generation and nucleocytoplasmic migration of HMGB1 to enhance its replication (21). Intracellular ROS also activates p53 (32). Therefore, we wondered whether p53 activation in PCV2-infected cells may also be mediated by the PERK-ROS axis. The PAM cells receiving perk silencing or N-acetylcysteine as a ROS scavenger were collected to examine changes of the total and phosphorylated forms of p53 and PERK by Western blotting and of intracellular ROS by flow cytometry. We found that the perk silencing or NAC treatment decreased both the total and phosphorylated form of p53 without affecting the ratio of p-p53 to t-p53 in PAM cells infected with PCV2 (Fig. 4A and B) and reduced PCV2 Cap protein level (Fig. 4A and C). Flow cytometry revealed that perk silencing markedly inhibited ROS production to a degree similar to that of NAC treatment (Fig. 4D and E). It is noteworthy that NAC treatment of the PCV2-infected PAM cells did not affect t-PERK or p-PERK but downregulated both t-p53 and p-p53 to an extent close to that of perk silencing (Fig. 4A to C). Similar results were obtained in PK-15 cells (Fig. 5). These findings indicate that PCV2-induced phosphorylation of p53 at Ser15 was mediated by ROS as a result of PERK activation during viral infection.

FIG 4.

Scavenging of ROS with N-acetylcysteine (NAC) of PCV2-infected PAM cells downregulated t-p53 and p-p53 similarly to perk silencing without affecting t-PERK and p-PERK. PAM cells were transfected with shPERK or shNC vector for 12 h and then infected with PCV2 at 37°C and 5% CO₂ for 36 h. For NAC treatment, the PAM cells were infected with PCV2 and then cultured in fresh medium containing NAC for 36 h. (A) Protein levels of t-PERK, p-PERK, t-p53, p-p53, Cap, and β-actin in PCV2-infected PAM cells with the above treatments were analyzed by Western blotting. (B) Ratios of p-p53 or t-p53 to β-actin and p-p53 to t-p53 from panel A. (C) Ratios of p-PERK to t-PERK and of t-PERK or Cap to β-actin from panel A. (D) PCV2-infected PAM cells were collected and incubated with 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA). Intracellular ROS levels were assessed by flow cytometry. Relative ROS levels in PCV2-infected PAM cells are shown as DCFH-DA fluorescence intensity with the mock-infected cells set at 100%. (E) Mean values of DCFH-DA fluorescence intensity from one typical experiment. The bar charts in panels B, C, and D show the means ± SD of three independent experiments from panels A and E. Statistical significance was analyzed by Student t test (*, P < 0.05; **, P < 0.01).

FIG 5.

Removal of ROS with N-acetylcysteine of PCV2-infected PK-15 cells decreased t-p53 and p-p53 similarly to perk silencing without affecting t-PERK and p-PERK. PK-15 cells were transfected with shPERK or shNC vector for 12 h and then infected with PCV2 at 37°C and 5% CO₂ for 36 h. For NAC treatment, the PCV2-infected cells were cultured in fresh medium containing NAC for 36 h. (A) Protein levels of t-PERK, p-PERK, t-p53, p-p53, Cap, and β-actin (as a loading control) in PCV2-infected cells with the above-mentioned treatments were analyzed by Western blotting. (B and C) Ratios of p-p53 or t-p53 to β-actin and p-p53 to t-p53 (B) and p-PERK to t-PERK and t-PERK or Cap to β-actin (C). (D) PCV2-infected cells were collected and incubated with DCFH-DA. Intracellular ROS levels were assessed by flow cytometry. Relative ROS levels in PCV2-infected cells are shown with the mock-infected cells set at 100%. The images in panels A are representative of three independent experiments. The bar charts in panels B, C, and D show the means ± SD of three independent experiments. Statistical significance was analyzed by Student t test (*, P < 0.05; **, P < 0.01).

Phosphorylated p53 is known to translocate to the nucleus as a transcription factor to regulate the downstream target genes (33, 34). We wondered if distribution of phosphorylated p53 is changed during PCV2 infection and how PERK signaling or elevated ROS impacts the subcellular localization of p-p53. Confocal microscopic imaging was used to examine such effects in PCV2-infected PAM cells with perk silencing or NAC treatment. Figure 6 shows that PCV2 infection enhanced p-p53 distribution within the nuclei compared with mock-infected cells (Fig. 6A). However, perk silencing and ROS inhibition by NAC led to apparent downregulation of p-p53, with reduced accumulation in the nuclei (Fig. 6A). Quantification of p-p53 positive signals in PCV2-infected cells also confirmed reduction of p53 accumulation in the nuclei upon perk silencing or NAC treatment (Fig. 6B and C). Taken together, our data demonstrate that PCV2 infection activated p53 and facilitated its nuclear translocation via the PERK-ROS axis.

FIG 6.

PCV2 infection enhanced p-p53 distribution within the nuclei by activating PERK and ROS production. PAM cells were transfected with shPERK or shNC vector for 12 h and then infected with PCV2 at 37°C and 5% CO₂ for 36 h. For N-acetylcysteine treatment, the PAM cells were infected with PCV2 and then cultured in fresh medium containing NAC for 36 h. The PAM cells were fixed and incubated with antibodies against p-p53 (green) and Cap (red). (A and B) Distribution of p-p53 detected by confocal microscopy. (C) Mean values of p-p53 fluorescence intensity (FL) in at least 30 PCV2-infected PAM cells analyzed by ImageJ software. The bar chart in panel C shows the means ± SD of three independent experiments from panels A and B. Statistical significance was analyzed by Student t test (*, P < 0.05).

Activation of the PERK-ROS axis during PCV2 infection induced cell cycle arrest at the S phase.

PCV2 infection was reported to induce S phase arrest through p53 signaling to promote its replication (12). Our results described above demonstrate that the PERK-ROS axis was involved in p53 activation. To further investigate the role of the PERK-ROS pathway in the cell cycle, we analyzed the ratio of cells in different phases and cell cycle-related proteins in PCV2-infected cells with perk silencing or NAC treatment. Flow cytometry analysis showed that PCV2 infection significantly increased the population of cells at the S phase compared with mock-infected cells (Fig. 7A and B). Downregulation of PERK by RNA interference in PCV2-infected cells or treatment with NAC markedly reduced the S phase cell population, with an increased cell proportion at the G₀/G₁ phase (Fig. 7A and B). To further confirm the observed effects of the PERK-ROS axis on cell cycle arrest, cyclins E1 and A2, involved in regulation of the cell cycle by promoting S phase entry and progression, were examined by Western blotting. We found that PCV2 infection led to significant increases of cyclin E1, cyclin A2, and ROS which could be downregulated by perk silencing or NAC treatment alone (Fig. 7C to F). These findings indicate that the PCV2-activated PERK-ROS axis contribute to cell cycle transition from G₀/G₁ to S phase when cellular enzymes are expressed to favor porcine circovirus DNA synthesis (35).

FIG 7.

PCV2 induced cell cycle arrest at S phase and upregulated cyclins E1 and A2, which could be counteracted by perk silencing or scavenging of ROS with N-acetylcysteine. PCV2-infected PAM cells with shPERK vector transfection or N-acetylcysteine treatment were collected for cell cycle analysis and relative protein detection. (A) Cells were stained with PI/RNase staining buffer, and cell cycle analysis was performed by flow cytometry. (B) Percentages of the cells at each phase were analyzed by ModFit LT. (C) Western blotting of S phase-related proteins cyclins E1 and A2 in whole-cell lysates. (D and E) Ratios of cyclin E1 to β-actin (D) and of cyclin A2 to β-actin (E). (F) Relative intracellular ROS level in each cell group with the mock-infected cells set at 100%. The results in panels A and C are representative of three independent experiments. The bar charts in panels B, D, E, and F are the means ± SD of three independent experiments. Statistical significance was analyzed by Student t test (*, P < 0.05; **, P < 0.01).

DISCUSSION

PCV2 is able to induce diverse cell stress responses, such as PERK-mediated ER stress, oxidative stress, autophagy, and p53-dependent cell cycle arrest, all of which are conducive to its replication (9–12). It is of great significance to understand how PCV2 takes advantage of such host responses to benefit its replication. Our recent work has revealed that HMGB1 within the nuclei is inhibitory to PCV2 DNA replication, while the virus can drive part of the nuclear HMGB1 into cytosol by ROS generated via ER stress-ERO1 signaling to favor its replication (21). We wondered if PERK activation by PCV2 is involved in p53 signaling and cell cycle changes and if there are mechanisms other than ROS-mediated nucleocytoplasmic HMGB1 translocation that might contribute to enhanced PCV2 replication. By downregulation of PERK or NAC treatment, we clearly demonstrated that PCV2 infection promotes cell cycle transition into the S phase via activation of the PERK-ROS-p53 nexus. Because PCV2 DNA replication is dependent on cellular enzymes expressed in the S phase (35), it is intriguing that PCV2 simply takes advantage of the simple ROS molecules to promote its replication by inducing PERK-mediated ER stress and oxidative stress (Fig. 8).

FIG 8.

Proposed model of the PCV2-activated PERK-ROS-p53 nexus in inducing S phase accumulation of the cell cycle to benefit viral replication. PCV2 infection induces PERK-mediated ER stress and oxidative stress, shown as increased generation of ROS, which activate p53 via Ser15 phosphorylation. PERK activation is required for maintaining p53 stability and its phosphorylation. The PERK-ROS-p53 nexus signaling during PCV2 infection leads to accumulation of the cell populations at the S phase when a multienzyme complex including DNA polymerase is initiated not only for duplication of host DNA but also for synthesis of viral DNA (black circles with or without protrusions) to benefit viral replication. Inhibition of PERK by RNA interference or scavenging of ROS by N-acetylcysteine treatment results in decreased phosphorylation of p53 at Ser15, with reduced nuclear translocation as well as reduced accumulation of cells in the S phase, unfavorable for PCV2 replication.

As one of the centers of intracellular signaling in response to stresses, p53 activity is closely regulated. In unstressed cells, p53 is usually maintained at a low level due to ubiquitination and subsequent degradation (36). A variety of stimulations, including DNA damage, oncogene activation, and nutrient fluctuation, could activate the p53 pathway to induce expression of a large number of downstream target genes, either to support cell survival or to mediate cell death (37, 38). Studies in recent years have shown that p53 also plays an important role in virus infection (39, 40). Different viruses may use distinct strategies to manipulate p53 and downstream pathways, resulting in different cell fates. For example, West Nile virus induces p53-mediated apoptosis through sequestration of HDM2 to the nucleoli (41). Orf virus utilizes its VIR protein to antagonize p53-mediated antiviral effects by degrading p53 (42). Porcine parvovirus, TGEV, and hepatitis B virus infections promote p53 phosphorylation to induce apoptosis (43), cell cycle arrest (44), or even p53 reactivation (45). Here, we report that PCV2 infection significantly increased p53 phosphorylation at Ser15 in both PAM and PK-15 cells (Fig. 1), suggesting that PCV2 infection activates p53 signaling. This is consistent with a previous report that PCV2 infection induces the cellular DNA damage response and subsequent p53 phosphorylation (46). However, we advance the understanding of the mechanism PCV2 employs to benefit its replication in the infected cells, by revealing that the oxidative stress as a result of PERK-mediated ER stress is responsible for p53 activation and subsequent accumulation of the S phase cells favorable for viral DNA replication.

Our previous work found that PCV2 infection selectively activates the PERK-eIF2α (α subunit of eukaryotic initiation factor 2) arm of ER stress and utilizes this pathway to facilitate its replication (9). ER stress is known to prevents p53 stabilization and p53-mediated apoptosis upon DNA damage in cells treated with either of the pharmacological inducers tunicamycin and thapsigargin (TG) (47). A recent report indicates that prolonged ER stress promotes apoptosis via a p53-dependent inhibition of BiP expression (48). However, there are few studies on whether ER stress caused by virus infection is involved in regulation of p53 activity. We show that activation of PERK signaling by PCV2 infection is involved in p53 stabilization and phosphorylation at Ser15, while perk silencing of PCV2-infected cells reduced total and phosphorylated form of p53. This seems to contradict the report by Baltzis et al., who showed that total p53 is increased in perk gene knockout human sarcoma HT1080 cells treated with TG for 4 h (49). This could be due to the difference in PERK activation by PCV2 infection and chemical induction. PCV2 infection is usually of long duration and readily induces PERK activation that could upregulate ERO1α expression, leading to ER-sourced ROS (21) that could activate p53 as shown in this study. In addition, TG induces perturbation of endoplasmic reticulum homeostasis by inhibiting SERCA (sarco/endoplasmic reticulum Ca²⁺ ATPase) and exerts multiple effects on cellular responses, including both PERK and IRE1 branches of ER stress, leading to cell apoptosis (50). In a study by Jeong et al., TG treatment of the human colorectal carcinoma cells was found to increase total p53 over extended period of time from 6 to 24 h (51). It is clear that the effect of ER stress on the fate of p53 is dependent on the type of inducers, duration, and other factors.

By gene silencing of perk in PCV2-infected PAM cells, we found that both p-p53 and t-p53 were reduced (Fig. 2A, E, and F). Phosphorylation of p53 at Ser15 is known to hinder binding of the E3 ubiquitin ligase MDM2, thus inhibiting p53 ubiquitination and proteasomal degradation (30, 31). We speculated that decreased phosphorylation of p53 by perk silencing might increase ubiquitination of p53 and subsequent degradation. Thus, nutlin-3 was used to inhibit MDM2-dependent p53 degradation (52). As shown in Fig. 3A and B, treatment with nutlin-3 of PCV2-infected cells with perk-silencing increased total p53, but without affecting its phosphorylated form, while nutlin-3 treatment of the PCV2-infected cells transfected with scrambled control RNA did not affect total and phosphorylated p53, both at similar levels to the PCV2-infected cells transfected with control (shNC) vector only. These findings suggest that PERK activation induced by PCV2 is involved in p53 stability and its activation and that downregulation of PERK promotes p53 degradation, which could be partly rescued by inhibition of p53 and MDM2 interaction with nutlin-3.

Intracellular ROS could be produced in the ER during protein folding (53, 54). ROS is well known in modulation of p53 activities either upstream or downstream (32). Saha et al. found that ROS could trigger apoptosis by inducing the p38 mitogen-activated protein kinase (MAPK) pathway to promote entry of phosphorylated p53 into the nucleus (55). ROS also regulates p53 activity as an upstream molecule in PEDV-induced apoptosis (56). The fact that perk silencing led to effective reduction of ROS in PAM and PK-15 cells upregulated by PCV2 infection (Fig. 4D and Fig. 5D) reinforced our recent finding that the ER is the source of ROS due to PCV2-mediated activation of the PERK-ERO1α axis (21). To further explore how PCV2-induced ER stress participated in p53 signaling, we examined the effects of perk silencing or NAC treatment on p53. We found that both total and phosphorylated forms of p53 at Ser15 were decreased in PCV2-infected cells subjected to either treatment (Fig. 4 and 5). Clearly, PCV2 activates the p53 pathway through the PERK-ROS axis.

In the viral life cycle, some viruses exploit certain specific phases of the cell cycle in favor of their own replication (57). Cyclins and cyclin-dependent kinases (CDKs) are fundamental molecules in cell cycle regulation. Cyclins A and E are mainly involved in transition of the cells from G₁ to S phase, while cyclin A is maintained at high levels and interacts with CDK2 during S phase (58). In the cell cycle, p53 activates the CDK inhibitor p21, which binds and inhibits complexes of cyclins with CDKs, thus repressing cell cycle progression during G₁ and S phases (58). Phosphorylation of p53 at Ser15 is involved in regulation of cell cycle in response to stresses, including virus infection (22–24). Some porcine viruses have been found to affect the cell cycle by activating p53. For example, classical swine fever virus induces cell cycle arrest at G₁ phase by mediating p53 phosphorylation at Ser15, with upregulation of p21 and downregulation of cyclin E (33). PEDV induces cell cycle arrest in the G₀/G₁ phase by activating p53 at Ser15 and p21, with subsequent upregulation of cyclin A (59). However, it remains unknown if PERK-ROS-induced p53 activation by PCV2 is linked to changes of the cell cycle. Our study shows that PCV2 infection increased nuclear localization of phosphorylated p53 and induced S phase accumulation with increased cyclin E1 and cyclin A2, while these effects could be effectively downregulated by perk silencing or NAC treatment (Fig. 6A to C and Fig. 7A to E), indicating that the PCV2-induced PERK-ROS-p53 nexus is involved in accumulation of the S phase cells. However, Xu et al. reported that PCV2 infection induces S phase arrest through the p53 pathway with increased cyclin E but decreased cyclin A (12). Such a discrepancy, increased cyclin A2 versus decreased cyclin A during PCV2 infection, may be due to the difference in cell states: a heterogenous cell population was used in this study, in contrast to the chemically synchronized cells used by Xu et al. (12). Theoretically, increased cyclin E, when combined with CDK2, could lead to activation of the E2F proteins, with subsequent expression of its target genes, including cyclin A (60).

In conclusion, our study provides a novel mechanism of PCV2-induced ER stress and enhanced viral replication: the PCV2-activated PERK-ROS-p53 nexus increases accumulation of the S phase, a cell cycle period favorable for PCV2 replication. This is a new addition to our recent report that PCV2 induces PERK activation and ER-sourced ROS that benefit viral replication through decreasing HMGB1 within the nuclei and derepressing its inhibition of viral DNA replication via enhancing translocation of HMGB1 out of the nuclei (21). Because oxidative stress is involved in both mechanisms of enhanced PCV2 replication via p53 and HMGB1, in farm situations where PCVAD is severe, use of antioxidants is justified to reduce the disease burden. Because PCV2 also induces mitophagy (61), a cellular process that may also generate ROS, further research is warranted to examine the relative contributions of the endoplasmic reticulum and mitochondria to increased generation of ROS that could be utilized by PCV2 to enhance its replication in infected cells.

MATERIALS AND METHODS

Cell lines and viruses.

Porcine alveolar macrophage line 3D4/31 (PAM) and a porcine kidney epithelial cell line (PK-15) free of PCV1 were preserved in our laboratory. PAM cells were cultured at 37°C with 5% CO₂ in RPMI 1640 (Thermo Fisher Scientific, MA, USA) supplemented with 10% fetal bovine serum (FBS; HyClone, UT, USA) and 1% penicillin-streptomycin (Thermo Fisher Scientific). PK-15 cells were grown in Dulbecco’s modified Eagle’s medium (DMEM; Thermo Fisher Scientific) supplemented with 4% FBS and 1% penicillin-streptomycin at 37°C and 5% CO₂. PCV2 strain YW (PCV2b; GenBank accession no. MG245866) was originally isolated from a pig with naturally occurring PCVAD in Zhejiang, China, and propagated in PK-15 cells. The ΔORF3 strain is an isogenic mutant of strain YW with inactivation of the ORF3 initiation codon. The virus titers of PCV2 strain YW and its ΔORF3 mutant were 1 × 10^5.6 50% tissue culture infective doses (TCID₅₀)/mL and 1 × 10^5.1 TCID₅₀/mL, respectively.

Reagents and antibodies.

N-Acetylcysteine (NAC; Beyotime, Shanghai, China) and nutlin-3 (Selleck Chemicals LLC, Houston, TX, USA) were used at the designated concentrations in appropriate experiments. Antibodies against total PERK (t-PERK), p-PERK, total p53 (t-p53), cyclin E1, and cyclin A2 were purchased from Proteintech (Rosemont, IL, USA). Antibodies against p-p53 (Ser15) and β-actin were obtained from Cell Signaling Technology (MA, USA). Mouse monoclonal anti-Cap antibody was produced in our laboratory. Horseradish peroxidase (HRP)-conjugated goat anti-mouse IgG and HRP-conjugated goat anti-rabbit IgG for Western blotting were purchased from Sigma-Aldrich (St. Louis, MO, USA). Alexa Fluor 555-conjugated donkey anti-mouse IgG and Alexa Fluor 488-conjugated donkey anti-rabbit IgG for confocal imaging were obtained from Thermo Fisher.

Virus infection and chemical treatments.

PAM or PK-15 cells grown to approximately 60% to 70% confluence were infected with PCV2 or its ΔORF3 mutant strain at a multiplicity of infection (MOI) of 1.0. After 2 h of absorption, the cell monolayers were rinsed with Hanks’ balanced salt solution (HBSS) to remove unattached viruses and then incubated in the presence of fresh medium containing 4% or 8% FBS at 37°C and 5% CO₂. Mock-infected cells were included as controls. To investigate the effects of oxidative stress and MDM2 inhibitors on intracellular ROS, cell cycle, p53, and relevant molecules, PAM or PK-15 cells were first infected with PCV2 and then cultured at 37°C and 5% CO₂ in fresh medium containing NAC (5 or 10 mM) or nutlin-3 (10 μM) for 36 h. Cells without chemicals were included as a control. According to the requirements of individual experiments, the infected and/or treated cells were collected at the indicated time points for the following assays.

RNA interference and transfection.

RNA interference of the perk (EIF2AK3) gene of Sus scrofa was based on the sequence (5′-GCAGTCATCAGTTAGAATTTC-3′) synthesized by GenePharma (Shanghai, China) and inserted into the pGPU6-shPERK vector. A scrambled sequence (5′-TTCTCCGAACGTGTCACGT-3′) was introduced into the vector (as pGPU6-shNC) as the negative control. To examine the effects of PCV2-induced PERK activation on p53 expression and distribution, intracellular ROS, and changes of cell cycle-relevant molecules, PAM or PK-15 cells were transfected with short hairpin RNA (shRNA) plasmids using ExFect transfection reagent (Vazyme, Nanjing, China) according to the manufacturer’s instructions. An equivalent volume of the transfection reagent was added to the corresponding control wells. After transfection for 12 h, the cells were infected with PCV2 and collected for analysis as described below.

Quantitative reverse transcription-PCR.

Total RNA of the cells was extracted with an RNA extraction kit (Bioteke, Beijing, China) and used as the templates for cDNA synthesis by the HiScript II Q RT SuperMix for qPCR (+DNA wiper) (Vazyme, Nanjing, China) according to the manufacturer’s instructions. The specific primers were used to detect the transcription of p53 (forward primer 5′-GCCACTGGATGGCGAGTATT-3′ and reverse primer 5′-TCCAAGGCGTCATTCAGCTC-3′) (62). The gapdh gene was used as an internal control (forward primer 5′-GCAAGTTCCACGGCACAGTCAA-3′ and reverse primer 5′-TCTCGCTCCTGGAAGATGGTGATG-3′). Quantitative reverse transcription-PCR (qRT-PCR) was performed by using the ChamQ universal SYBR qPCR master mix (Vazyme) and the Mx3005P qPCR system (Agilent Technologies, CA, USA). Relative expression was calculated using the threshold cycle (2^−ΔΔCT) method (63).

Cell lysis and Western blotting.

The cells in culture plates were washed twice with phosphate-buffered saline (PBS; pH 7.2; Beyotime), then treated with cell lysis buffer (Beyotime) supplemented with protease inhibitor cocktail (Roche, Mannheim, Germany), and scraped from the culture plate. The cell lysates were centrifuged at 4°C and 14,000 × g for 10 min, and the supernatant samples were collected. Protein concentrations were determined using the Bradford protein assay kit (Beyotime). Equal amounts of protein samples were separated on 10% or 12% SDS-PAGE gels and transferred to polyvinylidene fluoride membranes (Merck Millipore, MA, USA). After blocking for 1 h at 37°C in TBST buffer (20 mM Tris, 150 mM NaCl, and 0.05% Tween 20 [pH 7.6]) containing 5% nonfat milk powder, the membranes were incubated with different primary antibodies overnight at 4°C. Then the membranes were washed with TBST and incubated in HRP-conjugated secondary antibodies at 37°C for 1 h. Immunoreactive bands on the membranes were visualized by using an ECL kit (Cyanagen, Italy), and images were captured by the Gel 3100 chemiluminescent imaging system (Sagecreation, Beijing, China). Quantification of protein blots was performed by ImageJ software (National Institutes of Health, MD, USA). The ratios of target proteins to reference proteins were used for relative quantitative purposes.

Virus titration.

The effect of PERK knockdown on PCV2 replication was examined by virus titration using a conventional virology protocol (11).

Confocal fluorescence microscopy.

Confocal fluorescence microscopy was used to visualize the subcellular localization and quantify the fluorescence intensity of p-p53 induced by PCV2 infection, transfection, or chemical treatments. Specifically, PAM cells grown on coverslips (Cellvis, Mountain View, CA, USA) were infected with PCV2, transfected with shRNA plasmids or treated with NAC. After infection for 36 h, cells were washed twice with PBS, and 4% paraformaldehyde was added to fix cells for 20 min at room temperature. The cell samples were then subjected to Triton X-100 (0.1%) for permeabilization and blocked in PBS containing 5% goat serum for 15 min at room temperature to prevent nonspecific binding. After incubation with corresponding primary antibodies overnight at 4°C, the cells were washed three times, followed by labeling with Alexa Fluor-conjugated secondary antibodies at 37°C for 1 h. The cell nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI; Beyotime). The cells were observed with a confocal laser scanning microscope (IX81-FV1000; Olympus, Japan), and the images were captured. The fluorescence intensity of p-p53 in PCV2-infected cells was quantified by using ImageJ software.

Measurement of intracellular ROS.

Intracellular ROS were determined as described previously (64). In brief, the cells were collected after infection or treatment and incubated with 20 μM 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA; Sigma-Aldrich) at 37°C for 30 min in the dark. The cells were then washed and resuspended in HBSS for flow cytometric analysis by using a BD FACSCalibur flow cytometer (BD Biosciences, NJ, USA). FlowJo software version X (Tree Star, Inc., OR, USA) was used to analyze the data. The intracellular ROS levels are shown as relative fluorescence intensity with the mock-infected cells set at 100%.

Cell cycle analysis.

For analysis of cell cycles, PAM cells were collected and fixed with 70% cold ethanol overnight at −20°C. After sufficient washing with HBSS, the cells were incubated with propidium iodide (PI)/RNase staining buffer (BD Biosciences, San Diego, CA, USA) at room temperature for 15 min in the dark. Flow cytometric analysis was performed on the BD FACSCalibur flow cytometer, and data were analyzed using ModFit LT (Topsham, ME, USA).

Statistical analysis.

Statistical analyses were performed by using GraphPad Prism version 6.0 (GraphPad Software, San Diego, CA, USA). The two-tailed Student t test was used for comparisons between two groups or treatments. All data are expressed as the means ± standard deviations of three independent experiments. Differences were regarded as statistically significant at P values of <0.05 and highly significant at P values of <0.01.