Cholesterol 25-Hydroxylase Inhibits Porcine Reproductive and Respiratory Syndrome Virus Replication through Enzyme Activity-Dependent and -Independent Mechanisms

ABSTRACT

Cholesterol 25-hydroxylase (CH25H) has recently been identified as a host restriction factor that exerts antiviral effects by catalyzing the production of 25-hydroxycholesterol (25HC). CH25H can be rapidly induced upon infection with some viruses. Porcine reproductive and respiratory syndrome virus (PRRSV), an arterivirus, has ranked among the most important swine pathogens since it was discovered in the late 1980s. In this study, we found that PRRSV infection significantly downregulated the expression of CH25H in cells by a so-far unknown mechanism, suggesting that CH25H exerts antiviral activity against PRRSV. Indeed, overexpression of CH25H inhibited PRRSV replication, whereas knockdown of CH25H by short interfering RNA (siRNA) promoted PRRSV infection. The anti-PRRSV effect of 25HC operates via inhibition of viral penetration. Interestingly, a CH25H mutant (CH25H-M) lacking hydroxylase activity still inhibited PRRSV infection. Screening using a yeast two-hybrid system followed by coimmunoprecipitation and immunofluorescence colocalization analyses confirmed that both CH25H and CH25H-M interact with the nonstructural protein 1 alpha (nsp1α) of PRRSV. Unexpectedly, the expression of nsp1α decreased following coexpression with CH25H or CH25H-M. Detailed analyses demonstrated that CH25H/CH25H-M could degrade nsp1α through the ubiquitin-proteasome pathway and that site K169 in the nsp1α protein is the key site of ubiquitination. Taken together, our findings demonstrate that CH25H restricts PRRSV replication by targeting viral penetration as well as degrading nsp1α, revealing a novel antiviral mechanism used by CH25H.

IMPORTANCE PRRSV has been a continuous threat to the global swine industry, and current vaccines are insufficient to provide sustainable control. CH25H has been found to exert a broad antiviral effect; thus, it is an attractive target for the development of anti-PRRSV drugs. Here, we demonstrate that CH25H is an interferon-stimulated gene that is highly expressed in porcine alveolar macrophages. CH25H exerts its anti-PRRSV effect not only via the production of 25HC to inhibit viral penetration but also by degrading viral protein through the ubiquitin-proteasome pathway, suggesting that CH25H is a candidate for the development of antiviral therapeutics. However, PRRSV infection appears to actively decrease CH25H expression to promote viral replication, highlighting the complex game between PRRSV and its host.

INTRODUCTION

Porcine reproductive and respiratory syndrome (PRRS) is one of the most economically significant swine diseases and is characterized by severe reproductive failure in sows and respiratory distress in piglets and growing pigs (1, 2). Since its emergence in the late 1980s, PRRS has been a continuous threat to the global swine industry, causing high economic losses (3, 4). The causative agent, PRRS virus (PRRSV), is a single-stranded positive-sense RNA virus classified within the order Nidovirales in the family Arteriviridae. The PRRSV genome is approximately 15 kb, encoding at least 10 open reading frames (ORFs) (5, 6). Approximately three-quarters of the genome is occupied by two ORFs, ORF1a and ORF1b, which encode viral replicase polyproteins pp1a and pp1ab that are cleaved into 14 nonstructural proteins (nsps; nsp1α to nsp12) by viral proteases (7). The other ORFs encode eight structural proteins: glycoprotein 2 (GP2), envelop (E) protein, GP3, GP4, GP5, ORF5a protein, matrix (M) protein, and nucleocapsid (N) protein (8–10). Although many commercial vaccines against PRRSV have been made available, unfortunately, neither traditional control strategies nor commercial vaccines provide sustainable control of PRRS (11–13). A major obstacle in the development of a successful PRRS vaccine is that the pathogenic mechanisms of PRRSV are poorly understood (14). A better understanding of the virus-host interactions in PRRSV infection will facilitate development of more effective control measures.

Metabolism and innate immunity are increasingly found to be interrelated, sharing resources and displaying cross-regulation. Sterol is an important part of the cell membrane and is essential for a wide range of cellular functions. Thus, elaborate homeostatic mechanisms have been developed to control sterol metabolism as well as to regulate the metabolism of these sterol enzymes at multiple levels within and outside the cell. Several of these mechanisms involve 25-hydroxycholesterol (25HC), a naturally occurring oxysterol that is synthesized by cholesterol 25-hydroxylase (CH25H) (15). CH25H is a multitransmembrane endoplasmic reticulum (ER)-associated enzyme whose main function is to catalyze excessive cholesterol to produce 25HC to reduce cholesterol accumulation, while 25HC can also suppress sterol biosynthesis by the regulation of nuclear receptors and sterol-responsive element binding protein (SREBP) activity (16, 17). Clusters of histidine residues are essential for CH25H catalytic activity that uses di-iron–oxygen as a cofactor to catalyze the hydroxylation of hydrophobic substrates.

It is known that viruses can alter cellular lipid metabolism to promote their own proliferation, and importantly, inhibition of cholesterol and fatty acid biosynthetic pathways has been shown to reduce viral replication, maturation, and secretion (18–22). CH25H has been identified as a broadly antiviral factor that inhibits the replication of many viruses, including human immunodeficiency virus (HIV) (23), vesicular stomatitis virus (VSV) (24), Ebola virus (EBOV) (24), hepatitis C virus (HCV) (25, 26), murine cytomegalovirus (MCMV) (27), Nipah virus (24), human herpesvirus (HSV) (28), influenza virus (27), and Zika virus (ZIKV) (29), by producing 25HC. At present, the role of CH25H during PRRSV infection is unclear.

In this study, we investigated the expression and antiviral role of CH25H during PRRSV infection and found that PRRSV infection downregulated CH25H expression; however, overexpression of CH25H significantly inhibited PRRSV replication. We further investigated the underlying mechanism(s) used by CH25H to inhibit PRRSV infection. We found that CH25H not only uses 25HC to inhibit PRRSV penetration but also degrades viral nonstructural protein nsp1α through the ubiquitin-proteasome pathway, revealing a novel mechanism by which CH25H restricts virus replication.

RESULTS

PRRSV infection downregulates CH25H expression.

Previous studies showed that CH25H expression can be rapidly induced by some viruses, such as HCV and ZIKV (25, 29). To assess the response of CH25H to PRRSV infection, mRNA levels of CH25H first were evaluated with different doses of PRRSV in primary porcine alveolar macrophages (PAMs), the target cells of PRRSV infection in vivo, PK-15^CD163 cells, a pig kidney cell line stably expressing PRRSV receptor CD163 (gifted from En-ming Zhou at Northwest A&F University, China), or MARC-145 cells, a monkey kidney cell line highly permissive for PRRSV infection used extensively in studies of PRRSV and vaccine production. As shown in Fig. 1A, CH25H mRNA was significantly reduced after PRRSV infection, compared with the control groups in PAMs or PK-15^CD163 cells, in a dose-dependent manner. It should be noted that only low mRNA expression of CH25H could be detected in MARC-145 cells (data not shown), and CH25H expression in PAMs was significantly higher than that in PK-15^CD163 cells (Fig. 1A). Subsequently, CH25H protein levels were also determined after PRRSV infection at different doses. Similar to the mRNA expression pattern, PRRSV infection significantly decreased CH25H protein expression in PAMs and PK-15^CD163 cells (Fig. 1B), and CH25H expression was barely detectable in MARC-145 cells (data not shown). These results indicated the different expression levels of CH25H in different cells (high expression in PAMs, intermediate expression in PK-15^CD163 cells, and low expression in MARC-145 cells). Regardless of cell type, PRRSV infection significantly decreased CH25H expression at both the mRNA and protein levels.

FIG 1.

CH25H is downregulated in PRRSV-infected cells. PAMs or PK-15^CD163 cells were infected with different doses of PRRSV (multiplicity of infection [MOI] of 0.2, 1.0, and 5.0). At 24 h postinfection (hpi), cells were harvested to detect CH25H mRNA expression by qRT-PCR (A) or protein expression by Western blot analysis (B). The results depicted in panel A represent the means and standard deviations from three independent experiments.

CH25H inhibits PRRSV replication.

To assess the effect of CH25H on PRRSV replication, two strategies were employed: overexpression and knockdown by specific short interfering RNA (siRNA). Considering the higher expression levels of CH25H in PAMs, we designed three siRNAs targeting porcine CH25H and examined their interference effects in PAMs by Western blot analysis. The three siRNAs exhibited varied knockdown effects on CH25H expression, with siRNA1 displaying the highest knockdown efficiency (Fig. 2A). In PAMs transfected with these siRNAs, PRRSV mRNA and protein (nsp2) were examined by quantitative reverse transcription-PCR (qRT-PCR) and Western blot analysis. As shown in Fig. 2B and C, knockdown of CH25H by siRNA promoted PRRSV replication and nsp2 expression to some degree compared with cells transfected with the negative-control siRNA. Fifty percent tissue culture infectious dose (TCID₅₀) assays also showed that knockdown of CH25H significantly increased virus titers (Fig. 2D).

FIG 2.

Knockdown of CH25H promotes PRRSV infection. (A) PAMs were transfected with CH25H-specific siRNAs or control siRNA (NC). At 36 h after transfection, cells were harvested to determine the knockdown efficiency by Western blot analysis. (B to D) PAMs were transfected with CH25H-specific siRNAs or control siRNA. At 24 h after transfection, cells were infected with PRRSV (MOI of 1.0). Cells were harvested at 24 hpi to analyze PRRSV mRNA by qRT-PCR (B), PRRSV nsp2 expression by Western blot analysis using nsp2-specific MAb as primary antibody (C), and virus titers by TCID₅₀ assays (D). The results represent the means and standard deviations from three independent experiments.

To further investigate the effect of CH25H knockdown on PRRSV replication, we tested whether ectopic expression of CH25H could inhibit PRRSV replication. The CH25H eukaryotic expression construct (pCAGGS-CH25H) was transiently transfected into MARC-145 cells. CH25H expression was barely detectable in MARC-145 cells transfected with empty vector, whereas high expression was observed in pCAGGS-CH25H-tranfected cells (data not shown). As expected, in MARC-145 cells overexpressing CH25H, PRRSV replication was significantly inhibited at all time points tested: 12, 24, and 36 h postinfection (hpi) (Fig. 3A and B). We further investigated whether overexpression of CH25H could inhibit PRRSV replication in PAMs. As a result of the lower plasmid transfection efficiency in PAMs, a recombinant lentivirus expressing CH25H-mCherry fusion protein was constructed. No evident cytotoxicity could be detected in PAMs transduced with recombinant lentivirus expressing CH25H-mCherry (data not shown). Similar to the results observed in MARC-145 cells, PRRSV replication and titers were also significantly decreased in CH25H lentivirus-transduced PAMs (Fig. 3C and D), as demonstrated by qRT-PCR and TCID₅₀ assays.

FIG 3.

CH25H suppresses PRRSV replication. (A and B) MARC-145 cells were transfected with CH25H expression plasmid or empty vector. At 24 h after transfection, cells were infected with PRRSV (MOI of 1.0), and then 24 h later cells were harvested for a TCID₅₀ assay (A) or qRT-PCR (B). (C and D) PAMs were transduced with recombinant lentivirus expressing CH25H-mCherry fusion protein or control lentivirus expressing mCherry. At 24 h after transduction, cells were infected with PRRSV (MOI of 1.0), and then 24 h later cells were harvested to determine PRRSV titers by TCID₅₀ assay (C) or PRRSV mRNA by qRT-PCR (D). (E) PK-15^CD163 cells pretreated with IFN-α (1,000 U/ml) were transfected with siRNA1 or control siRNA. At 24 h after transfection, cells were infected with PRRSV (MOI of 1.0), and 24 h later cells were harvested for plaque assay to determine virus titers. Data are expressed as means and standard deviations from three independent experiments.

Because ectopic expression of CH25H significantly inhibited PRRSV infection, we wanted to know to what extent CH25H contributes to PRRSV inhibition by interferon (IFN) pretreatment of target cells. To this end, IFN-α-pretreated PK-15^CD163 cells were transfected with siRNA1 or control siRNA and virus titers were determined by a plaque reduction assay. The results showed that knockdown of CH25H in PK-15^CD163 cells rescued PRRSV infection from blockade by IFN-α pretreatment (Fig. 3E). Based on the above-described results from knockdown and overexpression assays, we concluded that CH25H is a host restriction factor for PRRSV.

25HC inhibits PRRSV replication at the step of penetration.

CH25H is an ER-associated hydroxylase that can catalyze cholesterol to produce 25HC (15). Previous studies suggested that CH25H exerts antiviral activity by producing 25HC (24–29). To investigate whether CH25H inhibition of PRRSV replication is dependent on its enzymatic activity, we first tested the cytotoxicity of synthetic 25HC to PK-15^CD163 cells to determine the optimal concentration of 25HC by an 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay. Cytotoxicity could not be detected when 12.5 μM 25HC was used (Fig. 4A). PK-15^CD163 cells then were infected with PRRSV, followed by incubation with various concentrations of 25HC for 8 h. A TCID₅₀ assay was performed to determine the virus titers of the infected cells at the indicated times postinfection. The results showed that treatment with 25HC significantly inhibited PRRSV proliferation in a dose-dependent manner (Fig. 4B). To further confirm the reduction in virus titer stimulated by treatment with 25HC, we also tested the expression of PRRSV nsp2 by Western blotting and the corresponding viral mRNA by qRT-PCR. The results showed that mRNA expression of the PRRSV N gene was notably inhibited (Fig. 4C) and the expression of PRRSV nsp2 was also significantly reduced (Fig. 4D).

FIG 4.

25HC inhibits PRRSV replication. (A) Determination of the cytotoxicity of 25HC to PK-15^CD163 cells. PK-15^CD163 cells were incubated with various concentrations of 25HC or ethanol as a control for 48 h prior to the MTT assay. (B to D) PK-15^CD163 cells were pretreated with 25HC at the indicated concentrations for 8 h prior to PRRSV infection (MOI of 1.0). The infected cells were cultured in the presence of different concentrations of 25HC and harvested at 24 hpi for the TCID₅₀ assay (B) or at 24 and 36 hpi for qRT-PCR (C) and Western blot analysis (D). Western blot analysis was performed using a specific antibody against PRRSV nsp2 protein, and β-actin was used as a loading control. Data are expressed as means and standard deviations from three independent experiments.

To further explore the mechanism by which 25HC inhibits PRRSV infection, we first tested whether 25HC is directly virucidal and thereby kills PRRSV particles. As shown in Fig. 5A, 25HC treatment (12.5 μM) failed to directly inactivate PRRSV. Plaque reduction assays then were employed to determine the effect of 25HC on PRRSV on the adsorption, penetration, replication, and release steps of the viral replication cycle as previously described (30). The adsorption assay was performed in PK-15^CD163 cells. The results, depicted in Fig. 5B, showed that 25HC (12.5 μM) treatment prior to PRRSV infection did not significantly block virus attachment to PK-15^CD163 cells at any of the infection doses tested (20, 100, and 200 PFU/well), indicating that 25HC does not inhibit PRRSV attachment to cells. We next evaluated the effect of 25HC on PRRSV internalization. As shown in Fig. 5C, 25HC blocked PRRSV internalization and infectious virus titers were reduced by approximately 100-fold when 25HC was added during the internalization stage. PRRSV negative-sense RNA was quantified by qRT-PCR to examine the effect of 25HC on the PRRSV replication step. However, there was no significant difference in PRRSV RNA levels between cells treated with 25HC and cells treated with ethanol (Fig. 5D), indicating that 25HC does not affect the replication step of PRRSV. To determine the role of 25HC in PRRSV release, a virus-budding assay was conducted in PK-15^CD163 cells. The results showed that 25HC failed to inhibit the release of the virus (Fig. 5E). Taken together, these results indicated that 25HC reduces PRRSV infection through inhibiting viral penetration but has no effect on the attachment, replication, and release steps of the viral replication cycle.

FIG 5.

25HC inhibits PRRSV penetration. (A) Inactivated assay. PRRSV and 25HC were incubated at 37°C for 3 h and PK-15^CD163 cells cultured in six-well plates were prechilled at 4°C for 1 h, and then the media were replaced by a mixture of 25HC (12.5 μM) or ethanol (EtOH) and PRRSV (2, 20, and 200 PFU/well). After incubation at 4°C for another 2 h, cells were washed with precooled PBS, covered with overlay medium (1.8% [wt/vol] Bacto agar mixed 1:1 with 2× DMEM containing 0.05 mg/ml neutral red), incubated at 37°C for a further 72 h, and examined by a plaque assay. N.S., not significant. (B) Adsorption assay. PK-15^CD163 cells cultured in six-well plates were prechilled at 4°C for 1 h, and then the media were replaced by a mixture of 25HC (12.5 μM) or ethanol (dilution) and PRRSV (3, 30, and 300 PFU/well). After incubation at 4°C for another 2 h, the cells were washed with precooled PBS, and then cells were treated as described for panel A and plaques were counted directly. (C) Penetration assay. PK-15^CD163 cells cultured in six-well plates were prechilled at 4°C for 1 h and then incubated for another 2 h at 4°C with PRRSV at different doses (1, 10, and 100 PFU/well). The virus-containing medium was replaced by fresh medium containing 25HC (12.5 μM) or ethanol, and the temperature was shifted to 37°C for 3 h. Cells then were treated as described for panel A and plaques were counted directly. (D) Replication assay. PK-15^CD163 cells were incubated with PRRSV (MOI of 1.0) for 6 h, the cell-free virus particles were removed, and cells were cultured in fresh medium containing 25HC (12.5 μM). At 7, 8, 9, and 10 hpi, the infected cells were collected for qRT-PCR to detect the level of negative-sense PRRSV RNA. (E) Release assay. PK-15^CD163 cells were infected with PRRSV (MOI of 1.0). At 18 hpi, the inocula were replaced by fresh medium containing 25HC (12.5 μM). Culture media were harvested at 15, 30, 45, and 60 min after medium switching and titrated by a plaque assay. All results are means ± standard deviations from three independent experiments performed in triplicate.

CH25H mutant lacking hydroxylase activity (CH25H-M) still inhibits PRRSV infection.

CH25H is a member of a small family of molecules containing clustered histidines residues, which are critical for their hydroxylase activity (15). It is of interest to investigate whether the inhibitory effect of CH25H on PRRSV infection depends on its enzyme activity. To this end, a pair of histidine (H) codons at positions 242 and 243 in the porcine CH25H were mutated to glutamine (Q) codons by site-directed mutagenesis, generating the mutant form of CH25H (CH25H-M), which is unable to produce 25HC (31).

Overexpression of CH25H-M by lentivirus-mediated or transient transfection was used to investigate the antiviral effect of CH25H-M on PRRSV in PAMs and MARC-145 cells, respectively. Ectopic expression of CH25H-M significantly decreased the titer of PRRSV in both PAMs (Fig. 6A) and MARC-145 cells (Fig. 6B) at all of the tested time points after infection. We also used qRT-PCR and Western blotting to analyze the replication and protein expression of PRRSV after CH25H-M overexpression. Consistent with the results from TCID₅₀ assays, overexpression of CH25H-M significantly decreased PRRSV RNA levels (Fig. 6C and D) and nsp2 expression (Fig. 6E and F) at 12, 24, and 36 hpi. These results indicated that CH25H can also inhibit PRRSV infection via a hydroxylase-independent mechanism.

FIG 6.

Mutant forms of CH25H lacking hydroxylase activity can inhibit PRRSV replication. (A, C, and E) PAMs were transduced with recombinant lentivirus expressing CH25H-M-mCherry fusion protein or control lentivirus expressing mCherry, followed by PRRSV infection (MOI of 1.0). At the indicated time points after PRRSV infection, cells were harvested to determine PRRSV titers by a TCID₅₀ assay (A), PRRSV mRNA by qRT-PCR (C), or viral protein expression by Western blot analysis (E) using a specific antibody against PRRSV nsp2. (B, D, and F) MARC-145 cells were transfected with the CH25H-M expression construct or empty vector, followed by PRRSV infection (MOI of 1.0). Cells were collected at the indicated time points postinfection and then subjected to a TCID₅₀ assay (B), qRT-PCR (D), or Western blot analysis (F). All results are means ± SD from three independent experiments performed in triplicate.

CH25H targets nsp1α for ubiquitin-proteasome degradation.

To further investigate the mechanism by which CH25H-M inhibits PRRSV infection independently of the hydroxylase activity of CH25H, we studied the interactions between CH25H and PRRSV-encoded proteins using a yeast two-hybrid system. As shown in Fig. 7A, PRRSV nsp1, nsp1α, and nsp7β are possible proteins interacting with CH25H. To further examine the interactions between CH25H and the viral proteins selected by the yeast two-hybrid screening, Flag-tagged CH25H and hemagglutinin (HA)-tagged nsp1α and nsp7β expression plasmids were cotransfected into HEK-293T cells and coimmunoprecipitation (co-IP) experiments were performed. The results showed that only nsp1α interacted with CH25H (Fig. 7B and C). To detect whether this interaction relied on the enzymatic activity of CH25H, the interaction between CH25H-M and nsp1α was tested by co-IP. Interestingly, CH25H-M still interacted with nsp1α (Fig. 7D and E), suggesting that this interaction is independent of its enzymatic activity. To test whether CH25H/CH25H-M and nsp1α colocalized inside cells, an indirect immunofluorescence assay was carried out. The results of two-color immunofluorescent staining showed that CH25H/CH25H-M and nsp1α colocalized in the cytoplasm (Fig. 7F). To further investigate whether nsp1α interacts with CH25H in PRRSV-infected cells, a monoclonal antibody against nsp1α was used to co-IP endogenous CH25H from the lysates of PRRSV-infected cells. As shown in Fig. 7G, the interaction between nsp1α and CH25H could be detected in PRRSV-infected cells.

FIG 7.

CH25H selectively interacts with PRRSV nsp1α. (A) Screening the possible interactions by yeast two-hybrid system. The blue clones indicate the possible interactions. Yeast cells contain both pGBKT7-T and pGADT7-p53 or pGADT7, used as a positive and negative control, respectively. (B to E) HEK-293T cells were transfected with expression constructs encoding HA-tagged nsp1α and Flag-tagged CH25H or CH25H-M. The cells were lysed 32 h after transfection and subjected to immunoprecipitation with an anti-Flag antibody (B and D) or an anti-HA antibody (C and E). The whole-cell lysates (WCL) and immunoprecipitation (IP) complexes were analyzed by immunoblotting (IB) using anti-Flag, anti-HA, or anti-β-actin antibodies. (F) HEK-293T cells were transfected with Flag-tagged CH25H or Flag-tagged CH25H-M, together with HA-tagged nsp1α expression plasmid. At 32 h posttransfection, cells were fixed for an immunofluorescence assay to detect nsp1α protein (red) and CH25H or CH25H-M (green) with anti-HA and anti-Flag antibodies, respectively. Nuclei were counterstained with DAPI (blue). Fluorescent images were acquired with a confocal laser scanning microscope (FluoView ver. 3.1; Olympus, Tokyo, Japan). (G) PAMs were infected with PRRSV, and then cells were lysed at 32 h after infection and subjected to immunoprecipitation with a MAb against PRRSV nsp1α. The WCL and IP complexes were analyzed by immunoblotting using anti-CH25H, anti-nsp1α, or anti-β-actin antibody, respectively.

When the co-IP experiments were performed (Fig. 7B to E), we were surprised to find that the expression level of nsp1α was decreased on cotransfection with CH25H or CH25H-M, indicating that CH25H/CH25H-M degrades nsp1α. The ubiquitin-proteasome system, autophagy, and apoptosis are three major intracellular protein degradation pathways in eukaryotic cells (32, 33). To test whether nsp1α is indeed degraded and by which pathway, the proteasome inhibitor MG132 (20 μM), the autophagy inhibitors LY294002 (1 mM) and 3-methyladenine (3-MA; 5 mM), and the apoptosis inhibitor Z-VAD-FMK (10 μM) were added to cells cotransfected with CH25H-M and nsp1α expression constructs. We found that only MG132 abolished the degradation of nsp1α, whereas the degradation of nsp1α was not affected by treatment with LY294004, 3-MA, or Z-VAD-FMK (Fig. 8A). To further confirm this observation, a dose-dependent experiment was performed. As shown in Fig. 8B, the degradation of nsp1α gradually increased with the increase in CH25H-M, whereas nsp1α degradation was abolished with MG132 treatment. We also tested whether nsp1α can be degraded by 25HC and found that 25HC did not degrade nsp1α (Fig. 8C). In addition, we demonstrated that CH25H-M and 25HC did not degrade nsp1β (Fig. 8C), another nonstructural protein of PRRSV, indicating that degradation of nsp1α by CH25H-M is specific.

FIG 8.

CH25H-M degrades nsp1α through ubiquitin-proteasome pathway. (A) PK-15^CD163 cells were cotransfected with expression constructs encoding HA-tagged nsp1α and Flag-tagged CH25H-M. At 16 h after transfection, cells were treated with MG132 (20 μM), Z-VAD-FMK (10 μM), 3-MA (5 mM), or LY294004 (1 mM) for 10 h. Cell lysates were prepared and analyzed by Western blotting to detect the expression of nsp1α and CH25H-M. (B) PK-15^CD163 cells were cotransfected with HA-tagged nsp1α and different concentrations of Flag-tagged CH25H-M expression plasmids. Cell lysates were prepared and analyzed by Western blotting at 30 h posttransfection. Cells treated with MG132 (20 μM) were used as controls. (C) PK-15^CD163 cells were transfected with expression constructs encoding Flag-tagged CH25H-M and HA-tagged nsp1α or nsp1β. Cell lysates were prepared and analyzed by Western blotting at 16 h posttransfection. Cells transfected with HA-tagged nsp1α expression plasmid and with 25HC added were used as controls. (D) HEK-293T cells were transfected with HA-tagged nsp1α and Flag-tagged CH25H-M expression plasmids. The cells were lysed 32 h after transfection and subjected to immunoprecipitation with anti-HA antibody. The whole-cell lysates (WCL) and immunoprecipitation (IP) complexes were analyzed by immunoblotting using anti-Flag, anti-HA, anti-Ub, or anti-β-actin antibodies. (E) HEK-293T cells were transfected with expression constructs encoding Flag-tagged CH25H-M and HA-tagged wild-type nsp1α or its mutants (K117A, K150A, and K169A). At 32 h after transfection, cells were collected for IP with an anti-Flag antibody. The WCL and IP complexes were analyzed by immunoblotting with anti-Flag, anti-HA, anti-K48-Ub, or anti-β-actin antibodies. (F) A mixture of MG132 (20 μM) and PRRSV (MOI of 1.0) was added to PK-15^CD153 cells pretreated with IFN-α (1000 U/ml), and 24 h later cells were harvested for plaque assay. All results are means ± SD from three independent experiments performed in triplicate.

Ubiquitin is a small-molecule polypeptide consisting of 76 amino acid residues that can be linked to the lysine residue of a protein in a covalent manner. Three possible ubiquitination sites, at lysine (K) residues 117, 150, and 169 in nsp1α, were predicted using the relevant software (http://www.ubpred.org/ and http://bdmpub.biocuckoo.org/prediction.php/), and K169 was the most probable site based on the obtained score. To confirm the ubiquitination sites in nsp1α, these three lysine (K) codons were mutated to alanine (A) codons by site-directed mutagenesis. When CH25H-M was cotransfected with the individual nsp1α mutant, we found that nsp1α K169A rescued the degradation of nsp1α by CH25H-M, whereas the other two mutants failed to do so (Fig. 8D and E). Therefore, these data collectively demonstrated that CH25H-M inhibits PRRSV replication by degrading nsp1α selectively and K169 is the ubiquitination site. In view of the fact that knockdown of CH25H could rescue PRRSV infection from blockade by IFN-α pretreatment of target cells (Fig. 3E), we also found that blockade of proteasomal degradation by MG132 partially rescues PRRSV infection from blockade by IFN-α pretreatment (Fig. 8F).

CH25H can be induced in PAM cells after IFN-α treatment.

Previous studies showed that mouse ch25h (CH25H in humans and ch25h in mice) is an interferon-stimulated gene (ISG) and that its expression can be induced by type I or type II IFN treatment in murine dendritic cells and macrophages (34). However, a recent study found that IFN-α treatment failed to induce the expression of human CH25H in human monocyte-derived macrophages, primary human hepatocytes, and Huh7 and A549 cells (26). To examine whether porcine CH25H can be induced by IFNs in PAMs, PAMs were stimulated with porcine IFN-α or Sendai virus, followed by detection of mRNA expression of CH25H at different points after stimulation/infection. The mRNA expression of ISG56, a classical ISG, was also detected from the same samples in parallel. As shown in Fig. 9, CH25H as well as ISG56 could be induced in PAMs upon IFN-α treatment. Similar results were observed in Sendai virus-infected PAMs. However, the fold induction of ISG56 was higher than that of CH25H, and IFN-α was a more effective CH25H inducer than Sendai virus. Together, these results demonstrated that porcine CH25H could be induced in PAMs after IFN-α treatment and that porcine CH25H is an ISG in PAMs.

FIG 9.

Porcine CH25H is an interferon-stimulated gene in PAMs. PAMs were treated with porcine IFN-α (1,000 U/ml) or infected with Sendai virus (MOI of 0.5) for 6, 8, or 10 h, and then cells were harvested for qRT-PCR to detect CH25H mRNA expression. All results are the means ± SD from three independent experiments performed in triplicate.

DISCUSSION

In recent years, identification of host antiviral restriction factors has gained increased attention, which will help us to understand the defense mechanisms against virus infection and will aid the development of novel antiviral drugs and vaccines (35–37). As a very successful virus, PRRSV has devastated the swine industry worldwide for nearly 30 years and has shown no sign of slowing down (38). Recent efforts have aimed at identifying anti-PRRSV host restriction factors and further characterizing their mechanisms of action (39–44). In this study, we demonstrated that CH25H can inhibit PRRSV infection by producing 25HC, a mechanism similar to that reported previously for other viruses. Furthermore, we identified another novel antiviral mechanism employed by CH25H to inhibit PRRSV infection that is independent of its enzyme activity and involves degradation of PRRSV nsp1α.

The antiviral effects of CH25H have been investigated for some viruses, and these studies suggested that CH25H has broad-spectrum antiviral effects that operate through the production of 25HC (29, 33, 45, 46). 25HC exerts its antiviral effects by multiple mechanisms. Because 25HC is a naturally occurring secreted oxysterol that can permeate membranes, it can block fusion of the viral envelope with cellular membranes, a critical step in the invasion of enveloped viruses (45). For example, 25HC restricts HCV replication through blockage via membranous web formation (25); 25HC directly modifies the cellular membrane to impede HIV entry but does not affect the transcription, translation, or budding processes of HIV (24); and 25HC blocks ZIKV internalization in a dose-dependent manner (29). Owing to its potent inhibitory effects on the maturation of SREBP2, a factor that controls sterol biosynthesis, 25HC was also reported to inhibit HCV infection at a postentry step by repressing SREBP2 activation (26). In addition, other antiviral mechanisms used by 25HC have been reported. For example, 25HC inhibits Lassa virus infection through aberrant GP1 glycosylation (47). Significant antiviral activities of 25HC have also been reported against nonenveloped viruses, such as human papillomavirus-16, human rotavirus, and human rhinovirus (46). In our present study, we elucidated that 25HC restricts PRRSV infection through suppressing viral penetration, which is similar to the inhibition mechanism of 25HC against HIV.

Besides its enzyme activity-dependent antiviral mechanism, a recent study demonstrated that the CH25H mutant (CH25H-M) lacking enzymatic activity could suppress HCV replication by inhibiting the formation of the HCV NS5A dimer (31), indicating that CH25H also inhibits viral infection by a 25HC-independent mechanism. In this study, we also found that CH25H-M interacts with PRRSV nsp1α. Overexpression of CH25H-M degrades nsp1α through the ubiquitin-proteasome pathway. PRRSV nsp1α is essential for viral replication and contributes to viral pathogenicity (48, 49). CH25H/CH25H-M-mediated nsp1α degradation therefore will affect viral replication and pathogenicity. We also identified K169 as the ubiquitination site of nsp1α, and this site was found to be highly conserved among different PRRSV isolates.

Previous studies have suggested that the expression levels of CH25H at steady state are low to undetectable in most cells and tissues (15, 34). By detecting the basal expression levels of CH25H in different cells through qRT-PCR and Western blotting, we found that the endogenous expression of CH25H is weak in MARC-145 and PK-15^CD163 cells, whereas it is significantly higher in PAMs, the target cells of PRRSV in the natural host. Evidence that CH25H exhibits anti-PRRSV effects and the highly abundant expression of CH25H in PAMs suggests that CH25H plays an important role in the host's defense against PRRSV infection.

There is a difference in the expression of CH25H in response to different viral infections. For example, CH25H is induced in response to VSV (24), ZIKV (29), HCV (26), and MCMV (27) infection; on the contrary, herpes simplex virus 1 (HSV-1) infection reduces the expression of CH25H (50). In this study, PRRSV infection significantly reduces CH25H expression in all three of the cell lines tested. Our results are consistent with a previous study in which the downregulated expression of CH25H was observed in PRRSV-infected PAMs by RNA sequence analysis (51). The mechanism by which PRRSV infection downregulates CH25H expression is unknown but is an issue worth studying. CH25H is a 31.6-kDa endoplasmic reticulum (ER)-associated glycoprotein encoded by an intronless gene (15). The molecular mechanisms that control the expression of CH25H have yet to be defined. Previous studies revealed that macrophages are the main cellular source of CH25H, and its transcript is selectively induced by TLR ligands (52). Park and Scott demonstrated that TLR-mediated mouse ch25h expression depends on TRIF, the production of type I IFN, and signaling through the IFNR/JAK/STAT1 pathway (34). Furthermore, consensus IFN-γ activation-site elements and IFN-stimulated response elements are located upstream of the ch25h ORF in the mouse genome (34). These observations suggested that mouse ch25h is a classical ISG; however, whether human CH25H is a classical ISG remains controversial (25–27, 53, 54). Our data showed that porcine CH25H can be defined as an ISG because it was greatly induced by IFN-α in PAMs (as shown in Fig. 9), which was similar to the results of mouse studies in macrophages and dendritic cells. Previous studies have revealed that PRRSV infection inhibits the host's IFN responses (55–61), and at least six proteins encoded by PRRSV have been identified as IFN antagonists: nsp1α, nsp1β, nsp2, nsp4, nsp11, and N protein (62–70). Whether these IFN antagonists contribute to PRRSV-induced CH25H downregulation is the subject of future studies in our laboratory.

Another interesting issue is the biological significance of the PRRSV-mediated decrease in CH25H. 25HC may regulate cholesterol homeostasis through its ability to activate live X receptors (LXRs) (71, 72). Oxysterol-mediated activation of LXR has been demonstrated to promote macrophage survival and immune regulation functions (73, 74). It is well known that PRRSV infection impairs the functions of PAMs. Whether the decrease in CH25H expression mediated by PRRSV in PAMs is associated with the impaired function of PAMs deserves further investigation.

In summary, we demonstrate, for the first time, that CH25H possesses antiviral activity against PRRSV, supporting the previous notion that CH25H operates as a broad-spectrum host antiviral restriction factor via the production of 25HC. We also identified a novel 25HC-independent antiviral mechanism of CH25H that involves binding and degrading PRRSV nsp1α through the ubiquitin-proteasome pathway. These findings indicate that CH25H harbors at least two methods, one 25HC dependent and the other 25HC independent, to suppress PRRSV infection. Further studies are warranted to more precisely elucidate the molecular mechanisms and significance of PRRSV-induced downregulation of CH25H in PAMs.

MATERIALS AND METHODS

Cell culture and viruses.

HEK-293T, PK-15^CD163, and MARC-145 cells were cultured in Dulbecco's modified Eagle's medium (Invitrogen, USA) containing 10% fetal bovine serum (FBS) at 37°C with 5% CO₂ in a humidified incubator. PAMs were cultured in RPMI 1640 (Sigma) supplemented with 10% heat-inactivated FBS at 37°C in a humidified 5% CO₂ incubator. PRRSV strain WUH3 is a highly pathogenic type 2 (North American) PRRSV, which was isolated from the brains of pigs suffering from high-fever syndrome in China at the end of 2006 (75). PRRSV was amplified, and the titer was determined in MARC-145 cells. Sendai virus was obtained from the Centre of Virus Resource and Information, Wuhan Institute of Virology, Chinese Academy of Sciences.

Plasmids.

Expression plasmids for wild-type porcine CH25H (amino acids [aa] 1 to 271) were constructed by PCR amplification of the cDNA from PK-15^CD163 cells and were cloned into pCAGGS-Flag vector to generate Flag-tagged expression plasmid pCAGGS-Flag-CH25H. Histidine codons at positions 242 and 243 of wild-type CH25H were converted to glutamine codons by site-directed mutagenesis to create the mutant form (CH25H-M) deficient in catalytic activity. The yeast screening plasmids for CH25H- or PRRSV-encoded proteins were constructed by cloning CH25H or PRRSV genes into the yeast vectors pGADT7 and pGBKT7, respectively. Full-length wild-type PRRSV nsp1α (aa 1 to 180) was cloned into pCAGGS-HA to generate an HA-tagged expression construct by standard molecular cloning techniques. Lysine codons at positions 117, 150, and 169 of nsp1α were converted to alanine codons by site-directed mutagenesis.

Reagents and antibodies.

25HC was purchased from Sigma-Aldrich (California, USA), reconstituted in ethanol at a concentration of 5,000 μM/liter, and stored at −20°C. Mouse or rabbit monoclonal antibodies (MAbs) against FLAG, HA, or β-actin were purchased from MBL. Rabbit polyclonal antibodies against CH25H were purchased from Santa Cruz Biotechnology (California, USA). Rabbit polyclonal antibodies against K48-Ub were purchased from Abclone (Wuhan, China). MAbs directed against PRRSV nsp1α and nsp2 were produced as described previously (76, 77).

Lentivirus packaging.

Lentiviral expression plasmids pLvx-mCherry-CH25H and pLvx-mCherry-CH25H-M were generated by cloning the cDNA of CH25H or CH25H-M into lentiviral vector pLvx-mCherry (TaKaRa, Japan). Lentivirus-assisted plasmids pLP1, pLP2, and pLP-VSV-G were kindly provided by Xing Liu at the Institute of Veterinary Medicine, Jiangsu Academy of Agricultural Sciences. To package the recombinant lentiviruses, HEK-293T cells were cotransfected with lentivirus-assisted plasmids pLP1, pLP2, pLP-VSV-G, and pLvx-mCherry-CH25H or pLvx-mCherry-CH25H-M using Lipofectamine 3000 (Invitrogen) according to the manufacturer's instructions. The supernatant was harvested at 48 h after cotransfection and stored at −70°C. Two milliliters of complete nutrient solution then was added to the original wells. The supernatant was harvested after 24 h and then mixed with the first harvested supernatant, followed by high-speed centrifugation (13,000 × g) for 4 h. The pellets were suspended in serum-free Dulbecco's modified Eagle's medium (DMEM) and stored at −70°C until use.

Cytotoxicity assay.

Different concentrations of 25HC were added to PK-15^CD163 cells and incubated for 48 h. Cell viability then was determined by an MTT assay.

Viral adsorption, penetration, replication, and release assays.

Adsorption, penetration, replication, and release assays of PRRSV were performed using PK-15^CD163 cells as described previously (30).

Immunofluorescence staining.

For immunofluorescence staining, cells were fixed with 4% paraformaldehyde and then permeabilized with methanol for 15 min prior to the addition of primary antibody (anti-mouse HA and anti-rabbit Flag) and incubation for 1 h at room temperature. Cells then were washed in Tris-buffered saline-Tween 20, followed by incubation with Alexa Fluor 488-conjugated donkey anti-rabbit or 594-conjugated donkey anti-mouse antibodies. Nuclei were stained with 4′,6-diamidino-2-phenylindole (DAPI). Fluorescent images were acquired with a confocal laser scanning microscope (Fluoview ver. 3.1; Olympus, Tokyo, Japan).

RNA extraction and quantitative real-time PCR.

Total RNA was extracted from cultured cells with TRIzol reagent (Invitrogen). RNA was then reverse transcribed into cDNA by reverse transcriptase (TaKaRa, Japan). Quantitative real-time PCR (qPCR) experiments were performed in triplicate. Relative quantification of mRNA expression levels was normalized to that of β-actin. Absolute quantitative mRNA levels were calculated using standard curves. Real-time PCR was performed using Power SYBR green PCR master mix (Applied Biosystems) in an ABI 7500 real-time PCR system (Applied Biosystems).

Coimmunoprecipitation.

Cells were collected and lysed 30 to 36 h after transfection of expression plasmids using 50 mM Tris-HCl, pH 8.0, 150 mM NaCl, and 1% Triton X-100. For immunoprecipitation, lysates were incubated with the appropriate antibodies for 4 h on ice, followed by precipitation with protein A+G agarose beads (Beyotime, China). Samples were separated by SDS-PAGE and transferred to polyvinylidene difluoride (PVDF) membranes (Millipore, USA). After blocking in phosphate-buffered saline (PBS) containing 0.1% Tween 20 and 5% skim milk, the blots were probed with the indicated antibodies. Western blot visualization was performed with enhanced chemiluminescence.

Western blot analysis.

Cells were cultured in 60-mm dishes and harvested with lysis buffer (Beyotime, China). The samples were then resolved by SDS-PAGE and transferred to PVDF membranes to determine protein expression. Cells were also treated with a broad caspase inhibitor (Z-VAD-FMK; Beyotime, China) or a proteasome inhibitor (MG132; Beyotime, China) at final concentrations of 10 μM and 20 μM, respectively, or an autophagy inhibitor (3-MA and LY294002; Sigma, USA) at final concentrations of 5 mM and 1 mM, respectively. The expression of nsp1α, nsp1β, and CH25H-M was analyzed using an anti-mouse HA and anti-mouse Flag antibody. PRRSV nsp2 antibody was used to detect virus replication. K48-Ub antibody was used to identify the ubiquitination site.

TCID₅₀ assay for PRRSV.

Briefly, MARC-145, PK-15^CD163, or PAM cells were seeded in 96-well plates and then infected with serial 10-fold dilutions of PRRSV samples in eight replicates. The plates were incubated for 72 to 96 h before virus titers were calculated. PRRSV titers were expressed as TCID₅₀ per milliliter using the Reed-Muench method.

Statistical analyses.

GraphPad Prism 5 software (GraphPad Software, San Diego, CA, USA) was used for data analysis using a two-tailed unpaired t test. Differences between groups were considered statistically significant when the P value was less than 0.05 (*, P = 0.05; **, P = 0.01; ***, P = 0.001).