SARS coronavirus papain-like protease induces Egr-1-dependent up-regulation of TGF-β1 via ROS/p38 MAPK/STAT3 pathway

SARS coronavirus (SARS-CoV) papain-like protease (PLpro) has been identified in TGF-β1 up-regulation in human promonocytes (Proteomics 2012, 12: 3193-205). This study investigates the mechanisms of SARS-CoV PLpro-induced TGF-β1 promoter activation in human lung epithelial cells and mouse models. SARS-CoV PLpro dose- and time-dependently up-regulates TGF-β1 and vimentin in A549 cells. Dual luciferase reporter assays with TGF-β1 promoter plasmids indicated that TGF-β1 promoter region between −175 to −60, the Egr-1 binding site, was responsible for TGF-β1 promoter activation induced by SARS-CoV PLpro. Subcellular localization analysis of transcription factors showed PLpro triggering nuclear translocation of Egr-1, but not NF-κB and Sp-1. Meanwhile, Egr-1 silencing by siRNA significantly reduced PLpro-induced up-regulation of TGF-β1, TSP-1 and pro-fibrotic genes. Furthermore, the inhibitors for ROS (YCG063), p38 MAPK (SB203580), and STAT3 (Stattic) revealed ROS/p38 MAPK/STAT3 pathway involving in Egr-1 dependent activation of TGF-β1 promoter induced by PLpro. In a mouse model with a direct pulmonary injection, PLpro stimulated macrophage infiltration into lung, up-regulating Egr-1, TSP-1, TGF-β1 and vimentin expression in lung tissues. The results revealed that SARS-CoV PLpro significantly triggered Egr-1 dependent activation of TGF-β1 promoter via ROS/p38 MAPK/STAT3 pathway, correlating with up-regulation of pro-fibrotic responses in vitro and in vivo.

via the degradation of ERK1 12 . In addition, we also indicated that SARS-CoV PLpro significantly triggers the up-regulation of Transforming growth factor beta 1 (TGF-β 1) and pro-fibrotic genes via ubiquitin proteasome, p38 MAPK, and ERK1/2-mediated signaling pathways in human promonocytes 13 . Therefore, PLpro modulates the innate immune response as well as involve in the pathogenesis of SARS-CoV-induced pulmonary fibrosis.

Materials and Methods
Cell culture and transient transfection with pSARS-PLpro. Human alveolar basal epithelial A549 cells were grown in Dulbecco's Modified Eagle's Medium (DMEM; HyClone Laboratories, Logan, Utah, USA) with 100 U/mL of penicillin and streptomycin, 2 mM L-glutamine, and 10% fetal bovine serum (FBS; Biological Industries, Kibbutz Beit Haemek, Israel). SARS-CoV PLpro gene, nt 4507-5840 of the SARS-CoV TW1 strain (GenBank accession no. AY291451) was amplified by RT-PCR, and then cloned into expression vector pcDNA3.1/ His C (Invitrogen), as described in our prior reports 12,13 . The empty vector pcDNA3.1 or pSARS-PLpro at the concentrations of 0, 0.5, 1, 2, 5, and 10 μ g/ml was transfected into A549 cells with Arrest-In transfection reagent (Thermo scientific). After 5-h incubation, transfected cells were maintained in DMEM medium containing 20% FBS. Transient expression of recombinant PLpro in A549 cells 2 days post transfection was analyzed using immunefluoresce staining and Western blotting with mouse polyclonal antibodies against anti-E. coli synthesized PLpro, as described in our prior reports 12,13 . Quantifying relative mRNA expression of fibrotic genes using real-time RT-PCR. To measure the expression of SARS PLpro, TGF-β 1, pro-fibrotic and pro-protein convertase genes in transfected cells or mouse lung tissues, total RNAs were extracted from transfected A549 cells with empty vector pcDNA3.1 or pSARS-PLpro 2 days post transfection using PureLink Micro-to-Midi Total RNA Purification System kit (Invitrogen). Relative mRNA levels were analyzed using two-step real time RT-PCR with SYBR Green I, as described in our prior reports 11,12 . Primer pairs of SARS PLpro, TGF-β 1, pro-fibrotic and pro-protein convertase genes were listed in Table 1. Quantification of specific PCR products was performed using the ABI Prism 7900HT Sequence Detection System (PE Applied Biosystems). Relative changes in mRNA level of indicated genes were normalized relative to GAPDH mRNA.
Immunofluorescence staining assay. For determining the expression of PLpro and TGF-β 1 as well as nuclear translocalization of transcription factors, A549 cells transiently transfected with pSARS-PLpro or empty vector grew on the glass coverslip in 6-well at 37 °C. After 2 days incubation, transfected cells were fixed with 3.7% formaldehyde in phosphate buffered saline (PBS) for 1 h, blocked with 1% bovine serum albumin (BSA) in PBS for the other 1 h, and then incubated with specific primary antibodies at 4 °C overnight, including mouse anti-SARS PLpro, rabbit anti-TGF-β 1 (Cell signaling), rabbit anti-NF-κ B p65 (Abcam), rabbit anti-Sp-1,
Dual-luciferase reporter assay of TGF-β1 promoter activation. To test the activation of TGF-β 1 promoter by SARS-CoV PLpro, PLpro-expressing and empty vector control cells were co-transfected with TGF-β 1 promoter firefly luciferase reporter plasmids and internal control Renilla luciferase reporter pRluc-C1, as we reported earlier 12  Approximately 100 μ l of 3% sucrose in PBS containing 50 μ g of pSARS-PLpro, empty vector or solvent alone were injected into a right chest of mouse using a 1-ml syringe with a 28-gage needle every 2 days. Each group of 5 eight-weeks-old BALB/c male mice was injected 15 times, and then sacrificed. The lung tissues of each mouse in indicated groups were collected for immunohistochemistry (IHC) staining, and SYBR Green real time RT-PCR, respectively. For IHC staining, mouse lung tissues were fixed in formaldehyde and dehydrated in 70% ethanol for 30 min, in 95% ethanol for 30 min, and finally in 100% ethanol for 30 min. The tissues were embedded in paraffin at 58 °C, then cut at 4-15 μ m thick section using a rotary microtome. Before staining, the sections were floated in a 56 °C water bath and mounted the sections onto slides. The slides with paraffin embedded section of mouse lung tissue were dewaxed in xylene 2 times for 5 min, rehydrated in 100% ethanol for 1 min, in 90% ethanol for 1 min, and finally in 80% ethanol for 1 min. Slides with mouse lung tissues were incubated in 3%H 2 O 2 for 1 min to remove endogenous peroxidase activity, washed with PBS, and heated with 100 °C EDTA pH 9.0 for 20 min to induce antigen retrieval. Subsequently, slides were blocked with a protein block solution, and incubated with primary antibodies for 30 min, including mouse anti-E. coli synthesized PLpro serum, anti-mouse CD11b, anti-mouse TGFβ 1 (Cell signaling), and anti-mouse vimentin (GeneTex). After washing with PBS, slides were reacted with Polymer-HRP for 20 min, developed using DAB substrate, and counterstained with hematoxylin. For collagen determination, the tissue sections were stained with Sirius red solution for 2 h, and then rinsed 10 times with 0.5% glacial acetic acid in PBS. After dehydrating with ethanol, stained sections were mounted on the glass slides, and then examined using light microscopy (Olympus, BX50). For quantitating relative mRNA levels, mouse tissue was homogenized, and performed as above described. Primer pairs of mouse TGF-β 1 and pro-fibrotic genes were listed in Table 1.
Statistical analysis. All data were calculated from 3 independent experiments. Student's t-test or χ 2 test was used to analyze all data. Statistical significance was considered at p < 0.05.

Result
SARS-CoV PLpro induced TGF-β1 production in human lung epithelial cells. Our prior study demonstrated SARS-CoV PLpro triggering the TGFβ 1 production in human promonocytes 13 , whether SARS PLpro induced TGF-β 1 production in human lung epithelial A549 cells was further examined. Transient transfection of A549 cells with empty vector or pSARS-PLpro was performed to analyze the TGF-β 1 production induced by SARS PLpro. Quantitative PCR, immunofluorescence staining, and Western blotting indicated transfection with pSARS-PLpro increasing the mRNA and protein levels of PLpro in A549 cells in a concentration-dependent manner, but not empty control vector ( Fig. 1A-D). Meanwhile, relative mRNA levels of TGF-β 1 in A549 cells were time-and concentration-dependently elevated following the transient transfection with pSARS-PLpro, but not vector control ( Fig. 2A). Immunofluorescence staining assays indicated the protein levels of TGF-β 1 obviously heightening in transfected cells with pSARS-PLpro compared to vector control (Fig. 2B). For examining the TGF-β 1 induction of SARS-CoV PLpro in different cell lines, Huh7 (human hepatocarcinoma), H1299 (human non-small cell lung carcinoma), and ca9-22 (human oral cancer) cells were also evaluated (Supplemental Fig. 1). Real-time RT PCR analysis of transfected cells with pSARS-PLpro indicated that a lower level of TGF-β 1 mRNA was detected in transfected H1299 cells compared to transfected A549 cells, but no significant level was found in transfected Huh7 and ca9-22 cells. In addition, comparison of the expression levels of PLpro and TGF-β 1 among transfected cells with empty vector, pSARS-PLpro, and pBAC-SARSCoVΔ ES (a non-infectious SARS-CoV replicon) was further performed (Supplemental Fig. 2). The expression level of PLpro in transfected cells with pSARS-PLpro was 25-fold higher than the cells transfected with pBAC-SARSCoVΔ ES. A dose-dependent increase of TGF-β 1 mRNA levels in A549 cells was induced by pSARS-PLpro and pBAC-SARSCoVΔ ES, respectively. Meanwhile, recombinant plasmids containing PLpro genes of MERS-CoV and HCoV NL63 (pMERS-PLpro, and pNL63-PLpro) were used for testing the specificity on the TGF-β 1 induction compared to pSARS-PLpro (Supplemental Fig. 3). Interestingly, only SARS-CoV PLpro, but not ERS-CoV and HCoV NL63 PLpro, dose-dependently up-regulated the mRNA expression of TGF-β 1. Therefore, the result demonstrated that SARS-CoV PLpro plays an important role in triggering a significant increase of TGF-β 1 mRNA and protein levels in human lung epithelial cells.
TGF-β1-dependent up-regulation of pro-fibrotic genes by PLpro. To evaluate the correlation between TGF-β 1 production and pro-fibrotic gene expression in PLpro-expressing cells, mRNA and protein levels of pro-fibrotic genes such as vimentin and glial fibrillary acidic protein (GFAP) in transfected cells were assessed using quantitative RT-PCR and Western blotting (Fig. 3). The mRNA levels of vimentin and GFAP were up-regulated in transfected cells with pSARS-PLpro, but not vector control (Fig. 3A,B). Besides, Western blotting showed the plasmid dose-dependent increase of vimentin proteins in PLpro-expressing cells, but not in vector controls (Fig. 3C,D). Next, a selective inhibitor of TGF-β 1 receptor (SB-431542) was used to test the association of pro-fibrotic gene up-regulation with the TGF-β 1 induction in PLpro-expressing cells (Fig. 3E). SB-431542 exhibited a dose-dependently inhibitory effect on SARS-PLpro-induced expression of vimentin (Fig. 3D). The result demonstrated SARS-CoV PLpro initiating TGF-β 1-dependent up-regulation of pro-fibrotic genes in human lung epithelial cells.
To further investigate the nuclear localization of NF-κ B, Sp-1 and Egr-1, both types of cells were analyzed by immunofluorescence staining with primary and FITC-conjugated secondary antibodies, plus DAPI nuclear counterstain (Fig. 5). Immunofluorescence imaging analysis indicated NF-κ B and Sp-1 were localized in the nucleus as well as Erg-1 was localized in the cytoplasm of vector control cells (Fig. 5A-C). However, PLpro stimulated the translocation of Erg-1 into the nucleus, but inactivated the NF-κ B and Sp-1 that were localized in the cytoplasm as an inactive complex (Fig. 5A-C). The finding correlated with the previous report in that SARS CoV PLpro has the inhibitory ability on the activation of NF-κ B into nucleus 10 . In addition, Western blotting indicated SARS-CoV PLpro causing the increased expression of Egr-1 in plasmid dose-dependent manners (Fig. 5D,E). Subsequently, gene silencing of Erg-1 by RNA interference was performed to examine the role of Egr-1 in    Sp-1mut)) constructs of TGF-β 1 promoter luciferase reporter were shown (A). A549 cells transfected with 10 μ g of pcDNA3.1 and pSARS-PLpro were subsequently co-transfected with dual-luciferase reporters, and harvested for dual-luciferase reporter assays 1 day post transfection. TGF-β 1 promoter-driven firefly luciferase and renilla luciferase were measured, and firefly luciferase activity normalized to renilla luciferase activity is reported (B). *p value < 0.05 compared with vector control cells.
PLpro-induced up-regulation of TGF-β 1 and pro-fibrotic responses (Fig. 6). Egr-1 siRNA, not non-targeting siRNA, definitely reduced mRNA and protein levels of Egr-1, TGF-β 1, vimentin and α -SMA in PLpro-expressing cells, but slightly decreased them in vector control cells. The result demonstrated gene silencing of Egr1 linked with the decrease of TGF-β 1 promoter activation and pro-fibrotic responses in PLpro-expressing cells. The finding revealed Egr-1 up-regulated by SARS-CoV PLpro playing a crucial role in the activation of TGF-β 1 promoter, as well as the induction of TGF-β 1-mediated pro-fibrotic responses in SARS pathogenesis.
ROS/p38 MAPK/STAT3 pathway was responsible for PLpro-induced Egr-1 dependent TGF-β1-mediated pro-fibrosis. Since the intracellular ROS generation was reported to modulate the expression of Egr-1 and TSP-1 26,27 , the involvement of ROS-mediated pathway in PLpro-induced Egr-1-depedent activation of TGF-β 1 and pro-fibrotic responses was explored (Figs 8-10). Intracellular ROS levels in PLpro-expressing and vector control cells were detected using flow cytometry with DCFH-DA staining (Fig. 8A). DCF fluorescence analysis indicated PLpro triggered the ROS generation in a dose dependent manner. Meanwhile, ROS inhibitor (YCG063) concentration-dependently reduced PLpro-induced up-regulation of Egr-1 and TSP-1 (Fig. 8B). For insight into the pathway of ROS-mediated Egr-1 up-regulation, the activity of MAP kinases and transcription factors were analyzed using Western blotting (Fig. 9A,B). ROS inhibitor (YCG063) significantly reduced PLpro-induced phosphorylation of p38 MAPK and STAT3. Furthermore, p38 MAPK inhibitor (SB203580) notably declined the PLpro-induced expression of Egr-1, TGF-β 1 and vimentin, as well as PLpro-induced phosphorylation of STAT3 (Figs 9C and 10A,B). STAT3 inhibitor (Stattic) also diminished PLpro-induced expression of Egr-1, vimentin, and Type I collagen (Fig. 10C). Therefore, results showed 1 or pSARS-PLpro were re-transfected with negative control or Egr-1siRNA were harvested 1 day post transfection for quantitative PCR (A) and Western blotting (B) assays. Relative mRNA levels of Egr-1, TGF-β 1, and vimentin were normalized by GAPDH mRNA, presenting as relative ratio. The Western blot was probed with indicated primary antibodies as an internal control, detected using HRP-conjugated secondary antibodies and chemiluminescent HRP substrate. Relative band intensity of indicated proteins was normalized by β actin, compared to the mock cell group, and quantified using imageJ based on triplicate replicates of each experiment (C). *p value < 0.05; **p value < 0.01 compared with vector control cells.   For analysis of ROS-mediated signaling, A549 cells transfected with 10 μ g of pcDNA3.1 or pSARS-PLpro were treated with or without ROS inhibitor (YCG063), and then harvested for Western blotting assays with specific primary antibodies against Egr-1, phospho-p38 MAPK, and phospho-STAT3 (A). Relative band intensity of indicated proteins was normalized by β actin, compared to the mock cell group, and quantified using imageJ based on triplicate replicates of each experiment (B). For conforming the role of p38 MAPK in ROS-dependent Egr-1 up-regulation, both types of cells were treated with p38 MAPK inhibitor (SB203580), and then harvested for quantitating relative changes of Egr-1, TGF-β 1, and vimentin mRNA, in which was normalized by GAPDH mRNA and presented as the relative ratio (C). *p value < 0.05; **p value < 0.01 compared with untreated cells. Pulmonary pro-fibrotic activity of SARS-CoV PLpro in a mouse model. A mouse model that direct injected with empty vector or pSARS-PLpro into the mouse lung was set up for examining the pro-fibrotic activity of PLpro in vivo (Fig. 11). The mice were injected 15 times with indicated plasmids into the chest every two days (Fig. 11A), and then sacrificed and collected the lung tissue for tissue immunohistochemistry stain and quantitative RT-PCR (Fig. 11B,C). Immunohistochemistry staining with anti-PLpro immunized sera indicated a significant expression of SARS-PLpro in lung tissues of mice injected with pSARS-PLpro, but not empty vector or solvent control (Fig. 11B). Lung infiltration of immune cells, particular CD11b monocytes, was identified in pulmonary alveoli expressing SARS-PLpro, but not the controls, using immunohistochemistry staining with anti-mouse CD11b mAb. In addition, PLpro, but not the controls triggered a significant increase of TGF-β 1 and vimentin protein levels in the lung tissues. Subsequently, real-time PCR confirmed PLpro raised the mRNA expression of Egr-1, TSP-1, TGF-β 1, and vimentin in mouse lung tissues versus empty vector or solvent control (Fig. 11C). Overall, result of the in vivo experiments was in accord with the in vitro data that SARS PLpro substantially stimulated Egr-1 dependent TGF-β 1 mediated pro-fibrotic responses.

Discussion
This study verified SARS-CoV PLpro inducing TGF-β 1 mediated pro-fibrotic responses in human lung epithelial cells and mouse lung tissues (Figs 1-3 and 11), according with the previous report in that PLpro up-regulated TGF-β 1 and its associated genes such as glial fibrillary acidic protein (GFAP) and vimentin 13 . Except SARS-CoV nucleocapsid 28 , PLpro was identified to generate the TGF-β 1 production that linked to activate the pro-fibrotic responses. Among SARS-CoV-induced cytokines 5,6 , TGF-β 1 could be associated with the induction of lung fibrosis. Therefore, SARS-CoV PLpro plays an important role in the TGF-β 1-mediated pulmonary fibrosis of SARS pathogenesis.
This study proved SARS-CoV-PLpro heightening the role of Egr-1 in TGF-β 1-mediated pro-fibrosis, in which increased the expression and nuclear translocalization of Egr-1, as well as strengthened the transcription of Egr-1-responsive genes (TGF-β 1 and TSP-1) (Figs 5-7). In addition, gene silencing by siRNA confirmed the importance of Egr-1 on TGF-β 1-mediated pro-fibrosis induced by PLpro (Fig. 6). Previous studies demonstrated Egr-1 exhibiting potent stimulatory action on fibrotic gene expression, and correlating with human fibrotic disorders like emphysema, pulmonary fibrosis, and systemic sclerosis 29,30 . Those results revealed TGF-β increasing Egr-1 protein and mRNA levels, stimulating Egr-1-depdent transcription of collagen, and then triggering Smad-independent fibrotic response. However, our study indicated SARS-CoV PLpro causing Egr-1-depdent transcription of TGF-β 1, in which was associated with TGF-β 1-mediated up-regulation of pro-fibrotic genes (such as vimentin, GFAP, and α -SMA). Of pro-protein convertases 19-21 , Egr-1 dependent up-regulation of TSP-1 in PLpro-expressing cells was discovered (Figs 7B and 8B), in which TSP-1 was suggested as responsible for the Figure 10. Analysis of STAT3 activity in p38 MAPK-mediated pathway of PLpro-induced Egr-1 up-regulation. For analysis of p38 MAPK-dependent signaling, A549 cells transfected with 10 μ g of pcDNA3.1 or pSARS-PLpro were treated with or without p38 MAPK, MEK, and ERK inhibitors (SB203580, U0126, and PD98059), and then harvested for Western blotting assays with specific primary antibodies against phospho-STAT3 (A). Relative band intensity of indicated proteins was normalized by β actin, compared to the mock cell group, and quantified using imageJ based on triplicate replicates of each experiment (B). For conforming the role of STAT3 in p38 MAPK-dependent Egr-1 up-regulation, both types of cells were treated with STAT3 inhibitor (Stattic), and then harvested for quantitating relative changes of Egr-1, vimentin, and type I collagen mRNA, in which was normalized by GAPDH mRNA and presented as the relative ratio (C). *p value < 0.05; **p value < 0.01 compared with vector control cells. The lung tissues were collected after 15-times of chest injection, embedded, and sectioned. The lung tissue sections were performed using IHC staining with anti-PLpro sera, anti-CD11b, anti-TGF-β 1, and anti-vimentin antibodies. (C) Total RNAs of the lung tissues were extracted; relative mRNA levels of Egr-1, TSP-1, TGF-β 1, and vimentin were quantitated using real-time PCR, normalized by GAPDH mRNA, and presented as the relative ratio. *p value < 0.05; **p value < 0.01 compared with the solvent control group.