Cell type-specific roles of PAR1 in Coxsackievirus B3 infection

Protease-activated receptor 1 (PAR1) is widely expressed in humans and mice, and is activated by a variety of proteases, including thrombin. Recently, we showed that PAR1 contributes to the innate immune response to viral infection. Mice with a global deficiency of PAR1 expressed lower levels of CXCL10 and had increased Coxsackievirus B3 (CVB3)-induced myocarditis compared with control mice. In this study, we determined the effect of cell type-specific deletion of PAR1 in cardiac myocytes (CMs) and cardiac fibroblasts (CFs) on CVB3-induced myocarditis. Mice lacking PAR1 in either CMs or CFs exhibited increased CVB3 genomes, inflammatory infiltrates, macrophages and inflammatory mediators in the heart and increased CVB3-induced myocarditis compared with wild-type controls. Interestingly, PAR1 enhanced poly I:C induction of CXCL10 in rat CFs but not in rat neonatal CMs. Importantly, activation of PAR1 reduced CVB3 replication in murine embryonic fibroblasts and murine embryonic cardiac myocytes. In addition, we showed that PAR1 reduced autophagy in murine embryonic fibroblasts and rat H9c2 cells, which may explain how PAR1 reduces CVB3 replication. These data suggest that PAR1 on CFs protects against CVB3-induced myocarditis by enhancing the anti-viral response whereas PAR1 on both CMs and fibroblasts inhibits viral replication.

Recently, we showed that PAR1 plays a role in CVB3 infection in mice 5,30 . We found that PAR1 -/mice expressed lower levels of CXCL10 in the early phase of CVB3 infection that resulted in an increased viral load, inflammation and cardiac injury a later phase after infection 5 . In vitro studies with murine embryonic (me) CFs demonstrated that PAR1 activation enhanced poly I:C activation of the p38 MAPK pathway and induction of CXCL10 expression 5 . These data indicated that PAR1 on meCFs contributes to the TLR3-dependent innate response to CVB3 infection. CVB3 can infect and replicate in both CMs and CFs, although replication is higher in CF 31 . In addition, CFs infected with CVB3 express inflammatory mediators that aggravated myocarditis 31 . Importantly, CVB3 uses autophagy to increase its replication 32 .
In this study, we determined the effect of cell type-specific deletion of PAR1 in either CFs or CMs on CVB3induced myocarditis. In addition, we analyzed the effect of PAR1 in CVB3 replication in murine embryonic fibroblasts (MEFs) and meCMs. Finally, we measured the effect of PAR1 activation on autophagy in MEFs and rat H9c2 cells.

Material and methods
Mice. To delete PAR1 in CMs (PAR1ΔCM), we used mice expressing the Cre recombinase under the control of the myosin light chain 2v (Mlc2v) promoter were purchased from Dr. Chien 33 . To delete PAR1 in CFs (PAR1ΔCF), we used mice expressing Cre recombinase under the transcription factor 21 (TCF21) obtained from Dr. Tallquist 34 . Female PAR1 fl/fl mice 35,36 were crossed with male Cre-expressing mice to generate PAR1 fl/ fl ;Mlc2v Cre and PAR1 fl/fl ;TCF21 Cre-ERT2 . To activate TCF21 Cre-ERT2 expression 34,37 in PAR1 fl/fl ;TCF21 Cre-ERT2 mice, 6 week old mice were gavaged with 2 mg of tamoxifen (Sigma-Aldrich, St. Louis, MO, USA) for 5 consecutive days and then mice used 5 days later. PAR1 fl/fl mice treated with tamoxifen were used as controls. In addition, we used wild-type (control) mice and mice with a global PAR1 deletion (ΔPAR1) 38 . All mice strains were on a C57Bl/6 J background. The studies were performed and approved in accordance with the guidelines of the Institutional Animal Care and Use Committee of the UNC-Chapel Hill and complied with National Institutes of Health and ARRIVE guidelines 39 . Virus infection and histology. Male mice were infected at 6-8 weeks of age intraperitoneal (i.p.) with 10 5 plaque forming units (pfu) CVB3 (cardiotropic Nancy strain) 5,16,40 . Hearts were fixed in 10% formalin and embedded in paraffin (Sigma-Aldrich). Sections from paraffin-embedded sections were cut and stained with H&E. Area of cellular infiltration was calculated from whole cardiac tissue area on H&E stained tissue sections by 2 blinded investigators as described previously 5 . For macrophage staining, slides were blocked with 5% Non-fat skim milk (Applichem, Council Bluffs, IA, USA) in phosphate-buffered saline (Thermo Fisher Scientific, Waltham, MA, USA) for 1 h at room temperature. Endogenous avidin and biotin were blocked with avidin solution and biotin solution, respectively (Vector Laboratories, Burlingame, CA, USA). Sections were incubated overnight at 4 °C with 0.4 μg/mL rabbit anti-mouse Iba1 antibody (FUJIFILM Wako Chemicals, Osaka, Japan, CAT#019-19741). Sections were washed and incubated with biotinylated goat anti-rabbit antibody (Vector Laboratories, CAT# BA-1000) for 1 h at room temperature. Avidin-peroxidase complex solution (Vector Laboratories, CAT# PK-6100) was added and Iba-1 staining (brown) was detected using the DAB substrate (Agilent Technology, Santa Clara, Ca, USA, CAT#K3468). Sections were counterstaining with hematoxylin (Thermo Fisher Scientific) for 1 min. Sections were evaluated with a 1 to 5 scale by two blinded investigators.

Measurement of autophagy.
To analyse autophagic flux, wild-type (control) or ΔPAR1 MEFs were seeded the day before in growth media (GM). The next day, GM was replaced with either serum-free media (SFM) alone or SFM supplemented with chloroquine (CQ, 10 µM). Rat H9c2 CMs (ATCC, Manassas, VA, USA) 42 were infected with CVB3 (MOI 5) for 30 min in SFM. After a 30 min incubation period, media was replaced with SFM with or without thrombin (10 nM) for 24 h. After 6 h or 24 h, cells were lysed in RIPA buffer and subjected to Western blot analyses as described 5,41 . Autophagic flux was determined using an anti-light chain 3A (LC3A) antibody (1:1000, #PM036, MBL International Corporation, Woburn, MA, USA), which detects LC3-I and LC3-II.
To analyse the effect of thrombin stimulation on autophagy, wild-type MEFs or H9c2 CMs were seeded in GM. The next day, cells were either further cultured in GM or in SFM with or without 10 nM thrombin for 24 h. Total cell lysates were subjected to Western blot analyses for LC3A, p62 (1:1000, #PM045, MBL International Corporation) and ZBTB16 (1:1000, ab39354, Abcam, Cambridge, MA, USA) 43 . Loading differences were normalized to β-actin protein levels (Abgent). Images of the Western blots were obtained with an Odyssey LI-COR imager (LI-COR Bioscience, Lincoln, NE, USA) with a 700 and 800 nm filter 5 . Specific band intensities were determined with ImageJ (National Institutes of Health, Bethesda, MA, USA) 5 .
Statistics. All statistical analyses were performed in GraphPad Prism 9.1 (GraphPad Software Inc., La Jolla, CA, USA). Data are represented as mean ± SEM. For 2-group comparison of normally distributed data, the 2-tailed Student's t test was used. For multiple-group comparison, normally distributed data were analyzed by 1-or 2-way ANOVA tests and were Bonferroni-corrected for repeated measures over time. P-value ≤ 0.05 was regarded as significant.

Results
Cell-specific deletion of PAR1 in the heart using cardiac fibroblast and cardiac myocyte specific Cre expression. To analyze the deletion rate of PAR1 in hearts from the two transgenic lines, we quantified levels of PAR1 mRNA expression in hearts of uninfected PAR1ΔCF, PAR1ΔCM mice and their respective control littermates by RT-PCR. Expression of TCF21-promotor controlled Cre recombinase after tamoxifentreatment for 5 days resulted in significantly reduced expression of PAR1 in hearts of PAR1ΔCF mice compared to the tamoxifen-treated control mice (Fig. 1A). In addition, cardiomyocyte-specific Cre recombinase expression significantly reduced PAR1 mRNA expression in hearts of PAR1ΔCM mice compared to control littermates (Fig. 1B).

Mice with PAR1 deleted in either cardiac fibroblasts or cardiac myocytes have increased levels of CVB3 genomes, inflammatory mediators and inflammatory cells in their hearts after CVB3 infection.
We determined the effect of deleting PAR1 in either CFs or CMs on CVB3-induced myocarditis. Effect of PAR1 activation on poly I:C induction of CXCL10 in cardiac fibroblasts and cardiac myocytes. We recently showed that PAR1 activation enhanced poly I:C induction of CXCL10 expression in meCFs 5 . In this study, we used rnCFs and rnCMs because larger quantities of cells can be isolated. We confirmed that either thrombin or a PAR1 AP enhanced poly I:C induction of CXC10 expression in rnCFs (Fig. 5A). In contrast, treatment of rnCMs with either thrombin or PAR1 agonist peptide did not enhance poly I:C induction of CXC10 expression (Fig. 5B).
Effect of PAR1 on adhesion and uptake of CVB3 in cardiac myocytes. For CVB3 adhesion experiments, serum-containing GM was removed and the rnCMs were infected with CVB3 multiplicity of infection (MOI) of 1 under serum-free conditions and incubated for 30 min on ice. Next, rnCMs were washed twice with ice-cold PBS and total RNA was isolated. Virus genomes were measured by RT-PCR 44 . We found that activation of PAR1 with an agonist peptide did not affect adhesion of CVB3 (data not shown).
To analyse CVB3 uptake, rnCMs cells were incubated with CVB3 (MOI 1) under serum-free conditions for 30 min on ice. Next, rnCMs were washed and incubated in SFM for 1 h at 37 °C. Trypsin was added to the cells for 1.5 min to remove the remaining virus from the cell surface. CMs were washed twice with PBS, total RNA was isolated and virus genomes measured by RT-PCR 44 . Activation of PAR1 with an agonist peptide did not affect uptake of CVB3 (data not shown). PAR1 reduces CVB3 replication in murine embryonic fibroblasts and murine embryonic cardiac myocytes. Next, we determined if PAR1 stimulation affects CVB3 replication. GM was aspirated from the cells and CVB3 solution (MOI 1) was added to MEFs or meCMs. After 30 min at 37 °C the medium was aspirated, and fresh GM was added to the cells. Levels of CVB3 genomes were analysed 6 h after infection by RT-PCR   www.nature.com/scientificreports/ and normalized using HPRT mRNA 5,30 . Active CVB3 virus was analysed after 24 h 5, 16 . We found that PAR-1 AP stimulation reduced CVB3 genome levels in MEFs (Fig. 6A). In addition, we also observed significantly reduced levels of CVB3 genomes in meCMs treated with a PAR1 agonist peptide (Fig. 6B) indicating that PAR1 in CMs limits CVB3 replication. We did not observe a decrease in CVB3 infection of meCMs lacking PAR1 compared to wild-type cells (data not shown). Treatment of meCMs with a PAR1 AP also significantly reduced the number of CVB3 pfu (Fig. 6C).
PAR1 activation reduces autophagy. CVB3 uses autophagy to increase its replication 32 . In addition, a recent publication showed that GPCR activation reduces autophagy 45 . We compared autophagic flux in wildtype and ΔPAR1 MEFs by analyzing the conversion of the cytosolic precursor form of microtubule-associated protein LC3-I to the autophagosome-localized, cleaved, and lipidated form LC3-II 43 . We found that MEFs isolated from ΔPAR1 exhibited an increased LC3-II/LC3-I ratio, which is an indicator for increased autophagic flux (Fig. 7A,B). We used the autophagy inhibitor CQ, which inhibits the last stage of autophagy by preventing protein degradation that leads to an accumulation of LC3-II. Again, ΔPAR1 cells exhibited an increase in autophagy flux in the presence of CQ compared with wild-type cells (Fig. 7A,B). Next, we analyzed if thrombin stimulation of wild-type MEFs can reverse the autophagy induction during serum-starvation. First, we analyzed the changes in LC3-II/LC3-I ratios in MEFs cultured under serum and serum-free condition with or without thrombin stimulation. As expected, serum deprivation leads to autophagy induction seen as an increase in the LC3-II:LC3-I ratio (Fig. 7C,D). Thrombin stimulation significantly reduced the LC3-II/LC3-I ratio induced by serum starvation (Fig. 7C,D). Next, we investigated the changes in p62 protein levels (also known as Sequestosome 1, or SQSTM1, a substrate degraded by autophagy). Again, serum deprivation drastically reduced p62 levels in MEFs and thrombin stimulation resulted in significantly higher  www.nature.com/scientificreports/ p62 levels compared to serum deprivation alone (Fig. 7C,E). Lastly, we analyzed the changes in ZBTB16 protein levels, also known as promyelocytic leukemia zinc finger (PLZF) or Zfp145, which was shown to be important for GPCR-mediated autophagy inhibition 45 . Serum deprivation caused a reduction in ZBTB16 in MEFs that was significantly reduced by thrombin (Fig. 7C,F). We found that thrombin significantly inhibited the loss of p62 expression induced by serum deprivation in the rat H9c2 cells (Fig. 8A,B). In addition, thrombin abolished the reduction on p62 in rat H9c2 cells infected with CVB3 (Fig. 8C,D).

Discussion
In this study, we found that mice lacking PAR1 in either CFs or CMs had increased CVB3-induced myocarditis compared with controls, which indicated that PAR1 on these different cell types regulates viral infection. Previously, we have shown that PAR1 activation in CFs enhances poly I:C induction of IFN-β and CXCL10 5 . Here, we found that PAR1 activation did not enhance poly I:C induction of CXCL10 in CMs. However, we found that PAR1 activation on CMs reduced CVB3 replication in MEF and CMs. Furthermore, PAR1 expression and activation reduced serum deprivation and CVB3 infection-induced autophagy in MEFs and rat H9c2 CMs. This suggests that PAR1 enhances the anti-viral response to CVB3 infection and inhibits its replication by reducing autophagy.
We found higher levels of PAR1 mRNA in CFs compared to CMs (data not shown). The fact that PAR1 enhances CXCL10 expression in CFs but not CMs is not surprising. Although CFs express lower basal levels of IFN-β compared to CMs, CFs have a larger induction of IFN-β after virus infection compared to CMs 46,47 . In addition, activation of the IFN-β-dependent signaling pathway in CFs was found to be more important to restrict virus replication compared to CMs 46 . Together, this suggests that CFs are more reliant on the induction of this antiviral pathway than CMs after viral infection. Therefore, PAR1 enhancement of the IFN-β-dependent signaling pathway in CFs will have a greater impact on CVB3-induced myocarditis than its expression in CMs 46 .
In line with the notion of a cell-specific response to PAR1 activation, Sabri et al. 10 showed that CFs exhibited a different intracellular signaling pattern compared to CMs after PAR1 stimulation. These differences were attributed to specific PAR1-dependent transactivation of the epidermal growth factor receptor (EGFR) via Src leading www.nature.com/scientificreports/ to enhanced p38 signaling in CFs 2,10 . PAR1 transactivation of EGFR is not observed in CMs 10 . Importantly, EGFR-dependent Src activity was required for effective TLR3-dependent immune responses 48 . Activated Src led to TLR3 phosphorylation leading to increased TRIF binding and subsequently enhanced TLR3:TRIF signaling 48 . We found that PAR1 activation has no influence on CVB3 infectivity with regards to virus adhesion and uptake into CMs. CVB3 mainly uses the coxsackievirus adenovirus receptor (CAR) and its co-receptor decayaccelerating factor (DAF) to infect cells 19 . Moreover, CMs are the primary site of CVB3 replication in the heart 49 . Interestingly, it was shown that uptake of the enveloped DNA virus, herpes simplex virus 1, into cells is increased by PAR1-dependent signaling 50 . However, there was no PAR1-dependency in the infection of cells with another enveloped virus, influenza virus A 51 . The lack of a PAR1 effect on CVB3 infection of CMs suggests that PAR1 does not affect CVB3 interaction with CAR and/or DAF. The differences in PAR1-dependency on uptake of these three viruses is surprising since they all use endocytosis for cell entry.
Importantly, we found that PAR1 stimulation reduced CVB3 replication in MEFs and CMs. A hallmark of RNA( +) viruses, such as CVB3, is that they use host membrane structures for their replication 19,32 . After CVB3 infects cells it associates with intracellular membrane structures (multivesicular bodies, [MVB]) via endosomes, which supports virus replication 19 . While CVB3 initiates early endosomal/autophagic processes it is able to block the maturation and fusion with lysosomes inhibiting its own degradation 19 . Here, we showed that PAR1 deficiency was associated with reduced autophagic flux in MEFs as measured by an increase in the LC3-II:LC3-I ratio. We showed that MEFs without stimulation release PAR1 activating proteases, such as matrix metalloproteinases 13 2,5 , causing paracrine and/or autocrine PAR1-dependent autophagy inhibition in wild-type but not PAR1 deficient cells. Interestingly, an increase in the conversion of LC3-I to LC3-II and autophagic flux was shown to enhance CVB3 replication 52,53 . A recent study showed that GPCRs activation reduces autophagy by stabilizing ZBTB16 protein levels 45 . Serum starvation was shown to cause ZBTB16 proteasomal degradation due to increased glycogen synthase kinase-3 (GSK3) activity, which subsequently leads to autophagy induction 45 . We showed that serum starvation increases ZBTB16 degradation and autophagy induction in MEFs. Interestingly, PAR1 activation with thrombin was able to reduce autophagy induction as determined by a lower LC3-II:LC3-I ratio and higher p62 and ZBTB16 protein levels. In addition, autophagy induced by either serum deprivation or CVB3 infection of H9c2 cells was significantly reduced by thrombin. This might be due to the known effect of PAR1 activation on GSK3 inhibition 54 , which could lead to increased inhibition of autophagy 45 . A recent study showed that CVB3 protease 2A can reduce p62 levels by direct cleavage causing increased CVB3 replication 55 . Furthermore, overexpression of p62 restricted whereas p62 knock-down supported CVB3 replication in vitro 55 . Thrombin can also activate PAR4. However, we found that activation of PAR4 does not affect CVB3 replication in CMs 41 , which suggested that PAR1 but not PAR4 influences CVB3 life cycle in CMs possibly by maintaining p62 levels. Together, the negative effects of PAR1 on autophagy might be the reason we see reduced virus replication in wild-type cells stimulated with thrombin. Our observation of PAR1-dependent autophagy inhibition might explain certain protective phenotypes in ΔPAR1 mice, such as in cardiac ischemia-reperfusion injury 11 , www.nature.com/scientificreports/ chronic angiotensin-II infusion 56 or doxorubicin induced heart failure 57 where increased autophagy was shown to be protective [58][59][60] .

Conclusion
In this study, we showed that deletion of PAR1 in either CFs or CMs was associated with increased CVB3-induced myocarditis. PAR1 activation enhanced poly I:C induction of in CFs but not CMs. PAR1 activation reduced CVB3 replication in MEFs and CMs. Finally, thrombin activation of PAR1 inhibited autophagy, which may explain its inhibitory effect on CVB3 replication.