Mesenchymal stem cells regulate M1 polarization of peritoneal macrophages through the CARD9‐NF‐κB signaling pathway in severe acute pancreatitis

Macrophages release large numbers of proinflammatory cytokines that trigger inflammatory cascade reactions, which promote the rapid development of severe acute pancreatitis (SAP) from local to systemic inflammation. The ability of mesenchymal stem cells (MSCs) to suppress inflammation is related to inhibition of M1 polarization of macrophages. Our previous studies revealed that caspase recruitment domain protein 9 (CARD9) was involved in SAP inflammation and activation of the CARD9‐NF‐κB signaling pathway plays an important proinflammatory role in SAP. At present, there is no effective treatment to control the inflammatory response in SAP. Therefore, the aim of the present study was to determine whether MSCs regulate the polarization of macrophages through the CARD9‐NF‐κB signaling pathway in SAP.


| INTRODUCTION
Severe acute pancreatitis (SAP) is an acute abdominal disease characterized by high prevalence, severe symptoms, complicated pathogenesis, and high mortality. Macrophage activation leads to overproduction and release of proinflammatory cytokines, including tumor necrosis factor (TNF)-α and interleukin (IL)-6, which play important roles in the early stage of acute pancreatitis 1 via M1 polarization of macrophages. 2,3 Although no effective treatment is currently available, new stem therapies can potentially reduce inflammation associated with SAP. Mesenchymal stem cells (MSCs) isolated from various tissues, including the stroma of bone marrow and adipose tissue, have been demonstrated to exert therapeutic effects in various inflammation-based diseases, such as ischemia/reperfusion injury associated with kidney disease, collagen-induced arthritis, and acute renal failure. [4][5][6] MSCs respond to inflammation by secretion of cytokines, exosomes, and apoptotic bodies targeting inflamed tissues, which provides local control of inflammation and facilitates tissue repair. [7][8][9] Previous studies have demonstrated that MSCs can alleviate damage to the pancreas and other organs by inhibiting inflammation in SAP. 10 Caspase recruitment domain-containing protein 9 (CARD9) is highly expressed in macrophages and closely associated with the immune response, immune cell activation, and inflammation. [11][12][13][14][15] Our previous study demonstrated that CARD9 expression was significantly positively correlated with the severity of SAP. 16 Small interfering RNAinduced inhibition of CARD9 was shown to obstruct NF-κB activation in pancreatic tissue and peritoneal macrophages, and reduce the release of TNF-α, IL-6, and other proinflammatory cytokines, thereby ameliorating inflammation associated with SAP and improving survival. 17,18 However, there is currently no treatment strategy to regulate CARD9 and, thus, control inflammation associated with SAP.
Therefore, the aim of the present study was to determine whether MSCs can reduce inflammation associated with SAP by regulating CARD9 expression through gene interference techniques in vivo and in vitro, and to further explore the mechanism underlying the regulatory roles of MSCs in macrophage polarization as a potential target for treatment of SAP.

| Animal model of SAP
Healthy male Sprague Dawley (SD) rats, 17 weighing 200-260 g, were obtained from Shanghai Jie Si Jie Laboratory Animal Co. Ltd (animal permit number: SCXK[Hu] 20130006) and acclimated for 1 week prior to experimentation. SAP was induced by retrograde perfusion of 5% sodium taurocholate at 1 mL/kg bodyweight. All rats were fasted with access to drinking water only for 12 hours prior to surgery.

| Transplantation of MSCs in vivo
Mesenchymal stem cells were obtained from the Shanghai Cell Bank (Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences). SD rats were randomly allocated to one of three groups (n = 6 each): a control group consisting of sham-operated rats, MSC group consisting of rats injected with 1 × 10 6 MSCs dissolved in 1 mL of phosphate-buffered saline (PBS) into the caudal vein, and a SAP group consisting of rats injected with 1 × 10 6 MSCs dissolved in 1 mL of PBS into the caudal vein plus 5% sodium taurocholate at 1 mL/ kg bodyweight 1 hour later. Following treatment, all rats were fasted with free access to drinking water. Peripheral blood and pancreatic tissues were collected 12 hours after modeling for detection of serum amylase and inflammation-induced injury to the pancreas. 18 The experimental protocols involving animals were approved by the Ethical and Research Committee of Shanghai Jiao Tong University (Shanghai, China) and conducted in strict accordance with the Guide for the Care and Use of Laboratory Animals.

| Isolation and culture of peritoneal macrophages
After establishment of the SAP rat model, peritoneal ascites was extracted via abdominal lavage with precooled PBS. The peritoneal lavage fluid and ascites were mixed evenly and centrifuged to obtained peritoneal macrophages.
The peritoneal macrophages were seeded into the wells of 12-well plates and incubated under an atmosphere of 5% CO 2 /95% air for 3 hours at 37°C. Afterward, the plates were washed twice with PBS to remove unattached cells. 18 After 24 hours of incubation, the culture medium was collected for an enzyme-linked immunosorbent assay (ELISA), a portion of the RAW 264.7 cells was collected for flow cytometry, and the remaining RAW 264.7 cells were used for isolation of total RNA and proteins for real-time polymerase chain reaction (RT-PCR) and western blot analysis, respectively.

| Transfection of M1 macrophages
with short hairpin RNA (shRNA) targeting CARD9 in vitro Adenoviral constructs carrying shRNA against CARD9 (reference sequence: NM_022303.3) were designed and synthesized by OBiO Technology Corp., Ltd. The target sequences were 5′-GCT TTC AGG ACA AAG ATA A-3′ (shCARD9) and 5′-TTC TCC GAA CGT GTC ACG T-3′ (shcontrol). 17,18 The concentration of the original viral solution was 1 × 10 11 plaque-forming units/mL and diluted to a multiplicity of infection (MOI) of 1:1, 1:10, 1:100, 1:1000, and 1:10 000. The diluted adenoviruses were mixed in 100 μL of medium without fetal bovine serum and added to each well. RAW 264.7 murine monocyte/ macrophage-like cells were obtained from the Shanghai Cell Bank. Following coculture of adenoviruses and RAW 264.7 for 12 hours, adenovirus transfection and cell viability were observed by inverted fluorescence microscopy to determine the optimal MOI for subsequent experiments. RT-PCR and western blot analysis were used to determine whether CARD9 expression was suppressed in RAW 264.7 cells.

| Coculture of MSCs and M1 macrophages in vitro
An earlier study confirmed that lipopolysaccharide (LPS) + interferon gamma (IFNγ) stimulated RAW 264.7 cells to induce M1 polarization. 19 Hence, LPS (100 ng/mL) + IFNγ (20 ng/mL) were used to stimulate RAW 264.7 cells to establish a M1 macrophage model in vitro. A noncontact coculture system was established with a 0.4 μmdiameter transwell chamber with MSCs in the upper compartment and RAW 264.7 cells in the lower compartment. As shown in Table 1, the RAW 264.7 cells were divided into five groups: (a) NC group, consisting of RAW 264.7 cells transfected with control shRNA (RAW 264.7-control shRNA); (b) NC1 group, consisting of RAW 264.7 cells transfected with control shRNA for 24 hours followed by stimulation with LPS and IFNγ, and co-culturing with MSCs for 24 hours (RAW 264.7-control shRNA + LPS + IFNγ); (c) KD1 group, consisting of RAW 264.7 cells transfected with CRAD9 shRNA for 24 hours, followed by stimulation with LPS and IFNγ, and culturing for 24 hours (RAW 264.7-CRAD9 shRNA + LPS + IFNγ); (d) NC2 group, consisting of RAW 264.7 cells transfected with control shRNA for 24 hours, followed by stimulation with LPS and IFNγ, and coculturing with MSCs for 24 hours (RAW 264.7-control shRNA + LPS + IFNγ + MSCs); and (e) KD2 group, consisting of RAW 264.7 cells transfected with CARD9 shRNA for 24 hours, followed by stimulation with LPS and IFNγ, and coculturing with MSCs for 24 hours (RAW 264.7-control shRNA + LPS + IFNγ + MSCs). Following treatment, the culture medium was collected for the ELISA, a portion of the RAW 264.7 cells was collected for flow cytometry, and the remaining RAW 264.7 cells were used for total RNA and protein isolation for RT-PCR and western blot analysis, respectively.

| Flow cytometry and western blot analysis were used to determine the polarization state of macrophages
In vivo, C-C motif chemokine receptor 2 (CCR2) and cluster of differentiation 206 (CD206) were used as protein markers of M1 and M2 macrophages, respectively. M1 and M2 macrophages were incubated with fluorescein isothiocyanate-conjugated antibodies against CCR2 and CD206 (Beijing Boaosen Biotechnology Co., Ltd.), respectively, for 30 minutes at 4°C, washed with PBS, and then detected by flow cytometry to confirm the polarization states of peritoneal macrophages. In vitro, antibodies against CD86 (Abcam) and CD206 (Beijing Boaosen Biotechnology Co., Ltd.) were used for western blot analysis to confirm the polarization states of M1 and M2 RAW 264.7 cells, respectively. All experiments were repeated three times.

| Histological analysis
The pancreatic tissues were fixed with 10% formalin, embedded in paraffin, then cut into sections, which were stained with hematoxylin and eosin for assessment of inflammation, edema, hemorrhage, and necrosis. 18

| RT-PCR analysis
Total RNA was extracted from frozen pancreatic tissues and cells from each group using TRIzol reagent (Takara Bio, Inc.) in accordance with the manufacturer's instructions. The mRNA expression levels of CARD9 in pancreatic tissues, peritoneal macrophages, and RAW 264.7 cells, and mRNA expression levels of TNF-α, IL-6, IL-10, and arginase 1 (Arg1) in peritoneal macrophages and RAW 264.7 cells were detected with the use of a two-step PCR reaction procedure with the primer sequences listed in Table 2. Relative changes in gene expression levels were analyzed with the 2 −ΔΔC t method.

| Western blot analysis
After processing, the pancreatic tissues, peritoneal macrophages, and RAW 264.7 cells were lysed in ice-cold radioimmunoprecipitation assay buffer (Beyotime Institute of Biotechnology) in accordance with the manufacturer's instructions. Total proteins were separated by electrophoresis with sodium dodecyl sulphatepolyacrylamide gels and transferred to polyvinylidene fluoride membranes (EMD Millipore Corporation), which were blocked with 5% nonfat milk for 1 hour and then incubated overnight at 4°C with primary antibodies against CARD9 (Santa Cruz Biotechnology, Inc.), p65/NF-κB (Abcam), and p-p65/NF-κB (Abcam) followed by secondary antibodies for 1 hour. After washing three times with PBS + 0.1% Tween 20 detergent, the relative quantities of proteins were visualized using a bioimaging analysis system (Bio-Rad Laboratories). Protein concentrations were normalized to β-actin as an endogenous control.

| ELISA
After treatment, blood of SD rats and culture medium of RAW 264.7 cells were collected. The serum and culture medium were stored in a freezer at −80°C. The level of TNF-α, IL-6, IL-10, Arg1, and TNF-stimulated gene-6 (TSG-6) in serum and TSG-6 in culture medium were measured using ELISA kits (eBioscience) in accordance with the manufacturer's instructions.

| Statistical analysis
The data are presented as the mean ± standard deviation of three independent experiments. Statistical differences were assessed using one-way analysis of variance followed by the Student-Newman-Keuls test with IBM SPSS Statistics for Windows, version 22.0 (IBM Corporation). All tests were two-tailed and a probability (P) value of <.05 was considered statistically significant.

| MSCs ameliorated the severity of pancreatitis in SAP rats
Successful establishment of a rat model of SAP was confirmed by analyses of amylase levels and histological changes to pancreatic tissues. As shown in Figure 1A, pathological changes and scores of the pancreatic tissues were significantly increased in the SAP group. Transplantation of MSCs through the caudal vein alleviated interstitial edema, hemorrhage, inflammatory cell infiltration, and focal necrosis of pancreatic tissues in the MSC group. TNF-α and IL-6 levels were correlated with the severity of SAP and pancreatic injury. As shown in Figure 1B, serum levels of TNF-α and IL-6 were significantly increased in the SAP group and alleviated by transplantation of MSCs. There were statistically significant differences in all comparisons (P < .05). These data confirmed that MSCs reduced the severity of pancreatitis in SAP rats.

| MSCs inhibited CARD9 expression in SAP rats
Our previous studies demonstrated that CARD9 expression was significantly positively correlated with the severity of SAP. In this study, the expression levels of CARD9 in peritoneal macrophages and pancreatic tissues were measured to determine the effect of MSCs on CARD9 expression. As shown in Figure 2, the mRNA and protein levels of CARD9 were significantly increased in peritoneal macrophages and pancreatic tissues of SAP rats, Interestingly, administration of MSCs significantly decreased the increased mRNA and protein levels of CARD9, indicating that MSCs can regulate the expression of CARD9 in SAP rats.

| MSCs inhibited M1 polarization of peritoneal macrophages in SAP rats
M1 macrophages mainly express the proinflammatory cytokines TNF-α and IL-6, while M2 macrophages mainly express the immunosuppressive cytokines IL-10 and Arg1. RT-PCR and ELISA were used to indirectly determine the polarization state of macrophages in vivo. As shown in Figure 3A, the expression levels of TNF-α, IL-6, IL-10, and Arg1 were significantly increased in the peritoneal macrophages and serum of SAP rats. Meanwhile, treatment with MSCs decreased the expression levels of TNF-α and IL-6, and pretreatment with MSCs increased the expression levels of IL-10 and Arg1. Flow cytometry was used to directly determine the polarization state of macrophages in vivo. As shown in Figure 3B, the expression levels of CCR2+ and CD206+ were significantly increased in the peritoneal macrophages in SAP rats and CCR2+ expression was notably elevated. Meanwhile, treatment with MSCs decreased the expression levels of CCR2+ and increased the expression of CD206+ in MSCs group. Collectively, these results indicate that the peritoneal macrophages of SAP rats mainly underwent M1 polarization, while the addition of MSCs reversed this trend.

| Coculture with MSCs inhibited M1 polarization in vitro
CARD9 shRNA and LPS + IFNγ were used in vitro with an MOI of 10 for follow-up tests. To determine changes in the polarization of RAW 264.7 cells induced by MSCs, western blot analysis was performed to detect changes in CD86+ and CD206+ expression patterns on the surfaces of RAW 264.7 cells. As demonstrated in Figure 4A, the in vitro results showed that the cell surface expression levels of CD86+ and CD206+ were significantly higher in the NC1 group as compared with NC group, while expression of CD86+ was significantly decreased and that of CD206+ was significantly increased in the NC2 group as compared with the NC1 group. Comparisons of the KD1 and NC1 groups revealed that adenovirus interference of CARD9 expression inhibited M1 polarization of RAW 264.7 cells, while comparisons of the NC2 and KD1 groups showed that MSCs and CARD9 shRNA had no significant effect on CD86+ and CD206+ expression by RAW 264.7 cells. In the KD2 group, the combination of MSCs and CARD9 shRNA did not increase the inhibition of CD86+ expression as compared with MSCs or CARD9 shRNA alone, while the combination of MSCs and CARD9 shRNA significantly enhanced expression of CD206+ by RAW 264.7 cells. These findings are consistent with the western blot F I G U R E 2 Mesenchymal stem cells (MSCs) inhibited CARD9 expression in SAP rats. Protein and mRNA expression levels of CARD9 in the (A) pancreatic tissues and (B) peritoneal macrophages of each group. The mRNA and protein expression of CARD9 were significantly increased in peritoneal macrophages and pancreatic tissues of the SAP group compared with the control group, and the level of CARD9 was decreased compared with the SAP group in the MSC group (*P < .05: SAP group vs MSC group; #P <. 05: SAP or MSCs group vs control group).

F I G U R E 3 Mesenchymal stem cells (MSCs) inhibited M1 polarization of peritoneal macrophages in vivo.
A, mRNA expression of TNF-α, IL-6, Arg1, and IL-10 in peritoneal macrophages. TNF-α and IL-6 were mainly secreted by M1 macrophages, IL-10 and Arg1 were mainly secreted by M2 macrophages. The expression of TNF-α and IL-6 were significantly increased in the peritoneal macrophages and serum of the SAP group, after the injection of MSCs, all these cytokines were decreased in the MSC group, while the levels of IL-10 and Arg1 were increased in the MSC group. B, Flow cytometry of isolated peritoneal macrophages in each group. CCR2+ and CD206+ are markers of M1 and M2 macrophages, respectively. The levels of CCR2+ and CD206+ were significantly increased in the peritoneal macrophages in the SAP group and CCR2+ expression was notably elevated. After the injection of MSCs,the expression level of CCR2+ was decreased, while the expression level of CD206+ was increased in the MSC group (*P < .05: SAP group vs MSC group; #P < .05: SAP or MSCs group vs control group). and RT-PCR results of changes to the expression patterns of TNF-α, IL-6, IL-10, and Arg1 in RAW 264.7 cells among the different groups used to indirectly determine the polarization state of macrophages in vitro. As shown in Figure 4B, the expression levels of TNF-α, IL-6, IL-10, and Arg1 mRNA were significantly higher in the NC1 group than the NC group, while the mRNA expression levels of TNF-α and IL-6 were significantly decreased and those of IL-10 and Arg1 were significantly increased in the NC2 group as compared with the NC1 group. Furthermore, comparisons of the KD1 and NC1 groups revealed that adenovirus interference of CARD9 expression inhibited F I G U R E 4 Coculture with mesenchymal stem cells (MSCs) inhibited M1 polarization of macrophages in vitro. A, Expression of CD86 and CD206 protein in RAW 264.7 cells in each group. CD86+ and CD206+ are the surface markers of M1 and M2 macrophages, respectively. Western blot was used to detect the above markers. The levels of CD86+ and CD206+ were significantly higher in the NC1 group as compared with the NC group, and the level of CD86+ was significantly decreased in the NC2 group which was cocultured with MSCs, as compared with the NC1 group. This change was consistent with the changes in the CARD9 shRNA group and the KD2 group, suggesting that MSCs could inhibit the M1 polarization of RAW264.7 macrophages, and CARD9 was involved in the process. B, Expression of TNF-α, IL-6, Arg1, and IL-10 mRNA in RAW 264.7 cells in each group. The levels of TNF-α and IL-6 were significantly decreased in the NC2 group which was cocultured with MSCs as compared with the NC1 group, and the level of IL-10 and Arg1 were significantly increased in the NC2 group; this change was consistent with the changes in the KD1 group and KD2 group. NC1: RAW264.7 + control shRNA + LPS + IFNγ; KD1: RAW264.7 + CARD9 shRNA + LPS + IFNγ; NC2: RAW264.7 + control shRNA + LPS + IFNγ + MSCs; KD2: RAW264.7 + CARD9 shRNA + LPS + IFNγ + MSCs; (#P < .05: NC1, KD1, NC2, or KD2 group vs control group; *P < .05: KD1, NC2, or KD2 group vs NC1 group; and &P < .05: KD1 or NC2 group vs KD2 group). mRNA expression of TNF-α and IL-6 mRNA and promoted that of IL-10 and Arg1 in RAW 264.7 cells, while comparisons of the NC2 and KD1 groups demonstrated that MSCs and CARD9 shRNA had no significant effect on the mRNA expression levels of TNF-α, IL-6, IL-10, and Arg1 in RAW 264.7 cells. Interestingly, in the KD2 group, the combination of MSCs and CARD9 shRNA increased inhibition of IL-6 mRNA expression as compared with CARD9 shRNA alone, while the combination of MSCs and CARD9 shRNA significantly enhanced the mRNA expression of IL-10 and Arg1 as compared with MSCs or CARD9 shRNA alone in RAW 264.7 cells. Collectively, these data demonstrate that coculture with MSCs inhibited M1 polarization in vitro.

| Coculture with MSCs downregulated CARD9 expression in vitro
Our previous studies confirmed the participation of the CARD9-NF-κB signaling pathway in SAP, while there were no significant differences in the expression levels of CD86+ and proinflammatory cytokine (TNF-α and IL-6), which indicate M1 polarization, between the KD1 and NC2 groups, suggesting that MSCs inhibit M1 polarization mainly by downregulation of the CARD9-NF-κB signaling pathway. So, the expression levels of CARD9 were measured in different groups. As shown in Figure 5A, the mRNA and protein levels of CARD9 were significantly increased in RAW 264.7 cells in the NC1 group as compared with the NC group, while treatment with MSCs or CARD9 shRNA inhibited the increased expression of CARD9 mRNA and protein in the KD1, NC2, and KD2 groups with no significant difference in CARD9 expression levels among these group. These data showed that coculture with MSCs downregulated CARD9 expression in vitro similar to the in vivo results.

| MSCs inhibited activation of NF-kB in vivo and in vitro
Our previous studies confirmed that shRNA-induced downregulation of CARD9-inhibited activation of NF-kB, reduced the release of inflammatory factors, and alleviated the inflammatory response in SAP. The results of the present study confirmed that MSCs inhibited CARD9 expression, thus the expression profile of NF-κB was investigated both in vivo and in vitro.
The focus of the in vivo investigation was differences in NF-κB mRNA and protein expression levels among the groups using RT-PCR and western blot analysis to further elucidate the mechanism underlying the ability of MSCs to alleviate SAP. As shown in Figure 5B, as compared with control group, the NF-κB mRNA and phosphorylated protein levels were significantly increased in the peritoneal macrophages in SAP rats, while treatment with MSCs inhibited these increases.
In vitro, we continued to detect the expression of NF-κB among the different groups. As shown in Figure 5C, as compared with the NC group, the NF-κB mRNA and phosphorylated protein expression levels were significantly increased in RAW 264.7 cells in the NC1 group, while treatment with MSCs or CARD9 shRNA inhibited these increases in the KD1, NC2, and KD2 groups. Interestingly, in the KD2 group, the combination of MSCs and CARD9 shRNA did not increase the inhibition of CARD9 mRNA and protein levels as compared with MSCs or CARD9 shRNA alone. These data showed that MSCs can effectively inhibit activation of NF-kB and noncontact coculture with MSCs downregulated NF-κB in vitro.

| MSCs regulated the inflammatory response of RAW 264.7 cells by secreting TSG-6 in vitro
Many studies have confirmed that MSCs exert biological effects by secreting cytokines. In this study, in the noncontact coculture environment, MSCs regulated the inflammatory response of RAW 264.7 cells, suggesting that the functions MSCs are dependent of the secretion of certain cytokines. Hence, the protein levels of TSG-6 in culture medium of RAW 264.7 cells were detected in the different groups. As shown in Figure 5D, TSG-6 expression was significantly higher in the NC2 and KD2 groups cocultured with MSCs than in the NC, NC1, and KD1 groups, indicating that TSG-6 secreted by MSCs regulate the inflammatory response of RAW 264.7 cells in vitro.

| DISCUSSION
The morbidity and mortality of patients with SAP remains high. 20 Owing to the lack of an effective and specific treatment for SAP, there is an urgent need for a better understanding of the pathophysiology of SAP. Many studies have confirmed that transplantation of MSCs conveys therapeutic effects in a variety of diseases, 21-23 demonstrating that MSCs have great potential in cell therapy because of the relatively low immunogenicity and immune regulation as well as the capacity of multidirectional differentiation, directional migration, tissue repair, and inhibition of inflammation. 24 MSCs can reduce the severity of SAP by interfering with the inflammatory response, 25 although the specific mechanism underlying the effect of MSCs on SAP remains unclear.
Numerous studies have confirmed that the activation of mononuclear macrophages plays an important role in the occurrence and development of SAP, as macrophages exhibit different functions in different microenvironments (ie, polarization). Proinflammatory mediators produced and secreted by M1 macrophages, such as TNF-α and IL-6, play an important role in the pathology of SAP. 26 Conversely, IL-10 and Arg1 are anti-inflammatory cytokines produced and secreted by M2 macrophages that can attenuate SAP. 27,28 In vivo, we investigated the possible mechanism underlying the ability of MSCs to alleviate SAP by intravenous administration of MSCs in a rat model of SAP. The results demonstrated that pancreatic injury was prominently improved as compared with the untreated control group. Transplantation of MSCs reduced CCR2 expression and production of the inflammatory cytokines TNF-αand IL-6. These findings, together the pathological and microscopic results, suggest that SAP could be mitigated by MSCs via suppression of M1 polarization. Therefore, an anti-inflammation approach could be useful for treatment of SAP. The results of this study showed that MSCs obviously decreased the expression levels of proinflammatory cytokines and increased those of anti-inflammatory factors. Furthermore, MSCs relieved SAP by decreasing the inflammatory response via inhibition of M1 polarization. Our previous studies confirmed the participation of the CARD9-NF-κB signaling pathway in peritoneal macrophages in SAP and inhibition of CARD9 expression inhibited the transcription activity of NF-κB and reduced the production and release of cytokines, such as TNF-α and IL-6, thereby reducing inflammation. [16][17][18] Therefore, the levels of CARD9 and p-NF-κB in peritoneal macrophages were detected. Treatment with MSCs inhibited CARD9 and p-NF-κB expression by peritoneal macrophages in SAP rats, indicating that MSCs inhibit M1 polarization by downregulation of the CARD9-NF-κB signaling pathway in vivo.
In order to determine whether MSCs regulate the polarization of macrophages through the CARD9-NF-κB signaling pathway, coculture and adenovirus interference studies were conducted in vitro. After establishment of an M1 macrophage model, CARD9 shRNA was transfected into M1 macrophages, which confirmed that CARD9 shRNA interfered with CARD9 expression, resulting in significant decreases in CD86+ expression on the surface of M1 macrophages, phosphorylation of NF-κB, and secretion of the proinflammatory cytokines TNF-α and IL-6 by M1 macrophages. These results confirmed that downregulation of the CARD9-NF-κB signaling pathway inhibited M1 polarization of macrophages. Interestingly, MSCs had similar effects as CARD9 shRNA in regulating the expression of CARD9, phosphorylation of NF-κB, and downstream secretion of proinflammatory cytokines and CD86+. Although there was no significant difference in the expression of CARD9, phosphorylation of NF-κB, and the expression levels of the proinflammatory cytokines and CD86+ between the KD1 and NC2 groups, the combination of MSCs and CARD9 shRNA did not increase the inhibition of CARD9 or expression of the proinflammatory cytokines and CD86+ in the KD2 group, as compared with MSCs or CARD9 shRNA alone. Collectively, these results showed that MSCs mainly regulate M1 polarization through the CARD9-NF-κB signaling pathway both in vivo and in vitro.
Macrophages, as the main effector cells, play a major role in the proinflammatory and anti-inflammatory imbalance in acute pancreatitis. 10 M1 macrophages kill intracellular pathogens and express the proinflammatory cytokines TNF-α and IL-6, which trigger and amplify the inflammatory response. M2 macrophages mainly produce immunosuppressive cytokines, such as IL-10 and Arg1, which are involved in wound repair and the immune response. 2,3,10 Previous studies have confirmed that MSCs play an anti-inflammatory role in the inflammatory response of SAP. However, there is a lack of understanding of how MSCs inhibit pancreatic inflammation. 10 The results of the present study found that MSCs can inhibit M1 polarization of macrophages and play an anti-inflammatory role by downregulating the CARD9-NF-κB signaling pathway. Notably, the expression levels of CD206+ and Arg1 were significantly elevated in response to treatment with MSCs. In addition, F I G U R E 5 Coculture with mesenchymal stem cells (MSCs) downregulated the CARD9-NF κB signaling pathway in vivo and in vitro. A, MSCs inhibited CARD9 mRNA and protein expression in RAW 264.7 cells in each group. The mRNA and protein levels of CARD9 were significantly increased in RAW 264.7 cells in the NC1 group as compared with the NC group, while treatment with MSCs or CARD9 shRNA could decrease the expression of CARD9 mRNA and protein in the KD1, NC2, and KD2 groups (#P < .05: NC1, KD1, NC2, or KD2 group vs control group; *P < .05: KD1, NC2, or KD2 group vs NC1 group; and &P < .05: KD1 or NC2 group vs KD2 group). B, MSCs inhibited phosphorylation of p65/NF-κB in each group in vivo. The NF-κB mRNA and phosphorylated protein levels were significantly increased in the peritoneal macrophages in SAP rats, and treatment with MSCs inhibited these increases in the MSC group (*P < .05: SAP group vs MSC group). C, MSCs inhibited mRNA expression of p65/NF-κB and phosphorylation of p65/NF-κB in vitro. As compared with the NC group, the NF-κB mRNA and phosphorylated protein expression levels were significantly increased in the NC1 group, while treatment with MSCs or CARD9 shRNA inhibited these increases in the KD1, NC2, and KD2 groups (#P < .05: NC1, KD1, NC2, or KD2 group vs control group; *P < .05: KD1, NC2, KD2 group vs NC1 group; and &P < .05: KD1 or NC2 group vs KD2 group). D, Serum levels of TSG-6 in each group. TSG-6 expression was significantly higher in the NC2 and KD2 groups cocultured with MSCs than in the NC, NC1, and KD1 groups. NC1: RAW264.7 + control shRNA + LPS + IFNγ; KD1: RAW264.7 + CARD9 shRNA + LPS + IFNγ; NC2: RAW264.7 + control shRNA + LPS + IFNγ + MSCs; KD2: RAW264.7 + CARD9 shRNA + LPS + IFNγ + MSCs; (#P < .05: NC1, KD1, NC2, or KD2 group vs control group; *P < .05: KD1, NC2, or KD2 group vs NC1 group). the expression levels of CCR2+, IL-10, and Arg1 were significantly increased in the NC2 and KD2 groups, similar to the effect of treatment with MSCs in vitro. These results confirmed that MSCs not only inhibit M1 polarization, but also promote M2 polarization of macrophages, thus improving the pancreatic inflammatory response in SAP.
Current studies on the mechanism of MSCs underlying the regulation of macrophage polarization have mainly focused on the following aspects: (a) MSCs secrete exosomes that promote the polarization of M2 macrophages and reduce the polarization of M1 macrophages; (b) MSCs secrete extracellular vesicles and apoptotic bodies that promote M2 polarization of macrophages in tissue repair; and (c) MSCs secrete the cytokines TSG-6 and hepatocyte growth factor that participate in macrophage polarization. Previous studies have reported that MSCs play a biological role in SAP mainly by homing to damaged tissues. 10 In the noncontact coculture system in this study, MSCs significantly increased TSG-6 expression in the culture medium of RAW 264.7 cells, indicating that MSCs secrete cytokines that regulate the polarization of macrophages in SAP.

| CONCLUSION
Transplantation of MSCs is potentially an effective treatment for SAP. MSCs act on peritoneal macrophages by secreting cytokines similar to TSG-6 that downregulate the CARD9-NF-κB signaling pathway to inhibit M1 polarization and promote M2 polarization of macrophages, thereby reducing the inflammatory response in SAP.