Antifibrotic TSG-6 Expression Is Synergistically Increased in Both Cells during Coculture of Mesenchymal Stem Cells and Macrophages via the JAK/STAT Signaling Pathway

The pro-inflammatory cytokines tumor necrosis factor-alpha (TNF-α) and interleukin (IL)-1β upregulate TNF-α-stimulated gene 6 (TSG-6); however, current knowledge about the optimal conditions for TSG-6 expression in mesenchymal stem cells (MSCs) is limited. Here, we investigated whether TSG-6 expression varies depending on the polarization state of macrophages co-cultured with adipose tissue-derived stem cells (ASCs) and analyzed the optimal conditions for TSG-6 expression in ASCs. TSG-6 expression increased in ASCs co-cultured with M0, M1, and M2 macrophages indirectly; among them, M1 macrophages resulted in the highest increase in TSG-6 expression in ASCs. TSG-6 expression in ASCs dramatically increased by combination (but not single) treatment of TNF-α, IL-1β, interferon-gamma (IFN-γ), and lipopolysaccharide (LPS). In addition, phosphorylation of signal transducer and activator of transcription (STAT) 1/3 was observed in response to IFN-γ and LPS treatment but not TNF-α and/or IL-1β. STAT1/3 activation synergistically increased TNF-α/IL-1β-dependent TSG-6 expression, and JAK inhibitors suppressed TSG-6 expression both in ASCs and macrophages. In LX-2 hepatic stellate cells, TSG-6 inhibited TGF-β-induced Smad3 phosphorylation, resulting in decreased α-smooth muscle actin (SMA) expression. Moreover, fibrotic activities of LX-2 cells induced by TGF-β were dramatically decreased after indirect co-culture with ASCs and M1 macrophages. These results suggest that a comprehensive inflammatory microenvironment may play an important role in determining the therapeutic properties of ASCs by increasing TSG-6 expression through STAT1/3 activation.

As a multifunctional protein, TSG-6 can regulate the growth of fibroblasts by suppressing the TGF-β signaling pathway, leading to antifibrotic outcomes [31][32][33]. rTSG-6 significantly decreases the viability and proliferation of capsule fibroblasts by suppressing the TGF-β/Smad2 signaling pathway in frozen shoulder and enhances cellular apoptosis concurrent with a reduction in Bcl-2 expression [31]. TSG-6 inhibits the proliferation of keloid fibroblasts by blocking the formation of the Smad2/3/4 complex and its nuclear translocation [33]. In addition, TSG-6 contributes to liver regeneration by suppressing the activation of hepatic stellate cells in CCl 4 -treated mice, suggesting that TSG-8 may have therapeutic potential in acute liver failure [23].
Taken together, MSCs exposed to an inflammatory milieu can regenerate damaged tissues by expressing TSG-6 to regulate the activity of inflammatory cells. Although inflammatory cytokines, including TNF-α and IL-1β, can increase TSG-6 expression, little is known about the relationship of TSG-6 expression between MSCs and macrophages. We hypothesized that there may be a difference in the expression of TSG-6 in MSCs associated with macrophages at different transition states. If this is the case, it is expected that the optimal timing of MSC transplantation in regenerative medicine can be determined based on exposure to inflammatory factors or macrophage transition status. Therefore, in this study, we want to identify which transition status of macrophages may be involved in TSG-6 expression in ASCs, which types of cytokines are important for TSG-6 expression, and which signaling pathways can regulate TSG-6 expression. To explore these issues, we investigated the expression of TSG-6 in both ASCs and M0, M1, and M2 macrophages after indirectly co-culturing them and the optimal condition of TSG-6 expression in ASCs. In addition, focusing on the anti-fibrotic effects of TSG-6, we investigated whether ASCs and macrophages expressing TSG-6 were able to alleviate TGF-β-induced fibrosis of hepatic stellate cells.

Increased Expression of TSG-6 mRNA in ASCs after Coculture with Macrophages
Since MSCs induce TSG-6 expression in an inflammatory milieu, we investigated whether TSG-6 expression in MSCs differs according to macrophage transition status. ASCs were mono-or cocultured with macrophages and treated with IFN-γ and LPS, or IL-4 and IL-13, which promote macrophage transition to M1 or M2 phenotypes, respectively. TSG-6 expression was analyzed by real-time quantitative polymerase chain reaction (qPCR). Macrophages, regardless of their transition status, increased TSG-6 expression in ASCs. M0 and M2 macrophages increased TSG-6 expression by 5-and 15-fold, respectively, whereas M1 macrophages significantly increased TSG-6 expression in ASCs by approximately 62-fold ( Figure 1). These results suggest that coculture with macrophages increases the expression of TSG-6 in ASCs and that M1 macrophages or pro-inflammatory conditions may play a particularly important role in TSG-6 expression in ASCs.

Increased Expression of TSG-6 mRNA in ASCs after Coculture with Macrophages
Since MSCs induce TSG-6 expression in an inflammatory milieu, we investiga whether TSG-6 expression in MSCs differs according to macrophage transition sta ASCs were mono-or cocultured with macrophages and treated with IFN-γ and LPS IL-4 and IL-13, which promote macrophage transition to M1 or M2 phenotypes, resp tively. TSG-6 expression was analyzed by real-time quantitative polymerase chain r tion (qPCR). Macrophages, regardless of their transition status, increased TSG-6 exp sion in ASCs. M0 and M2 macrophages increased TSG-6 expression by 5-and 15-f respectively, whereas M1 macrophages significantly increased TSG-6 expression in A by approximately 62-fold ( Figure 1). These results suggest that coculture with ma phages increases the expression of TSG-6 in ASCs and that M1 macrophages or pro flammatory conditions may play a particularly important role in TSG-6 expression ASCs. Next, we analyzed whether macrophages were essential for TSG-6 expression ASCs and whether TSG-6 was expressed in macrophages. ASCs and macrophages w mono-or cocultured and treated with IFN-γ and LPS (macrophage transition factors p moting the M1 phenotype). In ASCs treated with IFN-γ and LPS, TSG-6 expression increased by approximately 21-fold, while ASCs cocultured with M0 macrophages ex ited an approximately 10-fold increase in TSG-6 expression. Interestingly, when A were cocultured with M1 macrophages (differentiated with IFN-γ and LPS), TS mRNA increased approximately 63-and 46-fold in ASCs and M1 macrophages, resp tively ( Figure 2A black bars). Similar to the pattern of TSG-6 mRNA expression show Figure 2A, TSG-6 expression significantly increased in both cell types when M1 ma phage and ASCs were cocultured ( Figure 2B). Since a marked upregulation of TS mRNA was observed in phorbol ester-induced differentiation of THP-1 monocyte macrophages [34], we analyzed whether macrophages specifically were necessary TSG-6 expression in ASCs using murine macrophage cells Raw 264.7 to eliminate the of phorbol esters for transdifferentiation. Similar to M1 macrophages differentiated fr THP-1, Raw264.7 cells treated with LPS significantly induced TSG-6 expression in A ( Figure 2C). These results suggest that both macrophages and MSCs can overexpress T 6 if they coexist in an inflammatory or injured site. All qPCR reactions were performed in triplicate. GAPDH expression was used for normalization. The 2 −(∆∆Ct) method was used to calculate relative fold changes in mRNA expression. Data are presented as the mean ± standard deviation (SD) of three independent experiments. ** p ≤ 0.01 and *** p < 0.001. M0, macrophages differentiated from THP-1 monocytes; M1, M0 macrophages treated with IFN-γ and LPS; M2, M0 macrophages treated with IL-4 and IL-13.
Next, we analyzed whether macrophages were essential for TSG-6 expression in ASCs and whether TSG-6 was expressed in macrophages. ASCs and macrophages were monoor cocultured and treated with IFN-γ and LPS (macrophage transition factors promoting the M1 phenotype). In ASCs treated with IFN-γ and LPS, TSG-6 expression was increased by approximately 21-fold, while ASCs cocultured with M0 macrophages exhibited an approximately 10-fold increase in TSG-6 expression. Interestingly, when ASCs were cocultured with M1 macrophages (differentiated with IFN-γ and LPS), TSG-6 mRNA increased approximately 63-and 46-fold in ASCs and M1 macrophages, respectively ( Figure 2A black bars). Similar to the pattern of TSG-6 mRNA expression shown in Figure 2A, TSG-6 expression significantly increased in both cell types when M1 macrophage and ASCs were cocultured ( Figure 2B). Since a marked upregulation of TSG-6 mRNA was observed in phorbol ester-induced differentiation of THP-1 monocytes to macrophages [34], we analyzed whether macrophages specifically were necessary for TSG-6 expression in ASCs using murine macrophage cells Raw 264.7 to eliminate the use of phorbol esters for transdifferentiation. Similar to M1 macrophages differentiated from THP-1, Raw264.7 cells treated with LPS significantly induced TSG-6 expression in ASCs ( Figure 2C). These results suggest that both macrophages and MSCs can overexpress TSG-6 if they coexist in an inflammatory or injured site.

TSG-6 Expression in ASCs and Macrophages Treated with Cytokine Combinations
Since TNF-α and IL-1β secreted by macrophages can induce TSG-6 expression, we investigated whether IFN-γ and LPS, TNF-α, or IL-1β play a role in the increase of TSG-6 expression in ASCs. In ASCs treated individually with IFN-γ and LPS, TNF-α, or IL-1β, elevated TSG-6 levels were only detected in immunoblots exposed for a long period (300 s). However, in ASCs treated with a combination of IFN-γ, LPS, and TNF-α, a significant increase in TSG-6 expression was observed even in immunoblots exposed for short periods (22 s), with this increase dependent on the doses of TNF-α ( Figure 3A, lanes 5 and 6) but not IL-1β ( Figure 3A, lanes 7 and 8). However, maximum TSG-6 expression was observed when cells were treated with all four factors ( Figure 3A, lanes 9 and 10). These results suggest that the synergistic action of several inflammatory cytokines is required for optimal expression of TSG-6 in ASCs. In contrast, increased TSG-6 expression in macrophages was observed in response to combinatorial treatment with IFN-γ, LPS, and TNF-α, compared to cells treated with IFN-γ and LPS ( Figure 3B, lane 5). The expression pattern of TSG-6 secreted into the culture supernatant was quite similar to that observed in the cell lysate ( Figure 3). These results suggest that the conditions for inducing TSG-6 expression differ according to cell types, and that ASCs can express TSG-6 by responding more sensitively to inflammatory conditions than do macrophages.

TSG-6 Expression in ASCs and Macrophages Treated with Cytokine Combinations
Since TNF-α and IL-1β secreted by macrophages can induce TSG-6 expression, we investigated whether IFN-γ and LPS, TNF-α, or IL-1β play a role in the increase of TSG-6 expression in ASCs. In ASCs treated individually with IFN-γ and LPS, TNF-α, or IL-1β, elevated TSG-6 levels were only detected in immunoblots exposed for a long period (300 sec). However, in ASCs treated with a combination of IFN-γ, LPS, and TNF-α, a significant increase in TSG-6 expression was observed even in immunoblots exposed for short periods (22 sec), with this increase dependent on the doses of TNF-α ( Figure 3A, lanes 5 and ASCs and RAW 264.7 cells were seeded in lower and upper chamber, respectively, and the TSG-6 expression in ASCs was analyzed using immunoblotting. Relative expression was normalized with respect to GAPDH expression. Data are presented as the mean ± SD of three independent experiments. * p ≤ 0.05, ** p ≤ 0.01, and *** p < 0.001. TW, Transwell; A, ASCs; M, macrophages; R, RAW 264.7 murine macrophages.

Activation of the Signal Transducer and Activator of Transcription (STAT) Signaling Pathway for TSG-6 Expression
Next, we investigated whether the STAT pathway, which is activated by IFN-γ and LPS, is involved in TSG-6 expression.

Activation of the Signal Transducer and Activator of Transcription (STAT) Signaling Pathway for TSG-6 Expression
Next, we investigated whether the STAT pathway, which is activated by IFN-γ an LPS, is involved in TSG-6 expression.
In both ASCs and macrophages, IFN-γ + LPS treatment, but not TNF-α or IL-1β trea ment, induced STAT1/3 phosphorylation at 15 min (Figure 4). The Janus kinase (JAK) in hibitor that inhibits STAT activity reduced TSG-6 expression induced by the cytokin combination treatment ( Figure 5A). In addition, this decrease in TSG-6 expression wa also observed in both ASCs and macrophages cocultured indirectly ( Figure 5B,C). Thes results suggest that the STAT signaling pathway induced TSG-6 expression in ASCs an macrophages. Furthermore, STAT activity induced a synergistic expression of TSG-6 dur ing TNF-α + IL-1β treatment. Taken together, IFN-γ + LPS, TNF-α, or IL-1β can induc TSG-6 expression in ASCs and macrophages, and when ASCs and macrophages are ex posed to TNF-α or IL-1β under STAT-activated conditions, TSG-6 overexpression ma occur. In both ASCs and macrophages, IFN-γ + LPS treatment, but not TNF-α or IL-1β treatment, induced STAT1/3 phosphorylation at 15 min (Figure 4). The Janus kinase (JAK) inhibitor that inhibits STAT activity reduced TSG-6 expression induced by the cytokine combination treatment ( Figure 5A). In addition, this decrease in TSG-6 expression was also observed in both ASCs and macrophages cocultured indirectly ( Figure 5B,C). These results suggest that the STAT signaling pathway induced TSG-6 expression in ASCs and macrophages. Furthermore, STAT activity induced a synergistic expression of TSG-6 during TNF-α + IL-1β treatment. Taken together, IFN-γ + LPS, TNF-α, or IL-1β can induce TSG-6 expression in ASCs and macrophages, and when ASCs and macrophages are exposed to TNF-α or IL-1β under STAT-activated conditions, TSG-6 overexpression may occur.      (A) TSG-6 expression in ASCs after treatment with TNF-α, IL-1β, and/or IFN-γ + LPS. TSG-6 expression in ASCs (B) or macrophages (C) after coculture. The data shown represent one of three independent experiments. ** p ≤ 0.01 and *** p < 0.001. Inh., inhibitor.

Antifibrotic Effects of TSG-6 Expressed in ASCs and Macrophages
In addition to its representative anti-inflammatory properties, TSG-6 may modulate fibrosis by regulating the TGF-β/Smad pathway [31,33]. Therefore, we investigated whether TSG-6 expressed by ASCs and macrophages could regulate the fibrotic process in LX-2 cells induced by TGF-β. In LX-2 cells, TGF-β induced Smad3 phosphorylation and α-SMA expression, which were significantly decreased by TSG-6 ( Figure 6A,B). In addition, α-SMA expression induced by TGF-β was significantly reduced in LX-2 cells cocultured with ASCs and macrophages indirectly. We used phorbol ester 12-O-tetradecanoylphorbol-13-acetate (TPA) to differentiate THP-1 monocytes into macrophages; however, TPA is an activator of protein kinase C (PKC) that can regulate the TGF/Smad pathway [35,36]. Therefore, to confirm the antifibrotic effects of TSG-6 alone, excluding the effect of TPA in the coculture of ASCs and macrophages, Raw 264.7 macrophages were used. As shown in Figure 6D, ASCs and Raw 264.7 cells induced a decrease in α-SMA expression in LX-2 cells compared to TGF-β-treated cells.

Antifibrotic Effects of TSG-6 Expressed in ASCs and Macrophages
In addition to its representative anti-inflammatory properties, TSG-6 may mod fibrosis by regulating the TGF-β/Smad pathway [31,33]. Therefore, we investig whether TSG-6 expressed by ASCs and macrophages could regulate the fibrotic pro in LX-2 cells induced by TGF-β. In LX-2 cells, TGF-β induced Smad3 phosphorylation α-SMA expression, which were significantly decreased by TSG-6 ( Figure 6A,B). In a tion, α-SMA expression induced by TGF-β was significantly reduced in LX-2 cells c tured with ASCs and macrophages indirectly. We used phorbol ester 12-O-tetrad noylphorbol-13-acetate (TPA) to differentiate THP-1 monocytes into macrophages; h ever, TPA is an activator of protein kinase C (PKC) that can regulate the TGF/Smad p way [35,36]. Therefore, to confirm the antifibrotic effects of TSG-6 alone, excluding effect of TPA in the coculture of ASCs and macrophages, Raw 264.7 macrophages used. As shown in Figure 6D, ASCs and Raw 264.7 cells induced a decrease in α-S expression in LX-2 cells compared to TGF-β-treated cells.
Next, to determine whether the reduction of α-SMA expression in LX-2 was dep ent on TSG-6 expressed by ASCs and macrophages, we investigated the expression SMA after TSG-6 silencing using siRNA. ASCs and Raw 264.7 macrophages in the u chamber were treated with scrambled or TSG-6-specific siRNA and then cocultured LX-2 cells. TSG-6 expression was significantly reduced in TSG-6 siRNA-treated ASCs Raw 264.7 macrophages (Figure 7A), which resulted in the recovery of α-SMA expres in TGF-β-exposed LX-2 cells ( Figure 7B).
These results suggest that TSG-6 can decrease α-SMA expression by inhib Smad3 phosphorylation in LX-2 cells, and TSG-6 expressed by ASCs and macroph regulates the fibrotic activity of LX-2 cells.  Next, to determine whether the reduction of α-SMA expression in LX-2 was dependent on TSG-6 expressed by ASCs and macrophages, we investigated the expression of α-SMA after TSG-6 silencing using siRNA. ASCs and Raw 264.7 macrophages in the upper chamber were treated with scrambled or TSG-6-specific siRNA and then cocultured with LX-2 cells. TSG-6 expression was significantly reduced in TSG-6 siRNA-treated ASCs and Raw 264.7 macrophages ( Figure 7A), which resulted in the recovery of α-SMA expression in TGF-βexposed LX-2 cells ( Figure 7B).

Discussion
Although MSCs can express TSG-6, we found that M1 macrophages induced the highest increase in TSG-6 expression in ASCs. In addition, combination treatments of IFNγ + LPS, TNF-α, and IL-1β were responsible for a significant increase in TSG-6 expression in ASCs via phosphorylation of STAT1/3. STAT1/3 phosphorylation was observed in response to treatment with IFN-γ and LPS, but not TNF-α and/or IL-1β. STAT1/3 activation synergistically increased TNF-α/IL-1β-dependent TSG-6 expression, and JAK inhibitors suppressed TSG-6 expression in both ASCs and macrophages. TSG-6 inhibited TGF-βinduced Smad3 phosphorylation and resulted in decreased α-SMA expression. Moreover, fibrotic activities of LX-2 cells induced by TGF-β were dramatically decreased after indirect coculture with ASCs and M1 macrophages.
Although inflammatory mediators (e.g., IL-1β, TNF-α, LPS, TGF-β, and PGE2) have been shown to induce TSG-6 expression in leukocytes, stromal cells, and several tissues, optimal conditions of TSG-6 expression in each cell type have not been accurately identified [10,18,37,38]. In this study, a significant increase in TSG-6 expression was observed in both cocultured ASCs and macrophages. Macrophages secrete various inflammatory cytokines including TNF-α and IL-1β at the site of inflammation, and they can induce the expression of TSG-6 by ASCs. In addition, a significant increase in TSG-6 expression was observed in cocultured ASCs treated with the combination of IFN-γ, LPS, and TNF-α/IL-1β, but not in macrophages cultured alone. These results indicate that, although the stimuli that induce TSG-6 expression in ASCs and macrophages are different, the co-localization of ASCs and macrophages is important to express TSG-6 in an inflammatory or injured site. In macrophages, stimulation other than the combination of IFN-γ, LPS, and TNF-α is required for optimal TSG-6 expression. PGE2 induces M1-to-M2 transition of macrophages [39,40] and increases TSG-6 expression [11]. PGE2 production is increased in both cells during coculture of MSCs and macrophages [40,41]. In addition to PGE2, kynurenic acid (KYNA) can increase TSG-6 expression in monocytes [42]. KYNA is an These results suggest that TSG-6 can decrease α-SMA expression by inhibiting Smad3 phosphorylation in LX-2 cells, and TSG-6 expressed by ASCs and macrophages regulates the fibrotic activity of LX-2 cells.

Discussion
Although MSCs can express TSG-6, we found that M1 macrophages induced the highest increase in TSG-6 expression in ASCs. In addition, combination treatments of IFN-γ + LPS, TNF-α, and IL-1β were responsible for a significant increase in TSG-6 expression in ASCs via phosphorylation of STAT1/3. STAT1/3 phosphorylation was observed in response to treatment with IFN-γ and LPS, but not TNF-α and/or IL-1β. STAT1/3 activation synergistically increased TNF-α/IL-1β-dependent TSG-6 expression, and JAK inhibitors suppressed TSG-6 expression in both ASCs and macrophages. TSG-6 inhibited TGF-β-induced Smad3 phosphorylation and resulted in decreased α-SMA expression. Moreover, fibrotic activities of LX-2 cells induced by TGF-β were dramatically decreased after indirect coculture with ASCs and M1 macrophages.
Although inflammatory mediators (e.g., IL-1β, TNF-α, LPS, TGF-β, and PGE2) have been shown to induce TSG-6 expression in leukocytes, stromal cells, and several tissues, optimal conditions of TSG-6 expression in each cell type have not been accurately identified [10,18,37,38]. In this study, a significant increase in TSG-6 expression was observed in both cocultured ASCs and macrophages. Macrophages secrete various inflammatory cytokines including TNF-α and IL-1β at the site of inflammation, and they can induce the expression of TSG-6 by ASCs. In addition, a significant increase in TSG-6 expression was observed in cocultured ASCs treated with the combination of IFN-γ, LPS, and TNF-α/IL-1β, but not in macrophages cultured alone. These results indicate that, although the stimuli that induce TSG-6 expression in ASCs and macrophages are different, the co-localization of ASCs and macrophages is important to express TSG-6 in an inflammatory or injured site. In macrophages, stimulation other than the combination of IFN-γ, LPS, and TNF-α is required for optimal TSG-6 expression. PGE2 induces M1-to-M2 transition of macrophages [39,40] and increases TSG-6 expression [11]. PGE2 production is increased in both cells during coculture of MSCs and macrophages [40,41]. In addition to PGE2, kynurenic acid (KYNA) can increase TSG-6 expression in monocytes [42]. KYNA is an indoleamine 2,3-dioxygenase (IDO) metabolite, which is produced during tryptophan catabolism. MSCs exposed to IFN-γ and TNF-α express IDO, which can have anti-inflammatory and immunosuppressive functions [43]. Taken together, these data indicate that different cells express TSG-6 at different intensities depending on the treatment combination of inflammatory mediators (e.g., IL-1β, TNF-α, IFN-γ, LPS, TGF-β, PGE2, and KYNA). Therefore, when MSCs are applied to regenerative medicine, the optimal therapeutic effects can be expected by transplanting MSCs in consideration of the patient's inflammatory cytokine profile or the transition state of macrophages. More than 50 cytokines, growth factors, and hormones play important roles in development, metabolism, and healing by activating STAT-mediated signaling [44][45][46][47]. In this study, unlike TNF-α or IL-1β, IFN-γ and LPS induced M1 macrophage differentiation of TPA-treated monocytes, and phosphorylation of STAT1/3 in ASCs and macrophages. Inhibitors of JAKs (signaling molecules upstream of STAT) significantly reduced TSG-6 expression in ASCs and macrophages treated with cytokine combinations or cocultured. These results suggest that the STAT-mediated signaling pathway may play a critical role in directly increasing TSG-6 expression or regulating STAT-mediated IDO or COX-2 expression. To verify this, it will be necessary to assess the changes in TSG-6 expression in the presence of IDO and COX-2 inhibitors. Therefore, we plan further studies to understand the precise mechanisms by which the STAT-mediated signaling pathways increase TSG-6 expression.
In conclusion, considering the inflammatory environment during MSC transplantations will be very important to enhance the anti-inflammatory and antifibrotic properties of TSG-6 expressed by MSCs and macrophages.

Cell Culture
The hepatic stellate cell line LX-2 was purchased from Millipore (Burlington, MA, USA), and THP-1 monocytes and Raw 264.7 murine macrophage cells were purchased from the Korea Cell Line Bank (Seoul, Korea). LX-2 or Raw 264.7 cells were maintained in Dulbecco's Modified Eagle Medium (DMEM, Gibco BRL, Rockville, MD, USA) supplemented with 3% or 10% fetal bovine serum (FBS, Gibco BRL), respectively. ASCs were isolated from three healthy donors (24-38 years of age) with their written informed consent through elective liposuction procedures under anesthesia at the Wonju Severance Christian Hospital (Wonju, Korea). ASCs were maintained in DMEM supplemented with 10% FBS, and cells from passages 3-5 were used in all experiments. THP-1 cells were sub-cultured with Roswell Park Memorial Institute (RPMI) 1640 Medium (Gibco BRL) supplemented with 10% FBS. Penicillin/streptomycin (Gibco BRL) was also supplemented in culture media for maintaining cells at 37 • C and 5% CO 2 . For experiments, cells were seeded for 24 h, and treated with stimuli for the indicated time. Cells were exposed to JAK inhibitors 20 min prior to stimulus treatment.
For indirect coculture, Transwell plates (SPL, Pocheon, Korea) were used. To evaluate the anti-fibrotic effects of TSG-6, LX-2 cells were seeded in the lower chamber, and ASCs and macrophages were directly seeded into the upper chamber. Briefly, THP-1 monocytes were differentiated into macrophages for two days in the upper chamber, and LX-2 cells were plated in the lower chamber a day before coculture. To proceed with the coculture of ASCs (1 × 10 4 cells/cm 2 ), macrophages (2.5 × 10 4 cells/cm 2 ), and LX-2 cells (1 × 10 4 cells/cm 2 ), the upper chamber seeded with macrophages was assembled over the lower chamber, and ASCs were immediately added to the upper chamber.

qPCR
Total RNA was extracted using TRIzol reagent (Gibco BRL) according to the manufacturer's instructions. cDNA was synthesized from 1 µg of total RNA using Verso cDNA synthesis kit (ThermoFisher Scientific). TSG-6 mRNA was amplified using the sense and antisense primers 5 -TGGCTTTGTGGGAAGATACTGT-3 and 5 -TGGAAACCTCCAGCTGTC AC-3 , respectively. For GAPDH, sense and antisense primers of 5 -CAAGGCTGAGAACGG GAAGC-3 and 5 -AGGGGGCAGAGATGATGACC-3 , respectively were used. The reagents in a 10-µL reaction mixture included cDNA, primer pairs, and SYBR Green PCR Master Mix (Applied Biosystems, Dublin, Ireland), and PCR was conducted using a QuantStudio 6 Flex Real-time PCR System (ThermoFisher Scientific). All qPCR reactions were performed in triplicate. GAPDH expression was used for normalization. The 2 −(∆∆Ct) method was used to calculate relative fold changes in mRNA expression.

Immunoblotting
Cells were lysed in sample buffer [62.5 mM Tris-HCl, pH 6.8, 34.7 mM sodium dodecyl sulfate (SDS), 10% (v/v) glycerol, and 5% (v/v) β-mercaptoethanol], boiled for 5 min, subjected to SDS-polyacrylamide gel electrophoresis, and transferred to an Immobilon membrane (Millipore). After blocking with 5% skim milk in Tris-HCl-buffered saline containing 0.05% (v/v) Tween 20 (TBST) for 30 min, the membrane was incubated with primary antibodies against TSG-6, α-SMA, pSmad2/3, and GAPDH at a dilution of 1:1000 or STAT1, pSTAT1, STAT3, pSTAT3, and Smad2/3 at a dilution of 1:2000 at 4 • C overnight. The membrane was washed thrice for 5 min with TBST and then incubated with horseradish peroxidase-conjugated secondary antibodies (1:5000; 7074S and 7076S, Cell Signaling Technology) for 1 h. After washing thrice with TBST, protein bands were visualized using an EZ-Western Lumi Pico or Femto kit (Dogen, Seoul, Korea) and detected using a ChemiDoc XRS+ system (Bio-Rad, Hercules, CA, USA). To detect TSG-6 expression, the membrane made from ASCs treated with TNF-α or IL-1β was long exposed (approximately 300 s), whereas the membrane obtained from the ASCs after IFN-γ, LPS, and TNF-α treatment was sufficient with a short exposure (approximately 22 s). The intensity of immunoreactive bands was quantified by densitometry using ImageJ, and relative expression of proteins was normalized with respect to GAPDH expression.

Small Interfering RNA (siRNA) Treatment
To inhibit TSG-6 expression in ASCs and macrophages, ASCs/macrophages or LX-2 cells were seeded in the upper or lower chamber of Transwell wells, respectively. TSG-6 siRNA (Santa Cruz) or scrambled siRNA (Bioneer, Daejeon, Korea) was mixed with Lipofectamine RNAiMAX (Invitrogen, Carlsbad, CA, USA) according to the manufacturer's recommendations and applied to the cells, which were incubated for two days.

Statistical Analyses
All experiments were performed thrice. Data are expressed as the mean ± standard deviation. p values were determined using a paired 2-tailed Student's t-test (Mann-Whitney U test). All statistical analyses were performed using GraphPad Prism 7.0 software (GraphPad Inc., La Jolla, CA, USA). Significance was set at p ≤ 0.05.