FAIM Enhances the Efficacy of Mesenchymal Stem Cell Transplantation by Inhibiting JNK-Induced c-FLIP Ubiquitination and Degradation

Background The poor survival rates of transplanted mesenchymal stem cells (MSCs) in harsh microenvironments impair the efficacy of MSCs transplantation in myocardial infarction (MI). Extrinsic apoptosis pathways play an important role in the apoptosis of transplanted MSCs, and Fas apoptosis inhibitory molecule (FAIM) is involved in regulation of the extrinsic apoptosis pathway. Thus, we aimed to explore whether FAIM augmentation protects MSCs against stress-induced apoptosis and thereby improves the therapeutic efficacy of MSCs. Methods We ligated the left anterior descending coronary artery (LAD) in the mouse heart to generate an MI model and then injected FAIM-overexpressing MSCs (MSCsFAIM) into the peri-infarction area in vivo. Moreover, FAIM-overexpressing MSCs were challenged with oxygen, serum, and glucose deprivation (OGD) in vitro, which mimicked the harsh microenvironment that occurs in cardiac infarction. Results FAIM was markedly downregulated under OGD conditions, and FAIM overexpression protected MSCs against OGD-induced apoptosis. MSCsFAIM transplantation improved cell retention, strengthened angiogenesis, and ameliorated heart function. The antiapoptotic effect of FAIM was mediated by cellular-FLICE inhibitory protein (c-FLIP), and FAIM augmentation improved the protein expression of c-FLIP by reducing ubiquitin–proteasome-dependent c-FLIP degradation. Furthermore, FAIM inhibited the activation of JNK, and treatment with the JNK inhibitor SP600125 abrogated the reduction in c-FLIP protein expression caused by FAIM silencing. Conclusions Overall, these results indicated that FAIM curbed the JNK-mediated, ubiquitination–proteasome-dependent degradation of c-FLIP, thereby improving the survival of transplanted MSCs and enhancing their efficacy in MI. This study may provide a novel approach to strengthen the therapeutic effect of MSC-based therapy.


Introduction
Mesenchymal stem cells (MSCs) transplantation has been shown to be safe and beneficial in the treatment of myocardial infarction (MI) due to the low immunogenicity, multidirectional differentiation ability, and robust paracrine effects of these cells [1,2]. Several studies have revealed that the main factor that impairs the therapeutic efficacy of MSCs transplantation is the poor survival rate of transplanted cells [3], which is caused by cell death induced by a harsh microenvironment and the limited self-renewal rates of MSCs [4]. A previous study reported that Fas-FasL interactions have a significant effect on death receptor (DR) activation in implanted MSCs, and that inhibiting this interaction by recombinant Fas/Fc protein treatment improves cell retention and restores cardiac function [5], suggesting that extrinsic apoptosis pathways play an important role in transplanted MSCs apoptosis and that intervening in these pathways by genetic modification may be a new strategy to augment the longevity of transplanted MSCs.
Originally identified as an inhibitor of Fas signaling in Bcell receptor-activated lymphocytes [6], Fas apoptosis inhibitory molecule (FAIM) plays an essential role in apoptosis, metabolism, cell growth, and tumorigenesis [7]. Studies have revealed two splice variants of FAIM, and the short variant is ubiquitously expressed in multiple cell types [8][9][10]. Increasing evidence indicates that FAIM exerts an antiapoptotic effect on thymocytes and hepatocytes by suppressing T-cell receptor (TCR) signaling or DR signaling [11,12]. FAIM has been implicated in the regulation of heat-and oxidative stress-induced cell death, and FAIM knockout cells exhibit impaired resistance to cell death induced by stress [13]. A recent study reported that FAIM facilitates cell proliferation by suppressing autophagy [14]. We therefore hypothesized that FAIM might be an excellent candidate to promote MSCs survival and optimize the abilities of MSCs in a hostile ischemic environment.
Cellular FLICE inhibitory protein, also known as c-FLIP, is a strong regulator of the extrinsic apoptosis pathway and competitively inhibits the binding of caspase-8 with FADD, thus avoiding the activation of caspase-8 [15,16]. c-FLIP expression is markedly decreased in thymocytes and hepatocytes in the absence of FAIM [17]. To date, however, the underlying mechanisms linking FAIM to c-FLIP expression remain unclear, necessitating further exploration.
In this study, we discovered that FAIM overexpression promoted MSCs survival and enhanced their therapeutic effect in MI. The potential mechanisms involve reducing the ubiquitination and degradation of c-FLIP mediated by impairing JNK activation. Our results provide new insights and alternative treatment strategies for stem cell transplantation.
MSCs were characterized by flow cytometry to analyze the expression of membrane markers with a mouse MSCs surface marker assay kit (#MUXMX-09011, Cyagen Biosciences, Suzhou, China). In brief, MSCs were digested with trypsin (Genom Bio., Hangzhou, China) and resuspended in Hank's balanced salt solution (HBSS) (Genom Bio.). Subsequently, the single-cell suspensions were incubated with cell surface marker antibodies at 4°C for 30 min (mesenchymal surface markers: Scr-1, CD29, CD44, and CD90; hematopoietic stem cell marker: CD117; endothelial cell marker: CD31; and isotype-matched control). After being washed with HBSS, the MSCs were incubated with fluorescenceconjugated secondary antibodies for another 30 min. Surface marker expression was analyzed with an LSRFortessa flow cytometer (BD, USA). The MSCs were positive for CD29, CD44, CD90, and Sca-1 and negative for CD31 and CD117, and the MSCs were fusiform and arranged in a swirled pattern (Supplementary Figure S1(a)-1(b)).

Lentivirus Infection.
Recombinant lentiviruses expressing mouse FAIM-S labeled with GFP and Flag (GFP-FAIM) or mouse c-FLIP labeled with GFP (GFP-FLIP) and empty lentivirus labeled with GFP (GFP-Vec) were provided by GeneChem (Shanghai, China). MSCs were transduced with lentiviruses (MOI = 50) overnight with HitransG A (Gene-Chem, Shanghai, China) after being seeded in flasks. Then, the virus-containing supernatants were replaced with fresh medium for further culture. The infection efficiency was determined by measuring the expression of the target gene and the fluorescence of GFP.

Cell Apoptosis Model.
To simulate the microenvironment of the ischemic myocardium, a glucose/serumdeprived/hypoxia-induced apoptosis model (OGD) and an H 2 O 2 -induced apoptosis model were established as previously described [18]. For the OGD model, forty-eight hours after lentivirus infection, MSCs Vec (MSCs infected with vector) or MSCs FAIM (MSCs infected with murine FAIM) were washed with PBS and cultured with glucose-free DMEM under hypoxic conditions (0.3% O 2 ) for an additional 12 h until they were harvested. An H 2 O 2 -induced apoptosis model was also used in this study. In brief, after transfection with lentivirus, MSCs were challenged with 1 mM H 2 O 2 in glucose-free DMEM for 2 h until they were harvested.
2.5. TUNEL Assays. TUNEL assays were carried out using a TUNEL staining kit (#C1090, Beyotime, Shanghai, China). In brief, cells were fixed in 10% formalin and permeabilized with 0.5% Triton X-100, followed by incubation with TUNEL staining reagent at 37°C in the dark for 1 hour. Additionally, nuclei were stained with Hoechst 33258 (#H3569, Thermo Fisher, CA, USA). Fluorescence images were acquired, and the apoptotic rates were calculated as the percentages of TUNEL-positive cells.
2.6. Annexin V/Propidium Iodide (PI) Staining. Annexin V/ PI staining was carried out using an Annexin V-APC/PI apoptosis kit (#70-AP107, Liankebio, Hangzhou, China). Briefly, the culture medium was collected, and cells were digested with trypsin. The two components were then mixed and centrifuged. Then, the supernatant was discarded, and 1× binding buffer was added to resuspend the MSCs. The single-cell suspensions were incubated with an Annexin V-APC and PI mixture for 15 min at room temperature and then quantified with an LSRFortessa flow cytometer. The  Table 2.

Western
Blotting. Whole-cell lysate was isolated using RIPA buffer (#P0013B, Beyotime) supplemented with a 1× protease and phosphatase inhibitor cocktail mixture (#78442, Thermo Fisher), and the total protein concentration was quantified with a BCA assay kit (#FD200, Fude Biological, Hangzhou, China). Proteins were separated by SDS-PAGE and then transferred onto PVDF membranes. After blocking with 5% skim milk (#LP0031B, Thermo Fisher), the membranes were incubated overnight at 4°C with the specific primary antibodies listed in Table 3, followed by HRP-linked secondary antibody incubation. Finally, the proteins were detected by an Amersham ImageQuant 800 Western blot imaging system using ECL Western blotting substrate (#FD8020, Fude Biological). The images were quantified using ImageQuant TL 8.2 analysis software, and the protein expression levels were normalized to the housekeeping protein GAPDH or β-actin. 14, and 28 after MI surgery. Mice were anesthetized with 2% isoflurane in 95% O 2 . Serial B-mode and M-mode images were obtained using a Vevo® 2100 ultrasound system (Visual Sonics, Canada), followed by cardiac function analysis using Vevo® LAB software (version 3.1.0).
2.12. Immunofluorescence Staining. OCT-embedded hearts were sectioned in a cryostat at a thickness of 7 μm. The slices were fixed with 10% formalin for 10 min, permeabilized with 0.5% Triton X-100 for 15 min, and blocked with 3% BSA (#B265993, Aladdin, Shanghai, China) for 30 min. Subsequently, overnight incubation was performed with the specific primary antibodies listed in Table 4, followed by incubation with Alexa Fluor-conjugated secondary antibody. Nuclei were stained with DAPI mounting medium (#H-1200-10, Vector, CA, USA).
2.13. Tube Formation Assays. Tube formation assays were conducted to evaluate the angiogenic potential of MSCs in response to different treatments. Human umbilical vein endothelial cells (HUVECs) were digested, counted, and resuspended in different conditioned media. Then, singlecell suspensions of the HUVECs were seeded into 96-well plates (1:2 × 10 4 cells per well) coated with 50 μl of solidified Matrigel (#354230, Corning, MA, USA) and cultured for 6 h under standard conditions. Images were captured with Leica DM IL LED microscopes, and the branch points and capillary lengths were quantified with ImageJ software using the angiogenesis analyzer plugin.
2.14. Cycloheximide (CHX) Chase Assay. To evaluate protein stability, MSCs were treated with 10 μg/ml CHX (#S7418, Selleck) to arrest c-FLIP protein biosynthesis under hypoxic conditions for 0 h, 3 h, and 6 h. After the MSCs were harvested, c-FLIP protein expression was assessed by Western blotting.  To examine protein degradation, MG132 (10 μM, #HY-13259, MCE, Shanghai, China) and chloroquine (CQ, 50 nM, #HY-17589A, MCE) were used to inhibit the ubiquitin-proteasome and autophagy pathways, respectively. Cells were incubated with MG132 or CQ for the indicated times before being harvested, and c-FLIP protein expression was assessed by Western blotting.
For the ubiquitination assay, MSCs were treated with 10 μM MG132 under hypoxic conditions for 6 h before being harvested for immunoprecipitation. The lysates were incubated with anti-FLIP antibodies (#56343, CST, MA, USA) or normal rabbit IgG (#2729, CST), bound to magnetic protein A beads (Bio-Rad), washed thoroughly, and eluted by boiling in 2× loading buffer, followed by Western blotting to examine the level of ubiquitin bound to the c-FLIP protein.
2.16. Immunoprecipitation. Immunoprecipitation was performed as previously described [19]. Briefly, MSCs were lysed with NP-40 cell lysis buffer (#P0013F, Beyotime) supplemented with a 1× protease phosphatase inhibitor cocktail mixture, 10 mM N-ethylmaleimide (NEM) (# HY-D0843, MCE), and 1 mM EDTA. In total, 10% of the lysates was kept as input, and the remaining samples were used for immunoprecipitation. Rabbit monoclonal anti-FLIP antibodies (1 : 100, #56343, CST) or normal rabbit IgG was incu-bated with magnetic protein A beads (#1614013, Bio-Rad) for 1 h at room temperature. The supernatant was discarded, and the bead-antibody complexes were washed with lysis buffer 3 times. Then, the bead-antibody complexes were incubated overnight at 4°C with the cell lysates mentioned above. The bead-antibody-protein immunocomplexes were washed 3 times with lysis buffer and then boiled in loading buffer with SDS at 98°C for 5 min to elute the proteins from the beads. Immunoblotting was performed to analyze the ubiquitin levels.
2.17. Statistical Analysis. Representative experiments were repeated at least three times, and the results are depicted as the mean ± standard deviation (SD). Statistical differences between two sets of data were analyzed by Student's t test, and multiple groups were compared using one-way ANOVA with Tukey's posttest. All statistical analyses and graphing were performed with GraphPad Prism 8 software. Differences for which P < 0:05 were considered statistically significant.

FAIM Promoted the Survival of MSCs In Vitro and In
Vivo. FAIM expression was found to be upregulated in SRT1720-pretreated MSCs, and FAIM knockdown abolished the protective effects of SRT1720 on MSCs in our previous study [20]. Thus, we speculated that FAIM may have an important effect on the apoptosis of transplanted MSCs. To explore this hypothesis, two cell apoptosis models were used in vitro to mimic ischemia or ROS generation in the MI microenvironment: the OGD model and the H 2 O 2 model (Supplementary Figure S2). As shown in Figures 1(a) and 1(b), the mRNA and protein levels of FAIM were robustly decreased in the OGD model. The same reduction in FAIM protein expression was observed in response to H 2 O 2 exposure (Supplementary Figure S3(a)). Thus, a FAIM overexpression experiment was carried out to determine the antiapoptotic effect of FAIM in the OGD model (Supplementary Figure S4(a)-4(b)), and the cell apoptosis rate was evaluated      (Figure 1(c)), and this result was further supported by TUNEL staining (Figure 1(d)). Furthermore, the protein expression of cleaved caspase-3, the critical executioner of apoptosis, was examined by Western blotting. We found that FAIM overexpression significantly attenuated the activation of caspase-3 ( Figure 1(e)). To explore the antiapoptotic effect of FAIM on MSCs in vivo, MSCs were marked with GFP and injected into the peri-infarct region immediately after LAD ligation. The MSCs retention rate was assessed by fluorescence microscopy 3 days after engraftment (Supplementary Figure S2). The immunofluorescence results showed that while no significant differences were observed between the two groups of cells (Supplementary Figure S4(a)), more FAIM-overexpressing MSCs (MSCs FAIM ) than MSCs infected with vector (MSCs Vec ) remained in the target tissue ( Figure 1(f)). These results strongly suggested that FAIM inhibited caspase-3-dependent apoptosis under hypoxic conditions in vitro and in vivo.

FAIM-Overexpressing MSCs Improved Cardiac Function and Reduced Infarct Size.
To evaluate the effect of FAIM overexpression on the improvement in the therapeutic effect of MSCs, MSCs Vec and MSCs FAIM were transplanted into the peri-ischemic area immediately after MI model construction, and an equivalent volume of DMEM was injected into the ischemic hearts in the control group. Transthoracic echocardiography was conducted on days 3, 7, 14, and 28 after cell transplantation to evaluate cardiac function, and Sirius Red staining was conducted 28 days after MI to examine the fibrotic area (Supplementary Figure S2). The echocardiography data revealed that compared with the DMEM and MSCs Vec groups, cardiac function after MI was significantly improved in the MSCs FAIM group (n = 6 for the sham group, n = 7 for the DMEM and MSCs Vec groups, and n = 8 for the MSCs FAIM group) (Figures 2(a)-2(c)). Histological analysis also confirmed these results, and a considerable reduction in infarct scarring was observed in the MSCs FAIM group on day 28 after cell transplantation by Sirius Red staining (Figures 2(d)-2(e)).
Accumulating evidence indicates that the effects of MSCs are principally attributed to their abundant paracrine functions, which result in angiogenesis [21]. Thus, we hypothesized that the increased retention of MSCs FAIM would increase angiogenesis. To validate this hypothesis, the capillary density and microvasculature in the periinfarct area were measured. Compared with those in the DMEM group and MSCs Vec group, increased numbers of endothelial cells (CD31-positive) and vascular smooth muscle cells (α-SMA-positive) was observed in the MSCs FAIM group 28 days after MSCs engraftment (Figures 2(f)-2(i)). Furthermore, an in vitro tube formation assay was conducted to evaluate the proangiogenic capacity of conditioned medium from MSCs. The supernatants of MSCs Vec and MSCs FAIM cultured under normoxic conditions and hypoxic conditions were used to culture HUVECs. No clear differences were observed between the MSCs FAIM group and the MSCs Vec group under normoxic conditions; however, culture with the supernatants of hypoxia-treated MSCs Vec led to a significant decrease in angiogenesis, and FAIM overexpression rescued this reduction (Supplementary Figure S6). Taken together, these results suggested that MSCs FAIM more effectively promoted angiogenesis, improved cardiac function, and limited the fibrotic area than did the MSCs Vec .

FAIM Regulates the Expression of c-FLIP in MSCs under
OGD Conditions. The extrinsic apoptosis pathway, which is mediated by DRs, and the intrinsic apoptosis pathway, which is mediated by mitochondria, are the two main pathways through which apoptosis is induced. Caspase-8 and caspase-9 are the initiating caspases in these two pathways [22]. Several studies have demonstrated that FAIM strongly   Stem Cells International inhibits DR signaling and Fas-mediated apoptosis [12]; however, the specific role of FAIM in hypoxia-induced apoptosis is poorly understood. Thus, the protein expression of cleaved caspase-8 and cleaved caspase-9 was determined. As expected, FAIM overexpression reduced the cleavage of caspase-8 but had no significant effect on caspase-9 activation (Figures 3(a)  and 3(b)). c-FLIP regulates the activation of caspase-8, and FAIM-deficient hepatocytes exhibit decreased c-FLIP protein expression [17]. Therefore, to test whether c-FLIP is involved in hypoxia-induced apoptosis, and whether FAIM can regulate c-FLIP expression, the protein expression level of c-FLIP was examined after the knockdown or overexpression of FAIM and culture under hypoxic conditions by Western blotting. c-FLIP protein expression was notably decreased in the context of hypoxia, and FAIM overexpression blocked this reduction (Figures 3(c) and 3(d)). Not surprisingly, c-FLIP expression notably declined after FAIM knockdown (Figures 3(e) and 3(f)).  Figure S4(d)). FAIM was silenced by siRNA in both MSCs Vec and MSCs FLIP , which was followed by OGD exposure. Upon FAIM silencing, c-FLIP overexpression reduced the percentage of apoptotic MSCs, as shown by Annexin V/PI staining and TUNEL assays (Figures 4(a)-4(d)). Western blotting further confirmed that the protein levels of cleaved caspase-3 were significantly decreased by c-FLIP overexpression (Figures 4(e) and 4(f)). Furthermore, MSCs FAIM were transfected with either siRNA specific for c-FLIP (si-FLIP) or scrambled siRNA (si-Scr), followed by OGD challenge. Under OGD conditions, c-FLIP

FAIM Stabilizes c-FLIP Protein in MSCs by Inhibiting Its
Ubiquitination. To determine the underlying mechanisms, we examined the mRNA level of c-FLIP and found that FAIM overexpression did not alter the transcript expression of c-FLIP ( Figure 5(a)). To reveal whether FAIM posttranscriptionally regulates the expression of c-FLIP, a CHX chase assay was performed to test the change in protein stability of c-FLIP in response to FAIM knockdown and overexpression. We found that FAIM overexpression significantly inhibited the decrease in c-FLIP protein expression ( Figure 5(b)), but FAIM knockdown facilitated a reduction in c-FLIP expression in MSCs ( Figure 5(c)). These data suggested that the degradation of c-FLIP could be impaired by FAIM. Autophagy and proteasome-dependent proteolysis are the two main ways by which protein degradation is regulated. Consequently, hypoxia-challenged MSCs were treated with CQ to inhibit the autophagy-lysosome pathway and MG132 to impede the proteasome-dependent pathway. In comparison with that in response to DMSO, the expression of c-FLIP was strongly restored by MG132 but barely affected by CQ ( Figure 5(d)). Collectively, these data showed that FAIM upregulates c-FLIP protein expression largely by inhibiting its proteasome-dependent degradation. To further investigate the ubiquitination of c-FLIP, we treated OGDchallenged MSCs with MG132 to arrest the degradation of     10 Stem Cells International c-FLIP without affecting its ubiquitination 6 h before harvesting for analysis. We found that c-FLIP was ubiquitinated under hypoxic conditions, and FAIM overexpression markedly decreased the ubiquitination level of c-FLIP ( Figure 5(e)). This result complements our previous finding showing that FAIM modulates c-FLIP expression by inhibiting its ubiquitination-proteasome-dependent proteolysis.

FAIM Reduces c-FLIP Degradation by Blocking JNK
Activation. Phosphorylated JNK has been reported to be upregulated during apoptosis [23] and has also been reported to participate in the degradation of c-FLIP [24]. We hypothesized that the JNK pathway is involved in hypoxia-induced apoptosis, and that FAIM attenuates the degradation of c-FLIP by inhibiting the phosphorylation of JNK. To validate this hypothesis, phosphorylated JNK was assessed by Western blotting. As expected, we found that JNK phosphorylation was strongly inhibited by FAIM in hypoxic MSCs (Figures 6(a) and 6(b)), and FAIM silencing increased this phosphorylation (Figures 6(c) and 6(d)). We further treated FAIM-deficient MSCs with SP600125, a specific JNK inhibitor, and verify its impact on c-FLIP protein expression. We found that SP600125 upregulated c-FLIP protein expression (Figures 6(e) and 6(f)), verifying that FAIM inhibits c-FLIP degradation mainly by blocking JNK activation. Taken together, our results provide evidence that FAIM inhibits MSC apoptosis by inhibiting the ubiquitinationdependent degradation of c-FLIP, which is mediated by JNK inactivation, thereby improving the efficacy of MSC transplantation (Figure 7).

Discussion
In this study, we showed that FAIM was downregulated in MSCs under ischemia-like conditions, including hypoxia and oxidative stress, in vitro. Consequently, we overexpressed FAIM in MSCs and found that FAIM augmentation protected MSCs from apoptosis in a harsh microenvironment. Similarly, we observed improved MSCs survival in a mouse MI model. Furthermore, improvements in cardiac function and a reduced infarction size were detected in the MSCs FAIM group 28 days after transplantation, accompanied by increased angiogenesis, suggesting that FAIM augmentation improved the efficacy of MSCs transplantation in a mouse MI model. Mechanistically, our findings demonstrated that the antiapoptotic effect of FAIM is largely dependent on c-FLIP, which suppresses the activation of caspase-8. Next, we found that FAIM had no effect on c-FLIP transcription but altered the protein degradation of c-FLIP. Further experiments demonstrated that FAIM protects c-FLIP from ubiquitination-proteasome-dependent degradation, which is mediated by impaired JNK phosphorylation. These data provide a novel approach to improve the effects of stem cells engraftment on cardiovascular indications.
MI poses a grave threat to the health of elderly individuals, and the loss of cardiomyocytes after MI is the most important cause of poor prognosis [25]. Therefore, remuscularization of the heart is the major focus in this field, and countless efforts have been made to find solutions, including promoting the regeneration of cardiomyocytes, transplanting induced pluripotent stem cell (iPSC)-derived cardiomyocytes, exogenous engrafting of stem cells or their derived exosomes, and the application of novel functional materials such as engineered heart tissues and hydrogels [26][27][28][29]. It has been demonstrated that within one week of birth, the hearts of neonatal mice exhibit regenerative activity, and subsequent studies showed that regenerative activity also occurred in neonatal pigs [30,31]. Exploring the mechanism of cardiomyocyte regeneration and implementing treatments have become important research directions. A recent study reported that cardiac-specific knockout of PTEN facilitated the proliferation and renewal of cardiomyocytes [32]. Moreover, Salvador-siRNA delivered by adeno-associated virus 9 promoted the proliferation of cardiomyocytes after ischemia and reperfusion injury by inhibiting the Hippo pathway, which repressed cardiomyocyte proliferation after MI [33]. However, these studies are still in their early stages, and it could take decades to achieve   [34]. Patches were used to provide mechanical support to improve outcomes in the early days after MI, and subsequent studies reported the development of patches equipped with anti-inflammatory activity, but they could not compensate for the loss of cardiomyocytes [29]. However, the unique physical and electrical properties of the heart make it difficult for simple bioengineering materials to be of great use, and those engineering materials are often functionalized through synergistic application with stem cells or iPSCderived cardiomyocytes [35,36]. Taken together, these results highlight the importance of cell-based therapies for MI treatment. Tremendous improvements in cell-based therapies for MI have been made in recent decades, and these therapies aim to produce a sufficient number of cardiomyocytes to compensate for the loss of viable cardiomyocytes after MI [37]. A variety of cells are used in the treatment of MI, such as human iPSCs (hiPSCs), human embryonic stem cells (hESCs), and the widely used hMSCs [27,38]. Generating cardiomyocytes from hiPSCs are no longer difficult [39], and this strategy has been applied in porcine MI models [40,41]. However, there are still many challenges before iPSC-derived cardiomyocytes move to early trials, such as arrhythmias that can result from the injection of cardiomyocytes, the long-term immunosuppressant therapy required due to allogeneic-derived iPSCs, and the purity of the large number of iPSC-derived cardiomyocytes during induced proliferation and differentiation [26]. MSCs are ideal candidates for large-scale application in the clinical treatment of MI. Recent studies have shown that the therapeutic effects of MSCs are mainly exerted through their robust paracrine effects [21,42], which promote the proliferation and migration of endothelial cells and ultimately improve cardiac angiogenesis and left ventricular function [43]. However, the harsh microenvironment formed by conditions including ischemia, hypoxia, oxidative stress, and inflammation causes massive cell death. A large proportion of engrafted stem cells were reported to perish within 24 h after transplantation [44,45], making strategies that prolong MSCs lifespan an attractive way to improve the efficacy of MSCs [46]. Our previous study showed that pretreatment with SRT1720 improved the efficacy of MSCs in rodent and Whole-cell lysates were immunoprecipitated with anti-FLIP antibodies, followed by immunoblotting with antiubiquitin antibodies. The input (10% of the total) was analyzed by Western blotting using anti-FLIP and anti-FAIM antibodies. The data are shown as the mean ± SD. Ns indicates not significant; * denotes P < 0:05; * * P < 0:01.

13
Stem Cells International nonhuman primate MI models by prolonging MSCs survival [20,47]. Consistent with these findings, our results showed that FAIM augmentation improved implanted MSCs survival and ultimately promoted neovascularization and ameliorated cardiac function.
Two splicing variants of FAIM have been described; the longer variant, FAIM-L, which is exclusively expressed in neural tissue, and the ubiquitously expressed shorter variant, known as FAIM-S [10]. FAIM-L exerts antiapoptotic effects by suppressing TCR signaling or DR signaling [11,12]. However, to the best of our knowledge, this is the first study to reveal the protective effect of FAIM-L against apoptosis induced by hypoxia. Previous studies reported that c-FLIP expression was perturbed in FAIM-depleted cells [17]. Con-sistent with these findings, our results showed that the survival effect of FAIM was mainly achieved by restoring c-FLIP expression. However, to date, the effects of FAIM on c-FLIP expression are not fully understood. FAIM has been reported to exert its biological effects by reducing ubiquitination; FAIM was found to reduce the ubiquitination and degradation of XIAP via direct interaction [48], FAIM knockout led to the presentation of ubiquitinated aggregates in the retina [49], and FAIM protects glutaminase C from ubiquitination and induces cancer cell proliferation in lung adenocarcinoma [14]. Here, we found that FAIM could maintain c-FLIP stabilization under hypoxic conditions by protecting c-FLIP from ubiquitination and degradation, thus protecting MSCs from apoptosis. 14 Stem Cells International c-FLIP, a strong regulator of caspase-8 activation due to the pseudocaspase domain at its C-terminus [50], participates in various cell death pathways, including apoptosis, necroptosis, and autophagy [51,52]. Recent studies have shown that c-FLIP prevents cells from undergoing apoptosis by impairing p53-mediated PUMA upregulation and caspase-8 activation [53]. However, the functions of c-FLIP are contradictory and depend on its abundance in the death-inducing signaling complex (DISC) [15], which is assembled from Fas, FADD, procaspase 8, and c-FLIP [54]. Low concentrations of c-FLIP in the DISC can cause the formation of a procaspase-8/c-FLIP heterodimer and promote caspase-8 activation [55]. On the other hand, c-FLIP at a high concentration competitively suppresses the assembly of procaspase-8 into the DISC and thus reduces the activation of caspase-8 [56]. In the present study, we demonstrated that high protein expression of c-FLIP inhibited the activation of caspase-8; however, the caspase-8 activation status under low c-FLIP expression levels was not examined because we could not confirm whether the cleavage of procaspase-8 caused by the decrease in c-FLIP protein expression was due to a decrease in inhibition or an increase in activation. Regardless, this does not affect our conclusion that FAIM overexpression reduces MSCs apoptosis by ameliorating the decreased expression of c-FLIP under OGD conditions. In addition, accumulating evidence suggests that c-FLIP is a short-lived protein whose half-life is approximately 3 h [24]. Some studies have shown that the degradation of c-FLIP occurs mainly through the ubiquitin-proteasome pathway [57], and JNK activity may be a critical regulator of c-FLIP protein stability [24]. In agreement with previous findings, we revealed that the protein stability of c-FLIP is closely related to the presence of FAIM, and subsequent experiments showed that FAIM regulates c-FLIP protein stability by impairing JNK phosphorylation, which expands the current knowledge of the protein stability of c-FLIP.
There are some limitations of this study. First, our experimental setting could not completely mimic true clinical conditions. MSCs were injected into the peri-infarct zones immediately after LAD ligation, which is unrealistic in the real world. Second, unbiased analysis was not utilized in this study, and mass spectrometry-based proteomics analysis may be a more accurate way to determine the most predominant protein that regulates the antiapoptotic effects of FAIM in our model. Third, the mechanism by which FAIM regulates JNK activation was not explored in this work.

Conclusion
In conclusion, our study demonstrated that FAIM augmentation could effectively promote MSCs survival in vivo and ultimately ameliorate cardiac function and improve the prognosis of MI, providing a novel approach to improve the efficacy of MSC-based therapy.

Data Availability
The datasets used and/or analyzed to support the findings of this study are included within the article and supplementary information files.

Ethical Approval
All animal experiment protocols were approved by the Medical Ethics Committee of the Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, China.