SCARF1 promotes M2 polarization of Kupffer cells via calcium‐dependent PI3K‐AKT‐STAT3 signalling to improve liver transplantation

Abstract Objectives This study aimed to investigate the protective effect of SCARF1 on acute rejection (AR), phagocytic clearance of Kupffer cells (KCs), M2 polarization and the exact mechanism underlying these processes. Methods AAV was transfected into the portal vein of rats, and AR and immune tolerance (IT) models of liver transplantation were established. Liver tissue and blood samples were collected. The level of SCARF1 was detected via WB and immunohistochemical staining. Pathological changes in liver tissue were detected using HE staining. Apoptotic cells were detected using TUNEL staining. KC polarization was assessed via immunohistochemical staining. Primary KCs were isolated and co‐cultured with apoptotic T lymphocytes. Phagocytosis of apoptotic cells and polarization of KCs were both detected using immunofluorescence. Calcium concentration was determined using immunofluorescence and a fluorescence microplate reader. The levels of PI3K, p‐AKT and P‐STAT3 were assessed via WB and immunofluorescence. Results Compared to the IT group, the level of SCARF1 was significantly decreased in the AR group. Overexpression of SCARF1 in KCs improved AR and liver function markers. Enhanced phagocytosis mediated by SCARF1 is beneficial for improving the apoptotic clearance of AR and promoting M2 polarization of KCs. SCARF1‐mediated enhancement of phagocytosis promotes increased calcium concentration in KCs, thus further activating the PI3K‐AKT‐STAT3 signalling pathway. Conclusions SCARF1 promotes the M2 polarization of KCs by promoting phagocytosis through the calcium‐dependent PI3K‐AKT‐STAT3 signalling pathway.


| INTRODUC TI ON
Acute rejection (AR) and its related complications remain one of the most important factors affecting the long-term survival of patients who have undergone liver transplantation. The elimination of apoptotic cells is believed to play a role in the formation of a microenvironment supportive of immune tolerance (IT). 1 KCs. Based on this, we believe that SCARF1 may participate in the regulation KC transformation that may be related to the enhanced phagocytosis of APCs mediated by SCARF1. However, the exact mechanism underlying this process remains to be further studied.
It is believed that the activation of STAT3 plays an important role in the elimination of apoptotic cells and the formation of an IT microenvironment. [8][9][10] STAT3 also participates in the regulation of macrophage functions, particularly in immune function. 11 According to a previous study, 12  Previous studies have found that the PI3K-AKT signalling pathway plays an important role in the regulation of macrophage polarization, in particular by affecting the phosphorylation level of STAT3. 13,14 However, the PI3K-AKT signalling pathway itself is affected by changes in cytoplasmic calcium concentrations. 15,16 Notably, the intracellular calcium concentration was significantly increased during phagocytosis by APCs to clear apoptotic cells. 17 Based on the above study, we hypothesized that by recognizing apoptotic T lymphocytes, KCs can eliminate apoptotic T lymphocytes and transform into the M2 phenotype due to changes in STAT3 phosphorylation. Following this phenotype transformation, a large number of anti-inflammatory factors can be secreted and the microenvironment of IT can be formed. During this process, the SCARF1mediated calcium-dependent PI3K-AKT-STAT3 signalling pathway may play an important role.
In our study, we investigated the effect of upregulation of SCARF1 onKCs and on AR of the liver to further elucidate possible mechanisms underlying these processes and to aid in identifying new and efficient treatments to induce the formation of a microenvironment supportive of liver IT.  Two weeks after transfection, cell-specific identification of AAV transfection was detected according to living fluorescence imaging.

| Animalsandprotocols
The overexpression effect was also detected using RT-PCR and immunofluorescence staining. As shown in Figure S2A, the fluorescence intensity in liver tissue was significantly higher than that in other tissues (R). However, after blocking KCs in liver tissue by treatment with GdCl 3 , the fluorescence intensity in liver tissue decreased significantly (L). As shown in Figure S2B, the level of SCARF1 mRNA in KCs in the AR + OE group was significantly higher than that in the AR group or the AR + Ctrl group. As shown in Figure S2C, the number of SCARF1-positive KCs (level of co-localization of red fluorescence and green fluorescence) was significantly higher than that in the AR group or the AR + Ctrl group. This suggests that our transfection method can effectively overexpress SCARF1 in KCs.

| Isolation,cultivationandfunction identification of KCs
According to the three-step approach proposed by Li 18 that included digestion by collagenase IV, gradient centrifugation and selective adherence, KCs were isolated from normal liver samples. KCs were then cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin G and 100 U/mL streptomycin at 37°C in the presence of 5% CO 2 . After culturing for 24 hours, the phagocytosis of KCs was examined using the ink assay, and the surface molecules of KCs were detected via immunofluorescence staining ( Figure S3). As shown in Figure S3A, at 2 hours after culture, KCs were round and adhered well to the culture dish. As shown in Figure S3B, at 24 hours after culture, the morphology of KCs was typical fusiform. As shown in Figure S3C, numerous ink particles were observed in the KCs. As shown in Figure S3D, the percentage of F4/80-positive cells was greater than 90%. This indicates that our method can efficiently separate KCs from liver tissue while retaining normal phagocytosis.

| Co-cultureofKCsandapoptoticTlymphocyte
T lymphocytes were cultured in DMEM supplemented with 10% FBS, 100 U/mL penicillin G and 100 U/mL streptomycin at 37°C in the presence of 5% CO 2 for 24 hours. T lymphocytes were treated with 800 µM H 2 O 2 at room temperature for 1 hour to obtain apoptotic T lymphocytes. Next, the level of T cell apoptosis was detected using flow cytometry ( Figure S4). After 800 µM H 2 O 2 treatment, the level of T cell apoptosis was greater than 95%, indicating that our induction method was effective. KCs were then co-cultured with apoptotic T lymphocytes at a ratio of 10:1.

| Histopathologicalexamination
Liver samples from the different groups were collected after surgery. The liver tissues were fixed with 10% buffered formaldehyde for 24 hours. Next, all samples were embedded in paraffin, sectioned and stained with haematoxylin and eosin. The histopathological changes were observed using an inverted microscope, and the hepatic damage in each group was evaluated using the Banff score.
According to the rejection activity index (RAI), scores ranging from 0 to 3 were regarded as uncertain for AR, scores ranging from 3 to 5 were regarded as mild AR, scores ranging from 5 to 7 were regarded as moderate AR, and scores ranging from 7 to 9 were regarded as severe AR.

| Examinationofliverfunction
Blood samples from each group were collected after the operation.
The serum liver function markers, including alanine aminotransferase (AST) and aspartate transaminase (ALT), were detected using an automatic biochemical analyzer (Beckman CX7, Beckman Coulter, CA, USA) to evaluate liver function. cycles.

| Real-timereversetranscription-polymerase
All samples were normalized according to GAPDH expression.

| Total protein extraction
The total protein was extracted as follows: 1. Radioimmunoprecipitation assay buffer containing phenylmethane-sulfonyl fluoride (100 M) and sodium fluoride (100 M) was added to the suspension on ice for 30 minutes.
2. The suspension was centrifuged at 4°C and 12 000 rpm for 5 minutes. The loading buffer was added, and then, the suspension was heated at 95°C for 10 minutes.

| Electrophoresis
2. PVDF membranes were immersed in phosphate-buffered saline solution with 0.1% Tween 20 for 5 minutes. Next, the PVDF membranes were incubated at room temperature with the appropriate secondary antibody (1:5000) for 1 hour.
3. The signals were detected via chemiluminescence using a gel imaging system. The relative expression of the proteins of interest was normalized to GAPDH expression.

| ImmunostainingforTUNEL
TUNEL assays were performed on the paraffin sections of the hepatic tissues using the TUNEL kit to assess the number of apoptotic T lymphocytes. The paraffin sections were dewaxed with xylene, gradient eluted using ethyl alcohol and successively immersed in working buffer and blocking buffer. Next, TdT and streptavidin-HRP were successively added to each sample to complete the staining.
The staining of each group of cell nuclei was observed using fluorescence microscopy. Green cell nuclei were identified as TUNELpositive cells.

| Immunohistochemicalstaining
The paraffin sections were dewaxed using xylene and then gradient eluted with ethyl alcohol. After dewaxing and elution, the sections were digested with pancreatic enzymes at 37°C for 30 minutes.
Next, the sections were boiled in citrate buffer solution for 5 minutes. The paraffin sections were blocked with goat serum at 37°C for 10 minutes, incubated with primary antibody against SCARF1 ( The staining of each section was observed using inverse microscopy.

| Analysisusingflowcytometry
The degree of apoptosis in T lymphocytes was detected via flow cytometry using the Annexin V-FITC/PI apoptosis detection kit.

| ELISA
ELISA was performed to detect the levels of TNFα, IL-6 and IL-10 in blood. Antigen serum was added to an ELISA plate (100 μL/well)

| Statisticalanalysis
All results were analysed using SPSS 18.0 software (SPSS Inc, Chicago, USA). Normally distributed data are shown as mean ± SD.
Differences between groups were detected using a t test. The Shapiro-Wilk test was used to test normality. Data exhibiting a significance value > .05 were regarded as conforming to a normal distribution. Non-normally distributed data are shown as the median, and differences were detected using the rank-sum test. Differences with P values < .05 were regarded as statistically significant.

| SCARF1wasdecreasedinKCsduringARafter liver transplantation
To investigate whether SCARF1 is involved in the progression of liver transplantation AR, the expression of hepatic SCARF1 was detected in the AR and IT groups. As shown in Figure 1A,B, compared to that of the IT group, the expression of hepatic SCARF1 in the AR group was significantly decreased. Meanwhile, SCARF1-positive areas in the AR group were significantly lower than those in the IT group ( Figure 1C). It should be noted that both WB and immunohistochemical results demonstrated that the expression of SCARF1 in the liver was significantly decreased after KCs were depleted using GdCl 3 , particularly in the IT group. These results preliminarily suggest that SCARF1 in KCs may be involved in the process of acute liver transplantation rejection. We further identified the cell origin of SCARF1 in the liver. KCs were isolated from liver tissues in the AR and IT groups. The expression of SCARF1 in KCs from the AR group was significantly lower than that in the IT group ( Figure 1D,E). The KCs in the AR and IT groups were then labelled with F4/80 (red). We observed that the expression of SCARF1 (labelled by green) in the AR group was significantly lower than that in the IT group ( Figure 1F). Our results revealed that SCARF1 is expressed primarily in KCs and may be important in the context of acute liver transplantation rejection.

| SCARF1overexpressioninKCsalleviatedAR after liver transplantation
As the expression of SCARF1 in KCs was significantly decreased in the AR model after liver transplantation, we used CD68 labelled adeno-associated virus to overexpress SCARF1 in the AR group to observe whether overexpression of SCARF1 exerted a protective effect on AR after liver transplantation. Five days after liver transplantation ( Figure 2A,B), the expression of SCARF1 in the AR group was significantly lower than that in the IT group. Meanwhile, the expression of SCARF1 in the AR + Ctrl group was not significantly altered compared to that in the AR group, whereas SCARF1 expression in the AR + OE group was significantly higher than that in the AR group. HE staining ( Figure 2C) revealed that in comparison with the IT group, the liver tissue of the AR group exhibited typical AR markers, including a large number of inflammatory cell infiltration, hepatocyte necrosis, fibrous tissue hyperplasia and a significantly increased AR index ( Figure 2D). Concurrently, compared to the AR group, the above indexes ( Figure 2D) were not significantly altered in the AR + Ctrl group. In the AR + OE group, the degree of AR ( Figure 2C) in liver tissue was significantly reduced, and the AR index ( Figure 2D) was also significantly decreased. Liver function markers were detected in each group ( Figure 2E,F). Compared to levels in the IT group, serum ALT and AST levels in the AR group were significantly elevated. There was no significant change in these values in the AR + Ctrl group compared to those of the AR group; however, ALT and AST levels in the AR + OE group were significantly lower than those in the AR group. We further observed the expression of SCARF1 ( Figure 2G,H) and the histopathological changes ( Figure 2I), AR score ( Figure 2J), and liver function markers ( Figure 2K,L) in each group 15 days after liver transplantation. The results exhibited the same trend 5 days after liver transplantation.
In general, we observed that overexpression of SCARF1 in KCs alleviated the degree of AR after liver transplantation and improved the level of liver function markers, all of which were beneficial for alleviating the occurrence of AR after liver transplantation.

| SCARF1overexpressioninKCspromoted a microenvironment supportive of IT after liver transplantation
In the process of AR of liver transplantation, the production of apop-

| SCARF1-mediatedenhancementof phagocytosisinKCsisrequiredforapoptotic cell clearance
We next sought to confirm that the decrease in apoptotic cells in liver tissue after overexpression of SCARF1 is due to the enhanced phagocytic clearance of apoptotic cells by KCs and not because of a decrease in apoptotic cells themselves. Cytochalasin D (CD) was used to block the phagocytic function of KCs and to observe whether SCARF1-mediated apoptosis clearance was altered.
Fifteen days after liver transplantation ( Figure 4A,B), after phagocytosis was blocked, the number of apoptotic cells in the IT group increased significantly compared to that prior to blocking of phagocytosis. Concurrently, after phagocytosis was blocked, the number of apoptotic cells in the liver tissue of the AR and AR + Ctrl groups did not change significantly compared to that prior to blocking. It should be noted that after phagocytosis was blocked, apoptotic cells in the liver tissue of the AR + OE group were also significantly increased.
KCs were then isolated and co-cultured with apoptotic T cells in vitro. At 24 hours after co-culture, the suspended apoptotic cells were collected. As shown in Figure 4C-F, after phagocytosis was blocked, the levels of c-caspase3, Bax and Bid in the IT group increased significantly compared to that prior to blocking. Meanwhile, after phagocytosis was blocked, the levels of c-caspase3, Bax and Bid in the AR group and AR + Ctrl group did not change significantly compared to that observed prior to blocking. However, after phagocytosis was blocked, the levels of c-caspase3, Bax and Bid in the AR + OE group were also significantly increased. At 24 hours after co-culture, adherent KCs were also collected. As shown in Figure 4G,H, apoptotic cells were labelled with PKH26 (red) and KCs were labelled with F4/80 (green). After blocking phagocytosis, the number of apoptotic cells phagocytized by KCs (level of colocalization of red fluorescence and green fluorescence) decreased significantly in the IT group. Meanwhile, after phagocytosis was blocked, the number of apoptotic cells phagocytized by KCs in the AR group and AR + Ctrl group did not change significantly compared to that observed prior to blocking. However, after phagocytosis was blocked, the number of apoptotic cells phagocytized by KCs in the AR + OE group also decreased significantly.

In summary, SCARF1-mediated enhancement of phagocytosis in
KCs is required for apoptotic cell clearance after liver transplantation.

| SCARF1-mediatedenhancementof phagocytosisinKCspromotesM2polarization
Previous studies have demonstrated that KCs phagocytize apoptotic cells, and this is an important means for inducing cell polarity remodelling. Our results also revealed that the polarization state of These results confirmed our hypothesis that SCARF1-mediated enhancement of KCs phagocytosis was beneficial for the transformation of KCs from the M1 to the M2 phenotype.

| Thecalcium-dependentPI3K-AKT-STAT3 signallingpathwayexhibitedincreasedactivityduring phagocytosis
The change in intracellular calcium concentration is an important event in the process of phagocytosis. Previous studies also determined that an increase in intracellular calcium concentration can promote the activation of the PI3K-AKT signalling pathway. We speculate that the calcium-dependent PI3K-AKT signalling pathway may play an important role in SCARF1-mediated transformation of KCs induced by enhanced phagocytosis. First, at 24 hours after co-culture, the concentration of calcium in KCs was detected via immunocytochemistry and using a fluorescence microplate reader. As shown in Figure 6A,B, compared to the IT group, the calcium concentration in KCs in the AR and AR + Ctrl groups was decreased significantly. However, compared to the AR and AR + Ctrl groups, the calcium concentration in KCs in the AR + OE group was increased significantly. After blocking of phagocytosis, the calcium concentration in KCs in the IT group decreased significantly compared to that in unblocked cells. Meanwhile, the calcium concentration in KCs in the AR and AR + Ctrl groups did not change significantly compared to that in untreated cells. However, the calcium concentration in KCs in the AR + OE group was also significantly decreased compared to that in untreated cells.
We further examined the effect of calcium concentration on the activation of the PI3K-AKT-STAT3 signalling pathway. As shown in Figure 6C,D, the levels of PI3K, p-AKT and p-STAT3 in each group were consistent with the change in calcium concentration. To confirm that the activation of the PI3K-AKT-STAT3 signalling pathway is directly regulated by intracellular calcium, the calcium antagonist BAPTA-AM was used to block intracellular calcium. As shown in

| SCARF1-mediatedcalcium-dependent PI3K-AKT-STAT3signallingpathwaypromotesM2 polarization of KCs
To further confirm the effect of the calcium-dependent PI3K-AKT-STAT3 signalling pathway on the polarization of KCs, a PI3K inhibitor (IPI-549), an AKT activity inhibitor (GSK2141795) and a STAT3 phosphorylation inhibitor (Stattic) were all used to inhibit PI3K and the phosphorylation of AKT and STAT3, respectively. Thereafter, we observed the changes in the entire signalling pathway and the polarization state of KCs in the presence and absence of the above inhibitors.
As shown in Figure 7A-D, after inhibiting PI3K using IPI-549, the levels of PI3K, p-AKT and p-STAT3 in KCs in the IT group were decreased significantly compared to those of untreated cells.
Meanwhile, the levels of PI3K, p-AKT and p-STAT3 in KCs in the AR and AR + Ctrl groups did not change significantly compared to those of untreated cells. However, the levels of PI3K, p-AKT and p-STAT3 in KCs in the AR + OE group were also significantly decreased compared to those of untreated cells. Activated STAT3 enters the nucleus after phosphorylation, and based on this, the nuclear penetration of STAT3 was also detected using immunocytochemistry. As shown in Figure 7E,F, after inhibiting PI3K using IPI-549, the nuclear-positive rate for STAT3 in KCs in the IT group decreased significantly compared to that of untreated cells. Meanwhile, the nuclear-positive rate for STAT3 in KCs in the AR group and AR + Ctrl group did not change significantly compared to that of untreated cells. However, the nuclearpositive rate for STAT3 in KCs in the AR + OE group was also significantly decreased compared to that of untreated cells. As shown in Figure 7G, at 24 hours after co-culture, KCs were labelled with CD163 (red) and F4/80 (green). After inhibiting PI3K using IPI-549, the number of M2 KCs (level of co-localization of red fluorescence and green fluorescence) decreased significantly in the IT group compared to that of untreated cells. Meanwhile, the number of M2 KCs in the AR group and AR + Ctrl group did not change significantly compared to that of untreated cells.
However, the number of M2 KCs in the AR + OE group was also significantly decreased compared to that of untreated cells. Our results demonstrate that PI3K is located upstream of the entire signalling pathway, and inhibition of PI3K can block the activation of AKT and STAT3 and the M2 polarization of KCs.
As shown in Figure  As shown in Figure 9G, after inhibiting phosphorylation of STAT3  released that subsequently induce inflammation or lead to the development of autoimmune disease. 21,22 According to a preliminary study, this process is applicable to multiple diseases, including AR induced by transplantation, 23 oncogenesis 24 and autoimmune disease. 25 Therefore, the key to inducing IT in the context of liver transplantation is to promote the timely clearance of apoptotic cells by phagocytes (primarily macrophages within the liver). Moreover, after overexpression of SCARF1 in KCs, the clearance of apoptotic cells due to enhanced phagocytosis was significantly increased, despite the observation that apoptosis was not significantly reduced. Therefore, it is likely that SCARF1 may play a more important role in the clearance of apoptotic cells, a process that may facilitate the formation of a microenvironment supportive of IT.

| D ISCUSS I ON
The M2 polarization of KCs is conducive to the formation of IT after liver transplantation, and it is also regarded as an important indicator of the formation of an IT microenvironment after liver transplantation. 28 Our experiment demonstrated that after overexpression of SCARF1 in KCs, the polarization state of KCs transformed into the M2 type, and this was consistent with the decrease in apoptotic cells. Interestingly, previous studies have observed that the polarization state of KCs can transform into the M2 type in response to stimulation by apoptotic cells. 29 After phagocytizing apoptotic cells, the change in the polarization state of KCs is a complex process that may involve activation of the TGFβ/Smad signalling pathway, 30 LC3-associated phagocytosis, 31 and activation of the aryl hydrocarbon receptor. 32   Therefore, we demonstrated that the calcium-dependent PI3K-AKT-STAT3 signalling pathway promotes the polarization of KCs, and this process is initiated by SCARF1-mediated enhancement of phagocytosis.

| CON CLUS IONS
We believe that the upregulation of SCARF1 in KCs can suppress the damage caused by acute liver rejection and can promote IT.

ACK N OWLED G M ENTS
This study was supported by the National Natural Science Foundation of China (grant nos. 81670599, 81671580 and 81700573).

CO N FLI C T O F I NTE R E S T
None.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data included in this study are available upon request by contacting the corresponding author.