Saikosaponin-d Alleviates Renal Inflammation and Cell Apoptosis in a Mouse Model of Sepsis via TCF7/FOSL1/Matrix Metalloproteinase 9 Inhibition

ABSTRACT

Evidence exists reporting that saikosaponin-d (Sa) can prevent experimental sepsis, and this study aims to illustrate the molecular events underlying its renoprotective effects on lipopolysaccharide (LPS)-induced renal inflammation simulating sepsis. Through network pharmacology analysis and bioinformatics analysis, we identified that Sa may influence sepsis development by mediating TCF7. Dual luciferase reporter gene and chromatin immunoprecipitation (ChIP) assays were used to explore the interactions between TCF7, FOSL1, and matrix metalloproteinase 9 (MMP9). The experimental data suggest that Sa attenuated LPS-induced renal injury, as evidenced by the reduced production of proinflammatory cytokines as well as cell apoptosis in the renal tissues of LPS-induced mice. Mechanically, Sa inhibited FOSL1 by inhibiting TCF7, which reduced the expression of inflammatory factors in renal cells. TCF7 activated the FOSL1 expression and consequently promoted the expression of MMP9. Also, Sa reduced cell apoptosis and the expression of inflammatory factors by inhibiting the TCF7/FOSL1/MMP9 axis in vivo. In conclusion, Sa suppresses FOSL1 transcription by downregulating TCF7, thereby inhibiting MMP9 expression and ultimately reducing the renal inflammation and cell apoptosis induced by sepsis.

INTRODUCTION

Sepsis is regarded as a systemic inflammatory disorder induced by infection and can lead to organ failure (1). Unfortunately, sepsis poses one of the key reasons for intensive care unit admission (2). Sepsis is featured by early massive catabolism and loss of lean body mass, in addition to escalating hypermetabolism persisting for a long period (3). At present, there is a lack of effective therapy for sepsis, and platelets have been suggested as a potential target for treatment (4). Notably, the kidney is believed to be a common sepsis-injured organ, and sepsis may induce acute kidney injury and thus cause high morbidity and mortality (5).

Saikosaponin-d (Sa) is a crucial active component of the traditional Chinese medicinal herb, Bupleurum scorzonerifolium Willd (6). Sa plays an important therapeutic role in a variety of inflammation-related diseases (7). Intriguingly, it has been previously reported that Sa could prevent renal tubular epithelial cell from high-glucose-induced injury via regulation of SIRT3 (8). Of note, transcription factor 7 (TCF7) was found to be a differential gene in peritoneal cells in a mouse model of sepsis at early and late time points (9).

Intriguingly, a previous study revealed that TCF7 could promote the transcription of Fos-like antigen 1 (FOSL1) in perihilar cholangiocarcinoma (10). FOSL1 is identified as one of the AP-1 transcription factors and is upregulated in multiple human cancers (11). According to a recent study, FOSL1 is a novel mediator of deviant angiogenic signaling in pulmonary endothelial cells induced by sepsis (12). Intriguingly, downregulation of FOSL1 expression could block the expression of E2-induced matrix metalloproteinases 9 (MMP9) (13). MMP9 is regarded as one of the zinc ion-dependent proteinases and can exert a pathogenic role in chronic inflammatory autoimmune disorders (14). Downregulated MMP9 by pretreatment with AZD4547 was found to aid in protecting against excessive inflammatory damage in mice with sepsis (15). Taking the aforementioned findings into consideration, we therefore proposed a hypothesis in this study that Sa may mediate renal inflammation and cell apoptosis in sepsis through regulation of the TCF7/FOSL1/MMP9 axis.

RESULTS

Sa attenuated LPS-induced renal injury in mice.

In order to investigate whether Sa can alleviate renal injury in mice with sepsis, we established a mouse model of sepsis induced by intraperitoneal injection of lipopolysaccharide (LPS). The mice were given Sa by intragastric administration for 1 week before modeling. First, we investigated whether Sa affected the survival of mice. As shown in Fig. 1A, the survival rates of mice in the sham, Sa high-dose (Sa-H), and Sa low-dose (Sa-L) groups were still 100% on day 7, while that of mice in the LPS group was 0% on day 2. Mice in the Sa-L group died on day 4 after administration of Sa 1 week in advance, and the survival rate of mice in the Sa-H group dropped to 0% on day 6. Therefore, we concluded that Sa can significantly improve the survival rate of LPS. Subsequently, we further evaluated the effect of Sa on the pathological changes of kidneys in LPS-induced septic mice. First, we quantified the vascular leakage of mouse kidney by Evans blue staining. As shown in Fig. 1B, compared with the mice in the LPS group, the vascular leakage of LPS-induced septic mice in the Sa group was significantly reduced. As the concentration increased, the vascular leakage decreased significantly, which confirmed that Sa had a certain protective effect on the renal blood vessels of mice. After that, we detected the levels of blood urea nitrogen (BUN) and serum creatinine (SCr) in the sera of mice. As shown in Fig. 1C and D, the levels of BUN and SCr in the sera of mice in the LPS group were significantly higher than those in the sham group, while the levels of BUN and SCr in the sera of septic mice were decreased by intragastric administration of Sa in advance. In addition, we observed the renal tissue damage by histological staining. The results of hematoxylin and eosin (HE) and Periodic acid-Schiff (PAS) staining showed that tubule swelling, tubular nodule formation, brush edge loss, and inflammatory cell infiltration were observed in the extramedullary strip of renal tissue in the LPS group (Fig. 1E and F), while early administration of Sa could significantly reduce the degree of LPS-induced renal injury.

FIG 1.

Sa attenuates LPS-induced renal injury in mice. (A) Survival rate analysis of mice (10 mice in each group). (B) Vascular leakage of kidney in each group (10 mice in each group). (C) BUN level in the sera of mice in each group (10 mice in each group). (D) SCr level in the sera of mice in each group (10 mice in each group). (E) PAS staining of renal tissues in each group (10 mice in each group). (F) Quantitative analysis of HE staining for tissue damage in each group of mice (10 mice in each group). (G) Quantitative analysis of NGAL and lectin in renal tissue sections (10 mice in each group). Lectin was used as a marker of renal tubular epithelial cells. (H) Western blotting was used to detect the protein expression of NGAL in each group (10 mice in each group). (I) The protein expression of NGAL in panel H was quantified by the optical density method. (J) The expression of NGAL in renal tissue was detected by RT-qPCR (10 mice in each group). (K) KIM-1 expression in renal tissue was detected by RT-qPCR (10 mice in each group). The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

At the same time, immunofluorescence staining, Western blotting, and reverse transcription quantitative PCR (RT-qPCR) were conducted to analyze the biomarker of renal injury, namely, neutrophil gelatinase-associated lipocalin (NGAL). The results showed that Sa pretreatment could significantly reduce the NGAL protein and gene expression in LPS-induced kidney (Fig. 1G to J). At the same time, Sa significantly inhibited the KIM-1 gene expression induced by LPS (Fig. 1K). In conclusion, these data suggest that Sa can improve the renal function of mice.

Sa reduced the production of proinflammatory cytokines in the kidney of LPS-induced mice.

In order to evaluate whether Sa can eliminate the inflammatory responses induced by LPS, we used RT-qPCR to detect the mRNA expression of proinflammatory factors in kidney. As shown in Fig. 2A, the addition of Sa did not promote the mRNA expression of proinflammatory cytokines. Compared with the sham group, the mRNA expression of interleukin-6 (IL6), IL-1β, tumor necrosis factor alpha (TNF-α), and monocyte chemoattractant protein-1 (MCP-1) was decreased in the Sa-H and Sa-L groups. LPS stimulation significantly increased the mRNA expression of IL-6, IL-1β, TNF-α, and MCP-1 in renal tissues, while Sa pretreatment significantly reduced the corresponding gene expression. At the same time, as shown in Fig. 2B and C, compared with the sham group, the protein expression levels of TNF-α, IL-1β, IL-6, and MCP-1 in the renal tissues of mice were reduced in the Sa-H and Sa-L groups. However, the protein expression of TNF-α, IL-1β, IL-6, cyclooxygenase 2 (COX-2), inducible nitric oxide synthase (iNOS), and high-mobility group box 1 (HMGB1) in the renal tissues of mice in the LPS group was significantly increased, while Sa pretreatment could decrease the above-mentioned protein expression induced by LPS. The above results show that Sa can reduce the inflammation stimulated by LPS in renal tissues.

FIG 2.

Sa reduces the production of proinflammatory cytokines in the kidney of LPS-induced mice. (A) RT-qPCR was used to detect the mRNA expression of IL-1β, IL-6, TNF-α, and MCP-1 in renal tissues of mice (10 mice in each group). (B) Western blotting was used to analyze the protein expression of COX-2, iNOS, TNF-α, HMGB1, IL-6, and IL-1β in renal tissues (10 mice in each group). (C) Protein expression of COX-2, iNOS, TNF-α, HMGB1, IL-6, and IL-1β in panel B was quantified by the optical density method (10 mice in each group). The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

Sa reduced renal cell apoptosis in LPS-induced mice.

Renal cell apoptosis is another prominent feature in the pathogenesis of sepsis-associated nephritis. Intrarenal inflammatory infiltration can lead to apoptosis, which further promotes the loss of renal tubular epithelial cells and causes nephritis. Next, we studied the effect of Sa on the apoptosis of renal cells by terminal deoxynucleotidyl transferase-mediated dUTP-biotin nick end labeling (TUNEL) staining. As shown in Fig. 3A, the number of TUNEL-positive cells was decreased in the Sa-H and Sa-L groups relative to that in the sham group, with the Sa-H group showing a more pronounced decline. In addition, the number of TUNEL-positive cells in the LPS group was higher than that in the sham group, while pretreatment with Sa significantly reduced the number of TUNEL-positive cells in renal tissue sections of septic mice. Caspase-3 is a key executor of modifying apoptosis-related proteins. The expression of cleaved caspase-3 was detected by immunohistochemistry and Western blotting (16, 17). As shown in Fig. 3B and C, Sa pretreatment significantly reduced the protein expression of cleaved caspase-3 in the renal tissues of septic mice. At the same time, compared with the LPS group, Sa downregulated Bax protein expression and upregulated Bcl-2 protein expression in the renal tissues of septic mice. In addition, qPCR analysis further showed that Sa could reduce the expression of Bax in the renal tissues of septic mice (Fig. 3D) and increase that of Bcl-2. Together, the above experimental results showed that Sa had a certain antiapoptotic effect on the kidney of LPS-induced mice with septic nephritis.

FIG 3.

Sa reduces renal cell apoptosis in LPS-induced mice. (A) TUNEL staining was used to detect cell apoptosis in renal tissues of mice in each group. (B) Immunohistochemistry was used to detect the expression of cleaved caspase-3 in renal tissues of mice in each group. (C) Western blotting was used to detect the protein expression of Bax, Bcl-2, and cleaved caspase-3 in renal tissues of mice in each group. (D) RT-qPCR was used to detect the expression of Bax and Bcl-2 in renal tissues of mice in each group. There were 10 mice in each group. The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

Sa downregulated the expression of TCF7.

After proving that Sa can reduce the renal inflammatory responses in septic mice, we further explored the possible targets of Sa to alleviate sepsis-induced nephritis through network pharmacology analysis. A total of 65 target genes of Sa were obtained from the CTD database (http://ctdbase.org/), and 988 and 2,277 sepsis-related genes were obtained from the CTD database and the Phenolyzer database (http://phenolyzer.wglab.org/; screening condition, score > 0.14), respectively. A total of 35 potential targets for Sa treatment of sepsis were obtained by the intersection of Sa target genes and sepsis-related targets (Fig. 4A). The drug-target relationship network was drawn using Cytoscape software (https://cytoscape.org/) (Fig. 4B). The coexpression network of potential targets was obtained from the Coexpedia database (http://www.coexpedia.org/) (Fig. 4C), in which TCF7 was a transcription activator (18), and the CTD database indicated that Sa inhibited the expression of TCF7. Therefore, we first investigated whether Sa affected the expression of TCF7. As shown in Fig. 4D, compared with the sham group, TCF7 expression was upregulated in the renal tissues of septic mice in the LPS group. Compared to the LPS + oe-NC group, TCF7 expression was reduced in the LPS + Sa + oe-NC group and, in contrast, it was elevated in the LPS + oe-TCF7 group. In the LPS + Sa + oe-TCF7 group, the expression of TCF7 was significantly upregulated compared to that in the LPS + Sa + oe-NC group. (For descriptions of the mouse group designations, see “Animal experiments” in Materials and Methods.) As shown in Fig. 4E, the results on the protein expression of TCF7 determined by Western blotting were consistent with the trend in gene expression. These results demonstrate that Sa can effectively inhibit the expression of TCF7 and resist the increase of TCF7 expression induced by LPS.

FIG 4.

Sa downregulates the expression of TCF7. (A) Venn map of intersection of Sa target genes and sepsis-related targets. (B) Drug-target relationship network. (C) Coexpression network of potential targets. (D) TCF7 expression in renal tissues of mice in each group was detected by RT-qPCR (10 mice in each group). (E) TCF7 protein expression in renal tissues of mice in each group was detected by Western blotting (10 mice in each group). The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

Sa inhibited FOSL1 by inhibiting TCF7.

After proving that Sa can regulate the expression of TCF7, we further predicted the downstream regulatory factors of TCF7. A total of 213 regulatory factors of TCF7 were obtained from the ChIP-Atlas database (http://chip-atlas.org/target_genes), and 396 and 2,277 sepsis-related genes were obtained from the CTD and Phenolyzer databases, respectively. Following Venn diagram analysis, 21 candidate genes were found at the intersection (Fig. 5A). The functional relationship network among 21 candidate genes was obtained by using the GeneMANIA tool (http://genemania.org/). The scores of the PRKDC, PARP1, EGR1, and FOSL1 genes were higher than the scores of other genes (Fig. 5B). A search of the relevant literature showed no evidence of a relationship between FOSL1 and sepsis, which attracted our attention, and we thus investigated whether Sa could regulate FOSL1 expression through TCF7.

FIG 5.

Sa inhibits FOSL1 expression by suppressing TCF7 expression. (A) Venn map of intersection between TCF7 regulatory factors in ChIP-Atlas database and sepsis-related genes in CTD database and Phenolyzer database. (B) Functional relationship network between candidate genes. (C) Bioinformatics analysis of the binding sites between TCF7 and FOSL1. (D) Dual luciferase reporter gene assay confirmation of TCF7 binding to FOSL1 in MPC-5 cells. (E) ChIP-qPCR analysis of the TCF7 binding to the FOSL1 promoter fragments. (F) RT-qPCR (left) and Western blot (right) detection of FOSL1 expression in MPC-5 cells in each group. (G) qPCR detection of FOSL1 expression in renal tissues of mice in each group (10 mice in each group). (H) Western blot to detect FOSL1 protein expression in renal tissues of mice in each group (10 mice in each group). (I) Immunohistochemical detection of TCF7 and FOSL1 proteins in renal tissues of mice in each group (10 mice in each group). (J) Fitting curve of immunohistochemical score of FOSL1 and TCF7 in renal tissues of mice for panel I. The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups. Cell experiments were conducted three times independently.

Bioinformatics analysis results revealed possible binding sites between TCF7 and FOSL1 (Fig. 5C). Dual luciferase reporter gene assay results further verified that TCF7 could target FOSL1 and promoted the transcriptional activity of FOSL1 (Fig. 5D). Moreover, the results of chromatin immunoprecipitation (ChIP)-qPCR suggested that FOSL1 was hardly detected in the anti-IgG group, but the enriched FOSL1 promoter fragments could be detected in the anti-TCF7 group (Fig. 5E), indicating that TCF7 can directly bind to the FOSL1 promoter. Meanwhile, RT-qPCR and Western blot data suggested a decline in the expression of FOSL1 in MPC-5 cells in the shRNA against TCF7 (sh-TCF7) group relative to the shRNA negative control (sh-NC) group, while a contrasting result was observed in the overexpression of TCF7 (oe-TCF7) group relative to the oe-NC group (Fig. 5F). These results indicated that TCF7 can positively regulate the expression of FOSL1.

As shown in Fig. 5G, compared with the sham group, the expression of FOSL1 was significantly increased in the renal tissue of septic mice in the LPS group. Compared with the LPS + oe-NC group, the LPS + Sa + oe-NC group exhibited a significant decrease in the expression of FOSL1 and, conversely, the expression of FOSL1 was enhanced in the LPS + oe-TCF7 group. In addition, the LPS + Sa + oe-TCF7 group showed an increase in the expression of FOSL1 compared with the LPS + Sa + oe-NC group, indicating that the inhibiting effect of Sa on FOSL1 expression was blocked by oe-TCF7. Meanwhile, the protein expression of FOSL1 in mouse renal tissue was detected by Western blotting, and the result was consistent with the gene expression result (Fig. 5H). Additionally, as shown in Fig. 5I and J, the results of immunohistochemical staining analysis of the renal tissues of septic mice showed that FOSL1 expression was positively correlated with TCF7 expression. In conclusion, Sa can downregulate the expression of FOSL1 by inhibiting TCF7.

Sa inhibited TCF7/FOSL1 to reduce cell apoptosis and inflammatory factor expression in renal tissues.

After demonstrating that Sa can inhibit FOSL1 by inhibiting TCF7, we then tried to illustrate how the inhibition of TCF7/FOSL1 by Sa affects inflammation and cell apoptosis in renal tissues. As shown in Fig. 6A, compared with the sham group, the expression of inflammatory factors IL-1β, IL-6, and TNF-α was increased in the renal tissues of septic mice in the LPS group. In addition, compared with the LPS + oe-NC group, the expression of inflammatory factors IL-1β, IL-6, and TNF-α was decreased in the renal tissues of septic mice in the LPS + Sa + oe-NC group while the LPS + oe-TCF7 group showed opposite results. Additionally, the expression of inflammatory factors IL-1β, IL-6, and TNF-α in the renal tissues of septic mice in the LPS + Sa + oe-TCF7 group was elevated in contrast to the LPS + Sa + oe-NC group. Meanwhile, similar results were found in the protein expression of these factors by Western blotting (Fig. 6B). After that, TUNEL staining was utilized to investigate the effect of Sa on cell apoptosis by inhibiting the TCF7/FOSL1 axis. As illustrated in Fig. 6C, compared with the sham group, the cell apoptosis rate in renal tissues was increased in the LPS group. Compared to the LPS + oe-NC group, the cell apoptosis rate was attenuated in the LPS + Sa + oe-NC group while it was increased in the LPS + oe-TCF7 group. In addition, the LPS + Sa + oe-TCF7 group had a higher cell apoptosis rate than the LPS + Sa + oe-NC group. The results of Western blotting showed that the protein expression of cleaved caspase-3 and Bax was higher and that of Bcl-2 was lower in the LPS group than in the sham group. In contrast, the LPS + Sa + oe-NC group showed lower cleaved caspase-3 and Bax protein expression and higher Bcl-2 protein expression than the LPS + oe-NC group, but the LPS + oe-TCF7 group showed opposite results. In addition, the protein expression of cleaved caspase-3 and Bax was augmented while that of Bcl-2 was attenuated in the LPS + Sa + oe-TCF7 group relative to the LPS + Sa + oe-NC group (Fig. 6D). In conclusion, Sa can reduce the expression of inflammatory factors and cell apoptosis in septic mice by inhibiting the TCF7/FOSL1 axis.

FIG 6.

Sa inhibits the TCF7/FOSL1 axis to reduce cell apoptosis and inflammatory factor expression in septic mice. (A) The expression of proinflammatory factors IL-6, IL-β, and TNF-α in renal tissues of mice in each group was detected by qPCR. (B) The protein expression of proinflammatory factors IL-6, IL-β, and TNF-α in renal tissues of mice in each group was detected by Western blotting. (C) The cell apoptosis in renal tissues of mice in each group was detected by TUNEL staining. (D) The expression of cleaved caspase-3, Bax, and Bcl-2 in renal tissues of mice in each group was detected by Western blotting. There were 10 mice in each group. The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

TCF7 activates FOSL1 and upregulates MMP9.

We then sought to predict the downstream regulatory genes of FOSL1. Differential analysis of the GEO accession no. GSE28750 data set revealed 110 differentially expressed genes, including 67 highly expressed genes and 43 poorly expressed genes (Fig. 7A). A heat map of the obtained differentially expressed genes was drawn according to the differential gene expression profile (Fig. 7B). Following Venn diagram analysis of the differentially expressed genes and sepsis-related genes obtained from the DisGeNET database, MMP9 was found at the intersection (Fig. 7C). MMP9 was highly expressed in sepsis samples in the GSE28750 data set (Fig. 7D). A significant positive correlation was found between FOSL1 expression and MMP9 expression in renal tissues (Pearson coefficient r = 0.7381, P = 1.11e−5) (Fig. 7E). Subsequently, we investigated whether TCF7 regulates the expression of MMP9 by promoting the expression of FOSL1.

FIG 7.

TCF7 upregulates the expression of MMP9 by increasing the expression of FOSL1. (A) Volcano map of differential gene expression in the sepsis-related GSE28750 data set. The −log₁₀ (P value) is expressed on the abscissa, and the ordinate represents log2FC; green dots indicate significantly downregulated genes, and red dots indicate significantly upregulated genes. (B) The expression heat map of differential genes. The green to red color scale indicates the expression value from low to high, respectively. (C) Venn diagram of intersection of differentially expressed genes (DEGs) and sepsis-related genes in the DisGeNET database. (D) MMP9 expression in sepsis samples in the GSE28750 data set. (E) Coexpression of FOSL1 and MMP9 in renal tissues in the ChIPbase database. (F) Bioinformatics analysis of the binding sites between FOSL1 and MMP9. (G) Dual luciferase reporter gene assay confirmation of FOSL1 binding to MMP9 in MPC-5 cells. (H) ChIP-qPCR analysis of the FOSL1 binding to the MMP9 promoter fragments. (I) RT-qPCR (left) and Western blot (right) detection of MMP9 expression in MPC-5 cells in each group. (J) qPCR was used to detect MMP9 expression in renal tissues of mice in each group. (K) Western blotting was used to detect the protein expression of MMP9, FOSL1, and TCF7 in renal tissues of mice in each group (10 mice in each group). The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups. Cell experiments were conducted three times independently.

Potential FOSL1 binding sites in the promoter region of MMP9 were predicted following bioinformatics analysis (Fig. 7F), and the results of a dual luciferase reporter gene assay suggested that FOSL1 had a positive regulatory effect on the transcription of MMP9 (Fig. 7G). Further results of ChIP-qPCR verified the binding ability of FOSL1 to the MMP9 promoter fragment, as shown by the fact that MMP9 was not present in the anti-IgG group but the enriched MMP9 promoter fragment could be detected in the anti-FOSL1 group (Fig. 7H), proving that FOSL1 can bind to the MMP9 promoter directly. As depicted in Fig. 7I, MMP9 mRNA and protein expression was lowered in the MPC-5 cells in the sh-FOSL1 group relative to the sh-NC group while a contrasting result was found in the MPC-5 cells in the oe-FOSL1 group compared to the oe-NC group, suggesting that FOSL1 can positively regulate the expression of MMP9.

As shown in Fig. 7J, RT-qPCR data displayed higher MMP9 mRNA expression in the renal tissues of mice in the LPS group than in the sham group. Compared with the LPS + sh-NC + oe-NC group, MMP9 expression was reduced in the LPS + sh-TCF7 + oe-NC group while it was elevated in the LPS + oe-FOSL1 + sh-NC group. In addition, the LPS + sh-TCF7 + oe-FOSL1 group exhibited an enhancement in MMP9 expression compared to the LPS + sh-TCF7 + oe-NC group. At the same time, as shown in Fig. 7K, we detected MMP9 protein expression in mouse renal tissues by Western blotting and found that the trend of MMP9 protein expression was consistent with the mRNA expression trend. In addition, protein expression of TCF7 and FOSL1 was higher in the renal tissues of mice in the LPS group than in the sham group. In comparison to the LPS + sh-NC + oe-NC group, protein expression of TCF7 and FOSL1 was reduced in the LPS + sh-TCF7 + oe-NC group while TCF7 protein expression showed no changes, and FOSL1 protein expression was increased in the LPS + oe-FOSL1 + sh-NC group. In addition, the LPS + sh-TCF7 + oe-FOSL1 group exhibited unchanged TCF7 protein expression yet an enhancement in FOSL1 expression relative to the LPS + sh-TCF7 + oe-NC group. In conclusion, TCF7 can upregulate the expression of MMP9 by promoting the expression of FOSL1.

Sa inhibited the TCF7/FOSL1/MMP9 axis to reduce cell apoptosis and inflammatory factor expression in vivo.

After demonstrating that Sa can inhibit the TCF7/FOSL1/MMP9 axis, we then aimed to investigate the effect of Sa on inflammation and cell apoptosis in mice through the TCF7/FOSL1/MMP9 axis. As shown in Fig. 8A, compared with the sham group, the expression of inflammatory factors IL-1β, IL-6, and TNF-α in the renal tissues of septic mice in the LPS group was significantly increased. In comparison to the LPS + oe-NC group, their expression was reduced in the LPS + Sa + oe-NC group while the LPS + oe-MMP9 group had opposite results. In the LPS + Sa + oe-MMP9 group, the expression of IL-1β, IL-6, and TNF-α was enhanced in comparison to that in the LPS + Sa + oe-NC group. The results of Western blotting were consistent with those of the above-described gene expression determination (Fig. 8B). TUNEL assay results showed that the cell apoptosis rate was enhanced in the LPS group compared with the sham group. In comparison to the LPS + oe-NC group, the cell apoptosis rate was significantly attenuated in the LPS + Sa + oe-NC group, while it was enhanced in the LPS + oe-MMP9 group. In the LPS + Sa + oe-MMP9 group, the cell apoptosis rate was promoted in comparison to that in the LPS + Sa + oe-NC group (Fig. 8C). Furthermore, the results of Western blotting suggested an increase in the protein expression of cleaved caspase-3 and Bax yet a reduction in the protein expression of Bcl-2 in the renal tissues of septic mice in the LPS group relative to the sham group. Compared with the LPS + oe-NC group, the protein expression of cleaved caspase-3 and Bax in the renal tissues of septic mice in the LPS + Sa + oe-NC group was significantly decreased, and that of Bcl-2 was significantly increased. However, oe-MMP9 blocked the tendency of Sa to decrease cleaved caspase-3 and Bax and increase Bcl-2 protein expression (Fig. 8D). These results indicate that Sa can reduce the expression of inflammatory factors and cell apoptosis in renal tissues of septic mice by inhibiting the TCF7/FOSL1/MMP9 axis.

FIG 8.

Sa inhibits the TCF7/FOSL1/MMP9 axis to reduce cell apoptosis and inflammatory factor expression in mice. (A) The expression of proinflammatory factors IL-6, IL-β, and TNF-α in renal tissues of mice in each group was detected by qPCR. (B) The protein expression of proinflammatory factors IL-6, IL-β, and TNF-α in renal tissues of mice in each group was detected by Western blotting. (C) The cell apoptosis in renal tissues of mice in each group was detected by TUNEL staining. (D) The expression of cleaved caspase-3, Bax, and Bcl-2 in renal tissues of mice in each group was detected by Western blotting. There were 10 mice in each group. The measurement data are expressed as the mean ± standard deviation. *, P < 0.05, indicates data comparison between two groups.

DISCUSSION

Sepsis is considered to be a common lethal condition characterized by an uncontrolled and detrimental host reaction to microbial infection (19). In this study, we investigated the molecular events of Sa in sepsis, and our results revealed that Sa alleviated renal inflammation and cell apoptosis in sepsis via regulation of the TCF7/FOSL1/MMP9 axis (Fig. 9).

FIG 9.

Graphical summary of the molecular mechanism underlying renoprotective effects of Sa against sepsis. Sa inhibits FOSL1 transcription by downregulating the expression of TCF7, thereby inhibiting MMP9 expression, which ultimately reduces the renal inflammation and cell apoptosis induced by sepsis in mice.

In the first place, we found that Sa could contribute to the alleviation of renal inflammation and cell apoptosis in sepsis and that this effect was achieved by downregulating TCF7. Intriguingly, it has been previously reported that Sa could prevent renal tubular epithelial cells from high-glucose-induced injury by regulating SIRT3 (8). In addition, Ma et al. revealed that Sa could attenuate cisplatin-induced nephrotoxicity through the regulation of the c-Jun N-terminal kinase/nuclear factor kappa B (JNK/NF-κB) signaling pathway in part by diminishing the release of proinflammatory cytokines and inhibiting apoptosis of cisplatin-treated HK-2 cells (20). Notably, the regulation of TCF7 by Sa has been rarely reported. In the present study, our database-based bioinformatics analysis combined with Western blot assay found that Sa could effectively inhibit the transcription of TCF7 in sepsis. Importantly, the involvement of TCF7 in sepsis has been revealed. As previously reported, TCF7 was differentially expressed in peritoneal cells in mice with sepsis at early as well as late time points (9). Moreover, TCF7L2 has been suggested to participate in the differentiation and inflammation processes in chronic inflammatory diseases as well as sepsis (21). Overall, the aforementioned reports are supportive of our finding regarding the role of Sa in renal inflammation in sepsis through regulation of TCF7.

Mechanistically, the present study found that the downregulated expression of TCF7 could lead to inhibition of FOSL1 transcription, which thus inhibited the expression of MMP9, thereby reducing renal inflammation and cell apoptosis in septic mice. It has been revealed that TCF7 could increase FOSL1 transcription in perihilar cholangiocarcinoma (10). Strikingly, a recent study revealed that FOSL1 can function as a mediator of deviant angiogenic signaling in sepsis-stimulated pulmonary endothelial cells (12). Additionally, FOSL1 was identified through an upregulated inflammatory gene analysis as an important participator in the inflammation response during the aging processes in kidney (22). Moreover, miR-21-regulated FOSL1 was found to regulate apoptosis in kidney regeneration in fish (23). To our acknowledge, several studies have found that FOSL1 can regulate MMP9. For instance, as previously reported, FOSL1 could occupy the promoter regions of the MMP9 in trophoblast cells and silencing of FOSL1 could lead to disruption of the expression of MMP9 (24). Furthermore, an increasing number of studies have found that MMP9 is involved in sepsis and renal inflammation. Downregulation of MMP9 due to pretreatment with AZD4547 could exert a protective role against excessive inflammatory damage in a mouse model of sepsis (15). Besides, MMP9 has been revealed as one of the differentially expressed genes participating in the pathogenesis of sepsis (25). In addition, downregulation of MMP9 by saroglitazar could contribute to attenuation of unilateral ureteral obstruction-induced renal fibrosis in part by decreasing the inflammation response (26).

From the results obtained in the present study, we reached the conclusion that Sa inhibits FOSL1 transcription by downregulating TCF7, thereby inhibiting MMP9 expression, which ultimately reduces the renal inflammation and cell apoptosis induced by sepsis in mice. We hope that this finding may provide a novel direction for treatment of sepsis. Nevertheless, further verification is required to verify the clinical feasibility.

MATERIALS AND METHODS

Ethical approval.

This study was performed under the approval of the Animal Ethics Committee of the Affiliated People’s Hospital of Ningbo University and strictly performed according to the Guide for the Care and Use of Laboratory Animals published by the U.S. National Institutes of Health. Extensive efforts were made to ensure minimal suffering of the included animals.

Network pharmacology-based analysis.

The sepsis-related data set under accession no. GSE28750 was retrieved from the GEO database (https://www.ncbi.nlm.nih.gov/geo/), with the platform annotation file of GPL570. The GSE28750 data set contains 20 normal samples (control) and 10 sepsis samples (sepsis). Differential analysis was conducted using the R language “limma” package (http://www.bioconductor.org/packages/release/bioc/html/limma.html) with |log2FC| > 2, P < 0.05 set as the threshold. A heat map of the candidate differentially expressed genes was drawn using the R language “pheatmap” package (https://cran.r-project.org/web/packages/pheatmap/index.html). Next, the target genes of saidosaponin-d were predicted using the CTD database (http://ctdbase.org/) with “Saikosaponin D” set as the key word. The CTD (screening criteria: score > 40) and Phenolyzer (http://phenolyzer.wglab.org/) (screening criteria: score > 0.14) databases were adopted to search sepsis-related genes with “sepsis” set as the key word. The jvenn tool (http://jvenn.toulouse.inra.fr/app/example.html) was used to obtain the intersection results of the saidosaponin-d target genes and sepsis-related targets, in order to predict the potential target of saidosaponin-d in the treatment of sepsis. The Cytoscape 3.5.1 software (https://cytoscape.org/) was used to plot the drug-target network interaction map, and the Coexpedia database (http://www.coexpedia.org/) was used for coexpression analysis between potential targets. Colocalization analysis of transcriptional factors was conducted by use of the ChIP-Atlas database (http://chip-atlas.org/target_genes). The jvenn tool was used for the intersection analysis of transcription factors and disease-related genes. The functional relationship between genes was analyzed using the GeneMANIA tool (http://genemania.org/), and candidate factors were obtained according to the gene score on the website. Sepsis-related genes were retrieved by the DisGeNET database (https://www.disgenet.org/) (score > 0.5) and then subjected to intersection analysis with the differentially expressed genes in the sepsis-related GSE28750 data set to predict the downstream regulatory factors of FOSL1. The ChIPBase database (http://rna.sysu.edu.cn/chipbase/) was used to analyze the coexpression relationship between FOSL1 and the downstream factors in renal tissues. The JASPAR database (http://jaspar.genereg.net/) was employed to predict the binding sites between transcription factors and downstream target gene promoters.

Chemicals and antibodies.

Sa was purchased from China Chengdu Biotechnology Company, Ltd. (Chengdu, China). The primary antibodies are listed in Table 1.

TABLE 1.

Primary antibodies used in the study^a

Antibody	Company	Catalog no.
Anti-TCF7	Cell Signal Technology, USA	2203S
Anti-FOSL1	Abcam, UK	ab232745
Anti-MMP9	Abcam, USA	ab76003
Anti-IL-1β	Abcam, USA	ab9722
Anti-NGAL	Abcam, USA	ab63929
Anti-IL-6	HuaAn Biotechnology, China	EM170414
Anti-TNF-α	Affinity Biosciences, USA	AF7014
Anti-HMGB1	Abcam, USA	ab18256
Anti-COX-2	HuaAn Biotechnology, China	ET-1610-23
Anti-iNOS	Abcam, USA	ab3523
Anti-Bax	Abcam, USA	ab32503
Anti-Bcl-2	Cell Signaling Technology, USA	3498S
Anti-cleaved caspase-3	Proteintech Group, USA	66470-2-Ig
Anti-GAPDH	ZenBio Inc, USA	200306-7E4
Antilectin	Santa Cruz Biotechnology, Inc., USA	sc-324317

^aTCF7, transcription factor 7; FOSL1, Fos-like antigen 1; MMP9, matrix metalloproteinases 9; IL, interleukin; NGAL, neutrophil gelatinase-associated lipocalin; TNF-α, tumor necrosis factor alpha; HMGB1, high-mobility-group box 1 protein; COX-2, cyclooxygenase 2; iNOS, inducible nitric oxide synthase; Bax, Bcl-2-associated X protein; Bcl-2, B cell lymphoma/leukemia 2; GAPDH, glyceraldehyde-3-phosphate dehydrogenase.

Cell culture.

Mouse podocytes (MPC-5), purchased from Mingzhou Biotechnology Co., Ltd. (Ningbo, China), were cultured in Dulbecco’s modified Eagle’s medium (DMEM; Thermo Fisher Scientific, Inc., Waltham, MA) supplemented with 10% fetal bovine serum (FBS), 100 μg/ml streptomycin, and 100 IU/ml penicillin (both purchased from Sigma-Aldrich Chemical Company, St. Louis, MO) in a 5% CO₂ incubator at 37°C.

Lentiviral infection and establishment of stably transduced cell lines.

The human TCF7 gene was amplified from cDNA by PCR and cloned into pLenti-EF1a lentivirus vectors. In order to silence TCF7 in HEK293T cells, two TCF7 short hairpin RNAs (shRNAs) and packaging plasmids GV, pHelper1.0, and pHelper2.0 were used for transfection. MPC-5 cells were seeded into a 6-well plate at a density of 2 × 10⁶ cells/well and transfected with 10 μg plasmid. After 48 h, the lentiviral supernatant was collected and used to infect MPC-5 cells in the presence of 8 μg/ml polypropylene (Sigma-Aldrich). Stable cell lines were treated for 14 days with the medium containing 2 μg/ml puromycin and verified by Western blotting. The sequences of shRNA were as follows: control shRNA, 5′-TTCTCCGAACGTGTCACGT-3′; TCF7 shRNA-1, 5′-GGACATCAGCCAGAAGCAAGT-3′; TCF7 shRNA-2, 5′-CCAAGAAGCCAACCATCAA-3′; shFOSL1, 5′-CCTCAGCTCATCGCAAGAGTA-3′.

Dual luciferase reporter gene assay.

MPC-5 cells (3 × 10⁴ cells/well) were seeded in 24-well plates in triplicate and incubated for 24 h. Then, 1.5 ng pRL-TK Renilla plasmid was transfected into the cells using Lipofectamine 2000 transfection reagent (Invitrogen, Carlsbad, CA). At 48 h after transfection, luciferase reporter assay kits (Promega, Madison, WI) were used to determine the luciferase activity and Renilla activity according to the manufacturer’s instructions. The human FOSL1 and MMP9 promoter regions generated from MPC-5 cells were amplified by PCR and cloned into the NheI/BglII site of pGL3 basic dual luciferase reporter gene plasmid to generate FOSL1 or MMP9 luciferase reporter gene.

ChIP.

ChIP was performed using an EZ-Magna ChIP kit (17-371; Millipore, Billerica, MA). In brief, MPC-5 cells were subjected to ultrasonic treatment and centrifuged at 2,000 × g and 4°C for 10 min to remove the insoluble precipitate. Next, the cells were incubated with protein G-agarose at 4°C for 1 h and centrifuged at 5,000 × g for 1 min to obtain supernatant. Afterwards, 10 μl (1%) of the supernatant was taken out as “input,” and the remaining supernatant was divided into 2 parts and probed with TCF7 antibody (2203S, 1:50; Cell Signaling Technology, Beverly, MA) or FOSL1 antibody (ab278103, 1:30; Abcam, Cambridge, UK) and NC rabbit anti-human IgG (ab2410, 1:25; Abcam) overnight at 4°C for complete binding. The following day, incubation with the protein G-agarose was conducted at 4°C for 1 h to precipitate protein and DNA complexes, followed by centrifugation at 5000 × g for 1 min, with the supernatant discarded. The DNA-protein complex was eluted and incubated at 65°C overnight to relieve cross-linking. The DNA fragments were purified and recovered. Finally, the precipitated DNA was analyzed using RT-qPCR.

Animal experiments.

Male C57BL/6JNifdc mice (9 to 10 weeks old, weighing 25 to 27 g; purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd., Beijing, China) were housed in a temperature-controlled room (23 ± 2°C) with available food and water under a 12-h light/dark cycle. The experiment was conducted after 1 week of acclimatization.

The mice were randomly divided into six groups with 10 mice in each group: sham group (sham), Sa low-dose (Sa-L) group (5 mg/kg), Sa high-dose (Sa-H) group (20 mg/kg), model group (LPS group), LPS + Sa-L (5 mg/kg) group, and LPS + Sa-L (20 mg/kg) group. Intragastric administration of Sa was conducted 7 days before modeling, once a day. The LPS group and Sa group were injected intraperitoneally with 20 mg/kg LPS (E. coli LPS serotype 0111:B4; Sigma-Aldrich) to construct the model of renal inflammation in sepsis.

In addition, normal mice and model mice with LPS-induced septic nephritis were divided into four groups. The first grouping was used to detect the effect of Sa on TCF7 expression: sham group, LPS group (LPS was used to induce septic nephritis), LPS + oe-NC group (LPS mice injected with oe-NC via tail vein), LPS + Sa + oe-NC group (LPS mice receiving intragastric administration of 20 mg/kg Sa and tail vein injection of oe-NC), LPS + oe-TCF7 (LPS mice injected with oe-TCF7 via tail vein), and LPS + Sa + oe-TCF7 group (LPS mice receiving intragastric administration of 20 mg/kg Sa and tail vein injection of oe-TCF7), with 10 mice in each group (for all group designations, the “oe-” prefix represents overexpressed and NC represents negative control). The second grouping was used to detect the regulation of Sa on FOSL1 expression through TCF7, with the group setting being the same as the first grouping, with 10 mice in each group. The third grouping was used to detect the regulation of TCF7 on the expression of MMP9 through FOSL1: sham group (n = 10), LPS group (LPS was used to induce septic nephritis), LPS + sh-NC + oe-NC group (LPS mice injected with sh-NC and oe-NC via tail vein), LPS + sh-TCF7 + oe-NC group (LPS mice injected with sh-TCF7 and oe-NC via tail vein), LPS + oe-FOSL1 + sh-NC group (LPS mice injected with oe-FOSL1 and sh-NC via tail vein), and LPS + sh-TCF7 + oe-FOSL1 group (LPS mice injected with sh-TCF7 and oe-FOSL1 via tail vein), with 10 mice in each group (for all group designations, the “sh-” prefix represents shRNA). The fourth grouping was used to detect the effect of Sa on the cell apoptosis and production of inflammatory factors by inhibiting the TCF7/FOSL1/MMP9 axis: sham group, LPS group (LPS was used to induce septic nephritis), LPS + oe-NC group (LPS mice injected with oe-NC via tail vein), LPS + Sa + oe-NC group (LPS mice receiving intragastric administration of 20 mg/kg Sa and tail vein injection of oe-NC), LPS + oe-MMP9 group (LPS mice injected with oe-MMP9 via tail vein), and LPS + Sa + oe-MMP9 group (LPS mice receiving intragastric administration of 20 mg/kg Sa and tail vein injection of oe-MMP9), with 10 mice in each group. After incubation at room temperature for 15 min, the prepared solution was given to mice. The sequence of shMMP9 was synthesized by Invitrogen (Thermo Fisher Scientific, Inc.), and shFOSL1 was purchased from Santa Cruz Biotechnology (Santa Cruz, CA).

Measurement of renal vascular leakage.

Mice were injected with normal saline containing 1% Evans blue staining solution (Sigma-Aldrich) through the tail vein. After 40 min, the mice were euthanized and renal tissues were collected. The kidney weight was measured, and the kidney was placed in 1 ml of formamide (Avantor, Center Valley, PA) for 24 h at 60°C to extract Evans blue dye. The samples were centrifuged at a speed of 626 × g for 10 min, and the supernatant was collected. The concentration of Evans blue dye in the supernatant was quantified by measuring the absorbance at 620 nm and calculated by a flat-panel reader according to the standard curve.

Determination of BUN and SCr.

Twelve hours after intraperitoneal injection of LPS, whole blood was collected from the mice in each group and transferred to a test tube containing EDTA (BD Vacutainer, Franklin Lakes, NJ). The plasma was centrifuged at a rate of 1,570 × g for 30 min and stored at −80°C for analysis. BUN and SCr levels were detected by an automatic biochemical analyzer (TC6010 L; Tecom Science Corporation, Jiangxi, China). The SCr level was determined by the picric acid method, and the BUN level was determined by the urease method.

Histological examination of kidney.

The renal tissue was fixed in 10% neutral formalin buffer, embedded in paraffin, and cut into 4-μm-thick sections. After paraffin removal and fluid infusion, renal sections were stained with PAS or HE. The sections were observed by optical microscopy. For the semiquantitative analysis of morphological changes, two slices were randomly selected from each sample in each group (at least 3 samples), and 10 visual fields were randomly selected from each slice. Histopathological changes were assessed by the percentage of injured/damaged tubules as indicated by tubular dissolution, dilatation, rupture, and tubule formation. The tissue injury scores of 0, 1, 2, 3, and 4 corresponded to 0%, <25%, 26% to 50%, 51% to 75%, and ≥76% of the injury respectively.

TUNEL staining.

The renal tissue was fixed in 10% neutral formalin buffer, embedded in paraffin, and cut into 4-μm-thick sections. According to the experimental scheme, the DeadEnd TM fluorescence TUNEL system (G3250; Promega) was used for TUNEL staining. The slices were then incubated with DAPI (4′,6-diamidino-2-phenylindole) (D8200; Solarbio, Beijing, China) at a dilution of 1:500. The images were output by a fluorescence microscope. The positive cells were counted, and at least 10 visual fields were examined in each section of each sample.

Immunofluorescence staining.

The renal specimens were embedded in OCT compound, cut into 4-μm sections in a cryostat, and stored at −80°C for subsequent use. For immunofluorescence staining, the sections were rehydrated and sealed with phosphate-buffered saline (PBS) plus 5% normal goat serum for 1 h and then labeled overnight in a humidification chamber at –4°C. The sections were exposed to Cy5 red-labeled or fluorescein isothiocyanate (FITC) green-labeled secondary antibody (Jackson Immunoresearch, Inc., West Grove, PA). The nuclei were stained with DAPI (1:500; Life Technologies Corporation, OR). The images were taken using an AxioCam HRc digital camera (Carl Zeiss).

Immunohistochemical staining.

After being fixed overnight in 10% phosphate buffered formalin, the kidneys were dehydrated through a series of graded ethanol, embedded in paraffin, sliced into thin slices (5 μm), and fixed on the slide. The slides were sealed with 2.5% normal goat serum and incubated with the primary antibody at 4°C. The slides were washed three times in PBS and stained using Vectastain ABC kits (Vector, Burlingame, CA) according to the manufacturer’s instructions. Sections were stained with hematoxylin, and then the images were captured with an AxioCam HRc digital camera (Carl Zeiss).

RT-qPCR.

According to the experimental scheme, total RNA was extracted from kidney by using the total RNA extraction kit (BioTek, Winooski, VT). The mRNA concentration was measured by a ScanDrop 100 analyzer (AnalytikJena, Thuringia, Germany). After reverse transcription, quantitative real-time PCR was performed in a PCR system (CFX connect; Bio-Rad, Hercules, CA) using fast qPCR kits (Kapa Biosystems, Foster, CA). The primer sequences are listed in Table 2. The relative gene expression was normalized by glyceraldehyde-3-phosphate dehydrogenase (GAPDH) or β-actin and calculated by the 2^−ΔΔCT method.

TABLE 2.

Primer sequences for RT-qPCR^a

Gene	Sequence
	Forward primer	Reverse primer
TCF7	TACATGGAGAAGCCGAGGGA	ACTCTGGAAGTTTGTCCGGG
FOSL1	ATCCCCGACCTCTGACCTAT	CAAGGCGTTCCTTCTGCTT
MMP9	AAACCTCCAACCTCACGGAC	CTGAAGCATCAGCAAAGCCG
NGAL	GCAGGTGGTACGTTGTGGG	CTCTTGTAGCTCATAGATGGTGC
MCP-1	CATCCACGTGTTGGCTCA	GATCATCTTGCTGGTGAATGAGT
KIM-1	ACATATCGTGGAATCACAACGAC	ACTGCTCTTCTGATAGGTGACA
Bcl-2	GCGTCAACAGGGAGATGTCA	GCATGCTGGGGCCATATAGT
Bax	CTGGATCCAAGACCAGGGTG	GTGAGGACTCCAGCCACAAA
TNF-α	GATCGGTCCCCAAAGGGATG	CCACTTGGTGGTTTGTGAGTG
IL-6	TGGTCTTCTGGAGTACCATAGC	TGTGACTCCAGCTTATCTCTTGG
IL-1β	TGCCACCTTTTGACAGTGATG	TGATGTGCTGCTGCGAGATT
GAPDH	GTATGACTCCACTCACGGCAAA	GGTCTCGCTCCTGGAAGATG

^aTCF7, transcription factor 7; FOSL1, Fos-like antigen 1; MMP9, matrix metalloproteinase 9; IL, interleukin; NGAL, neutrophil gelatinase-associated lipocalin; TNF-α, tumor necrosis factor alpha; HMGB1, high-mobility-group box 1 protein; COX-2, cyclooxygenase 2; iNOS, inducible nitric oxide synthase; Bax, Bcl-2-associated X protein; KIM-1, kidney injury molecule 1; Bcl-2, B cell lymphoma/leukemia 2; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; RT-qPCR, reverse transcription-quantitative PCR.

Western blot assay.

The renal cortex of mice was dissected and homogenized in radioimmunoprecipitation assay (RIPA) lysis buffer (P0013B; Beyotime, Shanghai, China). After centrifugation at 4°C and 13,000 rpm for 15 min, the supernatant was collected, and the protein concentration was determined using bicinchoninic acid (BCA) kits (Beyotime). Bovine serum albumin was used as the standard. The same amount of protein lysate was directly loaded onto 10 to 12% sodium dodecyl sulfate polyacrylamide gel electrophoresis gel and transferred to a polyvinylidene fluoride (PVDF) membrane for Western blotting (0.2 μm; Bio-Rad). The membrane was sealed with 5% (wt/vol) skimmed milk powder containing Tris-buffered saline containing 0.1% Tween 20 (TBST) at room temperature for 1 h and then incubated with the indicated primary antibody at 4°C overnight. The membrane was then treated with horseradish peroxidase-labeled goat anti-rabbit antibody against IgG (1:2,000 in dilution; Beijing Biosynthesis Biotechnology Co., Ltd., Beijing, China) or goat anti-mouse antibody against IgG (1:2,000 in dilution; Beijing Biosynthesis Biotechnology Co., Ltd., Beijing, China) for 1 h. Immobilon Western chemiluminescence HRP substrate (Millipore Corporation, Billerica, MA) and a Bio-Rad ChemiDoc MP were used to observe the Western blotting results. The optical density was analyzed by ImageJ 6.0 software (National Institutes of Health, Bethesda, MD).

Statistical analysis.

All the data in this study were analyzed utilizing SPSS 21.0 statistical software (IBM Corp., Armonk, NY). The measurement data were expressed as the mean ± standard deviation. Data between two groups were compared using an independent sample t test. Data among multiple groups were compared using one-way analysis of variance (ANOVA), with Tukey’s post hoc tests. Data at different time points among multiple groups were compared by two-way ANOVA. P values <0.05 indicate statistically significant differences.