Statins enhanced sensitivity to chemotherapy in GC patients, and ILF3 as a potential target of statins.
GC patients who underwent preoperative neoadjuvant chemotherapy were divided into two groups treated with and without statins (≥ 6 months) (50 patients in each group). The Tumor Regression Grading (TRG) system (Table S1) is frequently employed to evaluate this treatment response of GC patients subjected to neoadjuvant therapies, typically chemotherapy or radiotherapy[30]. TRG grading of postoperative pathology were conducted between two groups of GC patients, which found that statins-treated groups exhibited more prominent tumor regression and were more sensitive to chemotherapy (Fig. 1A). The serum ILF3 was significantly reduced in statins-treated groups (Fig. 1B), with the basic clinical information for the two groups of GC patients (Table S2). Serum ILF3 was significantly elevated in GC patients compared to healthy people (Fig. 1C), with the basic clinical information of the two groups (Table S3). The Cancer Genome Atlas (TCGA) analysis found that the mRNA level of ILF3 was higher in GC tissues compared to normal tissues (Fig. 1D). Analysis of the clinical samples from our own hospital also found ILF3 expression was elevated in GC tissues compared to paracancerous tissues, which is consistent with the TCGA outcomes (Fig. 1E). The ILF3 expression was analyzed with hematoxylin-eosin (HE) staining between GC and paracancerous tissues (Fig. 1F) and immunohistochemistry (IHC) in different gastric cancer stages (Fig. 1G) found that ILF3 expression in advanced GC patients surpassed early GC patients (Fig. 1H). Scores 1 and 2 indicated low ILF3 expression, and scores 3 and 4 indicated high ILF3 expression. The Kaplan-Meier plotter and Gene Expression Profiling Interactive Analysis (GEPIA) analyses found that elevated ILF3 expression was associated with poorer overall survival (OS), post-progression survival (PPS), and disease-free survival (DFS) outcomes (Fig. 1I-K). Analysis of the clinical samples from our own hospital found that high ILF3 expression in GC patients had significantly a shorter survival period (Fig. 1L).
ILF3 modulated the expression of PD-L1 and regulated the killing effect of CD8 + T cells on GC cells.
GC patients who underwent preoperative neoadjuvant chemotherapy and immunotherapy were divided into two groups treated with and without statins (≥ 6 months), and then tested the expression ofILF3 and PD-L1, respectively. TRG grading of postoperative pathology found that, in statins-treated group, low-ILF3 expression and low-PD-L1 expression groups exhibited more prominent tumor regression and were more sensitive to chemotherapy (Fig. 2A), with the basic clinical information for GC patients (Table S4-6). The serum PD-L1 was significantly reduced in statins-treated group (Fig. 2B). The Tumor Immune Estimation Resource (TIMER) analysis found that ILF3 was positively correlated with PD-L1 expression in GC tissues (Fig. 2C). The western blot and qRT-PCR results showed a positive correlation between the expression of ILF3 and PD-L1 at both protein and mRNA levels (Fig. 2D-F). Tumor cells released PD-L1 positive extracellular vesicle (EV) to evade the immune system[25]. The ELISA results demonstrated that si-ILF3 reduced the secretion of PD-L1 and oe-ILF3 increased the secretion of PD-L1(Fig. 2G). In addition, immunofluorescence further confirmed the positive correlation between ILF3 and PD-L1 in GC cells (Fig. 2H, Figure S1A, B). CD8+ T cells were activated by co-stimulated with anti-CD3, anti-CD28 antibodies, and IL-2 (Fig. 2I). Activated CD8+ T cells were co-cultured with GC cells following knockdown and overexpression of ILF3 for 48 hours, and the flow cytometry analysis showed that si-ILF3 resulted in an elevated rate of apoptosis in GC cells, whereas oe-ILF3 led to a declined rate of apoptosis (Fig. 2J). Calcein/PI staining revealed the killing effect of activated CD8+ T on GC cells with fluorescence microscope, which showed that si-ILF3 enhanced the killing effect of activated CD8+ T on GC cells, and oe-ILF3 played the contrary effect (Fig. 2M, N). The crystal violet staining analysis revealed that the number of viable GC cells decreased in the si-ILF3 group and the number of oe-ILF3 increased when it was co-cultured with activated CD8+ T cells (Fig. 2K, L). The TIMER analysis found the elevated PD-L1 expression in multiple malignancies. Among them, the expression level of PD-L1 in GC exceeded that in normal gastric mucosal tissues (Figure S3A). Further investigation showed elevated PD-L1 was associated with poorer OS (Figure S3B). The TCGA analysis found that the mRNA level of PD-L1 was higher in GC tissues compared to normal tissues (Figure S3C). Analysis of the clinical samples from our own hospital also found PD-L1 expression was elevated in GC tissues compared to paracancerous tissues at both protein and mRNA levels (Figure S3D,3E). IHC showed the expressions of PD-L1 and CD8 in GC and paracanceerous tissues (Figure S3F).
Simvastatin inhibited PD-L1 via ILF3 and induced CD8 + T cell-dependent ferroptosis in GC cells.
Simvastatin, a lipophilic statin with enhanced lipid solubility, facilitates its permeation across cell membranes, thereby increasing the likelihood of exerting anticancer properties[31]. GC cells were divided into two groups cultured alone and with activated CD8+ T cells, and then treated with different concentrations of simvastatin, respectively. The CCK8 assay demonstrated a significant decrease in the activity of GC cells in two groups with the increased concentration of simvastatin, and the IC50 values were 35µM and 15µM, respectively (Fig. 3A). The IC50 values of MTT assays were 31 µM and 18 µM, respectively (Fig. 3B). The western blot and qRT-PCR analyses revealed a concentration-dependent suppression of ILF3 expression by simvastatin at the mRNA and protein levels (Fig. 3C). Calcein/PI staining showed the number of dead GC cells was significantly increased in si-ILF3 and simvastatin-stimulated groups compared to the nc-ILF3 group while co-culturing with activated CD8+ T cells (Fig. 3D, Figure S1A, B). To investigate the potential induction of ferroptosis in GC cells by simvastatin, simvastatin stimulation of GC cells was accompanied with ferrostatin-1 (Ferr-1; a ferroptosis inhibitor) treatment, as well as Z-VAD-FMK (Z-VAD; an apoptosis inhibitor), and necrostatin-1s (Nec-1s; a necrosis inhibitor). The CCK8 analysis showed a significant increase in cell activity only when the GC cells were co-stimulated with the ferroptosis inhibitor Ferr-1 (Fig. 3E). The hTFtarget database analysis showed that the critical ferroptosis-associated molecules (SLC7A11 and GPX4) were potential targets of ILF3. Moreover, a positive correlation was found between the protein expression levels of ILF3 and SLC7A11, as well as GPX4 (Figure S2C), and the protein expressions of SLC7A11 and GPX4 were decreased after si-ILF3 and simvastatin stimulation of GC cells when co-cultured with activated CD8+ T cells (Fig. 3F). A crucial characteristic of ferroptosis involves the diminishment or elimination of mitochondrial cristae, the rupture of the outer mitochondrial membrane, and its subsequent crumpling[32]. The different subgroups of GC cells were found under the electron microscope showed that mitochondria exhibited ferroptosis after the si-ILF3 and simvastatin stimulation of GC cells when co-cultured with activated CD8+ T cells (Fig. 3H, I). Flow cytometry analysis showed the increase level of lipid peroxidation (LPO) in GC cells after the si-ILF3 and simvastatin stimulation of GC cells when co-cultured with activated CD8 + T cells (Fig. 3J, K). The levels of ROS and Fe2+ were elevated in the si-ILF3 and simvastatin stimulation of GC cells when co-cultured with activated CD8 + T cells with fluorescence microscope (Fig. 3L-O).
To verify the induction of ferroptosis in GC cells by simvastatin through the inhibition of ILF3 when co-cultured with activated CD8+ T cells, the rescue experiment was conducted. The GC cells were divided into three groups: con-ILF3, oe-ILF3, and oe-ILF3 + simvastatin. The protein expressions of ILF3, SLC7A11, and GPX4 were reduced in the oe-ILF3 + simvastatin compared to oe-ILF3 group (Figure S2B). Calcein/PI staining showed the number of dead GC cells was increased in oe-ILF3 + simvastatin group compared to the oe-ILF3 group while co-culturing with activated CD8+ T cells (Figure S2C). The levels of ROS and Fe2+ were elevated in oe-ILF3 + simvastatin group compared to the oe-ILF3 group while co-culturing with activated CD8+ T cells with fluorescence microscope (Figure S2D, E). An increased MDA and decreased GSH levels were found in GC cells in oe-ILF3 + simvastatin group compared to the oe-ILF3 group while co-culturing with activated CD8+ T cells (Figure S2F, G). Flow cytometry analysis showed the increase level of LPO in oe-ILF3 + simvastatin group compared to the oe-ILF3 group while co-culturing with activated CD8+ T cells (Figure S2H).
Simvastatin inhibition of ILF3 regulated PD-L1 expression and enhanced the induction of ferroptosis in GC cells by activated CD8 + T cells
After stimulation of GC cells with different concentrations of simvastatin, the expression of PD-L1 at the mRNA and protein levels was gradually decreased (Fig. 4A). PD-L1 overexpression at the protein and mRNA levels was confirmed with plasmid transfection (Fig. 4B).The decreased protein expressions of PD-L1, SLC7A11, and GPX4 were found in the si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4C). The increased protein expression of ILF3, PD-L1, SLC7A11, and GPX4 were found in the simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4D). CFDA-SE/PI staining showed the number of dead GC cells was decreased in the simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4E, F). Flow cytometry analysis showed that simvastatin + oe-PD-L1 group had a decreased rate of apoptosis in GC cells compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4G). The crystal violet staining revealed that the number of viable GC cells was increased in simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4H). Previous studies demonstrated that inhibition of PD-L1 promoted CD8+ T cells to induce ferroptosis in tumor cells, providing a new direction for immunotherapy of malignant tumors[33, 34]. A decreased MDA and increased GSH levels were found in GC cells in simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4I, J). The levels of ROS and Fe2+ were decreased in simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4K-N). Flow cytometry analysis showed the decreased level of LPO in simvastatin + oe-PD-L1 group compared to si-ILF3 and simvastatin stimulation groups of GC cells when co-cultured with activated CD8+ T cells (Fig. 4O, P).
Simvastatin inhibited ILF3 expression by inhibiting H3K14ac, and ILF3 regulated PD-L1 expression through the DEPTOR/mTOR signaling pathway.
Swiss Target Prediction was utilized to predict potential targets of simvastatin by analysis of drug structure, which found that HDAC1, HDAC2, and HDAC6 had the potential to be targeted by simvastatin drugs. It means that simvastatin might inhibit ILF3 expression by inducing overexpression of the HDAC family and subsequently reducing ILF3 histone acetylation (Fig. 5A). The increased protein expression levels of HDAC1, HDAC2, and HDAC6 and the decreased protein expression of ILF3 were found under the stimulation of different concentration of simvastatin (Fig. 5B). Trichostatin (TSA) is a repressor of the HDAC family, leading to increased histone acetylation and promoting gene transcription[35]. Our study found the decreased protein expression levels of HDAC1, HDAC2, and HDAC6 and the increased protein expression of ILF3 under the co-stimulation of different concentration of simvastatin and TSA (Fig. 5C). Meanwhile, the total acetylation level in GC cells decreased with the increase of simvastatin concentration, and TSA could reduce the total acetylation level (Fig. 5D). Also simvastatin stimulation elevated the protein expressions of HDAC1, HDAC2, and HDAC6, but si-ILF3 did not impact HDAC1, HDAC2, and HDAC6 protein expression levels (Fig. 5E). Only the knockdown of HDAC6 resulted in an increase in ILF3 protein expression (Fig. 5F). The Cut & Tag Experiment showed an elevation in the acetylation at the H3K14 site in GC cells stimulated by simvastatin (Fig. 5G). A gradual decrease in acetylation at the H3K14 site was found in GC cells treated with varying concentrations of simvastatin (Fig. 5H). Additionally, simvastatin stimulation (20µM) in combination with different concentrations of TSA stimulation demonstrated a progressive increase in acetylation at the H3K14 site(Fig. 5I). After knocking down HDAC1, HDAC2, and HDAC6, respectively, the acetylation of the H3K14 site was increased when HDAC6 was knocked down only (Fig. 5J). KEGG enrichment analysis was performed on the whole genome sequencing results of previous nc-ILF3 and si-ILF3, which showed that ILF3 regulated the mTOR signaling pathway (Fig. 5K). The differential genes enriched in the mTOR signaling pathway were WNT8B, CAB39, DEPTOR, PRKCB, and RNF152etc. The qRT-PCR results showed the most significant differences in DEPTOR expression between nc-ILF3 and si-ILF3 groups (Fig. 5L). DEPTOR protein expression was elevated, and the p-mTOR expression level was decreased, after the knockdown of ILF3, and the results were reversed after overexpression of ILF3 (Fig. 5M). Previous studies demonstrated that DEPTOR was a suppressor molecule of mTOR, which was further validated in GC cells[36]. Western blot showed an upregulation of p-mTOR expression when knockdown the expression of DFPTOR (Fig. 5N), and a slightly elevated levels of p-mTOR and PD-L1 (Fig. 5O), and SLC7A11 and GPX4 (Fig. 5P) were found in the group with both ILF3 and DEPTOR knockdown compared to the group with only ILF3 knockdown. The protein expressions of ILF3, p- mTOR, PD-L1, SLC7A11, and GPX4 were reduced in the oe-ILF3 + LY294002, which was the inhibitor of the mTOR signaling pathway, compared to the oe-ILF3 group (Fig. 5Q).
DEPTOR and mTOR expression levels correlate with the prognosis of GC patients.
The Gene Set Cancer Analysis (GSCA) database analysis found that the mRNA level of ILF3 was lower in GC tissues compared to normal tissues, and the mRNA level of mTOR was higher in GC tissues compared to normal tissues (Figure S4A, B). IHC results showed that DEPTOR was highly expressed in normal gastric tissues, and when ILF3 was lowly expressed, DEPTOR showed high expression levels. Meanwhile, mTOR expression was elevated in GC tissues, and when ILF3 was highly expressed, mTOR also showed high expression levels (Figure S4C). A negative regulatory relationship was found between ILF3 expression and DEPTOR, as well as a positive regulatory association of ILF3 expression with mTOR (Figure S4D, E). The Kaplan-Meier plotter showed a poor prognosis when low expression of DEPTOR and high expression of mTOR in GC patients (Figure S4F, G).
Simvastatin induced ferroptosis in GC cells by inhibiting PD-L1 expression by ILF3 to achieve therapeutic effects on GC
Mouse Forestomach Carcinoma (MFC) stably expressing ILF3 were constructed by lentivirus vector transfection (Fig. 6A). The C57BL/6 mice were randomly assigned to six distinct groups: nc-ILF3, sh-ILF3, nc-ILF3 + simvastatin, con-ILF3, oe-ILF3, and oe-ILF3 + simvastatin, which revealed that the suppression of ILF3 and the administration of simvastatin effectively impeded the proliferation of subcutaneous tumors (Fig. 6B, C). Growth curves demonstrated that overexpression of ILF3 facilitated the proliferation of subcutaneous tumors and simvastatin inhibited the proliferation of subcutaneous tumors (Fig. 6D, F). The weight of subcutaneous tumors were consistent with growth curve (Fig. 6E, G). PD-L1 protein expression was positively correlated with ILF3 expression, while simvastatin inhibited ILF3 and PD-L1 expressions (Fig. 6H). IHC analysis revealed a decrease in 4-HNE expression alongside elevated levels of ILF3 and PD-L1, and elevated 4-HNE expression in simvastatin-treated groups with decreased ILF3 and PD-L1 expression (Fig. 6I-K).
Simvastatin inhibited ILF3 and enhanced anti-PD-L1 therapy efficacy by recruiting CD8+ T cells in GC
This study investigated the expression levels of ILF3 in GC tissues obtained from radical gastric cancer surgery. Subsequently, GC tissue samples with high and low expressions of ILF3 were separately implanted into NOD/SCID mice to establish a GC patient xenograft (PDX) model (Fig. 7A). The high ILF3-expression group was administered with simvastatin treatment and simvastatin combined with anti-PD-L1 therapy, followed by weekly measurements of tumor size. After 5 weeks, the tumors were extracted to generate tumor growth curves and histograms depicting tumor weight for each group (Fig. 7B-D). Immunohistochemistry was conducted on paraffin sections of tumors from four distinct groups to assess the expression levels of ILF3, DEPTOR, mTOR, PD-L1, CD8, and 4-HNE, which revealed that the regulation of PD-L1 expression occurred via the DEPTOR/mTOR signaling pathway, facilitated by ILF3. Additionally, simvastatin demonstrated the potential to enhance the therapeutic efficacy of anti-PD-L1 treatment by suppressing PD-L1 expression through ILF3 and promoting the recruitment of CD8+ T cells to induce ferroptosis in GC cells (Fig. 7E).