Abstract
Background/Aim: Advanced renal cell carcinoma (RCC) is commonly treated with immune checkpoint inhibitor (ICI)-based therapies. However, patient responses vary, and predictive biomarkers remain limited. Here, we demonstrate that Schlafen 11 (SLFN11), which has emerged as an important gene in immunity and drug response, acts as a key enhancer of antitumor immunity in advanced RCC.
Materials and Methods: We analyzed transcriptome data from in-house and public RCC patient cohorts to evaluate associations between SLFN11 expression, clinical outcomes, and immune contexture. Transcriptome profiling of RCC cell lines was performed to identify SLFN11-regulated cytokines. Functional assays included co-culture of wild-type or SLFN11-knockout RCC cells with M0 macrophages, immature macrophages, followed by cytokine quantification and phenotypic analyses. Multiplex immunofluorescence staining was applied to formalin-fixed paraffin-embedded samples from ICI-treated patients to validate immune cell infiltration in the tumor microenvironment.
Results: Clinically, patients with SLFN11-high RCC treated with ICI-based regimen showed significantly prolonged progression-free survival in both our cohort and public datasets. SLFN11-high RCC tumors exhibited increased infiltration of anti-tumoral M1 macrophages and up-regulation of immune-related pathways compared to SLFN11-low tumors. Mechanistically, transcriptome as well as proteome analysis revealed that, in RCC cells, SLFN11 significantly up-regulated the expression of colony stimulating factor 2 (CSF2) encoding granulocyte-macrophage colony-stimulating factor (GM-CSF), which promotes polarization from M0 macrophages to M1 macrophages. Consequently, SLFN11-positive RCC cells co-cultured with M0 macrophages secreted significantly higher levels of GM-CSF than SLFN11-negative cells and promoted M1 macrophage differentiation. Consistently, multiplex immunofluorescence staining of RCC patient sample analysis confirmed that the number of M1 macrophages was significantly higher in the tumor microenvironment of SLFN11-high tumor than in SLFN11-low tumor.
Conclusion: SLFN11 enhances antitumor immunity in RCC by increasing GM-CSF secretion and activating M1 macrophages, potentially improving ICI efficacy.
Introduction
Renal cell carcinoma (RCC) is a genitourinary malignancy diagnosed in nearly 400,000 people worldwide each year, leading to over 170,000 deaths annually (1). At the time of diagnosis, 25-30% of RCC cases already present with metastases, requiring subsequent therapeutic options. Given that distinct hyper-vascularity and immune cell infiltration, inhibitors of the vascular endothelial growth factor (VEGF) pathway and the programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) axis as monotherapy or in combination with immune checkpoint inhibitors (ICIs) (2) have resulted in significant improvement of clinical outcomes in patients with advanced RCC. However, not all patients respond (3-8) and some of them experience life-threatening immune-related adverse events (9). Thus, a better understanding of the molecular factors underlying clinical heterogeneity in patients with advanced RCC is essential to guide treatment selection strategy and clarify mechanisms of resistance.
Clear cell RCC (ccRCC), the predominant cell type in RCC, is characterized as a highly inflamed tumor type, with one of the highest immune cell infiltration scores in pan-cancer analysis and high expression of immune checkpoints, such as PD-L1 and CTLA-4 (10, 11). ICI treatments, particularly anti-PD-1 and anti-PD-L1 antibodies, act in part by reinvigorating a pre-existing tumor immune response. So far, multiple biomarkers including PD-L1 expression, tumor mutation burden, and specific mutational profiles in cancer tissues have been proposed to have a significant impact on the therapeutic response to ICIs (12, 13). Another potential predictor of the response to ICIs is the density of tumor-infiltrating lymphocytes (TILs) within a tumor tissue. Although CD8+ T cells, representative TILs, play a fundamental role in attacking cancer cells in the tumor microenvironment (TME) (14), the presence of CD8+ T cells sometimes produces contradictory outcomes depending on the content of ICI combination therapies in RCC (15, 16). These findings imply that immune cells beyond T cells including innate immune cells in the TME influence T cell function and affect the efficacy of ICI treatments.
The Schlafen 11 (SLFN11) gene has recently gained attention in cancer therapy for two major reasons. First, SLFN11 is emerging as a potent predictive biomarker for DNA-damaging agents, including platinum derivatives (cisplatin, carboplatin), topoisomerase inhibitors (etoposide, irinotecan), and replication inhibitors (cytarabine, gemcitabine), as it enhances cellular sensitivity to these anticancer agents (17-22). In cancers where SLFN11 is expressed in a certain proportion of patients (23, 24), SLFN11 has been validated as a potent predictive biomarker for chemotherapy, including in breast (18), ovary (25, 26), stomach (27), bladder (28), lung (29), esophagus (30), head and neck (31), prostate cancers (32), a subset of medulloblastoma (33), and acute myeloid leukemia (21). SLFN11 also enhances sensitivity to PARP inhibitors (e.g., olaparib, niraparib, talazoparib) (34-38). We and other groups have elucidated multiple mechanisms underlying SLFN11-dependent drug sensitivity, including its tRNA cleavage-, helicase domain-, and single-stranded DNA binding-dependent activities (37, 39-49). Second, SLFN11 has been reported to play roles in both innate and adaptive immune responses. It was initially identified as an interferon-stimulating gene with potent antiviral activity (50, 51) and has since been implicated in immune regulation in cancer (25, 52). SLFN11 activates an innate immune response by recognizing single-strand DNA outside the genome (53) or through the transcriptional activation of immediate early genes (e.g., JUN, FOS, NF-kB2) under the treatment with DNA-damaging agents (54). In the analysis using patient samples, SLFN11 expression is associated with immune activation in high-grade serous ovarian cancers (25) and acts as a predictive biomarker for the response to ICIs in hepatocellular carcinoma (55).
Our immunohistochemical analysis, in conjunction with other studies on RCC, has demonstrated that SLFN11 is highly expressed in more than half of ccRCC cases (23, 24). Nonetheless, the functional and clinical significance of SLFN11 in RCC has never been systematically investigated.
In this study, we investigate the impact of SLFN11 on clinical outcomes in RCC under different treatment regimens. We demonstrate that high SLFN11 expression in RCC is associated with a superior response to ICI and increased infiltration of M1 macrophages, as determined through analyses of both our cohort and public databases. We further elucidate the mechanism by which tumor-intrinsic SLFN11 promotes macrophage polarization towards the M1 phenotype. The biological and clinical insights gained from this study suggest a new strategy for using SLFN11 as a predictive biomarker for ICI therapy and for enhancing ICI efficacy through macrophage polarization. To our knowledge, this is the first study to clarify the biological role and clinical impact of SLFN11 in RCC.
Materials and Methods
Patient selection. A total of 121 patients diagnosed with ccRCC who underwent either radical or partial nephrectomy at our institution were included as the in-house cohort. Tumor tissue samples from these cases were used for RNA sequencing (RNA-seq) analysis. In addition, 28 patients with advanced RCC who had available formalin-fixed, paraffin-embedded (FFPE) tumor slides and received nivolumab as second-line systemic therapy were included as the in-house cohort for ICI-treated cases. Written informed consent was obtained from all patients, and the Ethics Review Board of the Osaka University Medical Hospital approved the study protocol (no. 13397-2, no. 14069-3).
Immunohistochemistry (IHC). Our step-by-step IHC protocol for SLFN11 is previously described (23). Briefly, IHC was performed using a Dako Envision+ Mouse Peroxidase Detection System (Dako Cytomation, Carpinteria, CA, USA). The sections were incubated with the first antibody for 1 h at room temperature, followed by incubation with Envision+ anti-mouse or rabbit peroxidase for 1 h. SLFN11 IHC scores were counted by a pathologist without knowledge of clinical and pathological parameters or patient outcome. According to the average values of ratio of positivity, SLFN11 IHC scores were considered 1+ (1-10%), 2+ (11-50%), or 3+ (51-100%). When IHC score of SLFN11 expression was ≧2+, immunostaining was considered to indicate high SLFN11 expression.
Public database analysis. The RNA-seq normalized data of the CheckMate 025 trial and the JAVELIN Renal 101 trial were downloaded from supplementary tables for the respective articles (15, 56). The infiltration fractions of 22 immune cell types in each sample were calculated using the CIBERSORTx algorithm 2 (57). The differential cell infiltration between The differential immune cell infiltration between the SLFN11–high (top 10%) and SLFN11–low (bottom 10%) expression groups was compared using the Mann–Whitney U test.
Cell lines, culture conditions and generation of SLFN11-knockout cells. The human renal cell carcinoma cell lines 786-O and TUHR10TKB were obtained from RIKEN BRC (Tsukuba, Ibaraki, Japan), while the human monocytic cell line THP-1 was purchased from the JCRB Cell Bank (Tokyo, Japan). All cell lines were maintained in RPMI 1640 medium supplemented with 10% fetal bovine serum. The cells were cultured in a humidified incubator at 37°C with 5% CO2. SLFN11-knockout (KO) cells in the RCC cell lines were generated using CRISPR/Cas9 methods as described previously (58).
Co-culture assays. To establish a co-culture system for macrophages and RCC cells in vitro, a 24 mm Transwell chamber with a 0.4 μm pore polycarbonate membrane (Corning Inc., Corning, NY, USA) was used. THP-1 monocytes (1×106/well) were first induced to differentiate into M0 macrophages with 150 nM phorbol 12-myristate 13-acetate (PMA) (Sigma-Aldrich, St. Louis, MO, USA) in the upper chamber, and RCC cells (2×105/well) were seeded in the lower chamber before co-culture. After 24 h, the upper chamber was placed onto the lower chambers for co-cultivation, and the macrophage cells on the upper chamber were collected after 72 h for analysis.
Olink® proteomics analysis. The Olink® Target 48 Cytokine panel (Olink Proteomics AB, Uppsala, Sweden) was used to quantify cytokine levels in cell-free supernatants collected from co-culture experiments of RCC cell lines (parental or SLFN11-KO) and M0 macrophages after 48 h of incubation. Supernatants were clarified by centrifugation and stored at −80°C until analysis. Protein concentrations were measured using proximity extension assay (PEA) technology according to the manufacturer’s instructions (Olink Proteomics AB). The assay was conducted by a certified service provider (FUJIFILM Wako Pure Chemical Corporation, Osaka, Japan). Results are presented as normalized protein expression (NPX) values, which represent relative log2-scale expression levels normalized using internal and external controls. Protein expression ratios (RCC parent vs. SLFN11-KO), calculated from NPX values, were obtained for each cytokine. The ratios derived from 786-O co-cultures were plotted on the X-axis, and the corresponding ratios from TUHR10TKB co-cultures were plotted on the Y-axis.
Flow cytometry. Macrophages derived from THP-1 cells and adherent on the upper chamber of the Transwell insert were harvested by trypsinization. The collected cells were resuspended and prepared in Stain Buffer (554656; BD Biosciences, San Jose, CA, USA). Non-specific Fc receptor binding was blocked using Human Fc Block (564219; BD Biosciences). Cell suspensions were aliquoted at 100 μl per tube and incubated with fluorochrome-conjugated antibodies in the dark at 4°C for 30 min. Flow cytometric analysis was performed using a BD FACSCanto II flow cytometer (BD Biosciences), and the data were analyzed with FlowJo software (FlowJo, LLC, Ashland, OR, USA). The following antibodies and isotype controls were used: FITC-conjugated anti-human CD80 antibody (305206; BioLegend, San Diego, CA, USA) and FITC Mouse IgG1, κ Isotype Control (400107; BioLegend), PE-conjugated anti-human CD163 antibody (PGI-PE-65169-100; Proteintech Group, Inc., Rosemont, IL, USA) and Mouse IgG1 κ Isotype Control (P3.6.2.8.1) (ab185801; Abcam, Cambridge, UK).
Quantification of granulocyte–macrophage colony-stimulating factor (GM-CSF) in cell culture supernatant. The concentration of GM-CSF in cell culture supernatants was measured using the Quantikine ELISA Kit (DGM00; R&D Systems, Minneapolis, MN, USA), following the manufacturer’s instructions. Absorbance was measured at 450 nm with a correction at 570 nm using Bio-Rad iMark Microplate Absorbance Reader (168-1130; Bio-Rad, Hercules, CA, USA). All measurements were performed in triplicate, and the final values represent the mean of these replicates. The duration between cell seeding and collection of culture supernatants was kept consistent across all experimental conditions.
Multiplex immunofluorescence staining. Multiplex immunofluorescence staining was performed on FFPE tumor tissue sections obtained from in-house ccRCC cases (n=28) treated with nivolumab. The staining was conducted using the Opal Multiplex IHC Kit (Akoya Biosciences, Marlborough, MA, USA) according to the manufacturer’s protocol. Primary antibodies were used to stain the following markers in the specified order: CSF2 (17762-1-AP, 1:400; Proteintech), CD80 (ab134120, 1:500; Abcam), CD8 (ab17147, 1:100; Abcam), SLFN11 (sc-515071, 1:100; Santa Cruz, Dallas, TX, USA), CD68 (ab844, 1:200; Abcam), and CA9 (MAB9551, 1:100; Abnova, Taipei, Taiwan, ROC). All primary antibodies were incubated at 4°C overnight. For each round of staining, HRP-conjugated secondary antibodies and fluorescent dyes (CSF2: Opal 650, CD80: Opal 540, CD8: Opal 620, SLFN11: Opal 690, CD68: Opal 520, CA9: Opal 570) were applied. Heat-induced antigen retrieval using AR6 buffer (provided in the kit) was performed between each staining round.
The sections were counterstained with 4′,6-diamidino-2-phenylindole dihydrochloride (DAPI, Thermo Fisher, Waltham, MA, USA) and mounted using Fluoro-KEEPER Antifade Reagent Non-Hardening Type (Nacalai Tesque, Kyoto, Japan). Imaging was performed using a confocal laser scanning microscope (LSM980 with Airyscan2 module; Carl Zeiss, Jena, Germany) in OFP mode. A 20x objective lens (Plan-Apochromat 20x/0.8 M27) was used, and excitation was achieved with lasers of 405 nm, 488 nm, 561 nm, and 639 nm wavelengths. Detector settings, gain, and pinhole size were optimized according to the manufacturer’s recommendations.
Images were analyzed using Zeiss ZEN software (Carl Zeiss). Cells were classified as follows: RCC cells (CA9+), SLFN11-positive RCC cells (CA9+SLFN11+), CSF2-positive RCC cells (CA9+CSF2+), CSF2-positive and SLFN11-positive RCC cells (CA9+CSF2+SLFN11+), macrophages (CD68+), and M1 macrophages (CD68+CD80+).
Statistical analysis. Clinical characteristics were compared using the Mann-Whitney U test. Progression-free survival (PFS) was estimated using the Kaplan-Meier method and Cox proportional hazard regression. All p-values were two-sided, and differences were considered statistically significant at p<0.05. The level of significance was depicted as *p<0.05, **p<0.01, ***p<0.001, and not statistically significant (n.s.). Data were analyzed using JMP Pro (v.17.0.0; SAS Institute, Cary, NC, USA).
Results
High SLFN11 expression is associated with favorable response to immune checkpoint inhibitors in RCC. To evaluate whether SLFN11 is expressed in RCC tissues, we first performed IHC staining on 81 tumor specimens obtained from patients who underwent nephrectomy at our institution. Representative IHC images are shown in Figure 1A, and the distribution of SLFN11 positivity is presented in Figure S1A. Consistent with multiple reports (23), SLFN11 staining was exclusively observed in the nuclei of RCC tumor cells. The proportion of patients with IHC scores of 0, 1+, 2+, and 3+ was 37.0%, 13.5%, 36.5%, and 21.2%, respectively, indicating that SLFN11 is frequently expressed in RCC tissues (Figure 1A).
SLFN11 is highly expressed in renal cell carcinoma tissue, and its high expression is associated with response to immune checkpoint inhibitors. (A) Immunohistochemical (IHC) staining of renal cell carcinoma (RCC) tissue sections for SLFN11, along with representative scoring and stratification into high and low expression groups. Scale bar, 100 μm. (B) Overall survival of RCC patients from our institutional cohort, stratified by SLFN11 expression. (C) Progression-free survival (PFS) of 24 patients with advanced RCC treated with second-line sunitinib, stratified by SLFN11 expression. (D) PFS of nivolumab-treated RCC patients from our cohort, stratified by SLFN11 expression. (E) SLFN11 localization in normal kidney (top) and RCC tumor tissue (bottom). The wavy line indicates the tumor–stroma boundary. Scale bar, 100 μm. (F) Paired comparison of SLFN11 mRNA expression between tumor and adjacent normal tissue in 121 RCC cases from our institutional RNA-seq dataset. Data are shown as paired values; statistical significance was assessed using a paired t-test. (G) PFS with everolimus treatment in the CheckMate 025 cohort, stratified by SLFN11 mRNA expression (high vs. low defined by median value). (H) PFS with nivolumab treatment in the CheckMate 025 cohort, stratified by SLFN11 mRNA expression. Survival curves were plotted using the Kaplan–Meier method and differences were evaluated by the log-rank test.
We stratified the patients in our institutional RCC cohort into a SLFN11-low group (IHC score 0 or 1+) and a SLFN11-high group (IHC score 2+ or 3+) and examined the association between SLFN11 expression and clinical prognosis. There was no significant difference in overall survival (OS) between the SLFN11-high and SLFN11-low groups [median OS: 76.0 months vs. 78.0 months, p=0.612, hazard ratio (HR)=1.25, 95% confidence interval (CI)=0.524-3.00; Figure 1B]. Among patients who received sunitinib (a tyrosine kinase inhibitor), PFS was also not significantly different between the two groups based on SLFN11 expression (median PFS: 11.0 months vs. 23.0 months, p=0.860, HR=1.11, 95% CI=0.342-3.62; Figure 1C). However, the PFS in response to nivolumab, a PD-1 ICI, was significantly longer in the SLFN11-high group compared to the SLFN11-low group (median PFS: 16.5 months vs. 3.00 months, p=0.0250, HR=0.257, 95% CI=0.0785-0.843; Figure 1D and Table S1).
Next, we extended our analysis of the association between SLFN11 expression and therapeutic response to ICIs to transcriptome databases. Although tumor-infiltrating cells such as stromal or lymphoid cells can express SLFN11 and potentially confound measurement in some cancers, these cells were almost entirely negative for SLFN11 in our RCCs and the adjacent normal tissue samples (Figure 1E). Moreover, transcriptome analysis of our cohort revealed that SLFN11 expression was significantly higher in tumor tissues compared to adjacent non-tumor tissues (Figure 1F), suggesting that SLFN11 expression is up-regulated during RCC tumorigenesis and that its high expression originates from tumor cells rather than from infiltrating immune or stromal cells.
Building on this understanding, we analyzed publicly available transcriptome data from the CheckMate 025 clinical trial, which evaluated the efficacy of nivolumab in patients with RCC. In the everolimus (mTOR inhibitor)-treated arm of this trial, SLFN11 expression was not associated with a significant difference in PFS (median PFS: 4.01 months vs. 4.19 months, p=0.436, HR=1.16, 95% CI=0.801-1.67; Figure 1G). However, in the nivolumab-treated arm, PFS was significantly prolonged in the SLFN11-high group compared to the low group (median PFS: 5.32 months vs. 3.71 months, p=0.0470, HR=0.725, 95% CI=0.528-0.996; Figure 1H). These results suggest that high SLFN11 expression in RCC may serve as a predictive marker for favorable response to ICI-based therapy.
High SLFN11 expression is associated with antitumor immune responses and predominantly involves M1 macrophages in the tumor microenvironment. Since a favorable response to ICI-based therapy is generally associated with an active immune response, we analyzed the transcriptome data cross three independent RNA-seq cohorts: our in-house RCC cohort, the CheckMate 025 cohort, and the publicly available JAVELIN Renal 101 cohort. Gene Ontology (GO) enrichment analysis revealed that immune-related pathways including “positive regulation of immune response” (GO:0050778), “positive regulation of cytokine production” (GO:0001819), and “adaptive immune response” (GO:0002250) were consistently enriched in SLFN11-high tumors across all three cohorts (Figure 2A-C). These results indicate that the TME in SLFN11-high RCC exhibits greater immune activation in than SLFN11-low RCC.
High SLFN11 expression is associated with anti-tumor immune responses predominantly involving M1 macrophages. (A-C) Gene Ontology (GO) enrichment analysis of differentially expressed genes between SLFN11-high and SLFN11-low tumors in three independent cohorts: the in-house renal cell carcinoma (RCC) cohort (A), the CheckMate 025 cohort (B), and the JAVELIN Renal 101 cohort (C). (D) Tumor-infiltrating immune cells analyzed by CIBERSORTx in in-house cases were compared between SLFN11-high and SLFN11-low groups. Statistical analyses were performed by a Mann–Whitney U-test. Bars represent the mean±SD. (E) Comparison of tumor-infiltrating immune cell subsets between SLFN11-high and SLFN11-low groups across three RCC cohorts. Colored circles indicate subsets showing significant differences (red: enriched in SLFN11-high; blue: enriched in SLFN11-low), with circle size proportional to −log10(p-value); white circles indicate no significant difference.
To identify which immune cell types contribute to immune activation, we estimated the immune cell composition using a deconvolution algorithm. We compared the abundance of immune cell subsets between the top and bottom 10% of tumors ranked by SLFN11 expression within each cohort (Figure 2D, Figure S2A, B) with a summary shown Figure 2E. While several immune cell fractions showed statistically significant differences, the most consistent finding across all cohorts was that SLFN11-high tumors harbored a significantly higher proportion of M1 macrophages compared to SLFN11-low tumors (p=0.0061, p=0.0100, p<0.0001, respectively). M1 macrophages are classically polarized from M0 macrophages by stimuli such as interferon-gamma, lipopolysaccharide, or GM-CSF, and enhance anti-tumor immunity partly by stimulating cytotoxic T cell activity (59). Notably, however, there was no significant difference in CD8+ T cell infiltration between high and low SLFN11 expression groups in all cohorts (p=0.403, p=0.899, p=0.153, respectively). These findings suggest that high SLFN11 expression in RCC may promote M1 macrophage polarization, which in turn activates antitumor immune response within the TME.
Loss of SLFN11 in RCC reduces CSF2 mRNA expression. To investigate whether tumor-intrinsic SLFN11 expression affects the polarization of M1 macrophages, we conducted several in vitro experiments. We first generated SLFN11-KO cells from two human RCC cell lines with high endogenous SLFN11 expression (786-O and TUHR10TKB) and confirmed SLFN11 deletion by immunoblotting (Figure 3A). To functionally validate the knockout of SLFN11, we assessed the viability of these cells in response to SLFN11-dependent drugs. As we expected, the absence of SLFN11 reduced sensitivity to PARP inhibitors (niraparib, olaparib) as well as to a topoisomerase I inhibitor (irinotecan), consistent with previous reports, confirming that the SLFN11-KO cells display the expected functional phenotype (Figure S3A, B) (37, 58).
SLFN11 knockout in renal cell carcinoma cells leads to down-regulation of CSF2. (A) Generation of SLFN11-knockout (KO) renal cell carcinoma (RCC) cell lines (786-O and TUHR10TKB). SLFN11 protein expression was confirmed using Western blotting. (B) Venn diagram showing differentially down-regulated genes upon SLFN11-KO in 786-O and TUHR10TKB RCC cell lines. The cutoff values were |Fold Change| >2 and p-value <0.05. (C) List of 26 genes commonly down-regulated by SLFN11-KO between 786-O and TUHR10TKB cell lines. (D) Gene Ontology analysis of differentially expressed genes commonly down-regulated by SLFN11-KO between 786-O and TUHR10TKB cell lines. (E) Volcano plots illustrating differentially expressed genes between parental cells and SLFN11-KO cells in 786-O (left) and TUHR10TKB (right). The cutoff values were |Fold Change| >2 and p-value <0.05. (F) RT-PCR analysis comparing CSF2 mRNA expression between parental and SLFN11-KO cells in 786-O and TUHR10TKB cell lines. Statistical analyses were performed using a Mann–Whitney U-test.
Next, we performed transcriptome analysis comparing 786-O parental and SLFN11-KO cells, and TUHR10TKB parental and SLFN11-KO cells. In 786-O and TUHR10TKB cell sets, 265 and 742 genes were significantly down-regulated by the deletion of SLFN11, respectively, with 26 genes commonly down-regulated between the two cell sets (Figure 3B and C). GO enrichment analysis of these 26 genes revealed strong enrichment of immune-related pathways, such as “cellular response to lipopolysaccharide” (GO:0071222) and “positive regulation of leukocyte proliferation” (GO:0070665), which are known to be closely associated with macrophage activation (Figure 3D).
Among 26 genes, CSF2, the gene encoding the macrophage-activating cytokine GM-CSF and promoting the transition of M0 macrophages towards either M1 or M2 macrophages, was one of the most down-regulated genes in both SLFN11-KO cells (Figure 3E). The reduction of CSF2 mRNA by SLFN11-KO was separately confirmed using real-time RT-PCR in both cells (Figure 3F). These results suggest that SLFN11-dependent CSF2 up-regulation in RCC contributes to the M0 macrophage activation.
Tumor-intrinsic SLFN11 promotes M1 macrophages polarization through increasing GM-CSF production. To examine whether SLFN11 in RCC cells influences M0 macrophage activation and polarization, we performed co-culture experiments using RCC cell lines (786-O and TUHR10TKB) and M0 macrophages. M0 macrophages were derived from the human monocytic cell line THP-1 by stimulating with phorbol myristate acetate (PMA) (Figure S4A). We used trans-well culture systems to co-culture M0 macrophages with or without either SLFN11-proficient parental cell lines or their corresponding SLFN11-KO counterparts (Figure 4A). We submitted the culture supernatants to the Olink® Target 48 Cytokine panel, which can absolutely quantify 48 proteins involved in cytokine signaling and immune-inflammatory responses (Figure 4B, Table S2). The quantification of the 48 proteins revealed that GM-CSF followed by IL4, IL33, CCL11 and IL27 was the most increased protein in the co-culture medium of M0 macrophage with SLFN11-proficient parental cells compared with that from co-cultures with SLFN11-KO cells in both cell lines (Figure 4C). GM-CSF was at an undetectable level in the culture medium of M0 macrophage monocultures (Figure 4D, E). SLFN11-proficient parental RCC cells alone secreted slightly higher levels of GM-CSF than their SLFN11-KO counterparts, although the absolute amounts were very low (Figure 4D and E). However, when M0 macrophages were co-cultured with SLFN11-proficient parental RCC cells, GM-CSF concentration increased markedly. GM-CSF levels also increased when M0 macrophages were co-cultured with SLFN11-KO RCC cells, yet the concentrations were significantly lower than in co-cultures with the parental cells (Figure 4D and E). These results indicate that the RCC cells (786-O and TUHR10TKB) have the capacity to produce GM-CSF, this capacity is enhanced in the presence of M0 macrophages, and that it is further augmented by tumor-intrinsic SLFN11.
SLFN11-positive renal cell carcinoma cells promote M1 polarization of M0 macrophages compared to SLFN11-knockout cells. (A) Schematic representation of the co-culture experiment with M0 macrophages and renal cell carcinoma (RCC) cell lines. (B) Schematic representation of multiplex immune assay by Olink® Target 48 panel. (C) Protein concentration analysis of co-culture supernatants using the Olink® Target 48 panel. Relative protein levels were plotted as the ratio of concentrations measured in co-cultures with parental RCC cells versus SLFN11-knockout (KO) RCC cells. (D, E) Enzyme-linked immunosorbent assay (ELISA)-based quantification of GM-CSF protein concentrations in culture supernatants under various co-culture and monoculture conditions using 786-O (D) and TUHR10TKB (E) cell lines. “MΦ” indicates macrophage. (F-I) Flow cytometry analysis of surface expression of CD80 (F, G) and CD163 (H, I) in co-cultured macrophages using the 786-O (F, H) and TUHR10TKB (G, I) cell lines. (D-I) All data represent results from six independent experiments. Statistical analyses were performed using the Mann–Whitney U test. Bars represent the mean±SD.
To examine the resulting polarization of M0 macrophages co-cultured with RCC SLFN11-positive and -negative RCC cells, we characterized the macrophages using several specific surface markers. First, we assessed CD80 positivity, a hallmark of M1 macrophage, using flow cytometry and found that both the fractions of CD80+ cells and the mean fluorescence intensity (MFI) of CD80 were higher in co-cultures with SLFN11-proficient parental cells than with their SLFN11-KO counterparts (Figure 4F and G). In contrast, the CD163 positivity – a hallmark of M2 macrophage – was largely comparable between the macrophages co-cultured with parental or SLFN11-KO cells (Figure 4H and I).
We then analyzed M1-specific markers (NOS2, CXCL9, TNF-α, CD80, CD86) and the M2-specific markers (ARG1, CD163, CD206, IL-10, TGF-β) using RT-PCR. As expected, the addition of recombinant GM-CSF to M0 macrophage monocultures up-regulated the expression of M1 markers (CD80, CD86) but not the M2 marker CD163 (Figure S4B). M0 macrophages co-cultured with SLFN11-proficient 786-O cells exhibited significantly higher expression of M1 markers (NOS2, CXCL9, TNF-α, CD80, CD86) than those co-cultured with SLFN11-KO counterparts, whereas the expression level of several M2 markers (ARG1, IL-10) were rather increased in the macrophages co-cultured with SLFN11-KO cells compared to the parental cells (Figure S4D). Collectively, these findings indicated that SLFN11 expression in RCC cells promotes the polarization of M0 macrophages toward the M1 phenotype through enhancing the production of GM-CSF.
SLFN11-high RCC shows high-GM-CSF production and enriches M1 macrophages in the TME. To relate these findings back to our cohort, which showed that patients in the SLFN11-high group responded favorably to nivolumab (as shown in Figure 1D), we performed multiplex immunofluorescence staining on all 28 cases. The staining included SLFN11, GM-CSF, tumor marker (CA9), a general macrophage marker (CD68), an M1 macrophage marker (CD80), an anti-tumor immunity T cell marker (CD8) and DNA (DAPI) (Figure 5A, B and Figure S5).
SLFN11-positive renal cell carcinoma tissues contain more GM-CSF-positive RCC cells and show increased M1 macrophage infiltration. (A, B) Higher-magnification multiplex immunofluorescence images of representative SLFN11-high and -low cases. Nuclei are counterstained with DAPI (blue) in all panels. Scale bar=50 μm. (A) The panels visualize SLFN11 (magenta) and GM-CSF (red). (B) The panels illustrate M1 macrophages co-expressing CD68 (cyan) and CD80 (green). (C-E) Quantitative comparison of the number or rate of tumor cells and immune cells between SLFN11-high and -low expression groups. (C) Number of RCC cells co-expressing CA9 and GM-CSF. (D) Number of CD68+ macrophages. (E) Number of CD68+CD80+ M1 macrophages. (F) The proportion of M1 macrophages among total CD68+ cells. (G) Number of CD8+ T cells. Statistical analyses were performed using the Mann–Whitney U-test. Bars represent the mean±SD.
Consistent with the in vitro results, SLFN11-high tumors contained a significantly greater number of RCC cells co-expressing CA9 and GM-CSF compared to SLFN11-low tumors (p<0.0001; Figure 5A-C). The total number of CD68+ macrophages did not differ between SLFN11-high and SLFN11-low groups (p=0.814, Figure 5A, B, D). However, the number of CD68+CD80+ double-positive M1 macrophages was significantly higher in SLFN11-high group (p=0.0024; Figure 5A, B, E). Accordingly, the proportion of CD68+CD80+ double-positive M1 macrophages among total CD68+ cells was also significantly higher in the SLFN11-high group (p=0.0054; Figure 5A, B, F). In contrast, there was no significant difference in the number of CD8+ T cells between SLFN11-high and SLFN11-low groups (p=0.165; Figure 5G).
These results confirm that tumor-intrinsic SLFN11 expression is closely associated with increased GM-CSF production and preferential recruitment or activation of M1 macrophages in the RCC tumor microenvironment. Furthermore, these findings imply that tumor-intrinsic SLFN11 may enhance anti-tumor immunity through M1 macrophage polarization and confer favorable response to nivolumab.
Discussion
Given the limited clinical impact of ICI treatment in RCC (56, 60-62), evidence-based patient selection namely, precision immunotherapy, is essential to improve outcomes. Although CD8+ T cells are the primary effectors of anti-tumor immunity and their ability to recognize and attack cancer cells is vital for effective immune surveillance, their abundance alone does not consistently predict ICI efficacy (15, 56, 63), which motivated us to pursue the precise characteristics of the TME that may influence responsiveness to ICIs.
In this study, we focused on SLFN11, which, despite its well-established role in enhancing sensitivity to DNA-damaging agents, has been largely unexplored in RCC. Our work is the first to demonstrate that SLFN11 in RCC is functionally linked to the response to ICIs, building on prior evidence from hepatocellular carcinoma (HCC) (55). Notably, while our clinical findings in RCC echo the association between SLFN11 and response to ICI reported in HCC, the underlying mechanisms appear distinct. For example, in HCC, SLFN11 is localized primarily in the cytoplasm and promotes an immunosuppressive M2-like macrophage phenotype through up-regulation of C-C motif chemokine ligand 2. In contrast, we found that in RCC, SLFN11 is localized exclusively in the nucleus and promotes M1 macrophage polarization by increasing GM-CSF expression and secretion. Our findings align with previous reports linking M1 macrophages to favorable responses to ICIs. Kim et al. demonstrated that a higher density of M1 macrophages was significantly associated with improved treatment response and longer PFS in patients with RCC treated with nivolumab plus ipilimumab (64). Consistently, transcriptomic analyses of ccRCC have revealed that tumors enriched with M1 macrophages are associated with better prognosis and increased responsiveness to ICIs (65, 66). These reports, along with our current results, underscore the critical role of preferential M1 polarization within the TME in modulating ICIs efficacy in RCC.
Tumors with higher M1 macrophage infiltration typically reflect a “hot” immune microenvironment, primarily by attracting T cells into the TME through the secretion of chemokines such as CXCL9, resulting in the responsiveness to ICIs (67). Our analysis, however, showed no significant difference in CD8+ T cells infiltration between SLFN11-high and -low tumors; instead, the key feature was enrichment of M1 macrophages (Figure 2E). This suggests that pro-inflammatory M1 macrophages, driven by GM-CSF, may enhance antigen presentation, cytokine production, and tumoricidal activity independent of changes in T cell density (68).
The precise mechanism by which SLFN11 up-regulates CSF2 in RCC remains to be clarified. Previous in silico analysis have shown that SLFN11 correlates with immune regulatory pathways and immune gene signatures in several types of cancer (25, 35, 52). Although no significant correlation was observed between SLFN11 and CSF2 expression in the Cancer Therapeutics Response Portal (CTRP) cell-line transcriptomic dataset, which includes a broad panel of human cancer cell lines (Figure S6A), we found a modest positive correlation specifically within kidney cancer cell lines (Figure S6B). Moreover, the TCGA database revealed that kidney cancer has the highest CSF2 expression among all cancers (Figure S6C). Considering the above, RCC are basically potent to express CSF2 and SLFN11 may indirectly regulate CSF2 expression with some additional stimulators in RCC. Recently, Zhang et al. reported that SLFN11 functions as a single-stranded DNA (ssDNA)- activated RNase. SLFN11 directly binds intracellular ssDNA enriched with CGT motifs, translocates to the cytoplasm, and initiates tRNA-cleavage–dependent innate immune responses (53). Given that CSF2 can be induced by endogenous ssDNA in SLFN11-positive HUVEC and HL60 cells subjected to immune stimulators (53), it is plausible that SLFN11 may be activated by intrinsic DNA damage, ssDNA or unknown factors in RCC, leading to CSF2 up-regulation under basal conditions.
In addition to its immunomodulatory role, SLFN11 is well known for enhancing the cytotoxicity of DNA-damage agents, including PARP inhibitors (37, 58). Although such agents are not routinely used for RCC treatment (69), the fact that approximately half of RCC express high SLFN11 level suggests a potentially therapeutic window. Indeed, we confirmed that SLFN11 in RCC cells increased sensitivity to camptothecin and PARP inhibitors (Figure S3A, B). These results raise the intriguing possibility that SLFN11-positive RCC could benefit from combined ICI and PARP inhibitor regimens – a strategy already been tested in clinical trials with agents such as niraparib (IMAGENE trial) and olaparib (WIRE trial) (70, 71).
Study limitations. First, while our findings indicate that GM-CSF plays a key role in M1 macrophage recruitment, further studies are needed to determine whether GM-CSF alone is sufficient to drive this effect. Second, we did not investigate whether SLFN11 directly influences other immune cell populations, such as cytotoxic or regulatory T cells, which could further modulate ICI response. Third, we did not investigate the regulatory factors that control SLFN11 expression or the precise molecular mechanisms linking SLFN11 loss to CSF2 down-regulation. Future studies should focus on identifying the upstream regulators of SLFN11 and characterizing its transcriptional network interactions. Finally, the lack of an orthologous SLFN11 gene in mice presents a challenge for assessing its role in antitumor immunity using standard in vivo models.
Conclusion
This study is the first to demonstrate that SLFN11 plays a pivotal role as an immune regulator in RCC through its role in GM-CSF-mediated M1 macrophage polarization. High SLFN11 expression correlates with increased M1 macrophage infiltration and improved outcomes in patients treated with ICIs, implying its promise as a predictive biomarker and therapeutic target. Future research should focus on validating these findings in larger clinical cohorts and exploring therapeutic strategies to modulate SLFN11 expression to enhance immunotherapy efficacy.
Acknowledgements
The Authors thank the patients and their families for their participation in this study. We also acknowledge the technical assistance provided by the Department of Urology, Osaka University Graduate School of Medicine.
Footnotes
Authors’ Contributions
T.K. and J.M. conceived the project and designed the study. Y.O. contributed to data collection and analysis, figure and table preparation, reference management, and manuscript writing. T.K. also contributed to data collection and analysis and supervised all activities. T.T., N.S., S.I., T.K., A.Y., M.T., Y.H., L.Y., S.N., T.O., T.U., A.Y., G.Y., Y.I., T.H., Y.Y., K.H., and A.K. analyzed and interpreted experimental data. N.N. supervised all aspects of the study. The first draft of the manuscript was prepared by Y.O., T.K., and J.M. All Authors reviewed and approved the final version of the manuscript.
Supplementary Material
Supplementary tables and figures are available at the following OSF repository: https://osf.io/z5t3q/overview
Conflicts of Interest
J.M. reports receiving lecture fees from AstraZeneca plc and Takeda Pharmaceutical Company Limited. All other Authors declare that they have no competing interests.
Funding
This work was supported by the Japan Society for the Promotion of Science, Grants-in-Aid for Scientific Research (JP23H02768 to J.M.), and by the Japan Science and Technology Agency (JST) under the FOREST Program (grant number JPMJFR2056 to J.M.).
Artificial Intelligence (AI) Disclosure
During the preparation of this manuscript, a large language model (ChatGPT, GPT-5.1 Thinking; OpenAI, San Francisco, CA, USA) was used solely for language editing and stylistic improvements in select paragraphs. No sections involving the generation, analysis, or interpretation of research data were produced by generative AI. All scientific content was created and verified by the authors. Furthermore, no figures or visual data were generated or modified using generative AI or machine learning–based image enhancement tools.
- Received December 2, 2025.
- Revision received December 10, 2025.
- Accepted December 15, 2025.
- Copyright © 2026 The Author(s). Published by the International Institute of Anticancer Research.
This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.
References
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.