Background

Colorectal cancer (CRC) ranks as the third most prevalent malignant tumor worldwide.1 Approximately 10% of CRC cases harbor the B-Raf proto-oncogene, serine/threonine kinase (BRAF)V600E mutation, which leads to constitutive activation of BRAF and downstream mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK) signaling in the mitogen-activated protein kinase (MAPK) signaling pathway.2 3 Patients with BRAFV600E-mutant CRC typically exhibit suboptimal responses to standard therapies and have a poorer prognosis compared with those with BRAF wild-type CRC.4–6

Monotherapy with the BRAF inhibitor (BRAFi) has yielded unsatisfactory therapeutic outcomes in treating BRAFV600E CRC, with response rates falling below 5%.7 8 Furthermore, clinical trials have investigated the efficacy of combinations of BRAFi with epidermal growth factor receptor (EGFR) inhibitor, BRAFi with MEK inhibitor (MEKi), and triplet therapy combining BRAFi, EGFR inhibitor, and MEKi.9–11 Despite these efforts to produce sustained MAPK suppression, the objective response rates of BRAFV600E metastatic CRC to these treatment regimens peak at only 26%, with a disappointing median progression-free survival of less than 5 months.12 Hence, novel and effective therapeutic approaches are urgently needed for patients with BRAFV600E-mutant CRC.

Immune checkpoint blockade therapy using programmed death 1 (PD-1) inhibitors shows promise in restoring the immune function of T cells by blocking the interaction between PD-1 and its ligands, programmed death ligand 1 (PD-L1) and PD-L2, potentially resulting in long-lasting antitumor responses.13 Patients with microsatellite instability/deficient mismatch repair (MSI/dMMR) CRC have shown encouraging responses to anti-PD-1 antibodies.14–16 However, the response rates for microsatellite stable/proficient mismatch repair (MSS/pMMR) CRC are notably low,15 emphasizing the urgent need to uncover treatment strategies that can enhance the response of MSS CRC to PD-1 inhibitors.

Histone deacetylase (HDAC) inhibitors, a class of epigenetic drugs, can inhibit the activity of histone deacetylase, inducing the acetylation of histone and some non-histone proteins, thus altering chromatin structure, transcriptional activity, and gene expression.17 Previous studies have demonstrated the efficacy of HDAC inhibitors (HDACi) in CRC by inhibiting tumor cell growth, inducing apoptosis, and causing cell cycle arrest.18 Additionally, HDAC inhibition has shown potential in enhancing the antitumor immune response.19 Meaningful clinical benefits of combining HDACi with anti-PD-1 antibodies have been observed in natural killer cell/ T-cell lymphoma, melanoma, and lung cancer.20–22

In this study, we investigated the potential synergistic effects of HDAC inhibition and MEK inhibition in combination with anti-PD-1 antibodies in BRAFV600E MSS CRC. We characterized the distinguishing features between BRAFV600E MSS CRC and BRAF wild-type MSS CRC. Immunocompetent mouse models were employed to evaluate the efficacy of combined HDAC, MEK, and PD-1 inhibition. Furthermore, various experimental approaches were used to elucidate the underlying cooperative mechanisms among these drugs.

MethodsPatients and samples

CRC tissue samples for this study were acquired from Peking University Shougang Hospital and Peking University Cancer Hospital and Institute.

Patient-derived organoid generation and culture

Patient-derived organoid (PDO) generation was performed as previously described.23 The PDOs were cultured in the IntestiCult Organoid Growth Medium (#06010, Stemcell). The viability of PDOs treated with drugs was assessed using the CellTiter-Glo 3D Cell Viability Assay (G9683, Promega).

Cell lines and cell culture

The human CRC cell lines HT-29, COLO 201, COLO 205, SW1417, SW480 and the murine CRC cell line CT26 were acquired from the American Type Culture Collection (USA). These cell lines are all MSS cell lines. Moreover, HT-29, COLO 201, COLO 205, and SW1417 harbor the BRAF V600E mutation, while SW480 and CT26 are BRAF wild-type cell lines. More detailed molecular backgrounds of these cell lines are provided in online supplemental table S1.

HT-29, COLO 201, COLO 205, and CT26 cells were cultured in Roswell Park Memorial Institute (RPMI)-1640 medium with 10% fetal bovine serum (FBS), while SW1417 and SW480 cells were cultivated in Dulbecco’s Modified Eagle’s Medium (DMEM) with 10% FBS, both at 37°C with 5% CO2. Cell lines were authenticated and tested for Mycoplasma.

Drug library and drug sources

To identify potential therapeutic agents for BRAFV600E MSS CRC, we intersected the Anti-cancer Compound Library (L3000, Selleck) and the Food and Drug Administration-approved Drug Library (L1300, Selleck), resulting in a custom subset of 768 drugs for screening. The biological activity and safety of these 768 drugs have been validated through clinical trials. Additionally, these drugs encompass a wide range of signaling pathways and targets related to antitumor therapies.

Drugs including trametinib, panobinostat, encorafenib, and dabrafenib were procured from Selleck. The anti-mouse PD-1 antibody was InVivoMAb anti-mouse PD-1 (catalog #BE0146, Bio X Cell).

Western blot

Proteins were extracted and separated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, then transferred to a polyvinylidene fluoride membrane. The membrane was blocked in non-fat milk and incubated overnight at 4°C with primary antibodies. After incubation with horseradish peroxidase (HRP) conjugated secondary antibody, proteins were visualized using the chemiluminescent HRP substrate (WBKLS0500, Millipore). Antibodies used are listed in online supplemental table S2.

Reverse transcription-quantitative PCR

Total RNA was isolated using TRIzol reagent (#15596026, Invitrogen) and then converted into complementary DNA through a reverse transcription kit (#6110, Takara). Quantitative PCR (qPCR) was performed on Applied Biosystems 7500 PCR system. The gene relative quantification was calculated using the 2−ΔΔCt method. Primers used are listed in online supplemental table S3.

Chromatin immunoprecipitation assay

Chromatin immunoprecipitation (ChIP) assay was performed using the ChIP Assay Kit (P2080S, Beyotime). Three experimental groups were established: Flag-homeobox C6 (HOXC6) transfected cells immunoprecipitated with anti-Flag antibody (#14793, CST), Flag-HOXC6 transfected cells immunoprecipitated with Normal IgG (#2729, CST), and Flag vector transfected cells immunoprecipitated with anti-Flag antibody. DNA obtained from the ChIP assay was quantified by qPCR assay.

DNA mismatch repair capacity detection assay

The DNA mismatch repair (MMR) capacity of cells was assessed using a plasmid-based and fluorescence-based detection method. Specifically, cells were transfected with a plasmid containing a C:A mismatch along with elements of EGFP and tdTomato. Correct repair of the C:A mismatch allowed EGFP expression. The tdTomato expression served as an internal control for transfection efficiency. The relative MMR capacity was calculated as the peak intensity of EGFP fluorescence divided by that of tdTomato fluorescence. The plasmid was provided by Hanyin Biotech (Shanghai, China).

Construction of CT26 BRAFV637E cell line

The CT26 BRAFV637E cell line was constructed by transfecting CT26 cells with lentivirus containing the murine BRAF V637E sequence (GeneChem, China), followed by puromycin selection.

Animal experiments

All mice used were obtained from Beijing Vital River Laboratory Animal Technology. Mice were housed in a pathogen-free facility and used in compliance with the institutional guidelines for animal care. At the conclusion of the experiment, all mice were euthanized using CO2.

To investigate the therapeutic efficacy of combining HDACi and MEKi with anti-PD-1 antibody in BRAFV600E MSS CRC, a subcutaneous xenograft tumor model using CT26 BRAFV637E cells in BALB/c mice was established. CT26 BRAFV637E cells (2.5×105 cells per injection) were injected into the axilla of mice (female BALB/c mice, 6–8 weeks old). When tumor volume reached 50–100 mm3, the mice were randomly divided into six groups based on the randomization table and subjected to the following treatments: (1) vehicle control (corn oil, oral gavage; PEG-300+Tween 80, intraperitoneal injection); (2) trametinib (1 mg/kg, oral gavage, once daily); (3) panobinostat (5 mg/kg, intraperitoneal injection, once daily); (4) trametinib and panobinostat; (5) anti-mouse PD-1 antibody (10 mg/kg, intraperitoneal injection, two times per day); and (6) trametinib, panobinostat, and anti-mouse PD-1 antibody. The entire treatment period spanned 2 weeks, administered for five consecutive days each week. Tumor volumes and mouse weights were recorded every 3 days by an investigator unaware of treatment. Tumor volume was calculated using the formula 0.5 × L × W2, where L represents tumor length and W represents tumor width. The order of treatments and measurements was randomized daily. At the conclusion of the experiment, the tumors were excised and weighed.

Statistical analysis

Statistical analysis in the study was performed using GraphPad Prism V.8.0 (GraphPad Software). For data that met the assumptions of normal distribution and homogeneity of variance, we used Student’s t-test for comparisons between two groups and one-way analysis of variance for comparisons among multiple groups. When the data did not meet the criteria for normal distribution and homogeneity of variance, we used the Mann-Whitney U test for non-parametric comparisons between two groups and the Kruskal-Wallis test for comparisons among multiple groups. P value<0.05 was considered statistically significant.

ResultsDrug screening identifies HDACi and MEKi as effective drugs for BRAFV600E MSS CRC

To identify effective treatments for BRAFV600E MSS CRC, we conducted extensive drug screening using two BRAFV600E MSS CRC PDOs and three BRAFV600E MSS CRC cell lines (figure 1A). Notably, the two PDOs were derived from patients with BRAFV600E MSS CRC who had undergone a series of pharmacological treatments but yielded suboptimal therapeutic outcomes. The detailed treatment courses for the two patients are depicted in online supplemental figure S1.

Figure 1

Drug screening results in BRAFV600E MSS CRC based on PDOs and cell lines. (A) Graphical overview of the study. (B) Heatmap of drug inhibition for the two PDOs. (C) Venn diagram showing the intersection of drugs with over 60% inhibition on the two PDOs. (D) Bar graph detailing specific inhibition rates of common drugs for the two PDOs. (E) Heatmap of drug responses on the three cell lines. (F) Venn diagram showing the intersection of drugs with over 60% inhibition on the three cells. (G) Bar graph showing specific inhibition rates of common drugs across the three cells. (H) Venn diagram showing the intersection of common drugs between the two PDOs and the three cell lines. BRAF, B-Raf proto-oncogene, serine/threonine kinase; CRC, colorectal cancer; HDAC, histone deacetylase; IHC, immunohistochemistry; MEK, mitogen-activated protein kinase kinase; MSS, microsatellite stable; PDO, patient-derived organoid; PD-1, programmed death 1; RNA-seq, RNA sequencing.

A total of 27 drugs exhibited more than 60% inhibition rates on both PDO-1 and PDO-2 (figure 1B–D). Additionally, we determined the half-maximal inhibitory concentration (IC50) values of several selected drugs on the two PDOs, as shown in online supplemental table S4. Furthermore, we identified that 11 drugs exhibited more than 60% inhibition rates on all three cell lines (figure 1E–G). Among these, HDACi panobinostat and MEKi trametinib demonstrated remarkable inhibitory effects in both BRAFV600E MSS CRC PDOs and cell lines (figure 1H).

HDACi and MEKi demonstrate superior synergistic inhibition in BRAFV600E MSS CRC

To ascertain the synergy of HDACi and MEKi in BRAFV600E MSS CRC, we examined their combined application across various concentrations (figure 2A) and calculated the HAS synergy scores (figure 2B). In HT-29 cells, the HSA synergy score peaked at 50.5, indicating strong synergy between the drugs (figure 2B). This synergistic effect was also observed in encorafenib-resistant HT-29 cells (HT-29_EnR, figure 2C) and two PDO models (figure 2D–F). The resistance phenotype of HT-29_EnR cells to encorafenib was confirmed in online supplemental figure S2A,B (IC50 of HT-29_EnR cells=116.3 µM, IC50 of HT-29 wild-type cells=6.1 µM).

Figure 2

HDACi and MEKi exhibit synergistic inhibition effects by inducing apoptosis and inhibiting the EGFR, MAPK, and phosphoinositide 3-kinase-AKT pathways. (A)-(B) Inhibition rates (A) and HSA synergy scores (B) of combining HDACi and MEKi at various concentrations in HT-29 cells, respectively. (C) Inhibition rates of combining HDACi and MEKi in HT-29_EnR cells. (D)-(E) Inhibition rates (D) and HSA synergy scores (E) of combining HDACi and MEKi in PDO-1, respectively. (F) Inhibition rates of combining HDACi and MEKi in PDO-2. (G) Representative flow cytometry plots showing apoptosis under different treatments in HT-29 cells. (H)-(J) Bar charts showing apoptosis in HT-29 (H), COLO 205 (I), and COLO 201 cells (J) under different treatments (trametinib, 0.1 µM; panobinostat, 0.1 µM), respectively. Results represent means±SD (n=3). **p<0.01; ***p<0.001; ****p<0.0001 (Student’s t-test). (K)-(N) Western blotting for HT-29 cells (K), PDO-1 (L), PDO-2 (M), and HT-29_EnR cells (N) with different treatments, respectively. HT-29 and HT-29_EnR cells: trametinib, 0.1 µM; panobinostat, 0.1 µM; PDO-1 and PDO-2: trametinib, 0.2 µM; panobinostat, 0.2 µM. HT-29_EnR, encorafenib-resistant HT-29 cells. AKT, protein kinase B; EGFR, epidermal growth factor receptor; HDACi, histone deacetylase inhibitors; MAPK, mitogen-activated protein kinase; MEKi, mitogen-activated protein kinase kinase inhibitor; PDO, patient-derived organoid; Pano, panobinostat; Tra, trametinib.

Furthermore, we compared the synergistic effects of the HDACi-MEKi combination to the conventional BRAFi-MEKi combination. In the PDO-1 model, the synergy between the HDACi panobinostat and the MEKi trametinib (HSA synergy score: max=34.4, mean=14.7) surpassed that of the BRAFi encorafenib with the MEKi trametinib (HSA synergy score: max=29.6, mean=5.8) and the BRAFi dabrafenib with the MEKi trametinib (HSA synergy score: max=32.3, mean=1.1) (figure 2D–E, online supplemental figure S3A). Similar findings were observed in HT-29 cells (figure 2A,B, online supplemental figure S3E–H), PDO-2 model (figure 2F, online supplemental figure S3I,J) and HT-29_EnR cells (figure 2C, online supplemental figure S3K). These findings underscored the superior efficacy of the HDACi-MEKi combination in treating BRAFV600E MSS CRC, significantly outperforming the conventional BRAFi-MEKi combination.

HDACi and MEKi synergistically promote cell apoptosis

To elucidate the mechanism underlying the synergistic inhibitory effect of the two agents, we conducted cell apoptosis detection. In HT-29 cells, we observed a significantly higher apoptosis rate in the combination group compared with the control group as well as either single-agent groups (figure 2G,H). Similarly, increased apoptosis with the combination treatment was noted in COLO 205 and COLO 201 cells (figure 2I,J, online supplemental figure S4A,B). These findings suggest that the HDACi and MEKi can synergistically promote cell apoptosis in BRAFV600E MSS CRC.

HDACi and MEKi synergize to inhibit EGFR, MAPK, and AKT pathways

Furthermore, we investigated the effects of the HDACi-MEKi combination on key signaling pathways. Western blotting showed that this combination significantly reduced EGFR protein expression, EGFR phosphorylation, protein kinase B (AKT) protein expression, AKT phosphorylation, and ERK1/2 phosphorylation in HT-29 cells (figure 2K), suggesting that the combination effectively inhibited the EGFR, MAPK, and AKT pathways. Concordant results were obtained in PDOs (figure 2L,M) and HT-29_EnR cells (figure 2N). In contrast, the combination of BRAFi (darafenib or encorafenib) with MEKi elevated the phosphorylation of EGFR (online supplemental figure S5A,B). Notably, the elevation of EGFR phosphorylation was a characteristic of cells resistant to encorafenib (online supplemental figure S5C). Furthermore, BRAFi encorafenib with MEKi was ineffective in inhibiting the AKT pathway (online supplemental figure S5D). These observations may partially elucidate why the HDACi-MEKi combination was more effective than the BRAFi-MEKi combination.

HDACi and MEKi synergistically suppress the key transcription factor HOXC6

Remarkably, western blotting also revealed that the combination therapy decreased the protein expression of the key transcription factor HOXC6 in HT-29 cells (figure 2K), PDOs (figure 2L,M) and HT-29_EnR cells (figure 2N). This reduction is crucial for treating BRAFV600E MSS CRC, as we found substantial evidence suggesting that HOXC6 plays a significant role in promoting tumor development and treatment resistance in this cancer subtype, resonating with several previous studies that have reported the pivotal role of HOXC6 in tumors.24 25

HOXC6 contributes to treatment resistance in BRAFV600E MSS CRC

As depicted, HOXC6 expression was higher in patients with BRAFV600E CRC compared with patients with BRAF wild-type CRC, particularly within the MSS subtype (figure 3A). This pattern was also observed in CRC cells from the CCLE database, with HOXC6 expression being higher in the BRAFV600E MSS CRC cell lines compared with the BRAF wild-type MSS counterparts (figure 3B). Moreover, HOXC6 expression in HT-29_EnR cells was significantly higher than that in HT29 wild-type cells (figure 3C; online supplemental figure S5C). After the knockdown of HOXC6, the IC50 value for encorafenib in HT-29_EnR cells was significantly decreased compared with the control (figure 3D).

Figure 3

HOXC6 contributes to treatment resistance in BRAFV600E MSS CRC and regulates the MAPK and AKT pathways and the MYC gene. (A) Left: HOXC6 expression in BRAFV600E and BRAF wild-type groups in TCGA COAD. ****p<0.0001 (Mann-Whitney U test). Right: HOXC6 expression in BRAFV600E dMMR, BRAF wild-type dMMR, BRAFV600E pMMR, and BRAF wild-type pMMR groups in TCGA COAD. ns, not significant; ***p<0.001 (Mann-Whitney U test). (B) HOXC6 expression in BRAFV600E MSS and BRAF wild-type MSS CRC cell lines from the CCLE database. **p<0.01 (Student’s t-test). (C) HOXC6 mRNA expression in HT-29 and HT-29_EnR cells. *p<0.05 (Student’s t-test). (D) IC50 values of encorafenib in HOXC6-knockdown HT-29_EnR cells and control cells. ***p<0.001 (Student’s t-test). (E) IC50 values of dabrafenib, encorafenib, and trametinib in HOXC6-overexpressing HT29 cells (HT29-HOXC6) and control cells (HT29-Control), respectively. Results represent means±SD (n=4). *p<0.05 (Student’s t-test). (F) Top 10 enriched pathways in HOXC6-overexpressing cells revealed by KEGG analysis. (G) Western blotting for HOXC6-overexpressing cells and control cells. (H) Western blotting for HOXC6-knockdown cells and control cells. (I) Cell viability profiles of four BRAFV600E MSS CRC cell lines treated with dabrafenib, encorafenib, and trametinib, respectively. Results represent means±SD (n=4). (J) Western blotting for the four BRAFV600E MSS CRC cell lines. (K) Peak annotation analysis indicating the distribution of binding sites. (L) Chromatin immunoprecipitation-quantitative results on the percentage of Input for MYC promoter. **p<0.01 (Student’s t-test). (M) Relative luciferase activity analyzed in the dual luciferase reporter assay to examine HOXC6 binding to the MYC promoter. MYC_WT, MYC_wild-type; MYC_Mut, MYC_mutant. Results represent means±SD (n=3). **p<0.01; ns, not significant (Student’s t-test). (N) Schematic of HOXC6 binding to the MYC gene. AKT, protein kinase B; BRAF, B-Raf proto-oncogene, serine/threonine kinase; COAD, colon adenocarcinoma; CRC, colorectal cancer; dMMR, deficient mismatch repair; ERK, extracellular signal-regulated kinase; HOXC6, homeobox C6; IC50, half-maximal inhibitory concentration; KEGG, Kyoto Encyclopedia of Genes and Genomes; MAPK, mitogen-activated protein kinase; MSS, mismatch repair; p-AKT, phosphorylated-AKT; p-ERK, phosphorylated-ERK; PI3K, phosphoinositide 3-kinase; pMMR, proficient mismatch repair; TCGA,The Cancer Genome Atlas; TPM, transcripts per kilobase million.

Furthermore, the drug sensitivity results revealed that the IC50 values of BRAFi dabrafenib and encorafenib, as well as MEKi trametinib, were significantly elevated in HT-29 cells overexpressing HOXC6 compared with the control cells (figure 3E), suggesting that HOXC6 overexpression conferred a reduction in the efficacy of BRAFi and MEKi.

HOXC6 regulates MAPK and AKT pathways and MYC gene in CRC

Moreover, we analyzed differentially expressed genes between HOXC6-overexpressing cells and control cells by RNA sequencing (RNA-seq). Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis of differentially upregulated genes revealed significant enrichment of both the MAPK pathway and the phosphoinositide 3-kinase (PI3K)-AKT pathway in HOXC6-overexpressing cells (figure 3F). Gene set enrichment analysis (GSEA) analysis on the MSS subtype samples from The Cancer Genome Atlas (TCGA) colon adenocarcinoma (COAD) (online supplemental figure S6A,B) and MSS subtype CRC cell lines in the CCLE database (online supplemental figure S6C,D) confirmed this enrichment in the HOXC6 high-expression group. Western blotting indicated increased phosphorylated-ERK1/2 (p-ERK1/2) and phosphorylated-AKT in HOXC6-overexpressing cells (figure 3G) and decreased levels in HOXC6-knockdown cells (figure 3H). Reverse transcription-quantitative PCR (RT-qPCR) results for HOXC6 overexpression and HOXC6 knockdown are shown in online supplemental figure S7A,B, respectively. These findings suggest a strong link between HOXC6 and these pathways. Additionally, a negative correlation between innate treatment sensitivity to BRAFi and MEKi (figure 3I) and HOXC6 expression as well as the activation of both the MAPK and AKT pathways (figure 3J) was observed.

To investigate target genes regulated by HOXC6, we employed CUT&Tag technology. Peak annotation analysis indicated that the majority (42.18%) of binding sites were located in promoter regions (figure 3K). When intersecting neighboring genes of the peaks identified by CUT&Tag with differentially expressed genes identified by RNA-seq, a set of 197 genes was identified, including MYC, which exhibited the highest peak score (online supplemental figure S8A,B). The result of the ChIP-qPCR assay revealed that the % of input for the MYC promoter in the Flag-HOXC6+anti Flag group was significantly higher compared with the Flag-HOXC6+Normal IgG group and Flag+anti Flag control group, suggesting specific binding between the HOXC6 protein and the MYC promoter (figure 3L). The dual luciferase reporter assay further confirmed that HOXC6 enhances luciferase activity linked to the wild-type MYC promoter, but not to a mutant promoter (figure 3M). The sequence of 5'-CATGAATTAACTA-3', strand= “-", appeared to be the site where HOXC6 binds to the MYC gene (figure 3N). Additionally, protein levels of MYC were increased in HOXC6-overexpressing cells (figure 3G) and decreased in HOXC6-knockdown cells (figure 3H). These results highlight the direct regulatory role of HOXC6 on MYC.

HOXC6 drives CRC growth and metastasis and predicts poor prognosis

Notably, knockdown of HOXC6 significantly suppressed the growth of HT-29 cells (online supplemental figure S9A) and diminished their colony formation capacity in vitro (online supplemental figure S9B,C). Conversely, HOXC6 overexpression promoted cell growth (online supplemental figure S10A–C). In vivo, shHOXC6 cell group showed reduced subcutaneous xenograft tumor volume and weight compared with the control cell group (online supplemental figure S9D–F). In contrast, the upregulation of HOXC6 fostered the in vivo growth of cells (online supplemental figure S10D–F). Furthermore, shHOXC6 cell group exhibited a lower incidence of liver metastasis (0/6) compared with the control HT-29 cell group (3/7) (online supplemental figure S9G–I). Conversely, the proportion of liver metastasis that occurred in the oe-HOXC6 SW480 cell group (2/8) was higher than that in the control cell group (1/9) (online supplemental figure S10G,H). These results emphasize HOXC6’s role in CRC growth and metastasis and the significance of HOXC6 inhibition by the HDACi-MEKi combination.

In addition, analysis of HOXC6 expression in 252 patients with CRC’s tissues revealed that low HOXC6 expression correlated with improved overall survival (p=0.0051, online supplemental figure S6J–L). The basic clinical information of the 252 patients with CRC is shown in online supplemental table S5. This was confirmed in TCGA COAD (online supplemental figure S9M), TCGA rectum adenocarcinoma (READ) (online supplemental figure S11A), and GSE39582 data set (online supplemental figure S11B). Specifically, this association was identified within pMMR CRC (online supplemental figure S11C,E) but not in dMMR CRC (online supplemental figure S11D,F). Furthermore, HOXC6 expression was elevated in CRC tumor tissues than normal tissues (online supplemental figure S11G), higher in right-sided tumors (online supplemental figure S11H), and increased in MSI-high types (online supplemental figure S11I). In MSS/MSI-low cases, HOXC6 expression was higher in stage IV tumors (online supplemental figure S11J). Interestingly, HOXC6 expression was higher in CMS1 and CMS4 types (online supplemental figure S11K). Similar patterns were identified in GSE39582 data set (online supplemental figure S11L–N).

HDACi and MEKi synergize to enhance immune activation gene expression

To investigate the regulatory effects of combination therapy on immune-related genes, including immune activation and antigen recognition, we compared the expression of these genes under different drug treatments. RT-qPCR results showed a significant increase in the expression of interferon gamma (IFN-γ) pathway-related genes, such as IFN regulatory factor 1 (IRF1), C-X-C motif chemokine ligand 9 (CXCL9), and C-X-C motif chemokine ligand 11, with the combined application of HDACi and MEKi (figure 4A–C). Moreover, the expression of antigen presentation-related genes, including major histocompatibility complex, class I, A (HLA-A), HLA-B, and HLA-C, was markedly elevated with the combined drug treatment (figure 4D–F).

Figure 4

histone deacetylase inhibitors and mitogen-activated protein kinase kinase inhibitor synergistically promote the expression of immune activation-related genes and reduce the mismatch repair (MMR) capacity of cells. (A)-(F) Gene expression changes of interferon-γ pathway-related genes IRF1 (A), CXCL9 (B) and CXCL11 (C), and antigen-presentation-related genes such as HLA-A (D), HLA-B (E) and HLA-C (F) in HT-29 cells under different treatments (trametinib, 0.1 µM; panobinostat, 0.1 µM), respectively. Results represent means±SD (n=3).*p<0.05; **p<0.01; ****p<0.0001 (Student’s t-test). (G) Western blotting for the cGAS/STING pathway in HT-29 cells with different treatments (trametinib, 0.1 µM; panobinostat, 0.1 µM). (H) Relative MMR capacity of HT-29 cells with different treatments (trametinib, 0.1 µM; panobinostat, 0.1 µM). Results represent means±SD (n=3). *p<0.05; **p<0.01; ****p<0.0001 (Student’s t-test). (I) Western blotting for MMR proteins in HT-29 cells with different treatments (trametinib, 0.1 µM; panobinostat, 0.1 µM). (J)-(K) Western blotting for MMR proteins in PDO-1 (J) and PDO-2 (K) with different treatments (trametinib, 0.2 µM; panobinostat, 0.2 µM), respectively. cGAS/STING, cyclic guanosine monophosphate–adenosine monophosphate synthase/stimulator of interferon genes; CXCL9, C-X-C motif chemokine ligand 9; CXCL11, C-X-C motif chemokine ligand 11; IRF1, interferon regulatory factor 1; mRNA, messenger RNA; Pano, panobinostat; PDO, patient-derived organoid; Tra, trametinib.

HDACi and MEKi synergistically activate the cGAS/STING pathway

Furthermore, we investigated the effects of the HDACi and MEKi combination on the cyclic guanosine monophosphate–adenosine monophosphate synthase/stimulator of interferon genes (cGAS/STING) pathway. Western blotting demonstrated that the co-administration of HDACi and MEKi in HT-29 cells markedly increased the phosphorylation of STING, TBK1, and IRF3—key indicators of the cGAS/STING pathway activation (figure 4G). These findings suggest that HDACi and MEKi act synergistically to activate the cGAS/STING pathway, which is beneficial to the endogenous antitumor immune response.

HDACi and MEKi synergistically reduce cells’ mismatch repair capacity

To evaluate the impact of the combination therapy on the DNA repair capacity of BRAFV600E MSS CRC cells, we employed a plasmid- and fluorescence-based detection method. Cells were transfected with a plasmid containing C:A mismatch to assess the effect of the combined treatment on MMR capacity. Fluorescence detection results revealed a significant inhibition of MMR capacity in HT-29 cells when treated with HDACi and MEKi together (figure 4H). Additionally, western blotting demonstrated suppression of MMR proteins such as PMS1 homolog 2 (PMS2) and mutS homolog 6 (MSH6) in HT-29 cells following combination treatment (figure 4I). Notably, PDO-based experiments showed a more pronounced inhibitory effect on MMR proteins with combination therapy (figure 4J–K).

The DNA MMR system is crucial for maintaining the accuracy of DNA replication and genome integrity.26 Deficiencies in DNA MMR lead to frequent polymorphism in short-sequence DNA repeats, known as microsatellite instability, and accumulation of mutations. This can increase the presence of mutation-associated neoantigens, potentially promoting infiltration of tumor-infiltrating lymphocytes and enhancing tumor sensitivity to immune checkpoint blockade.27

Immune microenvironment differs between BRAFV600E MSS CRC and BRAF wild-type MSS CRC

To elucidate differential characteristics between patients with BRAFV600E MSS and patients with BRAF wild-type MSS CRC, we compared their gene expression profiles using RNA-seq. A total of 29 patient samples were ultimately analyzed, comprising three cases of BRAFV600E MSS CRC and 26 cases of BRAF wild-type MSS CRC. Gene Ontology (GO) enrichment analysis indicated that genes upregulated in the BRAFV600E MSS group were prominently enriched in pathways related to immune regulation and cytokine production (figure 5A,B). GSEA analysis also revealed significant enrichment of immune-related gene sets in the BRAFV600E MSS group (figure 5C–F). Consistently, in TCGA COAD, immune-related pathways (online supplemental figure S12A,B) and immune-related gene sets such as PD-1 signaling and activation of immune response were enriched in the BRAFV600E MSS group (online supplemental figure S12C–E). Additionally, significant differences in single-sample enrichment scores of immunity-relevant gene sets, such as recognition of cancer cells by T cells and immune checkpoint, were also observed between the two groups (online supplemental figure S12F–K).

Figure 5

Immune microenvironment characteristics and immune cell infiltration differ between BRAFV600E MSS CRC and BRAF wild-type MSS CRC. (A) Volcano plot demonstrating differentially expressed genes between patients with BRAFV600E MSS CRC and patients with BRAF wild-type MSS CRC. (B) Top 10 significantly enriched pathways in BRAFV600E MSS CRC identified by GO enrichment analysis. (C)-(F) GSEA analysis revealing significant enrichment in pathways such as immunoregulatory interactions between a lymphoid and a non-lymphoid cell (C), MHC class II antigen presentation (D), Hallmark_IL-2_STAT5_signaling (E), and Hallmark_IL-6_JAK_STAT3_signaling (F) in BRAFV600E MSS CRC. (G)-(H) GSEA analysis illustrating significant enrichment of IL-17 signaling pathway (G) and IL-10 signaling pathway (H) in BRAFV600E MSS CRC cell lines. (I)- (J) Representative images of multiplex fluorescent immunohistochemistry for CD8+ cells, PD-1+ cells, and DAPI in BRAF wild (I) and BRAFV600E(J) MSS CRC tissues. Scale bars, 200 µm. (K) Quantification of PD-1+ cells in BRAFV600E MSS CRC and BRAF wild-type MSS CRC groups. *p<0.05 (Student’s t-test). (L) Quantification of CD8+ cells in BRAFV600E MSS CRC and BRAF wild-type MSS CRC groups. ns, not significant (Student’s t-test). (M) ssGSEA analysis showing the scores of immune cell infiltration in BRAFV600E MSS CRC and BRAF wild-type MSS CRC groups. *p<0.05; **p<0.01; ***p<0.001; ****p<0.0001 (Mann-Whitney U test). BRAF, B-Raf proto-oncogene, serine/threonine kinase; CRC, colorectal cancer; GO, Gene Ontology; GSEA, gene set enrichment analysis; IL, interleukin; KEGG, Kyoto Encyclopedia of Genes and Genomes; MHC, major histocompatibility complex; MSS, microsatellite stable; PD-1, programmed death 1; ssGSEA, single sample gene set enrichment analysis.

Moreover, gene expression profiles of BRAFV600E MSS CRC cell lines (n=6) and BRAF wild-type MSS CRC cell lines (n=38) in the CCLE database were compared. GSEA analysis indicated significant enrichment of the interleukin (IL)-17 signaling pathway and the IL-10 signaling pathway in the BRAFV600E MSS cell lines (figure 5G,H). These findings suggest potential differences in the immunogenicity of the tumor epithelial cells between BRAFV600E MSS CRC and BRAF wild-type MSS CRC.

Immune cell infiltration differs between BRAFV600E MSS CRC and BRAF wild-type MSS CRC

Subsequently, differences in immune cell infiltration in the tumor tissues between the two patient groups were investigated through multiplex fluorescent immunohistochemistry (mfIHC) of tissue sections (figure 5I,J). The results revealed a significantly higher number of PD-1+ cells in the BRAFV600E MSS group compared with the BRAF wild-type MSS group (figure 5K). Consistently, PD-1 expression was also significantly higher in BRAFV600E MSS CRC in the TCGA COAD data set (online supplemental figure S12L,M). Additionally, there was a trend towards increased CD8+ cell infiltration in the BRAFV600E MSS group, though it was not statistically significant (figure 5L). Furthermore, the results of single sample gene set enrichment analysis (ssGSEA) analysis demonstrated significant variations in the abundance of immune cell types between the BRAFV600E MSS and BRAF wild-type MSS groups, with higher levels of memory CD8+ T cells, type II helper T cells, macrophages, and neutrophils observed in the BRAFV600E MSS group (figure 5M). These findings indicate a more robust immune profile within the tumor microenvironment of patients with BRAFV600E MSS CRC, potentially influencing the efficacy of immunotherapy in these patients.

Combination of HDACi, MEKi, and anti-PD-1 antibody effectively inhibit tumor growth

The aforementioned study findings suggest that HDACi and MEKi have synergistic effects and potential for combination with immunotherapeutic agents such as anti-PD-1 antibody in BRAFV600E MSS CRC.

Prior studies have confirmed that the BRAFV637E mutation in mice mirrors the effects observed with the BRAFV600E mutation in humans.28 29 Therefore, we established the CT26 BRAFV637E cell line and validated its characteristics. As shown, the p-ERK1/2 level in the CT26 BRAFV637E cell line was significantly elevated compared with that in the wild-type C