C57BL/6 mice (6-9 weeks old) and BALB/c mice (6-9 weeks old) were purchased from Japan SLC (Shizuoka, Japan). Rag2^-/-OT-I and Rag2^-/-OT-II T cell receptor (TCR) transgenic mice (6-9 weeks old) were purchased from Taconic Biosciences. These mice were kept and bred in the animal unit at Kanazawa Medical University, an environmentally controlled and specific pathogen-free facility, in accordance with the guidelines for experimental animals defined by the facilities. Animal experimental protocols were approved by the Animal Research Committee at Kanazawa Medical University (Approval No.:2020-30, 2022-17). The procedures were carried out in accordance with the approved guidelines. Ovalbumin (OVA) _257-264 peptide SIINFEKL and OVA_323-339 peptide were purchased from Anaspec.

E.G7 tumor cell line was purchased from the American Type Culture Collection (ATCC) and B16 melanoma cell line was kindly gifted by Dr. Nobuo Yamaguchi (Kanazawa Medical University). E.G7 cells were cultured in Roswell Park Memorial Institute (RPMI)- 1640 medium modified to contain 2mM L-glutamine, 10mM HEPES, 1mM sodium pyruvate, 4500mg/l glucose, and 1500mg/l sodium bicarbonate (Sigma-Aldrich) supplemented with 10% fetal calf serum (FCS), B16 cells were cultured in Dulbecco’s modified eagle medium (DMEM)supplemented with 10% FCS.

The four small molecule inhibitors (YPPP) were first reconstituted into stock solutions; 10 mM Y27632 (Fujifilm Wako) was prepared in sterilized phosphate buffered saline (PBS), and 40 mM PD0325901 (Fujifilm Wako), 10 mM PD173074 (Fujifilm Wako) and 10 mM PD98059 (Fujifilm Wako) were prepared in dimethyl sulfoxide (DMSO, Nakarai tesque). Stock solutions were stored at −20°C until use. Y27632, PD0325901, PD173074 and PD98059 were used at final concentrations of 50, 0.04, 0.01, 6.3 μM.

BM cells were prepared from C57BL/6 mice and these cells were cultured with GM-CSF as we described previously (8). Briefly, BM cells were cultured in RPMI-1640 supplemented with 10% FCS, 5 x 10^-5M 2-mercaptoethanol, 100 U/ml penicillin, 100 μg/ml streptomycin and 25 ng/ml GM-CSF (Biolegend) and with DMSO or YPPP for 6 days. On day 6, CD11c⁺ cells were isolated using the magnetic-activated cell sorting (MACS) system with magnetic microbead-conjugated anti-CD11c antibody (Miltenyi Biotec). The purities of the sorted CD11c⁺ fractions by using CD11c microbeads were consistently ≥90% ( Supplementary Figure 2 ). In E.G7 tumor model, CD11c⁺ cells were stimulated with 10 ng/ml lipopolysaccharide (LPS) for 12 h and loaded with 10 μM OVA_257-264 peptide SIINFEKL for 2h prior to injection. In B16 melanoma model, CD11c⁺ cells were stimulated with 10 ng/ml lipopolysaccharide (LPS) for 12 h prior to injection.

Flow cytometry was conducted using the following antibodies (All purchased from Biolegend unless stated otherwise): anti-CD11c (N418), anti-I-A/I-E (M5/114.15.2), anti-CCR7 (4B12), anti-CD40 (3/23), anti-CD11b (M1/70), anti-CD80 (16-10A1), anti-CD86 (GL-1), anti-CD4 (GK1.5), anti-CD8 (53-6.7), anti-CD25 (PC61), anti-CD62L (MEL14), anti-CD69(H1.2F3) and anti-PD-L2 (TY25). Fc-mediated nonspecific staining was blocked with anti- CD16/32 (2.4G2 hybridoma culture supernatant). Events were acquired using a FACSCanto II (BD Biosciences), and the data of 10,000-100,000 events were analyzed using the FACSDiva (BD Biosciences) or FlowJo software programs (FlowJo). The surface molecule expressions were calculated by defining the delta mean fluorescence intensity between the specific antibody stain and the isotype-matched control antibody. CD11c⁺I-A/I-E^high cells and CD11c⁺I-A/I-E^int cells were isolated by using a cell sorting system (SH800; Sony Biotechnology) or FACS Aria fusion (BD Biosciences). Separation of these cells were shown in Supplementary Figures 5A, B . Briefly, BM lineage negative (Lin (–)) cells were immunomagnetically pre-enriched using PE-Cy5-conjugated antibodies against lineage antigens, including CD3ϵ (145-2C11), CD4 (GK1.5), CD8α (53-6.7), B220 (RA3-6B2), CD19 (MB19-1), CD11c (N418), CD11b (M1/70), Gr-1 (RB6-8C5), NK1.1 (PK136), anti-Cy5 microbeads and autoMACSpro (Miltenyi Biotec). BM- Lin (-) cells were cultured with 25 ng/ml GM-CSF (Biolegend) for 6 days. On day 3, BM-Lin (-) derived CD11c⁺ cells were isolated using the MACS system with magnetic microbead-conjugated anti-CD11c antibody (Miltenyi Biotec). BM- Lin (-) cells were then stained with FITC-anti-I-A/I-E (M5/114114.15.2) and PE-Cy7-anti-CD11c (N418) (all from BioLegend). CD11c⁺I-A/I-E^high cells and CD11c⁺I-A/I-E^int cells were sorted by gating on indicated cells from CD11c⁺ cells. The purity of the sorted populations was consistently ≥90%.

Sorted CD11c⁺ cells in mouse bone marrow culture with GM-CSF and DMSO/YPPP for 6 days were incubated in 96-well round-bottom tissue culture plates. For OT-I T cells stimulation, CD11c⁺ cells were incubated with OVA_257-264 peptide (1 nM) in the presence of 100 ng/ml of LPS. For OT-II T cells stimulation, CD11c⁺ cells were incubated with OVA_323-339 peptide (10 μM) in the presence of 100 ng/ml of LPS and co-cultured with naive Cytotell green-labeled OT-II T cell. For allogenic BALB/c T cells stimulation, CD11c⁺ cells were incubated with LPS (100 μg/ml). Various number of CD11c⁺ cells (7.3 x, 22.0 x or 66.0 x 10² cells were incubated with 2.0 x 10⁴ T cells. T cell proliferation and activation was assessed by flow cytometry after 3 days for OT-I T cells and after 5 days for OT-II and BALB/c T cells.

Cells were grown into 24 well plates. At day 6, 10 μl of CCK-8 solution (Dojindo) was added into each well and cultured for another 2 hours, and measured at 490 nm in a microplate reader Multiscan JX (Thermo scientific). The ratios of proliferation were calculated as follows: the absorbance of culture supernatants at 490 nm/absorbance of the control cell supernatants at 490 nm.

Cells were incubated with the indicated stimulators. The levels of cytokines in the culture supernatants were determined using ELISA kits, in accordance with the manufacturers’ instructions. The mouse IL-12p70, IL-12p40, IL-6, TNF-α ELISA MAX Standard kits (Biolegend) were used.

The cells for immunoblotting were prepared as we described previously (22). Briefly, Cells were solubilized in lysis buffer (1.0% NP-40, 50 mM HEPES, pH7.4, 150 mM NaCl), containing protease and phosphatase inhibitor (Thermo fisher scientific). Cell lysates were separated by sodium dodecyl sulfate (SDS)- polyacrylamide gel electrophoresis (PAGE), transferred to a Polyvinylidene difluoride (PVDF) membrane (Merck), and detected with the following antibodies using ECL substrate (Bio-Rad): rabbit anti-CCAAT enhancer binding protein (C/EBP)α anti-phospho-C/EBPα (Ser21), anti-phospho- C/EBPα (Thr221/226), anti-Erk1/2, anti-phospho- Erk1/2, anti-JNK, anti-phospho-JNK, anti-NF-κB(p65), anti-phospho-NF-κB(p65), mouse anti-IκBα (Cell Signaling Technology), mouse anti-IκBβ (Santa Cruz Biotechnology), mouse anti-βactin (Sigma-aldrich), horseradish peroxidase (HRP)-conjugated mouse IgG (Cell Signaling Technology), or rabbit IgG antibodies (Cell Signaling Technology). Digital images were obtained using an ImageQuant LAS4000 mini instrument (GE Healthcare). Densitometry was performed on scanned blots using the ImageQuant TL software program (GE Healthcare).

The protocol by which E.G7 or B16 melanoma tumor model were previously described (23, 24). These tumor cells were established by intradermally injecting 2 x or 10 x 10⁵ tumor cells (in 50 μl of DMEM) into the back flanks of 8-9 weeks old C57BL/6 mice (day 0). For dendritic cell vaccine (DCV) therapy, on days 7, 10, 13, 1.25 or 12.5 x 10⁴ syngeneic BM derived CD11c⁺ cells (in 50 μl DMEM/0.1% bovine serum albumin) were injected intratumorally with or without anti Programmed death (PD)-1 (Biolegend, 20μg/mouse) in 200 μl saline intraperitoneal injection. Tumor sizes were measured twice a week. Tumor volume was calculated by the following formula: tumor volume. 0.4 x length (mm) x [width (mm)]² (24). Mice were euthanized when they became moribund or when their tumors exceeded 20 mm in diameter.

The fragmented double-strand cDNA was synthesized using NEB Next Ultra II Directional RNA Library Prep Kit (New England BioLabs) according to manufacturer’s instructions. After library quality was assessed using the TapeStation 4200 with High Sensitivity D1000 ScreenTape (Agilent Technologies). All libraries were quantified using the HS Qubit dsDNA assay (Thermo Fisher Scientific). All libraries were sequenced (2 × 75 bp) using Illumina NextSeq 500 (Illumina) according to the standard Illumina protocol. The FASTQ files were generated using the bcl2fastq software (Illumina).

FASTQ files were aligned to the reference mouse genome (mm10) using HISAT2 (version 2.1.0) (25). The StringTie algorithm (v.1.3.4d) (26) was then used with default parameter settings to assemble RNA-Seq alignments into annotated transcripts to estimate their expression using the UCSC annotated mouse genome (mm10) assembly file. Subsequently, the transcript expression was normalized using the transcripts per million (TPM) algorithms. For differential expression analysis, we used the R package (edger) (27), as described previously (28, 29). Gene ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses were performed using Enrichr web server (https://maayanlab.cloud/Enrichr/) (30–32).

Statistical analysis was carried out by ANOVA using GraphPad Prism software. All experimental repetitions and numbers of specimens and mice are indicated in the figure legends. One-way and two-way ANOVA were used for the experiments with one variable and two variables, respectively. p < 0.05 was considered to be statistically significant. For survival analysis, p values were computed using the Log Rank test. p values < 0.05 was considered to be statistically significant.

Previous study in mouse BM culture with GM-CSF have demonstrated the presence of primarily two subsets: CD11c⁺I-A/I-E^high cells (GM-DCs) and CD11c⁺I-A/I-E^int cells (GM-Macs) (33). To determine the optimal combination for DC differentiation in a mouse BM-derived DC culture system, we initially examined 16 different combinations of the four small molecule compounds ( Figure 1A ). The optimal concentration of Y27632, PD0325901 and PD173074 were determined through preliminary tests ( Supplementary Figures 1A, B ), while effective concentration of PD98059 was demonstrated in previous study (21, 34). In this culture system, the addition groups of the 6th, 13th, and 16th combination of the small molecule inhibitors significantly increased the percentages of CD11c⁺I-A/I-E^high cells in the 40.1 ± 25.0, 44.8 ± 16.5, and 53.4 ± 11.7% compared to 17.4 ± 10.5% for the control ( Figures 1B, C ). No effect was observed with the other additional groups. On the other hand, the addition groups of the 3th, 5th 15th, and 16th combination of the small molecule inhibitors significantly decreased the percentages of CD11c⁺I-A/I-E^int cells in the 25.8 ± 2.40, 25.2 ± 0.49, 24.5 ± 0.57, and 30.0 ± 1.77% compared to 45.4 ± 6.35% for the control ( Figures 1B, E ). The total cell number of CD11c⁺I-A/I-E^high cells shows no significant difference between control and 16th group ( Figure 1D ). However, the total cell number of CD11c⁺I-A/I-E^int cells of 16th group is markedly reduced when compared with control ( Figure 1F ). These results indicate that the addition of these small molecule inhibitor cocktails significantly altered the percentage of CD11c⁺I-A/I-E^high cell, representing DCs, with the 16^th small molecule inhibitor cocktail YPPP, yielding cells of the highest purity ( Figures 1B-F ). It has been reported that GM-CSF induced differentiation of myeloid-derived suppressor cells (MDSCs) from BM cells, defined as Ly6G^highLy6c^int cells (35). Therefore, we examined the effect of YPPP on GM-CSF induced MDSCs. The YPPP did not have a significant impact on MDSCs differentiation ( Supplementary Figures 2A, B ).

The anti-tumor activity of DCV used in cancer immunotherapy relies on the quality of the DC-containing cell fraction CD11c⁺ cells in the DCV. Therefore, we examined cytokine production in DMSO- and YPPP-treated BM-derived CD11c⁺ cells, as LPS is known to be a potent inducer of the Th1 cytokine IL-12 by using magmatically purified CD11c⁺ cells ( Supplementary Figure 3 ). The level of proinflammatory cytokine IL-6 and IL-12 were measured after LPS stimulation. As shown in Figure 2A , YPPP-treated BM derived CD11c⁺ cells (YPPP- CD11c⁺ cells) produced high amount of IL-12p70 and IL-12p40 compared to DMSO-treated CD11c⁺ cells (DMSO- CD11c⁺ cells) after LPS stimulation, while no difference was observed in IL-6 and TNF-α levels between DMSO and YPPP-CD11c⁺ cells. Next, we investigated the activation of key signaling molecules in the Toll like receptor (TLR) 4-mediated signaling pathway, which was activated by LPS stimulation. Previous studies have shown that CCAAT enhancer binding protein alpha (C/EBPα), whose activation is regulated by extracellular signal-regulated kinase 1/2 (Erk1/2), negatively regulates IL-12 production upon LPS stimulation (36). Therefore, we examined the phosphorylation of Erk1/2 and C/EBPα (murine Ser21 and Thr222/226) upon LPS stimulation. While NF-κB and JNK phosphorylation were comparably increased in DMSO- and YPPP-CD11c⁺ cells, Erk1/2 phosphorylation was decreased in YPPP-CD11c⁺ cells ( Figure 2B , lanes 4, 5, 6, 7, 8 and 9). Surprisingly, pSer21 C/EBPα was decreased in YPPP-CD11c⁺ cells, whereas C/EBPα (Ser21) was phosphorylated in DMSO-CD11c⁺ cells, even after LPS stimulation ( Figure 2B , lanes 1 and 3, and Figure 2C ). In contrast, Thr222/226 C/EBPα was equally phosphorylated in both DMSO- and YPPP-CD11c⁺ cells ( Figure 2B , lane 2). IκBα and IκBβ levels were equivalently decreased in DMSO- or YPPP-CD11c⁺ cells ( Figure 2B , lanes 10 and 11). These results suggest that IL-12 production is enhanced by suppressing C/EBPα activation upon LPS stimulation in YPPP-CD11c⁺ cells.

The ability of DCV to promote T cell proliferation is a critical function for anti-tumor activity. Therefore, we examined the T cell proliferation activities of DMSO- or YPPP-CD11c⁺ cells in a mixed lymphocyte reaction (MLR). Cell tracker Cytotell green labeled Thy1⁺ splenic CD4⁺ or CD8⁺ T cells from OT-II or OT-I mice were co-cultured with DMSO- or YPPP-CD11c⁺ cells from B6-WT mice for 3 or 5 days in vitro. The proliferation was monitored using CCK-8 and flow cytometry. As shown in Figure 3A , we observed a significantly increased number of proliferating cells in the group in which YPPP-CD11c⁺ cells, as determined by the CCK-8 proliferation assay. Furthermore, flow cytometric (FCM) analysis revealed that a significant increase of the population of Cytotell green^lowCD25⁺CD8⁺ T cells, Cytotell green^lowCD69⁺CD8⁺ T cells, and Cytotell green^lowCD62L^lowCD8⁺ T cells at 1: 3 CD11c⁺/T cell ratio in the presence of YPPP-CD11c⁺ cells ( Figures 3B-G ).

Next, we examined whether YPPP-CD11c⁺ cells can efficiently expand splenic CD4⁺ T cells isolated from OT-II mice. As shown in Supplementary Figure 4A , we found a significantly increased number of proliferating cells in the group in which YPPP-CD11c⁺ cells, as determined by the CCK-8 assay. FCM analysis revealed that an increased population of Cytotell green^lowCD25⁺CD4⁺ T cells and Cytotell green^lowCD69⁺CD4⁺ T cells in co-culture groups with YPPP-CD11c⁺ cells ( Supplementary Figures 4B-E ). Consistent with previous results, the MLR assay using BALB/c T cells also demonstrated enhancement of T-cell proliferation and activation in the presence of YPPP-CD11c⁺ cells ( Supplementary Figures 5A-E ). These results indicate that the YPPP-CD11c⁺ cells are more efficient than DMSO-CD11c⁺ cells in priming and expanding T cells.

Finally, we compared IL-12p40 production and T cell activation ability between CD11c ⁺ I-A/I-E^high cells and CD11c⁺I-A/I-E^int cells treated with DMSO/YPPP. The YPPP-treated CD11c⁺I-A/I-E^high cells produced more IL-12p40 than the other cells upon LPS or CpG stimulation ( Figure 4A ). Additionally, CD11c ⁺ I-A/I-E^high cells demonstrated the highest CD8⁺ T cell proliferation and IFN-γ⁺ cell-inducing-abilities compared with other cells ( Figures 4B-E ).

To elucidate the mechanism of enhanced IL-12 production and T cell priming activity in YPPP-CD11c⁺ cells, we examined the effect of YPPP on the differentiation of DCs derived from mouse BM cells cultured with GM-CSF. We focused on the morphological analysis of CD11c⁺I-A/I-E^high cells. FCM analysis showed that an increased population of the CD11c⁺I-A/I-E^high cells in the YPPP treatment group and a decreased population of CD11c⁺I-A/I-E^int cells, representing macrophages ( Figure 5A ; Supplementary Figures 6A, B ). Diff-Quick staining did show marked differences between DMSO- and YPPP-CD11c⁺I-A/I-E^high cells ( Figure 5A ). There were no significant differences in the number of terminally differentiated cells with or without YPPP, and responsiveness to LPS stimulation ( Figure 5B ). Furthermore, FCM analysis of these DC surface makers revealed up-regulation of CD11c, CCR7, and CD80, and down-regulation of CD11b in the YPPP-treated CD11c⁺I-A/I-E^high cells ( Figure 5C ). Additionally, both DMSO-treated and YPPP-treated DCs did not express cDC1 marker XCR1 ( Supplementary Figure 7 ). These changes align with the general characteristics of GM-DCs (33).

To compare in vitro DC induced by YPPP, we performed RNA sequencing (RNA-Seq) of CD11c⁺I-A/I-E^high cells and CD11c⁺I-A/I-E^int cells differentiated under DMSO or YPPP conditions for 0, 3, and 6 days ( Figure 6A ). As a result, approximately 6,030, 5,620, 5,626 and 5,662 differentially expressed (DE) genes were identified between DMSO-CD11c⁺I-A/I-E^high cells and YPPP-CD11c⁺I-A/I-E^high cells on day 3, between DMSO-CD11c⁺I-A/I-E^int cells and YPPP-CD11c⁺ I-A/I-E^int cells on day 3, between DMSO-CD11c⁺I-A/I-E^high cells and YPPP-CD11c⁺I-A/I-E^high cells on day 6, and between DMSO-CD11c⁺I-A/I-E^int cells and YPPP-CD11c⁺ I-A/I-E^int cells on day 6 ( Figure 6B ; Supplementary Figures 8A-D ). Principal component analysis (PCA) based on differentially expressed genes clearly separated CD11c⁺I-A/I-E^high cells cultured with or without YPPP based on both location and genotype ( Figure 6C ). The outcomes of ChIP Enrichment Analysis (ChEA) revealed a significant enrichment of different expressed gene set, and the top 15 enriched gene set up- or down- regulated by YPPP are shown in Supplementary Figure 8 . And so some of genes known to be important for DC differentiation were shown ( Supplementary Figure 8E ). The outcomes of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis revealed a significant enrichment of different expressed gene set, and the top 10 enriched gene set up-regulated by YPPP are shown in Figure 6D . Within this gene set, Peroxisome Proliferator-Activated Receptor (PPAR) signaling pathway genes exhibited statistically significant upregulation in YPPP-treated CD11c⁺I-A/I-E^high cells when compared to DMSO-treated CD11c⁺I-A/I-E^high cells. The MA plot showed up-regulated several genes associated with PPARγ, and transforming growth factor (TGF)-β ( Figures 6E, F ). Consistent with RNA-seq analysis, immunoblotting demonstrated higher expression level of fatty acid binding protein (FABP) 4, FABP5 and PPARγ in YPPP-CD11c⁺ cells compared to DMSO-CD11c⁺ cells ( Figure 6G ). Previous reports have shown that FABP family plays an important role in DC function and T cell priming (37, 38). These results provide further insight in our data that the YPPP-treated CD11c⁺I-A/I-E^high cells produced more IL-12p40 than DMSO-treated CD11c⁺I-A/I-E^high cells upon LPS and CpG stimulation ( Figure 4A ). Additionally, no significant differences were observed in the expression of the major cDC1-related genes (Xcr1, Irf8, and Batf3) between DMSO- and YPPP-treated cells ( Supplementary Figure 8E ). These results suggested that YPPP instructs the direction of DC differentiation via the activating of PPARγ and TGF-β associated genes.

To investigate the in vivo anti-tumor potential of the DCV produced by adding YPPP, we evaluated its therapeutic efficacy against established E.G7 and B16 melanoma tumor models ( Figure 7A ). The mice were injected with E.G7 tumor cells (Day 0). On day 7, 10 and 13 post tumor inoculation, the mice were vaccinated with DMSO- or YPPP-CD11c⁺ cells cultured with GM-CSF. As shown in Figures 7B, C , a notable decrease in tumor growth and an improvement in survival is apparent within the DCV-administered group (both DMSO and YPPP treated-group), in comparison to the DCV-non-administered group. In addition, DMSO-CD11c⁺ cells (DMSO-DCV) were unable to suppress tumor growth completely, whereas YPPP-CD11c⁺ cells (YPPP-DCV) significantly reduced tumor growth, although they could not eradicate the tumor. These results indicated that YPPP-DCV has superior anti-tumor effects than DMSO-DCV ( Figures 7B-E ).

The combination of anti-PD-1, an immune checkpoint inhibitor, and other therapies has shown remarkable efficacy (39). Thus, we examined the combined anti-tumor activities of DCV with anti-PD-1 in the same tumor model as mentioned above. As shown in Figures 8A-C , tumor growth and mortality of the group treated with YPPP-DCV significantly reduced in the E.G7 tumor model (Log-rank test, p=0.0164), although these activities were not observed in the B16 melanoma model ( Figures 8D, E ). These results suggest that YPPP-DCV enhances the efficacy of anti-PD-1 therapy in some cancer models.