Fig 2: miR-15a carried by adMSCs-Evs inhibits KDM4B expression in CRC cells. After overexpression of miR-15a in adMSCs, Evs were extracted. (A) miR-15a expression carried in Evs determined by RT-qPCR. (B) The uptake of PKH-26-labeled Evs by HCT-116 and LoVo cells detected by fluorescence microscopy. (C) HCT-116 and LoVo cells were treated with Evs at different concentrations and cell viability was determined by CTG kit. (D) The mRNA expression of KDM4B in CRC cells measured by RT-qPCR. (E) The protein expression of KDM4B in CRC cells measured by western blot. (F) Cell proliferation assessed by EdU staining. (G) The apoptotic levels of HCT-116 and LoVo cells detected by flow cytometry. (H) Expression of pro-apoptotic protein Cleaved PARP and Bax and anti-apoptotic protein Bcl-2 determined by western blot. Subsequently, 100 μL of Evs was injected every 7 days after inoculation of HCT-116 and LoVo cells in vivo. (I) Growth curve of tumor. (J) Tumor weight of mice injected with PBS or Evs. (K) Immunohistochemical staining of KI67 protein in tumor tissues. (L) The protein expression of KDM4B, HOXC4, and PD-L1 in tumor tissues examined by western blot. In (A–H), each assessment was done in triplicate with three–time repetition to ensure minimum deviation, and in (I–L), each group contained five mice. The data in figures were all measurement data and presented as the mean ± standard deviation. Comparisons between multiple groups were analyzed by two-way ANOVA, followed by Tukey’s multiple comparison test. The experiment was repeated three times. *p < 0.05, **p < 0.01.

Fig 3: KDM4B induces PD-L1 expression by H3K27me3/HOXC4 axis. (A) The protein expression of KDM4B, H3K27me3 and HOXC4 in HCT-116 and LoVo cells after IFN-γ treatment. (B) HOXC4 expression in tumor (n = 275) and normal (n = 349) tissues in the TCGA-COAD database. We transfected shRNAs targeting KDM4B in IFN-γ treated cells. (C) KDM4B expression in CRC cells evaluated by RT-qPCR and western blot analysis. (D) The proportion of cells expressing PD-L1 by flow cytometry. (E) The protein expression of PD-L1 and Galentin-9 in CRC cells determined by western blot. (F) Measurement of specific lysis rate using an LDH kit. (G) HOXC4 binding sites to PD-L1 promoters and conserved binding sites predicted by Jaspar website. (H) The binding relationship between HOXC4 and PD-L1 promoter verified by ChIP-qPCR assays. (I) HOXC4 mRNA expression in CRC cells evaluated by RT-qPCR. (J) The proportion of cells expressing PD-L1 by flow cytometry. (K) The mRNA expression of PD-L1 and Galentin-9 in CRC cells determined by RT-qPCR. (L) The protein expression of PD-L1 and Galentin-9 in CRC cells evaluated by western blot. (M) Measurement of specific lysis rate using an LDH kit. The data in figures were all measurement data and presented as the mean ± standard deviation. Each assessment was done in triplicate with three–time repetition to ensure minimum deviation. Comparisons between multiple groups were analyzed by two-way ANOVA, followed by Tukey’s multiple comparison test. The experiment was repeated three times. *p < 0.05, **p < 0.01.

Fig 4: Overexpression of KDM4B or HOXC4 blocks the effects of Evs on immune evasion of CRC cells. IFN-γ-pre-treated HCT-116 and LoVo cells were transfected with overexpression vectors of KDM4B and HOXC4, respectively, and then co-cultured with Evs. (A) The expression of KDM4B in HCT-116 cells and the expression of HOXC4 in LoVo cells determined by western blot. (B) The proportion of cells expressing PD-L1 by flow cytometry. (C) The mRNA expression of PD-L1 and Galentin-9 in CRC cells determined by RT-qPCR. (D) The protein expression of PD-L1 and Galentin-9 in CRC cells determined by western blot. (E) Measurement of specific lysis rate using an LDH kit. (F) miR-15a carried by adMSCs-Evs maintained H3K27me3 expression and inhibited HOXC4 transcription by targeting KDM4B in CRC cells, thereby reducing PD-L1 expression and finally inhibiting immune evasion in CRC cells. The data in figures were all measurement data and presented as the mean ± standard deviation. Each assessment was done in triplicate with three–time repetition to ensure minimum deviation. Comparisons between multiple groups were analyzed by two-way ANOVA, followed by Tukey’s multiple comparison test. The experiment was repeated three times. *p < 0.05, **p < 0.01.