Targeting the KLF5/PI3K/AKT axis as a therapeutic strategy to overcome neoadjuvant chemoresistance in colorectal cancer

Figures

• 

  Figure 1. Single-cell atlas of CRC patients undergoing NAC treatment. (A) UMAP plot of broad cell clusters from all samples. (B) Boxplot showing the relative frequencies of different cell clusters stratified by CRC patient treatment strategies. (C) Violin plots illustrating the expression patterns of marker genes across the identified cell clusters. (D) Heatmap of DEGs identified among the nine major cell clusters, generated using the “FindMarkers” function in Seurat V5 (parameters: min.pct = 0.15, logfc.threshold = 0.2). (E) Radar chart comparing the proportion of cell clusters between NAC and Non-NAC groups.

• 

  Figure 2. Identification and characterization of NAC-resistant cell clusters. (A, B) Identification of malignant epithelial cells using the InferCNV algorithm. (A) InferCNV analysis reveals CNV profiles across all chromosomes, and (B) cells with high CNV scores were identified as malignant compared to immune cell types. (C) UAMP plot of malignant epithelial cell clusters. (D) Boxplot illustrating the proportions of tumor cell clusters in NAC and non-NAC groups. (E) GO term enrichment analysis reveals key biological processes in Epi_5. (F) KEGG pathway analysis identifies activated signaling pathways in Epi_5.

• 

  Figure 3. Cellular communication in the TME. (A) Communication counts and weight among various cell clusters in the TME. (B) Communication weights between Epi_5 and Epi_others and their interactions with other cell clusters in the TME. (C) Signaling pathways mediating communication within and between Epi_5 and Epi_others.

• 

  Figure 4. KLF5 regulon specifically activated in the Epi_5 cluster. (A) Identification of spatially activated transcription factors in Epi_5 using the SCENIC algorithm. (B) Interaction and regulatory network of transcription factors identified in Epi_5. (C, D) UMAP visualization (C) and Violin plot showing the AUCell scores of KLF5 regulon activity in various cell clusters. (E) Regulatory network between KLF5 and its target genes, including cell migration, Wnt signaling, Hippo signaling, and PI3K/AKT pathways.

• 

  Figure 5. Exploring therapeutic strategies for NAC-resistant patients driven by the KLF5 regulon. (A, B) Univariate and multivariate Cox regression analyses of the KLF5 regulon, respectively. (C) Kaplan-Meier survival analysis of risk score groups, with patients divided into high- and low-risk groups based on the median risk score. (D) Univariate Cox regression analyses of clinical factors. (E) GSEA analysis of upregulated genes in the high-risk score group. (F) Volcano plot illustrating DEGs between high- and low-risk score groups. (G, H) Tumor purity and immune scores were calculated using the ESTIMATE algorithm. Data were shown as mean ± SD. The statistical significance was determined by the Wilcoxon test (**p < 0.01). (I) Riskscore-related drug sensitivity screening.

• 

  Figure 6. GDC-0941 enhances chemosensitivity by inhibiting the PI3K/AKT/mTOR signaling pathway. (A, B) IC50 analyses of HCT116 (A) and HCT116L (B) cells treated with Oxaliplatin, determined by CCK-8 assay. (C) Western blot analysis of total AKT and P-AKT expression levels in HCT116 and HCT116L cells. (D) Western blot analysis of AKT and P-AKT expression levels in HCT116 cells treated with GDC-0941 and ML264. (E, F) Flow cytometry analysis of apoptosis in HCT116 cells treated with different conditions. Data were shown as mean ± SD. The statistical significance was determined by the Wilcoxon test (***p < 0.001).

• 

  Figure 7. The combination of GDC-0941 and oxaliplatin enhances antitumor immune responses and suppresses lung metastasis. (A) Schematic diagram of the experimental design for animal studies. (B) KM survival analysis of the four treatment groups in C57BL/6N mice. (C) H&E staining of lung tissue from the four groups. (D) Quantification of tumor area in the lungs from H&E staining results (C). Data were shown as mean ± SD. The statistical significance was determined by one-way ANOVA (***p < 0.001, ns, not significant: p>0.05). (E) KM survival analysis of the four treatment groups in nude mice.