Fig 2: Characterization of resistant tumor cells against combination therapy.A The UMAP plot shows five epithelial subtypes in distinct colors, with bar graphs indicating their quantities. B The UMAP plot (left) shows the distributions of control, resistant, and sensitive groups, with cell proportions (center) and subtype compositions (right). C Heatmap of top 10 signature genes for three groups. D CellChat circle plot shows differences in intercellular communication number and strength between two groups. E The UMAP plot shows 11 immune cell subtypes, with bar graphs indicating their numbers. F The UMAP plot shows 11 immune subtypes across two groups. G CellChat analysis compares epithelial-immune information flow in Resi and Sens groups, with relative (left) and absolute (right) flow shown. H Scatter plot of ANNEXIN pathway signaling in cell clusters from CellChat analysis. Bubble plots from CellChat (I) and CellPhoneDB (J) show epithelial subtype communication with Neu1 receptors and ligands. P values were calculated using a permutation test, assessing the significance of cell-cell communication by comparing observed mean expression with a null distribution generated by random permutations (I). P values were calculated using a permutation test, which generates a null distribution by randomly shuffling cell labels to determine the specificity of interactions (J). P values are one-sided and exact. K Violin plot (left) and FeaturePlot (right) show Anxa1 expression in epithelial cells across groups. The box plot elements are defined as described in the Statistics and Reproducibility section. Each group includes cells from 3 mice. P values were calculated one-way ANOVA. L Violin plot showing Anxa1 expression across epithelial cell subtypes. The box plot elements are defined as described in the Statistics and Reproducibility section. Each group includes cells from 9 mice. P values were calculated one-way ANOVA. M Representative IHC staining and score of Anxa1 in HNSCC. n = 6 control, n = 6 resistant, n = 8 sensitive mice. Scale bar = 65 μm. Data shown as as mean ± SD. P values were calculated one-way ANOVA with Tukey’s multiple comparisons test. Source data and exact p values are provided as a Source Data file.

Fig 3: Anxa1-Fpr1 axis regulated the mitochondrial fission of neutrophils.A GSEA was performed in the Resi; Fpr1cKO-Con and Resi-Fpr1cKO groups to compute an enrichment score (ES) for gene set overrepresentation in a ranked mRNA-seq list. A positive ES indicates top enrichment, while a negative ES indicates bottom enrichment. B The ridgeline plot displays GSEA core gene fold changes (positive: upregulated; negative: downregulated). P values were calculated using a permutation test, evaluating the significance of pathway enrichment by comparing the observed enrichment score (ES) with a null distribution generated through random permutations of gene sets. P values are one-sided and adjusted for multiple comparisons using the Benjamini-Hochberg method. C GO-BP enrichment of up/downregulated genes in Resi; Fpr1cKO-Con VS Resi-Fpr1cKO, annotated with p values (yellow numbers). P values were calculated using the hypergeometric test to evaluate the significance of gene set enrichment. The test was one-sided, and P values were adjusted for multiple comparisons using the Benjamini-Hochberg method. D Representative electron microscopy images of mitochondria from four groups. Scale bar, 400 nm. The experiment was independently repeated three times. Knockdown group selected using shFPR1-2. E Volcano plot of Resi; Fpr1cKO-Con VS Resi; Fpr1cKO mRNA-seq data highlights Bnip3(red) among differentially expressed genes (threshold: p < 0.05, |Log2FC| > 1). n = 3 mice in each group. P values were calculated using the Wald test to evaluate the significance. The test was two-sided. P values were adjusted for multiple comparisons using the Benjamini-Hochberg method. F Representative immunofluorescence staining of HL60 cells from four groups. Knockdown group selected using shFPR1-2. Cells were stained with anti-BNIP3 (green) and DAPI (blue) for nuclei. Scale bar = 15 μm. Right: percentages of BNIP3+ cells by area and number. The experiment was independently repeated three times. Data shown as mean ± SD. P values were calculated one-way ANOVA with Tukey’s multiple comparisons test. G BNIP3 protein levels (WB) in four groups, with α-Tubulin as loading control. H WB analysis of Cytochrome c in mitochondria-free cytoplasmic proteins from four groups, with α-Tubulin as loading control. Source data and exact p values are provided as a Source Data file.

Fig 4: Sox9 mediates resistance in HNSCC by acting as a transcription factor for Anxa1.A Venn diagram of genes identified by FindMarkers and SCENIC analyses. B Violin plot (left) shows Sox9 expression across groups, and FeaturePlot (right) depicts Sox9 expression in epithelial cells. The box plot elements are defined as described in the Statistics and Reproducibility section. Each group includes cells from 3 mice. P values were calculated one-way ANOVA. C Violin plot showing Sox9 expression across epithelial cell subtypes. The box plot elements are defined as described in the Statistics and Reproducibility section. Each group includes cells from 9 mice. P values were calculated one-way ANOVA. D Representative IHC staining and score of Sox9 in HNSCC. n = 6 control, n = 6 resistant, n = 8 sensitive mice. Scale bar = 65 μm. Data shown as mean ± SD. P values were calculated one-way ANOVA with Tukey’s multiple comparisons test. E ChIP of genomic DNA from resistant mouse HNSCC tissues was performed with a Sox9-specific antibody. Left: gel image, right: qPCR analysis of Anxa1 promoter ChIP signal relative to IgG group. Two primer pairs were used. The experiment was independently repeated three times. Data shown as mean ± SD. P values are presented by two-tailed unpaired Student’s t test. F Luciferase reporter assay assessed Anxa1 promoter activity in Moc1 cells. A schematic diagram shows cloning the Anxa1 promoter into the pZX001 vector to generate pGL3 luciferase plasmid. Binding motif information was retrieved from footprintDB database. G The relative activity of the Anxa1 promoter was detected by luciferase assay across the three groups. The experiment was independently repeated three times. Data shown as mean ± SD. P values were calculated one-way ANOVA with Tukey’s multiple comparisons test. H Representative immunofluorescence colocalization of Anxa1 (red) and Sox9 (green) in Resi mouse HNSCC tissues. Nuclei counterstained with DAPI (blue). Scale bar = 20 μm. Source data and exact p values are provided as a Source Data file.

Fig 5: Met deprivation significantly reduces ANXA1 expression in GBM and inhibits macrophage M2 polarization. A qRT-PCR analysis of ANXA1 expression in LN229 and U251 after treatment with MD; B Immunofluorescence image of LN229 cells ANXA1 expression after cultured with NC or MD medium for 48 h. The scale bar is 100 μm; C Protein translation level of ANXA1 in LN229 and U251 cells. Western blot analysis was performed 48 h after treatment with a Met-deprived medium and a standard medium as control. GAPDH was used as an internal control; D LN229 and U251 were treated in NC or MD medium for 48 h. Then, supernatants were obtained to evaluate the secretion of ANXA1 by ELISA. E Schematic of GBM cells indirect co-culture with macrophages induced from THP-1 cells; F qRT-PCR analysis of M1-(iNOS, TNFα) and M2 (CD163, IL10) marker gene expression in the indirect co-culture system; G Macrophages were treated as described in panel A. Macrophages were cultured for 48 h in the indirect co-culture systems. Then, supernatants were obtained to evaluate the secretion of IL10 and TNFα by ELISA; H Changes in macrophage polarization markers CD86 (M1) and CD206 (M2) were analyzed under three CM: Met-containing (NC) medium, Met-deprived (MD) medium, and MD medium supplemented with ANXA1 protein (MD + ANXA1) (100 ng/ml). Statistical significance was determined using the unpaired Student’s t-test (A–D, F–G) or one-way ANOVA followed by Tukey’s multiple comparisons test (H). *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001