Fig 2: PITPNM3 mediates CCL18-induced fibroblast activation via NF-?B signaling.A Representative western blotting for a-SMA, FAP, CD10 and GPR77 in untreated (UT) or CCL18-treated NBFs transduced with shRNA against GFP or PITPNM3 (n = 3). B Representative images(up) and diameters(down) of collagen gel contraction assay for NBFs with indicated treatment (n = 3). C Representative western blotting for total expression and phosphorylation of IKK, I?B and SMAD2/3 in NBFs with indicated treatment (n = 3). D (left)Representative images of p65 immunofluorescent staining and (right) quantification of p65 nuclear translocation in NBFs with indicated treatment (n = 3). Scale bars (white line), 50 µm; Scale bars (yellow line), 5 µm. E Localizations of p65 to the promoters of CD10 and GPR77 genes in indicated fibroblasts were analyzed by ChIP assay using anti-p65 Ab or control IgG. F, G QRT–PCR (F) and representative western blotting (G) for CD10 and GPR77 expression in untreated (UT) or CCL18-activated NBFs treated with DMSO, JSH-23 or BAY117082 (n = 3). H, I ELISA for IL-8 (H) and IL-6 (I) levels in the conditioned medium of untreated, CCL18 or TGF-ß treated NBFs in the presence of JSH-23, BAY117082 or transduced with GFP or p65 shRNAs (n = 3). JSH-23, an NF-?B inhibitor. BAY117082, an IKK inhibitor. Data expressed as mean ± SEM, ***P < 0.001 compared with NBFs co-cultured with untreated CCL18-activated NBFs (B, F, H, I) or TGF-ß treated NBFs (E) by two-tailed one-way ANOVA with Dunnett’s multiple comparison test.

Fig 3: Breast cancers with different chemotherapeutic responses exhibit conspicuously distinct cytokine profiles in their tumor microenvironment.A Representative images of H&E staining and CD10/GPR77 immunofluorescent staining in serial sections of the pre-treatment breast cancer biopsies of chemosensitive (n = 3) and chemoresistant (n = 3) patients. B, C Western blotting (B) for a-SMA, FAP, CD10, and GPR77 and flow cytometric analysis (C) for CD10 and GPR77 in primary normal breast fibroblasts (NBFs) treated with pre-treatment tumor conditional medium (CM). D Cytokine arrays of pre-treatment tumor CM, squares indicate the cytokines with significant changes. E Signal intensity of indicated cytokines in the cytokine arrays and their relative fold change between the chemoresistant (n = 3) and chemosensitive tumors (n = 3) F Flow cytometric analysis for CD10 and GPR77 in NBFs treated with chemosensitive or chemoresistant tumor CM added without or with neutralizing antibodies against IL6(aIL6), IL8(aIL8) or CCL18(aCCL18). The patients with complete remission (CR) or partial remission (PR) were classified as chemosensitive, while those with stable disease (SD) or progressive disease (PD) were chemoresistant.

Fig 4: CCL18 produced by TAMs mediates the chemoresistance-inducing phenotype polarization in NBFs.A Representative images(left) and diameters(right) of collagen gel contraction assay for NBFs with or without CCL18 and TGF-ß treatment was determined by three-dimensional collagen matrices (n = 3). B, C QRT–PCR(B) and western blotting(C) for a-SMA, FAP, CD10 and GPR77 expression in NBFs with or without CCL18 and TGF-ß treatment (n = 3). D Representative flow cytometric analysis for CD10 and GPR77 in NBFs with indicated treatment (n = 3). E Representative western blotting for a-SMA, FAP, CD10, and GPR77 in NBFs cultured alone (UT) or co-cultured with PBMC or TAMs isolated from different subtypes of breast cancer (n = 3). F Western blotting for a-SMA, FAP, CD10 and GPR77 in NBFs cultured alone (UT) or co-cultured with PBMC or TAMs pretreated with control IgG or neutralizing antibodies against CCL18(aCCL18) or TGF-ß(aTGF-ß) (n = 3). G Representative images(left) and diameters(right) of collagen gel contraction assay for NBFs treated as in F (n = 3). Data expressed as mean ± SEM, **P < 0.01, ***P < 0.001 compared with untreated NBFs (A, B) or NBFs co-cultured with untreated TAMs (G) by two-tailed one-way ANOVA with Dunnett’s multiple comparison test.

Fig 5: The intratumoral accumulation of CCL18+ tumor-associated macrophages is associated with the abundance of CD10+GPR77+ CAFs and chemoresistance.A Representative images of immunohistochemical staining for CCL18 and EPCAM in pre-treatment breast cancer biopsies (n = 259, n = 156 in chemosensitive group and n = 103 in chemoresistant group). Scale bars, 500 µm. B The counts of CCL18+ cells in pre-treatment breast cancer biopsies with different chemotherapeutic responses. Mean ± SEM, ***P < 0.001 by two-tailed Student’s t test (each dot represents the mean of the counts of at least five random fields from each tumor section). C Representative images of immunofluorescent staining of CD68, CCL18, CD10, GPR77 and a-SMA in serial sections of breast cancer samples with high or low CCL18 expression. Scale bars, 50 µm. The arrowheads denote the area of higher-magnification images shown in the top-right corner. D The correlation between the number of CD10+GPR77+ CAFs and CCL18+ TAMs in breast cancer samples. The Pearson’s correlation coefficient r value and P values were determined by two-tailed Pearson correlation coefficient test (n = 259). E (Upper) EPCAM expression on t-SNE map of combined scRNA-seq transcriptomes of total breast tissue cells from GEO: GSE161529 and Pan-cancer TME Blueprint, and (lower) the EPCAM-negative cells were reclustered to eight microenvironment populations according to their representative markers. F, G Expression plots of (F) fibroblast markers (PDGFRA, PDGFRB, ACTA2, PDPN, FAP and CD10) and (G) macrophage markers (CD14, CD68, CD163, and CCL18) were exhibited on t-SNE layout. H The correlation between the percentage of CD10 positive cells in fibroblasts with the percentage of CCL18 positive cells in macrophages in breast cancer samples based on scRNA-seq transcriptomes. The Pearson’s correlation coefficient r value and P values were determined by two-tailed Pearson correlation coefficient test (n = 46).