CXCL1-LCN2 paracrine axis promotes progression of prostate cancer via the Src activation and epithelial-mesenchymal transition

Abstract

Background: Mechanisms driving the progression of castration-resistant prostate cancer are believed to relate substantially to the tumor microenvironment. However, the cross-talks between tumor epithelial cell, stromal cells, and immune cells are yet to be fully elucidated. The present study aims to determine the role of chemokine and neutrophil derived cytokine paracrine axis in mediating the interaction between tumor cells, stromal myofibroblasts, and neutrophils in the tumor microenvironment of prostate cancer.

Methods: To identify myofibroblasts and neutrophil derived specific proteins affecting progression of prostate cancer, bioinformatics analyses were firstly performed in independent human prostate cancer gene expression data sets from the GEO data bank. Expression of stromal myofibroblasts secretory chemokine CXCL1 and neutrophil derived cytokine LCN2 was evaluated in prostate tissues via immunohistochemistry assay. We further investigated the effect of CXCL1 and LCN2 on prostate cancer using in vivo and in vitro models, and explored the underlying signal transduction pathways.

Results: A CXCL1-LCN2 paracrine network was confirmed in prostate cancer tissue samples, which was correlated with the biochemical recurrence of prostate cancer. Of note, CXCL1-LCN2 axis activates Src signaling, triggers the epithelial-mesenchymal transition (EMT), consequently promotes the migration of prostate cancer cells, leading to enhanced tumor metastasis.

Conclusions: Our findings may provide enhanced insight into the interactions of carcinoma-stromal cells and immune cells linked to prostate cancer progression, wherein CXCL1-LCN2 axis is a key contributor to prostate cancer cells migration. These data indicate tumor microenvironment and Src signaling pathway may be potential therapeutic targets of prostate cancer treatment.

Keywords: Chemokine; Cytokine; Tumor cell migration; Tumor metastasis; Tumor microenvironment.

Fig. 1

Expression of CXCL1, CD177 and LCN2 in prostate tissues and correlation with biochemical recurrence free survival in prostate cancer cases. a Representative images (magnification 200X) of CXCL1 (a~c), CD177 (d~f) and LCN2 (g~i) expression in normal prostate tissues, cancerous lesion and para-carcinoma tissue (PCT). Red arrows: tumor foci; yellow arrows: prostate stroma. b Representative images (magnification 200X) of AMACR (red, expressed in carcinoma epithelium), CK18 (white, expressed in prostate glandular epithelium), CXCL1 (cyan) and LCN2 (orange) co-staining in normal prostate tissues, carcinoma tissue and PCT observed by PerkinElmer Tyramide Signal Amplification system. c Semi-quantitative data of CXCL1, LCN2 and CD177 expression in normal prostate (n = 19) and prostate cancer (n = 118) are shown in scatter plots. p value was determined by Mann–Whitney test, asterisks indicate *p < 0.05, ***p < 0.001. d Kaplan-Meier and log-rank test were used to evaluate CXCL1 expression alone (a) or CXCL1-LCN2 co-expression (b) and biochemical recurrence free survival. High expression level of CXCL1 (CXCL1 High vs. Negative expression p < 0.001) and high co-expression of CXCL1-LCN2 (High-High/High-Medium vs. Negative/Low/Medium expression p < 0.0001) predicted early biochemical recurrence after radical prostatectomy

Fig. 2

CXCL1 contributes to myofibroblast promoted migration in prostate cancer cells. a Representative images (magnification 40X) of cells migration in prostate cancer cells control and those co-cultured with WPMY-1 in transwell (TW) system. b CXCL1 secretion in PCa, naïve WPMY-1 and PCa CM treated WPMY-1 was determined using ELISA. This cytokine concentration was normalized with numbers of cells. c Influence of CXCL1 on prostate cancer cells proliferation was determined using SRB. DHT was used as control to treat androgen-sensitive cell line LNCaP and androgen-insensitive PC-3. d Representative cells migration images (magnification 40X) of PC-3 and DU145 treated with different concentrations of CXCL1. e Representative images (magnification 40X) of cells migration in prostate cancer cells control and those treated with CXCL1, with or without supplement of CXCR1/2 blockage using specific antibodies or inhibitor. f Influence of CXCL1 on benign prostate epithelial cell RWPE-1 proliferation was determined using SRB. g Representative cells migration images (magnification 40X) of RWPE-1 treated with CXCL1. Data are represented as mean ± SD from a representative triplicate experiment. Histograms shows quantitative results and p value was determined by ANOVA or t test, asterisks indicate *p < 0.05, **p < 0.01

Fig. 3

Prostate cancer cell-neutrophil interaction in transwell systems and the influence of LCN2 on tumor cells. a Representative cells migration images (magnification 40X) of PC-3 and DU145 co-cultured with neutrophil. b LCN2 secretion in neutrophil, tumor cell lines and co-cultured system were determined using ELISA. The LCN2 level of neutrophil control was normalized as 1. c MMP9/LCN2 secretion in tumor cell lines and respective co-cultured system with neutrophil were measured using ELISA. d Influence of LCN2 on prostate cancer cells proliferation was determined using SRB. DHT was used as control to treat androgen-sensitive cell line LNCaP and androgen-insensitive PC-3. e Representative cells migration images (magnification 40X) of PC-3 and DU145 treated with different doses of LCN2. f DU145 was treated with low concentration of CXCL1/LCN2 separately or in combination for 96 h. g Influence of LCN2 on benign prostate epithelial cell RWPE-1 proliferation was determined using SRB. h Cells migration of RWPE-1 treated with LCN2. Representative images (magnification 40X) and quantitative data were shown. Data are represented as mean ± SD of triplicates from a representative experiment of n = 3. Histograms shows quantitative results and p value was determined by ANOVA or t test, asterisks indicate *p < 0.05, **p < 0.01

Fig. 4

CXCL1&LCN2 facilitate prostate cancer cells metastasis in mice model. a Lung tissues taken from 29 mice inoculated with control, CXCL1 or LCN2 treated DU145cells via tail vein injection. b Representative lung samples of three groups demonstrating tumor metastasis nodules indicated with blue arrows. c Comparison of the numbers of metastatic nodules in each mouse between CXCL1 or LCN2 treated group with control. p value was determined by ANOVA test

Fig. 5

Src family kinases activation in CXCL1 or LCN2 treated prostate cancer cells. a Activation of Src-FAK-Paxillin pathway was indicated by phosphorylation of these kinases in CXCL1 or LCN2 treated PC-3 and DU145 in time-dependent manner. b Phosphorylation of Src family members (Src, Fyn, Lck, Lyn, c-Yes) in CXCL1&LCN2 treated prostate cancer cells was measured using western-blotting. Representative blots from at least twice independent experiments are shown

Fig. 6

Src family kinases activation in CXCL1 or LCN2 treated prostate cancer cells mediates cancer cells migration. a Representative images (magnification 40X) of prostate cancer cells migration after treatment of CXCL1 or LCN2 with or without SFKs inhibitor PP2. Histograms shows quantitative results from three times experiments and p value was determined by Student’s t test. b Phosphorylation of Src-FAK-Paxillin in DU145 after treatment of CXCL1 or LCN2 with or without supplement of PP2 was detected using western blotting. c Representative cells migration images (magnification 40X) of DU145 stimulated with CXCL1 or LCN2 in addition of β-Catenin inhibitor XAV939. d Expression of EMT markers in DU145 after CXCL1/LCN2 treatment was determined using western blotting

Fig. 7

Schematic diagram of CXCL1-LCN2 paracrine axis in prostate cancer microenvironment. CXCL1 secreted from myofibroblast directly enhances prostate cancer cell migration (①); CXCL1 recruits neutrophils to prostate cancerous lesion (②); LCN secreted from neutrophil further promotes tumor cells migration (③); Activation of Src family kinases contributes to highly aggressive phenotype of prostate cancer cells stimulated by CXCL1 and LCN2 (④)