Invadopodia play a role in prostate cancer progression

Abstract

Background: Invadopodia, actin-rich structures that release metallo-proteases at the interface with extra-cellular matrix, in a punctate manner are thought to be important drivers of tumour invasion. Invadopodia formation has been observed in-vitro and in-vivo in numerous metastatic cell lines derived from multiple tumour types. However, prostate cancer cell lines have not been routinely reported to generate invadopodia and the few instances have always required external stimulation.

Methods: In this study, the invasive potential of primary prostate adenocarcinoma cell lines, which have never been fully characterised before, was investigated both in-vitro invadopodia assays and in-vivo zebrafish dissemination assay. Subsequently, circulating tumour cells from prostate cancer patients were isolated and tested in the invadopodia assay.

Results: Retention of E-cadherin and N-cadherin expression indicated a transitional state of EMT progression, consistent with the idea of partial EMT that has been frequently observed in aggressive prostate cancer. All cell lines tested were capable of spontaneous invadopodia formation and possess a significant degradative ability in-vitro under basal conditions. These cell lines were invasive in-vivo and produced visible metastasis in the zebrafish dissemination assay. Importantly we have proceeded to demonstrate that circulating tumour cells isolated from prostate cancer patients exhibit invadopodia-like structures and degrade matrix with visible puncta. This work supports a role for invadopodia activity as one of the mechanisms of dissemination employed by prostate cancer cells.

Conclusion: The combination of studies presented here provide clear evidence that invadopodia activity can play a role in prostate cancer progression.

Keywords: Circulating tumour cells; Invadopodia; Prostate cancer.

Fig. 1

Differential cadherin expression levels. A cancer cell lines stained for E-cadherin (green) and F-actin (red). Scale bar = 10 μm. B Percentage of cells was forming colonies (cells forming adhesions with at least two neighbour cells) exhibiting E-cadherin signal at cell:cell junctions. Note that whilst in the 1542 cell colony all cells form at least one E-cadherin positive junction (arrowheads indicate examples of E-cadherin positive cell: cell junctions) the 1535 cell colony is less compact overall and contains cells that are not forming any E-cadherin positive cell: cell junctions (indicated by *) C Expression level of E-cadherin. D Quantification of E-cadherin expression by densitometric analysis corrected for the loading control (GAPDH) E Cell growth curve repeated over four consecutive days. F Expression level of N-cadherin G Quantification N-cadherin expression by densitometric analysis corrected for the loading control (GAPDH). Membranes were cut prior to hybridisation cropped Figure C and F are taken from three replicate analysis. Statistical significance was calculated with One-way Anova followed by Tukey’s multiple comparisons test, *p < 0.05, **p < 0.01, ***p < 0.005. All data is representative of 3 independent experiments mean ± SEM

Fig. 2

Primary prostate adenocarcinoma cell lines form invadopodia. A Cells were seeded on Cy3-conjugated gelatin for 24 h and stained for F-actin and Cortactin. B co-localisation of cortactin, F-actin and gelatin degradation. C Gelatin degradation. D percentage of degraded area underneath total cell area. n = 90 cells. All Data presented represent 3 independent experiments, mean ± S.E.M. Significance was calculated with One-way Anova followed by Tukey’s multiple comparisons test, **p < 0.005. Scale bar = 10 μm

Fig. 3

Primary prostate adenocarcinoma cell lines disseminate in-vivo. A xenograft formation in zebrafish yolk-sac invasion assay including an example of a non xenografting cell line NIH 3T3. B metastatic dissemination in the tail 3 days post injection, indicated by arrows. C Quantification of the percentage of embryos exhibiting metastasis in the tail region. Data are representative of three independent experiment, with at least 15 embryos screened for metastasis. Data are presented as mean ± S.E.M. Significance was calculated with One-way Anova followed by Tukey’s multiple comparisons test. Scale bar = 200 µm

Fig. 4

Circulating prostate tumour cells exhibit invadopodia activity. A circulating tumour cells (top) and hematopoietic cells (bottom) stained for CD45 and actin, imaged from the same field of view. B number of CD45-negative cells isolated from each blood sample. C percentage of CD45-negative cells displaying puncta and/or degradation. D PCa CTCs subjected to invadopodia assay and stained for F-Actin and Cortactin after 24 h incubation. Cells exhibited localised matrix degradation overlapping with actin puncta (white arrow) and cortactin puncta co-localising with actin (yellow arrow). E percentage of CTCs exhibiting matrix degradation properties. Samples without any visible degradation underneath CTCs were excluded. F percentage of CTCs exhibiting invadopodia formation in samples found positive for matrix degradation. Samples without any visible invadopodia activity were excluded. G percentage of samples with matrix degradation and invadopodia. Data are presented as Mean ± SD. Scale bar = 10 µm