Elevated c-Src is linked to altered cell-matrix adhesion rather than proliferation in KM12C human colorectal cancer cells

Abstract

Elevated expression and/or activity of c-Src, the prototype of the Src family of protein tyrosine kinases, is associated with the development of human colon cancer. However, despite the known pleiotropic effects of these kinases in promoting (a) cell growth downstream of growth factor receptors, and (b) the dynamic regulation of integrin adhesions in fibroblast model systems, their precise role in epithelial cancer cells is unknown. Here we addressed whether elevated expression and activity of cellular Src alters cell proliferation and/or cell-matrix adhesion in cancer cells from the Fidler model of colorectal metastasis. Although elevated Src correlates with ability to metastasise to the liver after intrasplenic injection, we found that this was not linked to enhanced growth, either in vitro or in vivo as sub-cutaneous tumours. However, elevated Src was associated with enhanced attachment to extracellular matrix. In addition, adhesion to fibronectin, was suppressed by agents that inhibited Src activity, while enforced elevation of Src in non-metastatic cells was sufficient to stimulate adhesion to fibronectin and enhanced assembly of adhesion complexes, without influencing cell growth. Thus, we conclude that one role of elevated Src in human colon cancer cells is to modulate integrin-dependent cell-matrix attachment and formation of adhesion structures, which may, in turn, influence cell motility and integrin-dependent cellular responses.

Figure 1

(A) Immunoblots showing expression (upper panel) and activity (lower panel) of c-Src in the poorly metastatic cell line, KM12C, and in their highly metastatic derivative cell lines, KM12SM and KM12L4A. Activity was assessed by reactivity with a phospho-specific antibody raised against the region of c-Src containing tyrosine-416, the presumed site of autophosphorylation. (B) Comparison of paxillin tyrosine phosphorylation in KM12C and KM12L4A cells. Paxillin was immunoprecipitated, blotted and probed with phosphotyrosine-specific antibodies (upper panel) or anti-paxillin (lower panel).

Figure 2

(A) In vitro growth of KM12C, KM12L4A and KM12SM cells (seeded at 1×10⁵ cells in 35 mm dishes) was monitored for 14 days. (B) The ability of KM12C, KM12SM and KM12L4A cells (seeded at 5×10² cells per ml of medium containing 0.6% agar) to grow when deprived of anchorage was compared. As control, the colon adenoma cell line RGC2 that does not grow in semi-solid medium was used. Shown are representative areas on culture dishes. Over a number of experiments, there was no visible difference between the number, or size, of colonies formed by all three cell lines under anchorage-independent conditions. (C) Subcutaneous primary tumour growth of KM12C, KM12SM and KM12L4A cells was monitored by injecting 2×10⁶ cells into mice and measuring tumour dimensions at regular intervals. Tumour volumes (upper panel) and doubling times (lower panel) are shown. 4–6 mice were used in each experiment.

Figure 3

(A) Immunofluorescence staining (anti-vinculin) of KM12C and KM12L4A cells grown under normal culture conditions. Visualisation was by reaction of primary antibody with FITC-conjugated secondary antibody and confocal microscopy. Bars are 100 μm. (B) Anti-vinculin staining of KM12C and KM12L4A cells after plating suspended cells on to poly-L-lysine (left panels) or fibronectin (right panels) for 60 min. Bars are 100 μm. Similar images to those shown for KM12L4A cells were obtained for KM12SM cells. (C) The ability of KM12C, KM12SM and KM12L4A cells to attach to different matrix components was monitored using Chromium-51-labelled cells that were plated on to collagen, fibronectin, laminin or vitronectin for 45 min at 37°C. A known number of cells in suspension was also counted to calibrate the Chromium-51 counts per cell, allowing estimation of the number of attached cells. The means and standard errors of three replicates are presented.

Figure 4

The effects of the Src inhibitory agents PP2 and SU6656 in KM12L4A cells. (A) PP2 suppresses Src autophosphorylation. c-Src phosphorylation on tyrosine-416 in the presence (+) and absence (−) of 10 μm PP2 for 60 min was determined by immunoprecipitating c-Src, blotting and probing with anti-phospho-tyrosine-416 (upper panels). c-Src expression was checked by probing blots with anti-Src (lower panels). The ability of PP2 (at concentrations ranging from 0 to 50 μm) to interfere with KM12L4A cell attachment to poly-L-lysine or to fibronectin was examined by Chromium-51 labelling. Suspended, labelled cells were treated with PP2 for 60 min prior to attaching for 60 min (lower panels). (B) SU6656 suppresses c-Src kinase activity. The ability of c-Src from cells grown in the presence and absence of SU6656 to phosphorylate enolase was determined by in vitro kinase assays (upper panels). The ability of SU6656 to suppress cell adhesion to fibronectin in 60 min was visualised microscopically and was quantitated as percentage inhibition of adhesion (lower panels). Bars are 125 μm.

Figure 5

(A) c-Src expression and activity (monitored by auto-phosphorylation at tyrosine-416) in KM12C cell clones (2C3 and 2C4) stably expressing active c-SrcY527F, or vector control (2CV), was examined and compared with parental KM12C cells and their metastatic derivatives, KM12SM and KM12L4A. Paxillin phosphorylation in vector- and c-SrcY527F-expressing KM12C cells was also examined. (B) In vitro growth of KM12C cells stably expressing active c-SrcY527F (2C3 and 2C4; seeded at 1×10⁵ cells in 35 mm dishes) was compared with vector-transfected cells (2CV). (C) The doubling times of subcutaneous tumours formed after injection of 10⁶ KM12C cells transfected with either vector (2CV) or active c-SrcY527F (2C3 and 2C4) were determined.

Figure 6

The effect of increasing cellular c-Src expression and activity on the formation of adhesion structures in KM12C cells expressing either vector (2CV; A, B) or active c-SrcY527F (2C3 or 2C4; C–H) were plated on to fibronectin (A, B, E, F, G, H) or poly-L-lysine (C, D), fixed and stained with anti-vinculin (A, C, E and G) or anti-Src (B, D, F and H) and examined by immunofluorescence. Bars are 100 μm.