TNF-alpha increases human melanoma cell invasion and migration in vitro: the role of proteolytic enzymes

Abstract

Inflammatory mediators have been reported to promote malignant cell growth, invasion and metastatic potential. More specifically, we have recently reported that tumour necrosis factor alpha (TNF-alpha) increases melanoma cell attachment to extracellular matrix (ECM) substrates and invasion through fibronectin. In this study, we extend these investigations asking specifically whether the TNF-alpha effect on cell invasion and migration involves activation of proteolytic enzymes. We examined the effect of TNF-alpha on melanoma expression/activation of type IV gelatinases matrix metalloproteinases 2 and 9 (MMPs -2 and -9) and general proteolytic enzymes. Stimulation with TNF-alpha significantly increased both melanoma cell migration at 24 h (+21%) and invasion through fibronectin (+35%) but did not upregulate/activate the expression of latent MMP-2 constitutively produced by these cells and did not upregulate their general protease activity. However, the increased cell migration and invasion through fibronectin observed following stimulation with TNF-alpha were inhibited by the general protease inhibitor alpha(2) macroglobulin. These findings suggest that the promigratory and proinvasive effect of TNF-alpha on this melanoma cell line may be mediated to some extent by induction of localised cell membrane-bound degradative enzyme activity, which is not readily detected in biochemical assays.

Figure 1

(A) Effect of TNF-α on HBL melanoma cell invasion through fibronectin. A significant increase in invasion (+35% above control) was observed with TNF-α at 100 U ml⁻¹. Control invasion is expressed as 100%. Results are the mean±s.e.m. of four experiments and triplicate wells were used for each experimental condition in each experiment (^*P<0.05). (B) Time course of TNF-α action on HBL cell migration on plastic as assessed by the ‘scratch wound’ migration assay. A significant increase in the overall migration at 24 h was observed for the cells stimulated with TNF-α at 200 U ml⁻¹. The data shown are mean values±s.e.m. from n=8 experiments (^*P<0.05).

Figure 2

Representative gelatin zymograph of control and TNF-α-stimulated HBL melanoma cell extracts and cell conditioned medium. Cells constitutively expressed pro-MMP-2 (66.3 kDa band) in the culture medium and no upregulation or activation of the latent enzyme was observed after preincubation with TNF-α at 200 U ml⁻¹. Lanes 1 and 2 represent internal human purified MMP-2 and MMP-9 standards.

Figure 3

(A) The proteolytic activity of a series of trypsin concentrations was assessed by the quenched fluorescent substrate assay. Results shown are of one representative experiment. Duplicate wells were used for each trypsin concentration. (B) Inhibition of the proteolytic activity of trypsin (3.3 μg ml⁻¹) by the addition of the general protease inhibitor α₂ macroglobulin. Inhibition of 84% was observed with α₂ macroglobulin at a concentration of 2 U ml⁻¹. Results show means±s.e.m. (n=2, ^*P<0.05).

Figure 4

(A) Effect of the general protease inhibitor α₂ macroglobulin at a concentration of 2 U ml⁻¹ on HBL melanoma cell invasion through fibronectin. When added to non-TNF-α-stimulated cells, the invasion was not significantly reduced. In contrast, the increased invasion of the TNF-α-stimulated cells was completely inhibited by α₂ macroglobulin and the invasion of these cells reduced to −8% of the control cell invasion. Results are the mean±s.e.m. of three experiments, ^*P<0.05. (B) Effect of α₂ macroglobulin (2 U ml⁻¹) on the migration of control and TNF-α-stimulated cells. α₂ macroglobulin decreased the control level of migration and the increase in migration resulting from cell exposure to TNF-α was completely inhibited by the protease inhibitor (n=3, ^*P<0.05).