Membrane type-1 matrix metalloproteinase (MT1-MMP) as a target in cancer therapy.
Membrane type I-matrix metalloproteinase (MT1-MMP), a member of the highly active extracellular matrix (ECM)-degrading matrix metalloproteinases (MMPs), is known to be involved in connective tissue remodelling and embryogenesis, as well as tumour invasion and metastasis. Positioned on the leading edge of the invading cell, its proteolytic activity is enhanced by activation of proMMP-2 in a complex with tissue inhibitor of matrix metalloproteinase-2 (TIMP-2). The aim of this study was to attempt to produce highly specific and immunoinhibitory antibodies against human MT1-MMP and to test whether such antibodies are able to stop invasion by blocking MT1-MMP activity. As various MMP domains are highly conserved across species, and within specific MMP groups, and human cancer cells were to be used in invasion assays, sequence alignment of human MT1-MMP domains was used to identify the most variant and, hence, the optimal laboratory species for antibody production, and the chicken was subsequently selected. A hemopexin-like or collagen-binding domain, together with the catalytic domain (PEXcat), the hemopexin-like domain (PEX) and the propeptide and catalytic domain (PROcat) were selected as target domains for antibody production. Escherichia coli or Pichia pastoris expressed PEXcat and PEX domains, respectively, were obtained collaboratively, and an E. coli expression system was used to express the PROcat domain. Urea and β-mercaptoethanol successfully solubilised PROcat inclusion body protein, and Q- and S-Sepharose ion exchange chromatography, removed majority of the E. coli contaminating proteins, yielding >200 mg/litre expressed PROcat protein. Alum adjuvant and unrenatured soluble PEXcat and PEX proteins, or the less soluble, S-Sepharose purified PROcat protein was used for inoculations of chickens. The PROcat antigen, also injected as a homogenised band in acrylamide, proved to be inferior to S-Sepharose-purified PROcat antigen in alum, as it failed to induce an antibody response. The S-Sepharose-purified PROcat antigen, in alum adjuvant, produced the highest overall response, purified anti-PROcat IgY recognising recombinant forms of MT1-MMP (33 kDa and 50 kDa) and a 63 kDa protein in human blood, concluded to be either latent, soluble MT1-MMP or a non-specific protein. These antibodies, however, failed to detect native human and murine MT1-MMP (43 kDa) in cell line homogenates, suggesting that they possibly did not recognise the zinc-binding site of the catalytic domain in the 43 kDa processed MT1-MMP. In contrast to purified IgY, crude anti-PROcat IgY preparations recognised renatured PROcat MT1-MMP (29 kDa), indicating possible binding and removal of anti-human MT1-MMP antibodies during PEG purification. Despite this, the purified IgY resulted in higher immunoinhibition of the renatured PROcat antigen than the crude IgY. Anti-PEXcat antibodies had low titre, recognising native MT1-MMP in human cell (43 kDa) and mouse macrophage homogenates, but did not recognise the original recombinant PEXcat MT1-MMP antigen or PROcat MT1-MMP, possibly due to levels of loaded antigen being too low for detection in the western blots. Although these antibodies also did not seem to recognise the catalytic domain in the western blots, the high immunoinhibitory effect induced by these antibodies suggested otherwise. The PEX antigen induced the weakest antibody response, antibodies detecting only recombinant MT1-MMP (50 kDa). The anti-PROcat IgY, overall, produced denser labelling of the MCF10A and MCF10A-neoT cell lines, than the anti-PEXcat IgY, and these antibodies preferentially recognised the PRO domain of proMT1-MMP, focused in lamellipodia of the MCF10A cell line. Comparisons between the normal and cancer cell line, the anti-PROcat IgY labelled the MCF10A-neoT cells weaker than they labelled the MCF10A cells and the labelling was spread along the plasma membrane and the base of the cell. The anti-PEXcat IgY, in contrast, showed slightly more labelling of MT1-MMP in the MCF10A-neoT cells, compared to the MCF10A cells, which may promote the invasive phenotype of this cell line. In the fixed tissue, anti-PEXcat labelled all forms of MT1-MMP, as expected. Although similar labelling was observed with the anti-PROcat IgY, these antibodies were most likely recognising proMT1-MMP. Renaturation of Q-Sepharose purified PROcat antigen, using 0.5 mM ZnCl 2 gradient dialysis, produced catalytically active, renatured protein for immunoinhibition assays, although the observed higher Km in this study possibly suggested this procedure was not a successful as a one-step dialysis procedure previously reported. Despite this, immunoinhibition assays revealed a 88%, 70% and 34% inhibition of activity by the anti-PEXcat, PROcat and anti-PEX IgY, respectively, suggesting that the anti-PEXcat IgY would be most useful in invasion inhibition studies.