Metalloproteinases

The metastatic process involves tumor cell adhesion to basement membrane components, such as type IV collagen. A specific mitogen activated protein kinase cascade is activated by cell adhesion to type IV collagen. This activation causes the expression of proteolytic enzymes. These enzymes will then participate in compromising extracellular matrix components and enhance cell movement through them. To better understand tumor invasion of type IV collagen, we have constructed triple-helical peptide (THP) ligands for melanoma cell receptors, and used these ligands to determine if receptors such as CD44/CSPG and the alpha2beta1 integrin have unique matrix metalloproteinase (MMP) signaling pathways affected by the tyrosine kinase inhibitor genistein. MMP protein expression profiles were evaluated using the alpha2beta1 integrin ligand, and CD44/CSPG ligand. Results were indicative of specific activation sequences that tumor cells undergo upon binding to select regions of type IV collagen.

Metastatic cancers are problematic because they spread throughout the body. A crucial step in cancer metastasis is the separation of the cancer cells from their surrounding normal cells. This occurs due to suppression or destruction of cell adhesion molecules such as E-cadherin, occludin, and various claudins. The Snail and Slug transcription factors play a direct role in suppressing these cell adhesion molecules through their SNAG repression domain. We explored the possibility of developing an ELISA diagnostics capable of detecting soluble E-cadherin, occludin, and claudin fragments in the serum of cancer patients. Using several bioinformatics tools, unique extracellular antigenic sequences were identified on claudins-1, 4, 16, occludin, and E-cadherin. These sequences were cloned as GST fusion proteins, expressed, and purified in large quantities to raise antibodies. In parallel, expression profiling of metastatic cancer cell lines was carried out to derive a correlation between Snail-Slug expression and suppression of cell adhesion molecules.

The tissue inhibitors of metalloproteinases (TIMPs) are endogenous inhibitors of the matrix metalloproteinases (MMPs). Since unregulated MMP activities are linked to arthritis, cancer, and atherosclerosis, TIMP variants that are selective inhibitors of disease-related MMPs have potential therapeutic value. The structures of TIMP/MMP complexes reveal that most interactions with the MMP involve the N-terminal region of TIMP and the C-D B-strand connector which occupy the primed (right side of the active site) and unprimed (left side) regions of the active site. Substitutions for Thr2 of N-TIMP- 1 strongly influence MMP selectivity. In this study we found that Arg and Gly, which generally reduce MMP affinity, have less effect on binding to MMP-9. When the Arg mutation is added to the NTIMP-1 mutant with AB loop of TIMP-2, it produced a gelatinase-specific inhibitor with Ki values of 2.8 and 0.4 nM for MMP-2 and MMP-9, respectively. The Gly mutant has a Ki of 2.1 nM for MMP-9 and > 40 uM for MMP-2, indicating that engineered TIMPs can discriminate between MMPs in the same subfamily. In collaboration with Dr. Yingnan Zhang at Genentech, we have developed a protocol for the phage display of full-length human TIMP-2 to identify high-affinity selective inhibitors of human MMP-1, a protease that plays a role in cleaving extracellular matrix (ECM) components, connective tissue remodeling during development, angiogenesis, and apoptosis. We have generated a library containing 2x1010 variants of TIMP-2 randomized at residues 2-6 (L1), at residues 34-40 (L2) and 67-70 (L3).

Tissue inhibitors of metalloproteinases (TIMPs) comprise a family of four proteins in humans that modulate the turnover of the extracellular matrix by regulating the activities of endopeptidases that catalyze its degradation, especially the matrix metalloproteinases (MMP). In general, the four TIMPs are broad-spectrum tight binding inhibitors of MMPs with individual differences in specificity. In this study, we attempted to understand the basis of such variation by using membrane type-1 MMP (MT1-MMP) as a model, since it is inefficiently inhibited by TIMP-1 in contrast with the other TIMPs. We designed and engineered mutations in the N-domain of TIMP-1, based on current knowledge of TIMP interactions. By measuring inhibition levels of each mutant against several MMPs, including MT1-MMP, we were able to obtain a triple mutant with an vii improved affinity for MT1-MMP.