Cell surface serine proteases as regulators of epithelial development, repair, and malignancy (33% effort) Background: Proteases and protease inhibitors operate in the pericellular environment to provide proteolysis essential for cytokine/growth factor maturation, matrix remodeling, signaling receptor activation, ion channel activity, and more. Research accomplished: Head and neck squamous cell carcinoma (HNSCC) is associated with aberrant activity of the tyrosine kinase receptor, c-Met, caused by enhanced expression of its cognate ligand hepatocyte growth factor (HGF). ProHGF cannot induce receptor activation unless cleaved by a serine protease whose identity is debated. In collaboration with Silvio Gutkind and Alfredo Molinolo, OPCB, we explored if matriptase was endogenous activator of proHGF in the context of squamous cell carcinogenesis. Indeed, we found that matriptase amplified migratory responses of keratinocytes to proHGF through c-Met. Ablation of c-Met from matriptase-expressing keratinocytes negated the ability of the protease to promote squamous cell carcinogenesis. Matriptase-mediated carcinogenesis also was blocked by inhibition of the mammalian target of rapamycin (mTor). Our data identify matriptase as an activator of proHGF in squamous cell carcinoma and reveal mTor activation as a component of matriptase/c-Met-induced oncogenesis. Cellular trafficking of matriptase and prostasin in polarized epithelial cells Matriptase promotes tight junction formation and terminal differentiation of oral epithelium and epidermis through activation of the serine protease, prostasin. The two membrane-anchored serine proteases are co-expressed in multiple epithelia. In polarized epithelia, however, matriptase is located on the basolateral membrane, whereas prostasin is located on the apical membrane. To investigate how matriptase and prostasin interact, in collaboration with Lotte Vogel, University of Copenhagen, we mapped the subcellular itinerary of matriptase and prostasin in polarized epithelial cells. Matriptase was transported directly to the basolateral membrane, but prostasin was initially routed to the basolateral membrane, and only thereafter transcytosed to the apical membrane. Furthermore, prostasin was activated on the basolateral membrane before transcytosis. Thus, matriptase and prostasin briefly co-localize at the basolateral membrane providing an explanation for how matriptase may activate prostasin during transit to its apical destination. Expression and genetic loss of function analysis of the HAT/DESC cluster proteases Type II transmembrane serine proteases (TTSPs) phylogenetically belong to one of four subfamilies: matriptase, hepsin/TMPRSS, corin and HAT/DESC. A wealth of information has been gathered as to the functions the hepsin/TMPRSS, matriptase, and corin subfamilies, whereas little is known about the HAT/DESC subfamily. We therefore performed an expression and genetic loss of function analysis of this subfamily. The HAT/DESC proteases were coordinately expressed, suggesting a level of functional redundancy. Phenotypic analysis of mice deficient in two of the most widely expressed HAT/DESC proteases, TMPRSS11A and HAT, revealed that the two proteases are dispensable for development, health, and long-term survival in the absence of external challenges or additional genetic deficits. Our delineation of HAT/DESC protease expression and generation of TMPRSS11A- and HAT-deficient mutant mouse strains provide a valuable resource for the scientific community for exploration of HAT/DESC subfamily proteases. Urokinase receptor-associated protein (uPARAP) and intracellular collagen turnover (33% effort) Background: The urokinase plasminogen activator (uPA) receptor-associated protein (uPARAP) is part of the mannose receptor family of endocytic transmembrane proteins. uPARAP is critical for the uptake and lysosomal degradation of collagen by mesenchymal cells engaged in bone formation, and in pathological matrix remodeling by stromal cells during tumor invasion. Research accomplished: We previously showed that murine fibroblasts, osteoblasts, and chondrocytes deficient in uPARAP fail to endocytose and degrade collagen. By using uPARAP-deficient mice, a series monoclonal antibodies that block uPARAP-dependent collagen uptake was generated in collaboration with Dr. Niels Behrendt, University of Copenhagen. We used the antibodies to investigate the capacity of a large number of primary cells, established cell lines, and tumor cell lines to endocytose and degrade collagen. This study revealed a tight quantitative correlation between endocytic collagen uptake rates and uPARAP abundance, as well as a strict requirement of uPARAP for collagen uptake. An exception was primary macrophages that internalized collagen not through uPARAP, but via the closely related mannose receptor. Reengineered bacterial cytotoxins as antitumor agents (33% effort) Background: We are engaged in a long-standing collaboration with Steve Leppla, NIAID, Art Frankel, Scott &White, Temple, Texas and Chris Austin, NIH Chemical Genomics Center, NHGRI, on the development of reengineered bacterial cytotoxins as therapeutic agents for cancer and as tools for the imaging of cell surface proteolytic activity. Research accomplished: We previously generated an intercomplementing engineered anthrax toxin variant that requires cleavage by both uPA and matrix metalloproteinases (MMPs) and targets the ERK/MAPK pathway, thereby achieving high tumor selectivity. HNSCC is characterized by the upregulation of a large number of proteolytic enzymes, including uPA and MMPs. We therefore explored the use of our intercomplementing anthrax toxin for the treatment of HNSCC. Indeed, the toxin displayed strong systemic anti-tumor activity towards four different xenografted human HNSCC cell lines by inducing apoptotic and necrotic tumor cell death, and by impairing tumor cell proliferation and angiogenesis. Complete tumor regression was observed in a subset of mice, demonstrating that the intercomplementing toxin has the capacity to eradicate established tumors. All HNSCC cell lines were insensitive to the intercomplementing toxin when cultured ex vivo, suggesting that the toxin targets the tumor-supporting stromal cell compartment or that the tumor cell requirement for ERK/MAPK signaling differs in vivo and ex vivo. The intercomplementing toxin warrants further investigation as an anti-HNSCC agent. High-throughput assay for inhibitors of the generation of cell surface uPA activity We previously developed a simple and sensitive assay that uses non-toxic, reengineered anthrax toxins to specifically and quantitatively measure cell surface uPA activity. Because this imaging assay uniquely measures the activity of a cell surface protease in its natural milieu and is easy to miniaturize, we hypothesized that it would be ideally suited for the identification of new classes of uPA inhibitors that do not target the active site of uPA, but rather the micro-environmental factors required for the generation of cell surface uPA activity, and, thus, could aid the elucidation of the complex and poorly understood proteolytic pathways that govern the conversion of pro-uPA to active uPA. To test this proposition, in collaboration with the NIH Chemical Genomics Center, NHGRI, and supported by an RO3, we adapted the assay for fully automated quantitative high-throughput screening. The modified assay was miniaturized to a 1536-well plate, 8 μl reaction volume, 4000 cells/reaction format. Screening of 325,464 compounds, each at four titrated concentrations, yielded 913 compounds with demonstrated inhibitory activity towards uPA cleavage-dependent internalization of the anthrax toxin and subsequent cellular intoxication. These compounds have recently been obtained for hit validation.
Showing the most recent 10 out of 68 publications
