Matriptase is a type II membrane-anchored serine protease essential for epithelial integrity and with implication in carcinogenesis. Matriptase is co-expressed in epithelial cells of most tissues with the protease prostasin and their mutual inhibitor hepatocyte growth factor activator inhibitor 1 (HAI-1). All three proteins have crucial functions for epidermal integrity; and in the epidermis matriptase acts upstream of prostasin and is required for its activation.
The difference between the subcellular locations of matriptase and prostasin has so far been an enigma for matriptase-prostasin interaction. By mapping the subcellular trafficking of matriptase, prostasin and HAI-1, we demonstrate that the basolateral plasma membrane is part of the subcellular itinerary of both proteases and thus constitutes a location for interaction. In Caco-2 cells, zymogen matriptase is routed to the basolateral plasma membrane where it becomes activation site cleaved and activated. At steady state, prostasin is found to locate mainly to the apical plasma membrane although a minor fraction of prostasin can be detected at the basolateral plasma membrane. We show that prostasin is present in its active form on both the apical and basolateral plasma membrane and that prostasin as well as HAI-1 are endocytosed from the basolateral plasma membrane and transcytosed to the apical plasma membrane. Activation of matriptase on the basolateral plasma membrane is followed by efficient internalization as a matriptase-HAI-1 complex.
Deregulated or unopposed matriptase activity causes a perturbation of several biological functions e.g. developmental defects in mice. However, the molecular mechanisms behind these effects are poorly understood. We performed a genetic epistatic analysis to identify new components of matriptase-dependent pathways in embryogenesis. We show that prostasin is an indispensable component of the matriptase-dependent proteolytic cascade that causes early embryonic lethality in mice. In placental tissue prostasin acts upstream of matriptase and is required for the activation of matriptase. During embryonic development both HAI-1 and HAI-2 are essential inhibitors of matriptase. In this study we moreover find that both HAI-1 and HAI-2 are indispensable inhibitors of prostasin in placental tissue.
Matriptase activity is a tightly regulated protease; zymogen and HAI-1-complexed matriptase is abundantly present in most epithelial cell lines and tissues. There are no specific substrates or inhibitors of matriptase, which has led to challenges in detecting active matriptase. In order to assess the location and relative amount of active matriptase as compared to the total amount, we have established an assay for detection of active matriptase. This assay is based on a chloromethyl ketone peptide inhibitor with a predicted substrate sequence of matriptase. We show that active matriptase is present on the basolateral plasma membrane of Caco-2 cells and merely comprise a fraction of total matriptase. The conversion of zymogen matriptase to active matriptase can be induced by exposure to slight acidic conditions as well as the levels of HAI-1- complexed matriptase and active matriptase rises under these conditions in Caco-2 cells. Labeling of Caco-2 cells with the chloromethyl ketone peptide inhibitor delayed matriptase-HAI-1 complex formation. In addition to the peptide inhibitor being unable to bind HAI-1-complexed matriptase, we show that the matriptase zymogen has an intrinsic activity that enables it to bind the chloromethyl ketone peptide inhibitor. Moreover, this assay can be easily modified to detect active matriptase in other cell system, and we show that cultured primary murine keratinocytes contain low levels of peptidiolytic active matriptase.
Taken together, these results have improved our knowledge of the interplay between matriptase and prostasin towards a understanding of these components roles in epithelial integrity and present an assay for detection of active matriptase.