From Loop to Strand: Characterization of the Conformation and Dynamics of the Human Plasminogen Activator Inhibitor-1 Reactive Center

Plasminogen activator inhibitor-1 (PAI-1), with its cofactor vitronectin (VN), controls the rate of plasmin-mediated fibrin breakdown in blood clots by inhibiting tissue-plasminogen activator (tPA) and urokinase-plasminogen activator (uPA). The activity of PAI-1 is attributed to its reactive center loop (RCL), which is solvent-exposed in an active conformation, but inserts as an additional strand into its central β [beta]-sheet during transition to a latent state and during inhibition. VN slows the latency transition, and the rate at which PAI-1 inhibits the plasminogen activators (PAs) also differs. However, the steps during the latency transition, mechanism of VN stabilization, and basis for inhibitory rate differences are unclear, and all involve the RCL. To address these issues, this study combines computational methods with cysteine-scanning mutagenesis of the RCL for fluorescence and electron paramagnetic resonance (EPR) spectroscopy to investigate changes in the RCL due to interactions with these ligands. Homology modeling of the RCL indicates sampling of a limited energy-conformation landscape for this region. Fluorescence investigation of the latency transition suggest that RCL detachment to assume the latent conformation occurs within the first 10 minutes of the process, which typically has a half-life of about 1 hour. Equilibrium-binding studies indicate that VN, its N-terminal somatomedin B (SMB) domain, and a longer truncation involving an intrinsically disordered domain (SMB-IDD) increase the solvent exposure of the RCL in stabilizing PAI-1. Studies with active site-blocked PAs reveal that both dock at the RCL, but rest differently on its top, employing distinct exosite interactions and mobility constraints on the RCL that likely effect the kinetics of its interaction with PAI-1. Thereby, this study provides detailed structural information on the PAI-1 RCL, and new insights into the latency process and interaction with PAs. Such information is valuable in the development of inhibitors specific for the interaction of PAI-1 with either PA, and in targeting this biomarker in diseases states caused by the dysregulation of PAI-1. Overall, the results from this work reveal that ligand interactions fine-tune the activity of PAI-1 by affecting the conformation and dynamics of the RCL from its position as a solvent-exposed loop to an inserted β [beta]-strand.