Introduction: Blood clots kill more humans than any other single pathogenic cause and acute treatments to digest blood clots are limited and carry significant bleeding risk. Current standard of care includes the delivery of systemic and catheter based fibrinolytic treatments, such as recombinant tissue Plasminogen Activators (tPA), which rely on activation of the host fibrinolytic system converting nearly of all a patient’s endogenous plasminogen to clot digesting plasmin. A limitation of this approach is that widespread activation of this system leads to a 5-10% rate of major bleeding, including intracranial hemorrhage. While bypassing plasminogen activation by administering active plasmin as a direct fibrinolytic agent is appealing, the clot lysing efficacy of naked plasmin is limited due to nearly instantaneous inactivation by circulating α2-antiplasmin and α2-macroglobulin. Plasmin is a unique enzyme with both a serine protease active site to digest fibrin and five kringle domains that bind to fibrin. By utilizing multivalent subsite binding principles tranexamic acid (TXA, kringle site binder) in combination with competitive inhibitors (benzamidine) on a single scaffold plasmin activity can be controlled for potential in-vivo targeting and drug delivery applications. By physically blocking plasmin’s active site it can be shielded from recognition and subsequent inactivation by α2-antiplasmin. This study specifically explores plasmin inhibition by hetero-bivalent inhibitors of varying dPEG linker lengths (dPEG4 to dPEG36) corresponding to ~4-17 nm of separation to achieve simultaneous subsite binding to both plasmin’s active site and K1 kringle domain. Additionally, the hetero-bivalent inhibitors were tested on human thrombin, enzyme responsible for fibrin deposition.
Materials and
Methods: Hetero-bivalent inhibitors comprising both benzamidine and TXA were synthesized in the following manner. Amino-methyl benzamidine (AMB) was first reacted with Fmoc-dPEGx-NHS/TFP esters (x = 4,8,12,36) in a mixture of DMF and PBS. Fmoc was deprotected using 30% piperidine in DMF and the NH2-dPEGx-AMB obtained was purified on a HPLC using a semi-preparative Thermo Hypersil GOLD C18 column (5µm, 250 x 10mm) on a gradient of water and methanol with 0.1% trifluoroacetic acid (TFA). The masses were confirmed using an Agilent LC-MS QTOF mass spectrometer. NH2-dPEGx-AMB was then reacted with Fmoc-TXA in DMF at room temperature using 2-(1H-benzotriazol-1-yl)-1,1,3,3-tetramethyluronium hexafluorophosphate (HBTU) and N, N-Diisopropylethylamine (DIEA). Fmoc was again deprotected using 30% piperidine and the product was purified on HPLC, and the final masses were again verified with mass spectrometry. Inhibition assays were performed using a fixed human plasmin concentration (42.5 nM) with chromogenic substrate S-2251 concentrations of 100-500 µM, over a range of inhibitor concentrations (0-300 mM) to determine inhibition constants (Ki values) via Dixon plots analysis. Similarly, inhibition assays for human thrombin utilized fluorogenic Thrombin Substrate III (TSIII; Benzoyl-Phe-Val-Arg-AMC•HCl) and were carried out at TSIII concentrations of 20-50 µM, inhibitor concentrations of 0-300 mM, and a fixed human thrombin concentration of 0.25 U/mL. Thrombin activity was determined by monitoring the fluorescent AMC tag (λex: 370 nm and λem: 450 nm) released by hydrolysis of TSIII.
Results, Conclusions, and Discussions: Hetero-bivalent tranexamic acid (TXA) and benzamidine (AMB) inhibitors varying linker lengths from ~4-17 nm were synthesized and it was observed that TXA-dPEG4-AMB was the strongest hetero-bivalent inhibitor for both plasmin and thrombin with Ki values of 42 ± 12 and 14 ± 6.7 µM, respectively (Table 1). The planar separation distance between kringle domain lysine binding site (LBS) on K1 of plasmin and the active site was measured using Chimera and found to be ~8 nm. Therefore, increasing from dPEG4 to dPEG12 linker lengths was not sufficient to achieve multivalent subsite binding at K1 as they are < 8 nm in length. With increasing linker length, inhibition became weaker as seen previously with homo-multivalent benzamidine inhibitors owing to a decrease in effective concentration and an increase in entropic penalty due to the overall inhibitor molecule size and flexibility. However, TXA-dPEG36-AMB instead of being a weaker inhibitor due to an even longer linker length, was found to be a stronger inhibitor of plasmin than TXA-dPEG12-AMB (Ki = 207 ± 4.1 µM). Despite its significantly longer linker length of 16.6 nm, TXA-dPEG36-AMB exhibited improved inhibition of plasmin with a Ki value of 75 ± 4.3 µM. The longer linker allowed for simultaneous binding of benzamidine to the active site and TXA to the LBS on K1 of plasmin resulting in enhanced inhibition due to multivalent subsite binding. As thrombin does not have any kringle domains, inhibition steadily decreased with increasing linker lengths. This study identifies how hetero-bivalent inhibitors can be designed to drive multivalent subsite binding to modulate plasmin inhibition. By physically blocking plasmin’s active site it can potentially be shielded from recognition and subsequent inactivation by α2-antiplasmin and α2-macroglobulin in the body to provide for a possible drug delivery strategy of unmodified, full-length plasmin as a safer direct fibrinolytic therapy. Simultaneous binding of plasmin’s active site and K1 LBS would still allow for in-vivo fibrin binding of plasmin via the remaining four kringle domains with reversible inhibition of plasmin being tuned to allow for inhibitor release once bound to fibrin.
