ML359, a Small Molecule Inhibitor of Protein Disulfide Isomerase That Prevents Thrombus Formation and Inhibits Oxidoreductase but Not Transnitrosylase Activity

Vascular thiol isomerases comprise a family of enzymes including protein disulfide isomerase (PDI), ERp5, and ERp57 that are important in the process of thrombus formation. PDI is secreted at sites of vascular injury, and antibody-mediated PDI inhibition prevents thrombus formation in a mouse laser injury model. Our group has previously reported on the discovery of the small molecule PDI inhibitors quercetin-3-rutinoside and ML359. Identified as part of a high-throughput screen, ML359 is a second-generation PDI inhibitor that selectively blocks PDI oxidoreductase activity with approximately ten-fold the potency of quercetin-3-rutinoside. To better understand the mechanism of allosteric modulation of PDI by small molecules, we evaluated the association of ML359 with isolated domains of PDI, determined the effects of ML359 on a variety of PDI functions, and compared the activity of ML359 to that of quercetin-3-rutinoside. PDI is composed of four thioredoxin-like domains and an x-linker region in the sequence a-b-b’-x-a’. Major substrate binding is thought to occur in the b-b’ region while the a and a’ domains contain catalytically active cysteine motifs (CGHC) that mediate the oxidoreducase, nitrosylase, and thiol isomerase functions of PDI. In order to identify potential binding sites of ML359 on PDI, we constructed and expressed the domain fragments a, ab, abb’, b’xa’, and a’. These fragments were tested in the presence of 10 µM ML359 using an insulin turbidometric assay that measures the oxidoreductase activity of PDI. ML359 demonstrated full inhibition of oxidoreductase activity when full-length PDI and the b’xa’ fragment were used whereas no inhibition was observed with the other fragments assayed. These results are consistent with docking studies showing that ML359 likely binds in a pocket formed at the b’x interface. In contrast, when the same experiment was performed in the presence of 30 µM of quercetin-3-rutinoside, inhibition was only noted with full-length PDI and the abb’ and b’xa’ fragments, suggesting that binding was dependent on the b’ and not the x-linker region. To determine if ML359 has differential effects on the oxidoreductase and nitrosylase functions of PDI, we utilized a platelet-based assay in which fluorescence intensity stemming from the NO-sensitive intracellular dye DAF-FM was measured as an indicator of PDI-mediated translocation of NO from the extracellular surface into the cytosol (transnitrosylation). While quercetin-3-rutinoside potently inhibited PDI-mediated transnitrosylation activity, ML359 had no effect. These results are consistent with the idea that the transnitrosylase and oxidoreducase functions of PDI are separable and inhibition of either is specific to the small molecule used. We evaluated the ability of ML359 to inhibit thrombosis in a mouse laser injury model. Intravital microscopy was used to follow thrombus formation in mouse cremaster arterioles after laser-induced vascular injury. Infusion of ML359 resulted in inhibition of thrombus formation, in contrast to thrombosis seen after infusion of vehicle alone. In summary, ML359 is a second generation small molecule inhibitor of PDI that likely binds at the b’x interface of the enzyme. Furthermore, ML359 is able to selectively inhibit PDI oxidoreductase activity without affecting transnitrosylase activity. ML359 may prove a useful molecular probe to better understand the different functions of PDI relative to thrombus formation in vivo.