IDENTIFICATION OF NOVEL ANTAGONISTS OF PROTEIN DISULFIDE ISOMERASE FOR INHIBITION OF
THROMBUS FORMATION.
Authors: Robert Flaumenhaft ,Carol Khodier ,Christina Galinski ,Partha Nag ,Freda Passam ,Alissa Scalise
,Lotte van Hessem ,Suzanne Gunnink ,Sivaraman Dandapani ,Benito Munoz ,Daniel Kennedy
Background:
Protein disulfide isomerase (PDI) is a widely expressed oxidoreductase that is highly concentrated in
endoplasmic reticulum and is required for protein folding. However, it can also be released from cells
into the extracellular environment, where it catalyzes the rearrangement of intramolecular disulfide
bonds in receptors and secreted proteins. Several animal models demonstrate that PDI released from
platelets and endothelial cells is required for thrombus formation in vivo. Recently, we identified the
flavonoid rutin as an inhibitor of PDI. Rutin inhibited both platelet accumulation and fibrin generation
during thrombus formation in mouse models at concentrations commonly ingested as nutritional
supplements, providing proof-of-principle for PDI inhibitors as a new class of antithrombotics (J. Clin.
Invest., 122:2104). Although quercetin flavonoids may be useful as a first generation of PDI inhibitors,
they are only modestly potent for PDI (IC50 6 μM), poorly absorbed, and have activities other than PDI
inhibition.
Aims:
Our goal was to identify potent and selective inhibitors of PDI as novel antithrombotics.
Methods:
A high throughput screen of PDI activity was developed based on the ability of the enzyme to reduce the
insulin β-chain, resulting in insulin aggregation. This insulin turbidimetric assay was converted to a 1536well format and an industrial scale screen of approximately 350,000 compounds of the Molecular
Libraries Small Molecule Repository was performed.
Results:
The primary screen identified 443 putative PDI inhibitors, which were subsequently tested in 8-point
dose curves. Forty-one compounds were identified with IC50s <10 μM. Two compounds, a piperdine
(IC50 0.3-0.6 μM) and a bromo indole (IC50 0.6-0.8 μM), were selected for further analysis. To assess the
selectively of these compounds, their activity in other bioassays performed within the NIH Molecular
Libraries Probe Production Network was evaluated. The piperdine showed no confirmed activity at <10
μM in any of the other of 380 biological assays in which it has been tested. The bromo indole had
confirmed activity at <10 μM in only 2 other assays out of 473 in which it had been tested. Within the
thiol isomerase family, both compounds demonstrated selectivity for PDI, failing to inhibit either ERp5,
ERp57, or thioredoxin. The piperdine was soluble at 65 μM in aqueous solution and demonstrated 97%
stability in GSH at 48 h, indicating that the compound did not inhibit PDI by interacting with the CxxC
catalytic domain. The bromo indole was soluble at 81 μM and demonstrated 99% stability to GSH.
Consistent with this observation, inhibition of PDI by the piperdine and the bromo indole was entirely
reversible. Neither compound demonstrated toxicity in a HeLa cell assay at 20-fold their IC50s. Organic
synthesis of 31 analogs of the piperdine lead was performed and structure activity relationships
determined. In preliminary in vivo studies, the piperdine inhibited platelet accumulation in cremaster
arterioles following laser injury.
Conclusions:
The compounds identified in this screen will serve as leads for the development of a second generation
of PDI inhibitors as a new class of antithrombotics with improved potency and specificity compared with
quercetin flavonoids.