Deciphering the role of trehalose in hindering antithrombin polymerization

Abstract

Serine protease inhibitors (serpins) family have a complex mechanism of inhibition that requires a large scale conformational change. Antithrombin (AT), a member of serpin superfamily serves as a key regulator of the blood coagulation cascade, deficiency of which leads to thrombosis. In recent years, a handful of studies have identified small compounds that retard serpin polymerization but abrogated the normal activity. Here, we screened small molecules to find potential leads that can reduce AT polymer formation. We identified simple sugar molecules that successfully blocked polymer formation without a significant loss of normal activity of AT under specific buffer and temperature conditions. Of these, trehalose proved to be most promising as it showed a marked decrease in the bead like polymeric structures of AT shown by electron microscopic analysis. A circular dichroism (CD) analysis indicated alteration in the secondary structure profile and an increased thermal stability of AT in the presence of trehalose. Guanidine hydrochloride (GdnHCl)-based unfolding studies of AT show the formation of a different intermediate in the presence of trehalose. A time-dependent fluorescence study using 1,1'-bi(4-anilino)naphthalene-5,5'-disulfonic acid (Bis-ANS) shows that trehalose affects the initial conformational change step in transition from native to polymer state through its binding to exposed hydrophobic residues on AT thus making AT less polymerogenic. In conclusion, trehalose holds promise by acting as an initial scaffold that can be modified to design similar compounds with polymer retarding propensity.

Keywords: antithrombin; chemical chaperones; polymerization; serpin; thrombosis; trehalose.

Figure 1. Rate of thrombin inhibition by AT in the absence and presence of small molecules

Kinetics of thrombin inhibition by AT were determined by reacting AT in three different concentrations (100, 200 and 300 nM) with thrombin (30 nM) in the (A) absence and presence of (B) 1.5 M sorbitol, (C) 1 M trehalose, (D) 1.5 M mannose, (E) 1 M TMAO and (F) 1.25 M serine at given time-points in a 96-well plate. Absorbance was taken at 405 nm after the addition of thrombin substrate S-2238 (0.15 mM). Appropriate thrombin and S2238 controls/blanks with small molecules in the absence of protein were taken. Measurements were carried out at least three times and data were analyzed with linear regression function of GraphPad Prism v5.0.

Figure 2. Stability of AT in the absence and presence of small molecules

Relative ellipticity change at 222 nm is plotted against temperature over the range of 35–80°C of 3 µM of AT in the absence and presence of 1 M trehalose, 1.5 M sorbitol and 1.5 M mannose as described in methods. Relative ellipticities were calculated by using the relation [θ_obs− θ_min]/[θ_{max −}θ_min] [38] where θ_min and θ_max were fitted values of ellipticity at the lowest and highest temperatures used in the study, respectively, and θ_obs is the observed ellipticity at temperature T. Each curve is an average of three experiments.

Figure 3. Electron micrographs showing reduction in the size and shape of AT polymers by sorbitol and trehalose

AT Polymers were formed by heating 10 µM of the protein at 60°C for 90 min. Samples were stained negatively with 1.5% (w/v) uranyl acetate, and viewed with a magnification of up to ×50000. AT incubated at (A) 0 min, (D) 90 min; AT incubated in the presence of 1.5 M sorbitol at (B) 0 min and (E) 90 min and in the presence of 1 M trehalose at (C) 0 min and (F) 90 min. Insets in D–F shows enlarged view of polymers.

Figure 4. Effect of trehalose on native state of AT

(A) Bis-ANS fluorescence, (B) Far-UV CD and (C) Near-UV CD spectra of native AT in the absence and presence of 1 M trehalose. Experimental details are mentioned in Materials and methods section. All measurements were carried out at 25°C in 20 mM phosphate buffer containing 100 mM NaCl and 0.1% EDTA (PNE buffer). The concentration of protein was 1–20 µM and that of trehalose was 1 M. All the data were corrected for buffer effect. Each curve represents an average of three (for Bis-ANS and Far-UV CD) and two (for near-UV CD) independent experiments.

Figure 5. GdnHCl-based denaturation profile of AT

GdnHCl (0–6 M)-induced unfolding transition of native AT was measured by the change in the fluorescence signal. Representative tryptophan fluorescence spectra of AT were recorded in the (A) absence and (B) presence of 1 M trehalose. All the measurements were carried out with 500 nM of AT incubated with GdnHCl (0–6 M) for 2 h prior to fluorescence measurements at 25°C in PNE buffer. An excitation wavelength of 280 nm was used and the scan was recorded from 295 to 440 nm. (C) Emission maxima plot extrapolated from the chemical denaturation profile of AT by GdnHCl (0–6 M) monitored by intrinsic tryptophan fluorescence in the absence and presence of 1 M trehalose. Each plot is an average of measurements carried out at least three times.

Figure 6. Effect of trehalose on intermediate state of AT

(A) Far-UV CD spectra and (B) Bis-ANS fluorescence of AT in the absence and presence of 1 M trehalose was done to observe conformation of intermediate state. AT was incubated with 2 M GdnHCl for 2 h in the absence and presence of 1 M trehalose prior to CD and Fluorescence analysis. The concentration of AT was 1 µM and the molar ratio of AT to bis-ANS was 1:10. Data were corrected by subtracting the buffer contribution. All data points in each plot were obtained from the average value of atleast three independent experiments.

Figure 7. Time dependence of AT studied by bis-ANS fluorescence

Time course of AT polymerization in the (A) absence and (B) presence of 1 M trehalose. 2 µM AT in a total of 1 ml in PNE buffer was incubated at 60°C at indicated times in the absence and presence of 1 M trehalose. Samples were removed post-incubation and increase in fluorescence was measured at 485 nm after excitation of samples at 385 nm in bis-ANS. (C) Fluoresence intensity versus time plot. Average points from three separate experiments each with five scans were used to plot the graphs. Data were analyzed with two-way ANOVA (bonferroni post tests) and a P-value ≤0.01 was considered significant.