No abstract

Cytosolic glutathione S-transferases (GSTs) play a critical role in xenobiotic binding and metabolism, as well as in modulation of oxidative stress. Here, the high-resolution X-ray crystal structures of homodimeric human GSTA1-1 in the apo form and in complex with S-hexyl glutathione (two data sets) are reported at 1.8, 1.5, and 1.3A respectively. At this level of resolution, distinct conformations of the alkyl chain of S-hexyl glutathione are observed, reflecting the nonspecific nature of the hydrophobic substrate binding site (H-site). Also, an extensive network of ordered water, including 75 discrete solvent molecules, traverses the open subunit-subunit interface and connects the glutathione binding sites in each subunit. In the highest-resolution structure, three glycerol moieties lie within this network and directly connect the amino termini of the glutathione molecules. A search for ligand binding sites with the docking program Molecular Operating Environment identified the ordered water network binding site, lined mainly with hydrophobic residues, suggesting an extended ligand binding surface for nonsubstrate ligands, the so-called ligandin site. Finally, detailed comparison of the structures reported here with previously published X-ray structures reveal a possible reaction coordinate for ligand-dependent conformational changes in the active site and the C-terminus.

The binding of plasminogen activator inhibitor-1 (PAI-1) to serine proteinases, such as tissue-type plasminogen activator (tPA) and urokinase-type plasminogen activator (uPA), is mediated by the exosite interactions between the surface-exposed variable region-1, or 37-loop, of the proteinase and the distal reactive center loop (RCL) of PAI-1. Although the contribution of such interactions to the inhibitory activity of PAI-1 has been established, the specific mechanistic steps affected by interactions at the distal RCL remain unknown. We have used protein engineering, stopped-flow fluorimetry, and rapid acid quenching techniques to elucidate the role of exosite interactions in the neutralization of tPA, uPA, and ␤-trypsin by PAI-1. Alanine substitutions at the distal P4 (Glu-350) and P5 (Glu-351) residues of PAI-1 reduced the rates of Michaelis complex formation (k a ) and overall inhibition (k app ) with tPA by 13.4-and 4.7-fold, respectively, whereas the rate of loop insertion or final acyl-enzyme formation (k lim ) increased by 3.3-fold. The effects of double mutations on k a , k lim , and k app were small with uPA and nonexistent with ␤-trypsin. We provide the first kinetic evidence that the removal of exosite interactions significantly alters the formation of the noncovalent Michaelis complex, facilitating the release of the primed side of the distal loop from the active-site pocket of tPA and the subsequent insertion of the cleaved reactive center loop into ␤-sheet A. Moreover, mutational analysis indicates that the P5 residue contributes more to the mechanism of tPA inhibition, notably by promoting the formation of a final Michaelis complex.Plasminogen activator inhibitor-1 (PAI-1) 1 functions as the primary regulator of the fibrinolytic system by inhibiting the conversion of plasminogen into plasmin via its action on tissuetype plasminogen activator (tPA) and urokinase-type plasminogen activator (uPA) (1). This inhibitor also plays a vital role in other physiological processes, including tumor invasion and tissue remodeling (2). PAI-1 belongs to the serpin superfamily, which shares several unique structural features, including a five-stranded ␤-sheet motif and a flexible reactive center loop (RCL). The conformational changes associated with these structures have been linked to the inhibitory function of PAI-1 (3-9). Notably, the mechanism of inhibition depends on the structural changes accompanying the S (stressed) to R (relaxed) transition that results in complete insertion of the Nterminal (proximal) part of the RCL into ␤-sheet A as an additional ␤-strand, s4A. The proteinase, tethered by a covalent bond with the P1 residue of the serpin, is displaced from the initial docking site to the opposite end of the serpin molecule, separating the P1Ј and P1 residues by ϳ70 Å (10 -12). This mechanism efficiently distorts and inactivates the catalytic triad of the proteinase, stabilizing the serpin-proteinase complex at the acyl-enzyme intermediate stage.Unique to the PAI-1 inhibitory mechanism are the exosite i...

No abstract

Binding of a hydrophobic glutathione product conjugate to rGST A1-1 proceeds via a two-step mechanism, including rapid ligand docking, followed by a slow isomerization to the final [GST.ligand] complex, which involves the localization of the flexible C-terminal helix. These kinetically resolved steps have been observed previously by stopped-flow fluorescence with the wild-type rGST A1-1, which contains a native Trp-21 approximately 20 A from the ligand binding site at the intrasubunit domain-domain interface. To confirm this binding mechanism, as well as elucidate the effects of truncation of the C-terminus, we have further characterized the binding and dissociation of the glutathione-ethacrynic acid product conjugate (GS-EA) to wild-type, F222W:W21F, and Delta209-222 rGST A1-1 and wild-type hGST A1-1. Although modest kinetic differences were observed between the hGST A1-1 and rGST A1-1, stopped-flow binding studies with GS-EA verified that the two-step mechanism of ligand binding is not unique to the GST A1-1 isoform from rat. An F222W:W21F rGST A1-1 double mutant provides a direct fluorescence probe of changes in the environment of the C-terminal residue. The observation of two relaxation times during ligand binding and dissociation to F222W:W21F suggests that the C-terminus has an intermediate conformation following ligand docking, which is distinct from its conformation in the apoenzyme or localized helical state. For the wild-type, Delta209-222, and F222W:W21F proteins, variable-temperature stopped-flow experiments were performed and activation parameters calculated for the individual steps of the binding reaction. Activation parameters for the binding reaction coordinate illustrate that the C-terminus provides a significant entropic contribution to ligand binding, which is completely realized within the initial docking step of the binding mechanism. In contrast, the slow isomerization step is enthalpically driven. The partitioning of entropic and enthalpic components of binding energy was confirmed by isothermal titration calorimetry with wild-type and Delta209-222 rGST A1-1.