Several serotypes of Escherichia coli produce protein toxins closely related to Shiga toxin (Stx) from Shigella dysenteriae serotype 1. These Stx-producing E. coli cause outbreaks of hemorrhagic colitis and hemolytic uremic syndrome in humans, with the latter being more likely if the E. coli produce Stx2 than if they only produce Stx1. To investigate the differences among the Stxs, which are all AB 5 toxins, the crystal structure of Stx2 from E. coli O157:H7 was determined at 1.8-Å resolution and compared with the known structure of Stx. Our major finding was that, in contrast to Stx, the active site of the A-subunit of Stx2 is accessible in the holotoxin, and a molecule of formic acid and a water molecule mimic the binding of the adenine base of the substrate. Further, the A-subunit adopts a different orientation with respect to the B-subunits in Stx2 than in Stx, due to interactions between the carboxyl termini of the B-subunits and neighboring regions of the A-subunit. Of the three types of receptor-binding sites in the B-pentamer, one has a different conformation in Stx2 than in Stx, and the carboxyl terminus of the A-subunit binds at another. Any of these structural differences might result in different mechanisms of action of the two toxins and the development of hemolytic uremic syndrome upon exposure to Stx2.

Shigella dysenteriae is the pathogen responsible for the severe form of dysentery in humans. It produces Shiga toxin, the prototype of a family of closely related bacterial protein toxins. We have determined the structure of the holotoxin, an AB5 hexamer, by X-ray crystallography. The five B subunits form a pentameric ring, encircling a helix at the carboxy terminus of the A subunit. The A subunit interacts with the B pentamer via this C-terminal helix and a four-stranded mixed beta-sheet. The fold of the rest of the A subunit is similar to that of the A chain of the plant toxin ricin; both are N-glycosidases. However, the active site in the bacterial holotoxin is blocked by a segment of polypeptide chain. These residues of the A subunit would be released as part of the activation mechanism of the toxin.

The x-ray crystal structure of recombinant human renin has been determined. Molecular dynamics techniques that included crystallographic data as a restraint were used to improve an initial model based on porcine pepsinogen. The present agreement factor for data from 8.0 to 2.5 angstroms (A) is 0.236. Some of the surface loops are poorly determined, and these disordered regions border a 30 A wide solvent channel. Comparison of renin with other aspartyl proteinases shows that, although the structural cores and active sites are highly conserved, surface residues, some of which are critical for specificity, vary greatly (up to 10A). Knowledge of the actual structure, as opposed to the use of models based on related enzymes, should facilitate the design of renin inhibitors.

The design of inhibitors with enhanced potency against proteolytic enzymes has many applications for the treatment of human diseases. In addition to the optimization of chemical interactions between the enzyme and inhibitor, the binding affinity can be increased by constraining the inhibitor to the conformation that is recognized by the enzyme, thus lowering the entropic barrier to complex formation. We have structurally characterized the complexes of a macrocyclic pentapeptide inhibitor and its acyclic analogue with penicillopepsin, an aspartic proteinase, to study the effect of conformational constraint on the binding affinity. The phosphonate-based macrocycle PPi4 (Ki = 0.10 nM) is covalently linked at the P2-Asn and P1'-Phe side chains [nomenclature of Schechter and Berger, Biochim. Biophys. Res. Commun. (1967) 27, 157-162] via an amide bond, relative to the acyclic compound PPi3 (Ki = 42 nM). Comparisons of the high-resolution crystal structures of PPi4-penicillopepsin (0.95 A) and PPi3-penicillopepsin (1.45 A) reveal that the conformations of the inhibitors and their interactions with the enzyme are similar. The 420-fold increase in the binding affinity of PPi4 is attributed to a reduction in its conformational flexibility, thus providing the first rigorous measure of the entropic contribution to the binding energy in a protein-ligand complex and stressing the advantages of the design strategy.