Author: Manuel Baca

Publisher:

Published: 1994

Total Pages: 480

ISBN-13:

Total chemical synthesis has been used to explore structure-function relationships of the protease from human immunodeficiency virus 1 (HIV-1 protease). Analogues of this enzyme were prepared by conventional stepwise solid phase peptide synthesis and by chemical ligation of unprotected peptide segments. These analogues contained unnatural elements of structure, each designed to address a specific aspect of HIV-1 protease structure-function. Replacement of a surface [beta]-turn in each subunit of the homodimeric enzyme with a constrained, type II [prime] [beta]-turn mimic produced a fully active enzyme analogue with enhanced resistance to thermal inactivation. These results indicate that the precise geometry of this [beta]-turn is not critical for activity, and that replacement of native sequence with a rigid [beta]-turn mimic can lead to superior protein stability. Mass spectrometric and NMR experiments were designed to study the properties of a solution enzyme-inhibitor complex. No evidence for a specifically bound water molecule was found in the noncovalent ternary complex between the homodimeric enzyme and a reversible substrate-based inhibitor as detected by mass spectrometry. Bound water is reputedly involved in mediating hydrogen bonds between the enzyme flaps and substrate P2 and P1 [prime] carbonyls. The mode of binding and catalytic contribution of these enzyme-substrate hydrogen bonds was also investigated. An analogue of HIV-1 protease was designed in which the functionally relevant amide -CONH- linkage between residues Gly[49]-Ile[50] in each flap was replaced by an isosteric thioester -COS- bond. The catalytic activity (kcat) of the backbone modified enzyme was reduced approximately 3000 fold compared to the native amide bond-containing enzyme. This result confirms that backbone hydrogen bonds from the enzyme flaps to substrate are important for the catalytic function of the HIV-1 protease. These hydrogen bonds were further studied by removing the hydrogen bonding potential from one flap only. Existing peptide ligation methods were adapted to allow the linking of multiple peptide segments. Using these new methods, a tethered dimer analogue of HIV-1 protease was prepared in which the Gly[49]-Ile[50] amide bond was replaced in one subunit only by an ester -COO- bond. The kcat for this tethered dimer analogue was reduced only twofold. This suggests that both flap-substrate hydrogen bonds may originate from only a single flap, and not symmetrically from each flap as has been observed crystallographically. This hydrogen bonding pattern would coincide with that observed for the related pepsin-like aspartic proteases. Elucidation of the mechanistically relevant pattern of enzyme-substrate hydrogen bonding could have important applications to the design of HIV-1 protease inhibitors.