Milan Souček

Unusual Binding Mode of an HIV-1 Protease Inhibitor Explains its Potency against Multi-drug-resistant Virus Strains

Protease inhibitors (PIs) are an important class of drugs for the treatment of HIV infection. However, in the course of treatment, resistant viral variants with reduced sensitivity to PIs often emerge and become a major obstacle to successful control of viral load. On the basis of a compound equipotently inhibiting HIV-1 and 2 proteases (PR), we have designed a pseudopeptide inhibitor, QF34, that efficiently inhibits a wide variety of PR variants. In order to analyze the potency of the inhibitor, we constructed PR species harboring the typical (signature) mutations that confer resistance to commercially available PIs. Kinetic analyses showed that these mutated PRs were inhibited up to 1,000-fold less efficiently by the clinically approved PIs. In contrast, all PR species were effectively inhibited by QF34. In a clinical study, we have monitored 30 HIV-positive patients in the Czech Republic undergoing highly active antiretroviral therapy, and have identified highly PI resistant variants. Kinetic analyses revealed that QF34 retained its subnanomolar potency against multi-drug resistant PR variants. X-ray crystallographic analysis and molecular modeling experiments explained the wide specificity of QF34: this inhibitor binds to the PR in an unusual manner, thus avoiding contact sites that are mutated upon resistance development, and the unusual binding mode and consequently the binding energy is therefore preserved in the complex with a resistant variant. These results suggest a promising route for the design of second-generation PIs that are active against a variety of resistant PR variants.

Inhibitor binding at the protein interface in crystals of a HIV-1 protease complex.

Depending on the excess of ligand used for complex formation, the HIV-1 protease complexed with a novel phenylnorstatine inhibitor forms crystals of either hexagonal (P6(1)) or orthorhombic (P2(1)2(1)2(1)) symmetry. The orthorhombic form shows an unusual complexity of crystal packing: in addition to one inhibitor molecule that is bound to the enzyme active site, the second inhibitor molecule is bound as an outer ligand at the protein interface. Binding of the outer ligand apparently increases the crystal-quality parameters so that the diffraction data allow solution of the structure of the complex at 1.03 A, the best resolution reported to date. The outer ligand interacts with all four surrounding HIV-1 protease molecules and has a bent conformation owing to its accommodation in the intermolecular space. The parameters of the solved structures of the orthorhombic and hexagonal forms are compared.

Solid-phase peptide synthesis by fragment condensation: Coupling in swelling volume

The condensation of short peptides to resin-bound fragments was examined with respect to high coupling yields with only a small molar excess of a peptide in the reaction solution. The best results were achieved by the addition of reactants (C-unprotected peptide, DIC, and HOBt) dissolved in a so-called swelling volume of an appropriate solvent to a dry resin with an attached N-deprotected peptide chain. Each coupling step was followed by the end-capping of unreacted resin-bound peptide with 2,4-dinitrofluorobenzene. The substituted dinitroaniline chromophore formed in this reaction made the detection and separation of deletion peptides easy. Both conventional and ‘swelling volume’ methods were compared on parallel syntheses of the HIV-1 protease C-terminal 78–99 fragment. The yields of the isolated heneicosapeptide were 21 and 81% in favor of the ‘swelling volume’ procedure.

A distinct binding mode of a hydroxyethylamine isostere inhibitor of HIV-1 protease

Crystallization conditions for an HIV-1 protease-inhibitor complex were optimized to produce crystals suitable for X-ray diffraction experiments. The X-ray structure of the HIV-1 protease complex was solved and refined at 3.1 A resolution. In contrast to Saquinavir, the mimetic hydroxy group of the inhibitor Boc-Phe-Psi[(S)-CH(OH)CH(2)NH]-Phe-Glu-Phe-NH(2) is placed asymmetrically with respect to the non-crystallographic twofold axis of the protease dimer so that hydrogen bonds between the amino group of the inhibitor and the catalytic aspartates can be formed. The inhibitor binds in the centre of the active site by a compact network of hydrogen bonds to Gly27, Gly127, Asp25, Asp125 and via the buried water molecule W301 to Ile50 and Ile150.

Development and Testing of Inhibitors of Candida Aspartic Proteinases

Candida is a typical opportunistic pathogen which is often present in healthy individuals but under certain conditions it causes fungal infections.1 In addition to superficial infections, Candida is a major cause of deep systemic infections in immunocompromised patients such as HIV patients, transplant recipients, or cancer patients undergoing chemotherapy. An aspartic proteinase secreted by many members of the genus Candida has been often suggested to take part in the invasive character of the microorganism.2 The aspartic proteinases secreted by Candida represent a potential target for the drug intervention of the disease and the studies devoted to the understanding of these enzymes span from genetic studies3 to substrate specificity studies4 or crystallographic studies.5,6 It has been shown that there are differences in the specificity of secreted aspartic proteinases (SAP) from different Candida strains and that these proteinases are in general similar to eukaryotic aspartic proteinases with a deep active site cleft which accommodates at least eight residues of a substrate or inhibitor.