Abstract
Background Drug resistance mutations are commonly found in HIV protease (PR) and result from many factors allowing specific mutations to predominate, eg, single base pair changes resulting in functional yet drug-resistant enzymes. Drug exposure drives evolution by selecting for energetically stable and functional proteins. Flap region (residues 36-63) mutations in PR are of particular interest because they are distal from the active site and as they accumulate contribute significantly to resistance while preserving enzymatic function. The structural and protein dynamical aspects of how this occurs are poorly understood. We hypothesize that the flap is stuck in a partially open conformation in the resistant forms, which may improve protein stability even in the absence of bound PI while simultaneously impeding PI binding. We are developing a new structural biology technique, Raman crystallography, to study PR flap mutations. We present Raman spectroscopic and x-ray crystallographic data showing how the Phe 53 flap residue can be used to determine the flap position.
Methods “Wild-type” PR was crystallized and used in our experiments. Nonresonance Raman difference spectra were obtained with indinavir by soaking the crystals in inhibitor solution for 10 minutes. Crystals for x-ray crystallography were prepared as for Raman, and 2-3 Å resolution data obtained for the apo-structure and the indinavir-inhibited structure. Structures were solved by molecular replacement. Ab initio calculations were used to model the Raman spectra.
Results The intensity of the peaks in the Raman spectra is very sensitive to the orientation of the crystals. The phenylalanine transitions at ≈1,000-1,008 cm-1 in the difference spectra are especially informative. The frequency and intensity of these peaks are signatures of open—>closed conformational changes in the Phe53s on binding of inhibitor. There is also a contribution from the methylphenyl side chain of the bound inhibitor. Other PI transitions are also sensitive to crystal orientation. The preliminary crystal structure data indicate that the flap region is highly disordered in the apo-crystal (some open, some closed) and shows bound inhibitor when it is added. The Phe 53 orientation is still being refined.
Conclusions The ≈1,000-1,008 cm-1 region in Raman spectra of PR can be used to assess the flap conformation in WT PR on PI binding. Peak widths and intensity can give information of the stability of the complexes. We will study resistant PR next. Raman spectroscopy is a powerful tool to study flap mutations in PR.