Gln151 of HIV-1 Reverse Transcriptase Acts as a Steric Gate Towards Clinically Relevant Acyclic Phosphonate Nucleotide Analogues

Abstract

Background

In the treatment of HIV, the loose active site of the HIV-1 reverse transcriptase (RT) allows numerous nucleotide analogues to act as proviral DNA ‘chain-terminators’. Acyclic nucleotide phosphonate analogues (ANPs) represent a particular class of nucleotide analogue that does not possess a ribose moiety. The structural basis for their substrate efficiency regarding viral DNA polymerases is poorly understood.

Methods

Pre-steady-state kinetics on HIV-1 RT together with molecular modelling, were used to evaluate the relative characteristics of both the initial binding and incorporation into DNA of three different ANP diphosphates with progressively increasing steric demands on the acyclic linker: adefovir-diphosphate (DP), tenofovir-DP, and cidofovir-DP.

Results

The increase of steric demand in ANPs induced a proportional loss of the binding affinity to wild-type HIV-1 RT (K _d cidofovir-DP>>K _d tenofovir-DP>K _d adefovir-DP∼K _d dNTPs), consistent with the lack of HIV-1 inhibitory activity for cidofovir. We show that, starting from adefovir-DP, the steric constraints mainly map to Gln151, as its mutation to alanine provides cidofovir-DP sensitivity. Interactions between the Gln151 residue and the methyl group of tenofovir-DP further increase with the mutation Gln151Met, resulting in a specific discrimination and low-level resistance to tenofovir-DP. This alteration is the result of a dual decrease in the binding affinity (K _d ) and the catalytic rate (k _pol ) of incorporation of tenofovir-DP. By contrast, the tenofovir resistance mutation K65R induces a broad ‘k _pol -dependent’ nonspecific discrimination towards the three ANPs.

Conclusions

Overall, our results show that the efficiency of ANPs to compete against natural nucleotides as substrates for RT is determined by their close interaction with specific amino acids such as Gln151 within the RT active site. These results should help us to map and predict ANP sensitivity determinants in cellular and viral DNA polymerase active sites for which the understanding of different ANP sensitivity patterns are of medical importance.