Background:  CCR5 represents a major chemokine receptor, which R5-HIV exploits in its entry to target cells, thus serving as an attractive target for possible intervention of R5-HIV infection. Aplaviroc (AVC) is a potent experimental CCR5 inhibitor highly potent against R5 HIV. In this study, we examined the structural/molecular interactions of AVC and its analogs including AK317 and AK530 with CCR5. We also determined the relationships between their structures and anti-HIV activity.

Methods:  More than 35 mutant CCR5-expressing cell lines were generated. Using such cells and various ³H-labeled inhibitors, the binding affinity of each inhibitor to CCR5 was determined. CCR5 model was generated based on a crystal structure of bovine rhodopsin. Based on the data obtained, docking model of CCR5 and each inhibitor was generated.

Results: Binding assay and structural modeling revealed that the amino acids important for the binding of CCR5 inhibitors overlap, but substantially differ among CCR5 inhibitors. AK317, which has lower CCR5-binding affinity (K_D: ~16 nM) and less potent antiviral activity (IC₅₀: ~3 nM) than AVC (K_D: ~3.6 nM; IC₅₀: ~0.2 nM), had no interactions with E283, with which AVC had tight interactions. AK530, which has >2-fold greater CCR5-binding affinity (K_D: ~1.5 nM), but less potent antiviral activity (IC₅₀: ~3 nM) than AVC, had a tight association with Y37, with which AVC had no interactions. Moreover, AK530 had least interactions with K191, with which AVC had a close association. We also found that K191 and certain amino acids (eg, G163, T195), located in CCR5’s upper trans-membrane domain, form a tight hydrogen bond network and AVC is closely associated with the network, while AK530 is not. The tight interactions of AVC with residues involved in distinct hydrogen bond networks most likely cause targeted changes in the conformation of CCR5’s extra-cellular loop (ECL). Thus, not only the binding affinity of the compound but also the mechanism of allosteric changes of ECL should determine the potency of CCR5 inhibitors against R5-HIV.

Conclusions:  Structural modeling analysis combined with the data of CCR5 binding affinity of CCR5 inhibitors should help understand their structural/molecular interactions and help more rationally design more potent CCR5 inhibitors. The present data strongly suggest that not only the binding affinity of CCR5 inhibitors but the resultant allosteric changes of ECL determine antiviral potency of CCR5 inhibitors.