Background:  AMD11070 is a first-in-class, bioavailable small-molecule antagonist against the HIV co-receptor CXCR4. An ongoing safety and dose-finding study has demonstrated the activity of AMD11070 in HIV-infected patients harboring X4-tropic virus. This study aims to understand the interactions of AMD11070 with CXCR4 at the molecular level.

Methods:  Site-directed mutagenesis was employed to generate mutated forms of the hCXCR4 receptor that carry single amino acid substitutions. The mutated hCXCR4 receptors were expressed transiently in a canine thymus cell line (Cf2Th). Calcium mobilization and ¹²⁵I-SDF-1a homologous competition binding experiments were used to characterize the SDF-1a signaling response and binding affinity respectively of the mutated hCXCR4. In addition, ¹²⁵I SDF-1a competition binding experiments were performed using these recombinant Cf2Th cell lines to examine how the single site mutations affected the binding of AMD11070. Molecular modeling was used to illustrate the experimental results.

Results:  We generated 14 single-site mutants that span across the entire putative small-molecule antagonist binding site of CXCR4 receptor. The SDF-1a EC₅₀s in calcium flux signaling and IC₅₀ in homologous competition binding of these mutants were within 3- to 4-fold as compared to those of the wild type receptor. The mutations affected the binding of AMD11070 at various degrees. Among the hCXCR4 mutants studied, 4 of them (W94A, D97N, D171N, and E288A) decreased the potency of AMD11070 binding by >100-fold, while 2 of them (Y45A and D262N) decreased the potency by 10- to 50-fold. This study also identified an amino acid residue (D97) on the receptor that interacts specifically with AMD11070, but not its predecessor AMD3100. Comparing with previously published data, these findings suggested that AMD11070 and AMD3100 share partially overlapping binding sites.

Conclusions:  Binding studies using single site mutants of hCXCR4 showed that AMD11070 interacts in a similar, but distinct mode from its bicyclam predecessor AMD3100. This study suggests that it is possible to develop antiretroviral CXCR4 antagonists that have different binding mechanisms to the HIV co-receptor. These mechanistic studies might prove to be useful for the development of other structurally diverse CXCR4 antagonists with improved clinical pharmacology and safety profiles.