A study of the molecular mechanism of binding kinetics and long residence times of human CCR5 receptor small molecule allosteric ligands

Abstract

Background and purpose: The human CCR5 receptor is a co-receptor for HIV-1 infection and a target for anti-viral therapy. A greater understanding of the binding kinetics of small molecule allosteric ligand interactions with CCR5 will lead to a better understanding of the binding process and may help discover new molecules that avoid resistance.

Experimental approach: Using [(3) H] maraviroc as a radioligand, a number of different binding protocols were employed in conjunction with simulations to determine rate constants, kinetic mechanism and mutant kinetic fingerprints for wild-type and mutant human CCR5 with maraviroc, aplaviroc and vicriviroc.

Key results: Kinetic characterization of maraviroc binding to the wild-type CCR5 was consistent with a two-step kinetic mechanism that involved an initial receptor-ligand complex (RA), which transitioned to a more stable complex, R'A, with at least a 13-fold increase in affinity. The dissociation rate from R'A, k-2 , was 1.2 × 10(-3) min(-1) . The maraviroc time-dependent transition was influenced by F85L, W86A, Y108A, I198A and Y251A mutations of CCR5.

Conclusions and implications: The interaction between maraviroc and CCR5 proceeded according to a multi-step kinetic mechanism, whereby initial mass action binding and later reorganizations of the initial maraviroc-receptor complex lead to a complex with longer residence time. Site-directed mutagenesis identified a kinetic fingerprint of residues that affected the binding kinetics, leading to the conclusion that allosteric ligand binding to CCR5 involved the rearrangement of the binding site in a manner specific to each allosteric ligand.

Keywords: CCR5; HIV-1 infection; allosteric; binding kinetics; maraviroc; molecular mechanism of action; pharmacology.

Figure 1

Kinetic mechanisms. Both models are based upon the assumption that receptor binding corresponds to a simple reversible bimolecular process that obeys the law of mass action. Bound receptors can only adopt a single state/conformation in the ‘Single-step bimolecular binding process’. In the more complex ‘Two-step bimolecular binding process’, the initial binding step, yielding a loosely bound complex RA, is followed by a reversible isomerization step yielding a more tightly bound complex R'A. The rate constants that are associated with this latter model are usually defined as ‘microscopic’ and, to avoid potential confusion, they are denominated differently from those that are associated to the first model.

Figure 2

Potency and kinetics of MVC binding to human WT CCR5. (A) Saturation curve of [³H]-MVC binding to human WT CCR5 during 3 h reaction at room temperature. Specific binding was determined by subtracting non-specific binding from total binding. The r² for fitting to a one site hyperbolic function for specific binding was 0.993. Binding of [³H]-MVC shown in the ensuing panels is the value for specific binding. (B) Time dependence of association of [1–40 nM] [³H]-MVC with CHO cell membranes containing WT CCR5, the k_obs at each concentration was determined by fitting to a single exponential. (C) Kinetic replot of k_obs versus [³H]-MVC concentrations for the WT CCR5 and mutants Y108A and F109A. Kinetic constants were determined from the relationship k_obs = k_on·[[³H]-MVC] + k_off. As shown in the Supporting Information, the relationship between k_obs and [[³H]-MVC] is expected to be linear provided that RA state is sufficiently stable to survive the short wash step at the end of the incubation. (D) Pre-incubation of WT CCR5 with [³H]-MVC for 3 h followed by wash-out with 1 μM of unlabelled MVC. Data were fitted to a single and two-phase exponential decay model. (E) Effect of pre-incubation on the dissociation kinetics of MVC. Unlabelled MCV was pre-incubated for 2 min, 30 min and 2 h and dissociation measured continuously after the addition of [³H]-MCV using the SPA format in the Leadseeker. ‘Control’ refers to the occupancy of CCR5 by MVC at the onset of the washout. (F) ‘Kinetic shift’ protocol for semi-quantitative estimation of the dissociation rate of unlabelled MVC from WT CCR5. CCR5 were incubated with range of unlabelled MVC concentrations along with 4 nM [³H]-MVC (no pre-incubation) or 3 h prior (pre-incubation) to addition of 4 nM [³H]-MVC. After 2 h at room temperature, the amount of receptor bound [³H]-MVC was determined by filtration.

Figure 3

Kinetics of MVC interaction with CCR5 Y108A. Binding of [³H]-MVC in is specific. (A) Time dependence of association of [1–40 nM] [³H]-MVC with CHO cell membranes containing human CCR5Y108A. The k_obs at each concentration was determined by fitting to a single exponential. (B) ‘Kinetic shift’ protocol for semi-quantitative estimation of the dissociation rate of unlabelled MVC from human CCR5-Y108A. The experimental procedure and symbols are the same as outlined in the legend of Figure 2F.

Figure 4

Visual representation of kinetic shifts by different antagonists for WT and mutated CCR5. The ratio of K_I/K_I* quantitates the shift observed upon 3 h pre-incubation and provides a semi-quantitative evaluation of the unlabelled allosteric ligand's residence time. K_I and K_I* values are listed in Table 1. HS, hot spot.