The chemokine receptor CCR5 is essential for HIV infection and is thus a potential target for vaccine development. However, because CCR5 is a host protein, generation of anti-CCR5 antibodies requires the breaking of immune tolerance and thus carries the risk of autoimmune responses. In this study, performed in mice, we compared 3 different immunogens representing surface domains of murine CCR5, 4 different adjuvants, and 13 different immunization protocols, with the goal of eliciting HIV-blocking activity without inducing autoimmune dysfunction. In all cases the CCR5 sequences were presented as fusions to the Flock House virus (FHV) capsid precursor protein. We found that systemic immunization and mucosal boosting elicited CCR5-specific antibodies and achieved consistent priming in Peyer's patches, where most cells showed a phenotype corresponding to activated B cells and secreted high levels of IgA, representing up to one-third of the total HIV-blocking activity. Histopathological analysis revealed mild to moderate chronic inflammation in some tissues but failed in reporting signs of autoimmune dysfunction associated with immunizations. Antisera against immunogens representing the N terminus and extracellular loops 1 and 2 (Nter1 and ECL1 and ECL2) of CCR5 were generated. All showed specific anti-HIV activity, which was stronger in the anti-ECL1 and -ECL2 sera than in the anti-Nter sera. ECL1 and ECL2 antisera induced nearly complete long-lasting CCR5 downregulation of the receptor, and especially, their IgG-depleted fractions prevented HIV infection in neutralization and transcytosis assays. In conclusion, the ECL1 and ECL2 domains could offer a promising path to achieve significant anti-HIV activity in vivo.
IMPORTANCEThe study was the first to adopt a systematic strategy to compare the immunogenicities of all extracellular domains of the CCR5 molecule and to set optimal conditions leading to generation of specific antibodies in the mouse model. There were several relevant findings, which could be translated into human trials. (i) Prime (systemic) and boost (mucosal) immunization is the best protocol to induce anti-self antibodies with the expected properties. (ii) Aluminum is the best adjuvant in mice and thus can be easily used in nonhuman primates (NHP) and humans. (iii) The Flock House virus (FHV) system represents a valid delivery system, as the structure is well known and is not pathogenic for humans, and it is possible to introduce constrained regions able to elicit antibodies that recognize conformational epitopes. (iv) The best CCR5 vaccine candidate should include either extracellular loop 1 or 2 (ECL1 or ECL2), but not N terminus domains.