Introduction:  An effective HIV-1 vaccine that prevents transmission likely requires the stimulation of envelope (Env) antibodies capable of binding and neutralising a broad range of cell-free infectious virus particles, and to also stimulate antibodies that direct cellular-effector functions, such as antibody dependent cellular cytotoxicity (ADCC) and antibody dependent phagocytosis (ADP), to eliminate HIV cell-cell transmission. Authentic structured Env trimers have superior structural properties to elicit these antibodies; it is unclear, however, if more flexible uncleaved Env gp140 trimers are superior to cleaved SOSIP gp140 trimers for triggering, in one hand, neutralising activity by binding to epitopes such as the CD4 binding site (CD4bs) found on cell-free virus, and in other hand, antibodies targeting the CD4-induced epitopes present on cell-associated Env and utilized in ADCC or ADP.  Vaccination of bovines with HIV-1 Env trimer vaccines have demonstrated unusual capability to produce these useful antibodies (Kramski et al., 2012; Sok et al., 2017). Here we compared HIV Env gp140 vaccines with different primary amino acid sequence, different glycosylation levels and different structural rigidity for their capacity to elicit antibodies with broad cross-clade binding and neutralising activity, especially targeting the CD4bs, and antibodies directing cell-associated immune functions.

Methods: A group of 32 cows were vaccinated with HIV-1 Envs from clades A, B and C before conception, and boosted during pregnancy in a period of 59 weeks. The Env binding activity was determined by direct and by antigen capture ELISA. The neutralising activity of bovine antibodies was assessed against a panel of pseudoviruses in TZM-bl or CF2³ cells. Fc gamma receptor dimer ELISA was performed to investigate the interaction of bovine antibodies with human gamma receptors in immune cells. ADP-SHIP assay was also performed on high Fc receptor binder bovine antibodies.

Results: Analysis of sera pre- and post-vaccination and colostrum showed high titres of HIV Env-specific antibodies induced by the vaccination. These antibodies are IgG and demonstrate high binding affinity for their autologous envelope vaccine, and especially, cows that received clade B uncleaved Env gp140 exhibit binding breadth. These IgG targeted the CD4bs, V2 and V3 loops within Env. Bovine antibodies elicited by the clade A SOSIP Env gp140 vaccine yielded the highest neutralising activity against heterologous viruses. Neutralising activity against B-clade Env was also elicited by the more highly glycosylated uncleaved AD8 Env gp140, compared to a low glycosylated Env (PSC89), but this vaccine also induced IgG with the highest binding to FcgRIIa, mediating phagocytosis of beads coated with Env gp140 on the surface by THP1 cells.

Discussion: The analysis of different HIV-1 Env protein vaccines showed that, although the antibodies elicited with clade A SOSIP cleaved Env gp140 had lowest binding to HIV Env proteins, they could neutralise HIV pseudoviruses strongly. On the other hand, high binding of antibodies from cows vaccinated with the clade B, conformationally flexible, uncleaved Env gp140 trimers, did not guarantee efficient HIV-1 neutralisation, but these exhibited strongest binding to human FcgRIIa receptors on monocytes explaining the ADP activity in the THP1 model.

Conclusions: An optimal vaccine for stimulating antibody immunity to HIV may require both flexibility to present the Env epitopes triggered by interaction with CD4 on the surface of infected cells, and an elaborated structured neutralising epitopes present on cell-free virion.