We have established that heat-killed Brucella abortus conjugated to a V3- loop peptide from HIV-1 MN can elicit neutralizing antibodies and CTL in normal and mice depleted of CD4+ T-cells, systemically and at mucosal surfaces. The antibodies elicited in mice lacking CD4+ T-cells were mainly of the IgG2a isotype. Sera from these mice were more efficient than sera containing predominantly IgG1 in neutralizing HIV-1. Furthermore this conjugate can induce neutralizing antibody (systemic and mucosal) and cellular responses in Rhesus macaques. Secretions from several mucosal sites including vagina exhibited ant-HIV neutralizing activity. To determine whether IgG subclasses differ in their ability to neutralize HIV-1, polyclonal HIVIG was passed over protein A columns and the appropriate fractions collected and shown to be > 89% pure IgG1, IgG2 and IgG3 (IgG4 was not present in sufficient quantities for study). All three subclasses showed binding to all major HIV-1 proteins (IgG1 >IgG2 > IgG3). In contrast, IgG3 was more active than other subclasses in its ability to neutralize HIV-1 as assayed by syncytia inhibition and cell- free virus neutralization (IgG3 > IgG1 > IgG2). The ability of HIVIG and the subclasses to neutralize viruses also depended on the virus strain, so that T cell tropic (R4) were more susceptible than Macrophage tropic (R5) viruses. IgG3 differs from IgG1 and IgG2 by virtue of a longer and more flexible hinge region. To test whether this is important in neutralizing virus, the IgG fractions will be digested by papain into Fc and Fab fragments and the Fab fragments tested for activity. These findings have implications for manufacture and clinical use of HIVIG. Moreover they suggest that HIV-1 vaccines should induce high titer IgG3 responses.
