Anti-idiotypic antibody Ab2/3H6 mimics the epitope of the neutralizing anti-HIV-1 monoclonal antibody 2F5

Anti-idiotypic antibodies directed against potently neutralizing anti-HIV-1 antibodies may mimic epitopes on gp41 otherwise cryptic to the immune system. This study reports the generation of murine monoclonal antibody Ab2/3H6 blocking the binding of human Ab1 2F5 to the synthetic epitope and to gp160 in an enzyme-linked immunosorbent competition assay. Ab2/3H6 diminished the neutralizing potency of 2F5 in an in-vitro neutralization assay. Ab2/3H6 Fab fragments were capable of inducing neutralizing immune and 2F5-specific responses in B6D2F1 mice applying a simple prime-boost regimen of immunization. Two monoclonal antibodies 2F5 [1,2] and 4E10 [3] recognize epitopes on amino acids 651–657 and 661–680 of gp41. Both are broadly cross-clade neutralizing HIV-1. The synthetic peptide T20, which carries the epitope of 2F5 has been shown to prevent the entry of HIV-1 efficiently [4]. Surprisingly, both monoclonal antibodies (mAb), 2F5 and 4E10 even more show the synergistic effects of in-vitro neutralization when combined with T20. The region recognized by these antibodies is considered to have a major function for HIV entry, which is enabled by conformational changes that occur subsequent to binding to the CD4 cell receptor. The epitopes of these antibodies are cryptic to the humoral immune response, as has been revealed by extensive screening of sera from HIV-positive individuals. We have therefore chosen the anti-idiotypic antibody (Ab2) approach using 2F5 as the first immunogen. Female BALB/c mice were immunized and boosted twice with 2F5-Fab′ in incomplete Freund's adjuvant. Spleen cells were fused with X63-Ag-8.653 myeloma cells and the supernatants of several hybridomas were initially screened for mouse IgG production and idiotype binding by enzyme-linked immunosorbent assay (ELISA). One anti-idiotypic antibody, Ab2/3H6, significantly inhibited the binding of 2F5 to its synthetic epitope ELDKWA in equimolar application. Similarly, Ab2/3H6 preincubated with 2F5 in an in-vitro neutralization assay decreased the potency of 2F5 (Fig. 1). This assay was performed with HIV-1 strain NL4-3 infected AA-2 cells (NIAID, Bethesda, MD, USA) as indicator cells using p24 antigen as the virus replication marker [5]. 2F5 and another HIV-1 neutralizing mAb, 2G12, alone neutralize 90% of the virus at a concentration of 3 μg/ml and 99% of the virus at 12 μg/ml and 11 μg/ml, respectively. The 2F5 and 2G12 concentration was kept constant at 10 μg/ml in this experiment. Aliquots of 50 μl of 2F5 or 2G12 were incubated with serial dilutions of the Ab2/3H6 or unspecific mouse Ig. Ab2/3H6 was able to deplete the neutralization of 2F5 in a dose-dependent manner.Fig. 1.: The anti-idiotypic antibody Ab2/3H6 reduced neutralization of 10 μg/ml 2F5 by 95% at a concentration of 3.9 μg/ml. 3H6 is not able to reduce the neutralization of 2G12 at concentrations as high as 250 μg/ml as evidence of its specificity. —&◆— 2F5 + 3H6; —⋄— 2F5 + mouse IgG; —▴— 2G12 + 3H6; —▵— 2G12 + mouse IgG.We then addressed the question of whether Ab2/3H6 is capable of inducing 2F5-like antibodies in mice. Three B6D2F1 mice were immunized with 3H6 Fab′ in incomplete Freund's adjuvant. Serum samples of immunized mice were screened by ELISA in a peptide epitope binding study as well as in a 2F5 competition assay. All three sera specifically recognized gp160 and the peptide epitope with a serum cut-off value of 1 : 1260 (data not shown). Evidence for 2F5-like antibodies generated through Ab2/3H6 immunization was further demonstrated in a competition assay. ELISA binding of a constant amount of 2F5 was reduced with increasing amounts of the mouse sera as shown in Fig. 2 in more detail.Fig. 2.: The reduction of 2F5 binding to the precoated gp160 in terms of competition with the Ab3 serum samples. Gp160 was immobilized at the enzyme-linked immunosorbent assay plate, and a constant amount of 2F5 was incubated with increasing amounts of the serum samples. The 50% binding inhibition was defined as that serum dilution at which 2F5 binding reached half the maximum of initial optical density (1 : 70 at serum 1 and 1 : 100 at serums 2 and 3). —&◆— Serum 1; —▪— serum 2; —▴— serum 3; uu.uu.uu.uu.uu. – uu.uu.uu.uu.uu. 2F5 (25 ng/ml).In addition, the neutralizing properties of these murine sera were tested with HIV-1 laboratory strain RFNT. Those sera that most potently competed with 2F5 for binding to the antigen also exhibited in-vitro neutralization potency. In this test a virus titre of 102.75was applied. Serum obtained from mouse 1 did not show antiviral effect. Sera 2 and 3 reached an IC50 at dilutions of 1 : 14.1 and 1 : 11.9, respectively. To our knowledge this represents the first in-vivo experiment utilizing an anti-idiotypic anti-HIV-1 antibody, Ab2β, successfully mimicking a neutralizing epitope that is either not or is only poorly immunogenic on native HIV-1 during natural infection. In earlier studies the core epitope motif ELDKWA of 2F5 was integrated into different antigenic formats, such as on the haemagglutinin of live influenza [6] or fused to the hepatitis B surface antigen expressed in yeast [7]. However, the sera of such immunized animals, although leading to serum IgG binding to the ELDKWA motif, did not show any significant in-vitro neutralization potency. Peptide versions of the 2F5 epitope were also poorly immunogenic. In contrast, sera of Ab2/3H6-induced immune responses inhibited 2F5 binding to its antigen as well as showing significant in-vitro neutralization potency. We conclude that the anti-2F5-idiotypic Ab2/3H6 per se bears the potential as a vaccine candidate capable of inducing protective and neutralizing immune responses. Consequently, the Ab2β approach can be considered as a tool contributing to the design of an HIV-1 vaccine that also successfully induces protective and neutralizing B cell responses. Our own follow-up studies with Ab2/3H6 will be focussed on two topics. First, we will try to induce humoral neutralizing immune responses in non-human primates. Second we will chimerize Ab2/3H6 with human Fc or engineer other immunogenic molecules with Ab2/3H6 specificity. The antibodies 2F5 and 4E10, in addition to the characters described above, exhibit significant synergy of neutralization when applied in combinantion although their core epitopes have been mapped in close vicinity on a linear peptide [3]. We are curious to investigate whether Ab2β can be induced (or alternatively screened with libraries) that are capable of mimicking the epitopes of both antibodies and to represent a hyperantigenic neutralizing vaccinal candidate. Renate E. Kunerta Robert Weikb Boris Ferkob Gabriela Stieglera Hermann Katingera,b Acknowledgement The authors would like to thank Sabine Necina, Willibald Steinfellner and Azemina Zudjelovic for technical support.