SummaryHuman immunodeficiency virus (HIV) binds to cells via an interaction between CD4 and the virus envelope glycoprotein, gp120. Previous studies have localized the high affinity binding site for gp120 to the first domain of CD4, and monoclonal antibodies (mAbs) reactive with this region compete with gp120 binding and thereby block virus infectivity and synrytium formation . Despite a detailed understanding of the binding of gp120 to CD4, little is known of subsequent events leading to membrane fusion and virus entry. We describe two new mAbs reactive with the third domain ofCD4 that inhibit steps subsequent to virus binding critical for HIV infectivity and cell fusion. Binding of recombinant gp120 or virus to CD4 is not inhibited by these antibodies, whereas infection and synrytium formation by a number of HIV isolates are blocked . These findings demonstrate that in addition to virus binding, CD4 may have an active role in membrane fusion .C D4 is the high affinity cellular receptor for HIV and is found predominantly on thymus-derived T lymphocytes (1) . The structure predicted for CD4 is an extracellular region of four tandem domains having homology with members of the Ig superfamily, followed by a transmembrane domain and a cytoplasmic region (2, 3) . HIV binds CD4 via the major envelope glycoprotein, gp120, and recent genetic studies have localized the high affinity gp120 binding site on CD4 to -12 amino acids within the NH2-terminal, Ig-like (V1) domain (4-6). Similarly, residues have been defined in the COOH-terminal half of gp120 that are critical for CD4 binding (7,8). It is clear that the interaction between CD4+ human cells and cells expressing HIV envelope is sufficient to induce synrytium formation (9-11), and that virus entry is by pH-independent fusion ofthe virus and cell membranes (12, 13) . Little else is known, however, about the events subsequent to binding that lead to membrane fusion .Recently, evidence from two studies suggests that CD4 may play a role in cell-cell fusion (synrytium formation) additional to envelope binding . In the first study, single residue substitutions made in the CDR-3 analogous loop of the V1 domain of CD4 resulted in mutant molecules that were unable to facilitate synrytium formation, even though they mediated both virus binding and infectivity to an extent comparable with wild-type CD4 (14). In the second, chemical mutagenesis of a CD4+ T cell line resulted in CD4 molecules that bound gp120 but mediated synrytium formation poorly (15). Although the mechanism of inhibition of cell fusion was not elucidated in either study, the findings indicate that CD4 participates in events subsequent to gp120 binding that are required for membrane fusion. mAbs reactive with CD4/V1 are highly efficient at inhibiting the binding of gp120 to CD4, and thereby prevent infection and synrytium formation (6,(16)(17)(18)(19) . In contrast, a mAb reactive with the third or fourth CD4 domains has been shown not to inhibit gp120 binding (18,19) . To investigate whether regions of CD4 oth...

SummaryThe production of class-switched antibodies, particularly immunoglobulin (Ig)G1 and IgE, occurs efficiently in T cell receptor (TCP,)ot-/-mice that are congenitally devoid of or/J3 T cells. This finding runs counter to a wealth of data indicating that IgG1 and IgE synthesis are largely dependent on the collaboration between B and ot/[3 T cells. Furthermore, many of the antibodies synthesized in TCRot-/-mice are reactive to a similar spectrum of self-antigens as that targeted by autoantibodies characterizing human systemic lupus erythematosus (SLE). SLE, too, is most commonly regarded as an ot/[3 T cell-mediated condition. To distinguish whether the development of autoantibodies in TCP, ot-/-mice is due to an intrinsic de-regulation of B cells, or to a heretofore poorly characterized collaboration between B and "non-or/J3 T" cells, the phenotype has been reconstituted by transfer of various populations of B and non-or/J3 T cells including cloned ~//8 T cells derived from TCRcx-/-mice, to severe combined immunodeficient (SCID) mice. The results estabhsh that the reproducible production of IgG1 (including autoantibodies) is a product of non-a/j3 T cell help that can be provided by "y/8 T cells. This type of B-T collaboration sustains the production of germinal centers, lymphoid follicles that ordinarily are anatomical signatures of or/J3 T-B cell collaboration. Thus, non-or/J3 T cell help may drive Ig synthesis and autoreactivity under various circumstances, especially in cases of or/J3 T cell immunodeficiency.

T cells are essential for inducing clonal B cell expansion in germinal centers during T cell-dependent antibody responses. However, class-switched antibodies are readily detectable in TCR alpha-deficient mice that congenitally lack alpha beta T cells, including those such as IgG1 that are considered to be dependent on collaboration between B cells and alpha beta T cells. This observation suggests that a novel form of B:T collaboration may be evident in TCR alpha-/- mice. We report that germinal centers develop spontaneously in mice lacking T cell receptor alpha genes (TCR alpha-/-), despite the absence of alpha beta T cells. They are not seen in TCR beta-/- mice kept in similar conditions. Both strains of mice have gamma delta T cells, but it is a subset of T cells expressing TCR beta and CD4 that is dominant in the germinal centers of TCR alpha-/- mice. Exceptionally, germinal centers were associated with CD4+ gamma delta T cells. The expression of CD4 seems to be important, for few extrafollicular T cells have CD4 and CD4 is largely absent from TCR beta-/- T cells. The CD4+ TCR beta cells may help B cells produce autoantibodies that have been identified in TCR alpha-/- mice.

In mice and humans, T cells are characterized on the basis of T-cell receptor (TcR)

Anti-CD4 antibodies directed to the N terminus of CD4 can inhibit human immunodeficiency virus (HIV) infection. Therefore, it has been proposed that some of these reagents may contain idiotypic determinants which conformationally model the binding site expressed on gp120. In this report, we have selected a panel of anti-CD4 monoclonal antibodies as idiotypic mimics of gp120 by employing cross-blocking techniques, and CD4 epitope mapping using site-directed mutagenesis. These studies suggest that only 4 out of the original panel of 12 would be expected to represent suitable candidates for modelling the gp120 binding site. Nevertheless, anti-idiotypic antisera raised against these antibodies failed to inhibit gp120 binding to CD4. This negative result may reflect the incomplete modelling of the virus binding site by anti-CD4, or the lack of internal image antibody in the anti-idiotypic preparations. Alternatively, the binding site on gp120 may not be accessible to antibody neutralization, excluding the possibility of an idiotypic vaccine to HIV based on anti-CD4 antibody as surrogate antigen.