After incubation, the plate was washed, and goat anti-human IgG-Fc antibody was added

lizing Fab from the same affinity matured series differing in only 4 positions in the CDR-H2 loop, complexed to two mimetics of the pre-haipin intermediate revealed only subtle structural differences between complexes with the neutralizing and nonneutralizing Fabs. Moreover, despite differences in neutralizing activity of three or more orders of magnitude in Env-pseudotyped virus neutralization assays, the binding UPF 1069 web affinities of the neutralizing and non-neutralizing Fabs to various pre-hairpin intermediate mimetics differ by little more than an order of magnitude. This suggested to us that the ability to bind to the six-helix bundle, in addition to the exposed N-HR trimer of the pre-hairpin intermediate, may be critical to the neutralization activity of this particular series of Fabs. Interestingly, alanine scanning mutagenesis and Western blot analysis showed that this series of Fabs targets a structurally contiguous epitope on the N-HR that is solvent accessible and located in a shallow groove between two C-HR helices in the six-helix bundle . The same residues of the N-HR also comprise part of the binding site in the pre-hairpin intermediate mimetics . However, if the Fabs were to bind to the six-helix bundle in the same mode as seen in the crystal structures with the prehairpin intermediate mimetics, there would be steric clash and atomic overlap between the CDR-H1 and CDR-H2 loops of the Fab and one of the C-HR helices of the six-helix bundle. The simplest hypothesis for these observations is that binding of this series of Fabs to the six-helix bundle involves displacement of one of the C-HR helices. Here we investigate the binding of the 8066 and 8062 monoclonal antibodies, as well as of various point mutants within the CDR-H2 loop of 8066, in Fab and single chain-variable fragment formats to mimetics of the pre-hairpin intermediate and six-helix bundle conformation. We show that binding to the six-helix bundle does not involve displacement or fraying of a C-HR helix, indicating that these antibodies must interact somewhat differently with the six-helix bundle and the pre-hairpin intermediate mimetics, although they share a common epitope. Further, we show that there are large differences in binding affinity of neutralizing and non-neutralizing antibodies within this series to the six-helix bundle. Results and Discussion Binding affinities of Fab8066 and Fab8062 to pre-hairpin intermediate mimetics Both the neutralizing Fab8066 and the non-neutralizing Fab8062 recognize an epitope at the C-terminal end of the NHR timer of gp41 comprising a hydrophobic pocket formed at the interface of two N-HR helices . The relative binding affinities for Fab8066 and Fab8062 to various prehairpin intermediate mimetics, as determined by isothermal titration calorimetry, differ by factors ranging from 20-fold to only 2-fold. In the case of 5-helix, a single chain pre-hairpin mimetic that presents a single antibody binding site PubMed ID:http://www.ncbi.nlm.nih.gov/pubmed/19661801 in which the N-HR trimer is surrounded by two C-HR helices, the binding affinities differ by,20 ; for N35CCG-N13, an N-HR trimer construct stabilized by disulfide bridges that presents three antibody binding sites, the binding affinities differ by,8; and for CCIZN36, an N-HR trimer construct stabilized by a disulfide-linked leucine zipper, which also presents three antibody binding sites, the binding affinities differ by a factor of only 2 . This variation might be attributable to differences in conformational flexibility of th