Selecting HIV-1 Neutralizing mABs from Californian Donors

James Binley, Torrey Pines Institute for Molecular Studies

Neutralizing antibodies (nAbs) are likely to be an essential component of a successful AIDS vaccine. Neutralizing monoclonal antibodies (mAbs) provide paradigms for the kind of nAbs researchers aim to induce by vaccines. However, efforts to recover neutralizing mAbs from patients by many investigators for more than two decades have been met with limited success and very few broadly neutralizing monoclonal nAbs have been isolated. A fraction of patients generate measurable titers of Abs that are able to neutralize many primary isolates. An alternative approach to understanding nAbs is to map the neutralization patterns in these rare plasmas. It would be of great interest to know whether neutralization traces to Abs similar to the handful of known monoclonal nAbs, or other nAbs that defy such categorization. In the past, similar studies have been inconclusive perhaps because of the confounding dominance of non-neutralizing Abs present in HIV+ patient plasmas. However, we have recently developed novel methods to begin to map nAb specificities (see Appendix 1). Our Specific Aim is to try to isolate novel neutralizing mAbs by selecting them from combinatorial phage-displayed Ab libraries prepared from Californian HIV+ donors. Selection of these libraries with subunits of HIV Env like gp120 has so far retrieved neutralizing mAbs b12 (selected against monomeric gp120) and Z13 (selected against a gp41 fragment), but the vast majority of mAbs have been non-neutralizing. Soluble Env fragments therefore do not appear to effectively differentiate neutralizing and non-neutralizing Abs. The selection problems may also stem in part from the scarcity of nAbs in humans amid a background of non-neutralizing Abs. Because nAbs are distinct in their ability to bind to Env trimers, it should be possible to select libraries against native Env trimers in the form of infection- competent virus-like particles (VLPs) captured on ELISA plates. However, there are three technical problems that need to be overcome: 1) insufficient Env antigen to enrich the Env-binding fraction of phage; 2) non-functional forms of Env that bind phage-Abs, irrespective of their neutralizing activity; 3) non-Env membrane protein contaminants which may enrich phage-Abs that do not bind to Env. Therefore, we propose various modified phage selection methods involving native Env proteins to ensure that selection is tied to neutralization. In pilot experiments, we will screen mixtures of known mAb- phage clones of differing neutralizing activities. Here, the b12-phage derives from a well known neutralizing mAb and the b6-phage derives from a prototype non-neutralizing mAb. We will optimize selection protocols for b12-phage enrichment before we procede with screenings of whole phage libraries prepared from Californian donors. Using established protocols, we will then prepare soluble Fab fragments and determine their neutralization and epitope reactivities. New monoclonal nAbs would be a significant breakthrough and a good basis of an NIH-funded proposal to apply the information to vaccine development by tailoring immunogens able to elicit these nAbs. Considering how useful the few existing neutralizing mAbs have proved to be, any new neutralizing mAb would also open up extensive collaborative opportunities and may be worth evaluating as post-exposure prophylactic and therapeutic agents.