Research Project
8420190
DESCRIPTION (provided by applicant): HIV elicits the production of antibodies in the infected individual. Whereas for most, this response is highly specific for the particular viral isolate of infection, there are some individuals that mount broadly cross neutralizing antibodies that appear to keep the virus in check. It is assumed that such cross neutralizing activity is central i combating AIDS and would be a necessary attribute of a prophylactic AIDS vaccine. Therefore, it is highly desirable to be able to critically analyze the antibody response towards HIV and profile the specificities correlated with virus neutralization. Current serological methods, however, are tedious, laborious and of unsatisfactory resolution, classifying antibody specificities to gross chunks of the HIV envelope. Here we propose a novel and much improved technological platform that generates exceptionally high resolution profiles of the antibody specificities of polyclonal serum with relative ease, low cost and at high throughput. The essence of the technology is the combination of two new methods; a phage display library of a comprehensive array of HIV envelope peptides/domains ranging in length from 50 to 300 amino acids, and a unique vector system that converts phage display compatible with second generation DNA cyclic sequencing. Together, this methodology, coined Deep Panning of Domain Scan libraries, enables one to analyze the complete antibody repertoire of selected polyclonal sera. The data output generated consists of hundreds of thousands of affinity isolated peptides corresponding to HIV envelope segments. In this study feasibility of the methods and evaluation of their utility and productivity will be tested using a panel of >400 well characterize serum samples of HIV infected individuals that range in their neutralizing capacity from non-existent/weak to extraordinarily potent (top 2% of the collection). Selected sera will be subjected to Deep Panning and the data sets to be generated will be classified implementing machine learning computational methods. These will produce a profile of those epitopes that best correlate with HIV neutralization. It is proposed that this information will be extremely useful fo our understanding of mechanisms of HIV neutralization and as guides to rational vaccine design. Moreover, the proof of feasibility in this pilot study will serve as the basis for more elaborate and comprehensive studies targeted to better understand HIV correlates of immunity, disease progression and prevention of AIDS.
