Nanoscale imaging of cell surface HIV ENV immune complexes to optimize NK cell adcc antibody therapeutics

The National Institute of Allergy and Infectious Diseases awarded San Diego Biomedical Research Institute $992,920 on September 1, 2026, under the Allergy and Infectious Diseases Research program (CFDA 93.855) for research on nanoscale imaging of cell surface HIV envelope immune complexes to optimize natural killer cell antibody-dependent cellular cytotoxicity (ADCC) therapeutics.

The project integrates nanoscale imaging, machine learning, and in silico protein modeling to characterize the spatial organization and conformational state of envelope immune complexes and natural killer cell receptor coordination, with the goal of designing de novo antibodies with enhanced ADCC activity. Current antibody-based HIV therapies are limited by viral evasion strategies and incomplete understanding of how antibody stoichiometry and immune complex geometry influence natural killer cell activation. The research addresses gaps in knowledge that currently prevent rational engineering of potent ADCC activity into therapeutic antibodies, including broadly neutralizing antibodies in clinical trials. The institute will employ machine learning algorithms trained on structural and functional data to accelerate antibody development beyond traditional monoclonal antibody isolation and library screening methods.

Performance occurs in San Diego, California, with a period of performance from September 1, 2026, through August 31, 2031. The assistance type is a Project Grant.

Abstract

Natural killer (NK) cell-mediated antibody-dependent cellular cytotoxicity (adcc) is a powerful mechanism by which the immune system can clear HIV-1-infected cells and drive the long-term suppression of viral reservoirs, likely supported by the early antibody-based immune response. Despite promising advances in antibody-based therapies, NK cell activation and clearance of HIV infected cells remains limited by viral evasion strategies. The spatial organization and conformational state of envelope (ENV) immune complexes (ICS), along with NK cell receptor coordination, are critical determinants of adcc efficacy, yet remain poorly understood at the molecular level. We have shown that antibody stoichiometry substantially influences NK cell adcc. Further, we provide preliminary data that demonstrates that IC geometry can significantly alter NK cell adcc activity. However, our limited understanding of the basis of NK cell adcc abrogates our ability to predictably engineer potent in vivo adcc activity into antibody therapeutics, including broadly neutralizing antibodies (bnabs) currently in clinical trials. We have previously demonstrated that integrating machine learning (ML) with in silico protein modeling can enable the development of novel protein designs, including nanobodies. Rapid development in the artificial intelligence (AI) field has shown that such designs could be used to build full antibodies DE novo. By training new algorithms on structural and functional data, we can accelerate the production of new antibodies that go beyond traditional methods, such as monoclonal antibody isolation or library screening using biological assays. This proposal will address key gaps in our knowledge of NK cell adcc using integrated multidisciplinary expertise, building on promising preliminary data toward designing DE novo antibodies with enhanced adcc activity. In aim 1, we will evaluate the NK cell adcc activity of a panel of both bnabs and non-neutralizing antibodies that are known to enhance adcc. We will also use minflux nanoscopy to resolve the nanoscale structure of HIV ENV and ENV ICS on the surface of HIV ENV presenting cells. In aim 2, we will build upon our previous work defining the spatial organization of the IGG receptor CD16 within the NK cell synapse to determine how ENV ICS influences the spatial organization of CD16 as well as additional activating receptors within the immune synapse. In aim 3, we will create and validate a ML algorithm capable of accurately predicting the structural and functional properties of HIV-targeting nanobodies. Using this model, we will design new nanobody candidates in silico that target ENV epitopes predicted to elicit enhanced adcc and validate promising candidates. This proposal will define the molecular architecture of the adcc synapse against ENV presenting target cells and use this to create a ML framework linking ENV molecular features to NK cell adcc. Dual innovation in our approach (applying minflux nanoscopy to ENV immunobiology and utilizing cutting-edge ML and protein design to develop novel therapeutics) directly addresses the goals of PAR-25-300 to advance immune-based HIV cure strategies. The resulting insights from our studies will guide subsequent therapeutic antibody designs to target HIV reservoirs.