1) Structure/function studies of the HIV Env trimer (collaboration with Dr. Paolo Lusso, NIAID). Extending our findings that V1V2 masking of the V3 loop of gp120 occurs by an intraprotomer (cis) rather than interprotomer (trans) mechanism, we are pursuing a proposal for a novel gp120 conformation that we believe represents the unliganded structure in the native trimer prior to CD4 binding. Our evidence comes from a combination of studies with monclonal antibodies and synthetic peptides based on newly predicted sites of intramolecular interactions, as well as gp120 and CD4 constructs (soluble and membrane anchored forms) with engineered mutations in the regions of interest. Molecular dynamics analyses provide support the new model. 2) Immunotoxiins for targeted killing of HIV-infected cells. We have initiated collaborations using in vivo models to test the ability of immunotoxins directed against HIV-1 Env expressed on the surface of productively infected cells to deplete infected cell reservoirs persisting during ART. In collaboration with Dr. Tom North (recently moved from U. C. Davis to Emory U.), we are examining the effects of sCD4-PE40 in SIVmac239-infected rhesus macaques whose viral loads have been suppressed below detectable levels by combination ART;the goal is to test whether the immunotoxin significantly delays (or, less likely, prevents) the virus rebound typically observed after ART cessation. Analysis of the first phase of this study is near completion. We have also initiated a collaboration with Drs. Victor Garcia-Martinez and Shailesh K Choudhary (U. North Carolina) to test the activity of CD4-PE40 and 3B3-PE38 in HIV-1-infected humanized mice (containing human bone marrow, liver, thymus, i.e. BLT mice);again, we wish to test whether the immunotoxins significantly delay (or prevent) viral rebound after cessation of ART. 3) Enhanced chimeric antigen receptors (CARs) for adoptive transfer of autologous CD8 T cells genetically modified to target HIV-1 Env (collaboration with Drs. Steven Rosenberg and Rick Morgan, NCI). The immunotoxin approach suffers from the inherent problem of immunogenicity due to the foreign moiety of such proteins (bacterial in the case of PE-based agents);thus the immunotoxin treatment is limited to short duration, a major limitation in view of the presence of reservoirs of latently infected cells that can become activated at a later time. Thus a more durable """"""""targeted cell killing"""""""" method would seem preferable. Adoptive trasfer of engineered CD8 T cells represents an exciting option, particularly in view of the emerging successses of this strategy in the cancer filed. This has been tried for HIV infection more than a decade ago;while the transferred engineered T cells were found to populate many anatomical sites where HIV replication occurs, they persisted only at low levels, and no clinical benefits were obtained. We are introduciing two new features to the CAR constructs that designed to improve efficacy of this approach. The first includes enhancement of the intracellular signaling domains by including relevant motifs developed by cancer investigators in the CAR field namely regions of CD28 and 41BB;these are expected to enhance the functional activity and in vivo longevity of the adoptively transferred CD8 cells. Perhaps more importantly are modifications that we have introduced into the extracellular Env-targeting moiety. We have designed a modified variants of the 2-domain extracellualr region of CD4 (designated CD4-M) with two important features compared to the corresponding wild type motif: much higher binding efficiency to surface Env, and inability to support HIV entry into target cells expressing this molecule plus coreceptor CCR5 or CXCR4. Indeed we found that CAR constructs with wild type CD4 extracellular motifs used in previous clinical studies functioned very efficiently as entry receptors, suggesting the likely possibility that when transduced into T CD8 cells (which also express CCR5), such CAR constructs rendered the cells highly susceptible to HIV infection. We believe this is a likely scenario because CD8 T cells exert their cytotoxic activity against infected CD4 T cells by establishing intimate contacts with these targets;such conditions would be optimal for the undesired iinfection of the CD8 cells via cell-to-cell infection (""""""""virologic synapse""""""""). CARs containing the modified CD4-M extracellular motif were found to be completely devoid of of this undesired entry receptor activity. The second critical feature of CD4-M is that because of its greatly enhance binding to surface Env, it should significantly enhance the cytotoxic activity of the engineered CD8 T cells. Preliminary assays show only a modest improvement compared to the wild type CD4 moiety, but these assays were not performed under conditions expected to optimally reveal the improvements, i.e. they were performed with isolates whose Envs are inherently capable of efficiently binding WT CD4, and under conditions of high Env expression on the target cells (i.e. Env pseudotype viruses). More meaningful tests are underway using Envs from isolates known to bind CD4 much less efficiently (i.e. many clinical isolates), and under conditions of less abundant Env surface expression (as occurs during productive infection of primary CD4+ T cells). In our efforts to extend the CAR studies to a relevant animal model, we have begun studies to test the approach in rhesus macaques infected with SIVmac239 (collaboration with Dr. Mario Roederer, VRC). To this end, we have are designing CAR constructs with the corresponding rhesus-derived motifs, including the rhesus based CD4-M variant. We also plan to compare CAR efficacy with different subsets of rhesus CD8 cells, including the recently described memory CD8 T cells with stem cell-like properties.
Showing the most recent 10 out of 14 publications
