HIV-Receptor Interactions and Related Anti-HIV Strategies

National Institute Of Allergy And Infectious Diseases

Investigators

Linked publications, trials & patents

Abstract

We have continued development of an HIV functional cure strategy based on novel chimeric antigen receptors (CARs). Durable (life-long?) HIV suppression requires the CAR to display not only high potency, but also to approach the ideals of inescapabilty and non-immunogenicity. We have designed bispecific CARs containing CD4 linked to a second moiety that binds to a distinct highly conserved region of Env. The second moiety enhances potency, and also prevents the CD4 from acting as an HIV entry receptor. Particular success was achieved with the carbohydrate binding domain of human mannose binding lectin (MBL), which binds to the universally conserved high-mannose patch on gp120. The CD4-MBL bispecific CAR shows greatly enhanced potency compared to the monospecific CD4 CAR, with wide breadth for genetically diverse HIV-1 isolates; moreover, the lectin moiety inhibits the CD4 from acting as an entry receptor and rendering CAR-expressing CD8+ T cells susceptible to infection (Ghanem et al. 2018). Efforts during the past year have focused on design of improved CAR constructs, and on testing of the CAR-T cells in relevant animal models in collaboration with investigators in the network of Martin Delaney Collboratories for HIV Cure. 1) Comparative potencies of the CD4-MBL CAR to antibody-based alternatives. The HIV field has been greatly stimulated by the development of a wide range of broadly neutralizing monoclonal antibodies (bnAbs) that bind to different conserved regions of Env, and display great breadth and potency against genetical diverse HIV-1 variants. Several research groups are developing corresponding bnAb-based CARs. In collaboration with the group of Dr. Thor Wagner (defeatHIV Collaboratory, Seattle Childrens,), we generated analogous CAR constructs targeted by either CD4-MBL or an scFv of a highly studied bnAb. Our in vitro assay results indicated highly superior activity of the CD4-MBL CAR, likely leading defeatHIV to switch to this CAR. 2) Design of a trispecific CAR. With the goal of further enhancing CAR potency, we examined addition of a third human component of the targeting domain that binds to yet a distinct highly conserved region of HIV-1 Env. Success was obtained using a polypeptide region from the N-terminus of human CCR5, the entry coreceptor used by HIV-1 variants that predominate in infected individuals from the time of initial infection throughout the asymptomatic period. This trispecific CAR, designated CD4-MBL-R5Nt, displayed enhanced potency over CD4-MBL against the two HIV-1 primary isolates tested. Moreover, the presence of the third motif further blocked to undetectable levels the ability of the CD4 moiety from acting as an HIV entry receptor. Additional studies will test the breadth of the CD4-MBL-R5Nt CAR in vitro, and possibly expand to in vivo models. 3) CAR studies in humanized mouse models. Collaborative studies with investigators in the BELIEVE Collaboratory have yielded interesting results. Dr. Harris Goldstein (Yeshiva Med) tested our earlier bispecific construct, the CD4-17b CAR in their intrasplenic humanized mouse. Co-administration of the CAR at the time of HIV-1 challenge resulted in very strong virus suppression. With Dr. Brad Jones (Cornell/Weill), we are following up earlier experiments in his infused memory T cell humanized mouse model. Initial studies showed negligible effects of the CD4-MBL CAR when administered at the time of mouse challenge, whereas potent suppression was observed with an HIV-specific T cell product designated HST-NEETs, developed by Dr. Catherine Bollard (Childrens National). We hypothesized that the difference may be due to the very different mode of CAR-T cell expansion used by the Berger and Bollard labs. Consequently, the CD4-MBL CAR was transduced and the T cells were expanded by the Bollard protocol. In vitro analyses indicating excellent CAR-T cell expansion and CAR surface expression on the expanded cells. These CAR-T cells showed potent HIV suppression in vitro. This sets the stage for direct in vivo comparison of the HST-NEET cells with the similarly expanded CAR-T cells in the Brad Jones humanized mouse. 4. NHP studies. In collaboration with BELIEVE investigator Dr. Pamela Skinner (U. Minnesota) we previously generated an all-rhesus variant of CD4-MBL for co-expression with CXCR5 to facilitate homing of CAR-T cells to B cell follicles, where infected Tfh cells serve as major sites of HIV persistence. Previous findings indicated that CXCR5 promotes selective migration of CAR-T cells toward the chemokine ligand CXCL13 in vitro, as well their accumulation inside B cell follicles in rhesus lymphoid tissue explants ex vivo. Studies in rhesus macaques will require tracking of CAR-T cells in blood and various tissues. Immunostaining of CAR constructs containing c-myc tags in the ectodomains proved unsuitable due to high background staining. The Skinner lab has now adopted RNA-Scope technology, which will enable detection critical cell types in tissues, including CAR-T cells, HIV-infected cells, etc. In a preliminary experiment, the Skinner group observed that 2 of 3 rhesus macaques injected with CD4-MBL CAR-T cells showed significant control of SIV rebound after cessation of antiretroviral therapy, in contrast to 3 controls that did not receive CAR-T cells. This set the stage for examining numerous critical features associated with virus suppression and potential functional cure, including follicular trafficking of CAR/CXCR5-T cells in vivo, and their ability to deplete infected Tfh cells. A second study in rhesus macaques has been initiated with Dr. Mario Roederer (Vaccine Research Center, NIAID, NIH). The goal is to compare a diversity of CAR constructs varying in both targeting domains (CD4-MBL vs. scFvs derived from anti-SIV Env monoclonal antibodies) and also signaling domains (CD28 vs. 4-1BB). The initial focus is to compare the relative persistence of these diverse CAR constructs. Initial results will help focus on optimal designs for durable control of infection.