Ex vivo expanded HIV-Specific CD8+ T-cell (HST) immunotherapy offers great promise toward achieving an HIV cure. While plans to test HST therapies in humans are currently underway, a small animal model would enable the rapid and cost-effective pre-clinical evaluation of multiple approaches. We have developed a humanized mouse model reconstituted with only the memory subset of human CD4+ T-cells, which has greatly mitigated the effects of GvHD and allows for the in vivo analysis of autologous HST therapies.
NSG mice were engrafted with 5-10×106 memory CD4+ T-cells isolated from HIV- or HIV+ donor leukapheresis samples. Autologous HSTs were generated by stimulating T-cells with pools of overlapping Clade B consensus peptides or peptides representing only conserved viral epitopes. Four to six weeks post humanization, mice were simultaneously infected with JR-CSF and treated with autologous HSTs. Weekly blood samples were analyzed by flow cytometry to measure changes in the human CD4+/CD8+ levels as well as qRT-PCR to measure viral load. Plasma RNA was subsequently sequenced for the presence of viral escape mutations.
Mice engrafted with only the memory CD4+ T-cell subset survived significantly longer than mice engrafted with total CD4+ T-cells and were able to support robust HIV infection sustained out to 20 weeks post engraftment. Daily ARV injections resulted in viral suppression and CD4+ T-cell reconstitution, followed by viral rebound and CD4+ T-cell loss upon ARV cessation. Mice that received autologous HSTs saw significant, transient decreases in plasma viral load compared to the No Treatment group (p<0.0001). Sequencing analysis of plasma virus revealed a dominant escape mutation in one mouse in the HST group suggesting immunological pressure. Early results from an additional in vivo experiment demonstrated similar, significant decreases in viral loads with 66% of mice receiving HST therapy reaching an undetectable viral load 4 weeks post therapy initiation.
We have demonstrated that our novel memory CD4+ T-cell humanized mouse model accurately recapitulates many aspects of natural HIV infection while significantly reducing the effects of GvHD. Using this model, we have observed significant decreases in viral load in mice receiving clinically relevant HST products. This platform provides opportunities to assess a variety of immunotherapeutic strategies as well as immunomodulatory approaches in an in vivo, autologous target/effector setting.