Selection for an HIV-Resistant Immune System By Base-Edited CD45 CAR-T Cell Therapy

Introduction

Eradicating cellular reservoirs of human immunodeficiency virus (HIV) in patients on antiretroviral therapy (ART) is critical to curing HIV, since integrated provirus remains in CD4+ cells even in patients without detectable viral load on anti-retroviral therapy (ART). There are rare but compelling reports of HIV cures in patients who underwent allogeneic hematopoietic stem cell transplantation (alloHSCT) from CCR5Δ^32/Δ³² donors, wherein myeloablative chemotherapy and donor alloreactive T cells completely eradicated the recipient’s immunohematopoietic system and allowed it to be replaced by donor-derived hematopoiesis lacking the HIV co-receptor CCR5. While applying this strategy systematically as a treatment option is not feasible due to the paucity of HLA-matched CCR5Δ^32/Δ³² donors for HIV-infected patients and the high-risk nature of alloHSCT, these reports highlight that immunologic depletion of HIV reservoirs (via potent T cell alloreactivity) in combination with an HIV-resistant HSCT (via CCR5 deficiency) can achieve HIV eradication. To exert potent yet broad immune pressure against the cellular components of the viral reservoir we harnessed an HIV- and fratricide-resistant CAR-T cell therapy against the pan-leukocyte antigen CD45.

Results

Fratricide resistance of the anti-CD45 immunotherapy is achieved by epitope editing CD45 using an adenine base editing (ABE) as we recently described [PMID: 37651540]. At the same time, HIV-resistance is accomplished by either base editing the CCR5 start codon to silence CCR5 expression or mutating tyrosine residues (Y14 and Y15) in the N-terminal domain of CCR5 that are known to bind HIV gp120. Base editing CCR5 in primary human CD4 T cells is highly efficient with ~84% A to G editing at the CCR5 start codon mutation, successfully silencing CCR5 protein expression. Similarly, a single gRNA achieved ~90% editing for A5 (Y14H) and ~38% for A2 (Y15H), accomplishing mutagenesis of the sulfotyrosine binding residues for HIV gp120 docking while preserving CCR5 expression. To test whether CCR5 base-edited cells (CCR5^BE) are resistant to HIV infection, we infected CCR5^BE or unedited CD4 T cells with the CCR5-tropic HIV stain BaL. Both groups of CCR5^BE CD4 T cells showed reduced susceptibility to HIV BaL infection both in vitro and in vivo, whereas unedited control cells were readily infected. Resistance to HIV infection translated to a ~4-fold higher preservation of peripheral blood T cell numbers and a ~10-fold lower vRNA copy number in mice that received CCR5^BE T cells compared to unedited T cells. Furthermore, we observed a strong enrichment of CCR5 edited cells from ~60-80% to >99% upon HIV infection, suggesting that CCR5^BE T cells are protected from infection-induced cell death.

To generate CD45 CAR-T cells that are resistant to fratricide and HIV infection, on-target base editing must co-occur biallelically at both the CCR5 loci and the CD45 epitope with high efficiency. Using single-cell DNA sequencing, we show that biallelic, multiplex base editing is efficient in primary human T cells with editing efficiencies of ~75% (CCR5 start codon + CD45^BE and ~85% (CCR5 Y14/15 + CD45^BE) respectively.

Lastly, we demonstrate that multiplex base editing CD45 with CCR5 generates fratricide-and HIV-resistant CART45 cells that can eliminate HIV-infected CD4 T cells in xenografted mice while preventing the spreading of HIV particles from infected CD4 T cells to the CD4 positive CAR-T cells. This facilitated the persistence of HIV-resistant CART45, whereas CD45^BE only edited CAR-T cells were unable to persist due to HIV-induced T cell apoptosis.

Conclusion

Overall, our study demonstrates the potential of anti-CD45 immunotherapy to eradicate HIV infected cells. We speculate that when combined with an engineered autologous HSCT, this CD45/CCR5 multiplex engineering approach can regenerate an HIV- and anti-CD45 resistant hematopoietic system free of latent HIV infected cells.