Scientists Reprogram T Cells In Vivo Using CRISPR Gene Editing

CAR T cell therapy has revolutionized treatment of blood cancers, with seven FDA-approved products to date. However, standard manufacturing requires extracting a patient's T cells, engineering them outside the body over weeks, and reinfusing them—a process that is costly, time-consuming, and produces variable quality products. This landmark study presents a fully in vivo approach to generate precision-engineered CAR T cells without ever removing cells from the patient.

The team engineered a two-vector delivery system. The first vector consists of enveloped delivery vehicles (EDVs)—lipid nanoparticle-like structures displaying T cell-targeting ligands—loaded with CRISPR-Cas9 ribonucleoprotein (RNP) complexes. These EDVs are optimized to selectively fuse with and deliver their cargo specifically to T cells in circulation. The second vector is a recombinant adeno-associated virus (AAV) engineered to carry the CAR DNA donor template with homology arms flanking a T cell-specific genomic locus, enabling precise site-specific integration via homology-directed repair (HDR).

In humanized mouse models reconstituted with human immune cells, intravenous co-administration of the two vectors resulted in stable, site-specific CAR transgene integration in circulating T cells at therapeutically relevant frequencies. Because the CAR was inserted at a T cell-specific locus (analogous to the TRAC locus used in prior ex vivo studies), expression was tightly controlled and cell-type restricted, avoiding aberrant expression in non-T cells—a key safety advantage over random viral integration.

Functional validation across multiple disease models was compelling. In a model of B cell aplasia (mimicking autoimmune disease), in vivo-generated CAR T cells effectively depleted target B cells. In hematological malignancy models, the engineered cells demonstrated robust anti-tumor activity. Remarkably, efficacy was also demonstrated in a solid tumor model, a historically difficult frontier for CAR T therapy. These results suggest the platform is broadly applicable across cancer types and potentially beyond oncology.

The implications are significant for both accessibility and safety. By eliminating ex vivo manufacturing, this approach could dramatically reduce cost and production time, enabling same-day or near-immediate treatment. The site-specific integration strategy avoids genotoxic risks associated with random retroviral insertion. Caveats include the current reliance on humanized mouse models rather than clinical human data, questions about scalability of EDV and AAV production, immune responses to viral vectors, and whether HDR efficiency in humans will match murine results. Nonetheless, this work establishes a compelling proof-of-concept for in vivo precision T cell engineering.