Reprogramming T Cells: The Promise of In Vivo Site-Specific CAR Engineering

Background: The Manufacturing Bottleneck

CAR T-cell therapy has transformed the treatment landscape of hematologic malignancies, with seven FDA-approved products now available.¹ Yet despite their clinical efficacy, these therapies remain fundamentally constrained — not by biology, but by manufacturing.

Current standard-of-care requires leukapheresis, ex vivo T-cell activation and viral transduction, quality testing, cryopreservation, and shipment—a process that can take weeks and cost hundreds of thousands of dollars per patient.² During this interval, patients may clinically deteriorate, and access disparities become increasingly apparent.

The ex vivo model also introduces biological limitations. Retroviral and lentiviral vectors integrate semi-randomly into the genome, resulting in heterogeneous CAR expression across the final product.^2,3 Prolonged culture drives differentiation toward more exhausted phenotypes.^9,10 Rare but documented events—including clonal expansion linked to integration near oncogenes—highlight the safety risks inherent to random integration.^5,6

For years, the field has pursued a more direct strategy: engineering T cells within the patient. Prior approaches—including in vivo lentiviral delivery and lipid nanoparticle platforms—have either lacked durability or precision.^12,15 Both modalities have been validated in nonhuman primate models^17,18 and recently evaluated in a phase I trial.¹⁹ What has remained elusive is stable, site-specific integration of a full CAR construct in vivo. This study addresses that gap.

The Study: What Nyberg, Bernard, and Colleagues Did

In this Nature report, Nyberg, Bernard, and colleagues describe a dual-vector system capable of achieving stable, cell-specific, site-directed integration of a full CAR transgene directly in primary human T cells in vivo.¹

The system is built on two complementary components. The first is a CD3-targeted enveloped delivery vehicle (EDV) — a viral-like particle pseudotyped with a modified VSV-G glycoprotein and decorated with an anti-CD3 single-chain variable fragment.¹ This design confers T-cell specificity while simultaneously activating resting T cells, expanding the pool of cells capable of homology-directed repair (HDR). Encapsulated within the EDV is a Cas9 ribonucleoprotein targeting the TRAC locus — previously shown to drive uniform, physiologically regulated CAR expression after integration.⁴

The second component, AAV-hT7, is an engineered AAV capsid evolved for efficient transduction of primary human T cells and resistance to pre-existing neutralizing antibodies.¹ Genome-wide CRISPR knockout screening identified CD7 as a key co-receptor mediating its T-cell tropism, and confirmed detargeting from hematopoietic stem cells and malignant B cells. Packaged within AAV-hT7 is a promoterless homology-directed repair template encoding the CAR construct. Because the template is promoterless, CAR expression occurs only following correct integration downstream of the endogenous TRAC promoter — limiting expression exclusively to properly edited T cells.¹

Efficacy Results: Hitting the Therapeutic Threshold

In humanized NSG mouse models, the optimized EDV + AAV-hT7 system achieved TRAC-CAR integration in up to 19.7% of splenic T cells.¹ In contrast, earlier vector combinations yielded negligible editing, underscoring the contribution of both delivery components.

In B-ALL models, a single systemic administration resulted in complete responses in 18 of 20 mice across multiple donors.¹ Rechallenged animals maintained durable tumor control, demonstrating functional persistence of in vivo–generated CAR T cells.

Comparative studies against in vivo lentiviral delivery further highlighted the advantage of site-specific integration.^12,16 TRAC-CAR T cells expanded more rapidly and achieved complete tumor eradication, whereas lentiviral approaches produced only partial responses despite higher vector doses.¹ This likely reflects the uniform CAR expression driven by TRAC-locus integration, in contrast to the heterogeneous expression inherent to random integration at low in vivo multiplicity of infection.

Similar efficacy was observed in multiple myeloma models, with complete responses in all treated animals.¹ Notably, activity extended into solid tumor models, with responses observed in B7-H3–targeted sarcoma systems.²⁷ While preliminary, this represents an important proof of concept for in vivo CAR applications beyond hematologic malignancies.¹

Phenotypic Characterization: Quality Matters

In vivo–engineered TRAC-CAR T cells were enriched for less differentiated phenotypes, including CD45RA⁺CD62L⁺ populations associated with improved persistence and long-term antitumor activity.^1,9,10

These cells demonstrated increased proliferative capacity (Ki-67⁺) and features consistent with progenitor-exhausted T cells, including preservation of TCF1 expression.¹ Importantly, regulatory T-cell representation was reduced among CAR-positive populations—a finding that may have implications for overcoming resistance to CAR T therapy.²¹

Collectively, these findings suggest that bypassing ex vivo culture may not only simplify logistics but also yield a biologically superior cellular product.

Safety Considerations and Open Questions

No evidence of cytokine release was observed at early time points following administration in humanized models.¹ The platform incorporates multiple layers of specificity — CD3-targeted delivery, CD7-mediated tropism, and TRAC-dependent expression — designed to minimize off-target activity and risk of CAR expression in tumor cells, which could otherwise drive antigen-negative relapse.²²

Several critical questions remain, including the immunogenicity of AAV-hT7, the feasibility of repeat dosing in the setting of anti-vector neutralizing antibodies, and the extent of off-target editing in immunocompetent systems.²⁵ These will require evaluation in nonhuman primate models before first-in-human studies.

Additionally, TCR repertoire restriction and the dependence on actively cycling T cells raise important considerations for heavily pretreated or lymphopenic patients.¹

Clinical and Translational Perspective

The significance of this work extends beyond technical innovation. It directly addresses one of the most pressing limitations in cellular therapy: access. The current CAR T paradigm favors patients who can tolerate delays, access specialized centers, and undergo complex treatment workflows.² An in vivo approach has the potential to fundamentally shift this model—enabling faster treatment, reducing logistical complexity, and expanding delivery into community settings.

Unlike allogeneic strategies, which introduce risks of immune rejection and limited persistence^16-18, in vivo autologous engineering preserves biological compatibility while eliminating manufacturing constraints. This represents a conceptually distinct and potentially more scalable solution.

What Comes Next

Further evaluation in immunocompetent nonhuman primate models will be essential to define safety, pharmacokinetics, and immunogenicity.²⁵ While early signals in solid tumors are encouraging, the challenges of the tumor microenvironment—including antigen heterogeneity and immunosuppression—remain substantial.^1,27 Combination strategies will likely be required to achieve meaningful clinical impact.

Finally, regulatory and manufacturing frameworks for dual-vector biologics will need to evolve alongside the science.

The Bottom Line

Nyberg, Bernard, and colleagues have achieved a major milestone: precise, stable, T cell–specific CAR integration directly in vivo.¹ The implications are substantial. By eliminating ex vivo manufacturing, this approach has the potential to expand access, reduce cost, and fundamentally reshape how cellular therapies are delivered. The translational path is complex — but the direction is clear: cell therapy without the manufacturing bottleneck.

References

1. Nyberg WA, Bernard PL, Ngo W, et al. In vivo site-specific engineering to reprogram T cells. Nature. 2026. doi:10.1038/s41586-026-10235-x
2. Wang X, Riviere I. Clinical manufacturing of CAR T cells: foundation of a promising therapy. Mol Ther Oncolytics. 2016;3:16015. doi:10.1038/mto.2016.15
3. Riviere I, Gallardo HF, Hagani AB, Sadelain M. Retroviral-mediated gene transfer in primary murine and human T-lymphocytes. Mol Biotechnol. 2000;15(2):133-142. doi:10.1385/MB:15:2:133
4. Eyquem J, Mansilla-Soto J, Giavridis T, et al. Targeting a CAR to the TRAC locus with CRISPR/Cas9 enhances tumour rejection. Nature. 2017;543(7643):113-117. doi:10.1038/nature21405
5. Fraietta JA, Nobles CL, Sammons MA, et al. Disruption of TET2 promotes the therapeutic efficacy of CD19-targeted T cells. Nature. 2018;558(7709):307-312. doi:10.1038/s41586-018-0178-z
6. Shah NN, Qin H, Yates B, et al. Clonal expansion of CAR T cells harboring lentivector integration in the CBL gene following anti-CD22 CAR T-cell therapy. Blood Adv. 2019;3(15):2317-2322. doi:10.1182/bloodadvances.2019000219
7. Short L, Holt RA, Cullis PR, Evgin L. Direct in vivo CAR T cell engineering. Trends Pharmacol Sci. 2024;45(5):406-418. doi:10.1016/j.tips.2024.03.004
8. Hamilton JR, Doudna JA, et al. In vivo human T cell engineering with enveloped delivery vehicles. Nat Biotechnol. 2024;42(10):1684-1692. doi:10.1038/s41587-023-02085-z
9. Hinrichs CS, Borman ZA, Cassard L, et al. Human effector CD8+ T cells derived from naive rather than memory subsets possess superior traits for adoptive immunotherapy. Blood. 2011;117(3):808-814. doi:10.1182/blood-2010-05-286286
10. Sommermeyer D, Hudecek M, Kosasih PL, et al. Chimeric antigen receptor-modified T cells derived from defined CD8+ and CD4+ subsets confer superior antitumor reactivity in vivo. Leukemia. 2016;30(2):492-500. doi:10.1038/leu.2015.247
11. Cael B, Smets F, Cabrespine A, et al. Umbilical cord blood as a source of less differentiated T cells to produce CD123 CAR-T cells. Cancers. 2022;14(13):3168. doi:10.3390/cancers14133168
12. Pfeiffer A, Thalheimer FB, Hartmann S, et al. In vivo generation of human CD19-CAR T cells results in B-cell depletion and signs of cytokine release syndrome. EMBO Mol Med. 2018;10(11):e9158. doi:10.15252/emmm.201809158
13. Smith TT, Stephan SB, Bhatt DL, et al. In situ programming of leukaemia-specific T cells using synthetic DNA nanocarriers. Nat Nanotechnol. 2017;12(8):813-820. doi:10.1038/nnano.2017.57
14. Rurik JG, Tombácz I, Yadegari A, et al. CAR T cells produced in vivo to treat cardiac injury. Science. 2022;375(6576):91-96. doi:10.1126/science.abm0594
15. Parayath NN, Stephan SB, Koehne AL, Nelson PS, Stephan MT. In vitro-transcribed antigen receptor mRNA nanocarriers for transient expression in circulating T cells in vivo. Nat Commun. 2020;11(1):6080. doi:10.1038/s41467-020-19486-2
16. Michels KR, Sousa AA, Bhatt DL, et al. Preclinical proof of concept for VivoVec, a lentiviral-based platform for in vivo CAR T-cell engineering. J Immunother Cancer. 2023;11(2):e006292. doi:10.1136/jitc-2022-006292
17. Nicolai CJ, Haft RJF, Westein E, et al. In vivo CAR T-cell generation in nonhuman primates using lentiviral vectors displaying a multidomain fusion ligand. Blood. 2024;144(9):977-987. doi:10.1182/blood.2023023841
18. Hunter TL, Fonseca PA, Li C, et al. In vivo CAR T cell generation to treat cancer and autoimmune disease. Science. 2025;388(6751):1311-1317. doi:10.1126/science.adv3945
19. Xu J, Wang Y, Chen L, et al. In-vivo B-cell maturation antigen CAR T-cell therapy for relapsed or refractory multiple myeloma. Lancet. 2025;406(10445):228-231. doi:10.1016/S0140-6736(25)00576-3
20. Nyberg WA, Croze RH, Doudna JA, et al. An evolved AAV variant enables efficient genetic engineering of murine T cells. Cell. 2023;186(3):446-460. doi:10.1016/j.cell.2022.12.022
21. Good Z, Sarno J, Jager A, et al. Post-infusion CAR Treg cells identify patients resistant to CD19-CAR therapy. Nat Med. 2022;28(9):1860-1871. doi:10.1038/s41591-022-01916-x
22. Ruella M, Xu J, Barrett DM, et al. Induction of resistance to chimeric antigen receptor T cell therapy by transduction of a single leukemic B cell. Nat Med. 2018;24(10):1499-1503. doi:10.1038/s41591-018-0201-9
23. Booth C, Gaspar HB, Thrasher AJ. Treating immunodeficiency through HSC gene therapy. Trends Mol Med. 2016;22(4):317-327. doi:10.1016/j.molmed.2016.02.002
24. Pillay S, Meyer NL, Puschnik AS, et al. An essential receptor for adeno-associated virus infection. Nature. 2016;530(7588):108-112. doi:10.1038/nature16465
25. Boutin S, Monteilhet V, Veron P, et al. Prevalence of serum IgG and neutralizing factors against adeno-associated virus (AAV) types 1, 2, 5, 6, 8, and 9 in the healthy population: implications for gene therapy using AAV vectors. Hum Gene Ther. 2010;21(6):704-712. doi:10.1089/hum.2009.182
26. Nyberg WA, Dahl M, Porteus MH, et al. In vivo engineering of murine T cells using the evolved adeno-associated virus variant Ark313. Immunity. 2025;58(3):499-512. doi:10.1016/j.immuni.2025.01.009
27. Kontos F, Michelakos T, Kurokawa T, et al. B7-H3: an attractive target for antibody-based immunotherapy. Clin Cancer Res. 2021;27(5):1227-1235. doi:10.1158/1078-0432.CCR-20-2584