Researchers Develop a Novel Delivery System That Efficiently Generates CAR-T Cells In Vivo to Powerfully Attack Tumors

Using lipid nanoparticles to deliver nucleic acids and directly generate CAR-T cells in vivo is a highly promising new therapeutic strategy for a wide range of diseases. However, the relatively low efficiency of gene delivery to T cells in vivo remains a major obstacle to the clinical translation of this technology.

Recently, researchers from China published a research paper online in ACS Nano titled “β-Hydroxy Thioether-Derived Ionizable Lipids for Spleen-Tropic mRNA Delivery and In Vivo Chimeric Antigen Receptor T Cell Engineering”. In this study, the researchers used β-hydroxy thioether as a molecular scaffold to construct a structurally diverse combinatorial library of ionizable lipids. They synthesized 300 lipids with different structures in batches and systematically screened and evaluated their mRNA delivery capability in both in vitro and in vivo systems using a luciferase reporter gene assay.

The optimal lipid molecule identified through screening was 113-AA-C8C14. Nanoparticles assembled from this lipid showed excellent spleen-targeted delivery properties. Compared with classic organ-targeted lipid nanoparticles, these nanoparticles increased mRNA expression in the spleen by 57-fold. After loading the lipid nanoparticles with mRNA encoding a CD19-targeting CAR, the number of CAR-T cells induced in vivo increased by 2.7-fold. At the same time, the T-cell immune response was significantly enhanced, with interferon-γ and tumor necrosis factor-α secretion levels increased by 2.9-fold and 3.7-fold, respectively. In a mouse model of pancreatic cancer, this delivery system improved tumor growth inhibition by 2.3-fold. Overall, 113-AA-C8C14 lipid nanoparticles provide an efficient and feasible new approach for the in situ generation of CAR-T cells in vivo.

Figure 1. 113-AA-C8C14 LNP for in vivo CAR T cell therapy. (Guo Q, et al., 2026)

Chimeric antigen receptor T-cell therapy, or CAR-T therapy, is a breakthrough technology in the field of cancer immunotherapy and has achieved remarkable clinical efficacy in the treatment of hematological malignancies. However, the conventional ex vivo manufacturing process is complicated and expensive, requiring a series of steps including patient T-cell isolation, genetic modification, ex vivo expansion, and reinfusion. These requirements make large-scale adoption difficult. In addition, traditional viral vector-mediated gene editing carries the risk of insertional mutagenesis, and long-term genomic safety concerns remain to be fully addressed. These limitations have driven an urgent need in the scientific community to develop new technologies capable of directly modifying T cells in vivo, thereby eliminating the cumbersome ex vivo manufacturing process.

Lipid nanoparticles have become highly promising nucleic acid delivery vehicles because of their good biocompatibility, scalability, and precisely tunable molecular structures. They have already been widely used in the development of mRNA vaccines and various nucleic acid-based drugs, and they show broad potential in the field of in vivo cellular reprogramming. However, classic lipid nanoparticles represented by DLin-MC3-DMA, SM-102, and ALC-0315 generally exhibit preferential liver targeting. This makes it difficult for them to efficiently deliver mRNA to extrahepatic tissues or effectively target T cells in the spleen, greatly limiting progress in in vivo immune cell engineering. Previous studies have shown that structurally diverse lipid libraries constructed through combinatorial chemistry can be used to screen for novel ionizable lipids with specific tissue- or organ-targeting capabilities. β-Hydroxy thioether derivatives synthesized through a two-step reaction involving thiolactone ring opening and thiol-epoxy chemistry allow modular and flexible modification of head groups, hydrophobic chains, and linkers. These structural units directly determine the in vivo distribution of the delivery vehicle, making this chemistry an excellent synthetic platform for developing targeted delivery lipids. Building a combinatorial lipid library based on this reaction system could enable precise T-cell targeting and support the in situ generation of CAR-T cells in vivo.

In this study, the researchers constructed two types of β-hydroxy thioether ionizable lipid combinatorial libraries through a two-step thiol-epoxy reaction combined with esterification. Library 1 contained 50 lipids, while Library 2 contained a total of 250 lipids. Using a luciferase reporter gene model, the researchers carried out multiple rounds of systematic screening both in vitro and in vivo and successfully identified the targeted lipid 113-AA-C8C14, which enables precise spleen-targeted delivery of mRNA. They further used proteomics to investigate the underlying mechanism by which this lipid targets splenic T cells. In a mouse model of pancreatic cancer, 113-AA-C8C14 lipid nanoparticles loaded with mRNA encoding a CD19-targeting CAR efficiently induced the generation of functional CAR-T cells in vivo, enhanced the secretion of immune cytokines, and suppressed tumor proliferation and growth. This study demonstrates that rational design and optimization of ionizable lipid structures can greatly simplify the CAR-T manufacturing process, improve targeted delivery efficiency, and open a new path for the clinical translation of in situ CAR-T cell immunotherapy.