Back

An in situ generated CAR-M with IFNg and negative dominance Sirpa isoform augments hepatocellular carcinoma immunotherapy

Zhang, P.; Li, X.; Hu, J.; Zhang, T.; Li, H.; Wang, T.; Liu, Y.; Zhang, L.; Wei, Y.; Wei, S.

2025-12-11 cancer biology
10.64898/2025.12.08.693087 bioRxiv
Show abstract

Background and AimsChimeric antigen receptor (CAR) T cells have shown strong efficacy in hematological cancers but limited success in solid tumors. Macrophages, with their natural ability to infiltrate tumors, modulate immunity, and phagocytose cancer cells, offer a promising alternative when engineered with CARs. While studies have demonstrated the feasibility and anti-tumor activity of CAR macrophages (CAR-M), enhancing their persistence and phagocytic capacity remains a key challenge. Methods and ResultsBuilding on first-generation CD3{zeta}-based CAR-M, we developed a novel CAR targeting human GPC3, incorporating IFN-{gamma} and the extracellular domain of SIRP (SIRPECD) to improve CAR-M persistence and block the CD47-SIRP immune checkpoint. Following delivery via lipid nanoparticle-encapsulated mRNA (LNP-mRNA), the self-secreted IFN-{gamma} sustained M1 polarization through phospho-STAT1 activation. Meanwhile, the ectopically expressed SIRPECD competitively bound to CD47 on tumor cells, thereby blocking the endogenous SIRP-SHP2 interaction in a dominant-negative manner. This design enhanced pro-inflammatory activity and anti-tumor efficacy compared to CD3{zeta}-only CAR-M. Single-cell RNA sequencing and cellular analysis showed that in situ programmed CAR-M reprogrammed the tumor microenvironment toward inflammation in a murine HCC model. Moreover, CAR-M derived from human peripheral blood mononuclear cells (PBMCs) effectively phagocytosed human HCC organoids while sparing healthy tissues, indicating clinical potential. ConclusionsCollectively, our work presents a novel CAR design that enhances phagocytic function and sustains anti-tumor activity, offering a promising strategy for human solid tumor immunotherapy.

Published in Journal of Nanobiotechnology · training set

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.