The long-term existence of CAR T is closely related to its clinical treatment efficacy, and long-life memory CAR T provides continuous immune monitoring to prevent tumor recurrence [1,2]. Model animal studies have found that asymmetric cell division (ACD) is one of the important mechanisms for memory formation in CD8+T cells [3,4]. The daughter cells derived from parental cell division, which are close to antigen-presenting cells (APCs), inherit immune synapses and are more likely to differentiate into short-lived effector T cells (Teff). In contrast, distal terminal cells differentiate into long-lived memory T cells with unique transcriptional, epigenetic, and metabolic characteristics (see BioArt report: Expert review of epigenetic strategies for improving T cell function by Guo Ao/Huang Hongling/Chi Hongbo/Douglas Green, etc.) [4-6]. Given the broad prospects of CAR T for cancer treatment, researchers have maintained a strong interest in the phenotype and memory formation of CAR T over the past decade. However, the cellular mechanisms underlying the development of long-lived and short-lived CAR T, particularly whether ACD is involved, are still unclear.

Recently, the research team led by Christoph T. Ellebrecht from the Perelman School of Medicine at the University of Pennsylvania published a research article titled "Fate induction in CD8 CAR T cells through asymmetrical cell division" online in the journal Nature. This study reveals the different fate choices of CAR T cells caused by asymmetric division through scRNA seq, LIPSTIC, flow cytometry, and metabolic analysis, providing a new perspective for further understanding CAR T differentiation and improving treatment.

The team first used sorting enzyme labeled labeling of cell-cell contact labeling immune partners (LIPSTICs) labeling technology to trace the distal and proximal terminal cells of CAR T parental cells after the first division stimulated by APCs (see BioArt report: Nature | uLIPSTICs: A General Method for Recording Immune Cell Interactions in vivo). Flow cytometry analysis revealed that the progeny cells had equal populations of LIPSTICC+(near terminal generation of target cells) and LIPSTICC - (far terminal generation of target cells) cells, indicating the presence of ACD behavior during the first division of CAR T. Feature analysis shows that compared to the LIPSTIC cell population, LIPSTIC+cells have a larger volume, higher expression of CD25, and stronger proliferation ability. At the same time, metabolic analysis showed that LIPSTICC+cells rely on glycolytic metabolism, while the LIPSTICC cell population mainly relies on mitochondrial oxidative phosphorylation. Subsequently, the author found similar ACD phenomena in CD4 and CD8 derived CAR T cells, as well as in the initial (na ï ve, CD62L+CD45RO -), effector (CD62L -) derived CAR T cells, and the division of progeny CAR T cells after the first division and exposure to target cells again. In summary, these data indicate that CAR T activation leads to asymmetric division, with LIPSTIC+representing the proximal terminal generation of APCs and LIPSTIC - representing the distal terminal generation cell population of APCs, and offspring still exhibiting ACD upon exposure to target cells.

Afterwards, the authors adopted the sorted proximal and distal generation LIPSTICC+and LIPSTIC-CAR T cells into NOD scid gamma (NSG) mice and injected them with Nalm6 leukemia cells again. After 30 days, it was found that the mice receiving distal generation cells had more T cells in their peripheral blood and the progression of leukemia was effectively controlled, indicating that distal generation cells had a longer duration in vivo and better tumor control effects. At the same time, the author tested the cytotoxicity of the first dividing progeny cells in vitro. The results showed that the far terminal generation cells and the near terminal generation cells had similar killing ability on the first day after division, but the killing ability of the far terminal generation cells was lower than that of the near terminal generation cells after the fourth day of culture, indicating that the target cell activation rapidly induced the killing ability of the near terminal and far terminal generation cells. Although the far terminal generation cells had lower activation and metabolism, their life span was longer.

Subsequently, the author isolated resting, activated, pre division, and first division offspring CAR T cells for single-cell proteome, transcriptome, and TCR sequencing analysis. Firstly, UMAP analysis based on membrane surface sequencing data showed significant separation between resting CAR T and activated CAR T, indicating changes in the membrane surface protein profile induced by activation of CAR T; The dividing progeny cells are located between resting and activated cells, and the distal and proximal progeny cells are also separated, indicating that ACD of activated cells leads to different surface protein profiles of the progeny cell membrane. Specifically, proximal terminal cells inherit the membrane protein expression pattern of activated undifferentiated parental cells, while distal terminal cells are more similar to resting cells. Transcriptome analysis shows that distal and proximal terminal cells have different transcriptome characteristics. Specifically, compared to the initial T cell like proximal passage cells, distal passage cells expressed high levels of LEF1, TCF7, CCR7, IL7R, and KLF2, while proximal passage cells expressed high levels of MYC and MTORC1 target genes, consistent with the previous proximal cell dependent glycolysis metabolism. So, what are the reasons for the different transcriptomes? Is it the pre-existing uneven distribution of RNA caused by ACD or the upregulation and downregulation of fate related gene transcription after ACD? RNA rate analysis confirmed that both conditions simultaneously drive transcriptional differences after ACD.

Finally, the author analyzed the transcription factor related gene networks or regulators (regulon) between different cells using pySHENIC. The results showed that the upregulated regulatory subunits in proximal terminal cells were associated with apoptosis, proliferation, and effector cell differentiation, including TP73, E2F2/7/8, MYBL1, and YBX1; The upregulated regulatory domains in distal terminal cells are associated with self-renewal and anti proliferation, including IRF1, STAT1, KDM5B, REST, FLI1, MXD4, and IKZF1. Given that IKZF1 plays a role in limiting T cell proliferation, effector differentiation, and activation, and is not related to ACD, the authors will further investigate the impact of IKZF1 on the fate of distal terminal cells. After CRISPR-Cas9 mediated IKZF1 knockout, distal terminal cells exhibited a phenotype similar to proximal terminal cells, including reduced and increased CD45RA expression and effector differentiation. In vivo experiments also showed that compared to the control group, IKZF1 deficient distal terminal cells were introduced into NSG mice for 30 days, resulting in a decrease in peripheral T cell count and weakened control of Nalm6 tumors. In summary, these data indicate that CAR T ACD mediates differences in the transcriptome program of progeny cells and endows them with different fate choices. IKZF1, as a key transcription factor in this process, maintains cell longevity characteristics in distant progeny cells.

In summary, this study revealed the differences in membrane proteome, transcriptome, and metabolome of CAR T cell offspring caused by asymmetric division through methods such as LIPSTIC, proteome transcriptome TCR scRNA seq, and further promoted the differentiation and selection of offspring cell fate, providing a feasible framework for optimizing CAR T and other T cell-based immunotherapies.