A culture method with berbamine, a plant alkaloid, enhances CAR-T cell efficacy through modulating cellular metabolism

Abstract

Memory T cells demonstrate superior in vivo persistence and antitumor efficacy. However, methods for manufacturing less differentiated T cells are not yet well-established. Here, we show that producing chimeric antigen receptor (CAR)-T cells using berbamine (BBM), a natural compound found in the Chinese herbal medicine Berberis amurensis, enhances the antitumor efficacy of CAR-T cells. BBM is identified through cell-based screening of chemical compounds using induced pluripotent stem cell-derived T cells, leading to improved viability with a memory T cell phenotype. Transcriptomics and metabolomics using stem cell memory T cells reveal that BBM broadly enhances lipid metabolism. Furthermore, the addition of BBM downregulates the phosphorylation of p38 mitogen-activated protein kinase and enhanced mitochondrial respiration. CD19-CAR-T cells cultured with BBM also extend the survival of leukaemia mouse models due to their superior in vivo persistence. This technology offers a straightforward approach to enhancing the antitumor efficacy of CAR-T cells.

Fig. 1. Screening of chemical compounds using the iPS-derived T cells.

a T cell differentiation scheme from T-iPSCs for screening chemical compounds. Expanded T-iPS-T cells were stimulated with CD3/CD28 microbeads and then incubated with compounds. All data presented in Fig. 1 were acquired 6 d later. b Cell numbers based on intracellular dehydrogenase activity. Error bars indicate the standard deviation (SD) of five independent experiments. c Representative FACS plots 6 d after the addition of the compounds. d Percentage and e number of PI-viable cells determined with flow cytometry. f Percentage and g number of CCR7+CD45RA+ cells. Error bars indicate the SD of four independent experiments (c–f). Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (paired t-test).

Fig. 2. Effects of BBM on human healthy donor-derived T cells.

a Experimental scheme for characterizing the effect of BBM on primary CD3+ T cells. b Fold changes in cell number and c viability determined using the standard trypan blue method. Error bars indicate SD from six donors. d Representative FACS plots illustrating the expression of CCR7 and CD45RA in CD4+ cells (top) and CD8β+ cells (bottom) on day 14. e Estimated numbers of CD4+CCR7+CD45RA+ cells (left) and CD8β+CCR7+CD45RA+ cells (right) in cultured CD3+ T cells calculated from the total cell number and frequencies of the populations that are PI- on day 14. f, g Results of the CFSE assay in CD4+ (f) and CD8β+ (g) fractions within CD3+ cells on day 3. Error bars indicate SD from six (b, c) or four donors (e–g). Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (paired t-test).

Fig. 3. In vivo persistence of primary T cells derived from healthy donors.

a Experimental scheme designed to assess the in vivo persistence of BBM-treated primary CD8+ T cells. b Gating strategy employed to determine the expression of CCR7 and CD45RA in the CD45+CD8β+ fractions. c Frequencies of injected CD45+CD8β+ cells in the spleen (left), peripheral blood (center), and bone marrow (right) on day 14. Each data point represents one injected mouse and its corresponding donors: circle; donor A, triangle; donor B, square; donor C. All error bars indicate SD. Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (unpaired t-test). d Representative flow cytometry plots of CCR7 and CD45RA in CD45+CD8β+ cells in the spleen samples.

Fig. 4. Gene expression analysis in BBM-treated stem cell memory T cells.

a Gating strategy for FACS to purify the naïve and memory T cell fractions from healthy donor-derived CD8+ cells. b, c Gene ontology analysis of DEGs in BBM-treated T_SCM cells. Upregulated (b) or downregulated (c) genes from RNA-sequencing results on day 12. Genes were queried in the DAVID functional annotation database. d Transcripts per million (TPM) values of cell cycle-related and effector function-related genes. Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (n = 4, paired t-test). T_SCM stem cell memory T, T_CM central memory T, T_EM effector memory T.

Fig. 5. Effect of BBM on metabolism and signaling pathways in primary T cells.

a Metabolomics of BBM-treated T_SCM cells from healthy donor-derived CD8+ cells. Lipids commonly upregulated or downregulated in all three donors are shown. b Representative flow cytometry plots of Phosflow assay in BBM-treated CD3+ cells on day 10. Phosphorylated proteins in CD8+ T cells are shown. c Mean fluorescence intensity (MFI) of phosphorylated p38 in CD8+ T cells. Error bars indicate SD from four donors. d Oxygen consumption rate (OCR) across time for BBM-treated CD3+ T cells. e Spare respiratory capacity (SRC) calculated by OCRmax-OCRbasal. f Extracellular acidification rates (ECAR) across time for BBM-treated CD3+ T cells. Error bars indicate the standard error of the mean from six donors (d–f). Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (paired t-test) (c, e).

Fig. 6. Effector functions of second-generation CAR-T cells.

a CD19-CAR retrovirus vectors (top) and an experimental scheme for in vitro assays. b Assessment of cell proliferation among CD3+ T cells transduced with CD19-CAR (CAR-T cells). Fold expansion rates compared to Day 0 were shown. c Evaluation of in vitro cytotoxicity of CAR-T cells against CD19-expressing NALM6 cells. E/T ratio effector/target ratio. Error bars indicate SD from three donors (b, c). d Expression of CD107a and e IFNγ production in CAR-T cells that were stimulated with PMA plus Ionomycin (PMA/Ionomycin), NALM6 cells, or unstimulated control (T cell only). Circle; 19-28z, triangle; 19-BBz. Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001 (paired t-test).

Fig. 7. Berbamine improves the therapeutic effect of CAR-T cells.

a Experimental scheme of the in vivo assays. Bioluminescence in Luciferase-expressing NALM6 cells were monitored to measure the amount of leukemia burden. CD19-CAR with 4-1BB costimulatory domain (19-BBz) was transduced into primary CD3+ T cells and cultured with BBM or water. After 10 days of cultivation, CD19-CAR-T cells (2 × 10⁵ cells/mouse) were injected into each leukemia mouse model. b IVIS images to monitor NALM6 cells. The scale bar represents the bioluminescence signal in radiance (p/s/cm²/sr). c Representative flow cytometry plots and d percentages CD3+ GFP+ cells that are hCD45+ cells in peripheral blood on day 31. Statistics: Mann–Whitney test. e Kaplan–Meier survival analysis. Statistics: log-rank Mantel–Cox test. Statistical significance is denoted as follows: *p < 0.05; **p < 0.01; and ***p < 0.001.