A novel mitochondrial pyruvate carrier inhibitor drives stem cell-like memory CAR T cell generation and enhances antitumor efficacy

Abstract

Adoptive cell transfer with chimeric antigen receptor (CAR)-expressing T cells can induce remarkable complete responses in cancer patients. Therapeutic success has been correlated with central and stem cell-like memory T cell subsets in the infusion product, which are better able to drive efficient CAR T cell in vivo expansion and long-term persistence. We previously reported that inhibition of the mitochondrial pyruvate carrier (MPC) during mouse CAR T cell culture induces a memory phenotype and enhances antitumor efficacy against melanoma. Here, we use a novel MPC inhibitor, MITO-66, which robustly induces a stem cell-like memory phenotype in CD19-CAR T cells generated from healthy donors and patients with relapsed/refractory B cell malignancies. MITO-66-conditioned CAR T cells were superior in controlling human pre-B cell acute lymphoblastic leukemia in mice. Following adoptive cell transfer, MITO-66-conditioned CAR T cells maintained a memory phenotype and protected cured mice against tumor rechallenge. Furthermore, in an in vivo B cell leukemia stress model, CD19-CAR T cells generated in the presence of MITO-66 largely outperformed clinical-stage AKT and PI-3Kδ inhibitors. Thus, we provide compelling preclinical evidence that MPC inhibition with MITO-66 during CAR T cell manufacturing dramatically enhances their antitumor efficacy, thereby paving the way to clinical translation.

Keywords: CAR T cell manufacture; CAR T cell therapy; MT: Regular Issue; immunometabolism; memory T cell differentiation; mitochondrial pyruvate carrier.

Graphical abstract

Figure 1

MITO-66 induces a stem cell-like memory phenotype in CAR T cells (A) Experimental scheme depicting CD19-CAR T cell generation in the presence of MITO-66. (B) T cell expansion measured by machine-assisted trypan blue-based cell counting on days 5, 7, and 9 post-activation (five donors, pooled data from three independent experiments). (C and D) CD62L-positive, CD45RO-negative stem cell-like memory T cells out of CD4 (C) or CD8 (D) CAR T cells at day 9 post-activation (five donors, pooled data from three independent experiments). (E and F) CD62L-positive, CD45RO-negative stem cell-like memory T cells in the CD4 (E) or CD8 (F) CAR T populations at day 9 post-activation (four donors, pooled data from two independent experiments). (G and H) CAR-positive T cells in the CD4 (G) and CD8 (H) populations at day 9 post-activation (four donors, pooled data from two independent experiments). (I and J) CD4 (I) and CD8 (J) T cells out of total live cells at day 9 post-activation (four donors, pooled data from two independent experiments). (K) Fatty acid oxidation (FAO) and amino acid oxidation (AAO) measured by SCENITH in CD4 and CD8 CAR T cells at day 9 post-activation (four donors, pooled data from two independent experiments). (L and M) Representative western blot (L) and quantification of H3K27 acetylation, normalized by total histone H3 protein levels in T cells at day 9 post-activation (three donors, pooled data from two independent experiments). Data are represented as mean ± standard deviation (SD). Statistics are based on one-way ANOVA (B–D), on paired, two-tailed Student’s t test (E–J and M), or two-way ANOVA (K). See also Figure S1.

Figure 2

MITO-66 conditioning enhances CAR T cell expansion and cytotoxicity (A and B) CD45RO-negative, CD62L/CD45RA/CCR7/CD127-positive stem cell-like memory T cells in the CD4 (A) or CD8 (B) CAR T populations at day 9 post-activation (six donors, pooled data from four independent experiments). (C) Representative flow cytometry plots depicting CD45RO and CD95 in isolated T cells from a healthy donor before (left) and 9 days after (right) activation with anti-CD3:CD28 beads. (D and E) CD45RO-negative, CD62L/CD45RA/CCR7/CD127-positive stem cell-like memory T cells in the CD4 (D) or CD8 (E) CAR T populations at day 9 post-activation just before freezing or right after thawing (three donors, CAR T cell manufactured during three independent experiments). (F–M) CAR T cells were thawed and immediately exposed to NALM6 cells at a 1:2 effector-to-target ratio. Tumor cell killing (F and I), T cell expansion (G, H, J, and K), and cytokine expression (L and M) were analyzed by FACS after 72 h (F–H, L, and M) or 7 days (I–K) of co-culture. Data are represented as mean ± SD. Statistics are based on paired, two-tailed Student’s t test (A and B) or on one-way ANOVA (D–M).

Figure 3

MITO-66 conditioning during CAR T manufacturing enhances antitumor efficacy (A) Experimental scheme depicting CD19-CAR T cell treatment of NALM6 leukemia-bearing mice. CAR T cells were prepared from five different donors during three independent manufacturing experiments and frozen at the end of the culture. At the time of ACT, the CAR T cells from all donors were thawed and administered to one or two recipient mice. (B) Survival of mice receiving no treatment, or non-transduced T cells (NTD) or CAR T cells manufactured with DMSO or MITO-66. (C) Number (No.) of CD19-positive NALM6 cells in the blood of mice analyzed by flow cytometry at the indicated time points. (D) Mouse weight expressed as percentage of the weight at the beginning of the experiment. (B and D) Uuntreated, NTD DMSO, and NTD MITO-66: n = 3 mice, CAR DMSO: n = 9 mice, CAR MITO-66: n = 7 mice. (E and F) Number of CD4 (E) or CD8 (F) T cells in the blood at 11 days post-ACT. (G–J) Percentage of CD62L-positive cells in CD4 or CD8 T cell populations at day 11 post-ACT (G and H) or day 33 post-ACT (I and J). (E–H) Four donors, pooled data from two independent experiments. (E–J) Data from mice receiving CAR T cells from the same donor were pooled and average was calculated, represented here is the paired analysis of five donors (E–H) or three donors in surviving mice (I and J). Data are represented as mean ± SD. Statistics are based on two-way ANOVA (C) or on unpaired two-tailed Student’s t test (D) or on paired, two-tailed Student’s t test (E–J). See also Figure S2.

Figure 4

MITO-66-conditioned CAR T cells protect against cancer recurrence (A) Experimental scheme depicting CD19-CAR T cell treatment of NALM6 leukemia-bearing mice following in vivo rechallenge for three times with 2 × 10⁶ and finally 5 × 10⁶ NALM6 cells. CAR T cells were prepared from four different donors and performed in two independent experiments. CAR T cells from each donor were transferred in five to seven different NALM6-bearing mice. Throughout the experiment, some mice were lost because of graft-versus-host disease. (B–E) NALM6 tumor burden measured by bioluminescence at the indicated days post-initial NALM6 engraftment. (F–I) Number of transferred CD3 T cells in the blood of mice analyzed by flow cytometry at the indicated time points. (J and K) Percentage of CD62L-positive cells in the CD4 (E) or CD8 (F) T cell population in the blood at day 10 post-ACT. Data are represented as mean ± SD. Statistics are based on unpaired two-tailed Student’s t test (D) or two-way ANOVA (F–K).

Figure 5

Benchmarking MITO-66 against other small molecules influencing memory differentiation (A) Experimental scheme depicting CD19-CAR T cell generation in the presence of different small-molecule inhibitors (MITO-66 and UK-5099 [25 μM], AKT-VIII [AKTi, 1 μM], idelalisib [PI3KδI, 200 nM], rapamycin [mTORi, 100 nM], enasidenib [IDH2i, 5 μM], and 2-deoxyglucose [2-DG, 2 mM], molecule concentrations were selected based on previous literature). (B) T cell expansion measured by machine-assisted trypan blue-based cell counting at day 5, 7, and 9 post-activation. (C–F) CD62L/CD45RO double-positive cells out of CD4 (C) or CD8 (E) CAR T cells or CD62L-positive, CD45RO-negative stem cell-like memory T cells out of CD4 (D) or CD8 (F) CAR T cells at day 9 post-activation. (B–F) Six donors, pooled data from four independent experiments. (G) Survival of mice receiving no treatment or inhibitor-conditioned CAR T cells. Surviving mice out of total treated mice are indicated in brackets. p values indicate statistical difference versus CAR DMSO. (H) Number (No.) of transferred CD3 T cells in the blood of mice analyzed by flow cytometry at day 11 post-ACT. (I–K) Percentage of CD62L/CD45RO double-positive cells (I), CD62L-positive, CD45RO-negative cells (J), or PD1/TIM3 double-positive cells (K) out of transferred CD3 T cells at day 11 post-ACT. (G–K) Three human donors into 9–10 total mice, pooled data from 2 independent experiments, only flow cytometry data with >20 events were used for phenotypic analysis. Data are represented as mean ± SD. Statistics are based on one-way ANOVA (C–F and H–K) or log rank test (G). See also Figure S3.

Figure 6

MITO-66 induces a stem cell-like memory phenotype in CAR T cells from patients with B cell malignancies (A) T cell yield at day 10 post-activation, expressed as percentage of DMSO-conditioned CAR T cells. (B) Percentage of viable T cells at day 10 post-activation. (A and B) Eight patients, pooled data from three independent experiments. (C and D) Percentage of CD4 (C) and CD8 (D) T cells in live cells at day 10 post-activation. (E) Percentage of CAR-positive T cells in CD4 T cells at day 10 post-activation. (F–I) Percentage of CD127-positive (F), CD62L-positive (G), CD62L/CD45RO-positive T_CM (H), and CD45RO-negative CD62L-positive T_SCM (I) cells in CD4 CAR T cells at day 10 post-activation. (J) Percentage of CAR-positive T cells in CD8 T cells at day 10 post-activation. (K–N) Percentage of CD127-positive (K), CD62L-positive (L), CD62L/CD45RO-positive T_CM (M), and CD45RO-negative CD62L-positive T_SCM (N) cells in CD8 CAR T cells at day 10 post-activation. (C–N) Eleven patients, pooled data from four independent experiments. Data are represented as mean ± SD. Statistics are based on paired, two-tailed Student’s t test.