First-in-class transactivator-free, doxycycline-inducible IL-18-engineered CAR-T cells for relapsed/refractory B cell lymphomas

Abstract

Although chimeric antigen receptor (CAR) T cell therapy has revolutionized type B cancer treatment, efficacy remains limited in various lymphomas and solid tumors. Reinforcing conventional CAR-T cells to release cytokines can improve their efficacy but also increase safety concerns. Several strategies have been developed to regulate their secretion using minimal promoters that are controlled by chimeric proteins harboring transactivators. However, these chimeric proteins can disrupt the normal physiology of T cells. Here, we present the first transactivator-free anti-CD19 CAR-T cells able to control IL-18 expression (iTRUCK19.18) under ultra-low doses of doxycycline and without altering cellular fitness. Interestingly, IL-18 secretion requires T cell activation in addition to doxycycline, allowing the external regulation of CAR-T cell potency. This effect was translated into an increased CAR-T cell antitumor activity against aggressive hematologic and solid tumor models. In a clinically relevant context, we generated patient-derived iTRUCK19.18 cells capable of eradicating primary B cells tumors in a doxycycline-dependent manner. Furthermore, IL-18-releasing CAR-T cells polarized pro-tumoral macrophages toward an antitumoral phenotype, suggesting potential for modulating the tumor microenvironment. In summary, we showed that our platform can generate exogenously controlled CAR-T cells with enhanced potency and in the absence of transactivators.

Keywords: CAR-T cells; IL-18; Lent-On-Plus; MT: Delivery Strategies; TRUCKs; animal models; doxycycline; lentiviral vectors; lymphoma; regulation.

Graphical abstract

Figure 1

Generation and characterization of inducible IL-18-producing CAR-T cells (iTRUCK19.18) (A) CAR19 LV encoding for EF1α-A3B1-41BB-CD3ζ (top) and Dox-inducible LOP LVs expressing pro-IL-18 (bottom). (B) Representative histograms of CAR expression in non-transduced cells (NT) (top), CAR19 (middle), and iTRUCK19.18 −Dox (bottom). (C) Representative histograms of pro-IL-18 expression in CAR19 cells (top) and iTRUCK19.18 cells in the absence (middle) or presence (bottom) of 50 ng/mL Dox (48 h). Right: fold change of IL-18 expression from iTRUCK19.18 cells relative to basal background of NT (n = 17; −Dox: n = 15; +Dox: n = 17) (right). (D) Bioactive IL-18 secreted by iTRUCK19.18 without activation (left) and activating with TransAct (αCD3/CD28) (right) (n = 4). (E) Experimental procedure for CAR19 and iTRUCK19.18 cell generation (single or co-transduction with LVs) and their analysis in resting conditions after 10 days in the absence of stimuli (without Dox). (F) Phenotype of CAR19 and iTRUCK19.18 cells in the absence of Dox. Total T cells (CD3+) were analyzed for CD45RA and CD62L expression (CAR19 cells: n = 5; iTRUCK19.18: n = 7). (G) CD4/CD8 ratio of CAR19 and iTRUCK19.18 cells in the absence of Dox of the total populations (CD3+) (CAR19 cells: n = 5; iTRUCK19.18: n = 7). (H) Expression of AICD markers (TRAIL, FasL, Fas) and phosphorylation of CD3z in CAR19 vs. iTRUCK19.18 cells (−Dox) in CD3+ cells: TRAIL (CAR19 T cells: n = 5; iTRUCK19.18: n = 8), FasL (CAR19 T cells: n = 3; iTRUCK19.18: n = 6), Fas (CAR19 T cells: n = 4; iTRUCK19.18: n = 7), pCD3ζ (CAR19 T cells: n = 3; iTRUCK19.18: n = 7). (I) Index of proinflammatory cytokine-related secretion by CAR19 T cells vs. iTRUCK19.18 cells at basal state (n = 2). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001 (two-tailed paired t test).

Figure 2

Characterization of iTRUCK19.18 cells in absence and presence of Dox (A) Scheme for iTRUCK19.18 cell generation and analysis at basal state or after activation with αCD3/CD28. (B) Expression of AICD markers (TRAIL, FasL, Fas), phosphorylation of CD3z and apoptosis marker Annexin V (from left to right) in iTRUCK19.18 cells at basal state in the absence (light blue) or presence of 50 ng/mL Dox (dark blue): TRAIL (n = 7), FasL (n = 6), Fas (n = 7), pCD3ζ (n = 7), and Annexin V (n = 4). Analysis performed on total CD3+ T cells. (C) Phenotype of iTRUCK19.18 cells with (dark blue) and without Dox (light blue) at resting conditions (left) (n = 7) or after 4 h of stimulation (right) (n = 4). Analysis performed on total CD3+ T cells. (D) Percentage of positive cells in the CD3+ population for exhaustion markers PD1, LAG3, TIM3 of iTRUCK19.18 cells at basal state and after stimulation in the absence or presence of 50 ng/mL Dox (n = 4). (E) Fold change was calculated by dividing the percentage of +Dox population by the −Dox population (% at +Dox/% −Dox) for each of the activation markers analyzed (IFN-γ, TNF-α, IL-2, Granzyme B, and Perforin A) in iTRUCK19.18 cells, both without Dox (light blue) and with Dox (dark blue), at resting state (left) or after activation (right) (n = 4). (F) Secretion of proinflammatory cytokines from CAR19 (gray) and iTRUCK19.18 cells with (dark green) and without (light blue) Dox at basal state (n = 4) and after 24 h of stimulation with TransAct. Two-way ANOVA, multiple comparison Tukey’s test. ∗p < 0.05, ∗∗∗p < 0.001.

Figure 3

In vitro and in vivo evaluation of iTRUCK19.18 cells against B cell lymphoma model (A) Diagram of in vitro cytotoxicity assay: 50 ng/mL of Dox was added at the moment of the co-culture and the dose was refreshed in every challenge. (B) Percentage of surviving Namalwa cells after serial tumor encounters with CAR19 or iTRUCK19.18 cells without and with Dox (n = 4). (C) In vivo experimental procedure to evaluate iTRUCK19.18 at a therapeutic dose (1 × 10⁶ CAR-T cells/mouse). Dox (1,000 ng/mL) was added to drinking water after infusion into mice and refreshed twice a week. Two more tumor challenges with Namalwa cells were infused at days 25 and 42, respectively. (D) Proportion of circulating human T cells (left) and relative expression of IL-18 (right) by T cells obtained from blood 7 days after infusion of iTRUCK19.18 cells. (E) Bioluminescence images of tumor progression in mice treated with PBS, NT, CAR19, and iTRUCK19.18 without and with Dox. As control of re-challenges 1 and 2 (R1 and R2), novel mice were also infused with PBS at days +25 and +39. (F) Survival graph of mice treated with PBS, NT, CAR19, and iTRUCK19.18 without (−Dox, blue line) and with (+Dox, green line) Dox. (G) Percentage of viable tumor cells in different organs (spleen, bone marrow, brain, and blood, from left to right) of mice treated with NT, CAR19, and iTRUCK19.18 without (−Dox) and with Dox (+Dox), at final point (PBS: N = 5; NT: N = 4; CAR19: N = 6; iTRUCK19.18 −Dox: N = 5; iTRUCK19.18 +Dox: N = 5). (H) Diagram representing the infusion of a subtherapeutic dose (3 × 10⁵ CAR-T cells) into mice 6 days post-tumor. Dox (500 ng/mL) was added to drinking water after the infusion of the CAR-T cells into mice and was refreshed twice a week. (I) Tumor progression determined by bioluminescence (photons/s) of the different experimental groups (NT, CAR19, iTRUCK19.18 −Dox, and iTRUCK19.18 +Dox. (J) Percentage of surviving tumor cells in the spleen (left) and bone marrow (right) of the mice from the different experimental groups. (K) Percentage of tumor-infiltrated T cells (hCD3+) in spleen (left) and bone marrow (right) of mice at the time of sacrifice. (L) Proportion of T_{naive/SCM+TCM} cells in the spleen (left) bone marrow (right) of mice at endpoint (NT: N = 4; CAR19: N = 5; iTRUCK19.18 −Dox: N = 6; iTRUCK19.18 +Dox: N = 5). ∗p < 0.05, ∗∗p < 0.01, ∗∗∗∗p < 0.0001 (one-tailed paired t test for B; log rank test for F; one-tailed unpaired t test for D, G, I, J, and L).

Figure 4

Characterization and lytic capacity of patient-derived iTRUCK19.18 cells against primary B tumors (A) Representative histograms of CD19 expression of primary tumor sample-derived MZL (patient 1, leukemic mantle cell lymphoma), CLL (patients 2 and 3, chronic lymphocytic leukemia), and Namalwa cells (from Burkitt’s lymphoma). (B) Scheme of the generation and analysis of patient-derived CAR-T cells. (C) Percentage of CAR+ cells of patient-derived CAR19, iTRUCK19.18, and cTRUCK19.18 cells (n = 3). (D) Fold change of pro-IL-18 expression of patient-derived CAR19 and iTRUCK19.18 in the absence (−Dox) or presence (+Dox) of 50 ng/mL Dox and cTRUCK19.18 (n = 3). (E) Surviving CD19+ tumor cells following encounter with NT, patient-derived iTRUCK19.18 in the absence (−Dox) or presence (+Dox) of 50 ng/mL Dox and cTRUCK19.18 cells, at an E:T ratio of 1:5 and measured after 13 h of co-culture. (F) Representative histograms (left) and graph (right) showing the percentage of Tox expression of patient-derived NT, CAR19, or iTRUCK19.18 cells without and with Dox and cTRUCK19.18 cells. Analysis performed on total CD3+ T cells (NT, iTRUCK19.18 −Dox, and +Dox: n = 5; CAR19: n = 5; cTRUCK19.18 cells: n = 3). ∗p < 0.05, ∗∗p < 0.01 (two-tailed paired t test).

Figure 5

In vitro and in vivo efficacy of iTRUCK19.18 cells against CD19+ pancreatic ductal adenocarcinoma tumor model (A) Experimental procedure of the in vitro cytotoxicity assay with an artificial model of PDAC cells, MIA-PaCa2-CD19+. Left: histogram showing CD19 expression in MIA-PaCa2 WT (gray) or CD19+ (blue), as target of CAR19 and iTRUCK 19.18. Right: iTRUCK19.18 cells in the absence or presence of Dox (50 ng/mL) were co-cultured at an effector:target ratio of 1:2 with MIA-PaCa2-CD19+ cells. Tumor re-challenges and FACS analysis were performed every 48 h. (B) Specific lysis over four tumor encounters of iTRUCK19.18 −Dox (light blue) and +Dox (dark blue) compared with NT (n = 5). (C) Proportion of T_naive/SCM and T_CM from −Dox (light blue) and +Dox (dark blue) iTRUCK19.18 after 48 h of every tumor encounter (n = 5). Analysis performed on total CD3+ T cells. (D) Proportion of PD1+ TIM3+ cells inside the CD3+ population, analyzed at every tumor encounter (n = 5). (E) Bioluminescence of tumor progression in vivo up to day +37. PBS (N = 3), NT (N = 3), CAR19 (N = 4), iTRUCK19.18 (−Dox, N = 4), and iTRUCK19.18 (+Dox, N = 4). Dox was added at the moment of the inoculation. (F) Tumor progression (photons/s) in mice treated with NT, CAR19, or iTRUCK19.18 without and with Dox. (G) Percentage of tumor cells in the pancreas of mice treated with NT, CAR19, or iTRUCK19.18 without and with Dox. ∗p < 0.05, ∗∗p < 0.01 (two-tailed paired t test for B and C, and one-tailed unpaired t test for G).

Figure 6

Dox addition to iTRUCK19.18 cells induces polarization of primary human M2 macrophages toward an M1 phenotype (A) Experimental diagram of the generation of iTRUCK19.18 and M2-polarized macrophages from the same donor. When indicated, Dox (50 ng/mL) was added at the beginning of the co-culture. (B) Left: representative histograms corresponding to CD206 expression of macrophages co-cultured with the different groups of T cells. Center and right: percentage of CD206+ M2 (center) and M1 (right) macrophages when co-cultured with NT cells, iTRUCK19.18 cells without and with Dox, and with MIA-PaCa2 CD19+ cells at ratio 1:1 MIA-PaCa2 vs. CAR-T (M2: n = 5; M1: n = 3). (C) Representative dot plots and quantification of the proportion of T_naive/SCM and T_CM cells of iTRUCK19.18 cells in the presence of macrophages with and without Dox after 5 days of the co-culture (n = 3). Analysis performed on total CD3+ T cells. (D) Percentage of PD1 and TIM3+ cells inside the CD3+ population of iTRUCK19.18 cells after the co-culture in the presence (dark blue) or absence (light blue) of Dox (n = 3). Representative histograms and fold change of (E) IFN-γ and (F) TNF-α expression (compared with those without Dox) of iTRUCK19.18 cells in co-culture with macrophages with and without Dox (n = 3). ∗p < 0.05, ∗∗∗p < 0.001 (two-tailed paired t test for B; one-tailed paired t test for C, D, E, and F).