Akt inhibition at the initial stage of CAR-T preparation enhances the CAR-positive expression rate, memory phenotype and in vivo efficacy

Abstract

The adoptive transfer of chimeric antigen receptor-modified T (CAR-T) cells is a novel cancer treatment that has led to encouraging breakthroughs in the treatment of haematological malignancies. The efficacy of infused CAR-T cells is associated with a high CAR-positive expression rate, a strong proliferative response and the persistence of CAR-T cells in vivo. Manufacturing CAR-T cells is a process usually associated with the decreased CAR-positive expression rate and terminal differentiation of the infused CAR-T cells, which causes decreased proliferation and persistence of CAR-T cells in vivo. Therefore, the preparation of a high CAR-positive expression rate and few differentiated CAR-T cells is particularly important for clinical cancer treatment. In this study, we transduced and expanded CAR-T cells targeting the epithelial cell adhesion molecule (EpCAM) in the presence of an Akt inhibitor (MK2206) during the initial stage of CAR-T cell preparation. We show that the Akt inhibitor did not suppress the proliferation or effector function of the EpCAM-CAR-T cells but increased the CAR-positive expression rate and decreased the number of terminally differentiated EpCAM-CAR-T cells. Furthermore, EpCAM-CAR-T cells prepared using this protocol appeared to have enhanced antitumor activity in vivo. Taken together, these findings suggest that Akt inhibition during the initial stage of CAR-T cell preparation could improve the performance of CAR-T cells.

Keywords: Akt inhibitor; CAR-T; CAR-positive expression rate; MK2206; memory phenotype.

Figure 1

The Akt inhibitor MK2206 significantly suppresses Akt phosphorylation at the initial stage of CAR-T cell preparation. A. Schematic illustration of EpCAM-specific CAR. Sequences encoding EpCAM-scFv with a c-myc tag at the N-terminus were fused with the transmembrane domain of human CD8α and the cytoplasmic regions of 4-1BB and CD3ζ. EF1α: promoter; SP: signal peptide; scFv: single-chain variable fragment. B. Schematic diagram of CAR-T cell preparation. PBMCs were activated with CD3/CD28 Dynabeads in medium containing 200 IU/mL rhIL-2 and a vehicle control or 0.5-4 μM Akt inhibitor MK2206. Two days later (day 2), the activated cells were transduced with a lentivirus encoding EpCAM-specific CAR (MOI = 5). Five days later (day 5), Dynabeads and the MK2206 were removed, and the lentivirus-infected T cells were continuously expanded for another 5-9 days. LV Td, lentivirus transduction. C. The inhibitory effects of MK2206 on Akt phosphorylation in T cells were detected by flow cytometry. T cells treated with MK2206 or the vehicle were collected on days 2, 5 and 10 and intracellularly stained with antibodies against phosphorylated Akt (p-Akt). D. The inhibitory effects of MK2206 on Akt phosphorylation in T cells as detected by western blotting. T cells treated with MK2206 or the vehicle were collected on days 2, 5 and 10. Total proteins were extracted. Pan Akt and p-Akt were detected with antibodies against Akt and p-Akt, respectively. GAPDH was also detected as the internal control. E. The quantitative analysis results of p-Akt in the western blot analysis. The intensity of each strip was analyzed by ImageJ software. The average intensities of p-Akt were standardized to GAPDH. Shown are combined data from 3 independent experiments with mean ± SEM. **P < 0.01; ***P < 0.001; ns, not significant.

Figure 2

Akt inhibition enhanced the CAR-positive expression rate of EpCAM-CAR-T cells through cell cycle arrest rather LDL receptor expression. (A) Representative data of CAR expression were detected by flow cytometry. The EpCAM-specific CAR-T cells were prepared as described in Figure 1B, with treatment with the vehicle or 0.5~4 μM MK2206. CAR-T cells were collected on days 7 and 14. CAR expression was detected by flow cytometry with an anti-c-myc tag antibody and an Alexa Fluor 647-conjugated second antibody. (B) Statistical results of the CAR expression in CAR-T cells treated with the vehicle or 2 μM MK2206 on day 7 and day 14. (C, D) Representative and statistical data of the cell cycle analysis of MK2206-treated T cells by flow cytometry. PBMCs were activated with CD3/CD28 Dynabeads in medium containing 200 IU/mL rhIL-2 and a vehicle or 2 μM MK2206. Two days later, the cell cycle of the T cells was analysed by flow cytometry. (E, F) Determination of LDL receptor (LDL-R) expression by flow cytometry. The vehicle or MK2206-treated T cells were incubated with a mouse anti-human LDL-R monoclonal antibody or isotype control antibody and subsequently stained with an Alexa Fluor 488-conjugated goat anti-mouse secondary antibody. Representative data (E) and quantitative results (F) are shown. n = 3. (G) The LDL-R expression was determined by western blotting. Total proteins were extracted from the vehicle or MK2206-treated T cells. The LDL-R was detected with the mouse anti-human LDL-R monoclonal antibody. GAPDH was also detected as the internal control. The results showed a consistent trend. Shown are combined data from 3 independent experiments with mean values ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001.

Figure 3

MK2206 treatment generates EpCAM-CAR T cell populations with memory-like characteristics. (A, B) Representative data of flow cytometry (A) and comparative analysis results (B) of memory phenotype analysis of CAR-T cells. EpCAM-specific CAR-T cells were prepared as described in Figure 1B under treatment with the vehicle or 0.5~4 μM MK2206. CAR-T cells were collected on days 7 and 14. The memory phenotype of CAR-T cells was analysed by flow cytometry with anti-CCR7 and CD45RA antibodies. T_SCM, T memory stem cell; T_CM, central memory T cell; T_EM, effector memory T cell; T_TE, terminally differentiated effector T cells. (C) FoxO1 in the nucleus and cytoplasm was detected by western blot. PBMCs were activated with CD3/CD28 Dynabeads in medium containing 200 IU/mL rhIL-2 and vehicle or 2 μM MK2206. Five days later, the activated T cells were collected. The nuclear and cytoplasmic proteins were extracted using the nuclear and cytoplasmic protein extraction kit. FoxO1 in the nucleus and cytoplasm was analysed with an anti-FoxO1 antibody. PARP in the nucleus and β-tubulin in the cytoplasm were also detected as internal controls. (D) The quantitative analysis results of FoxO1 in the western blot analysis. The intensity of each strip was analyzed by ImageJ software. The average intensities of FoxO1 were standardized to β-Tubulin for cytoplasmic or PARP for nuclear. Shown are combined data from 3 independent experiments with mean ± SEM; ***P < 0.001. (E) The intracellular distribution of FoxO1 was analysed by immunofluorescence. Cell samples were prepared as described in (C). Five days after treatment, cells were collected and fixed with 4% paraformaldehyde. FoxO1 was probed with a primary anti-FoxO1 antibody and a secondary antibody. Hoechst 33342 counterstain was used to mask the nucleus. The images were taken with a confocal microscope.

Figure 4

MK2206 treatment did not suppress the proliferation or ameliorate cell-mediated cytotoxicity. The EpCAM-CAR-T cells were prepared as described in Figure 1B. (A, B) Effects of MK2206 treatment on the proliferation of EpCAM-CAR-T cells. The cells were counted every day from day 6 to day 10 (A) by a Countess II Automated Cell Counter (Thermo Fisher). On Day 11, a total of 1 × 10⁶ cells was taken from each group and continuously expanded. The cells were continuously counted until Day 15 (B). (C) The EpCAM expression in human normal colon cell lines and colon cancer cell lines were analysed by flow cytometry. The cell samples were incubated with an isotype control antibody (red open histogram) or an FITC-conjugated anti-EpCAM antibody (blue open histogram) and analysed by flow cytometry. (D) The release of cytokine was analysed by ELISA. The CAR-T cells treated with a vehicle or 2 μM MK2206 co-cultured with target cells at effector/target (E:T) = 1:1 for 24 h. The released IFN-γ and granzyme B in the supernatant were detected by ELISA. (E) The cytotoxicity was analysed by RTCA. The target cells were seeded at a density of 1 × 10⁴ cells/well in an E-Plate and then transferred to the incubator. When the target cells reached a logarithmic growth phase, the empty vector-transduced T cells, the vehicle-treated CAR-T cells or the MK2206-treated CAR-T cells were added into the plate at E:T = 1 in duplicate. The proliferation of the target cells was continuously monitored by the RTCA machine. Ctrl-T, empty vector transduced T cell; Vehicle, vehicle-treated EpCAM-CAR-T cells; MK2206, MK2206-treated EpCAM-CAR-T cells. *P < 0.05; ***P < 0.001.

Figure 5

EpCAM-CAR-T treatment can prolong the survival of NPG mice with metastatic tumor from human colon cancer. (A) Schematic diagram showing the treatment programme of the mice. NPG mice were injected with 2 × 10⁶ HCT116 cells via the tail vein to establish a metastasis model of human colon cancer. On day 7, the mice were randomly assigned into 3 groups (n = 6). The Ctrl-T group received 1 × 10⁷ untransduced T cells. The CAR-T group received 1 × 10⁷ EpCAM-CAR-T cells. All mice were intraperitoneally (i.p.) administered IL-2 (2000 IU/mouse) daily during the treatment. The experiment ended on day 50. (B, C) Results from the analysis of CAR-T cell persistence in vivo were based on flow cytometry. Blood (50 µL) was obtained from the tail vein on day 14. After red blood cell lysis, the cell samples were stained with anti-human CD45 and anti-CD3 antibodies and analysed by flow cytometry. Representative data (B) and statistical results (C) are shown. (D) Overall survival of the NPG mice bearing the established metastatic model of human colon cancer following Ctrl-T or EpCAM-CAR-T treatment. ***P < 0.001.

Figure 6

MK2206-treated EpCAM-CAR-T cells exhibited better antitumor efficacy against a metastatic model of human colon cancer established in NPG mice. (A) Schematic diagram showing the treatment programme of the mice. The NPG mice were injected with 2 × 10⁶ HCT116^luc+ cells via the tail vein to establish a metastasis model of human colon cancer. On day 7, the mice were randomly assigned to 2 groups (n = 5). The vehicle group received 3 × 10⁶ vehicle-treated EpCAM-CAR-T cells. The MK2206 group received 3 × 10⁶ 2 μM MK2206-treated EpCAM-CAR-T cells. All mice were intraperitoneally (i.p.) administered IL-2 (2000 IU/mouse) daily during the treatment. The experiment ended on day 40. (B, C) Results from the analysis of CAR-T cell persistence in vivo by flow cytometry. Blood (50 µL) was obtained from the tail vein on day 21. After red blood cell lysis, the cell samples were stained with anti-human CD45 and anti-CD3 antibodies and analysed by flow cytometry. Representative data (B) and statistical results (C) are shown. (D, E) The therapeutic efficacy was evaluated through luminescence imaging with an IVIS system. Luminescence images (D) and quantitative results (E) of the tumor luminescence intensity showed the tumor burden on day 28. (F) The overall survival of the NPG mice bearing the established metastatic model of human colon cancer following injection of the vehicle or MK2206-treated EpCAM-CAR-T cells. *P < 0.05.