Targeted delivery of a PD-1-blocking scFv by CD133-specific CAR-T cells using nonviral Sleeping Beauty transposition shows enhanced antitumour efficacy for advanced hepatocellular carcinoma

Abstract

Background: CD133 is considered a marker for cancer stem cells (CSCs) in several types of tumours, including hepatocellular carcinoma (HCC). Chimeric antigen receptor-specific T (CAR-T) cells targeting CD133-positive CSCs have emerged as a tool for the clinical treatment of HCC, but immunogenicity, the high cost of clinical-grade recombinant viral vectors and potential insertional mutagenesis limit their clinical application.

Methods: CD133-specific CAR-T cells secreting PD-1 blocking scFv (CD133 CAR-T and PD-1 s cells) were constructed using a sleeping beauty transposon system from minicircle technology, and the antitumour efficacy of CD133 CAR-T and PD-1 s cells was analysed in vitro and in vivo.

Results: A univariate analysis showed that CD133 expression in male patients at the late stage (II and III) was significantly associated with worse progression-free survival (PFS) (P = 0.0057) and overall survival (OS) (P = 0.015), and a multivariate analysis showed a trend toward worse OS (P = 0.041). Male patients with advanced HCC exhibited an approximately 20-fold higher PD-L1 combined positive score (CPS) compared with those with HCC at an early stage. We successfully generated CD133 CAR-T and PD-1 s cells that could secrete PD-1 blocking scFv based on a sleeping beauty system involving minicircle vectors. CD133 CAR-T and PD-1 s cells exhibited significant antitumour activity against HCC in vitro and in xenograft mouse models. Thus, CD133 CAR-T and PD-1 s cells may be a therapeutically tractable strategy for targeting CD133-positive CSCs in male patients with advanced HCC.

Conclusions: Our study provides a nonviral strategy for constructing CAR-T cells that could also secrete checkpoint blockade inhibitors based on a Sleeping Beauty system from minicircle vectors and revealed a potential benefit of this strategy for male patients with advanced HCC and high CD133 expression (median immunohistochemistry score > 2.284).

Keywords: CAR-T; CD133; Hepatocellular carcinoma; Minicircle; PD-1; Sleeping Beauty system.

Fig. 1

Representative image of CD133 and PD-L1 detection in different stages in HCC patients. A HE images and corresponding CD133 expression in primary tumour tissue in liver cancer patients and lung metastases. B The percentage of CD133 expression in primary tumours in early (stage I) and advanced stages (stage II and III) and metastases in the lung. C Kaplan–Meier curve depicting survival analyses in stage II and III male patients. The P value was calculated by the log-rank test. D Representative image of PD-L1 expression at different stages and in lung metastases of HCC patients. E The PD-L1 expression at different stages and in lung metastases of HCC patients were evaluated using the CPS score method and compared. Mean ± SD, n = 19, two-tailed unpaired Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant

Fig. 2

Construction of CD133 CAR-T and PD-1 s cells based on sleeping beauty transposons from minicircles. A Schematic diagram of CAR-T cell preparation and sequences of SB100, CD133 CAR-T and CD133 CAR-T and PD-1 s cells. B The minicircles from CD133 CAR, CD133 and PD-1 s CAR and SB100 cells were evaluated by agarose gel electrophoresis. C Representative flow cytometry analyses of CAR-T cell expression using a c-Myc-tag as shown in the surrogate. D The cell viability of CAR-T cells from 3 different donors. E CAR expression detected at different time points from 3 different healthy donors. One colour represents one patient sample. F Western blot detection of the secreted PD-1 blocking scFv in the supernatant of Mock T cells, CD133 CAR-T cells and CD133 CAR-T and PD-1 s cells

Fig. 3

Enhanced specific cytotoxic effects of CD133 CAR-T and PD-1 s cells against CD133 + HCC cell lines. A CAR expression in Mock, CD133 CAR-T and CD133 CAR-T and PD-1 s cells was assessed by flow cytometry after enrichment with anti-c-Myc-tag-biotin and anti-biotin microbeads. B RT-QPCR and western blot analysis of the CAR gene number and CD3ζ expression in Mock T, CD133 CAR-T, and CD133 CAR-T and PD-1 s cells. C The CD8-positive subset of CD133 CAR-T and PD-1 s cells and their CCR7 and CD45R expression by flow cytometric analysis. D Flow cytometry analysis of CD133 expression in HCC cell lines (SK-Hep-1 and Hep3B). E The upper images show Mock T, CD133 CAR-T and CD133 CAR-T and PD-1 s cells targeting Hep3B cells. The red arrow represents the CAR-T cells, and the yellow arrow represents the tumour. The lower image shows the proportion of Mock T, CD133 CAR-T cells or CD133 CAR-T and PD-1 s cells incubated with CD133 + HCC cell lines (Hep3B) after 24 h. F Specific lysis (ratio of 7AAD-positive cells to CD133 + tumour total cells) after 3 days of incubation of CAR-T cells with tumour cells. G Representative flow cytometric image of CD107a expression in Mock T cells, CD133 CAR-T cells, and CD133 CAR-T and PD-1 s cells incubated with tumour cells for 4 h. H The frequency of CD107a expression on Mock T, CD133 CAR-T, CD133 CAR-T and PD-1 s cells alone or incubated with Hep3B cells by flow cytometric analysis. I Cytokine production (IFNγ and TNFα) by Mock T cells, CD133 CAR-T cells, and CD133 CAR-T and PD-1 s cells with or without tumour cell incubation for 4 h was measured by ELISA. Mean ± SD, n = 4, two-tailed unpaired Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001, ns, not significant

Fig. 4

The PD-1 blocking scFv secreted by CD133 CAR-T and PD-1 s can bind bystander PD-1 + T cells and restore T cell function. A Schematic diagram of the PD-1 overexpression lentivirus with a GFP flag and a puromycin selection marker. B Flow cytometry analysis of PD-1 and GFP expression in T cells or T cells transfected with a PD-1 overexpression lentivirus and then selected with puromycin. C Flow cytometry analyses of the secreted PD-1 blocking scFv binding to normal T cells or PD-1 T cells. D Cellular immunofluorescence assay of PD-1 (GFP) and PD-1 blocking scFv (anti c-Myc-tag-647, Red) expression on normal T cells, PD-1 T cells, and PD-1 T cells incubated with the condensed supernatant from CD133 CAR-T and PD-1 s cells. E Representative flow cytometric images of granzyme B and TNFα expression in PD-1 T cells, PD-1 T cells incubated with Hep3B, and PD-1 T cells incubated with Hep3B and PD-1 blocking scFv. F, G Quantification of the frequency of granzyme B/PD-1 T cells and TNFα/PD-1 T cells by flow cytometric analysis of PD-1 + T cells, PD-1 + T cells incubated with Hep3B cells and PD-1 + T cells incubated with Hep3B cells and PD-1-blocking scFv secreted by CD133 CAR-T and PD-1 s cells. Mean ± SD, n = 3, two-tailed unpaired Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant

Fig. 5

CD133 CAR-T and PD-1 s cells have better antitumour effects than CD133 CAR-T cells in vivo after intravenous injection. A Schematic diagram of CAR-T cell treatment and tumour detection. NCG mice with Hep3B xenograft tumours were intravenous injected with two doses of 5 million Mock T cells, CD133 CAR-T cells or CD133 CAR-T and PD-1 s cells once Hep3B xenograft tumours reached 50–150 mm³, as revealed by bioluminescence imaging. B The tumours of Mock T cell-, CD133 CAR-T cell- and CD133 CAR-T and PD-1 s cell-treated groups were monitored weekly by bioluminescence imaging. C Growth curves of Mock T cell-, CD133 CAR-T cell- and CD133 CAR-T and PD-1 s cell-treated groups in Fig. 5B. D The tumour volumes of the Mock T cell-, CD133 CAR-T cell- and CD133 CAR-T and PD-1 s cell-treated groups were measured by bioluminescence imaging. Mean ± SD, n = 5, two-tailed unpaired Student’s t test, *P < 0.05, **P < 0.01, ***P < 0.001, ns, not significant. E Kaplan–Meier survival curves of Mock T, CD133 CAR-T and CD133 CAR-T and PD-1 s cell-treated groups. **P < 0.01 (log-rank test); ns, not significant. F Representative images of HE, CD133 and CD3 staining of the Mock T, CD133 CAR-T and CD133 CAR-T and PD-1 s cell-treated groups. G Percentage of CD3 expression in tumours in all the mice in Fig. 5F belonging to the Mock T, CD133 CAR-T and CD133 CAR-T and PD-1 s cell-treated groups

Fig. 6

CD133 CAR-T and PD-1 s cells eliminate tumour cells in a Hep3B metastatic and in situ xenograft tumour mouse models. A Schematic diagram of CAR-T cell treatment. In brief, NCG mice were intravenously injected with 2.5 million Hep3B-luc cells. Metastasis was confirmed by bioluminescence imaging, and the groups were then treated with a tail veil injection of 5 million Mock T, CD133 CAR-T, or CD133 CAR-T and PD-1 s cells 55 days after inoculation. B Bioluminescence imaging of Mock T cell-, CD133 CAR-T cell- or CD133 CAR-T and PD-1 s cell-treated groups in a metastatic model of HCC. C The growth curves of the Mock T cell-, CD133 CAR-T cell- and CD133 CAR-T and PD-1 s cell-treated groups are shown in Fig. 6B. Mean ± SD, n = 3, two-tailed unpaired Student’s t test, **P < 0.01, ns, not significant. D Schematic diagram of CAR-T cell treatment. In brief, 2.5 million Hep3B-luc cells were injected into the livers of NCG mice. The tumour was confirmed by bioluminescence imaging and then treated with a tail veil injection of 5 million Mock T, CD133 CAR-T or CD133 CAR-T and PD-1 s at 22 days after inoculation. E The bioluminescence imaging of Mock T cell-, CD133 CAR-T cell- or CD133 CAR-T and PD-1 s cell-treated groups in an situ xenograft tumour model of HCC. Mean ± SD, n = 5, two-tailed unpaired Student’s t test, *P < 0.05, ns, not significant. F, G The total flux and the metastasis flux of the Mock T cell-, CD133 CAR-T cell- and CD133 CAR-T and PD-1 s cell-treated groups in Fig. 6E. Mean ± SD, n ≥ 2, two-tailed unpaired Student’s t test, **P < 0.01, ns, not significant. H The bar graph shows the frequency of CAR-T cells 16 days after the last instance of Mock T, CD133 CAR-T and CD133 CAR-T and PD-1 s cell therapy identified by flow cytometry using anti-human CD3