PSMA-specific CAR-engineered T cells eradicate disseminated prostate cancer in preclinical models

Abstract

Immunology-based interventions have been proposed as a promising curative chance to effectively attack postoperative minimal residual disease and distant metastatic localizations of prostate tumors. We developed a chimeric antigen receptor (CAR) construct targeting the human prostate-specific membrane antigen (hPSMA), based on a novel and high affinity specific mAb. As a transfer method, we employed last-generation lentiviral vectors (LV) carrying a synthetic bidirectional promoter capable of robust and coordinated expression of the CAR molecule, and a bioluminescent reporter gene to allow the tracking of transgenic T cells after in vivo adoptive transfer. Overall, we demonstrated that CAR-expressing LV efficiently transduced short-term activated PBMC, which in turn were readily stimulated to produce cytokines and to exert a relevant cytotoxic activity by engagement with PSMA+ prostate tumor cells. Upon in vivo transfer in tumor-bearing mice, CAR-transduced T cells were capable to completely eradicate a disseminated neoplasia in the majority of treated animals, thus supporting the translation of such approach in the clinical setting.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. CAR development and validation.

(A) Anti-hPSMA CAR map. EC: extracellular domain; TM: transmembrane domain; IC: intracellular domain. (B) Cytofluorimetric profile of CAR expression in transfected 293 T cells. 293 T cells were transfected with pcDNA3.1 vector bearing CAR sequence and analysed by flow cytometry after 24 hours (dark line). Untransfected cells (grey plot) served as control. (C) CAR expression as assessed by Western blotting at 24 hours (line 1) and 7 days (line 2) post 293 T cell transfection. Negative controls were the medium alone (line 3) and untransfected 293 T cells (line 4). Line 5, molecular weight markers. (D) Linear map of the recombinant viral vector containing the minCMVPGK bidirectional promoter (22). In particular, the reporter gene (eGFP or Luciferase) is under the control of minCMV promoter, while the CAR gene is under control of hPGK promoter. (E) Transduction of Jurkat cells with LV CAR anti-hPSMA/eGFP. Co-expression of c-myc and eGFP in LV-transduced Jurkat cells, as assessed by flow cytometry. Dot plot reports the events gated on total viable cells; more than 90% of cells co-express both c-myc and eGFP. (F) Transduction of Jurkat cells with LV CAR anti-hPSMA/Luciferase. Left panel, c-myc expression (black) in Jurkat cells at 72 h post transduction, as assessed by flow cytometry. Grey plot represents the isotype control. Right panel, Assessment of luciferase activity in LV-transduced Jurkat cells (2×105/well) by BLI.

Figure 2. Generation, expansion and phenotipic properties of T-bodies.

(A) Set up of transduction protocol and growth kinetics. Left panel, PBMC were activated with OKT3 at time 0 and transduced concomitantly (Td0), or after 48-h (Td2) or 72-h (Td3) hours. Flow cytometry analysis of c-myc tag expression was performed 72 hours post infection. Central panel reports the expression of CD27, CD28 and CD62L in the three different conditions of activation/infection reported in the left panel. Right panel, Expansion of T-body-hPSMA/fluc and T-body-hPSMA/eGFP transduced after 48 h activation in response to weekly restimulation with PC3-PIP cells. Untransduced PBMC were used as control. (B) Phenotype and CAR expression in transduced T cells upon stimulation. Left panel, Expression of surface markers in T-body-anti-hPSMA/fluc at different time points (3, 11 and 18 days) post transduction, as assessed by flow cytometry. Data are representative of 3 independent experiments. Overlapping results were obtained in T-body-anti-hPSMA/eGFP. Right panel, percentage of c-myc+ cells in LV CAR hPSMA/fluc and LV CAR hPSMA/eGFP-transduced T cell populations at different time points post transduction. Figure shows mean +/− SD of at least three independent experiments. (C) Kinetics of CD4 and CD8 subsets in transduced T cell populations. Expression of CD4 and CD8 markers in c-myc+-gated T-body-hPSMA/eGFP (left panel) and T-body-hPSMA/fluc (right panel) populations at different time points post transduction, as assessed by flow cytometry. Data are from a representative experiment out of three that produced similar results.

Figure 3. Functional characterization of T-bodies.

(A) Lytic activity of T-body-hPSMA/eGFP and T-body-hPSMA/fluc. Cytotoxicity was analysed at 3, 11 and 60 days post-transduction; PC3, PC3-PIP and LNCaP cells were used as target cells. Untransduced PBMC served as negative control. In the upper left corner of each panel the percentages of c-myc+ T cells are reported. Figure shows mean +/− SD of 4 independent experiments. (B) IFN-γ secretion upon antigen stimulation. IFN-γ production was analyzed at different time points after PBMC transduction by stimulating T-body-hPSMA/fluc with PC3-PIP hPSMA+ or PC3 hPSMA− cancer cell lines. T-bodies unstimulated or treated with PMA/Ionomycin represented the negative and positive controls, respectively. Similar results were obtained with T-body-hPSMA/eGFP. (C) c-myc expression in T-body-hPSMA/fluc populations tested for IFN-γ production.

Figure 4. Assessment of T-body in vivo loco-regional therapeutic efficacy.

(A) Winn Assay. PC3-PIP (left panel) and PC3 (right panel) tumor cells were inoculated s.c. in SCID mice, alone or mixed 1∶1 with T-body-hPSMA/eGFP at opposite flanks of the same animal. Tumor growth was monitored over time by caliper measurement. Number of mice per group, n = 6. (B) Loco-regional therapy. T-body-hPSMA/eGFP at 72 hours post transduction were administered intralesionally and perilesionally in SCID mice 4 days after s.c. injection of PC3-PIP tumor cells (n = 6); untreated animals served as control group (n = 6). Left panel shows tumor volumes, while right panel reports Kaplan-Meyer survival curves of treated and untreated mice. (C) Expression of hPSMA antigen in prostate tumors. PC3-PIP tumor cells from in vitro cultures (first quadrant) or isolated ex-vivo from control or treated mice (second and third quadrant, respectively; dark line) were evaluated for hPSMA expression by flow cytometry. The grey plot corresponds to the isotype control.

Figure 5. Assessment of T-body in vivo systemic therapeutic efficacy against disseminated prostate carcinoma.

Rag2−/−/γc−/− mice (n = 6) and NOD/SCID mice (n = 6) bearing established bioluminescent PC3-PIP tumors were injected i.v. with 2×107 T-body-hPSMA/eGFP for 3 times at a two day interval. Untreated animals (n = 6 for both mouse strains) were used as controls. (A) Pictures show two representative Rag2−/−/γc−/− and NOD/SCID mice (left and right panels, respectively) imaged by BLI at different time points, whereas (B) and (C) graphs report cumulative results of the regions of interest (ROI) in lungs and in total body, respectively. Tumor growth was monitored as photon flux and quantified as photon * sec−1 * cm−2 * sr−1. Graphs show mean ± SD of three independent experiments. *: P<0.05. The t-Test was used for statistical analysis. (D) Cumulative Kaplan-Meier survival curves of Rag2−/−/γc−/− (left panel; untreated mice, black line; median survival = 54 days; treated mice, red line; median survival = 74 days; P = 0.046) and NOD/SCID mice (right panel; untreated mice, black line; median survival = 60 days; treated mice, red line; median survival = not evaluable; P<0.001).