Blockade of CD7 expression in T cells for effective chimeric antigen receptor targeting of T-cell malignancies

Abstract

Effective immunotherapies for T-cell malignancies are lacking. We devised a novel approach based on chimeric antigen receptor (CAR)-redirected T lymphocytes. We selected CD7 as a target because of its consistent expression in T-cell acute lymphoblastic leukemia (T-ALL), including the most aggressive subtype, early T-cell precursor (ETP)-ALL. In 49 diagnostic T-ALL samples (including 14 ETP-ALL samples), median CD7 expression was >99%; CD7 expression remained high at relapse (n = 14), and during chemotherapy (n = 54). We targeted CD7 with a second-generation CAR (anti-CD7-41BB-CD3ζ), but CAR expression in T lymphocytes caused fratricide due to the presence of CD7 in the T cells themselves. To downregulate CD7 and control fratricide, we applied a new method (protein expression blocker [PEBL]), based on an anti-CD7 single-chain variable fragment coupled with an intracellular retention domain. Transduction of anti-CD7 PEBL resulted in virtually instantaneous abrogation of surface CD7 expression in all transduced T cells; 2.0% ± 1.7% were CD7⁺ vs 98.1% ± 1.5% of mock-transduced T cells (n = 5; P < .0001). PEBL expression did not impair T-cell proliferation, interferon-γ and tumor necrosis factor-α secretion, or cytotoxicity, and eliminated CAR-mediated fratricide. PEBL-CAR T cells were highly cytotoxic against CD7⁺ leukemic cells in vitro and were consistently more potent than CD7⁺ T cells spared by fratricide. They also showed strong anti-leukemic activity in cell line- and patient-derived T-ALL xenografts. The strategy described in this study fits well with existing clinical-grade cell manufacturing processes and can be rapidly implemented for the treatment of patients with high-risk T-cell malignancies.

Graphical abstract

Figure 1.

CD7 expression in T-ALL. (A) The percentage of ALL cells expressing CD7 at diagnosis, relapse, and during chemotherapy (MRD); the number of bone marrow samples studied at each stage is shown. (B) CD7 mean fluorescence intensity (MFI) in T-ALL cells and residual normal T cells from the same samples (n = 19; ****P < .0001 by paired Student t test). (C) CD7 MFI in T-ALL cells at diagnosis or relapse (D/R) and in follow-up bone marrow samples with MRD (n = 18). (D) Flow cytometric contour plots illustrate CD7 expression in T-ALL cells (CD3 negative) and normal T cells (CD3 positive) at diagnosis, MRD, and relapse in 1 representative patient.

Figure 2.

Design, expression, and signaling of the anti-CD7 CAR. (A) Schema of the anti-CD7–41BB-CD3ζ construct. (B) Flow cytometric analysis of Jurkat cells transduced with either GFP alone (Mock) or GFP plus anti-CD7 CAR. Dot plots illustrate GFP fluorescence, and CAR expression after staining with biotin-conjugated goat anti-mouse F(ab′)2 antibody and streptavidin-APC (Jackson ImmunoResearch Laboratories). (C) Western blot analysis of CAR expression in Jurkat cells. Cell lysates of mock- and CAR-transduced Jurkat cells were separated on a 10% polyacrylamide gel under reducing or nonreducing conditions. The blotted membrane was probed with mouse anti-human CD3ζ antibody (8D3; BD Biosciences) and goat anti-mouse immunoglobulin G conjugated to horseradish peroxidase (R&D Systems). Antibody binding was revealed with Clarity Western ECL Substrate (Bio-Rad). (D) Anti-CD7 CAR induces expression of activation markers on ligation. Bars show the mean (± SD) of CD25 and CD69 MFI in CAR- and mock-transduced Jurkat cells after 24 hours with or without CD7⁺ MOLT-4 cells. P values by Student t test are shown for significant differences (*P = .016; ***P < .001). (E) Representative flow cytometric histograms of the experiments shown in panel D.

Figure 3.

Expression of anti-CD7 CAR in human peripheral blood T cells results in fratricide, which is prevented by CD7 downregulation. (A) The percentage of viable T cells recovered 24 hours after electroporation with or without anti-CD7 CAR mRNA (n = 7). Viable cells were counted by flow cytometry. (B) The percentage of viable T cells recovered 24 hours after CAR transduction with a retroviral vector as compared with cells from the same donors transduced with GFP alone (Mock) (n = 10). (C) The percentage of viable CAR- or mock-transduced T cells recovered during the week after transduction. Shown are follow-up results for 5 of the 10 experiments shown in panel B. (D) The percentage of CD107a in T cells after electroporation with or without anti-CD7 CAR mRNA. Mean (± SD) of triplicate measurements are shown. (E) Schematic representation of anti-CD7 PEBL constructs. (F) Representative flow cytometric histograms illustrate CD7 expression in T lymphocytes after retroviral transduction of 3 anti-CD7 PEBLs or mock-transduced GFP alone (Mock). T cells were stained with anti-CD7–PE (M-T701; BD Biosciences). (G) The percentage of CD7 expression in T cells retrovirally transduced with the anti-CD7 PEBL-1 or mock-transduced (n = 5). (H) Flow cytometric dot plots illustrate downregulation of CD7 expression in T cells by PEBL transduction together with expression of anti-CD7–41BB-CD3ζ CAR 12 hours after electroporation with or without CAR mRNA. Cells were stained with biotin-conjugated goat anti-mouse F(ab′)2 antibody and streptavidin-APC (Jackson ImmunoResearch Laboratories). (I) The percentage of viable T cells transduced with anti-CD7 PEBL recovered 24 hours after electroporation of anti-CD7 CAR mRNA as compared with cells electroporated with the anti-CD7 CAR mRNA, but transduced with a vector without anti-CD7 PEBL (n = 6). The number of viable cells was measured by flow cytometry. **P < .01; ***P < .001.

Figure 4.

CD7 downregulation by PEBL does not alter T-cell phenotype, proliferation, and functionality. (A) The percentage of CD4 and CD8 cells 7 to 14 days after retroviral transduction with either anti-CD7 PEBL or GFP alone (Mock). Each symbol corresponds to a different T-cell donor. (B) The growth rate of PEBL- and mock-transduced T cells (from 3 donors) maintained with 200 IU/mL IL-2 for 14 days. Symbols represent the mean (± SD) of triplicate measurements. (C) PEBL- and mock-transduced T cells were electroporated with either anti-CD19–41BB-CD3ζ CAR mRNA or no mRNA. Flow cytometric dot plots illustrate GFP and CAR expression 12 hours after electroporation. CAR was detected with biotin-conjugated goat anti-mouse F(ab′)2 antibody and streptavidin-APC (Jackson ImmunoResearch Laboratories). (D) Cytotoxicity of PEBL- or mock-transduced T cells, electroporated with or without anti-CD19 CAR mRNA, against CD19⁺ ALL cells (OP-1). Bars show the mean (± SD) of 4-hour cytotoxicity at a 1:1 E:T ratio. (E) CD107a expression in T cells from experiments identical to those described in panel D. (F) IFN-γ production in PEBL- or mock-transduced T cells, electroporated with or without anti-CD19 CAR mRNA, and cocultured with OP-1 for 6 hours at an E:T ratio of 1:1. Bars represent the mean (± SD) of triplicate experiments. ***P < .001; ****P < .0001.

Figure 5.

T cells with downregulated CD7 by PEBL acquire powerful cytotoxicity against CD7⁺leukemic cells after expression of anti-CD7 CAR. (A) Cytotoxicity of anti-CD7 PEBL-transduced T cells electroporated with or without anti-CD7 CAR mRNA against CD7⁺ cell lines. Shown are data for 4-hour assays at a 1:1 E:T ratio. Symbols indicate the mean of 3 measurements each with T cells from 4 donors for MOLT-4, CCRF-CEM, and Jurkat, and 5 donors for Loucy and KG1a (P < .001 for each comparison). (B) Cytotoxicity of anti-CD7 PEBL-transduced T cells electroporated with or without anti-CD7 CAR mRNA against primary leukemic cells from patients with T-ALL. Shown are data for 4-hour assays at the indicated E:T ratio. Symbols refer to the mean (± SD) of 3 measurements. (C) Overall specific cytotoxicity of T cells transduced with either anti-CD7 PEBL or GFP alone (Mock) after electroporation with anti-CD7 CAR mRNA against the 5 CD7⁺ cell lines. T cells from 3 donors were tested at a 1:1 E:T ratio in 4-hour assays. Each symbol represents the specific percentage of cytotoxicity against the CD7⁺ cell line after subtraction of the percentage of cytotoxicity obtained with the same T cells electroporated without mRNA. Horizontal bars indicate the median for each group. (D) Anti-CD7 PEBL- or mock-transduced T cells from 3 donors were electroporated with or without anti-CD7 CAR mRNA. Cytotoxicity against MOLT-4 was tested at a 1:1 E:T ratio in 4-hour assays. Shown is the MFI of anti-CD107a–PE (H4A3; BD Biosciences). Bars represent the mean (± SD) of triplicate experiments. (E) Anti-CD7 PEBL-transduced T cells were retrovirally transduced with either anti-CD7 CAR or mock-transduced and tested against primary leukemic cells from patients with T-ALL. Each symbol represents the mean (± SD) of triplicate experiments. (F) Mock- or PEBL-transduced T cells, sequentially transduced with or without anti-CD7 CAR, were cultured alone or in the presence of Streck-treated MOLT-4 cells, added weekly and 120 IU/mL IL-2. Symbols indicate the mean (± SD) percentage of cell recovery relative to the number of input cells in triplicate cultures. **P < .01; ***P < .001; ****P < .0001.

Figure 6.

PEBL-transduced T cells expressing an anti-CD7–41BB-CD3ζ CAR exert antitumor activity in xenografts. NOD-SCID-IL2RGnull mice were infused IV with 1 × 10⁶ CCRF-CEM cells labeled with luciferase. A total of 2 × 10⁷ PEBL-CAR T cells were administered IV on day 7 (A) or on day 3 and day 7 (B) after leukemic cell infusion to 3 and 5 mice, respectively. The remaining mice received either mock-transduced T cells or RPMI 1640 instead of cells (Control). All mice received 20 000 IU IL-2 once every 2 days IP. Shown is in vivo imaging of leukemia cell growth after d-luciferin IP injection. Ventral images of mice on day 3 in panel B are shown with enhanced sensitivity to demonstrate CCRF-CEM engraftment in all mice. The complete set of luminescence images is shown in supplemental Figure 7. (C) Leukemia cell growth in mice shown in panels A and B is expressed as photons per second. Each symbol corresponds to bioluminescence measurements in each mouse, normalized to the average of ventral plus dorsal signals in all mice before CAR T-cell infusion. (D) Kaplan-Meier curves show overall survival of mice in the different groups (8 in each group). Mice were euthanized when the total bioluminescence signal reached 1 × 10¹⁰ photons per second. The P values were calculated by log-rank test. N.S., not significant.

Figure 7.

PEBL-CAR T-cell activity against ETP-ALL in a patient-derived xenograft (PDX) model. (A) Primary ETP-ALL cells, previously propagated in NOD-SCID-IL2RGnull mice, were infused IV in 10 NOD-SCID-IL2RGnull mice at 2 × 10⁶ cells per mouse. Five mice (Control) were left untreated. The remaining 5 mice received a single IV infusion of PEBL-CAR T cells (2 × 10⁷ in PEBL-CAR#1 mouse, 2 × 10⁶ in the remaining 4 mice) at the indicated time point (blue arrow), as well as 20 000 IU IL-2 IP every 2 days; IL-2 was also administered to 2 of the 5 control mice. Red symbols (left y-axes) indicate the number of ETP-ALL cells per milliliter counted in peripheral blood. Blue symbols (right y-axes) show the numbers of PEBL-CAR T cells. The mice were euthanized when the percentage of ETP-ALL cells among blood mononucleated cells reached ≥80%. (B) The percentage of ETP-ALL (denominator, total human plus mouse CD45⁺ cells) in various organs of the 5 untreated mice. (C) Blood smears of treated (PEBL-CAR#1) and untreated ETP-ALL 7 days after infusion of T cells; smudge cells were prominent in blood after infusion of PEBL-CAR T cells. Original magnification ×40, Wright stain. (D) Flow cytometric dot plots show the presence of CD7⁺ CD3^– ETP-ALL cells in the tissues of an untreated control mouse with ETP-ALL and of CD7^– CD3⁺ PEBL-CAR T cells in the PEBL-CAR#1 mouse treated with PEBL-CAR T cells. No ETP-ALL (<0.01%) was detected in the treated mouse. The events shown were normalized to the events acquired for the corresponding plots shown in the control mouse. (E) Spleens of treated (PEBL-CAR#1) and untreated mice. BM, bone marrow.