Anti-CD37 chimeric antigen receptor T cells are active against B- and T-cell lymphomas

Abstract

Chimeric antigen receptor (CAR) T cells have emerged as a novel form of treatment of patients with B-cell malignancies. In particular, anti-CD19 CAR T-cell therapy has effected impressive clinical responses in B-cell acute lymphoblastic leukemia and diffuse large B-cell lymphoma. However, not all patients respond, and relapse with antigen loss has been observed in all patient subsets. Here, we report on the design and optimization of a novel CAR directed to the surface antigen CD37, which is expressed in B-cell non-Hodgkin lymphomas, in chronic lymphocytic leukemia, and in some cases of cutaneous and peripheral T-cell lymphomas. We found that CAR-37 T cells demonstrated antigen-specific activation, cytokine production, and cytotoxic activity in models of B- and T-cell lymphomas in vitro and in vivo, including patient-derived xenografts. Taken together, these results are the first showing that T cells expressing anti-CD37 CAR have substantial activity against 2 different lymphoid lineages, without evidence of significant T-cell fratricide. Furthermore, anti-CD37 CARs were readily combined with anti-CD19 CARs to generate dual-specific CAR T cells capable of recognizing CD19 and CD37 alone or in combination. Our findings indicate that CD37-CAR T cells represent a novel therapeutic agent for the treatment of patients with CD37-expressing lymphoid malignancies.

Conflict of interest statement

Conflict-of-interest disclosure: I.S. and M.V.M. are listed as inventors on patents related to this work and held by the Massachusetts General Hospital and Partners Health Care. The remaining authors declare no competing financial interests.

© 2018 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

CD37 is highly expressed on lymphoma cell lines and patient-derived samples. (A) Fluorescence-activated cell sorter (FACS) plots of tumor cell lines stained with CD37 and CD19 antibodies. NALM6, acute lymphoblastic leukemia cell line; K562 expressing CD19 and CD37, positive control; JEKO-1, MCL cell line; RAJI, Burkitt lymphoma cell line. (B) FACS plots of samples derived from 3 patients with MCL. (C) MFI of CD19 and CD37 on MCL patient’s cells. Each dot represents a separate xenograft sample (n = 3; medians shown). (D) CD19 and CD37 expression on PBMC from patients with CLL gated on CD3− lymphocytes (n = 20; mean ± SD shown; ****P < .0001 by Student t test). (E) Distribution of CD19 and CD37 antigens on CLL PBMC cells. Antibodies bound per cell gated on CD3− cells from each patient sample is shown. (F) Median of CD19 and CD37 antigen density is shown. CD19 range, 24 396 to 70 952; mean, 43 238; CD37 range, 8992 to 46 550; mean, 23 989. (G) CD37 immunohistochemistry in primary ALK-negative (left) and ALK-positive (right) ALCL specimens from the tissue microarray. Original magnification ×100.

Figure 2.

In vitro CAR-37 generation and expansion. (A) Two anti-CD37 second-generation chimeric antigen receptors were constructed with different orientations of a humanized murine antibody-derived single-chain variable fragment: the light-to-heavy orientation (CAR-37 L-H, top) and the heavy-to-light (CAR-37 H-L, bottom). (B) Representative flow plots of primary human T cells transduction efficiency after 10 days of activation with CD3/CD28 beads. (C) Expanded T cells from 3 healthy donors included variable CAR-37 expression with a mean of 38% (L-H) and 75% (H-L). (D) Ex vivo expansion of CD3/CD28 bead-activated and target-stimulated T-cells using static culture conditions in 3 healthy donors for 38 days. Each arrow represents antigen stimulation with K562 cells transduced to express CD37 and CD19. (E) Activation of Jurkat reporter (NFAT-Luc) T cells transduced with different CAR constructs and cocultured with tumor cells. Luciferase activity was measured after 16 hours. (CD3-CD28 beads: positive control). (F) Number of carboxyfluorescein succinimidyl ester–labeled, unstimulated target T cells measured by flow cytometry after 24 hours of coculture at indicated E:T ratio with CAR-37H-L, CAR-19, or UTD T cells. (G) Number of carboxyfluorescein succinimidyl ester–labeled T cells stimulated with PMA/ionomycin for 6 hours, then measured by flow cytometry after 24 hours of coculture at indicated E:T ratio with CAR-37H-L, CAR-19, or UTD T cells. (H) Number of Jeko-1 CBG−GFP cells measured by flow cytometry after 24 hours of coculture at indicated E:T ratio with CAR-37H-L, CAR-19 or untransduced T cells. Bars indicate mean ± standard error of the mean (SEM) count of triplicates from 1 normal donor, representative of 3 normal donors. (I) CD107a and IFN-γ production relative to media by CAR-37H-L, CAR-19 T cells incubated with primary immune cells for 6 hours at 1:1 E:T ratio was analyzed by flow cytometry. Bars show mean ± SEM percentage of the 3 normal donors analyzed.

Figure 3.

CAR-37 T cells exhibit robust in vitro effector functions in response to CD37 positive tumor cells. (A) Cytotoxic capacity of CAR-37 T cells was measured after overnight coculture with targets. CAR T cells were cocultured at indicated E:T ratios with indicated tumor cell lines. Increasing concentration of CAR-37 and CAR-19 T cells led to specific killing, whereas no killing was observed in the control group (UTD). The cytotoxicity assay is representative of 3 independent experiments conducted with different healthy donors. Cytokine production by CAR-37H-L, CAR-19, or UTD T cells incubated with primary CLL (B) or MCL PDX (C) tumor samples. CAR T cells were incubated with target cells for 24 hours at a 1:1 E:T ratio, and culture supernatants were analyzed by Luminex assay. Data are plotted as mean ± SEM for 3 donors.

Figure 4.

In vivo CAR-37 mediated tumor clearance of a MCL model. (A) Experiment schematic: NSG mice were injected IV with 1 × 106 JEKO-1(CBG−GFP) cells and monitored by BLI for tumor burden at different points. At day 0, mice were randomly assigned on the basis of tumor burden (BLI) to receive 2 × 106 control T cells (UTD), CAR-37, or CAR-19. (B) Representative bioluminescent images of JEKO-1 growth over time. (C) Average flux (photons/s) of whole mice in the 3 groups at different points. Graph is representative of 2 experiments with 5 mice per group, conducted with CAR T cells obtained from 2 different healthy donors. Mean ± SD shown. ***P < .001 by 2-way analysis of variance. (D) Absolute numbers of CAR T cells were monitored by bleeding and flow cytometric detection. Absolute CAR T-cell counts in peripheral blood at day 14 after CAR T injection are shown (Student t test, *P < .05).

Figure 5.

CAR-37 mediated tumor clearance of MCL patient-derived xenograft. (A) Protocol schema: NSG mice were injected intravenously with 1 × 106 MCL patient-derived cells and monitored for tumor burden by bioluminescent imaging (BLI) over time. At day 0, mice were randomly assigned on the basis of tumor burden to receive 3 × 106 control T cells (UTD), CAR-37, or CAR-19. (B) Representative BLI of MCL xenografts over time. (C) Average flux (photons/s) of whole mice in the 3 groups at different points. Graph is representative 2 simultaneous experiments of 5 mice per group, conducted with CAR T cells obtained from 2 different healthy donors, and pooled data. Mean ± SD shown (Student t test, *P < .05). (D) Absolute numbers of CAR T cells were monitored in peripheral blood using flow cytometry. Absolute counts of CAR T cells are plotted at day 14.

Figure 6.

CAR-37 in vitro activity against T-cell lymphoma and leukemia. (A) CD37 expression on PTCL tumor cell lines. (B) Representative FACS plots from patient-derived samples. (C) CD69 expression and (D) CD107a degranulation of CAR T cells, as evaluated by flow cytometry after 6 hours of coculture with indicated tumor cells at 1:1 E:T ratio. Degranulation is relative to PMA positive control; representative normal donor is shown. Cytotoxic capacity of CAR-37 T cells was measured after overnight coculture with (E) Hut78 and (F) FEPD target cells at a different E:T ratio. The cytotoxicity assay is representative of 3 independent experiments conducted with different healthy donors. Mean ± SEM shown.

Figure 7.

Bispecific CD19 CD37 CAR T cells. (A) Two bispecific second-generation CARs were constructed with a different order of the single-chain variable fragments: CAR-19-37 (top) and CAR-37-19 (bottom). (B) Activation of Jurkat reporter (NFAT-Luc) T cells transduced with different CAR constructs and cocultured with tumor cells. Luciferase activity was measured after 16 hours. CD3-CD28 beads: positive control. (C) Ex vivo expansion of CD3/CD28 bead-activated and target-stimulated T cells in 2 healthy donors for 30 days. (D) Cytotoxic capacity of bispecific CAR T cells was measured after overnight coculture with K562 targets transduced with CD37, CD19, or both at indicated E:T ratios. The cytotoxicity assay is representative of 2 independent experiments conducted with different healthy donors. (E) NSG mice were injected intravenously with 1 × 106 JEKO-1 (CBG−GFP) cells and monitored by BLI for tumor burden over time. At day 0, mice were randomly assigned on the basis of tumor burden (BLI) to receive 2 × 106 control T cells (UTD), CAR- 37, CAR-19, CAR-19-37, or CAR-37-19. All CAR T-cell groups were normalized to have the same % CAR+ cells and untransduced cells. Average flux (photons/s) of whole mice at different points is shown. Graph shows 1 experiment with 6 mice per group. (F) CAR T cells were enumerated in peripheral blood by flow cytometry at the indicated points. Absolute counts of CAR T cells are shown as mean ± SEM.