Preclinical development of CD37CAR T-cell therapy for treatment of B-cell lymphoma

Abstract

T cells modified to express chimeric antigen receptor (CAR) targeting CD19 (CD19CAR) have produced remarkable clinical responses in patients with relapsed/refractory B-cell acute lymphoblastic leukemia. CD19CAR T-cell therapy has also demonstrated prominent effects in B-cell non-Hodgkin lymphoma (B-NHL) patients. However, a subset of patients who relapse after CD19CAR T-cell therapy have outgrowth of CD19- tumor cells. Hence, development of alternative CARs targeting other B-cell markers represents an unmet medical need for B-cell acute lymphoblastic leukemia and B-NHL. Here, we confirmed previous data by showing that, overall, B-NHL has high expression of CD37. A second-generation CD37CAR was designed, and its efficacy in T cells was compared with that of CD19CAR. In vitro assessment of cytotoxicity and T-cell function upon coculture of the CAR T cells with different target B-cell lymphoma cell lines demonstrated comparable efficacy between the 2 CARs. In an aggressive B-cell lymphoma xenograft model, CD37CAR T cells were as potent as CD19CAR T cells in controlling tumor growth. In a second xenograft model, using U2932 lymphoma cells containing a CD19- subpopulation, CD37CAR T cells efficiently controlled tumor growth and prolonged survival, whereas CD19CAR T cells had limited effect. We further show that, unlike CD19CAR, CD37CAR was not sensitive to antigen masking. Finally, CD37CAR reactivity was restricted to B-lineage cells. Collectively, our results demonstrated that CD37CAR T cells also can effectively eradicate B-cell lymphoma tumors when CD19 antigen expression is lost and support further clinical testing for patients with relapsed/refractory B-NHL.

Conflict of interest statement

Conflict-of-interest disclosure: E.B.S. owns stock in Nordic Nanovector. H.H., G.K., E.B.S., J.H.M., E.M.I., and S.W. have applied for a patent related to the results. The remaining authors declare no competing financial interests.

© 2019 by The American Society of Hematology.

Figures

Graphical abstract

Figure 1.

Surface expression of CD19, CD20, CD22, and CD37 in multiple subtypes of B-NHL. Flow cytometry was used to analyze single-cell suspensions from FL, DLBCL, MCL, CLL, and MZL, as well as tonsils from healthy donors. Tumor cells were identified by gating on CD3−CD20+ B cells, followed by gating on tumor-restricted Ig light chain (Ig light chain–negative B cells in some cases). (A) Expression of CD37, CD19, CD20, and CD22 in tumor cells from NHL patients and in B cells from healthy donor tonsils. Relative protein expression was calculated through Cytobank using arcsinh transformation of median fluorescence intensity (MFI) of the cell population of interest as follows: fold change = arcsinh (MFI of protein in B cells/scale argument) − arcsinh (MFI of protein in T cells/scale argument). FL: n = 5 or n = 18 (CD19 and CD20), DLBCL: n = 18, MCL: n = 10, CLL/MZL: n = 8, tonsils: n = 8 or n = 15 (CD19 and CD20). (B) Association of CD37 and CD19 tumor cell expression levels in B-NHL. (C) Examples of CD19 and CD37 expression in tumor cells from 3 DLBCL patients. (D) Percent expression of CD19 and CD37 in individual samples from NHL patients. *P < .05, **P < .01, ****P < .0001, Mann-Whitney U nonparametric test.

Figure 2.

Expression of CD19CAR and CD37CAR in human T cells. (A) A schematic representation of the CAR constructs. (B) mRNAs generated from CD19CAR and CD37CAR constructs (A) were electroporated into T cells. CD19CAR and CD37CAR expression was detected by Protein L staining and anti-mouse Fab antibody, respectively, 18 hours after electroporation.

Figure 3.

Comparison of CD19CAR T cells and CD37CAR T cells for antilymphoma activity in vitro. BLI-based measurement of cytotoxicity mediated by CD19CAR T cells or CD37CAR T cells when cocultured at an E:T ratio of 25:1 with B-cell lymphoma cell lines. Lysis was analyzed after 2, 4, 6, 10, or 22 hours of coculture. Data represent mean ± standard deviation of quadruplicates. Data from 1 of 2 experiments are shown. *P < .05, **P < .01, ****P < .0001, Student t test with Bonferroni correction. n.s., not significant.

Figure 4.

CD37CAR T cells exhibit enhanced effector functions, including specific killing of CD19−U2932 cells. (A) A representative analysis of CD19 and CD37 expression in Jurkat, BL-41, and U2932 cells by flow cytometry. (B) Specific killing mediated by CD19CAR or CD37CAR T cells against BL-41 and U2932 cells after 7 hours of coculture at the indicated E:T ratios. Data represent mean ± standard deviation of triplicates. Data from 1 of 2 experiments are shown. (C) Detection of intracellular cytokine production in CD19CAR or CD37CAR T cells after coincubation with target cells BL-41 or U2932 for 24 hours at an E:T ratio of 1:2. Data from 1 of 2 experiments are shown. Data represent mean ± standard deviation of quadruplicates. (D) Detection of degranulation in CD19CAR or CD37CAR T cells after coincubation with target cells BL-41 or U2932 for 6 hours at an E:T ratio of 1:2. Data from 1 of 2 experiments are shown. Data represent mean ± standard deviation of quadruplicates. (E) Flow cytometric analysis of CD19 and CD37 on the surface of GFP+ U2932 cells 5 hours after coculture with mock T cells, CD19CAR T cells, or CD37CAR T cells at an E:T ratio of 10:1. *P < .05, **P < .01, ***P < .001, ****P < .0001, Student t test with Bonferroni correction.

Figure 5.

CD37CAR T cells and CD19CAR T cells have comparable antitumor efficacy in vivo. NSG mice were engrafted with GFP/Luc+ BL-41 tumors or GFP/Luc+ U2932 tumors subcutaneously; 4 or 12 days, respectively, after tumor inoculation, mice were randomized and received intratumoral injections of mock T cells, CD19CAR T cells, or CD37CAR T cells (n = 5 for each group) every 3 days for 2 weeks. Tumor size was measured using a caliper. These experiments were reproduced twice. (A) BL-41 tumor growth curves after mock T cell, CD19CAR T cell, or CD37CAR T cell transfer. (B) Kaplan-Meier survival curves of mice shown in panel A. (C) Proportion of tumor-infiltrating lymphocytes, as detected by CD3 staining of tumor tissue single-cell suspension, using flow cytometry. The tumor samples were obtained at day 27. Data represent mean ± standard deviation of duplicates. (D) Percentage of tumor-infiltrating lymphocytes expressing PD-1 exhaustion marker. Data represent mean ± standard deviation of triplicates. (E) U2932 tumor growth curves after mock T cell, CD19CAR T cell, or CD37CAR T cell transfer. (F) Kaplan-Meier survival curves of mice shown in panel E. These experiments were reproduced twice. Survival curves were analyzed with a Mantel-Cox (log-rank) test. Data represent mean ± standard deviation. *P < .05, **P < .01.

Figure 6.

CD37CAR expression in BL-41 does not mask CD37. (A) Flow cytometric analysis of BL-41 cells transduced with CD19CAR or CD37CAR. Expression of CD19 and CD37 was detected by commercial antibody clones or by the corresponding antibody clones from which the CAR constructs were derived. The expression of CARs was also validated by anti-mouse Fab and anti–c-Myc antibodies. Data from 1 of 2 experiments are shown. (B) BLI-based measurement of cytotoxicity mediated by mock T cells, CD19CAR T cells, or CD37CAR T cells when cocultured at an E:T ratio of 10:1 with target cells BL-41, BL-41 CD19CAR, or BL-41 CD37CAR. Lysis was analyzed after 1, 2, 3, 4.5, 7, and 9 hours of coculture. Data represent mean ± standard deviation of quadruplicates. Representative data from 1 of 3 experiments are shown. ****P < .0001, Student t test with Bonferroni correction between effector condition and its respective negative control.

Figure 7.

CD37CAR T cells are B-cell lineage specific. (A) Primary CD19+ B cells, CD3+ T cells, CD14+ monocytes, and CD56+ NK cells were isolated from healthy donor PBMCs and cocultured with effector mock T cells, CD19CAR T cells, or CD37CAR T cells generated from the same donors at E:T ratios of 1:2 and 1:5. Activated T cells, used as target, were generated via CD3/CD28 stimulation. Six hours after coculture, CD107a expression on effector T cells was assessed by flow cytometry. Data represent mean ± standard deviation of duplicates. (B) Bone marrow progenitor cells were cocultured with CD19CAR or CD37CAR autologous T cells from a healthy donor for 6 hours at an E:T ratio of 5:1. The cells were then plated in semisolid methylcellulose progenitor culture for 14 days and scored for the presence of red (CFU erythroid), white (CFU with granulocytes, macrophages, or cells of both lineages), and total (CFU with granulocyte, erythroid, macrophage, megakaryocyte) colonies. Data represent mean ± standard deviation of hexaplicates. Representative data from 1 of 3 experiments are shown, P > .5 for all data. Cytokine and chemokine secretion was measured by Bio-Plex assay of supernatants from T cells from 3 healthy donors, transfected with CD19CAR or CD37CAR and activated by coculture with BL-41 cells (C) or U2932 cells (D) for 24 hours at an E:T ratio of 1:2. Data represent mean ± standard deviation of triplicates. Data from 1 of 2 experiments are shown.