T cells redirected against CD70 for the immunotherapy of CD70-positive malignancies

Abstract

T-cell therapy with genetically modified T cells targeting CD19 or CD20 holds promise for the immunotherapy of hematologic malignancies. These targets, however, are only present on B cell-derived malignancies, and because they are broadly expressed in the hematopoietic system, their targeting may have unwanted consequences. To expand T-cell therapies to hematologic malignancies that are not B cell-derived, we determined whether T cells can be redirected to CD70, an antigen expressed by limited subsets of normal lymphocytes and dendritic cells, but aberrantly expressed by a broad range of hematologic malignancies and some solid tumors. To generate CD70-specific T cells, we constructed a chimeric antigen receptor (CAR) consisting of the CD70 receptor (CD27) fused to the CD3-ζ chain. Stimulation of T cells expressing CD70-specific CARs resulted in CD27 costimulation and recognition of CD70-positive tumor cell lines and primary tumor cells, as shown by IFN-γ and IL-2 secretion and by tumor cell killing. Adoptively transferred CD70-specific T cells induced sustained regression of established murine xenografts. Therefore, CD70-specific T cells may be a promising immunotherapeutic approach for CD70-positive malignancies.

Figures

Figure 1

CD70-CAR generation, cell-surface expression, and transduction of human T cells. (A) CD70-CAR was generated by fusing full-length CD27 to the signaling domain of the CD3-ζ chain, an IRES sequence, and tCD19 was included for the detection of genetically modified T cells. (B) 293T cells transfected with CD70-CAR constructs express both CD27 and the marker gene tCD19. (C) CD70-CAR expression on transduced human T cells was 45% ± 6%, as determined by staining tCD19. (D) Both CD4 and CD8 T cells were genetically modified.

Figure 2

CD70 is overexpressed on several tumor cell lines but not on normal lymphocytes. Less than 5% of B and T lymphocytes from the peripheral blood of healthy donors express CD70. K562 and K562.70 served as negative and positive controls, respectively. CD70 overexpression was observed on non-Hodgkin (Daudi, Raji, SNK6, and SNT16), Hodgkin (L1236), ALL (CCL-120), and multiple myeloma (U266) cells.

Figure 3

CD70-specific T cells release IFN-γ and IL-2 and proliferate in response to CD70-positive target cells. (A) T cells from 3 donors were transduced with CD70-CAR (black) or nontransduced (gray) and cocultured with K562.70 and K562 and various CD70-expressing tumor cell lines for 48 hours before performing IFN-γ ELISA. Black and gray rectangles represent mean IFN-γ release of CD70-CAR transduced and nontransduced T cells, respectively. CD70-CAR T cells were specific for CD70, because significantly more (P < .03) IFN-γ was released in the presence of K562.70 compared with K562 cells. CD70-CAR T cells also released significantly more (P < .0001) IFN-γ than nontransduced T cells when cocultured with CD70-expressing tumor cell lines. (B) Same coculture experiments but assayed for the presence of IL-2. CD70-CAR T cells released significantly more (P < .0001) IL-2 than nontransduced T cells in the presence of CD70-expressing tumors. (C) T cells were labeled with CFSE and cocultured for 5 days with K562, K562.70, SNT16, Raji, or Daudi in the absence of exogenous IL-2, and CFSE dilution was analyzed by flow cytometry. CD70-CAR T cells proliferated when cocultured with the CD70-overexpressing targets K562.70, SNT16, and Raji, but not CD70-dim Daudi cells or CD70-negative K562 cells.

Figure 4

CD70-specific T cells kill CD70-positive tumor cell lines. (A) CD70-CAR T cells (solid lines) killed K562.70 cells but not parental K562 cells. Nontransduced control T cells (dashed lines) did not kill either target. (B) CD70-CAR T cells (solid lines) killed CD70-positive Daudi, U266, SNK6, and SNT16 tumor cell lines; control T cells (dashed lines) did not. (C) CD70-specific T cells or nontransduced T cells were labeled with CFSE and cocultured with SNT16 cells at a ratio of 2:1. CD70-specific T cells proliferated and killed SNT16 cells, as shown by CFSE dilution of CD3+ cells and by the lack of CD3/CFSE-negative cells in the culture compared with nontransduced T cells. (D) In all coculture experiments, only CD70-specific T cells eliminated the CD3/CFSE–negative CD70+ tumor cells Daudi, U266, SNK6, and SNT16.

Figure 5

CD27 costimulation enhances T-cell viability. (A) In coimmunoprecipitation experiments, only full-length CD27-ζ associated with TRAF2. (B) T cells expressing CD70-CAR or ΔCD70-CAR showed equivalent killing of CD70+ LCL and U266 cells, but did not kill CD70− K562 cells in 51Cr-release assays. (C) Microscopic evaluation (10×) of T cells expressing CD70-CAR or ΔCD70-CAR activated with autologous fibroblasts genetically modified to express CD70 revealed larger “T-cell clumps” of T cells expressing CD70-CAR; however, CFSE dilution analysis showed no significant differences in proliferation between groups. (D) The viability of ΔCD70-CAR T cells was 35% ± 16% that of T cells expressing CD70-CAR (n = 5). (E) Intracellular staining for Bcl-xl was performed on T cells 3 days after stimulation with CD70-transgenic autologous fibroblasts. Bcl-xl expression was consistently increased in CD70-CAR T cells compared with ΔCD70-CAR T cells (n = 3). One representative FACS analysis is shown.

Figure 6

CD70-specific T cells recognize and kill primary CD70-positive lymphomas. (A) CD70-overexpressing tumor cells from 3 patients with B-cell lymphoma and 1 patient with T-cell acute lymphoblastic leukemia were cocultured with CD70-specific or nontransduced T cells from healthy donors for 48 hours before performing IFN-γ ELISA. In all cases, CD70-specific T cells released IFN-γ in the presence of patient tumor cells, whereas nontransduced cells released little or no IFN-γ. (B-C) Coculture assays were performed with primary tumor cells and CFSE-labeled T cells to distinguish effector and target cells by FACS analysis. Only CD70-specific (CD3/CFSE–positive) T cells were able to eradicate patient tumor cells (P = .036).

Figure 7

CD70-specific T cells exhibit in vivo antitumor activity in an IP and systemic xenograft model of lymphoma. (A-B) Daudi cells (5 × 105) expressing the eGFP-FFLuc gene were injected IP into SCID mice, and tumor growth was measured as increasing light signal (p/s/cm2/sr). On days 10, 11, and 17, mice were injected with 1 × 107 CD70-specific or nontransduced T cells. Tumors treated with CD70-specific T cells regressed, whereas tumors treated with nontransduced T cells did not (P = .002) 7 days after treatment. Panel A shows images of representative animals. Panel B shows quantitative bioluminescence imaging. In panels C and D, Raji cells (2 × 105) were injected intravenously into SCID mice. On days 4, 5, and 11, mice were injected with 1 × 107 CD70-specific or nontransduced T cells. (C) Systemic tumors were enumerated using bioluminescence imaging. At weeks 3 and 4 after tumor cell injection, there was a significantly higher tumor burden in mice receiving nontransduced T cells than CD70-specific T cells (week 3, P = .012; week 4 [n = 12], P = .010). (D) Mice treated with CD70-specific T cells displayed a significant survival advantage over those receiving nontransduced T cells (P < .05).