T lymphocytes coexpressing CCR4 and a chimeric antigen receptor targeting CD30 have improved homing and antitumor activity in a Hodgkin tumor model

Abstract

For the adoptive transfer of tumor-directed T lymphocytes to prove effective, there will probably need to be a match between the chemokines the tumor produces and the chemokine receptors the effector T cells express. The Reed-Stemberg cells of Hodgkin lymphoma (HL) predominantly produce thymus- and activation-regulated chemokine/CC chemokine ligand 17 (TARC/CCL17) and macrophage-derived chemokine (MDC/CCL22), which preferentially attract type 2 T helper (Th2) cells and regulatory T cells (Tregs) that express the TARC/MDC-specific chemokine receptor CCR4, thus generating an immunosuppressed tumor environment. By contrast, effector CD8(+) T cells lack CCR4, are nonresponsive to these chemokines and are rarely detected at the tumor site. We now show that forced expression of CCR4 by effector T cells enhances their migration to HL cells. Furthermore, T lymphocytes expressing both CCR4 and a chimeric antigen receptor directed to the HL associated antigen CD30 sustain their cytotoxic function and cytokine secretion in vitro, and produce enhanced tumor control when infused intravenously in mice engrafted with human HL. This approach may be of value in patients affected by HL.

Figure 1

Improved migration of activated T lymphocytes genetically modified to overexpress CCR4. (A) The expression of CCR4 in control (CTR) T cells and in T lymphocytes transduced with a retroviral vector encoding CCR4. Surface expression of CCR4 was evaluated by FACS analysis of CD3⁺, CD4⁺, and CD8⁺ T lymphocytes. represent the mean ± SD of control T cells; ■, mean ± SD of CCR4⁺ T cells. The data summarize the results of T-cell lines generated from 7 healthy donors. (B) A representative phenotypic analysis. (C) The migration of CTR () and CCR4⁺ (■) T cells toward TARC gradients, using a transwell migration assay. T-cell migration was evaluated using culture supernatants collected from 2 HL-derived cell lines (HDLM-2 and L428) that physiologically produce high amounts of TARC, and against the Karpas-299 cell line genetically modified to produce TARC (K/TARC). K/wt was used as a control. The panel indicates that migration toward TARC is significantly improved if T cells are genetically modified to overexpress CCR4. The data are the mean ± SD for T-cell lines generated from 7 healthy donors. (D) The improved migration of CCR4⁺ T cells (■) is TARC mediated as it is inhibited by addition of anti-TARC antibodies but not by the addition of an isotype control.

Figure 2

Improved migration of CD8⁺ T lymphocytes genetically modified to overexpress CCR4. (A) The expression of CCR4 by CTR and transduced CD8⁺ T cells, using FACS analysis. The data summarize the results of T-cell lines generated from 4 healthy persons. (B) A representative phenotypic analysis. Dotted lines indicate isotype control. (C) The migration of CTR () and CCR4⁺ (■) CD8⁺ T cells toward TARC gradients, using a transwell migration assay. The panel indicates that migration toward TARC is significantly increased when CD8⁺ T cells overexpress CCR4. The data are the mean ± SD for CD8⁺ T-cell lines generated from 4 healthy donors.

Figure 3

Immunophenotype and function of T cells overexpressiong CCR4 are retained. (A) The phenotypic composition of NT () and CCR4⁺ (■) T cells. The data are mean ± SD of 5 healthy donors. T-cell markers are shown on the x-axis. No significant differences were observed if cells overexpressed CCR4. (B) Expression of naive, central memory, and effector memory surface markers on CTR () and CCR4⁺ (■) T cells is not significantly different. The data are mean ± SD of 4 donors. (C) The production of Th1 (IFN-γ and TNF-α) and Th2 (IL-10 and IL-4) cytokines by CTR () and CCR4⁺ (■) T cells 24 hours after stimulation with OKT3. No significant differences in cytokine production were detected, suggesting that the transgenic expression of CCR4 does not induce the acquisition of a Th2 profile. (D) CCR4⁺ T cells do not acquire inhibitory function. The inhibitory activity of T cells was evaluated using a CFSE-based suppression assay in which PBMCs labeled with CFSE are stimulated with irradiated allogeneic PBMCs and OKT3 in the absence (top left graph) or in the presence of freshly isolated CD4⁺CD25^bright cells (top right graph), CTR (bottom left graph), or CCR4⁺ T cells (bottom right graph). The panel indicates a significant number of divisions (CFSE partitioning) of T cells in the absence or presence of CTR or CCR4⁺ T cells (evident in the left quadrants). In contrast, the divisions are significantly reduced in the presence of Treg cells. Shown is one of 3 donors studied, illustrative of results from all.

Figure 4

Expression and improved migration of CCR4⁺ T cells established from persons affected by HL. (A) The expression of CCR4 by CTR () and transduced (■) T cells established from 4 persons with HL. The data are mean ± SD. (B) A representative phenotypic analysis. (C) The migration of CTR () and CCR4⁺ (■) T cells toward TARC gradients, using a transwell migration assay. T-cell migration was evaluated using culture supernatants collected from HDLM-2 and L428, which physiologically produce high amounts of TARC, and against Karpas genetically modified to produce TARC (K/TARC). K/wt represents migration toward Karpas wild-type, which produces a negligible amount of TARC, as a negative control. Migration is significantly increased for CCR4⁺ T cells compared with CTR T cells. In addition, the figure shows that the improved migration of CCR4⁺ T cells (■) is TARC mediated as it is inhibited by addition of anti-TARC antibodies but not by addition of an isotype control.

Figure 5

Improved in vivo migration of CCR4⁺ T cells. CTR and CCR4⁺ T cells were transduced to express eGFP-FFLuc to monitor their migration in vivo in SCID mice, using the IVIS imager system. (A) The expression of eGFP-FFLuc on CTR (top plot) and CCR4⁺ (bottom plot) T cells evaluated by GFP. (B) The bioluminescence signal from CTR and CCR4⁺ T cells in 3 representative SCID mice/group engrafted with TARC⁻ tumor (K/wt) on the left side and the TARC⁺ tumor (K/TARC) on the right side. Whereas no significant expansion of the bioluminescent signal was observed to either site of tumor in mice receiving CTR T cells (top picture in each pair), an increase in bioluminescence was observed in mice receiving CCR4⁺ T cells (bottom picture in each pair) only at the site of the tumor producing TARC. (C) The fold change of bioluminescence signal between K/wt and K/TARC sites for CTR (○) and CCR4⁺ (■) T cells. Data are mean ± SD of 6 mice. (D) The immunophenotype of K/wt (left plot) and K/TARC (right plot) tumors isolated from one representative mouse that received CCR4⁺ T cells, killed on day 9. After removal, the tumors were homogenized and cells stained to distinguish T cells (using an antihuman CD3 antibody) from Karpas tumor cells (using an antihuman CD30 antibody). As shown in the panel, the proportion of cells detectable at the site of tumor-secreting TARC (5.6%) was more than 10-fold higher compared with K/wt (0.5%). This correlated with the increase in bioluminescence signal at the site of K/TARC tumor (2.1 × 10⁶ p/sec/cm²/sr) compared with the site of K/wt tumor (1.4 × 10⁵ p/sec/cm²/sr).

Figure 6

Improved migration of activated T lymphocytes genetically modified to overexpress CCR4 and a CAR targeting the CD30 antigen expressed by HL. A bicistronic vector encoding CCR4 and CAR-CD30 was constructed and used to transduce T cells from 8 healthy donors. (A) The expression of CCR4 and CAR-CD30 by T cells transduced with a retroviral vector encoding CAR-CD30 (□) or with a bicistronic retroviral vector encoding CCR4 and CAR-CD30 (■). Surface expression of the CCR4 and CAR was evaluated by FACS analysis. (B) The migration of CAR-CD30⁺ (□) and CCR4⁺CAR-CD30⁺ (■) T cells toward TARC gradients, using a transwell migration assay. T-cell migration was evaluated using culture supernatants collected from HDLM-2 and L428, which physiologically produce high amounts of TARC, and against Karpas genetically modified to produce TARC (K/TARC). Karpas wild type (K/wt) was used as a control. The panel indicates that migration toward TARC is significantly improved for T cells genetically modified to overexpress CCR4 using the bicistronic vector CCR4(I)CAR-CD30. This improved migration was TARC mediated as it was inhibited by addition of anti-TARC antibodies but not by addition of an isotype control. (C) Killing of CD30⁺ (HDLM-2, ; Karpas, ■) and CD30⁻ (□) tumor cells by CAR-CD30⁺ and CCR4⁺CD30-CAR⁺ T cells. (D) The measurement of IL-2 cytokine released in the supernatant of T cells cocultured with or without HDLM-2 and assessed using a specific ELISA assay. T cells were transduced with either CAR-CD30 or CCR4(I)CAR-CD30, where CAR molecules also incorporate the CD28 endodomain. As a control, T cells were also transduced with the same CAR targeting CD30 but lacking the CD28 endodomain (CD30CARζ). As anticipated, we observed enhanced production of IL-2 by T cells transduced with the CAR containing the CD28 endodomain (CAR-CD30), regardless of coexpression of CCR4, but not by T cells transduced with CD30CARζ lacking CD28. The figure indicates that T cells transduced with CCR4(I)CAR-CD30 vector produce IL-2 in amounts comparable with that of T cells transduced with the vector encoding CAR-CD30 alone, confirming that the CD28 pathway is not impaired by the coexpression of CCR4. Data are mean ± SD of 10 donors.

Figure 7

Improved in vivo migration and antitumor activity of T cells transduced with a bicistronic vector encoding both CCR4 and CAR-CD30. For in vivo experiments, T cells transduced with CAR-CD30 or the bicistronic vector encoding CCR4 and CAR-CD30 were further transduced to express eGFP-FFLuc to monitor their migration in vivo in SCID mice using the IVIS system. (A) The expression of eGFP-FFLuc on CAR-CD30⁺ and CCR4⁺CAR-CD30⁺ cells in one representative donor. (B) The bioluminescence signal from T cells in 3 representative SCID mice engrafted with TARC⁻ tumor (Karpas/wt) on the left side and the TARC⁺ tumor (Karpas/TACR) on the right side. Whereas no significant expansion of the bioluminescent signal was observed at either tumor site in mice receiving CAR-CD30⁺ T cells (top panels), a significant increase of bioluminescence was observed in mice receiving CCR4⁺CAR-CD30⁺ T cells (bottom panels) at the site of tumor-producing TARC. (C) The fold change of bioluminescence signal between K/wt and K/TARC site for CAR-CD30⁺ (□) and CCR4⁺CAR-CD30⁺ (■) T cells. Data are mean ± SD of 8 mice. Data confirm that CCR4 expressed by T cells using a bicistronic vector remain functional in vivo. To evaluate in vivo antitumor activity of T cells transduced with the bicistronic vector, the HL-derived cell line HDLM-2 was further transduced with FFLuc and inplanted subcutaneously into NOG-SCID mice (7 mice/group). Tumor growth was monitored by measuring bioluminescence signals with the IVIS system. Mice then received intravenous T cells transduced with retroviral vectors encoding either CCR4 or CAR-CD30, or the bicistronic vector CCR4(I)CAR-CD30. (D) The bioluminescence signal of tumor cells in 4 representative mice/group. In mice receiving CCR4⁺ or CAR-CD30⁺ T cells, the bioluminescence signal, and thus tumor, increased over time. In contrast, in mice that received CCR4⁺CAR-CD30⁺ T cells, the signal remained stable, indicating tumor control. (E) The data from all 7 mice/group.