Induction of tumor-specific T cell immunity by anti-DR5 antibody therapy

Abstract

Because tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) preferentially induces apoptosis in tumor cells and plays a critical role in tumor surveillance, its receptor is an attractive target for antibody-mediated tumor therapy. Here we report that a monoclonal antibody (mAb) against the mouse TRAIL receptor, DR5, exhibited potent antitumor effects against TRAIL-sensitive tumor cells in vivo by recruiting Fc receptor-expressing innate immune cells, with no apparent systemic toxicity. Administration of the agonistic anti-DR5 mAb also significantly inhibited experimental and spontaneous tumor metastases. Notably, the anti-DR5 mAb-mediated tumor rejection by innate immune cells efficiently evoked tumor-specific T cell immunity that could also eradicate TRAIL-resistant variants. These results suggested that the antibody-based therapy targeting DR5 is an efficient strategy not only to eliminate TRAIL-sensitive tumor cells, but also to induce tumor-specific T cell memory that affords a long-term protection from tumor recurrence.

Figure 1.

Characterization of MD5-1 in vitro. (A) Reactivity with mouse DR5, Fas, and Fn14 transfectants and mouse tumor cell lines. The thick histograms represent MD5-1 staining and the thin histograms represent control Ig staining. (B) Cytotoxic activity of biotinylated MD5-1 against 4T1 cells in the presence (□) or absence (○) of streptavidin. (C) Cytotoxic activity of MD5-1 against 4T1 cells in the presence of P815 cells with (○) or without (□) anti-FcR mAb (2.4G2). (D) Cytotoxic activity of mTRAIL/2PK-3 against 4T1 cells in the presence of MD5-1 or anti-TRAIL mAb (N2B2). (E) Effect of z-VAD-fmk on MD5-1–mediated cytotoxicity against 4T1 tumor cells. (F) Cytotoxic activity of MD5-1 and mTRAIL/2PK3 against R331-mock and R331-FLIP cells. Similar results were obtained in two or three independent experiments.

Figure 2.

Cytotoxic activity of MD5-1 in the presence of FcR-bearing cells. (A) FcR-mediated cytotoxic activity of MD5-1 against 4T1 tumor cells in the presence of fresh or fixed effector cells. CMA was added to inhibit perforin-dependent cytotoxicity and 2.4G2 was used to block FcγRII and FcγRIII. (B) Cytotoxic activity of MD5-1 against R331-mock and R331-FLIP tumor cells in the presence of various effector cells. (C) Cytotoxic activity of MD5-1 against 4T1 tumor cells in the presence of rabbit or mouse complement.

Figure 3.

No apparent toxicity of MD5-1 in vivo. (A) Serum AST and ALT levels 20 h after MD5-1 injection. Data are represented as the mean ± SE of 10 mice. Similar results were obtained in two independent experiments at several time points (not depicted). (B) Histological examination (H/E) of liver (×10) and thymus (×5) from MD5-1–treated mice.

Figure 4.

Contribution of macrophages and NK cells to tumoricidal effect of MD5-1. SCID mice (A) or wild-type BALB/c mice (B) were s.c. inoculated with 4T1 cells, and wild-type BALB/c mice (C) were s.c. inoculated with R331-mock cells. The mice were then treated with MD5-1 (solid symbols) or control Ig (open symbols). Some groups of mice were treated with anti-ASGM1 Ab (triangles), anti-CD11b mAb (circles), or control Igs (squares). (D) Wild-type BALB/c mice were s.c. inoculated with 4T1 cells. The mice were then treated with MD5-1 from day 0 (▪), 15 (•), or 25 (▴). One group of mice was treated with control Ig from day 0 (□) as the control. (E) Wild-type (squares), TRAIL^−/− (circles), perforin^−/− (triangles), and IFN-γ^−/− (upside down triangles) BALB/c mice were s.c. inoculated with R331-mock cells. The mice were then treated with MD5-1 (solid symbols) or control Ig (open symbols). (F) Wild-type (squares), FcRγ^−/− (circles), or FcγRII^−/− (triangles) BALB/c mice were s.c. inoculated with 4T1 cells. The mice were then treated with MD5-1 (solid symbols) or control Ig (open symbols). Some groups of wild-type mice were treated with both anti-ASGM1 Ab and anti-CD11b mAb (upside down triangles). All data are represented as the mean ± SE of 5–10 mice. Similar results were obtained in two or three independent experiments.

Figure 5.

MD5-1 induces infiltration of macrophages and DCs in tumor site. Tumor was removed 8 d after tumor inoculation from control Ig-, MD5-1–, or MD5-1– and anti-CD11b mAb (5C6)–treated mice, and then stained with H/E (×5), anti-F4/80 mAb (×10), or anti-CD205 mAb (×10). The marginal region of s.c. tumor is shown. A massive infiltration of F4/80⁺ macrophages (stained brown) was observed in the marginal region of tumor from MD5-1–treated mice. The arrows are indicated to distinguish CD205⁺ DCs from nonspecific staining of mast cells. Representative of five tumors in each group.

Figure 6.

Induction of tumor-specific T cell memory by MD5-1–mediated tumor rejection. (A) BALB/c mice were inoculated with 4T1 cells and then treated with MD5-1 mAb (▪; n = 25) or control Ig (□; n = 5) (left). Simultaneously to secondary challenge with 4T1 cells, some mice were treated with anti-CD4 mAb (•), anti-CD8 mAb (▴), anti-CD4 and anti-CD8 mAbs (▾), or control Ig (▪) (middle). Naive mice were inoculated with 4T1 cells as the control (□). Irrelevant R331-mock cells were inoculated into 4T1-rejecting mice (♦) or naive mice (⋄) (right). (B) BALB/c (squares) and SCID (circles) mice were inoculated with 4T1 cells and then treated with MD5-1 mAb (solid symbols; n = 25) or control Ig (open symbols; n = 5) (left). 5 wk after the first inoculation, the MD5-1–treated BALB/c (▪) and SCID (•) mice that had rejected 4T1 were secondarily inoculated with 4T1 cells (middle). Naive BALB/c (□) and SCID (○) mice were inoculated with 4T1 cells as the control. 2 d after inoculation of 4T1 cells, some SCID mice were i.v. transferred with splenic T cells from the MD5-1–treated BALB/c mice that had rejected 4T1 (•) or naive BALB/c mice (▵) (right). Naive SCID mice (○) and MD5-1–treated wild-type BALB/c mice that had rejected 4T1 (▪) were inoculated with 4T1 cells as the control. (C) Naive BALB/c mice (▿), MD5-1–treated wild-type BALB/c mice that had rejected live 4T1 (□), BALB/c mice immunized by irradiated 4T1 cells precoated with control Ig (▵), or BALB/c mice immunized by irradiated 4T1 cells precoated with MD5-1 (○) were inoculated with the indicated number of live 4T1 tumor cells. (D) Cytotoxic activity of splenic T cells against 4T1 (left) or R331-mock (right) target cells. Effector splenic T cells were prepared from BALB/c mice that had rejected live 4T1 cells by MD5-1 treatment (□), BALB/c mice that had been immunized by irradiated 4T1 cells precoated with MD5-1 (○) or control Ig (▵), or naive BALB/c mice (▿). All data are represented as the mean ± SE of 5–10 mice. Similar results were obtained in three to five independent experiments.

Figure 7.

Induction of perforin- and FasL-mediated T cell cytotoxicity that can also eliminate TRAIL-resistant variants. (A) BALB/c mice were inoculated with R331-mock (squares) or R331-FLIP (circles) cells, and then treated with MD5-1 mAb (solid symbols; n = 20) or control Ig (open symbols; n = 5) (left). The cured mice were secondarily challenged with R331-mock (▪) or R331-FLIP (•) cells (middle). Naive mice were inoculated with R331-mock (□) or R331-FLIP (○) cells, and the mice preimmunized by irradiated R331-mock cells were inoculated with R331-mock cells (▵) as the controls (middle). Simultaneously to the secondary challenge with R331-mock cells, some mice were treated with anti-CD8 mAb (▵), anti-CD4 and anti-CD8 mAbs (▿), or control Ig (□) (right). (B) R331-FLIP cells were inoculated into the left flank (•) at the indicated time point after R331-mock cell inoculation into the right flank (▪) and commencement of MD5-1 treatment on day 0. As a control, R331-FLIP cells were inoculated into naive mice (○) at the same time as the R331-FLIP inoculation into the treated mice. (C) BALB/c mice that had rejected R331-mock by MD5-1 treatment were secondarily challenged with the indicated number of R331-mock or R331-FLIP cells (○). Naive mice were inoculated with the same numbers of R331-mock or R331-FLIP cells and treated with MD5-1 (▪) or control Ig (□) for comparison. (D) Wild-type (squares) or perforin^−/− (circles) BALB/c mice were inoculated with R331-mock cells, and then treated with MD5-1 mAb (solid symbols; n = 20) or control Ig (open symbols; n = 5) (left). The cured mice were secondarily challenged with R331-mock cells. Simultaneously to the secondary challenge, some wild-type (middle) and perforin^−/− (right) mice were treated with anti-FasL neutralizing mAb (○), anti-TRAIL neutralizing mAb (▵), anti-FasL and anti-TRAIL mAbs (▿), or control Ig (□). Naive wild-type or perforin^−/− mice were inoculated with the same number of R331-mock cells as a control (⋄). All data are represented as the mean ± SE of 5–10 mice. Similar results were obtained in two independent experiments.