Tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) is an inhibitor of autoimmune inflammation and cell cycle progression

Abstract

The tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) induces apoptosis of tumor cells but not normal cells; its role in normal nontransformed tissues is unknown. We report here that chronic blockade of TRAIL in mice exacerbated autoimmune arthritis, and that intraarticular TRAIL gene transfer ameliorated the disease. In vivo, TRAIL blockade led to profound hyperproliferation of synovial cells and arthritogenic lymphocytes and heightened the production of cytokines and autoantibodies. In vitro, TRAIL inhibited DNA synthesis and prevented cell cycle progression of lymphocytes. Interestingly, TRAIL had no effect on apoptosis of inflammatory cells either in vivo or in vitro. Thus, unlike other members of the tumor necrosis factor superfamily, TRAIL is a prototype inhibitor protein that inhibits autoimmune inflammation by blocking cell cycle progression.

Figure 1

Recombinant sDR5 blocks TRAIL-induced apoptosis of tumor cells. Human Jurkat T cells (clone E6-1) and K562 B cells (10⁶/ml) were cultured in RPMI 1640 medium containing various concentrations of TRAIL (Biomol^® Research Laboratories), with or without 5 μg/ml of sDR5. 18 h later, apoptosis was determined by flow cytometry using annexin V–FITC according to the manufacturer's instructions (PharMingen). Jurkat cells were treated with either (A) 5 μg/ml of BSA, (B) 100 ng/ml of TRAIL, (C) 5 μg/ml of sDR5, or (D) 100 ng/ml of TRAIL plus 5 μg/ml of sDR5. Each histogram represents 10,000 events, with the apoptotic cells gated. (E) K562 cells were treated with various concentrations of TRAIL with (○) or without (▪) 5 μg/ml of sDR5. The percentage of apoptotic cells was determined as shown above in A–D. In parallel experiments, we also tested whether sDR5 blocked Fas- and ultraviolet-induced apoptosis of Jurkat cells; no effect of sDR5 on apoptosis was detected in these systems (our unpublished data).

Figure 2

Exacerbation of collagen-induced arthritis by TRAIL blockade. Groups of DBA/1 mice (four to six per group) were immunized on days 0 and 21 with chicken type II collagen as described in Materials and Methods. Starting from the second immunization, mice received daily intraperitoneal injections of 50–300 μg sDR5 or BSA in 0.5 ml PBS for a total of 21 d (in parallel experiments, other control proteins such as HSA were also used; similar results as reported here were observed). (A) Disease courses in mice treated with 100 μg of BSA (□) or sDR5 (•). Each data point represents a mean ± SD from a total of five (for sDR5-treated group) or six (for BSA-treated group) mice. The experiments were repeated five times with similar results. The differences between the two groups are statistically significant (P < 0.001) as determined by Mann-Whitney test. (B) Disease scores of individual feet 12 d after the second immunization. A total of four groups of mice are shown: one was treated with 100 μg BSA, and the other three were treated with 50–300 μg sDR5. Each data point represents an individual foot, with 16–24 feet per group. The differences between BSA- and sDR5-treated groups are statistically significant as determined by ANOVA (P < 0.05 for mice treated with 50 μg sDR5, and P < 0.01 for mice treated with 100 or 300 μg sDR5).

Figure 3

Inhibition of collagen-induced arthritis by TRAIL gene transfer. DBA/1 mice (four per group) were immunized with type II collagen as described in Materials and Methods. 6 d after the disease onset (10 d after the second immunization), mice were injected intraarticularly and periarticularly to the ankle and tarsal joints of the hind feet with 1 10 μl PBS (control), 2 10¹⁰ particles of Ad vector (the Ad vector contains no TRAIL gene but is otherwise identical to the TRAIL virus), 3 10⁹ particles of TRAIL virus, or 4 10¹⁰ particles of TRAIL virus in 10 μl of PBS (reference 17). Starting from the day of the viral injection, two groups of mice, one nontreated and the other injected with 10¹⁰ particles of TRAIL virus, were subjected to daily intraperitoneal injections of 100 μg sDR5 for a total of 14 d. Data presented are disease scores of individual hind feet 6 d after viral injection. A total of eight hind feet per group are shown. Only mice that received 10¹⁰ particles of TRAIL virus showed significant improvement (P < 0.001 as determined by ANOVA). Results are representative of two experiments.

Figure 4

Histochemical profiles of arthritic joints. DBA/1 mice (eight to nine per group) were immunized for arthritis and treated with sDR5 or BSA as described in the legend to Fig. 2 A. For A–D, G, and H, mice were killed 32 d after the second immunization, and their ankle joints were analyzed for histology and apoptosis as described in Materials and Methods. For E and F, mice were injected intraperitoneally with BrdU as described in Materials and Methods, and were killed 21 d after the second immunization. BrdU staining was performed as described in Materials and Methods. (A) An ankle joint of a BSA-treated mouse with a pathology score of 2 (HE staining; original magnification: ×20). Arrow indicates signs of synovitis. (B) An ankle joint of an sDR5-treated mouse with a pathology score of 4 (HE staining; original magnification: ×20). Arrows indicate severe synovitis, hyperplasia, and cartilage and bone destruction. (C) An ankle joint of a BSA-treated mouse with a pathology score of 1 (HE staining; original magnification: ×100). Arrow indicates signs of synovitis. (D) An ankle joint of an sDR5-treated mouse with a pathology score of 4 (HE staining; original magnification: ×100). Arrows indicate severe synovitis, hyperplasia, and cartilage and bone destruction. (E) An ankle joint of a BSA-treated mouse with a disease score of 2 (BrdU staining; original magnification: ×400). Arrows indicate BrdU⁺ nuclei, which are shown in brown. (F) An ankle joint of an sDR5-treated mouse with a disease score of 3 (BrdU staining; original magnification: ×400). (G) An ankle joint of a BSA-treated mouse with a pathology score of 4 (apoptotic staining; original magnification: ×200). Arrows indicate apoptotic cells, which are shown in brown. (H) An ankle joint of an sDR5-treated mouse with a pathology score of 4 (apoptotic staining; original magnification: ×200).

Figure 5

Pathological and apoptotic studies of arthritic joints. Mice were treated and killed as described in the legend to Fig. 4, and their paws were examined for the degrees of inflammation and apoptosis as described in Materials and Methods. A minimum of three comparable synovial sections per mouse were analyzed. A total of eight (for sDR5-treated group) and nine (for BSA-treated group) mice were used. For pathology scores (A), the differences between the two groups are statistically significant as determined by ANOVA (P < 0.01). For apoptotic index (B), the differences between the two groups are not statistically significant as determined by ANOVA (P = 0.21).

Figure 6

Effect of TRAIL blockade on anticollagen immune responses. DBA/1 mice (five to six per group) were treated as described in the legend to Fig. 2 A. To test cellular immune responses (A–C), mice were killed 32 d after the second immunization, and their inguinal lymph nodes were collected. For cytokine assays, lymph node cells (1.5 × 10⁶ per well) were cultured in 0.2 ml of serum-free medium (X-vivo 20; BioWhittaker) with or without 10–50 μg/ml of chicken type II collagen (CII) or 1 μg/ml of Con A. Culture supernatants were collected 40 h later, and IL-2/IFN-γ concentrations were determined by sandwich ELISA as described (reference 46). To test lymphocyte proliferation (A), lymph node cells (0.5 × 10⁶ cells per well) were first cultured for 72 h, then pulsed with [³H]thymidine for an additional 16 h. Radioactivity (presented as cpm) was determined using a Wallac β-plate counter. To test humoral immune responses (D and E), mice were bled retroorbitally 14 and 32 d after the second immunization, and anticollagen IgG1 and IgG2a antibodies were determined by ELISA using chicken type II collagen as antigen. Each data point represents a mean ± SD from five (for sDR5-treated group) or six (for BSA-treated group) mice. The experiments were repeated three times with similar results. When a similar ELISA was performed using sDR5 as detecting antigen, no anti-sDR5 antibodies were detected (data not shown). In parallel experiments, mice were killed 14 and 21 d after the second immunization, and were tested as described in A–C. Similar differences between control and sDR5-treated mice were observed (data not shown). Additionally, sDR5 had no effect on LPS-induced proliferation of splenocytes in vitro (our unpublished data). (A) Proliferative responses as determined by [³H]thymidine incorporation. (B) IL-2 production. (C) IFN-γ production. (D) Anticollagen IgG2a titers. (E) Anticollagen IgG1 titers. White bars, mice treated with BSA; black bars, mice treated with sDR5.

Figure 6

Effect of TRAIL blockade on anticollagen immune responses. DBA/1 mice (five to six per group) were treated as described in the legend to Fig. 2 A. To test cellular immune responses (A–C), mice were killed 32 d after the second immunization, and their inguinal lymph nodes were collected. For cytokine assays, lymph node cells (1.5 × 10⁶ per well) were cultured in 0.2 ml of serum-free medium (X-vivo 20; BioWhittaker) with or without 10–50 μg/ml of chicken type II collagen (CII) or 1 μg/ml of Con A. Culture supernatants were collected 40 h later, and IL-2/IFN-γ concentrations were determined by sandwich ELISA as described (reference 46). To test lymphocyte proliferation (A), lymph node cells (0.5 × 10⁶ cells per well) were first cultured for 72 h, then pulsed with [³H]thymidine for an additional 16 h. Radioactivity (presented as cpm) was determined using a Wallac β-plate counter. To test humoral immune responses (D and E), mice were bled retroorbitally 14 and 32 d after the second immunization, and anticollagen IgG1 and IgG2a antibodies were determined by ELISA using chicken type II collagen as antigen. Each data point represents a mean ± SD from five (for sDR5-treated group) or six (for BSA-treated group) mice. The experiments were repeated three times with similar results. When a similar ELISA was performed using sDR5 as detecting antigen, no anti-sDR5 antibodies were detected (data not shown). In parallel experiments, mice were killed 14 and 21 d after the second immunization, and were tested as described in A–C. Similar differences between control and sDR5-treated mice were observed (data not shown). Additionally, sDR5 had no effect on LPS-induced proliferation of splenocytes in vitro (our unpublished data). (A) Proliferative responses as determined by [³H]thymidine incorporation. (B) IL-2 production. (C) IFN-γ production. (D) Anticollagen IgG2a titers. (E) Anticollagen IgG1 titers. White bars, mice treated with BSA; black bars, mice treated with sDR5.

Figure 7

Inhibition of DNA synthesis and cell cycle progression by TRAIL. Splenocytes were prepared from 6–8-wk-old BALB/c mice (The Jackson Laboratory) and cultured in DMEM for 3 d in the presence of 10% FBS and 2.5 μg/ml of Con A. Live cells were then purified through a Ficoll gradient and cultured in 96-well plates at 3 × 10⁵ cells per well in 200 μl of DMEM containing 10% FBS, with or without the following reagents: 100 ng/ml of TRAIL, 5 μg/ml of sDR5, 5 μg/ml of anti-CD95L mAb (MFL-3), and anti–mouse CD3 mAb (which was coated on the plate by preincubating the plate with 10 μg/ml of the antibody at 4°C for 16 h). For apoptosis and cell cycle analyses (A and B), cells were cultured for a total of 24 h, fixed in 70% ethanol, and stained with 50 μg/ml of propidium iodide. For thymidine incorporation assays (C and D), cells were cultured for 24 h, then pulsed with 10 μCi/ml of [³H]thymidine for an additional 16 h. Cells were then harvested, and radioactivity was determined using a Wallac β-plate counter. (A) Percentage of apoptotic cells as determined by flow cytometry. The spontaneous apoptotic rate in cultures containing no anti-CD3 mAb was 12%, which was subtracted from the data presented here. For cultures that contained TRAIL and anti-CD3 mAb, the percentage of anti-CD3–induced apoptosis was 16 ± 1%, which was comparable to that of the cultures treated with anti-CD3 mAb alone. The differences between anti-CD95L mAb–treated culture and all other cultures are statistically significant as determined by ANOVA (P < 0.0001). (B) Number of S-G2/M cells per well as determined by flow cytometry. For cultures that contained TRAIL and anti-CD3 mAb, the number of cells in the S-G2/M phases was 70–90% of that of the anti-CD3 mAb–treated culture. The total numbers of live cells per well recovered from each group were as follows: cultures with anti-CD3 mAb alone, 0.3 × 10⁵; cultures with anti-CD3 mAb plus sDR5, 1.6 × 10⁵; cultures with anti-CD3 mAb plus anti-CD95L mAb, 1.1 × 10⁵; cultures with anti-CD3 mAb plus sDR5 plus TRAIL, 0.6 × 10⁵. The differences between all four groups are statistically significant as determined by ANOVA (P < 0.01). (C and D) DNA synthesis as determined by [³H]thymidine incorporation. Data presented are means cpm ± SD of triplicate cultures. For cultures containing anti-CD3 mAb, the differences between the two groups are statistically significant as determined by ANOVA (P < 0.01). The experiments were repeated twice with similar results. The concentrations of TRAIL and sDR5 used in these experiments were selected based on the dose-dependency studies performed in our laboratory. When 0.2–1 μg/ml of sDR5 was used, less significant effects on cell cycle progression were observed. Similarly, when 50–150 ng/ml of TRAIL was tested, much less significant effects on cell cycle were detected (data not shown).