Gene transfer of IL-27 to tumor cells has been proven to inhibit tumor growth in vivo by antiproliferation, antiangiogenesis, and stimulation of immunoprotection. To investigate the nonimmune mechanism of IL-27 that suppresses lung cancer growth, we have established a single-chain IL-27-transduced murine Lewis lung carcinoma (LLC-1) cell line (LLC-1/scIL-27) to evaluate its tumorigenic potential in vivo. Mice inoculated with LLC/scIL-27 displayed retardation of tumor growth. Production of IL-12, IFN-γ, and cytotoxic T cell activity against LLC-1 was manifest in LLC/scIL-27-injected mice. Of note, LLC-1/scIL-27 exhibited decreased expression of cyclooxygenase-2 (COX-2) and PGE2. On the cellular level, the LLC/scIL-27 transfectants had reduced malignancy, including down-regulation of vimentin expression and reduction of cellular migration and invasion. The suppression of tumorigenesis by IL-27 on lung cancer cells was further confirmed by the treatment with rIL-27 on the murine LLC-1 and human non-small cell lung carcinoma (NSCLC) cell lines. PGE2-induced vimentin expression, movement, and invasiveness were also suppressed by the treatment with rIL-27. Our data show that IL-27 not only suppresses expression of COX-2 and PGE2 but also decreases the levels of vimentin and the abilities of cellular migration and invasion. Furthermore, inoculation of LLC/scIL-27 into immunodeficient NOD/SCID mice also exhibited reduced tumor growth. Our data indicate that IL-27-induced nonimmune responses can contribute to significant antitumor effects. Taken together, the results suggest that IL-27 may serve as an effective agent for lung cancer therapy in the future.
Lung cancer is the leading cause of human cancer deaths worldwide. Survival rates of lung cancers can be increased by successful early detection and improved systemic therapies in early-stage diseases. Unfortunately, most patients are diagnosed with the advanced, unresectable disease and have a poor prognosis. Despite available treatments of lung cancers, the overall 5-year survival rate is only 8–14% and has been improved only marginally during the past 25 years. Novel therapeutic modalities, such as molecular targeted therapies and immunotherapeutic strategies, are needed for lung cancers (1, 2).
IL-27 is a novel member of the IL-6/IL-12 family (3, 4, 5). IL-27 is a heterodimeric cytokine composed of the EBV-transformed gene 3 and p28. IL-27 can be produced by activated APCs (3, 4, 5). The IL-27 receptor, IL-27R, consists of WSX-1 (also known as the T cell cytokine receptor) and gp130 subunits and is expressed in immune cells and some nonimmune cells such as endothelial cells and epithelial cells (3, 4, 5). IL-27 activates JAK-STAT and enhances proliferation in naive CD4+ T cells but does not induce similar effects on memory CD4+ T cells (3, 4, 5, 6, 7). IL-27 also induces the expression of T-bet (6, 7) and IL-12Rβ2 and further stimulates Th1 differentiation. Moreover, IL-27 and IL-12 synergize to enhance IFN-γ production (6, 7). Previous studies have showed that IL-27-transduced colon carcinoma (8), as well as neuroblastoma (9), can increase CD8+ T cell-dependent IFN-γ production, cytotoxicity, and tumor clearance. In addition, IL-27 has antiangiogenic and antiproliferative activities that inhibit tumor growth and metastasis in murine melanoma (10, 11). The antiangiogenic activities of IL-27 are through the production of IFN-γ that induces endothelial expression of IP-10 and MIG, the antiangiogenic CXC chemokines (10). It has been shown that activation of WSX-1/STAT-1 signaling is essential for the IL-27-regulated antiproliferative effects on melanoma (11). However, the direct effects of IL-27 on lung cancer cells remain to be determined.
Cyclooxygenases (COX)3 are the key enzymes in the biosynthesis of prostaglandins. COX-1 is constitutively expressed in most tissues to mediate normal physiologic function. COX-2, an inducible form of COX, is overexpressed in early and advanced lung cancer tissues. Expression of COX-2 is associated with poor prognosis of lung cancer (12, 13, 14, 15). Elevated levels of tumor COX-2 and its metabolite, PGE2, contribute to induction of tumor angiogenesis (13, 16), augmentation of cancer invasiveness (13, 17), resistance to apoptosis (13, 18), and suppression of antitumor immunity (13, 19, 20, 21). It has been shown that inhibition of COX-2 can decrease tumor volume in several lung cancer models (16, 21).
The epithelial-mesenchymal transition (EMT) is a cellular process in which epithelial cells are converted into motile mesenchymal cells. In addition to its role in normal embryonic development, EMT has been found in a variety of solid tumors and is closely correlated with tumor invasion, metastasis, and unfavorable prognosis (22, 23, 24). Common molecular markers for EMT are the increased expression of vimentin and loss of E-cadherin (22). Atypical high vimentin expression is observed in many different types of human cancers and has frequently been applied to grade and diagnose different types of tumors (25). Vimentin-transduced cancer cells exhibit increased motile activity, invasiveness, and tumorigenicity (25). The loss of E-cadherin has been reported to be associated with a poor clinical outcome in non-small cell lung cancer (NSCLC) (23, 26, 27). Available studies in NSCLC have shown that COX-2 increases invasive capacity via PGE2-mediated regulation of CD44 and matrix metalloproteinase-2 (17). In addition, inhibition of tumor COX-2 can increase E-cadherin expression and loss of cell aggregation in NSCLC. These effects are accomplished via suppression of PGE2 that mediates induction of the E-cadherin transcriptional repressors ZEB-1 and Snail (SNAI-1) (28). Thus, COX-2 can regulate lung cancer tumorigenesis (12, 13).
IL-27 has been known to induce activation of immune cells and endothelial cells. In the present study, we investigated the direct effects of IL-27 on tumor growth in lung cancer. We found that IL-27Rs were expressed by lung cancer cells and that single-chain IL-27 (scIL-27)-overexpressed Lewis lung carcinoma-1 (LLC-1) induces antitumor effects in vivo. In addition to the generation of specific cellular and humoral immunity in vivo, we also found that both overexpression of IL-27 and treatment with rIL-27 can directly inhibit expression of vimentin, COX-2, and its metabolite, PGE2, on lung cancer cells. These effects result in the reduction of cancer migration and invasion. The IL-27-dependent nonimmune effects that contribute to antitumor activities were validated in immunodeficient mice. These results suggest that IL-27 may be an effective agent for the treatment of human NSCLC.
Materials and Methods
Cells and mice
LLC-1 (H-2b29) were cultured in RPMI 1640 medium containing 10% FBS, 1% penicillin-streptomycin, 2 mM l-glutamine, and 1% nonessential amino acids. These cells were cultured in a humidified incubator containing 5% CO2 at 37°C. Male C57BL/6 (H-2b) mice were purchased from the National Laboratory Animal Center (Taipei, Taiwan), and male nonobese diabetic (NOD)/SCID (H-2d) mice were purchased from Animal Center of National Taiwan University (Taipei, Taiwan). The animals were raised under specific pathogen-free conditions in the Animal Center of National Yang-Ming University (Taipei, Taiwan) according to the regulations of the Animal Care Committee of National Yang-Ming University.
2 and anti-β-actin (AC-15) Abs were obtained from Sigma-Aldrich. Anti-COX-2 (M-19), anti-phospho-ERK1/2 (T202/Y204), anti-NF-κB p65 (C-20), anti-phospho-Akt (S473), and HRP-conjugated anti-goat IgG Abs were from Santa Cruz Biotechnology. Anti-phospho-STAT3 (Y705), anti-ERK1/2 (137F5), anti-phospho-NF-κB p65 (S536), anti-phospho-JNK (T183/Y185), and anti-phospho-p38 (T180/Y182) Abs were purchased from Cell Signaling Technology. Anti-vimentin (V9) Abs were from Neomarker. Anti-phospho-STAT1 (Y701) Abs were obtained from Zymed Laboratories. HRP-conjugated anti-rabbit IgG Abs were from Amersham/GE Healthcare. HRP-conjugated anti-mouse IgG, HRP-conjugated anti-rat IgG, and FITC-conjugated anti-mouse IgG Abs were from Jackson ImmunoResearch Laboratories.
Generation and growth of LLC-1/scIL-27 transduced cell lines
scIL-27, which contains a fusion sequence encoding the mature coding sequences for murine EBI3 followed by a synthetic linker (GSTSGSGKPGSGEGSTKG) and the mature coding sequence of murine p28, was amplified by PCR from the original template, pCR-Blunt-scIL-27 (M.-H. Tao, unpublished observations), and cloned into the XhoI and Hin4/ml) were cultured in 10% FBS-DMEM for 7 days. The number of cells was measured daily by a CellTiter 96 aqueous non-radioactive cell proliferation assay (Promega).
Anchorage-independent growth assay
Briefly (30), 1 × 103 LLC-1 and transfectants were suspended in 0.33% Bacto agar (Sigma-Aldrich) dissolved in DMEM with 10% FBS and antibiotics. Suspended cells were layered over 0.5% Bacto agar in the same medium in a 60-mm dish. On day 18, cells were fixed and then stained for the examination of colony growth, colonies > 60 μm (>60 cells) were counted, and results were expressed as the mean number of colonies ± SD.
RT-PCR for detection of scIL-27 and selected genes
Total cellular RNA was extracted using the Ultraspec RNA isolation system (Biotecx Laboratories). Each extracted RNA sample (5 μg) was reverse transcribed into cDNA by following the manufacture’s protocol for RT-PCR (Promega). The amplification was conducted for 25–35 cycles. The forward and reverse primers were synthesized by Mission Biotech (supplemental Table I).4
RT-quantitative PCR (RT-qPCR)
2-microglobulin mRNA of mouse cell lines or GAPDH mRNA of human cell lines.
Detection of secreted scIL-27, TGF-β1, and PGE2
Each transfectant (5 × 106 cells/10 ml) was cultured in DMEM for 24 h and the supernatants from the cells were collected and precipitated by Con A-Sepharose 4B (Amersham/GE Healthcare) for the detection of scIL-27. After washing the precipitates with PBS, the complexes were separated by electrophoresis on 10% SDS-polyacrylamide gels and transferred to nitrocellulose (Millipore). The membrane was then blotted with anti-mIL-27 Ab (R&D System) followed by a secondary anti-rat IgG Ab (Jackson ImmunoResearch Laboratories). In addition, the commercially available mouse IL-27 p28 ELISA kit (R&D System), the multispecies TGF-β1 ELISA kit (BioSource International), and the PGE2 EIA kit (Cayman Chemicals) were used for the measurement of corresponding substance concentrations.
Total cellular proteins from transfectants and treated cells were extracted in the lysis buffer (50 mM Tris (pH 7.4), 10% glycerol, 150 mM NaCl, 1% Triton-X, 1.5 mM MgCl2, 50 mM NaF, 1 mM Na3VO4, 1 mM PMSF, 1 mM EDTA, and 2 μg/ml aprotinin). Equal amounts of protein (60 μg) were loaded onto 10% SDS-polyacrylamide gels and transferred to nitrocellulose (Millipore) for Western blot analysis.
In vivo tumor growth
Mice were s.c. inoculated with different transduced tumor cell lines (2 × 105/100 μl). Tumor growth was measured with a caliper and calculated as length × width × height (in mm3) at intervals of 3 days. Mice bearing tumors with sizes > 2 cm3 were considered dead and survival rates were calculated.
IL-12, IFN-γ, and IL-10 production by spleen cells
LLC-1 cells were incubated with mitomycin-C (15 μg/ml; Kyowa) at 37°C for 60 min and used as tumor Ags for in vitro culture. C57BL/6 mice were s.c. injected with the LLC-1 cells and the transduced cell variants (2 × 105/100 μl) for 7 days. After removing the spleen from each mouse, the RBCs were lysed with the ACK lysis buffer (0.15M NH4Cl, 10 mM KHCO3, and 0.1 mM Na2 EDTA) at room temperature for 5 min. The spleen cells were washed in triplicate with RPMI 1640 medium supplemented with 10% FBS. The splenocytes (2 × 106/ml) were cultured in the presence of 2-ME (20 μg/ml) and LLC-1 tumor Ag (1 × 106/ml mitomycin-treated cells) at 37°C for 3 days (31). The concentrations of IL-12, IFN-γ, and IL-10 in the supernatant were measured by commercially available mouse cytokine ELISA kits (BioSource International).
Cell-mediated cytotoxicity assay (CTL)
After immunizing C57BL/6 mice with LLC-1 or transduced cells (2 × 105/100 μl) for 7 days, the spleen cells (1.6 × 106/ml) isolated from each mice were restimulated in vitro by coculture with mitomycin-C-treated LLC-1 tumor Ag (4 × 105 cells/ml) and IL-2 (20 IU/ml; Chiron) at 37°C for 4 days and used as effector cells in a lactate dehydrogenase release assay. After adding the LLC-1 target cells (104/100 μl) to different numbers of effector cells (100 μl) in 96-well U-bottom plates, the plates were centrifuged at 250 × g for 2 min and then cultured at 37°C in a 5% CO2 incubator for 16 h. The plates were then centrifuged at room temperature and lactate dehydrogenase activity in the supernatants of each well was detected by the CytoTox 96 non-radioactive cytotoxicity assay kit (Promega) (31). The percentage of lysis was calculated as follows: percentage lysis = (experimental release − spontaneous release)/(total release − spontaneous release) × 100.
ELISA for serum IgG against LLC-1
Sera were obtained from the mice 14 and 21 days after immunizing with different tumor cell lines (2 × 105/100 μl). LLC-1 cells (2 × 105/100 μl) were cultured in 96-well plates at 37°C for 16–18 h and then fixed with methanol at room temperature for 10 min. Serum samples were diluted by 400-fold, and the measurement of serum IgG was followed by our established ELISA procedures (31).
Invasion assay and migration assay of tumor cells
Transwell inserts (8-μm pore; Corning Costar) were placed into the wells. In an invasion assay, BD Matrigel (BD Biosciences) was hydrated with 10% FBS-DMEM. A total of 5 × 104 cells in 0.5% FBS-DMEM were placed in the upper chamber and 10% FBS-DMEM were added in the lower chamber. After a 20-h incubation at 37°C in a humidified 5% CO2 atmosphere, the cells that had invaded another side of the membrane were fixed with methanol and the Matrigel and noninvaded cells were mechanically removed with a cotton swab. The adherent cells on the membrane were stained with Liu’s stain (Muto Pure Chemicals). Cells were examined under microscopy at ×100 original magnification. In the migration assay, Transwell inserts were not coated with Matrigel and the procedures were the same as those of the invasion assay protocol.
Cells (5 × 104) were fixed in PBS containing 4% formaldehyde and 400 mM sucrose followed by permeabilization in 1% Triton X-100. After washing, cells were blocked in 5% BSA, incubated with anti-vimentin Abs, and then incubated with a FITC-conjugated anti-mouse IgG Ab and counterstained with 4′,6-diamidino-2-phenylindole (Sigma-Aldrich). Cells were mounted in Immunofluor mounting medium (ICN Biochemicals) and detected by scanning confocal microscopy (FV-1000; Olympus).
Data were expressed as mean ± SD and statistical significance was assessed by the Student t test.
Generation and characterization of LLC-1/scIL-27
To test whether IL-27 could enhance the antitumor effects of lung carcinoma, we prepared LLC-1 cells that were stably transfected with scIL-27 cDNA (LLC-1/scIL-27) or empty vector (LLC-1/pZeoSV). The expression of scIL-27 was confirmed by RT-PCR and Western blotting (Fig. 1⇓A). Among the seven clones, clone 18 and clone 19 (O18 and O19) expressed the high RNA and protein levels of scIL-27 (Fig. 1⇓A). The supernatants from O18 and O19 also contained a large amount of soluble IL-27 p28 (118.24 and 127.55 pg/5 × 105 cells/ml for 24 h) as shown by ELISA (supplemental Table III). We used O18 and O19 in subsequent experiments.
Overexpression of scIL-27 in lung carcinoma inhibited tumor growth in vivo
First, we examined whether the overexpression of scIL-27 in tumor cells affected tumorigenicity in vivo (Fig. 1⇑, C and D). In comparison with mice injected with parental LLC-1 (8142 mm3/33 days) and LLC-1/pZeoSV (7072 mm3/33 days), mice injected with scIL-27 transfectants (O18, 1974 mm3/33 days; O19, 2496 mm3/33 days) showed reduced tumor growth rates (Fig. 1⇑C) and prolonged life spans (Fig. 1⇑D). To rule out the cytotoxic effects of scIL-27 on lung cancer cells, we compared the cellular properties of LLC-1 and transduced cells in vitro. As compared with parental LLC-1, scIL-27 transfectants showed similar effects on monolayer growth (Fig. 1⇑B, left panel) and three-dimensional growth in soft agar (Fig. 1⇑B, right panel). We found that the expressions of IL-27 receptors (gp130 and WSX-1) in our test cells and the levels of IL-27 receptors were not changed among all test cells (see Fig. 3⇓). The results indicate that scIL-27 overexpression generates antitumor activity in vivo, excluding the direct cytotoxic effects.
scIL-27 stimulated a specific antitumor immune response against lung cancer cells
Previous studies show that IL-27 can generate tumor-specific Th1 immune responses. IL-27-induced antitumor immunity is through the production of IFN-γ that enhances tumor-specific CTL activity (8, 9, 10). We then examined whether IL-27 similarly influenced cellular and humoral immune responses against LLC-1 in vivo (Fig. 2⇓). Mice were s.c. injected with scIL-27 transfectants followed by the measurement of cytokine production, CTL activity, and tumor-specific Abs. In contrast to no change in IL-1β production (data not shown), Th1 cytokine (IFN-γ and IL-12) production (Fig. 2⇓, A and B) and CTL activity (Fig. 2⇓D) were increased. However, Th2 cytokine (IL-10) secretion (Fig. 2⇓C) was reduced in splenocytes from mice injected with LLC-1/scIL-27. In addition, increased tumor-specific IgG (Fig. 2⇓E) was detected in the sera of mice injected with LLC-1/scIL-27. Therefore, overexpression of scIL-27 in LLC-1 lung cells is able to induce specific Th1 antitumor immunity and humoral immunity in vivo.
COX-2 and TGF-β1 were down-regulated by overexpression of scIL-27 in LLC-1 cells
Studies show that COX-2 (13) metabolite and TGF-β1 production (32) by tumor cells causes immune escape via decreasing levels of IL-12 and increasing levels of IL-10 in immune cells. We then hypothesized that overexpression of IL-27 in lung cancer cells may stimulate the antitumor capacities of immune cells via decreased expression of COX-2 and TGF-β1, thereby reducing IL-10 and increasing IL-12 levels in vivo (Fig. 2⇑, B and C). Expressions of COX-2 and TGF-β1 in LLC-1/scIL-27 and control cells were further examined by RT-PCR and RT-qPCR. In comparison with control cells (LLC-1/pZeoSV), LLC-1/scIL-27 transfectants had low levels of COX-2 and TGF-β1 transcripts (Fig. 3⇓). However, there were no differences in the expressions of the urokinase plasminogen activator receptor and the angiogenic factor (vascular endothelial growth factor (VEGF)) (Fig. 3⇓, A and B). Decreased protein levels of COX-2 (Fig. 4⇓A, top panel) and TGF-β1 (Fig. 4⇓B, bottom panel) were confirmed in LLC-1/scIL-27 cells, and secreted levels of COX-2-metabolized PGE2 were also reduced in LLC-1/scIL-27 (Fig. 4⇓B, upper panel). Furthermore, COX-2 protein expression was significantly decreased in tumor mass from mice injected with LLC-1/scIL-27 as compared with control groups (Fig. 4⇓A, bottom panel). These results show that tumor expression of COX-2 and TGF-β1 can be reduced by overexpression of scIL-27 in LLC-1 cells.
scIL-27 overexpression in LLC-1 cells reduced level of vimentin and invasive ability
Overexpressions of COX-2 and PGE2 are associated with enhanced cellular invasion and metastasis (13). Thus, we evaluated whether LLC-1/scIL-27 cells have reduced abilities of cellular migration and invasion. We found a significant reduction in the number of migrated and invasive cells of scIL-27 transfectants as compared with parental control cells (Fig. 4⇑C). Quantitative analysis showed 23.5 and 32% reduction in cell movement and invasiveness, respectively, of LLC-1/scIL-27 cells (Fig. 4⇑C). It has been reported that reduction of E-cadherin and induction of vimentin are highly correlated with increased cell motility and invasiveness of cancer cells (22, 25). We further examined whether the overexpression of scIL-27 was able to change levels of E-cadherin and vimentin in LLC-1. Overexpression of scIL-27 in LLC-1 displayed slightly higher levels of E-cadherin and significantly lower levels of vimentin (Fig. 4⇑D) as compared with parental control cells. Decreased levels of vimentin in LLC-1/scIL-27 were also observed by immunofluorescence analysis (Fig. 4⇑E). In addition, RNA levels of the E-cadherin repressors ZEB-1, SNAI-1, and SNAI-2 were repressed by scIL-27 overexpression (Fig. 4⇑F). These results indicate that overexpression of scIL-27 in lung cancer cells can restrain the expression of vimentin and the abilities of migration and invasion.
rIL-27 suppressed COX-2, PGE2, TGF-β1, and vimentin in LLC-1 cells
To further confirm the direct effects of IL-27 on down-regulating the expressions of COX-2 and TGF-β1 in LLC-1, we analyzed levels of COX-2, TGF-β1, and VEGF transcripts in LLC-1 cells treated with various doses of rIL-27. As expected, rIL-27 suppressed expressions of COX-2 and TGF-β1 in LLC-1 in a dose- (Fig. 5⇓A, left panel) and time-dependent (Fig. 5⇓A, right panel) manner, whereas it had no effect on the levels of VEGF (Fig. 5⇓A). Protein levels of COX-2 (Fig. 5⇓D, upper panels) and secreted PGE2 and TGF-β1 (Fig. 5⇓C) were also reduced in cells treated with rIL-27. Treatment of rIL-27 could slightly promote E-cadherin expression and significantly repress vimentin expression (Fig. 5⇓D, upper panels). In addition, rIL-27 suppressed the expressions of ZEB-1, SNAI-1, and SNAI-2 (Fig. 5⇓D, bottom panel) that contribute to stable expression of E-cadherin (Fig. 5⇓D, upper panels). Moreover, down-regulation of COX-2, PGE2, TGF-β1, and vimentin by rIL-27 was reversed upon the addition of anti-IL-27 function-blocking Abs (Fig. 5⇓, B and C). Therefore, rIL-27 treatment can lead to suppression of COX-2, PGE2, TGF-β1, and vimentin in lung cancers.
Reduced PGE2 was involved in the down-regulation of vimentin by rIL-27 in LLC-1 cells
The role of PGE2 regulated by IL-27 on the expressions of E-cadherin, vimentin, and TGF-β1 in LLC-1 cells was further examined. When LLC-1 cells were treated with PGE2 and enhanced E-cadherin repressors (ZEB-1 and SNAI-1) (Fig. 5⇑E, bottom panels), slightly decreased E-cadherin (Fig. 5⇑E, upper panels) and increased vimentin (Fig. 5⇑E, upper panels) were found. PGE2-induced vimentin was diminished when LLC-1 cells were pretreated with rIL-27 (Fig. 5⇑E, upper panels). However, we did not see a change in TGF-β1 expression from the PGE2 treatment (Fig. 5⇑E, bottom panels). In addition, TGF-β1 alone could not change the levels of COX-2, PGE2, E-cadherin, and vimentin (Fig. 5⇑F). Therefore PGE2 plays an important role in suppression of vimentin by rIL-27.
Reduced PGE2 was involved in the suppression of migration and invasion of LLC-1 by rIL-27
Next, we examined the effects of rIL-27, PGE2, and TGF-β1 on the cellular migration and invasion of LLC-1 cells. LLC-1 cells treated with rIL-27 had a low expression level of vimentin in the cytoplasm as shown by immunofluorescence analysis (Fig. 6⇓A). When cells were pretreated with anti-IL-27 blocking Abs, the effect of rIL-27-induced down-regulation of vimentin was diminished (Fig. 6⇓A). In addition, rIL-27 decreased the cellular migration and invasion of LLC-1 cells, and the effects were reversed when cells were treated with anti-IL-27 Abs (Fig. 6⇓, B and C). Furthermore, PGE2, but not TGF-β1 increased the cellular migration and invasion of LLC-1 cells, and the PGE2-induced effects could be reduced when cells were treated with rIL-27 (Fig. 6⇓, B and C). Therefore rIL-27 decreases the level of vimentin and restrains cellular migration and invasion in lung cancers. We further confirmed the nonimmune effects of IL-27 directly on LLC-1 by inoculation of these IL-27 transfectant cells into immunodeficient mice (NOD/SCID mice). Overexpression of scIL-27 (3712 mm3/33 days) was still capable of suppressing the growth of LLC-1 cells in NOD/SCID mice compared with the LLC-1 (5878 mm3/33 days) and LLC-1/pZeoSV (5710 mm3/33 days) control groups (Fig. 6⇓D). Taken together, these results indicate that IL-27 directly inhibits expressions of COX-2, PGE2, and vimentin, as well as the abilities of migration and invasion. Both nonimmune and immune effects contribute to IL-27 antitumor activity in vivo.
rIL-27 suppressed COX-2, PGE2, and vimentin in human NSCLC cancer cells
In addition to murine lung cancer cells, the reduction of tumorigenesis by IL-27 was investigated in human lung cancer cells. We checked three human NSCLC cell lines. CL1-5 is a highly invasive cell line derived from parental CL1-0 cells (29). We found that the invasive A549 as well as CL1-5 lung cancer cell lines have high expression levels of COX-2 and vimentin (Fig. 7⇓A). We then further examined the direct effects of IL-27 on A549 and CL1-5 cells. When these cells were stimulated with rIL-27, the reductions of mRNA for COX-2 was observed in the A549 and CL1-5 cells by RT-qPCR (Fig. 7⇓B, left panel). Decreased protein levels of COX-2 and vimentin (Fig. 7⇓B, right panels) were also found in rIL-27-treated A549 and CL1-5 cells, respectively. Furthermore, the secretion of PGE2 (Fig. 7⇓C) as well as the abilities of cellular invasion and migration (Fig. 7⇓D) were all restrained by the addition of rIL-27 to A549 and CL1-5 cells. Therefore rIL-27 treatment can lead to suppression of COX-2, PGE2, and vimentin, resulting in the reduction of migration and invasion of human lung cancer cells.
The most frequent type of human lung carcinomas is NSCLC. Surgery, radiotherapy, and chemotherapy are the primary treatments for advanced NSCLC (33). However, the successful targeting therapy for advanced NSCLC is now considered to be combined modality treatments. Novel approaches, such as molecular targeted therapies and immunotherapeutic strategies, are urgently required for advanced NSCLC (1, 2). In the present study, we show that scIL-27-overexpressing lung cancer cells augment cellular and humoral antitumor immunity. We provide the first evidence that IL-27 directly suppresses tumorigenicity through down-regulation of COX-2 and PGE2 in tumor cells. The reductions of COX-2 and PGE2 by IL-27 result in inhibition of vimentin expression and suppression of cellular migration and invasion. The IL-27-induced responses can cause retardation of tumor growth in vivo and consequently prolong survival rates of mice.
Lung cancer cells elaborate a mechanism of immune escape through the release of immunosuppressive mediators including type 2 cytokine (IL-10), PGE2, and TGF-β, and the responses lead to tumor progression (12, 32). It has been shown that the expression of COX-2 is implicated in inflammation and cancer (34). Over-production of PGE2 that is derived from COX-2 induction can promote tumor progression (35). Down-regulation of the epithelial adhesion molecule E-cadherin by treatment of PGE2 has been found in a variety of solid tumors (13). In this study, we show that overexpression of and treatment with IL-27 in LLC-1 can significantly inhibit production of COX-2 and PGE2, resulting in changes in the expression levels of vimentin and E-cadherin (Figs. 4⇑ and 5⇑). Reductions of migration and invasion abilities are found in IL-27-transduced cells and IL-27-treated cells (Figs. 4⇑B and 6⇑, B and C). Administration of rIL-27 inhibits the PGE2-induced migration and invasion of LLC-1 cells (Fig. 6⇑, B and C). The IL-27-induced effects can be validated in that IL-27 overexpression significantly suppresses in vivo tumorigenesis in our LLC-1 model using immunodeficient mice (Fig. 6⇑D). We also find similar effects with IL-27-treated human lung cancer cell lines (Fig. 7⇑, B–D). Therefore, IL-27 can directly restrain tumorigenesis of human lung cancers. In addition, these nonimmune antitumor effects have not been reported in other immunomodulatory cytokines such as IL-12 and GM-CSF in an immunotherapeutic approach for treating lung cancers (36).
IL-27 has been reported to activate the JAK-STAT pathway in immune cells, endothelial cells, and melanoma cells (3, 4, 5, 6). In addition, IL-27-induced Th1 differentiation is dependent on activation of the MAPK pathway (37). Previous studies demonstrate that activation of STAT1 and inactivation of MAPK and PI3K/Akt pathways participate in the reduction of COX-2 expression (34, 35, 38, 39). In this study, treatment of rIL-27 enhances the activation of STAT1 as well as the reduction of phosphorylation of ERK1/2 and NF-κB (supplemental Fig. 1), which could result in decreased levels of COX-2 expression. The molecular mechanism underlying IL-27-induced COX-2 suppression merits further investigation.
IL-27 is well known to stimulate Th1 differentiation and IFN-γ production of naive T cells, leading to the generation of specific immune responses against cancer cells (6, 7, 8, 9). Studies show that PGE2 down-regulates the Th1 cytokines (TNF-α, IFN-γ, and IL-12) and up-regulates the Th2 cytokines (IL-4, IL-10, and IL-6) of immune cells, thus helping cancers evade immunity (21, 35). Interestingly, we can detect higher IL-12 and lower IL-10 expression levels in cocultured immature dendritic cells with the supernatants of LLC-1/scIL-27 cells (data not shown). The effects are consistent with the in vivo finding of increased Th1 cytokine production (IL-12) as well as reduced Th2 cytokine secretion (IL-10) in splenocytes from mice injected with LLC-1/scIL-27 cancer cells (Fig. 2⇑, B and C). However, significant changes in IL-12 and IL-10 secretion are not found between rIL-27-treated and -untreated dendritic cells (data not shown). Thus, down-regulation of PGE2 in lung cancers may also play a role in IL-27-mediated antitumor immunity.
Although TGF-β1 is a potent inhibitor of epithelial cell proliferation in the early phases of tumorigenesis, evidence shows that increased secretion of TGF-β1 by tumors is involved in the enhancement of cancer progression in the late phase (40). The functional alteration of TGF-β1 is due to changes in the degree of tumor differentiation and sensitivity to TGF-β1 (41). High levels of TGF-β1 are produced by many types of tumors and are associated with poor prognosis in cancer patients (32). Large amounts of TGF-β1 affect the proliferation and differentiation of cells as well as migration, invasion, angiogenesis, and immune surveillance (41). We first demonstrated that overexpression and stimulation of IL-27 inhibit the expression of TGF-β1 (Figs. 4⇑B and 5⇑C). Although a change in the expressions of COX-2 and PGE2 takes place earlier than the secretion of TGF-β1 (Fig. 5⇑, C and D), rIL-27-regulated TGF-β1 secretion is not dependent on PGE2 (Fig. 5⇑E). In addition, TGF-β1 treatment does not increase abilities of cellular migration and invasion in lung cancer cells (Fig. 6⇑, B and C). Therefore, decreased levels of TGF-β1 by IL-27 may be involved in the up-regulation of antitumor immunity, but not in changes of E-cadherin and vimentin in LLC. However, the role of TGF-β1 in lung cancer tumorigenicity merits further investigation.
It has been demonstrated that IL-12 promotes Th1 cellular responses and CTL activities against tumors (42). In light of inducing Th1 responses, IL-27 can be considered a better candidate anticancer agent. It has been shown that IL-27 not only augments the expression of Th1-specific factors (T-bet and IL-12Rβ2) and activities of CTLs, but also down-regulates expression of the Th2-specific factor GATA-3 in vitro and in vivo (4). Studies show that the antitumor effects by IL-27 are not dependent on IL-12-mediated signaling (43). IL-12 can cause severe cytotoxicity in vivo such as splenomegaly and liver injury due to the prolonged induction and maintenance phases of Th1 effectors (42, 44). In contrast to IL-12, IL-27 acts on naive CD4+ T cells and regulates only the initiation phase of Th1 responses. Thus, IL-27-treated mice are not observed with any apparent adverse effects (8, 45).
Immune-based therapies are shown to be successful for many cancers. However, lung cancer has not been considered an immune-sensitive malignancy. Available studies show that specific immune responses against lung cancer can be triggered in certain contexts (1). Our findings suggest that IL-27 may be an excellent molecule for immunotherapy of lung cancer. In addition to the regulatory effects on immune cells, IL-27 is shown to directly inhibit expression of the tumor-promoting genes COX-2, PGE2, and vimentin by lung cancer cells, along with reduced abilities of cancer migration and invasion. Taken together, IL-27 may be an attractive candidate applicable to lung cancer treatment.
We thank Dr. Chi-Ying F. Huang (Institute of Clinical Medicine, National Yang-Ming University, Taipei, Republic of China) for providing the CL1-0 and CL1-5 cell lines.
The authors have no financial conflict of interest.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
↵1 The work was supported by grants from the National Science Council (NSC 95-2320-B-010-022-MY3 and NSC97-2811-B010-014), the Veterans General Hospital University System of Taiwan (VGHUST) Joint Research Program, Tsou’s Foundation (VGHUST 97-P6-31), Taipei City Hospital, and the Ministry of Education (Aim for the Top University Plan), Republic of China.
↵2 Address correspondence and reprint requests to Dr Kuang-Hui Sun, Department of Biotechnology and Laboratory Science in Medicine, National Yang-Ming University, 155 Section 2, Lie-Nong Street, Taipei, Taiwan 112, Republic of China. E-mail address: or Dr. Shye-Jye Tang, Institute of Bioscience and Biotechnology, National Taiwan Ocean University, Keelung, Taiwan 202, Republic of China. E-mail address:
↵3 Abbreviations used in this paper: COX, cyclooxygenase; EMT, epithelial-mesenchymal transition; LLC-1, Lewis lung carcinoma-1; NOD, nonobese diabetic; NSCLC, non-small cell lung cancer; RT-qPCR, RT-quantitative PCR; scIL-27, single-chain IL-27; VEGF, vascular endothelial growth factor.
↵4 The online version of this article contains supplemental material.
- Received April 22, 2009.
- Accepted August 5, 2009.
- Copyright © 2009 by The American Association of Immunologists, Inc.