Abnormal Interleukin 10Rα Expression Contributes to the Maintenance of Elevated Cyclooxygenase-2 in Non-Small Cell Lung Cancer Cells¹ Free

COX³ catalyzes the biosynthesis of PGs such as PGE₂ and thromboxane from free arachidonic acid (1). Two COX isoforms have been identified: a constitutively expressed enzyme COX-1 and an inducible isoenzyme COX-2. Overexpression of COX-2 has been detected in numerous tumors, including NSCLC (2, 3). However, the mechanisms that regulate the expression of COX-2 in lung cancer are not completely understood. Studies indicate that COX-2 expression is induced in cancer cells at different levels in response to growth factors, cytokines, tumor promoters, and mutational events, and several signaling pathways have been implicated (2, 4, 5, 6). Type 2 cytokines (IL-4, IL-10, and IL-13) have been reported to be negative regulators of COX-2 expression and PGs production in normal and tumor cells (7, 8, 9, 10). For example, IL-10 decreases induced COX-2 expression as well as PGE₂ synthesis in astrocytes, chorion trophoblast cells, and monocytes (8, 9, 10). In the lung tumor environment IL-10, which has been implicated in tumor-mediated immunosuppression, is the predominant type 2 cytokine. Indeed, IL-10 is produced by lung tumor cells themselves as well as by host immune cells such as monocytes and lymphocytes in response to tumor COX-2-dependent production of PGE₂ (3, 11, 12). Because tumor-derived PGE₂ triggers the synthesis of IL-10 that is known as a potent suppressor of COX-2, we investigated the ability of IL-10 to regulate COX-2 expression in NSCLC cells. Our results show that IL-10 does not regulate COX-2 in NSCLC because of lack of tumor cell surface expression of IL-10Rα. Deficiency of the IL-10-mediated COX-2 regulatory feedback loop in NSCLC cells may contribute to COX-2 overexpression and maintenance of high level PGE₂ in the lung cancer microenvironment.

Clonetics NHBE cells were obtained from BioWhittaker (Walkersville, MD). Human large cell lung carcinoma H460 was obtained from the National Cancer Institute (Bethesda, MD). Lung adenocarcinoma line A549 and THP-1 human monoblastic leukemia cell line were obtained from American Type Culture Collection (Manassas, VA). The human squamous cell carcinoma line RH2 was established in our laboratory (11). The tumor cells were grown at 37°C in an atmosphere of 5% CO₂ in RPMI 1640 (Mediatech, Inc., Herndon, VA) supplemented with 10% FBS (Gemini Bio-Products, Woodland, CA), 100 units/ml penicillin/streptomycin and 2 mm glutamine. The NHBE cells were grown in the same conditions in bronchial epithelium cell growth media (BioWhittaker) supplemented with 5.2 μg/ml bovine pituitary extract, 5 μg/ml insulin, 0.5 pg/ml human recombinant epidermal growth factor, 0.5 μg/ml hydrocortisone and epinephrine, 10 μg/ml transferrin, 0.1 pg/ml retinoic acid, 6.5 pg/ml triiodothyronine, 50 μg/ml gentamicin, and 50 pg/ml amphotericin-B.

For COX-2 assays, tumor cells (0.5 × 10⁶ NSCLC cells in a 6-well plate or 0.5 × 10⁶ THP-1 cells/ml in a 12-well plate) cultured in complete medium were pretreated with different amounts of IL-10 (Peprotech, Inc., Rocky Hill, NY) before stimulation with IL-1β (280 units/ml; BD Biosciences PharMingen, San Diego, CA) or LPS (1 μg/ml; Sigma Chemical Co., St. Louis, MO) for 24 h.

Total RNA was isolated from each cell type using RNeasy Mini Kit (Qiagen, Inc., Valencia, CA) according to the manufacturer’s instructions. Total RNA (1.5 μg) was reverse transcribed using 200 units of SuperScript II RNase H Reverse Transcriptase (Invitrogen Corp., Carlsbad, CA) following the manufacturer’s instructions. A total of 3.3 μl of the cDNA was amplified by PCR using TaqDNA Polymerase (Invitrogen Corp). Reactions were performed under the following conditions: IL-10Rα, 35 cycles of denaturation at 94°C for 30 s, annealing at 68°C for 30 s, and extension at 72°C for 20 s; and IL-10Rβ, 30 cycles of denaturation at 94°C for 30 s, annealing at 52°C for 30 s and extension at 72°C for 40 s. The following primers were used: IL-10Rα sense primer 5′-ATGCTGCCGTGCCTCGTAGTGC-3′; IL-10Rα antisense primer 5′-ACTCTGGCCCG GTAGCCATTGC-3′ as previously described (13); IL-10Rβ sense primer 5′-CAAGATAAATGCATGAATAC-3′; and IL-10Rβ antisense primer 5′-GAAAGGAGAAAAACAGAAG-3′. The β-actin gene expression was used as an internal control for the quality of RNA specimens used as templates and as a standard to evaluate IL-10R expression in the different cell lines.

Tumor cells were cultured for 24 h in different conditions, washed in PBS, and then lysed at 4°C for 20 min in lysis buffer [50 mm Tris-HCl (pH 7.4), 1% NP40, 0.25% sodium deoxycholate, 150 mm NaCl, 1 mm EDTA, 1 mm phenylmethylsulfonyl fluoride, and 1× complete protease inhibitor mixture (Roche Diagnostics Corp., Indianapolis, IN)]. The protein concentration in the cell lysates was determined using a bicinchoninic acid assay (Pierce Chemical Co., Rockford, IL). Cell lysate proteins were separated on a 7.5 or 10% SDS-PAGE and transferred on polyvinylidene difluoride membranes (Millipore Corp., Bedford, MA). The proteins were immunodetected with anti-human COX-2 monoclonal antibody (Cayman Chemical Company, Ann Arbor, MI) or antihuman IL10-Rα polyclonal antibody (Santa Cruz Biotechnology, Inc., Santa Cruz, CA). The membranes were developed using an enhanced chemiluminescence-plus detection system (Amersham Biosciences, Piscataway, NJ). Equal protein loading was confirmed by immunodetecting the membranes with anti-actin antibody. Ramos cell lysate (Santa Cruz Biotechnology, Inc.) was used as positive control to evaluate IL10-Rα protein expression.

The human COX-2 monoclonal EIA kit was obtained from IBL Co., Ltd. (Gunma, Japan), and COX-2 EIA was performed according to the manufacturer’s instructions. Briefly, COX-2 standard protein or 15–30 μg of total protein from tumor cells were added to each well of an EIA plate precoated with an anti-hCOX-2 monoclonal antibody. After 1 h of incubation at 37°C, the plate was washed seven times, and an horseradish peroxidase-conjugated anti-hCOX-2 polyclonal antibody was added to each well. After 30 min of incubation at 4°C, the plate was washed, and the 3,3′,5,5′-tetramethylbenzidine substrate buffer was added to each well. Reactions were stopped by adding sulfuric acid solution, and the absorbance was read at 450 nm in a microtiter plate spectrophotometer.

The PGE₂ monoclonal EIA was obtained from Cayman Chemical Company and performed as described previously (3).

Cells cultured in complete medium for 24 h were recovered, washed with PBS 1×, and incubated in PBS 1× + 2% FBS (fluorescence-activated cell sorting buffer) containing phycoerythrin-human IL-10Rα monoclonal antibody (BD Biosciences PharMingen) or human IL-10Rβ polyclonal antibody (R&D Systems) for 30 min at room temperature. When unlabelled antibodies were used, a second incubation with a FITC-conjugated (Fab′)₂ fragment rabbit antigoat IgG (Jackson Immunoresearch Laboratories, Inc., West Grove, PA) was performed. Matched isotype immunoglobulin were used in control samples. After the last washing step, cells were resuspended in 500 μl of fluorescence-activated cell sorting buffer containing propidium iodide (1 μg/ml; Sigma Chemical Co.) or 7-amino-actinomycin D (1 μg/ml; Calbiochem, San Diego, CA) for dead cell exclusion.

To detect intracellular IL-10Rα expression, cells were fixed and permeabilized with 0.2% Tween 20 prior the staining. Cells were analyzed by a FACScan using CellQuest software (Becton Dickinson, Mountain View, CA).

The unpaired two-tailed Student’s t test was used to compare differences in NSCLC COX-2 expression and P = 0.05 were considered significant.

To evaluate the ability of IL-10 to regulate COX-2 in NSCLC, we performed in vitro assays using three different NSCLC cell lines (A549, H460, and RH2) and the monocyte cell line THP-1 as a positive control. IL-1β was used to increase COX-2 expression in NSCLC cells, and THP-1 cells were stimulated with LPS as described previously (3, 10, 14).

As shown in Fig. 1,A, in the absence of stimuli, NSCLC and THP-1 cell lines produced either undetectable or low levels of COX-2. Addition of IL-10 alone did not affect constitutive COX-2 production in the different cell lines. As anticipated from previous reports, after stimulation with IL-1β or LPS, both NSCLC and THP-1 cells produced higher levels of COX-2 (3, 14). In THP-1 cells, treatment with IL-10 (≥10 ng/ml) markedly inhibited LPS-induced COX-2 production (Fig. 1,A), demonstrating that the regulatory effect of exogenous IL-10 was intact in this monocyte cell line. Consistent with these findings for COX-2 levels, IL-10 potently down-regulated PGE₂ production in THP-1 cells (Fig. 1,C). In contrast, addition of IL-10 did not decrease IL-1β-induced COX-2 overexpression in the three NSCLC cell lines as shown by Western blotting (Fig. 1,A). These results were supported by quantification of COX-2 production in NSCLC cell lines using ELISA, which demonstrated that IL-10 did not have any significant effect on either constitutive or IL-1β-induced COX-2 expression (Fig. 1,B). As shown in Fig. 1 C, PGE₂ levels in the culture medium of NSCLC remained unchanged when IL-10 was added, indicating that COX-2-dependent production of PGE₂ is not affected by IL-10. Taken together, these findings suggest that in contrast to monocytes, COX-2 expression in NSCLC cells is not decreased by exogenous IL-10.

Cellular responses to IL-10 are mediated through its receptor (IL-10R) composed of two cell surface subunits: an IL-10 binding chain IL-10Rα, which is abundantly expressed in hematopoietic cells, and an accessory subunit IL-10Rβ, constitutively expressed in most cells and tissues (9).

To assess whether unresponsiveness of NSCLC cells to IL-10 results from absence of expression of either one or both IL-10R subunits, total RNA was purified from the different cell lines, and IL-10Rα and IL-10Rβ mRNA expression profiles were analyzed by RT-PCR using β-actin as an internal control (Fig. 2,A). Both IL-10Rα and IL-10Rβ mRNAs were detected in THP-1 and NSCLC cell lines under our experimental conditions (Fig. 2 A). Similar results were obtained when IL-10R mRNA expression was analyzed in nine other NSCLC cell lines (data not shown), indicating that IL-10Rα and IL-10Rβ are expressed in all of the NSCLC cell lines tested. Whereas both THP-1 and NSCLC cells showed abundant IL-10Rβ mRNA expression, IL-10Rα mRNA expression was relatively low in NSCLC compared with the monocyte cell line.

To determine whether the protein synthesis of IL-10Rα correlates with its mRNA expression, cell lysates from THP-1 and NSCLC cell lines were prepared and analyzed by Western blot with an anti-human IL-10Rα polyclonal antibody. Consistent with the results obtained for mRNA expression by RT-PCR, we found that IL-10Rα protein was expressed in both monocyte and NSCLC cell lines (Fig. 2 B). Differences observed in the apparent molecular weight of IL-10Rα and immunodetection of several bands in the protein from different cell lines extracts may be because of heterogeneity of protein glycosylation.

Colocalization of both IL-10Rα and IL-10Rβ on the cell surface is required for cellular responses to IL-10 (9, 15). Thus, to gain insight into the mechanisms underlying the unresponsiveness of NSCLC cells to IL-10, cell localization of both IL-10R subunits was examined in NSCLC and THP-1 cell lines by flow cytometry. As shown in Fig. 3, whereas THP-1 cells displayed cell membrane expression of both IL-10R subunits, only IL-10Rβ was detected on the surface of NSCLC cells. Consistent with their unresponsiveness to IL-10, NSCLC displayed no surface expression of IL-10Rα (Fig. 3). However, the intracellular IL-10Rα protein was found in permeabilized cells (Fig. 3), supporting the results obtained by Western blot (Fig. 2 B). Additional analysis of IL-10R expression in three other NSCLC cell lines by flow cytometry confirmed the lack of cell surface IL-10Rα (data not shown).

To determine whether the deficit of IL-10Rα surface expression is present in nontransformed lung epithelial cells, we examined its expression in NHBE cells. Interestingly, IL-10Rα was detected both on the surface and intracellularly in NHBE cells (Fig. 4,A), whereas IL-10Rα surface expression in A549 remained undetectable when the cells were cultured in the enriched medium required for NHBE growth (Fig. 4 B).

Studies have shown that COX-2 is overexpressed in numerous epithelial derived tumors, including NSCLC (2, 3). Tumor COX-2 overexpression has been implicated in enhanced angiogenesis, apoptosis resistance, invasion, and suppression of antitumor immunity (16, 17, 18, 19). Thus, based on its importance in regulating the malignant phenotype, investigations have focused on delineating the mechanisms mediating tumor COX-2 overexpression.

A variety of compounds, growth factors, cytokines, or tumor promoters have been revealed to up-regulate COX-2 expression at the transcriptional level (2, 5). For example, these include transforming growth factor β, IL-1β, and epidermal growth factor, which are all highly represented in the lung tumor microenvironment. Mutational events have been also implicated to contribute to transcriptional up-regulation of COX-2 in cancer cells (4, 6). Moreover, basal transcription has been shown altered in murine lung cancer (20). Modulation of COX-2 protein levels can also be achieved via posttranscriptional mechanisms involving 3′ regulatory elements and translational effects regulating the rate of protein synthesis and/or degradation (2, 5, 6). As previously reported in certain type of cells, IL-1β-dependent regulation of COX-2 expression can involve both transcriptional mechanisms and posttranscriptional mRNA stability (21). On the other hand, type 2 cytokines, including IL-4, IL-10 and IL-13, have been reported to suppress either LPS- or IL-1β-induced COX-2 expression in various cell types, thereby inhibiting PGE₂ synthesis (7, 8, 9). Indeed, in normal cells such as monocytes and epithelial or dendritic cells, the COX-2-dependent production of PGE₂ potently up-regulates IL-10 production by both lymphocytes and macrophages within the inflammatory microenvironment (22, 23). Increased levels of IL-10, in turn, have the capacity to potently inhibit induced COX-2 expression and subsequent PGE₂ production in normal cells. This autocrine/paracrine regulatory loop in normal cells has been suggested to serve as an important checkpoint in the balance of PGE₂-mediated regulation of Th1/Th2 cytokine homeostasis in normal host immune responses (23). In the lung tumor environment, IL-10 is the predominant type 2 cytokine, expressed by either lung tumor cells or host immune cells, and this induction occurs, in part, through a COX-2-dependent PGE₂-mediated mechanisms (3, 11, 12). Moreover, clinical studies have reported that both IL-10 and COX-2 overexpression are correlated with poor prognosis of NSCLC (13, 24). In this study, we have shown that IL-10 effectively down-regulates both COX-2 and PGE₂ production in the THP-1 monocyte cell line consistent with previous studies in normal cells (8, 9, 10). In contrast, we found that IL-10 has no effect on either constitutive or induced COX-2 expression and COX-2-dependent PGE₂ production in NSCLC cells. This result indicates that NSCLC COX-2 expression is not inhibited by IL-10 in the regulatory feedback loop operative in normal cells. Thus, defective functioning of this homeostatic mechanism could result in maintenance of high-level COX-2 expression and PGE₂ synthesis as well as heightened IL-10 production in lung cancer. The persistent tumor COX-2 elevation may cause alteration of immune responses, enhanced angiogenesis, apoptosis resistance, and invasion, leading to the promotion of tumorigenesis.

Loss or decrease of receptor expression is a common mechanism associated with abrogation of response to certain ligand in human cancer (25, 26). IL-10 exerts its action through a heterodimeric membrane receptor formed by a binding subunit IL-10Rα and an accessory chain IL-10Rβ. Although IL-10Rβ is constitutively expressed in most cells and tissues, IL-10Rα expression is often inducible and largely restricted to hematopoietic cells. In this study, we found that all of the NSCLC cell lines tested express mRNA for both IL-10R subunits, suggesting that NSCLC unresponsiveness to IL-10 is not a consequence of a transcriptional defect inhibiting IL-10R mRNA expression. This finding regarding IL-10Rα mRNA expression in NSCLC cells is consistent with a recent report showing that >95% of NSCLC tissue samples expressed IL-10Rα at the RNA level (13). In contrast Naruke et al. (27), who analyzed xenografts of NSCLC by RT-PCR, only detected IL-10Rα mRNA in 18% of the samples. These apparently discrepant observations may be because of the conditions used to amplify IL-10Rα mRNA. Indeed, our studies showed that the number of RT-PCR cycles seems to be a critical parameter for IL-10Rα detection.

Because IL-10Rα mRNA expression was found to be relatively low in NSCLC cells compared with the monocyte cell line, we analyzed its expression at the protein level. Consistent with the results obtained for mRNA expression by RT-PCR, we found that IL-10Rα protein was expressed in both monocyte and NSCLC cell lines. However, qualitatively, our results indicate that there does not seem to be a direct correlation between IL-10Rα transcript level and protein generation in the different cell lines tested (Fig. 2). Potential explanations for these results include differences in posttranscriptional regulation (2, 5, 6).

Known mechanisms underlying cell resistance to ligand effect, in normal and pathological conditions, include synthesis of nonfunctional receptor and alteration of receptor trafficking. In particular, recent studies have shown that unresponsiveness to IL- 10 may be the consequence of either synthesis of an IL-10R-truncated protein in colon epithelial cells or reduced membrane IL-10Rα expression in mature dendritic cells (7, 28). In the current study, we found that IL-10Rβ was expressed on the surface membrane of NSCLC cells but IL-10Rα was not. The lack of IL-10Rα on the cell membrane was neither because of the cell growth conditions as a monolayer nor to the presence of serum (data not shown).

Cell surface expression of IL-10Rα is a critical factor in cellular response to IL-10, and its colocalization with IL-10Rβ on the plasma membrane localization is required for optimal IL-10 signaling (9, 15). Therefore, lack of cell surface IL-10Rα may explain the defect of COX-2 down-regulation by IL-10 in lung tumor cells. Detection of intracellular protein in NSCLC suggests that deficiency of membrane IL-10Rα expression is unlikely the result of a translation defect but may instead be secondary to posttranslational events. These posttranscriptional mechanisms may result in either alteration of receptor trafficking from intracellular vesicles to the membrane or synthesis of a soluble form of IL-10Rα. However, there have been no reports of detection of soluble IL-10Rα in vivo (9). In the case of deregulation of IL-10Rα trafficking, our preliminary studies suggest that it is neither because of the alteration of the cell polarity nor to the presence of any factors in the culture medium. Studies are currently under way to determine the mechanisms underlying alteration of IL-10Rα cell surface trafficking; these findings may be valuable from a diagnostic and therapeutic standpoint.

In contrast to our findings in lung cancer cells, we found that IL-10Rα is expressed on the surface of normal lung epithelial cells. Recent reports showed aberrant chemokine receptor expression and disruption of transforming growth factor β receptor, TβIIR, signaling in malignant hematopoietic cells compared with normal cells (25, 26). Therefore, the mechanism involved in membrane IL-10Rα deficiency may be attributed to the malignant transformation of the lung epithelial cells. Additional experiments are required to confirm this hypothesis.

In conclusion, our results provide evidence that NSCLC responsiveness to IL-10 may be regulated through modulation of cell surface IL-10Rα expression. This is the first report of abnormal IL-10R cell surface expression as a contributor to the maintenance of elevated COX-2 expression in human cancer. A more complete understanding of the regulatory pathways leading to the up-regulation and maintenance of tumor COX-2 may assist in developing therapeutics or chemoprevention targeting COX-2 in lung cancer.

We thank Dr. Felicita Baratelli for critical review of the manuscript.