Lung cancer is one of the leading causes of death, and lung adenocarcinoma incidence has been increasing. Current oncostatic treatments for lung cancer are ineffective, and new therapies are needed for both treatment and prevention.1 Patients with NSCLC may present symptoms such as cachexia, leukocytosis and hypercalcemia, known collectively as paraneoplastic syndromes. It can be hypothesized that these symptoms are mostly due to SIRs.
NSAIDs have been studied in both chemoprevention and treatment of tumors, with emphasis in colorectal cancer.2 The rationale for the use of NSAIDs in oncology was to inhibit the enzyme COX, which is the key step in the conversion of arachidonic acid to PGs. Constitutive COX-1 is responsible of physiologic PG levels, whereas inducible COX-2 is expressed upon stimulation and accounts for high PG levels. COX-2 is overexpressed in a number of human and murine tumors and cell lines.3, 4, 5 Inhibition of COX-2 PG synthesis is thought to account for the beneficial effects of NSAIDs. PGs, in particular PGE2, promote tumor growth. They stimulate the proliferation and/or inhibit the apoptosis of tumor cells,6, 7 increase the angiogenic response8, 9 and promote the release of MMPs and/or stimulate the migration of tumor cells, increasing their invasive ability and spread in metastases.10, 11 Moreover, PGs can alter the balance between cytokines, leading to immune suppression.12 Despite this, NSAID treatment could be beneficial by mechanisms that are independent of PG synthesis inhibition13, 14 and still could exert antitumoral action by COX-independent mechanisms.15
Both indo and cxb are NSAIDs with known chemopreventive and antitumor properties. Indo inhibits nonselectively both COX-1 and –2, while cxb is a selective COX-2 inhibitor. They reduce tumor development and metastatic dissemination.16, 17 Selective COX-2 inhibitors have proven gastrointestinal safety compared to classical NSAIDs. However, indo activates PPAR-γ, a member of the superfamily of steroid hormone nuclear receptors with antitumor properties.18
Factors released by the tumor may be involved in the development of paraneoplastic syndromes. Tumor growth with subsequent development of cachexia is mediated by complex interactions of cytokines and growth factors, partly communicated by eicosanoids. Inappropriate production of inflammatory cytokines such as TNF-α, IL-6 and IL-1β is involved in cancer cachexia.19, 20
We have been studying a murine lung adenocarcinoma, P07, that induces in the host leukocytosis, cachexia and hypercalcemia.21 We detected GM-CSF and IL-6 in tumor conditioned media and PTHrP in plasma samples of tumor-bearing mice.22 These factors might partly account for the above-mentioned syndromes.
Our study was designed to determine the effect of SIR inhibition by NSAIDs on LP07 lung adenocarcinoma growth, metastasis outcome and paraneoplastic syndrome development and to identify underlying mechanisms.
Male and female BALB/c mice, 2–4 months old and weight-matched, were obtained from our Animal Care Division. Mice were housed under a 12:12 hr light–dark cycle with access to water and rodent chow ad libitum.
Indo was obtained from Sigma (St. Louis, MO). Cxb was a gift from Microsules & Bernabó (Buenos Aires, Argentina). Drugs were dissolved in DMSO and then diluted in medium for in vitro experiments. The final concentration of DMSO was maintained at 0.1%. For in vivo experiments, indo was dissolved in 100% ethanol (5 mg/ml) and then given to mice at a concentration of 10 μg/ml in drinking water. Cxb was mixed with the chow formula at a concentration of 1,500 ppm.
The LP07 cell line was obtained in vitro from an s.c. passage of the original spontaneous P07 tumor.23 P07 and LP07 cells share almost the same in vivo behavior with respect to paraneoplastic syndrome development and lung metastatic incidence. LP07 cells (2 × 105) were s.c.-inoculated in the right flank of female BALB/c mice and measured twice a week in 2 perpendicular diameters using a Vernier caliper. Tumor volume (mm3) was calculated as (a × b2)/2, where a and b are the largest and smallest diameters, respectively. Lung surface metastases were counted at autopsy. Body weight was recorded immediately before both LP07 cell inoculation and autopsy. Excised tumors were also weighed. Weight loss was evaluated as follows:
LP07 cell culture conditions
Cells were cultured in MEM supplemented with 10% FCS, 2 mM L-glutamine and 80 μg/ml gentamycin at 37°C in a humidified 5% CO2–air atmosphere.
Blood cell counts
Blood cell counts were performed on day 30 after s.c. tumor cell implantation. Blood was obtained from the retroorbital sinus of anesthetized mice. Leukocyte concentration was determined using a Coulter (Hialeah, FL) counter.
Effect of NSAIDs on tumor cell viability
Cells were seeded in 96-well, flat-bottomed plates at a density of 104/100 μl and allowed to attach overnight. Cells were then incubated with indo (25 and 50 μM), cxb (10 and 25 μM) or DMSO 0.1% in 100 μl of MEM supplemented with 10% FCS for 24 hr at 37°C in a 5% CO2 humidified atmosphere. Cell viability was determined by the Cell Titer 96 Aqueous Non-Radioactive Cell Proliferation Assay (Promega, Madison, WI).
Serum levels of IL-6 and IL-1β were measured by ELISA, using commercial kits (Endogen, Boston, MA) according to the manufacturer's specifications.
In vitro migration
Wounds of 300 μm width were made in LP07 subconfluent monolayers and the cells allowed to migrate into the cell-free area for 24 hr. Migration ability was evaluated in the presence of indo or cxb (10 μM), DMSO 0.1% or 1% FCS-supplemented medium. Cell migration was quantified by calculating the percentage of the area occupied by the migratory cells in the original cell-free wounded area. Three wounds per well (3 wells/treatment) were recorded.
Transwell (Corning, Corning, NY) cell culture chambers were used for invasion assays. Filters with 8 μm pores were previously coated with 0.1% gelatin on the lower side and with Matrigel (250 μg/ml; Becton Dickinson, Bedford, MA) on the upper side. Then, 2 × 105 cells in 0.1 ml MEM supplemented with 1% FCS and 0.1% BSA plus indo or cxb (10 μM) were seeded on the upper face of the chamber. The lower chamber contained lung conditioned media as chemoattractant and the same drugs. Cells were incubated for 24 hr at 37°C in a humidified 5% CO2 atmosphere. Membranes were fixed in Carnoy's solution and stained with Hoechst 33258 (Sigma); nuclei were counted in ×400 fields under a fluorescence microscope (Eclipse E400; Nikon, Tokyo, Japan). Data were expressed as percentages of control cells without treatment. Assays were done in triplicate.
Detection and quantification of MMP-9 activity by zymography
Enzymatic activity was determined on substrate-impregnated gels. Briefly, samples were collected and separated on 9% SDS polyacrylamide gels containing 1 mg/ml of gelatin (Sigma), under nonreducing conditions. After electrophoresis, gels were washed for 30 min in 2.5% Triton X-100, incubated for 48 hr at 37°C in 0.25 M TRIS-HCl, 1 M NaCl and 25 mM CaCl2 (pH 7.4). For nonspecific activity detection, gels were incubated in the same buffer solution with 40 mM EDTA. Gels were fixed and stained with 0.5% Coomassie brilliant blue G-250 (Bio-Rad, Richmond, CA) in methanol/acetic acid/H2O. Activity bands were visualized by negative staining. Gelatinolytic bands were measured with a digital densitometer (GS-700, Bio-Rad). Data were expressed as arbitrary units and normalized regarding control cell values.
Angiogenesis assay was performed as previously described.24 Briefly, LP07 cells were cultured in vitro with 25 μM indo, 10 μM cxb or 0.1% DMSO for 24 hr. LP07 cells (5 × 104/0.1ml MEM) were inoculated i.d. in both flanks of male BALB/c mice (n = 5/group). Controls received vehicle or medium alone. After 5 days, mice were killed. The inner skin was exposed and placed under a dissecting microscope at ×64 magnification. Vascularization (vessels/mm2 of skin) was quantified by measuring the vessel density as described previously.
All experiments were performed in triplicate. Data were expressed as means ± SEM. Statistical comparisons were performed by one-way ANOVA with Tukey's post-tests using Prism v3.00 for Windows (GraphPad Software, San Diego, CA).
Two-way ANOVA and repeated-measures 2-way ANOVA followed by Bonferroni's post-tests were done when indicated. The level of statistical significance for all tests was p < 0.05.
Growth of primary tumor and metastases
To study the effect of NSAID therapy on primary tumor and metastasis development, mice were given indo (10 μg/ml) or cxb (1,500 ppm) beginning the day of tumor inoculation.
Treatment with either indo or cxb reduced the growth of LP07 s.c. tumors when overall growth curves were analyzed by repeated-measures ANOVA (Fig. 1a).
Lung metastases, evaluated after 30 days of s.c. tumor growth, were also reduced in number and size by treatment with indo or cxb (Fig. 1b,c).
Development of paraneoplastic syndromes
Total leukocyte counts were reduced in both indo- and cxb-treated LP07 TBM 30 days after tumor cell inoculation (Fig. 2a). Spleen weights were reduced in NSAID-treated compared to untreated mice (data not shown).
Indo and cxb did not reduce serum levels of ionic calcium (data not shown).
Cachexia, evaluated as weight loss, was significantly reduced in treated groups. Control mice lost about 10% of their body weight, while indo- and cxb-treated mice maintained or slightly gained weight (Fig. 2b).
Serum levels of IL-6 and IL-1β
The cytokines IL-6 and IL-1β are involved in the onset of an SIR that is thought to play a role in the development of cachexia. Levels of these 2 cytokines were determined in serum samples of LP07 TBM treated or not with indo (10 μg/ml).
Fifteen days after tumor cell inoculation, before the onset of cachexia, serum IL-6 levels were significantly reduced in indo-treated mice compared to untreated mice. After 30 days of tumor growth, when cachexia was established in the control group, no differences in IL-6 levels were observed (Fig. 3a).
Serum levels of IL-1β were significantly reduced in control LP07 TBM compared to indo-treated mice after 15 days of LP07 cell inoculation. As tumors continued growing, a pronounced increase in IL-1β levels in the control group was detected. IL-1β levels in indo-treated mice throughout tumor progression were closer to values in normal mice (Fig. 3b).
In vitro viability
Viability tests were performed on LP07 cells to study if inhibition of tumor growth and metastasis outcome could be due to a direct cytotoxic effect of the drugs on tumor cells.
Tumor cell viability was significantly reduced with 50 μM indo and 25 μM cxb. No differences were observed with lower concentrations of both drugs (Fig. 4).
Migration and invasion
Given the reduced number of lung metastases in NSAID-treated mice, migration and invasion assays were done with LP07 cells in vitro. Migration of LP07 cells to wounded areas in wound repair assays was significantly reduced in cxb- and indo-treated cultures (Fig. 5a).
Matrigel invasion assays also showed a decrease in the invasiveness of LP07 cells treated with either indo or cxb (both at 10 μM, Fig. 5b).
The activity of MMP-9, involved in the metastatic process, was determined in conditioned media from indo- or cxb-treated LP07 cells. The activity of MMP-9 (Fig. 5c) was reduced when cells were treated with either indo or cxb.
The effect of NSAIDs on LP07 tumor-driven angiogenesis was analyzed. The number of blood vessels was significantly smaller when cells were treated with either 25 μM indo or 10 μM cxb compared to the control DMSO group. It was closer to the number of vessels in mice inoculated with medium alone as a basal control (Fig. 6).
The murine LP07 adenocarcinoma is useful to study lung cancer since it shares some characteristics with NSCLCs. The most interesting feature of this cell line is that hosts develop paraneoplastic syndromes such as cachexia, leukocytosis and hypercalcemia when developing s.c. tumors.23
Cytokines such as IL-6 and GM-CSF, but not TNF-α, were detected in tumor conditioned media; and PTHrP was found in plasma samples of LP07 TBM. These factors are thought to play key roles in the development of such symptoms.22 Inflammatory responses, when they become systemic and sustained, may enhance tumor growth, metastasis outcome and paraneoplastic syndrome development.
NSAIDs have been widely studied in colorectal cancer prevention and treatment in animal models and human patients.2, 25 The bulk of information suggests beneficial effects of NSAIDs in cancer treatment. Most research has focused on specific COX-2 inhibition as it was shown to be the main isoform that is upregulated in a great number of cancers other than colorectal tumors.3, 4, 5
Our rationale for NSAID use was to treat tumor-associated syndromes given that inflammatory components (GM-CSF, IL-6, leukocytosis) are present along tumor progression.22 Moreover, the role of eicosanoids in the settlement and progression of cancer cachexia has been documented, making NSAID treatment a good approach to deal with this syndrome.19, 26
Treatment with both NSAIDs delayed LP07 s.c. tumor growth (Fig. 1a) and metastasis outcome (Fig. 1b,c). This result is in agreement with previous findings in murine lung tumor models and other malignancies,27, 28, 29 although no effect of cxb on the size and number of tumors was reported for a carcinogen-induced murine lung tumor.30 A reduction in total leukocyte counts (Fig. 2a) and inhibition of weight loss associated with LP07 tumor progression (Fig. 2b) were observed. The effect on weight loss was already described for NSAIDs in different tumor models and in patients with cancer-related cachexia.26, 31, 32
PGE2 produced by the tumor can stimulate the secretion of IL-6. Thus, by diminishing PGE2 synthesis with NSAID treatment, the contribution of IL-6 to the onset of cachexia could be prevented.
In our study, NSAID treatment reduced serum levels of IL-6 on day 15 after tumor inoculation, when cachexia begins to develop (Fig. 3a), but failed to keep down the production of this cytokine on day 30. However, IL-6 levels in treated mice did not reach the levels of untreated mice by day 15. Despite this fact, cachexia was delayed with NSAID treatment, suggesting that inhibition of IL-6 production by anti-inflammatory therapy at early stages of tumor progression could be beneficial in dealing with this paraneoplastic syndrome.
IL-1β is a cytokine linked to an SIR and cancer-related cachexia.20 Its serum levels in indo-treated mice remained nearly constant and close to levels in normal mice (Fig. 3b). Nontreated LP07 TBM instead showed an early fall in IL-1β levels as if an immune response could not be initiated. By day 30, IL-1β reached levels characteristic of chronic inflammation that exceeded those in indo-treated mice. IL-1β is involved not only in cachexia development but also in cytokine, chemokine and growth factor production. It negatively or positively affects tumor growth depending on its levels.33
Reduced levels of inflammatory cytokines in response to NSAID treatment may account for the delayed weight loss in LP07 TBM. This could implicate IL-1β and IL-6 in LP07-related cachexia and suggest that NSAID treatment slows down its onset.
The action of NSAIDs on the outcome of lung metastases (Fig. 1b) could be ascribed to inhibition of invasion, migration and angiogenesis. Indo and cxb reduced LP07 cell migration in wound repair assays (Fig. 5a). Inhibition of the migration of murine mammary tumor cells by indo or the COX-2-specific inhibitor NS39813 has been reported,13 but conversely, no effect on migration of NS398 or nonselective ibuprofen was found in prostate epithelial cell lines.10 However, invasion was inhibited by NSAID treatment in both reports. Our results showed that indo and cxb reduced LP07 cell invasiveness (Fig. 5b). Different mechanisms have been described for this inhibition; it was pointed out that NSAID treatment can affect motility of tumor cells or release of MMPs, mainly MMP-9. While indo and NS398 failed to reduce MMP-9 activity in conditioned media from mammary tumor cells,13 ibuprofen and NS398 significantly decreased the activity of pro-MMP-9 in conditioned media from prostate tumor cells.10 We found low MMP-9 activity in conditioned media from LP07 cells treated with 10 μM cxb (Fig. 5c). A concentration of indo higher than what is inhibitory of cell invasion was necessary to reduce MMP-9 activity. Reduced migration of tumor cells and MMP-9 activity could account for the inhibition of the invasive ability of LP07 cells.
Angiogenesis is a key process for primary tumors and metastases to grow beyond the limits set by nutrient availability and hypoxic conditions. Indo and cxb reduce angiogenesis by acting either on tumor or endothelial cells.13, 34 We observed a reduction in the number of blood vessels developed when cells were previously treated in vitro with indo or cxb (Fig. 6). MMP-9 activity is also necessary for endothelial cells to invade tumor stroma and reach nonvascularized areas.35 Inhibition of MMP-9 activity by NSAIDs could affect not only cell invasiveness but also tumor and metastasis growth through angiogenesis inhibition.
A direct action of NSAIDs on tumor cells was also observed as indo and cxb could reduce the viability of LP07 cells in vitro (Fig. 4). However higher doses than those necessary to observe inhibitory effects in other in vitro experiments were needed. In vivo interactions with other host factors are probably important for the growth of tumor cells.
These findings are relevant, taking into account that the inhibitory concentrations of indo and cxb in in vitro experiments are usually under the plasmatic levels that can be achieved for both drugs in mice. Reported plasmatic concentrations of approximately 17 and 18 μM were determined for 10 μg/ml indo in drinking water17 and cxb 1,000 ppm in food,30 respectively. These concentrations exceed the dose of 10 μM that was inhibitory in our experiments, in particular with cxb.
Despite their differences in COX-2 inhibition specificity, both indo and cxb yield similar outcomes, although these NSAIDs could not act on the same targets. COX-2-independent actions of NSAIDs have been described. The PPAR-γ receptors can be activated by indo and other NSAIDs.18 PPAR-γ is expressed in a number of tumors, and its activation was inhibitory of tumor growth, invasion and angiogenesis.36 Moreover, it is involved in adipose tissue differentiation and could be another target for NSAIDs implicated in the inhibition of weight loss.
Immunohistochemical analysis of s.c. growing LP07 tumors and Western blotting of LP07 cell lysates showed high expression of COX-2 and PPAR-γ (data not shown).
Our results showed that several processes involved in cancer progression could be modulated by NSAID treatment, providing an alternative to deal with lung cancer. We observed that migration, invasion and angiogenesis are affected in LP07 lung adenocarcinoma when indo and cxb were used; and they protected mice from the weight loss and leukocytosis associated with tumor progression. In lung cancer patients, additional NSAID therapy can further improve performance status due to cachexia inhibition, allowing an otherwise impossible more aggressive conventional therapy.
We thank Dr. M.A. Jasnis for reviewing this article.