Simultaneous targeting of 5-LOX-COX and ODC block NNK-induced lung adenoma progression to adenocarcinoma in A/J mice

Abstract

Lung cancer is the leading cause of cancer deaths worldwide. Targeting complementary pathways will achieve better treatment efficacy than a single agent high-dose strategy that could increase risk of side effects and tumor resistance. To target COX-2, 5-LOX, and ODC simultaneously, we tested the effects of a dual 5-LOX-COX inhibitor, licofelone, and an ODC inhibitor, DFMO, alone and in combination, on NNK-induced lung tumors in female A/J mice. Seven-week-old mice were treated with NNK (10 μmol/mouse, single dose, i.p.) and randomized to different treatment groups. Three weeks after injection, mice were fed control or experimental diets (DFMO 1500/3000 ppm, licofelone 200/400 ppm, or a low-dose combination of 1500 ppm DFMO and 200 ppm licofelone) for 17 or 34 weeks. Both agents significantly inhibited tumor formation in a dose-dependent manner. As anticipated more adenomas and adenocarcinomas were observed at 17 and 34 weeks, respectively. Importantly, low dose combination of DFMO and licofelone showed more pronounced effects at 17 or 34 weeks in inhibiting the total tumor formation (~60%, p < 0.0001) and adenocarcinoma (~65%, p < 0.0001) compared to individual high dose of DFMO (~44% and 46%, p < 0.0001) and licofelone (~48% and 55%, p < 0.0001). DFMO and combination-treated mice lung tumors exhibited modulated ODC pathway components (Oat, Oaz, SRM, SMS, and SAT, p < 0.05) along with decreased proliferation (PCNA, Cyclin D1 and Cyclin A) and increased expression of p53, p21 and p27 compared to mice fed control diet. Both DFMO and licofelone significantly inhibited tumor inflammatory markers. Our findings suggest that a low-dose combined treatment targeting inflammation and polyamine synthesis may provide effective chemoprevention.

Introduction

Lung cancer is the primary cause of cancer deaths in the US and worldwide: the 5-year survival rate is only 15% [1]. Eighty-five percent of cases of lung cancer can be attributed to tobacco smoking [1]. Tens of millions of smokers are at high risk for developing this malignancy. Although there has been a modest improvement in survival, the advances observed in other malignancies have not translated to lung cancer. The marked lethality of this disease is likely due to the high frequency of advanced stage at diagnosis and the disease’s inherent therapeutic resistance. With advances in biologically targeted therapeutic agents, treatment of advanced lung cancer should continue to improve [2], but early diagnosis and effective chemopreventive interventions are crucial. Thus, intervention strategies and effective chemopreventive approaches targeting early lesions, adenoma, and adenocarcinoma may be useful for high-risk ex-smokers or those who are diagnosed with lung adenomas [3].

Clinical and preclinical studies have demonstrated that eflornithine (DFMO) is a potential chemopreventive agent [4-7]. The rationale for the use of DFMO, an ornithine decarboxylase (ODC) inhibitor, as a cancer chemopreventive agent has been strengthened over the years because ODC is found to be activated in cancer, which is followed by an increase in concentrations of the polyamines putresine, spermine, and spermidine that are associated with tumor promotion and progression. ODC is also modulated by important genes of the polyamine biosynthesis pathway, such as arginase (Arg1), ornithine aminotransferase (Oat), ODC antizyme (Oaz), and spermidine/spermine N(1)-acetyltransferase (SAT1). A link between polyamine levels and tumor grade and stage is well established [8,9]. Some studies have reported that lung carcinomas have greatly elevated polyamine levels [10,11].

Several experimental studies have indicated that the licofelone, a dual competitive inhibitor of cyclooxygenases COX-1 and COX-2 and 5-lipoxygenase (5-LOX), has analgesic, anti-inflammatory, anti-asthmatic and anti-platelet effects, with greater gastrointestinal safety than traditional non-steroidal anti-inflammatory drugs (NSAIDs); [12-15]. Clinical studies reported that licofelone was as effective as celecoxib, a COX-2-specific inhibitor, but had better gastrointestinal tolerability [12,16]. We demonstrated that licofelone was more effective than celecoxib in suppressing colonic tumors in APC^min mice [17]. It is widely accepted that selectively blocking COX-2 will shift arachidonic acid metabolism towards the 5-LOX pathway, overproducing leukotrienes and leading to increased prothrombotic effects [14,18,19]. This might be a mechanism of action for licofelone’s low gastric damage and anti-arthritic activity [20,21]. Preclinical studies show that NSAIDs may inhibit lung cancer both in vitro and in vivo. However, these studies are yet to be translated to clinical practice.

From published data it is assumed that combination of NSAIDs and polyamine inhibitors due to their suppressive effects might lead to increased anti-carcinogenic results [22]. Further, combinations of these drugs could offer the potential of efficacy at much lower concentrations than would have been required if they were used independently. Most previous research was conducted in vitro, evaluating the effect of either DFMO or licofelone alone at high concentrations. To our knowledge, this is the first study evaluating the effect of a combination of DFMO and licofelone on progression from adenoma to adenocarcinoma in NNK-induced lung cancer in A/J mice.

Materials and methods

Chemicals, antibodies, and reagents

NNK (>99% purity) was synthesized and kindly provided by the laboratory of Dr. Shantu Amin (Department of Pharmacology, Penn State Hershey, PA). Sodium orthovandate, Triton X-100, EDTA, EGTA, Igepal CA-630 (I7771), and protease inhibitor cocktail (P8340) were purchased from Sigma Chemical Co. (St Louis, MO, USA). Sodium fluoride, β-glycerophosphate, and phenylmethanesulfonylfluoride were purchased from Alfa Aesar (MA, USA). Primary antibodies against p-p38, pAKT, and MMP-2 were obtained from Bioss Antibodies (Massachusetts, USA). Primary antibody against PCNA and anti-rabbit, -goat, and -mouse horseradish-peroxidase (HRP)-conjugated secondary antibodies were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies against Cyclin D1 and NFκB were purchased from Cell Signaling Technology (Beverly, MA, USA). Histostain®-Plus 3rd Gen IHC Detection Kits were purchased from Life Technologies (NY, USA).

Animals, diet, and care

All animal experiments were performed in accordance with National Institute of Health (NIH) guidelines and the approval of the University of Oklahoma Health Sciences Center Institutional Animal Care and Use Committee. Six-week-old female A/J mice were obtained from Jackson Laboratory (Bar Harbor, ME). Mice were housed in ventilated cages under standardized conditions (21°C, 60% humidity, 12-h light/12-h dark cycle, 20 air changes per hour) in the pathogen-free University Rodent Barrier Facility. Semi-purified modified AIN-76A diet ingredients were purchased from Bioserv, Inc. (Frenchtown, NJ) and were stored at 4°C before preparation. Diets were prepared based on the modified American Institute of Nutrition (AIN)-76A diet as described earlier [23]. DFMO and licofelone were kindly supplied by the National Cancer Institute, Division of Cancer Prevention Repository (Bethesda, MD). Test agents were premixed with a small quantity of casein and were then blended into bulk diet using a Hobart Mixer. Both control and experimental diets were prepared weekly and stored in a cold room. Mice were allowed ad-libitum access to the diets and automated tap water purified by reverse osmosis; and checked routinely for signs of weight loss, toxicity, or any abnormalities. Food cups were replenished with fresh diet twice weekly during experimental period.

Experimental bioassay

The experiment was designed to evaluate the effects of short-term (during the progression of adenomas/early adenocarcinomas) and long-term (adenoma to adenocarcinomas) exposure to DFMO and licofelone on NNK-induced lung carcinogenesis in A/J mice. The experimental design is summarized in Figure 1. Female A/J mice were randomly categorized by weight into control and experimental groups. At 8 weeks of age, the mice intended for carcinogen treatment received a single intraperitoneal dose of 10 μmol NNK. Three weeks after NNK injection, groups of mice (25/group) were fed control AIN-76A or experimental diets containing DFMO (1500 or 3000 ppm), licofelone (200 or 400 ppm), or a combination of DFMO (1500 ppm) and licofelone (200 ppm), until study termination. Mice were euthanized by CO₂ asphyxiation after 17 weeks (10 mice/group) or 34 weeks (15 mice/group) of exposure to the test agents. All mice were weighed once every two weeks until termination of the study. After euthanasia, lungs were lavaged, perfused, and either snap-frozen in liquid nitrogen for protein and mRNA analysis or fixed in phosphate-buffered formalin, transferred within 2 days to 70% alcohol, and evaluated under a dissecting microscope for the number of tumors and tumor size. Tumors on the lung surface were counted under a dissecting microscope by at least two experienced readers who were blinded to sample identifiers. Tumor diameters were measured using Fisher brand digital callipers.

Figure 1.

Experimental design for evaluating the chemopreventive effect of DFMO, licofelone, and the combination in NNK-induced lung cancer in the A/J mouse model. Six-week-old mice were injected intraperitoneally with NNK (10 μmol/mouse). Mice were maintained on AIN-76A diet for 3 weeks. Three weeks after NNK injection, mice were fed with control diet (AIN-76A) or experimental diets containing 1500 or 3000 ppm of DFMO, 200 or 400 ppm licofelone, or a low-dose combination of DFMO and licofelone. Treatment continued until the end of the study. At 17 weeks (10 mice/group) and at 34 weeks (15 mice/group) after NNK treatment excluding 3 weeks of wash out period after NNK treatment, mice were killed and lungs were harvested for evaluation of lung tumors. Detailed information is provided in the materials and methods section.

Tumor histology

Fixed lung samples were embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). H & E-stained lung sections from three predetermined depths were evaluated in a blinded manner by a board-certified pathologist to obtain the number of adenomas and adenocarcinomas. Tumors were categorized according to the criteria of the Mouse Models of Human Cancers Consortium [24].

Protein immunoblotting

Lungs were frozen immediately in liquid nitrogen and stored at -80°C for further analysis. For marker analysis, tumors were excised from the lungs and total cell lysates were prepared using a previously described cell fractionation procedure [25]. Proteins were resolved on 8-12% SDS-PAGE, were transferred to a nitrocellulose membrane, and were probed with specific antibodies overnight at 4°C. The membranes were washed three times with Tris-buffered saline (TBS; pH 7.4) or phosphate-buffered saline (PBS; pH 7.4) for 15 min. Then, membranes were incubated with anti-rabbit or anti-goat or anti-mouse horseradish peroxidase-conjugated secondary antibody (1:5000, 2.5% skimmed milk in TTBS) and visualized using chemiluminescence reagent from Thermo Scientific (USA), followed by autoradiography. β-actin was used as the loading control. Densitometry of various analyte proteins and their respective loading controls from the same blot was performed using ImageJ 1.43 (NIH) software. Relative optical density was calculated by dividing the densitometry of analyte(s) protein with the respective loading control.

Immunohistochemical staining

For immunohistochemical (IHC) staining, normal and tumor tissues were fixed in 10% buffered formalin and expression was evaluated in adenocarcinomas by immunohistochemistry (IHC) as described in our previous paper [26]. Briefly, formalin-fixed, paraffin-embedded, 5-μm-thick tissue sections were mounted on white frosted, positive-charged (Denville Scientific, USA) glass slides. Sections were deparaffinised with xylene, rehydrated through a gradient series of alcohol, and washed in PBS. Antigen retrieval was carried out by heating sections in citrate buffer (10 mM, pH 6) for 30 minutes in a boiling water bath. Detection was conducted using a Histostain^®-Plus 3rd Gen IHC Detection Kit (Life Technologies, USA). Endogenous peroxidase activity was quenched by incubating the sections with peroxidase quenching solution for 30 min in darkness. Nonspecific binding sites were blocked with blocking solution for 1 h. Sections were then incubated with specific antibodies overnight at 4°C. The next day, the slides were washed with PBS and incubated with biotinylated secondary antibody for 1 h in a humidified chamber. After three 5-minute washes, the slides were incubated with streptavidin-peroxidase conjugate. After rinsing with PBS, diaminobenzidine was employed as the chromogenic substrate. The slides were counterstained with Mayor’s hematoxylin. For negative or isotype controls, the primary antibody was replaced with TBS or the respective antibody serum (used at the appropriate concentration). Slides were observed under an Olympus microscope IX71. Digital computer images were recorded with an Olympus DP70 camera.

Immunofluorescence staining

Formalin-fixed, paraffin-embedded, 5-μm-thick tissue sections were used to examine protein expression, as described elsewhere [27]. After deparaffinization, rehydration, antigen unmasking, and endogenous peroxidase activity blocking, sections were treated with freshly prepared 1% sodium borohydrate in distilled water for 30 min at room temperature. Sections were subsequently incubated with 1% bovine serum albumin (BSA) in PBS for 1 h to prevent nonspecific binding. After incubation with primary antibody, sections were incubated for 1 h with anti-rabbit or anti-goat Alexa Fluor^® 488 (Life Technologies) secondary antibodies. Slides were washed three times for 10 min with PBST and were then incubated with 4’, 6’-diamidino-2-phenylindole (DAPI; 100 ng/ml) for 20 min. The slides were mounted with SlowFade Gold antifade reagent (Life Technologies) with coverslips and were sealed with nail enamel. Slides were observed under an Olympus microscope IX71. Digital computer images were recorded with an Olympus DP70 camera.

Quantitative real-time PCR analysis

Total RNA from normal and tumor samples was extracted using TRizol reagent for total cellular RNA (Invitrogen, Grand Island, NY) per the manufacturer’s instructions. Equal amounts of DNA-free RNA were used for reverse transcription (RT) reactions to make cDNA using an iScript cDNA synthesis kit (Bio-Rad) per the manufacturer’s protocol. Real-time PCR was carried out in a 12-µl reaction volume containing 5 µl of diluted cDNA (50 ng) and FastStart Universal SYBR Green master (Roche) and primers (Invitrogen; Supplementary Table 1). All PCRs were performed in a Bio-Rad iCycler iQTM5 real-time PCR detection system. The fluorescence threshold values (Ct) were calculated. Relative mRNA levels were assessed by standardization to actin or GAPDH. Results are expressed as a relative fold difference in gene expression compared with control. Relative gene expression was calculated using the 2-ΔΔCT formula [28]. Specific primers for different genes were designed (see Supplementary Table 1). PCR conditions were as follows: denaturation at 94°C for 10 minutes, followed by 40 cycles at 95°C for 30 seconds, 60°C for 30 seconds, and 72°C for 45 seconds. All experiments were performed in triplicate.

Statistical analysis

Differences in body weights among groups were analyzed using analysis of variance. Adenoma and adenocarcinoma multiplicities (number of tumors/mouse) are expressed as means ± SEM. Protein expression and proliferative indices are expressed as means ± SEM and were analyzed by unpaired t-test with Welch’s correction. Dose-response effects were analyzed by linear regression analysis. Differences were considered statistically significant at p < 0.05. All statistical analysis was performed using Graphpad Prism 6.0.

Results

General observations

Dose selection was based on previously published toxicity studies. DFMO doses of up to ~5000 ppm and licofelone doses of up to ~600 ppm were tolerated in rodent models when administered with AIN-76A diet for 6 weeks [29]. In the present study, we applied 30% (1500 ppm) and 60% (3000 ppm) of the maximum tolerated dose (MTD) of DFMO and 30% (200 ppm) and 60% (400 ppm) of the MTD of licofelone to assess chemopreventive efficacy. We also administered a low-dose combination of 1500 ppm DFMO and 200 ppm licofelone to evaluate the efficacy of this combination. No significant changes in body weight or any other histologic toxicity in major organ sites were observed for either drug at any dose or time course (data not shown). The drugs did not induce any overt toxicities when administered through diet.

The combination of DFMO and licofelone significantly inhibits lung adenocarcinoma and delays the progression of adenoma to adenocarcinoma

NNK-treated mice fed on control diet developed 100% tumor incidence at 17 or 34 weeks (excluding wash-out period of 3 weeks after the carcinogen treatment, Figure 2A and 2B). Figure 2 shows the effects of both drugs and the combination on lung tumors, lung adenoma, and lung adenocarcinoma multiplicities at 17 and 34 weeks after treatment. Control-diet-fed mice that were killed at 17 weeks formed an average of 15.2 ± 1.29 lung tumors, while those killed at 34 weeks formed an average of 18.4 ± 1.4 lung tumors (Figure 2A and 2B). Administration of 1500 ppm and 3000 ppm of DFMO significantly suppressed lung tumor formation by 25% (p < 0.0001) and 32% (p < 0.001) at 17 weeks post-administration and by 30% (p = 0.0012) and 44% (p < 0.0001) at 34 weeks post-administration (Figure 2A and 2B). Treatment with 200 ppm and 400 ppm of licofelone significantly suppressed lung tumor formation by 17% (p < 0.0001) and 30% (p < 0.0001) at 17 weeks post-treatment and 32% (p = 0.0008) and 48% (p < 0.0001) at 34 weeks post-treatment (Figure 2A and 2B). The low-dose combination of DFMO and licofelone significantly suppressed lung tumor formation by 51% (p < 0.0001) at 17 weeks post-treatment and 60% (p < 0.0001) at 34 weeks post-treatment (Figure 2A and 2B).

Figure 2.

Effect of dietary feeding of DFMO, licofelone, and the combination on NNK-induced lung tumor growth in A/J mice. A. Lung tumors from different groups at the 17-week post diet treatment stage. B. Lung tumors from different groups at the 34-week post diet treatment stage. Effect of DFMO, licofelone, and the combination on NNK-induced lung adenoma in A/J mice. C. Lung adenoma from different groups at the 17-week post diet treatment stage. D. Lung adenoma from different groups at the 34-week post diet treatment stage. Values are means ± SE with N = 10. Effect of DFMO, licofelone, and the combination on NNK-induced lung adenocarcinoma in A/J mice. E. Lung adenocarcinoma from different groups at the 17-week post diet treatment stage. F. Lung adenocarcinoma from different groups at the 34-week post diet treatment stage. Values are means ± SE with N = 15. Differences between control and treatment groups were analyzed by one-tailed t-test with Welch correction and 95% confidence interval. G. Representative photomicrograph of hematoxylin and eosin (H&E) staining of lung tumor sections from mice in different experimental groups.

Figure 2G displays representative H&E staining of lung tumors from mice fed with control diet, DFMO, licofelone, or the combination. Tumors were classified as adenoma or adenocarcinoma, as described earlier [24]. Adenocarcinomas in control-diet-fed mice had varying degrees of differentiation and were characterized as the complete loss of normal alveolar architecture compared with adenocarcinomas seen in mice fed with DFMO, licofelone, or the combination (Figure 2G). Our observations suggest that a low-dose combination of DFMO and licofelone delays the adenoma-to-adenocarcinoma progression. We observed a reduced amount of cytoplasm and a small degree of pleomorphism without loss of alveolar architecture, thus retaining the adenoma structure in combination-treated mice (Figure 2G) at 34 weeks, compared with controls and mice treated with either drug alone.

Further histopathologic analysis showed that control-diet-fed mice had 9.4 ± 0.7 (means ± SEM) adenomas and 5.8 ± 0.5 lung adenocarcinomas, reflecting 61.8% of adenomas and 38.2% of adenocarcinomas, at 17 weeks post diet treatment excluding 3 weeks of wash out period after NNK treatment (Figure 2C and 2E). Control-diet-fed mice had only 2.2 ± 0.2 adenomas (11.9%) and 16.2 ± 0.7 adenocarcinomas (88.1%) at 34 weeks post diet treatment excluding 3 weeks of wash out period after NNK treatment weeks, reflecting a significant increase in the progression of adenomas to adenocarcinomas (Figure 2D and 2F). Administration of 1500 ppm and 3000 ppm of DFMO significantly suppressed NNK-induced lung adenoma by 23% (p < 0.001) and 42% (p < 0.0001) and suppressed lung adenocarcinoma by 27% (p < 0.01) and 44% (p = 0.0002), respectively, at 17 weeks (Figure 2C and 2E). However, at 34 weeks, the suppression was 13% (not significant) and 31% (p < 0.002) in lung adenoma and 33% (p < 0.0001) and 46% (p < 0.0001) in lung adenocarcinoma (Figure 2D and 2F). At 17 weeks, 200 ppm and 400 ppm of licofelone significantly suppressed NNK-induced lung adenoma by 28% (p < 0.003) and 38% (p < 0.0004) and lung adenocarcinoma by 31% (p < 0.005) and 37% (p < 0.0001), respectively (Figure 2C and 2E). However, at 34 weeks post-treatment, mice treated with either dose of licofelone had slightly more lung adenomas (31% and 4%, not significant) and significantly fewer lung adenocarcinomas (40% and 55%, p < 0.0001), reflecting some delay in the progression from adenoma to adenocarcinoma (Figure 2D and 2F). The combination treatment significantly suppressed lung adenoma by 46% (p < 0.0001) and lung adenocarcinoma by 60% (p < 0.0001) at 17 weeks post-treatment (Figure 2C and 2E). At 34 weeks post-treatment, the suppression was 13% in lung adenoma and 64% (p < 0.0001) in lung adenocarcinoma, suggesting that the low-dose combination of DFMO and licofelone was more effective than were higher doses of DFMO or licofelone alone (Figure 2D and 2F).

A combination of DFMO and licofelone inhibits the expression of COX-2, 5-LOX, mPGES1, and FLAP

We used real-time PCR and immunohistochemistry to examine mRNA and protein levels of pro-inflammatory enzymes COX-2 and 5-LOX in lung tumors. DFMO was less effective in inhibiting the expression of COX-2 and 5-LOX at mRNA and protein levels compared to licofelone and the combined treatment (Figures 3 and 4). We observed a significant reduction in mRNA expression of COX-2, 5-LOX, mPGES1, and FLAP in tumors from mice fed with licofelone and the combined treatment compared with tumors from mice fed with control diet (Figure 3). Only COX-2 mRNA expression was significantly inhibited by DFMO (Figure 3). Further, the low-dose combination of DFMO and licofelone was found to be more effective in decreasing the immunohistochemical staining of COX-2 and 5-LOX than were controls and either drug alone (Figure 4).

Figure 3.

Effect of DFMO, licofelone, and the combination on ODC, 5-Lox, Cox-2, mPGES-1, and FLAP mRNA expression in NNK-induced lung tumor from mice in different treatment groups. Relative mRNA expression of ODC, 5-Lox, Cox-2, mPGES-1, and FLAP in RNA samples from different treatment groups as analyzed by real-time PCR and normalized to β-actin. Significance of differences between control and treatment groups was analyzed by one-tailed t-test with Welch correction and 95% confidence interval. Data with means ± SE. p ≤ 0.05 was considered statistically significant.

Figure 4.

Effect of DFMO, licofelone, and the combination on ODC, 5-Lox, and Cox-2 expression in NNK-induced lung tumor sections from mice in different treatment groups. Representative photomicrographs showing immunohistochemical detection of ODC, 5-Lox, and Cox-2 (magnification ×600; inset shows 400× magnification).

A low-dose combination of DFMO and licofelone modulates the ODC signaling in NNK-induced lung tumors

ODC catalyzes the first step in the polyamine biosynthetic pathway, the decarboxylation of ornithine to putrescine. ODC is modulated by other signaling molecules of the polyamine pathway, including Oat, Oaz, SRM, SMS, and SAT1; its activation can lead to tumor promotion and progression. The expression of ODC and its related signaling molecules was first analyzed in normal lung and tumor tissues using real-time quantitative PCR assays (qPCR). We observed that mRNA levels of ODC, SRM, and SMS significantly increased in lung tumor tissue compared with normal tissue. The ODC-inhibitory enzyme Oaz, the polyamine-degrading enzyme SAT1, and Oat were significantly downregulated in lung tumor tissue compared with normal lung (Supplementary Figure 1).

Further, ODC pathway signaling molecules were measured using IHC and qPCR in lung tumor tissues. Markedly reduced mRNA and protein expression of ODC was observed in combination-treated mice compared with single-drug-treated and control-diet-fed mice (Figures 3 and 4). mRNA expression levels of SRM and SMS were reduced and mRNA expression levels of Oat, Oaz, and SAT1 were increased in lung tumors from combination-treated mice compared with control-diet-fed mice (Figure 5).

Figure 5.

Effect of DFMO, licofelone, and the combination on the polyamine synthesis pathway. Relative mRNA expression of spermidine synthase (SRM), spermine synthase (SMS), spermidine N(1)-acetyl transferase (SAT), ornithine amino transferase (Oat), and antizyme (Oaz) in RNA samples from different treatment groups as analyzed by real-time PCR and normalized to β-actin. Differences between control and treatment groups were analyzed by one-tailed t-test with Welch correction and 95% confidence interval. Data represent means ± SE. p ≤ 0.05 was considered statistically significant.

A low-dose combination of DFMO and licofelone inhibits tumor cell proliferation and induces p21, p27, and p53

We used immunoblotting/IHC and qPCR techniques to examine protein and mRNA expression of Cyclin D1, PCNA, and Cyclin A1 in mouse lung tumor tissues. The low-dose combination of DFMO and licofelone significantly decreased PCNA and Cyclin D1 protein expression compared with either drug alone or controls (Figure 6A and 6C). The combination was more effective in decreasing the intensity of IHC staining of PCNA and Cyclin D1 than was either drug alone. We also observed similar alterations in mRNA levels of PCNA in tissues obtained from mice fed with the low-dose combination (Figure 6B). Further, levels of Cyclin A protein, a cell cycle regulator, were inhibited in lung tumors from mice receiving the low-dose combination compared with tumors from mice fed control diet (Figure 6C).

Figure 6.

Effect of DFMO, licofelone, and the combination on cell proliferation markers in NNK-induced lung tumors from mice in different treatment groups. A. Representative western blot of PCNA and Cyclin D1 measured in total cell protein of tumor. B. Relative mRNA expression of PCNA in RNA samples from different treatment groups as analyzed by real-time PCR and normalized to β-actin. C. Representative photomicrographs showing immunohistochemical detection of PCNA, Cyclin D1, and Cyclin A (magnification ×600; inset shows 400× magnification). Differences between control and treatment groups were analyzed by one-tailed t-test with Welch correction and 95% confidence interval. Data with means ± SE. p ≤ 0.05 was considered statistically significant.

Further, we measured p53, p21, and p27 protein/mRNA levels and observed a significant increase in their expression accompanying the low-dose combination of DFMO and licofelone (Figure 7). Taken together, these results suggest that the low-dose combination inhibits tumor progression, which is associated with reduced tumor cell proliferation. These findings are consistent with our in vivo tumor inhibition findings.

Figure 7.

Effect of DFMO, licofelone, and the combination on p53, p21, Bax, and p27. A. Representative photomicrographs showing immunohistochemical detection of p53 and p21 (magnification ×600; inset shows 400× magnification). B. Relative mRNA expression of p53, p21, Bax, and p27 in RNA samples from different treatment groups as analyzed by real-time PCR and normalized to β-actin. Differences between control and treatment groups were analyzed by one-tailed t-test with Welch correction and 95% confidence interval. Data represent means ± SE. p ≤ 0.05 was considered statistically significant.

Dietary administration of DFMO, licofelone, and the low-dose combination inhibits inflammatory cytokines and matrix metalloproteinases

Inflammation within the tumor microenvironment is correlated with increased invasiveness and poor prognosis. The cytokines interleukin-1 beta (IL-1β), interleukin-6 (IL-6), and tumor necrosis factor alpha (TNF-α) are critical mediators of the inflammatory response. Therefore, we evaluated the mRNA expression of IL-1β, IL-6, and TNF-α in control and drug-treated lung tumor tissues. Mice fed with DFMO, licofelone, and a low-dose combination displayed reduced mRNA levels of IL-1β, IL-6, and TNF-α (Figure 8A).

Figure 8.

Effect of DFMO, licofelone, and the combination on pro-inflammatory cytokine and MMPs. Relative mRNA expression of (A) IL-1β, IL-6, TNF-α, and (B) MMP-2, MMP-7, MMP-9 and MMP-13, in RNA samples from different treatment groups as analyzed by real-time PCR and normalized to β-actin. Differences between control and treatment groups were analyzed by one-tailed t-test with Welch correction and 95% confidence interval. Data represent means ± SE. p ≤ 0.05 was considered statistically significant.

Matrix metalloproteinases (MMPs) are involved in tissue remodelling and disease processes, including cancer [30]. In addition to modifying extracellular matrix proteins, MMPs are intimately involved in the regulation of the activities of cytokines and cytokine receptors. Further, the control and drug-treated lung tumor tissues were analyzed for mRNA expression of MMP-2, MMP-7, MMP-9, and MMP-13. Compared with controls, mice fed with DFMO, licofelone, and the low-dose combination had lower mRNA levels of MMP-2, MMP-7, MMP-9, and MMP-13, as measured by real-time PCR (Figure 8B). Further, we used IHC/IF to compare MMP-9 levels in control and drug-treated mice. Our results showed that MMP-9 was inhibited by the low-dose combination of DFMO and licofelone, compared with controls (Supplementary Figure 2).

Discussion

Increased polyamine synthesis and inflammation, which are risk factors for cancer progression in humans, have long been associated with intraepithelial neoplasia. Targeting polyamine metabolism with polyamine synthesis inhibitors or catabolism activators and inflammation with NSAIDs has been studied in many malignancies, including lung, colon, prostate, bladder, and skin cancers [4,13,14,31]. Recent literature provides a basis for developing a combined treatment of chemopreventive and chemotherapeutic agents. In recent studies, inhibiting multiple molecular targets by administering a combination of their specific inhibitors was shown to block the pathogenesis of highly invasive and metastatic cancers more effectively than did single-agent-based therapeutic strategies [29,32]. Inflammatory eicosanoids and leukotrienes derived from COX-2 and 5-LOX pathway positively influence lung tumor growth. Prominent over-expression of ODC, COX-2, and 5-LOX enzymes activity and polyamines has been associated with proliferation and progression of lung cancer; expression of these proteins is correlated with tumor grade and invasion in some reports [13,14,17,22,33]. It has been widely accepted that selectively blocking COX-2 will shift arachidonic acid metabolism towards the 5-LOX pathway, overproducing leukotrienes and leading to increased prothrombotic effects [14,18,19]. Targeting both tumor inflammatory molecules and polyamines synthesis may provide additive/synergistic antitumor effects. Hence, we chose licofelone as an anti-inflammatory agent due to its ability to simultaneously inhibit COX and 5-LOX (16) and to limit undesirable cardiovascular and gastrointestinal side effects, and selected DFMO as the ODC inhibitor to study their individual and combined effects on lung tumor progression.

The total number of tumors decreased by >51% and adenocarcinomas decreased by >60% in mice receiving treatment with the low-dose combination of DFMO and licofelone for 17 or 34 weeks. These reductions were even greater than those observed with administration of higher doses of DFMO (>32% in total tumors, >44% in adenocarcinomas) and licofelone (>30% in total tumors, >37% in adenocarcinomas) alone. These results are consistent with studies using individual DFMO or licofelone in different organ-site cancers [4,13,14]. Our findings suggest that a low-dose combination of DFMO and licofelone significantly inhibits the number of lung tumors, slowing the progression of NNK-induced lung lesions to adenoma and then to adenocarcinoma. There are limited reports on the evaluation of COX-2 inhibitors, and no reports on dual COX/5-LOX inhibitors used in combination with ODC inhibitors against an NNK-induced mouse model of lung cancer. This is the first report on the chemopreventive effects of DFMO and licofelone on NNK-induced lung carcinogenesis in the A/J mouse strain.

We observed that lung tumors from NNK-induced A/J mice have elevated levels of ODC, COX-2, and 5-LOX, compared to normal tissue. Mice fed with the low-dose combination of DFMO and licofelone showed reduced expression of ODC, COX-2, and 5-LOX at protein and mRNA levels compared with mice receiving higher doses of either drug alone. This finding suggests that the inflammation and polyamine pathways play significant roles in lung tumor progression. Thus, a combined treatment that inhibits polyamines and COX-2 and LOX-5 is a valid approach. mPGES-1 and FLAP share significant sequence homology and protein structure, having pathological roles in inflammatory diseases, and require glutathione (GSH) as an essential cofactor for their activity [34]. Our results showed that the low-dose combination of DFMO and licofelone also decreased the mRNA levels of mPGES-1 and FLAP, thereby reducing the inflammation.

We examined the mRNA expression of 1) Oaz, an enzyme that regulates polyamine synthesis by binding and inhibiting ODC, 2) Oat, an enzyme that regulates the conversion of ornithine into proline via Glutamate-γ-semialdehyde, and 3) SAT1, a rate-limiting enzyme in the catabolic pathway of polyamine metabolism. These constituents of the ODC pathway were significantly lower in lung tumors than in normal lung tissues. Inhibition of ODC and modulation of several components of polyamine biosynthesis were observed in lung tumors from mice fed with high doses of DFMO or licofelone alone. However, the low-dose combination of DFMO and licofelone led to a significant reduction in ODC, SRM, and SMS expression, and increased Oaz and Oat expression. These results are consistent with previous reports on polyamine biosynthesis in various cancers [14,35] and in lung cancer progression [22].

Uncontrolled proliferation is a hallmark of cellular transformation and is accompanied by the deregulation of proteins involved in the control of cell cycle checkpoints. Inflammation and polyamines are known to play a role in the transition of the cell cycle from G2/M to the G1 phase; this process is delayed by a reduction in polyamine levels [36]. Our results showed reduced PCNA, Cyclin A and Cyclin D1, and significantly increased p53 and p21, in mice fed with the low-dose combination of DFMO and licofelone compared with mice fed with a high dose of either drug alone. These findings are consistent with our earlier reports demonstrating that PCNA-positive cells and Cyclin D1 decreased after treatment with either DFMO or licofelone in colon, bladder, and pancreatic cancer [4,6,13,14].

Transcription of Cyclin A is indirectly regulated by the tumor suppressor protein p53. Activated p53 turns on several downstream pathways, including p21, leading to cell cycle arrest [37]. Our study also showed reduced Cyclin A and increased p53 and p21 in mice-fed with a low-dose combination of DFMO and licofelone. This finding was consistent with those of other researchers [4,13,14,38].

TNF is one of the main mediators of acute and chronic inflammation: it can induce IL1β and IL6 expression [39]. High circulating levels of IL-6 were reported to be associated with poor prognosis in patients with lung cancer [40]. However, IL-1β is a pluripotent cytokine that promotes angiogenesis, tumor growth, and metastasis in experimental models; its presence in some human cancers is associated with aggressive tumor biology [41]. Our results indicate that high doses of DFMO and licofelone alone and a low-dose combination of the two drugs significantly block the expression of IL-1β, IL-6, and TNF-α in NNK-induced lung tumors in A/J mice, as reported in earlier in vitro and in vivo studies [42].

Further, we observed the role of interleukins (ILs) in regulation of MMP. As in many immunological responses, cell type-specific activation of various MMPs is principally mediated by ILs, other cytokines, eicosanoids like prostaglandin E (PGE) and leukotriene B (LTB), and other inflammatory mediators [43]. Boileau et al. showed that licofelone dose dependently inhibited the IL-1β stimulated production and expression of MMP-13 [44]. Our results suggest that the low-dose combination of DFMO and licofelone inhibits the expression of MMP-2, -7, -9, and -13 in lung tumors, as was also reported in in vitro and in vivo studies and research into human cancers of different origins [5,44,45].

In summary, this is the first report of the chemopreventive effects of a low-dose combination of DFMO and licofelone on NNK-induced lung carcinogenesis in the A/J mouse strain. Both agents were more effective at inhibiting adenocarcinoma multiplicity and reducing the total number of tumors when mice were treated for 17 and 34 weeks. Our observations suggest that inflammation and polyamine synthesis modulating agents delay the adenoma to adenocarcinoma progression. Approximately 60% inhibition of adenomas and adenocarcinomas was observed in the mice fed with the low-dose combination of DFMO and licofelone, compared with the mice fed with high doses of either drug alone or control diet.

Disclosure of conflict of interest

None.

Abbreviations

Arg1

arginase

BSA

bovine serum albumin

HRP

horseradish peroxidase

IHC

immunohistochemical

Oat

ornithine amino transferase

Oaz

ornithine decarboxylase antizyme

PBS

phosphate-buffered saline

PVDF

polyvinylidene difluoride

qPCR

quantitative real-time

SAT

spermidine N(1)-acetyl transferase

SDS-PAGE

sodium dodecyl sulfate polyacrylamide gel electrophoresis

SMS

spermine synthase

SRM

spermidine synthase

TBS

tris-buffered saline

TBST

TBS containing 0.1% Tween-20