Chemoprevention of mouse lung and colon tumors by suberoylanilide hydroxamic acid and atorvastatin

Authors

  • Michael A. Pereira,

    Corresponding author
    1. College of Medicine, Division of Medical Oncology, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    • Comprehensive Cancer Center, The Ohio State University, 333 W. 10th Avenue, Columbus, OH 43210
    Search for more papers by this author
    • Tel.: +[614-222-0405], Fax: +[614-293-7710]

  • Blake M. Warner,

    1. College of Public Health, Division of Environmental Health Sciences, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    Search for more papers by this author
  • Thomas J. Knobloch,

    1. College of Medicine, Division of Medical Oncology, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    2. College of Public Health, Division of Environmental Health Sciences, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    Search for more papers by this author
  • Christopher M. Weghorst,

    1. College of Medicine, Division of Medical Oncology, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    2. College of Public Health, Division of Environmental Health Sciences, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    Search for more papers by this author
  • Ronald A. Lubet,

    1. Division of Cancer Prevention, National Cancer Institute, Rockville, MD
    Search for more papers by this author
  • Vernon E. Steele,

    1. Division of Cancer Prevention, National Cancer Institute, Rockville, MD
    Search for more papers by this author
  • Bruce C. Casto

    1. College of Medicine, Division of Medical Oncology, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    2. College of Public Health, Division of Environmental Health Sciences, and Comprehensive Cancer Center, The Ohio State University, Columbus, OH
    Search for more papers by this author

Abstract

Atorvastatin and suberoylanilide hydroxamic acid (SAHA) were evaluated for chemoprevention of mouse lung tumors. In Experiment 1, lung tumors were induced by vinyl carbamate in strain A/J mice followed by 500 mg/kg SAHA, 60 or 180 mg/kg atorvastatin, and combinations containing SAHA and atorvastatin administered in their diet. SAHA and both combinations, but not atorvastatin, decreased the multiplicity of lung tumors, including large adenomas and adenocarcinomas with the combinations demonstrating the greatest efficacy. In Experiment 2, lung tumors were induced by 4-(methylnitrosoamino)-1-(3-pyridyl)-1-butanol in strain A/J mice followed by 180 mg/kg atorvastatin, 500 mg/kg SAHA, or both drugs administered in the diet. SAHA and the combination of both drugs, but not atorvastatin alone, decreased the multiplicity of lung tumors and large tumors, with the combination demonstrating greater efficacy. In Experiment 3, lung tumors were induced by 1,2-dimethylhydrazine in Swiss-Webster mice followed by 160 mg/kg atorvastatin, 400 mg/kg SAHA, or a combination of both drugs administered in the diet. SAHA and the combination, but not atorvastatin, decreased the multiplicity of lung tumors with the combination demonstrating greater efficacy. The multiplicity of colon tumors was decreased by SAHA, atorvastatin, and the combination, without any significant difference in their efficacy. mRNA expression analysis of lung tumor bearing mice suggested that the enhanced chemopreventive activity of the combination is related to atorvastatin modulation of DNA repair, SAHA modulation of angiogenesis, and both drugs modulating invasion and metastasis pathways. Atorvastatin demonstrated chemoprevention activity as indicated by the enhancement of the efficacy of SAHA to prevent mouse lung tumors.

Lung cancer is the leading cause of cancer deaths in both men and women in the USA,1 therefore it is exceedingly important to develop chemopreventive drugs that prevent and/or slow the progression of lung tumors. A preclinical mouse lung tumor model has been used by us and others to investigate the effect of chemopreventive agents on the occurrence and progression of lung tumors.2 This model involves the initiation of lung tumors by a carcinogen followed by treatment with chemopreventive agents. Developing lung tumors in this model have demonstrated wide variations in their response to chemopreventive agents, that is, some emerging tumors are highly sensitive, while others are more or less resistant. This is comparable to the situation in humans, where established lung tumors vary in their sensitivity to therapeutic drugs. Studies in the mouse model have shown that chemopreventive agents appear to decrease the yield of lung tumors by delaying their occurrence or by slowing their growth and progression. However, even in the continued presence of highly efficacious agents, at least some tumors eventually occur and progress to carcinomas. Thus, many chemopreventive agents appear to delay the occurrence and progression of lung tumors, rather than completely prevent them. Hence, at nontoxic dose levels, highly effective chemopreventive agents have not yielded complete prevention or cessation of growth of lung tumors.3

One way to increase chemopreventive efficacy is the use of combinations containing two or more drugs. This would be most advantageous, should the drugs have different mechanisms of action and potential toxicities. Hence, the different toxicities for the combined drugs would likely not be additive. Furthermore, should the drugs target different molecular pathways it would be expected that their activity would be efficacious to different subpopulations of emerging tumors, that is, those tumor cells dependent on pathways specifically modulated by each drug. This would result in additive or synergistic effects, since a greater population of tumors may be prevented. We propose that one way to enhance efficacy and to further delay the occurrence and progression of lung tumors is to use co-chemopreventive agents. Co-chemopreventive agents can include those with chemopreventive activity as well as those with minimal or no efficacy, but when administered along with an effective chemopreventive agent can enhance its efficacy in preventing tumors. Atorvastatin (a statin that inhibits 3-hydroxy-3-methylglutaryl CoA reductase) is a likely co-chemopreventive agent, since it appears to have minimal activity in preventing mouse lung tumors, while enhancing the activity of Polyphenon E.4 Suberoylanilide hydroxamic acid (SAHA, a histone deacetylase inhibitor) has been reported to prevent mouse lung tumors induced by 4-(methylnitrosoamino)-1-(3-pyridyl)-1-butanol [NNK].5 We report here the co-chemopreventive activity of atorvastatin as demonstrated by its enhancement of the efficacy of SAHA to decrease the occurrence of lung tumors and to delay their growth and progression to carcinomas.

Material and Methods

Carcinogens, chemopreventive agents and diet

1,2-Dimethylhydrazine (DMH) was purchased from Sigma Chemical Co. (St. Louis, MO), and NNK and vinyl carbamate from Toronto Research Chemicals (North York, Ontario, Canada). Atorvastatin and SAHA were obtained from the National Cancer Institute, Division of Cancer Prevention Repository (Rockville, MD). The AIN-76A diet was purchased from Dyets (Bethlehem, PA).

Animal experiments

Experiments with combinations containing atorvastatin and SAHA to prevent lung tumors were performed in three animal studies; each with a different carcinogen to induce lung tumors. Female strain A/J mice at 7–8 weeks of age were purchased from the Jackson Laboratory (Bar Harbor, ME) and female Swiss-Webster mice, also at 7–8 weeks of age, from Charles River Laboratories (Wilmington MA). The mice were maintained in The Ohio State University Laboratory Animal Facility under Institutional Animal Use and Care Committee approved protocols. AIN-76A pelleted diet and water were provided ad libitum. Mice were administered the chemopreventive drugs (SAHA and atorvastatin) and combinations containing the two drugs in their AIN-76A diets and at the completion of the studies were euthanized by CO2 asphyxiation followed by cervical dislocation. At necropsy, the lungs were harvested, fixed in phosphate buffered formalin, transferred within 2 days to 70% alcohol, and evaluated under a dissecting microscope for the presence and size of tumors. The lungs were then embedded in paraffin, sectioned, and stained with hematoxylin and eosin (H&E). H&E stained sections of the lungs were evaluated for the number of tumors and for histopathologic diagnosis of adenomas and adenocarcinomas. Carcinomas were distinguished from adenomas by the presence of large undifferentiated cells, cellular atypia, loss of normal alveolar architecture, increased nuclear/cytoplasmic ratio, and nuclear pleomorphism.

Experiment 1: vinyl carbamate-induced lung tumors

Female strain A/J mice at 8–9 weeks of age were given intraperitoneal injections of 16 mg/kg body weight of vinyl carbamate in sterile saline, once a week for three consecutive weeks. Two weeks after the last injection of vinyl carbamate, the mice were weighed and randomly assigned to one of the treatment groups listed in Table 1. The 500 mg/kg diet dose level of SAHA was chosen because subchronic treatment of A/J mice demonstrated an Maximum Tolerated Dose (MTD) of 600 mg/kg diet for SAHA and the 160 mg/kg diet dose of atorvastatin was chosen as the high dose level because it was not toxic in our subchronic study, and was higher than a dose that was reported to have efficacy in the ApcMin mouse.6 The mice were euthanized 21 and 32 weeks after beginning treatment with the chemopreventive agents. Our previous experience indicated that sufficient numbers of adenomas and carcinomas would be found at 21 and 32 weeks, respectively, to determine prevention by the chemopreventive agents.3

Table 1. Experimental protocols for experiments 1–3
inline image

Experiment 2: NNK-induced lung tumors

Female strain A/J mice at 8–9 weeks of age were administered 100 mg/kg body weight NNK in sterile saline by intraperitoneal injections, once a week for four consecutive weeks. Two weeks after the fourth dose of NNK, the mice were weighed and randomly assigned to one of the following four treatments groups listed in Table 1: (i) control AIN-76A diet, (ii) 500 mg/kg SAHA, (iii) 180 mg/kg atorvastatin and (iv) 500 mg/kg SAHA + 180 mg/kg atorvastatin. Each treatment group contained 20 mice, except for the Control Diet Group that contained 25 mice. The mice were euthanized after 26 weeks of treatment with the chemopreventive agents. We have previously found (at 26 weeks after NNK) a sufficient number of lung tumors to demonstrate prevention by a chemopreventive agent.7

Experiment 3: DMH-induced colon and lung tumors

DMH was used to induce both lung and colon tumors. Female Swiss-Webster mice at 8–9 weeks of age were administered 24 intraperitoneal injections, each containing 10 mg/kg body weight of DMH and at a rate of three injections per week for 8 weeks. One week after the last dose of DMH, the mice were weighed and randomly assigned to one of the following four treatments groups listed in Table 1: Group 1) Control Diet, Group 2) 400 mg/kg diet of SAHA, Group 3) 160 mg/kg diet of atorvastatin and Group 4) 400 mg/kg diet of SAHA + 160 mg/kg diet of atorvastatin. The dose level of SAHA was decreased to 400 mg/kg diet because in the above studies the 500 mg/kg dose was of sufficient efficacy in preventing tumors that it at times obscured the ability to demonstrate enhancement by atorvastatin. Each treatment group contained 25 mice, except for the Control Diet Group that contained 30 mice. The mice were euthanized after 20 weeks of treatment with the chemopreventive agents, at which time we have previously found sufficient numbers of colon and lung tumors to demonstrate prevention by chemopreventive agents.8

Statistical analysis for the three chemoprevention assays

Body weights were analyzed by an ANOVA followed by a Dunnett's test, tumor multiplicity by an ANOVA followed by the Bonferroni t-test, and the incidence of animals with tumors by an ANOVA on Ranks followed by the Dunnett's test. For all tests a p-value ≤0.05 was used for significance.

mRNA expression of cancer-related genes

To determine the effect on mRNA expression of cancer-related genes, mice from the vinyl carbamate-induced lung tumor study were administered Control Diet, 180 mg/kg atorvastatin, 500 mg/kg SAHA and the combination containing these dose levels of the two drugs for 14 days before sacrifice at week 32 (Table 1). Gene expression (mRNA) was evaluated using tumors from five mice of each of the three 14-day treatment groups, noninvolved lung tissue from five vinyl carbamate-treated mice, and lung tissue from five naive mice (not administered vinyl carbamate or a chemopreventive agent). At sacrifice, the tumors and lungs were placed in RNAlater (Ambion; Austin, TX) and stored at 4°C. Subsequently, a dissection microscope was used to excise and pool individual tumors from each lung. Tumors and lungs were placed in impact-resistant 2.0 mL vials containing 1.0 mL of RLT Plus Buffer (Qiagen; Valencia, CA) with 0.3 mL of 1.0 mm zirconia/silica beads (BioSpec Products; Bartlesville, OK). Following the manufacturer's protocol for the All Prep DNA/RNA Mini Kit (Qiagen), tissues were homogenized using a MiniBeadBeater high-energy tissue mill (BioSpec Products) for 3 min at room temperature and nucleic acids were extracted. Total RNA was quantified using the NanoDrop 2000 (Thermo Fisher Scientific; Waltham, MA) spectrophotometer and 1 μg of total RNA was reverse transcribed using the RT2 First-Stand cDNA Synthesis Kit (SABiosciences; Frederick, MD). Gene expression datasets were generated using the Cancer Pathway Finder RT2 Profiler Polymerase Chain Reaction (PCR) Array (SABiosciences) following the manufacturers recommended protocol. Briefly, 384-well PCR plates were preloaded with 84 cancer-associated gene primer sets, five “housekeeping” gene primer sets, and seven quality control primer sets in quadruplicate. The latter groups of primer sets were included for data normalization, to assess first-strand synthesis efficiency, genomic DNA contamination and technical robustness. Each sample cDNA was mixed with SYBR Green qPCR Mastermix (SABiosciences) and water, and were subsequently loaded onto qPCR arrays, with four samples loaded per card. qPCR arrays were amplified for 40 cycles on 7900HT real-time PCR machines (Life Technologies; Foster City, CA). Raw Ct values were calculated using Sequence Detection Software (v2.4; Life Technologies). Raw Ct values were compiled across all samples and uploaded into an integrated web-based software package (SABiosciences) to assign appropriate reference genes, compute standard ΔΔCt fold-changes from raw threshold cycle data, determine significance, and graphically visualize the data by heat map following unsupervised hierarchical clustering. Significant differences were determined by calculating p-values based on the Student's t-test of the replicate 2-ΔCt values for each gene in the control group and treatment groups. Significance was considered with p-values ≤0.05. Comparisons were made between: (i) normal untreated lung tissues and control tumors and (ii) control lung tumors and the lung tumors from each of the chemoprevention treatment groups in the short-term bioassay.

Results

Experiment 1: vinyl carbamate-induced lung tumors

Neither of the chemopreventive agents nor the combinations caused overt toxicity or altered the body weight of the mice, except for the 500 mg/kg diet of SAHA and the high dose (180 mg/kg diet) of atorvastatin, both of which increased the final body weight of the mice. The body weights of the SAHA, 180 mg/kg atorvastatin and Control Diet treatment groups were 26.4 ± 1.67, 24.3 ± 0.73 and 21.0 ± 0.85 gm, respectively (p-value <0.01). The final body weights of the remaining treatment groups did not differ significantly (p-value >0.05) from the Control Diet Group, that is, 60 mg/kg atorvastatin, 22.2 ± 0.88 gm; SAHA + 180 mg/kg atorvastatin, 22.0 ± 0.73 gm; SAHA + 60 mg/kg atorvastatin, 20.6 ± 1.0.

The effect of SAHA, atorvastatin (60 and 180 mg/kg diet) and two combinations on the multiplicity of total lung tumors, large adenomas and carcinomas is presented in Figure 1 with the results at week 21 depicted in Figures 1a, 1c and 1e and those of week 32 in Figures 1b, 1d and 1f, respectively. SAHA significantly decreased the multiplicity of total lung tumors at both weeks 21 and 32, while neither concentration of atorvastatin was effective (Figs. 1a and 1b). Both combinations (500 mg/kg SAHA with either 60 or 180 mg/kg atorvastatin) significantly reduced the multiplicity of total lung tumors at 21 weeks over that observed with SAHA acting alone (Fig. 1a), but did not cause a greater reduction in multiplicity of total tumors at 32 weeks (Fig. 1b).

Figure 1.

Effect of SAHA, atorvastatin and combinations containing the two drugs on lung tumors induced by vinyl carbamate. Week 21: total tumors (a); large adenomas that are >1 mm in diameter (c); carcinomas (e). Week 32: total tumors (b); large adenomas that are >1 mm in diameter (d); carcinomas (f). Results are means ± SE. The “a” indicates statistically different from the Control Diet Group, and the “b” indicates statistically different from both SAHA and atorvastatin treatment groups.

The multiplicity of large adenomas (>1.0 mm) was significantly decreased by SAHA and by both combinations, but not by either concentration of atorvastatin at weeks 21 or 32 (Figs. 1c and 1d). At week 21, the multiplicity of the large adenomas was decreased to a greater extent by the combination containing the high dose of atorvastatin than by SAHA administered alone (Fig. 1c). At week 32, the multiplicity of lung tumors in the SAHA treated group had increased to the extent that it was no longer significantly different from the Control Group. Hence, at this time point both combinations now resulted in a significant reduction in the multiplicity of large adenomas relative to the Controls as well as to SAHA administered alone (Fig. 1d).

At week 21, SAHA and two combinations significantly decreased the multiplicity of carcinomas to the same extent, that is, carcinomas were completely prevented by all three treatments (Fig. 1e). At week 32, the yield of carcinomas in the SAHA-treated mice increased to such an extent that both combinations now decreased the multiplicity of carcinomas significantly more than that caused by SAHA alone (Fig. 1f). Neither concentration of atorvastatin was effective in reducing carcinomas at either time point. The inability to demonstrate a greater reduction of carcinomas by the two combinations at week 21 was due to the total inhibition of carcinomas by SAHA and the two combinations. With this exception, the multiplicity of the three classifications of neoplastic lesions (total lung tumors, large adenomas and carcinomas) was reduced to a greater extent by the combinations than by SAHA administered alone.

Experiment 2: NNK-induced lung tumors

Neither of the treatment groups exhibited toxicity nor were there any alterations in the body weights of the mice. The final mean body weights of the mice in the different treatment groups were 25.9 ± 0.71 for SAHA, 26.2 ± 0.41 for atorvastatin, 22.9 ± 0.58 for the combination and 24.5 ± 0.92 for the Control Diet. As in Experiment 1, 500 mg/kg diet of SAHA, but not 180 mg/kg atorvastatin, significantly decreased the multiplicity of lung tumors and of adenomas greater than 1.0 mm in diameter (Figs. 2a and 2b). The combination of 500 mg/kg SAHA and 180 mg/kg atorvastatin also significantly decreased the multiplicity of all lung tumors (Fig. 2a) and decreased the multiplicity of large adenomas to a significantly greater extent than SAHA administered alone (Fig. 2b).

Figure 2.

Effect of SAHA, atorvastatin and the combination on lung tumors induced by NNK. (a) Total tumors and (b) large adenomas >1 mm diameter. Results are means ± SE. The “a” indicates statistically different from the Control Diet Treatment Group, the “b” indicates statistically different from atorvastatin and the “c” indicates statistically different from both SAHA and atorvastatin treatment groups.

Experiment 3: DMH-induced colon and lung tumors

As in the other two studies, none of the treatments appeared to cause any overt toxicity or to alter the body weights of the mice. Similar to our previous study, the lung tumors induced by DMH were classified as adenomas.7 SAHA (400 mg/kg diet), but not atorvastatin (160 mg/kg diet) significantly decreased the multiplicity of lung tumors (Fig. 3a). The combination of 400 mg/kg SAHA and 180 mg/kg atorvastatin was significantly more efficacious than SAHA administered alone in decreasing the multiplicity of lung tumors (Fig. 3a).

Figure 3.

Effect of SAHA, atorvastatin and the combinations on lung and colon tumors induced by DMH. (a) Lung adenomas and (b) colon adenomas. Results are means ± SE. The “a” indicates statistically different from the Control Diet Treatment Group, and the “b” indicates statistically different from both SAHA and atorvastatin treatment groups.

DMH also induced colon adenomas in the mice (Fig. 3b). The multiplicity of colon tumors was significantly decreased by all three treatments, that is, SAHA, atorvastatin and the combination. There was no significant difference between the three treatments in their ability to decrease the yield of colon tumors.

mRNA expression of cancer-related genes

Of the 84 cancer-related genes measured, 74 genes exhibited increased expression, whereas only three genes exhibited decreased expression in mouse lung tumors relative to lung tissues from untreated animals (p-value < 0.05). When compared to uninvolved carcinogen-treated lung tissues, control lung tumors showed significantly increased expression of 18 genes: Bad, Birc5 (survivin), Cdkn2a, Chek2, Col18a1 (endostatin), Fgfr2, Igf1, Kiss1, Map2k1, Met, Muc1, S100a4, Serpinb2, Tert, Tgfbr1, Tnfrsf10b, Twist1 and Vegfb; and significantly decreased expression of 25 genes: Akt1, Akt2, Angpt1, Apaf1, Bcl2, Cflar, E2f1, Egfr, Ets2, Fgf1, Figf, Hgf, Itga2, Itga3, Itga4, Itgb3, Jun, Mmp2, Mmp9, Pdgfa, Pdgfb, Pik3r1, Plaur, Pten and Tgfb1 (p-value <0.05). Unsupervised cluster analysis and visualization by heatmap demonstrated that the three groups can be discriminated between normal lung, carcinogen exposed noninvolved lung and control lung tumors (10/10, 5/5 and 5/5; Fig. 4).

Figure 4.

Unsupervised hierarchical cluster analysis and heat map visualization of 84 qRT-PCR probe sets across normal lungs (n = 10), carcinogen exposed noninvolved lungs (n = 5), and control lung tumors (n = 5) using Cancer Pathway Finder RT2 Profiler PCR Arrays (SABiosciences). Columns: samples; Rows: genes; Color key indicates gene expression value, green: Lowest, red: highest.

The genes significantly modulated (p-value <0.05) following the 14-day administration of individual or the combination of chemopreventive agents before sacrifice are presented in Table 2. Atorvastatin modulated the expression of four genes, decreasing the expression of Brac1, Muc1 and Pdgfa, while increasing the expression of Tgfb1. The increased expression of Tgfb1 was unique in being the only gene with a significantly greater relative expression by the chemopreventive agents atorvastatin, SAHA, or the combination. SAHA decreased the expression of 12 genes in lung tumors: Bad, Col18a1 (endostatin), Igf1, Kiss1, Met, Mta1, Muc1, Pdgfa, Raf1, Tert (telomerase), Vegfb and Vegfc. Five of the 12 genes modulated by SAHA are genes in the angiogenesis pathway: Col18a1 (endostatin), Igf1, Pdgfa, Vegfb and Vegfc; and four of the 12 genes are genes associated with invasion and metastasis pathways: Kiss1, Met, Mta1 and Muc1.

Table 2. Effect of atorvastatin, SAHA and their combination on mRNA expression of genes in mouse lung tumors
inline image

The combination of atorvastatin and SAHA decreased the expression of 11 genes in the lung tumors: Atm (ataxia telangiectasia mutated protein), Bax, Brac1, Chek2 (Rad53), Muc1, Nme4, Pdgfa, Serpinb2, Twist1, Vegfb and Vegfc. The combination decreased the expression Brac1, Muc1 and Pdgfa in common with atorvastatin and Muc1, Pdgfa, Vegfb and Vegfc in common with SAHA (Table 2). The combination decreased the expression of three genes in DNA repair pathways: Atm, Brac1 and Chek2 (Rad53) of which atorvastatin also decreased Brac1. Three of the genes decreased by the combination in common with SAHA are genes in the angiogenesis pathway: Pdgfa, Vegfb and Vegfc. With respect to the invasion and metastasis pathway, the combination treatment decreased the expression of four genes (Muc1, Nme4, Serpinb2 and Twist1) with one of the genes, Muc1 also being decreased by atorvastatin and SAHA treatments.

Discussion

A logical way to enhance the efficacy in chemoprevention studies is to treat with combinations containing chemopreventive drugs having different pharmacologic and toxicologic properties. One of the first examples of combination treatment for the prevention of lung cancer was reported by el-Bayoumy et al.9 who found that a mixture of 1,4-phenylenebis(methylene) selenocyanate, phenethyl isothiocyanate, indole-3-carbinol and d-limonene in combination was significantly more inhibitory than indole-3-carbinol given alone for reduction of mouse lung tumor multiplicity. In other experiments (where each agent alone was not quantified), Witschi et al.10 showed that a combination of myoinositol and dexamethasone significantly inhibited lung tumors induced in A/J mice by 5 months of treatment with sidestream and mainstream cigarette smoke. Each of these agents had previously been examined individually and in combination by Estensen and Wattenberg11 who found that 3% myoinositol and 0.5 μg/g dexamethasone in the diet reduced adenoma multiplicity induced by BaP by 40 and 57%, respectively, and when used in combination appeared to be additive (68% reduction) for tumor inhibition. In similar experiments, Wattenberg et al.12 found that myoinositol given in combination with aerosolized budesonide or beclomethasone was more effective than any of the agents given alone for inhibition of BaP induced lung tumors in A/J mice. Administering Budesonide in the diet in combination with oral gavage of a farnesyl transferase inhibitor (Zarnestra MT) was found by Alyaqoub et al.13 to prevent vinyl carbamate-induced lung tumors more effectively than the individual agents at weeks 20 and 36. In fact although the combination of budesonide and zarnestra was more efficacious at week 36, Zarnestra alone was ineffective. They also reported that budesonide, zarnestra, and the combinations reversed global hypomethylation of DNA in the tumors. D'Agostini et al.14 have demonstrated a synergistic relationship between N-acetylcysteine and ascorbic acid for inhibition of urethane-induced mouse lung tumors. Hecht et al.15 used a combination of N-acetyl-S-(N-2-phenethylthio-carbamoyl)-l-cysteine and myoinositol given at different times when the carcinogens, BaP + NNK, were being administered to induce lung tumors in mice. N-acetyl-S-(N-2-phenethylthiocarbamoyl)-l-cysteine failed to reduce tumor multiplicity in the postcarcinogen or the 75% carcinogen treatment phases, but in combination with myoinositol, all treatment phases yielded reductions in tumors with the combination giving significantly more reduction than either agent acting alone. A follow-up to these studies was performed16 in which N-acetyl-S-(N-2-phenethylthiocarbamoyl)-l-cysteine, indole-3-carbinol and myoinositol were given individually and in combination beginning after 50% of the carcinogen treatment or 1 week after carcinogen treatment. All of the individual agents significantly reduced tumor multiplicity and all of the combinations gave significantly greater reductions in tumor multiplicity, but the enhanced reductions were not synergistic.

Histone deacetylase inhibitors such as SAHA, and statins, including atorvastatin, have been investigated for the prevention of cancer. SAHA has been reported to prevent N-methylnitrosourea-induced mammary cancer in rats17, 18 and lung tumors in mice, including NNK-induced lung tumors.5 For the statins, the evidence for inhibition of cancer in humans is inconsistent.19 Large retrospective studies involving the use of high and low dose levels of atorvastatin, simivastatin, lovastatin and fluvastatin indicated that high dose levels may have been associated with a reduction in cancer cases;20 however, other studies did not find compelling support for reductions in cancer risk by statins.21, 22 Interestingly, combinations containing a statin with another chemopreventive agent have suggested synergistic activity in lowering the risk of cancer in humans. Hoffmeister et al.23 found that use of either statins or low dose aspirin (100 mg) resulted in a lowering of the risk for colorectal cancer, but also found that the use of low-dose aspirin in concert with statins caused a greater reduction in risk than either agent acting alone. Correspondingly, among those individuals involved in the California Men's Health Study, Flick et al.24 found that use of statins for more than 5 years resulted in a 28% reduction in risk of prostate cancer.

Although the evidence for the prevention of cancer by atorvastatin is limited and inconsistent, the studies presented herein, demonstrated that atorvastatin does not prevent lung tumors induced in mice by three different carcinogens (vinyl carbamate, NNK or DMH). In contrast, SAHA did prevent the occurrence and the progression of mouse lung tumors. Furthermore, combinations of atorvastatin and SAHA decreased the yield of lung tumors to a greater extent than SAHA acting alone; regardless of the carcinogen used to initiate the tumors, demonstrating the co-chemopreventive activity of atorvastatin. Sacrifice of mice with vinyl carbamate-induced tumors at both weeks 21 and 32 revealed that SAHA and the combination with atorvastatin, but not atorvastatin alone, reduced the total yield of lung tumors, large adenomas and carcinomas. Similar inhibition of lung tumors by the combination of SAHA and atorvastatin and the ineffectiveness of atorvastatin was observed when either NNK or DMH was used to initiate tumor formation. The inability of atorvastatin to inhibit the occurrence of lung tumors in A/J mice has also been demonstrated by Lu et al.,4 who reported that neither the green tea polyphenol, polyphenon E, nor atorvastatin was capable of inhibiting lung tumor formation, whereas their combination decreased both lung tumor multiplicity and tumor burden.

Although, we could not demonstrate that atorvastatin (160 mg/kg diet) inhibited mouse lung tumors, we did find that it significantly reduced the formation of DMH-induced colon tumors in Swiss-Webster mice. SAHA and the combination containing atorvastatin and SAHA also prevented DMH-induced colon tumors to a similar extent as atorvastatin. Atorvastatin (150 mg/kg diet) has been reported by Reddy et al.25 to prevent azoxymethane-induced colon tumors in F344 rats. Furthermore, they found that administering 100 ppm of atorvastatin in combination with either aspirin or celecoxib resulted in a greater inhibition of azoxymethane-induced colon tumors than any of the agents administered alone. In addition, atorvastatin has been shown to act synergistically with celecoxib to inhibit xenografts of prostate PC-3 cells in Severe Combined Immunodeficiency (SCID) mice26 and to inhibit the formation of colon and small intestinal polyps in ApcMin mice.6 However, Huang et al.27 could not demonstrate a reduction in polyps in the Min mouse model using 222 mg/kg atorvastatin in the diet, but did find an inhibition of growth of colon tumor cell xenografts. In summary, atorvastatin has consistently been found to act synergistically with Non-Steroidal Anti-Inflammatory Drugs (NSAIDs) in preventing colon cancer.

To determine whether the enhanced chemoprevention of lung tumors exhibited by the combination containing atorvastatin and SAHA was related to their interaction in modulating the mRNA expression of genes, we screened the ability of the combination to modulate the expression of cancer-related genes in lung tumors. The effects of Atorvastation, SAHA or the combination on the mRNA expression of cancer-related genes in lung tumors demonstrated that the chemopreventive agents primarily decreased the expression of genes rather than increasing them. Only one gene, Tgfb1 was increased as the result of treatment with a chemopreventive agent (atorvastatin). These results are similar to two other chemopreventive drugs, bexarotene and budesonide, which also inhibit mouse lung tumors. Bexarotene, a Retinoic Acid Receptor (RAR) agonist, prevents vinyl carbamate-induced lung tumors in mice. Global expression profiling using microarray demonstrated that bexarotene significantly decreased the expression of 45 genes while increasing the expression of only two genes: Apolipoprotein D, a biomarker of its hyperlipidemic effect, and CYP26b, a member of the cytochrome P450 family of drug metabolizing enzymes induced by RAR agonists.28 Budesonide, a glucocorticoid steroid, also prevents the formation of mouse lung tumors and concomitantly decreases the mRNA expression of many more genes than it increases.4, 29, 30 Thus, the chemopreventive activity of four drugs with very differing pharmacologic activity, that is, atovastatin, bexarotene, budesonide and SAHA appears to involve the attenuation of the mRNA expression of genes in cancer-related pathways.

Atorvastatin decreased the mRNA expression of three genes in lung tumors, SAHA decreased 12 genes and the combination decreased the expression of 11 genes. All three treatments decreased the expression of (i) Pdgfa, a gene involved in angiogenesis that was also increased in mouse lung tumors from mice not administered one of the treatment regimens and (ii) Muc1, a gene involved in invasion and metastasis. Both atorvastatin and the combination decreased the expression of Brac1, a gene involved in DNA repair. The combination also decreased the expression of the Atm, and Chek2 (Rad53) genes that are also involved in DNA repair. SAHA did not alter the expression of any of the 15 genes of the array that were involved in DNA repair. This would suggest that the contribution of atorvastatin is related to the modulation of DNA repair pathway in the enhanced chemopreventive activity of the combination.

Three of the genes decreased by the combination in common with SAHA are members of the angiogenesis pathway: Pdgfa, Vegfb, and Vegfc. SAHA also decreased the expression of two other genes involved in angiogenesis but not decreased by the combination: Col18a1 (endostatin) and Igf1. These results support the current literature that the chemopreventive activity of SAHA is, in part, related to its antiangiogenic effects.31 SAHA has also been shown to have an effect on invasion and metastasis pathways in models of breast and prostate cancers.32, 33 Our results demonstrate decreased expression of four genes in the invasion and metastasis pathway: Kiss1, Met, Mta1 and Muc1. The combination treatment with SAHA and atorvastatin decreased the expression of four genes in the invasion and metastasis pathway: Muc1, Nme4, Serpinb2 and Twist1. The only gene decreased in common with the two chemopreventive drugs is Muc1. The lack of modulation of these two pathways by the individual drugs would suggest that the enhanced chemopreventive activity of the combination may be related to modulation of invasion and metastasis that is lacking for the individual drugs. Hence, it would appear that the enhanced chemopreventive activity of the combination may be due to: the contribution of atorvastatin to the modulation of DNA repair, SAHA contributing to the modulation of angiogenesis, and both drugs contributing to the modulation of invasion and metastasis pathways.

In summary, synergy was indicated for the combined treatment with SAHA and atorvastatin since it caused a significantly greater reduction in vinyl carbamate-, NNK- and DMH-induced lung tumors than either SAHA or atorvastatin. As atorvastatin alone was ineffective, administering atorvastatin together with SAHA would not be expected to be additive for the reduction in lung tumors already caused by SAHA. However, a significantly greater reduction in lung tumors was found when atorvastatin was administered along with SAHA to prevent lung tumors. Synergy was confirmed between SAHA and atorvastatin with respect to the prevention of (i) vinyl carbamate-induced total lung tumors and large adenomas at week 21 and large adenomas and carcinomas at week 32, (ii) NNK-induced large adenomas and (iii) DMH-induced lung tumors. Our results demonstrate the advantage of using a co-chemopreventive drug (atorvastatin, that does not prevent mouse lung tumors) in combination with a chemopreventive drug (SAHA, that is efficacious in preventing mouse lung tumors) to enhance the prevention of lung tumors.

Ancillary