Background
Colorectal cancer is the second most common form of cancer in incidence and mortality in developed countries [
1]. While its etiology remains unknown, it is thought that both genetic and environmental factors play critical roles in causing colorectal cancer. Epidemiological studies consistently show that physical activity is associated with a reduced risk of colon cancer [
2‐
5]. However, the protective effect of exercise against colon carcinogenesis has not been consistently demonstrated in animal models [
6‐
9]. The inconsistent results observed in animal studies is likely due to the nature of the animal model as well as the type of exercise employed. For example, voluntary exercise was shown to inhibit 1,2-dimethylhydrazine- or azoxymethane-induced colon tumorigenesis in rats [
6,
7], whereas forced exercise was shown not to inhibit intestinal tumorigenesis in
Apc
Min/+ mice [
8]. Mehl
et al., however, noted that intestinal tumor formation in
Apc
Min/+ mice was inhibited by forced exercise but not by voluntary exercise [
9]. Colbert
et al. reported that voluntary exercise combined with food deprivation and weight loss resulted in a reduction in tumor formation in male
Apc
Min/+mice [
10]. Therefore, additional studies are required to evaluate the potential protective effects of physical exercise against intestinal tumor formation.
Apc
Min/+ mice carry a dominant heterozygous nonsense mutation at codon 850 of the mouse homologue of the human tumor suppressor gene,
APC (
Adenomatous polyposis coli). This mutation is significantly implicated in both sporadic and inherited human colorectal carcinogenesis [
12‐
14].
Apc
Min/+ mice, therefore, are recognized as a genetically relevant animal model mimicking human intestinal carcinogenesis and have been utilized extensively for various chemoprevention studies [
15]. Azoxymethane (AOM)-induced and dextran sulfate sodium (DSS)-promoted colon tumorigenesis in mice is an inflammation-associated colon carcinogenesis model that has been used in chemoprevention studies [
11]. The present study was designed to investigate whether voluntary exercise inhibits intestinal tumorigenesis in two different animal models,
Apc
Min/+ mice and AOM/DSS-treated mice. To further elucidate the mechanisms of the inhibition of tumorigenesis, the effects of exercise on insulin-like growth factor (IGF)-1 and IGF binding protein (IGFBP)-3 ratio, aberrant β-catenin signaling, and arachidonic acid metabolism were also investigated in the
Apc
Min/+ mice.
Discussion
The present study indicates that the opportunity to engage in voluntary running wheel exercise decreases the number of small intestinal tumors in female
Apc
Min/+ mice maintained on AIN93G or a high fat diet by ~30% (Table
1) and the number of colon tumors (per tumor bearing mice) in male AOM/DSS-treated mice maintained on a high-fat diet by 51% (Table
4). Effects of voluntary or forced exercise on colon tumor formation in AOM/DSS-treated mice have not been reported, and the present study is the first that shows the inhibitory effect of voluntary exercise on colon tumorigenesis in this model.
Voluntary exercise has been shown previously to inhibit chemically-induced colon tumorigenesis in rats [
6,
7], but this beneficial effect has not been shown previously in the
Apc
Min/+ mouse model [
9]. Forced exercise was shown to inhibit tumorigenesis in male
Apc
Min/+ mice in one study [
9], but was without beneficial effect in a second study [
8]. In addition, in previous studies, neither forced nor voluntary exercise were found to inhibit tumorigeneisis in female
Apc
Min/+ mice [
8,
9]. The reasons for the inconsistent results from the different studies in
Apc
Min/+mice are not clear. One possible reason is the large variation in tumor yield among individual
Apc
Min/+ mice (high variability across mice is often observed in our laboratory as well as in other research groups), which by chance might have resulted in false-negative results when small numbers of animals are used. Colbert
et al. reported that voluntary exercise combined with food deprivation and weight loss resulted in a reduction in tumor formation in male
Apc
Min/+ mice [
10]. The authors concluded that the negative energy balance caused by the combination of food deprivation and exercise was the underlying mechanism for tumor reduction. In the present study, we found that voluntary exercise significantly inhibited small intestinal tumorigenesis in female
Apc
Min/+ mice on a normal AIN93G or a high-fat diet as well as colon tumorigenesis AOM/DSS-treated male mice on a high-fat diet. Since mice in the present study were not food deprived, we conclude instead that exercise caused the inhibition of tumorigenesis under normal physiological conditions.
It was recently reported that voluntary running wheel exercise decreased the thickness of the dermal fat layer and inhibited ultraviolet B light-induced skin carcinogenesis in SKH-1 mice [
18]. Also, a highly significant correlation between dermal fat thickness away from tumors and the skin tumor multiplicity was found across individual mice [
18,
19], suggesting that the exercise-induced reduction in dermal fat contributed to the inhibition of skin tumorigenesis. In order to assess the relationship between regional body fat and the number of intestinal tumors in individual mice in our study, we performed linear regression analysis. However, a statistically significant association between the weight of regional body fat (omental or retroperitoneal) and the number of intestinal tumors (in different sizes and locations) per mouse was not found (data not shown). In our study,
Apc
Min/+ mice maintained on the high-fat diet had greater omental and retroperitoneal body fat weight/body weight (by 21% and 87%) than the
Apc
Min/+ mice on AIN93G diet, but intestinal tumor numbers were not significantly different. Moreover, we previously reported that administration of caffeine decreased the two regional body fat weights, but did not inhibit intestinal tumorigenesis in the
Apc
Min/+ mouse model [
16]. Taken together, these results suggest that the reduction of body fat levels may not be responsible for the inhibition of intestinal tumorigenesis caused by voluntary exercise in the
Apc
Min/+ mouse model.
High circulating levels of IGF-1 and low levels of IGFBP-3 (a major IGFBP that accounts for ~75% of IGF-1 bound) have been shown to be associated with an increased risk of several types of cancers, including colorectal cancer [
20,
21]. We found that voluntary exercise increased serum IGFBP-3 levels and decreased IGF-1/IGFBP-3 ratio (Table
2). The decrease in serum IGF-1 levels was not statistically significant (Table
2), and this is consistent with previous studies in the
Apc
Min/+mouse model [
8,
9] as well as other models [
22,
23]. In order to assess the relationship between IGFBP-3, IGF-1, or IGF-1/IGFBP-3 ratio and intestinal tumor numbers in individual mice, linear regression analysis was performed with data from 38 mice (in Experiment 2), but no significant associations were found (data not shown). It is still of interest, however, to further investigate how voluntary exercise increases IGFBP-3 levels and how this effect could contribute to the inhibitory action on intestinal tumorigenesis.
Table 2
Effect of voluntary exercise on serum IGF-1 and IGFBP-3 levels in Apc
Min/+ mice on high-fat dieta
Control (19) | 312.4 ± 14.5 | 53.5 ± 2.8 | 22.8 ± 1.6 |
Exercise (19) | 283.0 ± 18.8 | 63.9 ± 3.4* | 16.8 ± 1.6** |
The observed increase of E-cadherin levels and decrease of nuclear β-catenin levels in tumors (Figure
1) may contribute to the inhibition of tumorigenesis in
Apc
Min/+ mice by exercise because aberrant β-catenin signaling is a key molecular event in the development of intestinal tumors in this model [
24]. Similar events have been associated with the cancer preventive activities of (-)-epigallocatechin-3-gallate (EGCG), the major polyphenolic compound found in green tea [
16]. The observed decrease in levels of PGE2 in serum and tumors (Table
3) may also contribute to the inhibition of intestinal tumorigenesis. It has been well demonstrated that reduction of PGE2 production decreases colorectal carcinogenesis [
25]. A high level of leisure-time physical activity has been reported to be associated with low PGE2 levels in rectal mucosa [
26].
In the present study, five mice were group-housed in each cage; therefore, we could not determine how much each mouse ran each day. We did observe that 2–4 mice could run on the running wheel together. Nonetheless, it is possible that dominant mice gained more assess to the wheel than subordinates. However, if this were the case, we would predict high variability in the exercise group on measures of tumor number, body weight, fat weight, etc. In the present study, the variability observed in these parameters in the exercise groups was similar to that observed in the corresponding control groups. Therefore, although there were five mice per cage, we conclude they each contributed about equally to the number of wheel revolutions.
Competing interests
The authors declare that they have no competing interests.
Authors' contributions
JJ participated in designing experiments, breeding & genotyping mice, and sacrificing animals, and performed tumor-scoring, tissue-harvesting, Western blotting, ELISA, data analyses, and manuscript writing & submission. BM&MC performed animal experiments and participated in sacrificing animals and data analyses. MB participated in breeding, genotyping, and sacrificing mice. YL performed two-way ANOVA and normality tests in statistical analyses. GCW co-conceived the study with CSY and helped from designing experiments to writing manuscript. CSY conceived the study, coordinated different experiments, and revised the manuscript. All authors read and approved the manuscript.