|Home | About | Journals | Submit | Contact Us | Français|
Colorectal cancer, responsible for 50,000 deaths per year, is a contributing factor for considerable mortalities in the United States. Consumption of well-done red meat and saturated fats rich in polycyclic aromatic hydrocarbons may be one of the causative factors for sporadic colon cancer. The objective of this study was to investigate whether the formation of colon tumors in adult ApcMin mice was influenced by the ingestion of different types of fat containing benzo(a)pyrene [B(a)P], a polycyclic aromatic hydrocarbon compound. Treatment consisted of 50 and 100 μg B(a)P/kg body weight dissolved in peanut or coconut oil (representatives of unsaturated and saturated fats, respectively) administered daily to six-week-old male ApcMin mice via oral gavage for sixty days. At the end of exposure, mice were sacrificed; jejunum and colons were retrieved and preserved in 10% formalin for observation for gross pathological changes. An increased prevalence of adenomas in colons of mice that ingested B(a)P through saturated dietary fat compared to unsaturated fat and controls (p < .05) was noticed. Interestingly, we also observed adenomas with high-grade dysplasia in the B(a)P + saturated fat group, and these incidences were frequent at the 100 μg/kg B(a)P dose. On the other hand, the B(a)P-alone and unsaturated-fat groups did not show significant differences in the numbers of adenomas and invasive tumors in the both jejunum and the colon. Our studies established that dietary fat, especially saturated fat, potentiates the development of colon tumors caused by B(a)P in the ApcMin mouse.
Colorectal cancer (CRC) is one of the most common cancers in the western world. In the United States alone, nearly 108,070 new cases of CRC have been diagnosed recently, and 49,960 deaths were attributed to this cancer (NCI-SEER 2008). Both hereditary (familial) and sporadic forms of CRC occur. Of the familial forms of CRC, familial adenomatous polyposis (FAP; associated with germ line mutations in genes such as adenomatous polyposis coli [APC]) and hereditary nonpolyposis colon cancer (HNPCC; associated with DNA mismatch repair enzymes) are the most common. Familial adenomatous polyposis cancer is inherited as an autosomal dominant disorder. Although FAP contributes to 1% of the colon cancer cases, the HNPCC contributes annually to 3% to 4% of cases (Rustgi 2007). Individuals with FAP develop large numbers of adenomas in their colon and rectum during their late teens or early twenties. The adenomas are capable of progressing to carcinomas and becoming invasive, and they can metastasize. The gene responsible for FAP has been identified and designated as APC (Joslyn et al. 1991; Kinzler and Vogelstein 1996). Most sporadic colorectal cancers have the same genetic etiology as FAP, but the mutations in the APC gene are somatically acquired during the life of an individual. Sporadic gene damage seems to play an important role in the development of tumors the in colon, as in 90% of colon cancer cases; there is no familial history of colon cancer. Dietary factors and environmental agents have been suspected of causing sporadic gene mutations and therefore of being involved in the induction of sporadic colon carcinomas (World Cancer Fund and the American Institute for Cancer Research 1997).
Epidemiological studies also have shown that environmental factors, and especially diet, play an important role in colon cancer susceptibility (Cross and Sinha 2004; Martinez 2005; Sinha, Kulldorff, et al. 2005; Sinha, Peters, et al. 2005). Among the several environmental toxicants, benzo(a)pyrene (B[a]P), a polycyclic aromatic hydrocarbon (PAH) compound, has generated the most interest, as this chemical is formed in red meat cooked at high temperatures (Butler et al. 2005; Murtaugh et al. 2004). In addition to its formation during cooking at high temperatures (Lijinsky 1991), B(a)P is also contributed to the diet from several environmental sources, thus making it a potential candidate to contribute significantly to dietary contamination, human intake, and development of colorectal cancer (Ramesh et al. 2004; Sachse et al. 2002).
Because of the prevalence of B(a)P in the environment and the great potential for human exposure, studies are warranted to see the pathological changes that ensue following oral B(a)P treatment using animal models. The high incidence of tumors formed in the distal colon of some mouse models and the sequential development of adenomas to adenocarcinomas make the murine models ideal to recapitulate colon carcinogenesis in humans (Heyer et al. 1999; Rosenberg, Giardina, and Tanaka 2009). Of all the murine models, ApcMin has been widely used in recent years to study the onset and progression of gastrointestinal (GI) cancers. This paper examines the relationship between the administration of B(a)P, a known carcinogen, and dietary fat using the ApcMin mouse model. Also, it explores the interaction between B(a)P and dietary fat (saturated vs. unsaturated) in the etiology of colon cancer.
Six-week-old male ApcMin mice (Jackson Labs, Bar Harbor, ME, USA) weighing approximately 24 g were used in this study. The animals were housed in groups of two or three per cage, maintained on a twelve-hour light/dark cycle (lights on at 6:00 a.m.), and allowed free access to rodent chow (2016 Teklad Global 16% protein rodent diet [3.5% fat]; Harlan Laboratories, Indianapolis, IN, USA) and water. All mice involved in this study were housed in polycarbonate cages (Lab Products, Inc., Seaford, DE, USA) with laboratory grade 7089 Teklad Diamond Soft Cellulose (Harlan Laboratories) material as bedding. Mice were housed in an Association for the Assessment and Accreditation of Laboratory Animal Care International (AAALAC)-accredited animal care facility. This facility is under the oversight of Institutional Animal Care and Use Committee (IACUC), which ensures that animal care conforms with the National Institutes of Health (NIH) guidelines for the humane care and use of laboratory animals (NRC 1996). All animals were allowed a seven-day acclimation period prior to being randomly assigned to a control (n = 10 per each time point) or treatment group (n = 10 per each time point). Treatment consisted of two doses (50 and 100 μg/kg) of B(a)P (97% pure, unlabeled; Sigma Chemical Co., St. Louis, MO, USA) dissolved in research-grade peanut and coconut oils (representatives of unsaturated and saturated fat, respectively; Sigma); control groups were given only the oils without B(a)P. Mice that received no fat and B(a)P served as negative controls, whereas mice that received no fat, but received B(a)P through tricaprylin (vehicle for [B(a)P]) served as positive controls. The test chemical was administered daily via a single oral gavage for sixty days. As B(a)P is a potential carcinogen and mutagen, it was handled in accordance with NIH guidelines (NIH 1981). The experimental design for B(a)P dosing is illustrated schematically in Figure 1.
We chose the B(a)P doses based on the following rationale. Dietary exposure of humans to B(a)P varies. Although some studies reported B(a)P intake of 2.8 μg/person/day (Hattemer-Frey and Travis 1991), other studies reported 8.4 μg/person/day (Falco et al. 2003) and 17 μg/person/day (deVos et al. 1990). To calculate the B(a)P dose that was to be administered to mice, we used published data for human dietary intake of B(a)P. For dose computation, we chose the highest daily exposure of 17 μg/person/day for an average male who weighs 70 kg. Using this value, the human intake of B(a)P translates to 0.24 μg/kg bw/day. The doses of 50 and 100 μg B(a)P/kg bw given to ApcMin mice in our study are around 200 to 400 times the highest average daily exposure reported for humans. However, because of the increasing environmental contamination by B(a)P, and inconsistencies surrounding the human intake of B(a)P on a daily basis, and peak human exposure to B(a)P is unknown, we chose to use high doses of B(a)P in this study. Dosing vehicles containing peanut and coconut oils were selected because of their fatty acid composition. For example, peanut oil represents mono-unsaturated dietary fat (48%; composition of fat type/100 g of food); coconut oil represents saturated dietary fat (85%; composition of fat type/100 g of food; Holland et al. 1992).
All the ApcMin mice from control and experimental groups were observed twice a day for moribundity and mortality (once a day on weekends and holidays). Mouse body weight and food consumption were monitored periodically. At the end of sixty days of exposure, mice were sacrificed and tissues of interest such as jejunum, colon, and liver were harvested from control and experimental mice according to the guidelines of Ruehl-Fehlert et al. (2003), as follows. The intestines were separated from the mesentery with a pair of fine scissors. Briefly, the small intestine was excised from the duodenum to the ileocecal valve. The large intestine was excised into proximal (from the cecum through the transverse colon), distal (from the splenic flexure through the sigmoid colon), and rectum (first 2 cm from the anal opening) portions. Using a pair of fine scissors, the intestine was longitudinally opened; the mucosal layers were spread out in a petri dish with wet tissue paper and were gently rinsed with physiological saline to flush the ingesta. Before the intestines were processed for histopathological studies, the size, location, and number of adenomas in the small intestine and colon were documented. Adenomas were identified by the naked eye as tumorlike excrescences that stood out from the surrounding mucosa. The intestines were preserved for histopathological examination using the Swiss roll technique (Boivin et al. 2003) and fixed in 10% formalin prior to routine histologic processing. The pathologist examined the slides without knowledge of the treatment group.
Tumors were classified as adenoma or carcinoma, and the degree of dysplasia in each adenoma was graded. The neoplastic lesions were evaluated by size (>2.5 mm and >5 mm), number (single and multiple), and type (adenoma, carcinoma, degree of dysplasia). The scheme adopted for classification of lesions is shown in Table 1.
The number of polyps and polyp sizes were analyzed by using two-way analysis of variance to determine the effect of dietary fat type, dose, and fat × dose interaction. When p for the interactions was <.05, means of fat groups were separated using the Student-Newman-Keuls multiple-range test. When p was <.05 for the effects of fat type or dose but not for the interactions, overall means for fat type or dose groups were separated by using the Student-Newman-Keuls multiple-range test.
No B(a)P treatment–related deaths occurred in mice during administration of either B(a)P or the vehicle. Also, no treatment related changes were seen in either body weight (Table 2) or food consumption (Table 3) of mice administered with B(a)P compared to controls. Although some minor changes were noticed, the data were not statistically significant.
The effect of various treatments on tissue weights are depicted in Table 4. The ApcMin mice that received B(a)P through saturated fat had 4.4% and 7.4% heavier jejunum, compared to those that received B(a)P through unsaturated fat and tricaprylin, respectively, at the 50-μg/kg dose. Likewise, mice that received B(a)P through saturated fat had 4.3% and 5.8% heavier jejunum, compared to those that received B(a)P through unsaturated fat and tricaprylin, respectively, at the 100-μg/kg dose. A similar trend was noticed in the case of colon with B(a)P plus saturated fat, which registered a 4.3% and 13% heavier colon mass, relative to its B(a)P plus unsaturated fat and tricaprylin treatment categories at the 50-μg/kg dose. Mice that received B(a)P through saturated fat had a 4% and 20% heavier colon, compared to those that received B(a)P through unsaturated fat and tricaprylin respectively at the 100-μg/kg dose.
No appreciable differences in polyp size were detected among vehicle (peanut oil, coconut oil) only–treated ApcMin mice that received a standard diet. The average colon tumor size per mouse was larger in the B(a)P-treatment group, compared to the controls. Representative gross photographs of a B(a)P + dietary fat–induced tumor are shown in Figure 2.
Colon polyps from vehicle (peanut oil and coconut oil) only–treated mice were found to be similar in number to polyps from the ApcMin mice that received a standard diet (data not shown). The trend in polyp numbers in the jejunum was in consonance with that observed for the colon across the doses and B(a)P-treatment categories. The only striking difference was the polyp number, which was greater in the colon compared to jejunum, and the data were statistically significant (p < .05).
Data on the number of polyps in the colon of ApcMin mice is presented in Figure 3. Colonic polyp formation showed a treatment-related effect in both the 50 and the 100 μg/kg dose groups. The numbers of polyps were higher in B(a)P-treated mice compared to controls (vehicle-treated), and the difference was statistically significant (p < .05). A similar trend was observed in mice that received B(a)P through unsaturated and saturated fat individually, compared to their respective controls (p < .05). The differences in tumor numbers between mice that received B(a)P alone and B(a)P in unsaturated fat were not statistically significant. However, the differences in tumor numbers between mice that received B(a)P alone and B(a)P in saturated fat were statistically significant (p < .05). Similarly, colon polyp numbers induced by B(a)P between the two dietary fat categories were statistically significant (p < .05), with mice that received B(a)P through saturated fat registering more polyps than their unsaturated-fat counterparts. Also, a clear B(a)P dose–related effect was seen in polyp numbers, with the 100-μg/kg dose group showing more polyps relative to the 50-μg/kg dose group for both the dietary fat categories tested (p < .05).
Regional distribution of colonic polyps was affected by interactive effects of B(a)P and dietary fat, and the results are shown in Table 5. More polyps were seen in the distal colon compared to proximal and medial regions in all the three treatment groups (p < .05). No differences in the total number of polyps among the three regions were noted between mice that received B(a)P only and B(a)P plus unsaturated fat. On the other hand, the differences were statistically significant between B(a)P plus saturated fat group and the other two treatment groups. A dose-response relationship was seen in polyp regional distribution, with the 100 μg/kg showing more polyps than 50 μg/kg in the distal colon than the proximal colon.
Histopathology of the colon, jejunum, and liver subsequent to subchronic B(a)P exposure in saturated fat is shown in Figure 4. Histopathology of jejunum, colon, and liver tissues from B(a)P-alone-treated mice and B(a)P + unsaturated fat–treated mice were not shown because their pathological characteristics were not different from those mice that received B(a)P through unsaturated fat. Colon polyps from vehicle (peanut oil and coconut oil) only–treated mice were found to be similar in histology to polyps from the ApcMin mice that received a standard diet. Most of the tumors observed in both jejunum and colon of ApcMin mice were adenomas. A few invasive adenocarcinomas with invasion of the muscularis mucosae and deeper layers of the bowel wall were also found in the 100-μg/kg dose group.
Studies in animal models are necessary to identify the mechanistic underpinnings of environmentally induced colon cancers. The ApcMin mouse is one such model that is reliable and robust, and thus it can be used to study environmental toxicant–induced gastrointestinal tract colon carcinogenesis and the effects of diet and chemopreventive agents (Dashwood 2003; Femia and Caderni 2008; Halberg et al. 2008; Hu et al. 2006; Sattar et al. 1999; Shen et al. 2007). This is a mutant mouse with multiple intestinal neoplasias (Min; Su et al. 1992). It has a mutated adenomatous polyposis coli (Apc), a tumor-suppressor gene, similar to that in patients with familial adenomatous polyposis. Another interesting feature of the ApcMin mouse is that it mimics the rapid development of adenomatous polyps that affect humans. In this mouse, the loss of a single copy of the Apc gene predisposes the mice to the development of a large number of polyps, in both the small and the large intestine (Lamprecht and Lipkin 2003). One caveat to using this mouse as a model is that most ApcMin mice develop tumors in the small intestine rather than the colon, unlike humans (Yu et al. 2001). This demerit notwithstanding, these mice are used extensively as animal models for FAP and sporadic colorectal cancers (McCart, Vickaryous, and Silver 2008; Radtke and Clevers 2005). Thus, we used this mouse model to study dietary fat modulation of B(a)P-induced colorectal carcinogenesis.
The increase in weight of the jejunum and colon of B(a)P-treated mice through saturated fat compared to unsaturated fat and controls could be owing to the presence of a greater number of polyps in the intestines. However, this variation is small for both the jejunum and the colon and not statistically significant, except for the B(a)P + saturated fat group at the 100-μg/kg dose. The lack of variation in overall well-being of the mouse (as reflected by the body weight and food consumption data) indicate that B(a)P exposure did not impose any undue stress on the animal, except at a higher dose and in saturated lipid. Compared to the controls, per mouse, the average tumor size in the colon was larger in the B(a)P-treatment group. Genotoxic chemicals such as B(a)P act both as an initiator and a promoter of the carcinogenic process (Luch 2006), which may have contributed to the larger tumor size in the B(a)P-treatment groups. The majority of the intestinal tumors observed in ApcMin mice thus far in our study were typical exophytic adenomas; invasive adenocarcinomas were less frequent. Our findings were in agreement with those of Leclerc et al. (2004), who also rarely observed invasive adenomas in ApcMin mice.
It is apparent from the polyp incidence data in the jejunum and colon that B(a)P exposure increased the development of polyps in the saturated-fat group more so than in the unsaturated-fat group. For carcinogenic PAHs ingested through a lipid-rich diet, the intestine is the first sorting station where digestion, dispersion, and membrane and cytosolic transport occur before the lipids are packaged for delivery to the general circulation (Laher et al. 1984). In this context, it should be noted that the polyp numbers are a reflection of increased uptake of B(a)P depending on the vehicle of choice. The lipophilic nature of B(a)P (Librando, Sarpietro, and Castelli 2003) probably facilitates its absorption through the GI tract. Absorption of B(a)P from the GI tract of mice appears to be enhanced when it is administered via vehicles possessing lipophilic and hydrophobic properties (dietary fats) compared to controls, as demonstrated by Vetter, Carey, and Patton (1985) for benzo(a)pyrene, and Walker et al. (2007) and Harris, Hood, and Ramesh (2008) for fluoranthene. Also, a direct relationship between PAH absorption and fat absorption in higher mammals was demonstrated by the studies of Laurent et al. (2001, 2002) and Cavret et al. (2003). It is therefore conceivable that prolonged exposure to B(a)P through dietary fat contributes to an increased bioavail-ability of B(a)P and/or its metabolites in the GI tract and cause localized damage that is B(a)P dose dependent. How the kinetics of disposition of ingested B(a)P and subsequent accumulation of reactive metabolites and their residence time is modulated by dietary lipids is currently under investigation.
Dietary fat per se has high energy content per unit weight (Vogel et al. 2003) and has been shown to have a tumor-promoting effect (Guthrie and Carroll 1999; Reddy 1992) through increased lipid peroxidation (Hietanen et al. 1990), oxidative damage (Vogel et al. 2003), modulation of cell signaling pathways, and immune system function (Chapkin, McMurray, and Lupton 2007). These characteristics raise a question whether ApcMin mice can develop colon tumors when they are given unsaturated and saturated fat without B(a)P. We have encountered colon tumors in mice that received both types of fats; however, their numbers were statistically insignificant compared to the mice that received B(a)P through dietary fat. Our findings therefore strongly suggest that B(a)P in concert with dietary fat contributes to colon carcinogenesis.
The differences in polyp numbers between proximal and distal colon imply that tumor development is region and cell specific. These findings are interesting, given the fact that the dysplastic polyp load was predominantly in the distal colon, which is considered to be more comparable to the human colon (Chang, Chapkin, and Lupton 1997). The differential response in polyp numbers was assumed to be a consequence of increased apoptosis in the distal colon relative to the proximal colon (Hu et al. 2002). Predominance of sporadic tumors in the proximal colon were thought to result from microsatellite instability, whereas the ones in the distal colon were the result of chromosomal instability (Lindblom 2001). According to a model proposed by Breverick and Gaudemack (1999), the proximal tumors were attributed to methylating carcinogens and distal tumors were attributed to bulky-adduct–forming carcinogens. Relevant to this view are the published reports that B(a)P is capable of forming bulky DNA adducts in target tissues (Nesnow et al. 1993; Ramesh and Knuckles 2006). We hope to bring to light more information on the factors that underlie the differential distribution of polyps once our studies on spatial and temporal distribution of B(a)P-DNA adducts in the ApcMin mouse that received B(a)P in saturated versus unsaturated fat are completed.
To summarize, our findings suggest that the dose of B(a)P and the nature of dietary lipid (unsaturated vs. saturated) influence colon tumor development in the ApcMin mouse model. In the present study, colonic tumors were observed after short-term (nine weeks) administration of the carcinogen B(a)P through dietary fat. Dietary fat may have accelerated an early onset of colon carcinogenesis in ApcMin mice. Had we sacrificed a subset of mice every two weeks from all the controls and treatment groups, we would have been able to identify the specific window of colorectal neoplasia in these mice. Also, we did not prolong the exposure beyond the sixty-day period, as these mice have been reported to have a finite life span of 120 days (Moser et al. 1992). The mice were 102 days old when the exposure was terminated; any extension of the study may have resulted in deaths of mice owing to obstruction of the GI tract by polyps and lack of function (Moser et al. 1992), thus missing an opportunity to see whether B(a)P can cause colorectal tumors. At this time, we do not know whether extension of the period of exposure to B(a)P and dietary fat may have resulted in more diverse tumor phenotypes than the ones we had already observed. Nonetheless, our initial findings set the stage for additional studies to establish the conditions under which dietary fat in this mouse model enhances B(a)P induction of colon tumors in murine models. Studies on the mechanism of colon tumorigenesis as a result of B(a)P exposure through dietary fat are ongoing in our laboratory.
This research was supported by the National Institutes of Health (NIH) grants 1S11ES014156-01A1 (DBH, AR), 1RO3CA130112-01 (AR), 5T32HL007735-12 (Meharry) and 1F31ES017391-01 (DLH). The authors also thank Mr. Anthony Frazier, Lab Manager, Human Tissue Acquisition & Pathology Core of Vanderbilt-Ingram Cancer Research Center for his assistance with histopathology.