- Email: [email protected]
Reduction of aberrant crypt foci induced in rat colon with azoxymethane or methylnitrosourea by feeding cholic acid
Cancer Letters, 68 (1993) 15 - 23 Elsevier Scientific Publishers Ireland
15 Ltd.
Reduction of aberrant crypt foci induced in rat colon with azoxymethane or methylnitrosourea by feeding cholic acid Bernadene Department
(Received (Accepted
A. Magnuson of Foods
and Nutrition,
15 October 19 October
and Ranjana
P. Bird
University of Manitoba,
Recent studies in our laboratory have demonstrated that feeding cholic acid (CHA) to rats treated with a single dose of azoxymethane (AOMI reduces the growth of putative preneoplastic lesions, aberrant crypt foci (AU), in a dose-dependent manner [I]. This finding was unexpected since CHA has been reported to promote colon cancer in rats receiuing multiple treatments of the colon carcinogen, methylnitrosourea (MNU). The main objective of the present inuestigation was to eualuate the effect of the type of carcinogen and treatment protocol on the induction and growth ofACF in conjunction with CHA treatment. Male Sprague - Dawley rats receiued 0, 1 or 2 treatments with AOM or MNU and were fed either the AIN- 76A or AIN- 76A plus 0.2 % CHA diet for 4 weeks. The total number and auerage size oj ACF were significantly reduced in CHA-fed animals regardless of the type or number of treatments of carcinogen. The greatest reduction ofACF due to CHA-feeding was seen in the distal colon. The auerage crypt to: Bernadene A. Magnuson, Department of
Foods and Nutrition, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2. Abbreuiations: ACF, aberrant crypt foci; AHF, altered hepatic foci; ANOVA, analysis of variance; AOM, azoxymethane; CHA, cholic acid; CH, crypt height; MF, number of mitotic figures;
MNU, methylnitrosourea.
0304-3835/92/$05.00 Printed and Published
Manitoba R3T 2N2 (Canada)
1992) 1992)
Summary
Correspondence
Winnipeg,
0 1992 Elsevier Scientific Publishers in Ireland
multiplicity (number of crypts in each ACFj was not altered by diet or carcinogen treatment. Colonic cell proliferation (crypt height and number of mitotic figures) was significantly increased in CHA-fed animals compared to control diet animals. Therefore, feeding CHA for 4 weeks reduced the number and size of ACF in rats induced by 1 or 2 injections of AOM or MNU, despite stimulation of colonic cell proliferation. These findings suggest further investigation is needed to understand the mechanism of promotion by cholic acid and the value of number and growth characteristics of ACF as a biological endpoint in the pathogenesis of colon cancer.
Kqlwords: aberrant colon carcinogenesis
crypts;
cholic acid;
Introduction In 1987, Bird [2] proposed that aberrant crypt foci (ACF) identified in unsectioned colonic mucosa represented preneoplastic lesions of colon cancer and may be a useful measure of neoplastic events. Evidence is accumulating to support this hypothesis. ACF have been: induced specifically with colon carcinogens in a dose-dependent manner [3,4]; inhibited with a colon cancer inhibitor, disulfiram [5]; promoted by a high fat diet [3]; Ireland Ltd.
16
observed to display increased Iabelling index [6], nuclear atypia [7] and decreased hexosaminidase activity [8]. A higher number of ACF have recently been identified in human colonic mucosa of patients with colon cancer compared to patients without colon cancer
PSOI. It has been proposed that enumeration of the number and growth of ACF can be used to identify potential modulators of colon carcinogenesis. In order to use ACF as a biological endpoint in the study of colon carcinogenesis, it is important to evaluate their induction and growth characteristics employing various modulators of colon carcinogenesis and treatment protocols at different time intervals. Based on the contention that ACF are preneoplastic lesions, it was expected that their growth and number would be increased in the presence of a colonic tumor promoting diet containing bile acids. Bile acids have been implicated as promoters of colon cancer based on epidemiological [ 1 l] and animal model studies [12,13]. The addition of 0.2% cholic acid (CHA) a primary bile acid, to the diets of rats treated intrarectally with methylnitrosourea (MNU) resulted in increased colonic tumor incidence and multiplicity [ 14,151. However, Bird [l] recently reported that feeding 0.2% CHA to rats consistently resulted in a significantly reduced number and size of ACF. The discrepancy between the effect of CHA feeding on tumor incidence and ACF might be explained by differences in experimental protocols as Bird [l] utilized azoxymethane (AOM) in a one-injection protocol and introduced the CHA diet 1 week after carcinogen injection while McSherry et al. [14] and Cohen et al. [15] fed the CHA diet during multiple intrarectal installations of MNU. There are two differences in these protocols which may be responsible for the unexpected effect of CHA-feeding on ACF. AOM is an indirect carcinogen requiring metabolism in the liver, while MNU is a direct-acting carcinogen [ 161. Secondly, bile acids have been reported to be cocarcinogenic, ,enhancing the uptake of carcinogen into cells [ 171, therefore the effect of
feeding CHA during multiple injections versus following a single injection may differ. The objectives of this study were to determine if similar growth reduction of ACF would occur utilizing protocols which differed in the type of carcinogen and number of carcinogen treatments in rats fed CHA. As one mechanism of promotion by CHA that has been proposed is stimulation of cell proliferation [18], the effect of experimental protocol on cell proliferation measurements was also investigated. Materials
and Methods
Animals Male weanling Sprague - Dawley rats were obtained from Campus Breeding, Dept. of Animal Care, University of Manitoba, Winnipeg, Canada. Animals were housed 3 per cage in wire cages with a 12/12 h (light/dark) cycle. Temperature and humidity were controlled at 22OC and 50%, respectively. Animals were given laboratory chow and water ad Iibitum until initiation of the experiment. Carcinogens
As cholic acid feeding may affect weight gain of animals, carcinogens were administered on a body weight basis. MNU (Sigma Chemical Co., St. Louis, MO, USA) in sterile saline was instilled intrarectally at a dose of 40 mg/kg body weight. AOM (Sigma Chemical Co.) in sterile saline was injected i.p. at a dose of 20 mg/kg body weight. Animals in the single exposure group were given the carcinogen on day 1 only, animals in the multiple exposure group were given the above dose of carcinogen on day 1 and on day 4. Control animals were given sterile saline. Diets
Diets were formulated based on the composition of the AIN-76Adiet [19,20]. The control diet (CO) contained 200 g vitamin-free casein, 3 g DL-methionine, 500 g dextrose, 150 g cornstarch, 50 g corn oil, 50 g cellulose, 10 g AIN- vitamin mix, 35 g AIN- mineral
17
mix and 2 g choline bitartrate per kg diet. The cholic acid diet (CHA) was the AIN-76A diet plus 0.2% cholic acid by weight. Cholic acid (Sigma Chemical Co.) was approximately 98% pure as determined by the manufacturer.
crypt and lumen were visible, were used for Animals cell proliferation measurements. which displayed extensive mucosal damage preventing accurate measurements were eliminated.
Study design Animals were randomly assigned 6 per group to one of the following 10 experimental single dose MNU, CO diet groups: (MNUlCO); single dose MNU, CHA diet (MNUlCA); multiple dose MNU, CO diet (MNUZCO); multiple dose MNU, CHA diet (MNUZCA) ; single dose AOM, CO diet (AOMlCO); single dose AOM, CHA diet (AOMlCA), multiple dose AOM, CO diet (AOMZCO); multiple dose AOM, CHA diet (AOMZCA); control group, CHA diet (OCA) and control group CO diet (OCO). Animals were allowed to eat the designated diets ad libitum starting on day 1 for 4 weeks. Initial and weekly body weights were recorded.
Statistical analysis SAS statistical software for microcomputers was used for all statistical analysis. Two-way analysis of variance (ANOVA) procedure was used to analyze the effect of diet and carcinogen treatment on body weight, number of ACF, crypt height and number of mitotic figures. Duncan’s multiple range test was used to separate treatment means when ANOVA indicated significant differences. Log transformation of the number of ACF was conducted to increase the validity of ANOVA assumptions. A nested ANOVA design was used to analyze crypt multiplicity and size of ACF, due to large differences in number of ACF in treatment groups [21]. In all cases, a value of P I 0.05 was considered significant.
Aberrant
crypt
and
cell
proliferation
measurements
At the end of 4 weeks, animals were injected with colchicine (1 mg/kg body weight, Sigma Chemical Co.) 2 hours prior to killing by carbon dioxide asphyxiation. Colons were removed and prepared for enumeration as previously described [3,7]. The number, size, distribution and crypt multiplicity of ACF were determined for the entire length of the colon. To determine size, an ocular grid was used to measure the approximate area occupied by the ACF as viewed at 100 x magnification. To determine distribution, the number of ACF in every 2 cm starting from the rectal end were recorded. To determine crypt multiplicity, the number of crypts in each focus of AC was recorded. To determine cell proliferation, longitudinal sections from three Z-cm segments of the colon from the rectal end were processed for histology. Ten crypts per animal per segment were scored for the number of arrested metaphase cells and for crypt height. Only animals in the control and one injection groups, and only crypts in which the complete
Results Body weight Two-way ANOVA determined that the diet did not have an effect on body weight at any time point (data not shown). The effect of carcinogen treatment on body weight is shown in Table I. Throughout the experiment, body weights of animals receiving one injection of MNU (MNUl) were not significantly different from saline animals, whereas animals receiving 2 injections of MNU (MNUZ) had significantly lower body weights than saline animals, but were not significantly different from MNUl animals. In contrast, animals receiving one injection of AOM (AOMl) had significantly lower body weights than saline or MNUl animals at all time points. Body weights of animals given two injections of AOM (AOM2) were significantly lower than all other groups at all time points. Characteristics of ACF The effect of carcinogen treatment and diet
18 Table 1. Body weights” of animals receiving saline, AOM or MNU.
Saline MNUl MNU2 AOMl AOM2
Week 0
Week 1
Week 2
82.8 84.7 84.2 85.0 83.3
124.7 127.4 113.5 101.9 74.8
180.7 184.1 168.0 161.9 123.8
* 1.5” l 1.8” zt 1.6” zt 1.6” f 1.9”
l 4.6” zt 3.2” zt 2.8b +z 2.6’ f 4.3d
l
* zt zt zt
4.2a 4.9ab 4.1k 2.9’ 7.1d
Week 3
Final weight
243.8 249.7 227.3 222.6 181.5
301.1 310.6 282.4 281.6 239.1
zt 5.6” zt 6.9”b zt 5.6b” l 5.1’ f 8.7d
f zt f * l
9.1” 9.5”b 7.1b 5.9b 9.6’
“Grams; mean +ZS.E. Means within a column with different superscripts differ at P I 0.05.
on the total number of ACF is shown in Fig. 1. In all cases, rats fed 0.2% CHA had significantly (P I 0.0001) fewer ACF than rats fed the CO diet. There were no ACF in salinetreated animals. The effect of one versus two carcinogen treatments of total ACF was not consistent. Rats treated with two installations of MNU and fed either the CO or the CHA diet had higher total ACF than rats treated with only one installation, but the difference was not statistically significant. In AOM-treated animals, the effect of one versus two injections
varied with the diet group (Fig. 1). CO-fed animals had significantly more ACF when treated with one injection compared to two AOM injections and the reverse occurred in CHA-fed animals. Fewer ACF with a high versus low dose of AOM in rats has been reported previously [ZZ] and may be attributed to the toxicity of AOM which is evident in this experiment from the body weight data. Figure 2 illustrates the effect of protocol on the distribution of ACF throughout the colon. The greatest number of ACF were found in the
@
160’
co CA
120-’ t;
a 5 $ ? z’
loo-’ 80-’ 60-’
MNUI
MNU2 AOMl Carcinogen Treatment
AOM2
Fig. 1. The effect of carcinogen treatment and diet on total number of ACF in the colon. Treatments are described in Materials and Methods. In all treatment groups, CHA-fed rats had significantly lower total ACF compared to CO-fed rats (P 5 0.001).
19
first 6 cm from the rectum for both MNU- and AOM-treated animals fed the CO diet. Reduction in the number of ACF due to CHAfeeding occurred throughout the colon, but was greatest in the first 6 cm from the rectum, resulting in a more even distribution of ACF throughout the colon in CHA fed groups. The number of carcinogen treatments had little effect on distribution (Fig. 2). Neither the number of carcinogen treatments or diet had a significant effect on the crypt multiplicity after 4 weeks in AOM- or MNU-treated rats (Table II). In all carcinogen treatments the mean size of ACF was smaller in rats fed the CHA diet compared to rats fed the CO diet (Table III), however this difference reached statistical significance only in rats treated with AOM. The number of carcinogen treatments did not have an effect on the size of ACF after 4 weeks.
Cell proliferation measurements It was determined that only diet had a significant effect on cell proliferation measurements (Table IV). Crypt height (CH) and number of mitotic figures (MF) were significantly higher in CHA-fed compared to CO-fed animals in all segments of the colon (Table IV) in all treatment groups (data not shown). There was no significant difference in the CH or MF among saline- or AOM- or MNU-treated animals, except for a lower MF in MNU-treated animals in the segment of colon 4- 6 cm from the rectum. Discussion The main finding of this study is that feeding a diet containing 0.2% CHA to rats treated with a colon carcinogen reduces the number and size of ACF, regardless of the type of
m
b
25
a 3 5
20
e 3
10
AOM2CA
5
AOM2CO
(a)
AOMl CA AOMl CO
15
Distancefrom
rectum (cm)
III t; 50 a % $
40
2
20
*
10
MNUlCO
30
0
Fig. 2. described
lb)
MNUl CA
m MNU2CA MNU2CO l-2
2-4 5-6 7-8 9-10 Distance from rectum (cm)
The effect of carcinogen treatment in the Materials and Methods.
11-12
and diet on the distribution
of ACF throughout
the colon. Treatments
are
20 Table IL
Effect of carcinogen treatment and diet on crypt multiplicity of ACF”.
Diet
Number of treatments
Carcinogen MNU
co CHA co CHA
1 1 2 2
2.31 2.52 2.09 2.40
AOM f i f zt
0.06 0.18 0.04 0.11
(540) (140) (854) (165)
1.95 2.04 2.03 1.82
zt 0.04 f 0.10 zt 0.05 ztz0.06
(609) (122) (372) (219)
“Mean number of crypts/ACF f SE. (number of ACF evaluated). Crypt multiplicity of ACF in different dose and diet groups are not different (P Z+G 0.05).
Table III.
Diet
Effect of carcinogen treatment and diet on average size of ACF”. Number of treatments
Carcinogen MNU
co CHA co CHA
1 1 2 2
“Mean size x lo-‘mm2 P 5 0.05.
Table IV.
Diet
CHA CO Carcinogen AOM MNU Saline
4.33 3.35 3.85 3.18
AOM i zt f zt
0.22 0.24 0.08 0.12
(532)” (139)” (860)” (159)”
3.15 2.60 3.30 2.57
zt f f zt
0.09 0.13 0.12 0.10
(568)” (123)b (370)” (211)b
f S.E. (number of ACF evaluated). Means within a column with different superscripts differ
Effect of carcinogen and diet on cell proliferation. Nd
Mitotic figureb distance from rectum
Crypt height” distance from rectum’ o-2
2.1-4
4.1-60
o-2
2.1-4
4.1-6
16 16
43.6 +z 1.6’ 32.3 ztz0.6’
42.5 f 1.6’ 34.2 * 1.0’
43.8 * 1.4’ 36.8 zt 0.8’
12.9 f 1.4” 5.3 f 0.5’
12.5 * 1.1’ 6.9 zt 0.7’
15.1 i 1.3’ 6.6 zt 0.6’
10 11 12
38.1 zt 2.gg 39.2 f l.gg 36.6 l l.gg
40.1 zt 2.1g 38.4 f l.gg 37.1 zt 2.3g
40.7 zt 2.0g 40.5 zt 1.2g 39.8 * 2.0g
8.1 zt 1.6g 11.0 * 2.4g 8.6 * 1.6g
10.4 zt 1.8g 9.4 f 1.29 9.5 f 1.39
12.4 zt 1.7g 7.7 f 1.3 h 12.3 zt 2.0g
“Number of cells per mid-axial crypt. bNumber mitotic figures per crypt. ‘Distance in cm from rectal end of colon; all values are means f S.E. dNumber of animals scored, 10 crypts per animal per section of colon. e-hMeans within a column with different superscripts differ at P I 0.05.
21
cinogen or number of carcinogen treatments. Although these results corroborate previous findings [l], they are not consistent with our expectation of growth enhancement of preneoplastic lesions in the presence of tumor promoters. There are three possible explanations for this inconsistency: (1) ACF are not preneoplastic lesions, (2) CHA is not acting as a tumor promoter under our conditions, or (3) the number and/or size of ACF after 4 weeks is not a useful measure of tumor promotion. The possibility that ACF are not preneoplastic lesions seems remote considering the evidence to the contrary described in the introduction. We must question whether CHA was acting as a tumor promoter in our experiment. Cruse and colleagues [23] could find no evidence of promotion by intragastric instillation of CHA in rats treated with dimethylhydrazine for 20 weeks. The tumor promotional activity of CHA has been demonstrated by studies in which CHA, fed in the diet or administered intrarectally, increased the number and incidence of colon tumors in rats treated with MNU [14,15] or with methyl-N’-nitro-Nnitrosoguanidine [13]. The protocol for the MNU2 groups in our experiment was similar to the tumor-promoting protocol used by McSherry et al. [14] with the exception that Sprague - Dawley rats instead of Fisher F344 rats were used in this experiment. It is unlikely that this difference would account for the observed reduced number of ACF as Sprague - Dawley rats have been reported to be either equally [24] or more [25] susceptible to development of colon tumors compared to F344 rats. Therefore, although it may be possible that CHA was not acting as a colonic tumor promoter, it does not seem probable due to the similarities with previous tumor-promoting protocols. Lastly, it is possible that the number and/or size of ACF after 4 weeks is not a useful measure of promotional effects of a given treatment. In experimental hepatocellular carcinoma, the precursor lesions, altered hepatic
foci (AHF) , have been extensively studied and there is general agreement that the number of AHF represents the effectiveness of the initiation, the volume or total size of AHF represents the degree of promotion and the phenotype of the AHF characterizes the promotion [26]. As ACF are yet in the initial stages of investigation, the significance of their various characteristics in the multistep nature of colon carcinogenesis is not clearly established. Various investigators have employed ACF as a biological endpoint in their studies and have interpreted an increased number of ACF as evidence of promotion [3,27,28], or a decreased number as evidence of prevention [29] of colon carcinogenesis. Recently, Zhang and colleagues [30] have reported that a change in crypt multiplicity, or the average number of crypts per focus, of ACF measured after 100 days was associated with tumor promotion. Hardman and colleagues [31] measured ACF present at the time of tumor development (32 weeks) in rats given various dietary treatments and could find no correlation with tumor incidence and either number or crypt multiplicity of ACF. Also, Hardman and colleagues measured ACF in only 2 cm of the colon. The change in the distribution of ACF in CHA-fed compared to CO-fed animals in this study demonstrates that the effect of a dietary treatment on ACF may not be consistent throughout the colon. It must be noted that the time of promotion before measurement of ACF varied greatly in these studies. Possibly, the promotion by CHA is evident in ACF growth only at later time points. The ability of CHA diet to increase colonic cell proliferation has been suggested to be a possible mechanism for tumor promotion by CHA [18]. Interestingly, studies which have demonstrated a significant increase in cell proliferation due to CHA have not employed a carcinogen [18,32], whereas studies comparing colonic kinetics in carcinogen-treated animals have failed to find significant differences in CHA-fed compared to control groups [15,33]. In this study, CHA-feeding
22
resulted in significantly increased MF and CH in all 3 groups; saline, MNU and AOM. One possible reason for our ability to demonstrate increased cell proliferation in carcinogentreated animals is the lower dose and short duration of carcinogen treatment used in this experiment, as carcinogen treatment alone had no effect on cell proliferation after 4 weeks (Table IV). It is also noted that although intrarectal instillation of bile acids has been used in tumor studies demonstrating tumor promotion [12,13], the effect of pulsing high levels of bile acids versus the constant level achieved with feeding on cell proliferation has not been investigated. The finding that the number and size of ACF was reduced in the presence of enhanced cell proliferation raises some interesting questions. It may be speculated that perhaps only a small percentage of the ACF represent preneoplastic lesions with the potential to develop into carcinomas and that increased cell proliferation may enhance sloughing of ‘meaningless’ or benign ACF and selectively promoted remaining ACF. To test this possibility, a nested ANOVA (which accounts for variation due to differences in the number of ACF among groups) was used to compare the size and crypt multiplicity of ACF but no evidence of enhanced growth of the remaining ACF in CHA-fed animals could be demonstrated within the time period studied. If CHA is acting as a promoter in this protocol and we accept the hypothesis that ACF are preneoplastic lesions, then there must be other mechanisms of promotion of CHA, as clearly enhanced cell proliferation is not simply enhancing the growth of all ACF throughout the promotion period. Other mechanisms by which bile acids may play a role in colon including cancer have been postulated enhancement of lipid peroxidation [34], modulation of mucosal immune function [35] and modification of DNA [36]. In summary, feeding 0.2% CHA to rats receiving one or two treatments of either AOM or MNU resulted in the reduction of the size and number of ACF, putative preneoplastic le-
sions, in the colon after 4 weeks. These findings raise questions regarding the early predictive ability of ACF in the pathogenesis of colon cancer. However, it must be clearly established as to whether CHA is acting as a promoter with our protocol. A long-term study is underway in our laboratory to investigate the relationship between the effects of CHA on the growth of ACF in the early stages of colon carcinogenesis and eventual tumor incidence. Acknowledgments The financial support for this study was provided by the National Cancer Institute of Canada and the Natural Sciences and Engineering Council of Canada to R.P.B. The authors thank Dietlinde Tober and Valerie Netterfield for their technical assistance and Dr. Carl Schwartz for advice on the statistical analysis. Manitoba Health Research Council Studentship to B.A.M. is gratefully acknowledged. Reprint requests should be addressed to Dr. R.P. Bird, Department of Foods and Nutrition, University of Manitoba, Winnipeg, Manitoba, R3T 2N2, Canada. References Bird, R.P. (1991) Effect of cholic acid on the number and growth of aberrant crypt foci: putative preneoplastic lesions. Proc. Am. Assoc. Cancer Res., 32, 147. Bird, R.P. (1987) Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen : preliminary findings. Cancer Lett., 37, 147 - 151. McLellan, E.A. and Bird, R.P. (1988) Aberrant crypts: preneoplastic lesions in the murine colon. Cancer Res., 43, 6187-6192. Tudek, B., Bird, R.P. and Bruce, W.R. (1989) Foci of aberrant crypts in the colons of mice and rats exposed to carcinogens associated with foods. Cancer Res., 49, 1236 - 1240. McLellan, E.A. and Bid, R.P. (1991) Effect of disulfiram on 1,2-dimethyl-hydrazine and azoxymethane-induced aberrant crypt foci. Carcinogenesis, 12, 969 - 972. McLellan, E.A., Medline, A. and Bird, R.P. (1991) Dose response and proliferative characteristics of aberrant crypt foci: putative preneoplastic lesions in rat colon. Carcinogenesis, 12, 2093 - 2098. McLellan, E.A. and Bird, R.P. (1991) Sequential analyses of the growth and morphological characteristics of aberrant
23
8
9
10
11 12
13
14
15
16
17
18
19
20
21
22
crypt foci: putative preneoplastic lesions. Cancer Res., 51, 5270 - 5274. Pretlow, T.P., O’Riordan, M.A., Kolman, M.F. and Jurcisek, J.A. (1990) Colonic aberrant crypts in azoxymethane treated F344 rats have decreased hexosaminidase activity. Am. J. Pathol., 136, 13 - 16. Pretlow, T.P., Barrow, B.J., Ashton, W.S., O’Riordan, M.A., Pretlow, T.G., Jurcisek, J.A. and Stellato, T.A. (1991) Aberrant crypts: putative preneoplastic foci in human colonic mucosa. Cancer Res., 51, 1564- 1567. Roncucci, L., Stamp, D., Medline, A., Cullen, J.B. and Bruce, W.R. (1991) Identification and quantification of aberrant crypt foci and microadenomas in the human colon. Hum. Pathol., 22, 287 - 294. Hill, M.J. (1990) Bile flow and colon cancer. Mutat. Res., 238, 313-320. Reddy, B.S., Narasawa, T., Weisburger, J.H. and Wynder, E.L. (1976) Promoting effect of sodium deoxycholate on colon adenocarcinomas in germ-free rats. J. Natl. Cancer Inst., 56, 441-442. Reddy, B.S., Watanabe, K., Weisburger, J.H. and Wynder, E.L. (1977) Promoting effect of bile acids in colon carcinogenesis in germ-free and conventional F344 rats. Cancer Res., 37, 3238-3242. McSherry, C.K., Cohen, B.I., Bokkenheuser, V.D., Mosbach, E.H., Winter, J., Matoba, N. and Scholes, J. (1989) Effects of calcium and bile acid feeding on colon tumors in the rat. Cancer Res., 49, 6039-6043. Cohen, B.I., Raicht, R.F., Deschner, E.E., Takahashi, M., Sarwal, A.N. and Fazzini, E. (1980) Effect of cholic acid feeding on N-methyl-N-nitrosourea-induced colon tumors and cell kinetics in rats. J. Natl. Cancer Inst., 64, 573 - 578. Weisburger, J.H. (1971) Colon carcinogens: their metabolism and mechanism of action. Cancer, 28, 60 - 70. Sugezawa, A. and Kaibara, N. (1991) Possible mechanism for the cocarcinogenic effect of bile acids: increased intracellular uptake of methylcholanthrene by C3H/lOT1/2 fibroblasts in vitro. Oncology, 48, 138- 143. Deschner, E.E., Cohen, B.I. and Raicht, R.F. (1981) Acute and chronic effect of dietary cholic acid on colonic cell proliferation. Digestion, 21, 290 - 296. American Institute of Nutrition (1977) Report of the AlN ad hoc committee on standards for nutrition studies. J. Nutr., 107, 1340 - 1348. American Institute of Nutrition (1980) Second report of the ad hoc committee on standards for nutrition studies. J. Nutr., 110, 1726. Kleinbaum, D.G., Kupper, L.L. and Muller, K.E. (1988). Applied Regression Analysis and Other Multivariable Methods. 2nd Edn. PWS Kent Publishing Co., Boston. Corpet, D.E., Stamp, D., Medline, A., Minkin, S., Archer, M.C. and Bruce, W.R. (1990) Promotion of colonic microadenoma growth in mice and rats fed cooked sugar or cooked casein and fat. Cancer Res., 50, 6955- 6958.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Cruse, J.P., Lewin, M.R. and Clark, C.G. (1981) The effects of cholic acid and bile salt binding agents on 1,2 - dimethylhydrazine-induced colon carcinogenesis in the rat. Carcinogenesis, 2, 439 - 443. Nauss, K.M., Bueche, D. and Newbeme, P.H. (1987) Effect of beef fat on DMH-induced colon tumorigenesis: influences of rat strain and nutrient composition. J. Nutr., 117, 739-747. Nigro, N.D. (1985) Animal model for colorectal cancer. In: Carcinoma of the Large Bowel and Its Precursors, pp. 161- 173. AR Liss Inc., Place. Pitot, H.C. (1990) Altered hepatic foci: their role in murine hepatocarcinogenesis. Annu. Rev. Pharmacol. Toxicol., 30, 465-500. Kendall, C.W., Jane.&, S.A., Friday, D. and Rao, A.V. (1992) Dietary cholesterol enhances preneoplastic aberrant crypt formation and alters cell proliferation in the murine colon treated with azoxymethane. Nutr. Cancer, 17, 107 - 114. Shivapurkar, N., Tang, Z.C. and Alabaster, 0. (1992) The effect of high-risk and low-risk diets on aberrant crypt and colonic tumor formation in Fischer-344 rats. Carcinogenesis, 13, 887 - 890. Periera, M.A. and Khoury, M.D. (1991) Prevention by chemopreventive agents of azoxymethane-induced foci of aberrant crypts in rat colon. Cancer Lett., 61, 27 - 33. Zhang, X.-M., Stamp, D., Minkin, S., Medline, A., Corpet, D.E., Bruce, W.R. and Archer, M.C. (1992) Promotion of aberrant crypt foci and cancer in rat colon by thermolyzed protein. J. Natl. Cancer Inst., 84, 1026 - 1030. Hardman, W.E., Cameron, I.L., Heitman, D.W. and Contreras, E. (1991) Demonstration of the need for end point validation of putative biomarkers: failure of aberrant crypt foci to predict colon cancer incidence. Cancer Res., 51, 6388 - 6392. Robblee, N.M., McLellan, E.A. and Bird, R.P. (1989). Measurement of the proliferative status of colonic epithelium as a risk marker for colon carcinogenesis: effect of bile acid and dietary fibre. Nutr. Cancer, 12, 301- 310. Weidman, W.F., Deschner, E.E., Cohen, B.I. and DeCosse, J.J. (1985) Acute effects of dietary cholic acid and methylazoxymethanol acetate on colon epithelial cell proliferation; metabolism of bile salts and neutral sterols in conventional and germfree SD rats. J. Natl. Cancer Inst., 74, 665 - 670. DeLange, R.J. and Glazer, A.N. (1990) Bile acids: antioxidams or enhancers of peroxidation depending on lipid concentration. Arch. Biochem. Biophys., 276, 19 - 25. Elitsur, Y., Bull, A.W. and Luk. G.D. (1990) Modulation of human colonic lamina propria lymphocyte proliferation: effect of bile acids and oxidized fatty acids. Dig, Dis. Sci., 35, 212 - 220. Cheah, P.Y. and Bernstein, H. (1990) Modification of DNA by bile acids: a possible factor in the etiology of colon cancer. Cancer Len., 49, 207 - 210.
15 Ltd.
Reduction of aberrant crypt foci induced in rat colon with azoxymethane or methylnitrosourea by feeding cholic acid Bernadene Department
(Received (Accepted
A. Magnuson of Foods
and Nutrition,
15 October 19 October
and Ranjana
P. Bird
University of Manitoba,
Recent studies in our laboratory have demonstrated that feeding cholic acid (CHA) to rats treated with a single dose of azoxymethane (AOMI reduces the growth of putative preneoplastic lesions, aberrant crypt foci (AU), in a dose-dependent manner [I]. This finding was unexpected since CHA has been reported to promote colon cancer in rats receiuing multiple treatments of the colon carcinogen, methylnitrosourea (MNU). The main objective of the present inuestigation was to eualuate the effect of the type of carcinogen and treatment protocol on the induction and growth ofACF in conjunction with CHA treatment. Male Sprague - Dawley rats receiued 0, 1 or 2 treatments with AOM or MNU and were fed either the AIN- 76A or AIN- 76A plus 0.2 % CHA diet for 4 weeks. The total number and auerage size oj ACF were significantly reduced in CHA-fed animals regardless of the type or number of treatments of carcinogen. The greatest reduction ofACF due to CHA-feeding was seen in the distal colon. The auerage crypt to: Bernadene A. Magnuson, Department of
Foods and Nutrition, University of Manitoba, Winnipeg, Manitoba, Canada R3T 2N2. Abbreuiations: ACF, aberrant crypt foci; AHF, altered hepatic foci; ANOVA, analysis of variance; AOM, azoxymethane; CHA, cholic acid; CH, crypt height; MF, number of mitotic figures;
MNU, methylnitrosourea.
0304-3835/92/$05.00 Printed and Published
Manitoba R3T 2N2 (Canada)
1992) 1992)
Summary
Correspondence
Winnipeg,
0 1992 Elsevier Scientific Publishers in Ireland
multiplicity (number of crypts in each ACFj was not altered by diet or carcinogen treatment. Colonic cell proliferation (crypt height and number of mitotic figures) was significantly increased in CHA-fed animals compared to control diet animals. Therefore, feeding CHA for 4 weeks reduced the number and size of ACF in rats induced by 1 or 2 injections of AOM or MNU, despite stimulation of colonic cell proliferation. These findings suggest further investigation is needed to understand the mechanism of promotion by cholic acid and the value of number and growth characteristics of ACF as a biological endpoint in the pathogenesis of colon cancer.
Kqlwords: aberrant colon carcinogenesis
crypts;
cholic acid;
Introduction In 1987, Bird [2] proposed that aberrant crypt foci (ACF) identified in unsectioned colonic mucosa represented preneoplastic lesions of colon cancer and may be a useful measure of neoplastic events. Evidence is accumulating to support this hypothesis. ACF have been: induced specifically with colon carcinogens in a dose-dependent manner [3,4]; inhibited with a colon cancer inhibitor, disulfiram [5]; promoted by a high fat diet [3]; Ireland Ltd.
16
observed to display increased Iabelling index [6], nuclear atypia [7] and decreased hexosaminidase activity [8]. A higher number of ACF have recently been identified in human colonic mucosa of patients with colon cancer compared to patients without colon cancer
PSOI. It has been proposed that enumeration of the number and growth of ACF can be used to identify potential modulators of colon carcinogenesis. In order to use ACF as a biological endpoint in the study of colon carcinogenesis, it is important to evaluate their induction and growth characteristics employing various modulators of colon carcinogenesis and treatment protocols at different time intervals. Based on the contention that ACF are preneoplastic lesions, it was expected that their growth and number would be increased in the presence of a colonic tumor promoting diet containing bile acids. Bile acids have been implicated as promoters of colon cancer based on epidemiological [ 1 l] and animal model studies [12,13]. The addition of 0.2% cholic acid (CHA) a primary bile acid, to the diets of rats treated intrarectally with methylnitrosourea (MNU) resulted in increased colonic tumor incidence and multiplicity [ 14,151. However, Bird [l] recently reported that feeding 0.2% CHA to rats consistently resulted in a significantly reduced number and size of ACF. The discrepancy between the effect of CHA feeding on tumor incidence and ACF might be explained by differences in experimental protocols as Bird [l] utilized azoxymethane (AOM) in a one-injection protocol and introduced the CHA diet 1 week after carcinogen injection while McSherry et al. [14] and Cohen et al. [15] fed the CHA diet during multiple intrarectal installations of MNU. There are two differences in these protocols which may be responsible for the unexpected effect of CHA-feeding on ACF. AOM is an indirect carcinogen requiring metabolism in the liver, while MNU is a direct-acting carcinogen [ 161. Secondly, bile acids have been reported to be cocarcinogenic, ,enhancing the uptake of carcinogen into cells [ 171, therefore the effect of
feeding CHA during multiple injections versus following a single injection may differ. The objectives of this study were to determine if similar growth reduction of ACF would occur utilizing protocols which differed in the type of carcinogen and number of carcinogen treatments in rats fed CHA. As one mechanism of promotion by CHA that has been proposed is stimulation of cell proliferation [18], the effect of experimental protocol on cell proliferation measurements was also investigated. Materials
and Methods
Animals Male weanling Sprague - Dawley rats were obtained from Campus Breeding, Dept. of Animal Care, University of Manitoba, Winnipeg, Canada. Animals were housed 3 per cage in wire cages with a 12/12 h (light/dark) cycle. Temperature and humidity were controlled at 22OC and 50%, respectively. Animals were given laboratory chow and water ad Iibitum until initiation of the experiment. Carcinogens
As cholic acid feeding may affect weight gain of animals, carcinogens were administered on a body weight basis. MNU (Sigma Chemical Co., St. Louis, MO, USA) in sterile saline was instilled intrarectally at a dose of 40 mg/kg body weight. AOM (Sigma Chemical Co.) in sterile saline was injected i.p. at a dose of 20 mg/kg body weight. Animals in the single exposure group were given the carcinogen on day 1 only, animals in the multiple exposure group were given the above dose of carcinogen on day 1 and on day 4. Control animals were given sterile saline. Diets
Diets were formulated based on the composition of the AIN-76Adiet [19,20]. The control diet (CO) contained 200 g vitamin-free casein, 3 g DL-methionine, 500 g dextrose, 150 g cornstarch, 50 g corn oil, 50 g cellulose, 10 g AIN- vitamin mix, 35 g AIN- mineral
17
mix and 2 g choline bitartrate per kg diet. The cholic acid diet (CHA) was the AIN-76A diet plus 0.2% cholic acid by weight. Cholic acid (Sigma Chemical Co.) was approximately 98% pure as determined by the manufacturer.
crypt and lumen were visible, were used for Animals cell proliferation measurements. which displayed extensive mucosal damage preventing accurate measurements were eliminated.
Study design Animals were randomly assigned 6 per group to one of the following 10 experimental single dose MNU, CO diet groups: (MNUlCO); single dose MNU, CHA diet (MNUlCA); multiple dose MNU, CO diet (MNUZCO); multiple dose MNU, CHA diet (MNUZCA) ; single dose AOM, CO diet (AOMlCO); single dose AOM, CHA diet (AOMlCA), multiple dose AOM, CO diet (AOMZCO); multiple dose AOM, CHA diet (AOMZCA); control group, CHA diet (OCA) and control group CO diet (OCO). Animals were allowed to eat the designated diets ad libitum starting on day 1 for 4 weeks. Initial and weekly body weights were recorded.
Statistical analysis SAS statistical software for microcomputers was used for all statistical analysis. Two-way analysis of variance (ANOVA) procedure was used to analyze the effect of diet and carcinogen treatment on body weight, number of ACF, crypt height and number of mitotic figures. Duncan’s multiple range test was used to separate treatment means when ANOVA indicated significant differences. Log transformation of the number of ACF was conducted to increase the validity of ANOVA assumptions. A nested ANOVA design was used to analyze crypt multiplicity and size of ACF, due to large differences in number of ACF in treatment groups [21]. In all cases, a value of P I 0.05 was considered significant.
Aberrant
crypt
and
cell
proliferation
measurements
At the end of 4 weeks, animals were injected with colchicine (1 mg/kg body weight, Sigma Chemical Co.) 2 hours prior to killing by carbon dioxide asphyxiation. Colons were removed and prepared for enumeration as previously described [3,7]. The number, size, distribution and crypt multiplicity of ACF were determined for the entire length of the colon. To determine size, an ocular grid was used to measure the approximate area occupied by the ACF as viewed at 100 x magnification. To determine distribution, the number of ACF in every 2 cm starting from the rectal end were recorded. To determine crypt multiplicity, the number of crypts in each focus of AC was recorded. To determine cell proliferation, longitudinal sections from three Z-cm segments of the colon from the rectal end were processed for histology. Ten crypts per animal per segment were scored for the number of arrested metaphase cells and for crypt height. Only animals in the control and one injection groups, and only crypts in which the complete
Results Body weight Two-way ANOVA determined that the diet did not have an effect on body weight at any time point (data not shown). The effect of carcinogen treatment on body weight is shown in Table I. Throughout the experiment, body weights of animals receiving one injection of MNU (MNUl) were not significantly different from saline animals, whereas animals receiving 2 injections of MNU (MNUZ) had significantly lower body weights than saline animals, but were not significantly different from MNUl animals. In contrast, animals receiving one injection of AOM (AOMl) had significantly lower body weights than saline or MNUl animals at all time points. Body weights of animals given two injections of AOM (AOM2) were significantly lower than all other groups at all time points. Characteristics of ACF The effect of carcinogen treatment and diet
18 Table 1. Body weights” of animals receiving saline, AOM or MNU.
Saline MNUl MNU2 AOMl AOM2
Week 0
Week 1
Week 2
82.8 84.7 84.2 85.0 83.3
124.7 127.4 113.5 101.9 74.8
180.7 184.1 168.0 161.9 123.8
* 1.5” l 1.8” zt 1.6” zt 1.6” f 1.9”
l 4.6” zt 3.2” zt 2.8b +z 2.6’ f 4.3d
l
* zt zt zt
4.2a 4.9ab 4.1k 2.9’ 7.1d
Week 3
Final weight
243.8 249.7 227.3 222.6 181.5
301.1 310.6 282.4 281.6 239.1
zt 5.6” zt 6.9”b zt 5.6b” l 5.1’ f 8.7d
f zt f * l
9.1” 9.5”b 7.1b 5.9b 9.6’
“Grams; mean +ZS.E. Means within a column with different superscripts differ at P I 0.05.
on the total number of ACF is shown in Fig. 1. In all cases, rats fed 0.2% CHA had significantly (P I 0.0001) fewer ACF than rats fed the CO diet. There were no ACF in salinetreated animals. The effect of one versus two carcinogen treatments of total ACF was not consistent. Rats treated with two installations of MNU and fed either the CO or the CHA diet had higher total ACF than rats treated with only one installation, but the difference was not statistically significant. In AOM-treated animals, the effect of one versus two injections
varied with the diet group (Fig. 1). CO-fed animals had significantly more ACF when treated with one injection compared to two AOM injections and the reverse occurred in CHA-fed animals. Fewer ACF with a high versus low dose of AOM in rats has been reported previously [ZZ] and may be attributed to the toxicity of AOM which is evident in this experiment from the body weight data. Figure 2 illustrates the effect of protocol on the distribution of ACF throughout the colon. The greatest number of ACF were found in the
@
160’
co CA
120-’ t;
a 5 $ ? z’
loo-’ 80-’ 60-’
MNUI
MNU2 AOMl Carcinogen Treatment
AOM2
Fig. 1. The effect of carcinogen treatment and diet on total number of ACF in the colon. Treatments are described in Materials and Methods. In all treatment groups, CHA-fed rats had significantly lower total ACF compared to CO-fed rats (P 5 0.001).
19
first 6 cm from the rectum for both MNU- and AOM-treated animals fed the CO diet. Reduction in the number of ACF due to CHAfeeding occurred throughout the colon, but was greatest in the first 6 cm from the rectum, resulting in a more even distribution of ACF throughout the colon in CHA fed groups. The number of carcinogen treatments had little effect on distribution (Fig. 2). Neither the number of carcinogen treatments or diet had a significant effect on the crypt multiplicity after 4 weeks in AOM- or MNU-treated rats (Table II). In all carcinogen treatments the mean size of ACF was smaller in rats fed the CHA diet compared to rats fed the CO diet (Table III), however this difference reached statistical significance only in rats treated with AOM. The number of carcinogen treatments did not have an effect on the size of ACF after 4 weeks.
Cell proliferation measurements It was determined that only diet had a significant effect on cell proliferation measurements (Table IV). Crypt height (CH) and number of mitotic figures (MF) were significantly higher in CHA-fed compared to CO-fed animals in all segments of the colon (Table IV) in all treatment groups (data not shown). There was no significant difference in the CH or MF among saline- or AOM- or MNU-treated animals, except for a lower MF in MNU-treated animals in the segment of colon 4- 6 cm from the rectum. Discussion The main finding of this study is that feeding a diet containing 0.2% CHA to rats treated with a colon carcinogen reduces the number and size of ACF, regardless of the type of
m
b
25
a 3 5
20
e 3
10
AOM2CA
5
AOM2CO
(a)
AOMl CA AOMl CO
15
Distancefrom
rectum (cm)
III t; 50 a % $
40
2
20
*
10
MNUlCO
30
0
Fig. 2. described
lb)
MNUl CA
m MNU2CA MNU2CO l-2
2-4 5-6 7-8 9-10 Distance from rectum (cm)
The effect of carcinogen treatment in the Materials and Methods.
11-12
and diet on the distribution
of ACF throughout
the colon. Treatments
are
20 Table IL
Effect of carcinogen treatment and diet on crypt multiplicity of ACF”.
Diet
Number of treatments
Carcinogen MNU
co CHA co CHA
1 1 2 2
2.31 2.52 2.09 2.40
AOM f i f zt
0.06 0.18 0.04 0.11
(540) (140) (854) (165)
1.95 2.04 2.03 1.82
zt 0.04 f 0.10 zt 0.05 ztz0.06
(609) (122) (372) (219)
“Mean number of crypts/ACF f SE. (number of ACF evaluated). Crypt multiplicity of ACF in different dose and diet groups are not different (P Z+G 0.05).
Table III.
Diet
Effect of carcinogen treatment and diet on average size of ACF”. Number of treatments
Carcinogen MNU
co CHA co CHA
1 1 2 2
“Mean size x lo-‘mm2 P 5 0.05.
Table IV.
Diet
CHA CO Carcinogen AOM MNU Saline
4.33 3.35 3.85 3.18
AOM i zt f zt
0.22 0.24 0.08 0.12
(532)” (139)” (860)” (159)”
3.15 2.60 3.30 2.57
zt f f zt
0.09 0.13 0.12 0.10
(568)” (123)b (370)” (211)b
f S.E. (number of ACF evaluated). Means within a column with different superscripts differ
Effect of carcinogen and diet on cell proliferation. Nd
Mitotic figureb distance from rectum
Crypt height” distance from rectum’ o-2
2.1-4
4.1-60
o-2
2.1-4
4.1-6
16 16
43.6 +z 1.6’ 32.3 ztz0.6’
42.5 f 1.6’ 34.2 * 1.0’
43.8 * 1.4’ 36.8 zt 0.8’
12.9 f 1.4” 5.3 f 0.5’
12.5 * 1.1’ 6.9 zt 0.7’
15.1 i 1.3’ 6.6 zt 0.6’
10 11 12
38.1 zt 2.gg 39.2 f l.gg 36.6 l l.gg
40.1 zt 2.1g 38.4 f l.gg 37.1 zt 2.3g
40.7 zt 2.0g 40.5 zt 1.2g 39.8 * 2.0g
8.1 zt 1.6g 11.0 * 2.4g 8.6 * 1.6g
10.4 zt 1.8g 9.4 f 1.29 9.5 f 1.39
12.4 zt 1.7g 7.7 f 1.3 h 12.3 zt 2.0g
“Number of cells per mid-axial crypt. bNumber mitotic figures per crypt. ‘Distance in cm from rectal end of colon; all values are means f S.E. dNumber of animals scored, 10 crypts per animal per section of colon. e-hMeans within a column with different superscripts differ at P I 0.05.
21
cinogen or number of carcinogen treatments. Although these results corroborate previous findings [l], they are not consistent with our expectation of growth enhancement of preneoplastic lesions in the presence of tumor promoters. There are three possible explanations for this inconsistency: (1) ACF are not preneoplastic lesions, (2) CHA is not acting as a tumor promoter under our conditions, or (3) the number and/or size of ACF after 4 weeks is not a useful measure of tumor promotion. The possibility that ACF are not preneoplastic lesions seems remote considering the evidence to the contrary described in the introduction. We must question whether CHA was acting as a tumor promoter in our experiment. Cruse and colleagues [23] could find no evidence of promotion by intragastric instillation of CHA in rats treated with dimethylhydrazine for 20 weeks. The tumor promotional activity of CHA has been demonstrated by studies in which CHA, fed in the diet or administered intrarectally, increased the number and incidence of colon tumors in rats treated with MNU [14,15] or with methyl-N’-nitro-Nnitrosoguanidine [13]. The protocol for the MNU2 groups in our experiment was similar to the tumor-promoting protocol used by McSherry et al. [14] with the exception that Sprague - Dawley rats instead of Fisher F344 rats were used in this experiment. It is unlikely that this difference would account for the observed reduced number of ACF as Sprague - Dawley rats have been reported to be either equally [24] or more [25] susceptible to development of colon tumors compared to F344 rats. Therefore, although it may be possible that CHA was not acting as a colonic tumor promoter, it does not seem probable due to the similarities with previous tumor-promoting protocols. Lastly, it is possible that the number and/or size of ACF after 4 weeks is not a useful measure of promotional effects of a given treatment. In experimental hepatocellular carcinoma, the precursor lesions, altered hepatic
foci (AHF) , have been extensively studied and there is general agreement that the number of AHF represents the effectiveness of the initiation, the volume or total size of AHF represents the degree of promotion and the phenotype of the AHF characterizes the promotion [26]. As ACF are yet in the initial stages of investigation, the significance of their various characteristics in the multistep nature of colon carcinogenesis is not clearly established. Various investigators have employed ACF as a biological endpoint in their studies and have interpreted an increased number of ACF as evidence of promotion [3,27,28], or a decreased number as evidence of prevention [29] of colon carcinogenesis. Recently, Zhang and colleagues [30] have reported that a change in crypt multiplicity, or the average number of crypts per focus, of ACF measured after 100 days was associated with tumor promotion. Hardman and colleagues [31] measured ACF present at the time of tumor development (32 weeks) in rats given various dietary treatments and could find no correlation with tumor incidence and either number or crypt multiplicity of ACF. Also, Hardman and colleagues measured ACF in only 2 cm of the colon. The change in the distribution of ACF in CHA-fed compared to CO-fed animals in this study demonstrates that the effect of a dietary treatment on ACF may not be consistent throughout the colon. It must be noted that the time of promotion before measurement of ACF varied greatly in these studies. Possibly, the promotion by CHA is evident in ACF growth only at later time points. The ability of CHA diet to increase colonic cell proliferation has been suggested to be a possible mechanism for tumor promotion by CHA [18]. Interestingly, studies which have demonstrated a significant increase in cell proliferation due to CHA have not employed a carcinogen [18,32], whereas studies comparing colonic kinetics in carcinogen-treated animals have failed to find significant differences in CHA-fed compared to control groups [15,33]. In this study, CHA-feeding
22
resulted in significantly increased MF and CH in all 3 groups; saline, MNU and AOM. One possible reason for our ability to demonstrate increased cell proliferation in carcinogentreated animals is the lower dose and short duration of carcinogen treatment used in this experiment, as carcinogen treatment alone had no effect on cell proliferation after 4 weeks (Table IV). It is also noted that although intrarectal instillation of bile acids has been used in tumor studies demonstrating tumor promotion [12,13], the effect of pulsing high levels of bile acids versus the constant level achieved with feeding on cell proliferation has not been investigated. The finding that the number and size of ACF was reduced in the presence of enhanced cell proliferation raises some interesting questions. It may be speculated that perhaps only a small percentage of the ACF represent preneoplastic lesions with the potential to develop into carcinomas and that increased cell proliferation may enhance sloughing of ‘meaningless’ or benign ACF and selectively promoted remaining ACF. To test this possibility, a nested ANOVA (which accounts for variation due to differences in the number of ACF among groups) was used to compare the size and crypt multiplicity of ACF but no evidence of enhanced growth of the remaining ACF in CHA-fed animals could be demonstrated within the time period studied. If CHA is acting as a promoter in this protocol and we accept the hypothesis that ACF are preneoplastic lesions, then there must be other mechanisms of promotion of CHA, as clearly enhanced cell proliferation is not simply enhancing the growth of all ACF throughout the promotion period. Other mechanisms by which bile acids may play a role in colon including cancer have been postulated enhancement of lipid peroxidation [34], modulation of mucosal immune function [35] and modification of DNA [36]. In summary, feeding 0.2% CHA to rats receiving one or two treatments of either AOM or MNU resulted in the reduction of the size and number of ACF, putative preneoplastic le-
sions, in the colon after 4 weeks. These findings raise questions regarding the early predictive ability of ACF in the pathogenesis of colon cancer. However, it must be clearly established as to whether CHA is acting as a promoter with our protocol. A long-term study is underway in our laboratory to investigate the relationship between the effects of CHA on the growth of ACF in the early stages of colon carcinogenesis and eventual tumor incidence. Acknowledgments The financial support for this study was provided by the National Cancer Institute of Canada and the Natural Sciences and Engineering Council of Canada to R.P.B. The authors thank Dietlinde Tober and Valerie Netterfield for their technical assistance and Dr. Carl Schwartz for advice on the statistical analysis. Manitoba Health Research Council Studentship to B.A.M. is gratefully acknowledged. Reprint requests should be addressed to Dr. R.P. Bird, Department of Foods and Nutrition, University of Manitoba, Winnipeg, Manitoba, R3T 2N2, Canada. References Bird, R.P. (1991) Effect of cholic acid on the number and growth of aberrant crypt foci: putative preneoplastic lesions. Proc. Am. Assoc. Cancer Res., 32, 147. Bird, R.P. (1987) Observation and quantification of aberrant crypts in the murine colon treated with a colon carcinogen : preliminary findings. Cancer Lett., 37, 147 - 151. McLellan, E.A. and Bird, R.P. (1988) Aberrant crypts: preneoplastic lesions in the murine colon. Cancer Res., 43, 6187-6192. Tudek, B., Bird, R.P. and Bruce, W.R. (1989) Foci of aberrant crypts in the colons of mice and rats exposed to carcinogens associated with foods. Cancer Res., 49, 1236 - 1240. McLellan, E.A. and Bid, R.P. (1991) Effect of disulfiram on 1,2-dimethyl-hydrazine and azoxymethane-induced aberrant crypt foci. Carcinogenesis, 12, 969 - 972. McLellan, E.A., Medline, A. and Bird, R.P. (1991) Dose response and proliferative characteristics of aberrant crypt foci: putative preneoplastic lesions in rat colon. Carcinogenesis, 12, 2093 - 2098. McLellan, E.A. and Bird, R.P. (1991) Sequential analyses of the growth and morphological characteristics of aberrant
23
8
9
10
11 12
13
14
15
16
17
18
19
20
21
22
crypt foci: putative preneoplastic lesions. Cancer Res., 51, 5270 - 5274. Pretlow, T.P., O’Riordan, M.A., Kolman, M.F. and Jurcisek, J.A. (1990) Colonic aberrant crypts in azoxymethane treated F344 rats have decreased hexosaminidase activity. Am. J. Pathol., 136, 13 - 16. Pretlow, T.P., Barrow, B.J., Ashton, W.S., O’Riordan, M.A., Pretlow, T.G., Jurcisek, J.A. and Stellato, T.A. (1991) Aberrant crypts: putative preneoplastic foci in human colonic mucosa. Cancer Res., 51, 1564- 1567. Roncucci, L., Stamp, D., Medline, A., Cullen, J.B. and Bruce, W.R. (1991) Identification and quantification of aberrant crypt foci and microadenomas in the human colon. Hum. Pathol., 22, 287 - 294. Hill, M.J. (1990) Bile flow and colon cancer. Mutat. Res., 238, 313-320. Reddy, B.S., Narasawa, T., Weisburger, J.H. and Wynder, E.L. (1976) Promoting effect of sodium deoxycholate on colon adenocarcinomas in germ-free rats. J. Natl. Cancer Inst., 56, 441-442. Reddy, B.S., Watanabe, K., Weisburger, J.H. and Wynder, E.L. (1977) Promoting effect of bile acids in colon carcinogenesis in germ-free and conventional F344 rats. Cancer Res., 37, 3238-3242. McSherry, C.K., Cohen, B.I., Bokkenheuser, V.D., Mosbach, E.H., Winter, J., Matoba, N. and Scholes, J. (1989) Effects of calcium and bile acid feeding on colon tumors in the rat. Cancer Res., 49, 6039-6043. Cohen, B.I., Raicht, R.F., Deschner, E.E., Takahashi, M., Sarwal, A.N. and Fazzini, E. (1980) Effect of cholic acid feeding on N-methyl-N-nitrosourea-induced colon tumors and cell kinetics in rats. J. Natl. Cancer Inst., 64, 573 - 578. Weisburger, J.H. (1971) Colon carcinogens: their metabolism and mechanism of action. Cancer, 28, 60 - 70. Sugezawa, A. and Kaibara, N. (1991) Possible mechanism for the cocarcinogenic effect of bile acids: increased intracellular uptake of methylcholanthrene by C3H/lOT1/2 fibroblasts in vitro. Oncology, 48, 138- 143. Deschner, E.E., Cohen, B.I. and Raicht, R.F. (1981) Acute and chronic effect of dietary cholic acid on colonic cell proliferation. Digestion, 21, 290 - 296. American Institute of Nutrition (1977) Report of the AlN ad hoc committee on standards for nutrition studies. J. Nutr., 107, 1340 - 1348. American Institute of Nutrition (1980) Second report of the ad hoc committee on standards for nutrition studies. J. Nutr., 110, 1726. Kleinbaum, D.G., Kupper, L.L. and Muller, K.E. (1988). Applied Regression Analysis and Other Multivariable Methods. 2nd Edn. PWS Kent Publishing Co., Boston. Corpet, D.E., Stamp, D., Medline, A., Minkin, S., Archer, M.C. and Bruce, W.R. (1990) Promotion of colonic microadenoma growth in mice and rats fed cooked sugar or cooked casein and fat. Cancer Res., 50, 6955- 6958.
23
24
25
26
27
28
29
30
31
32
33
34
35
36
Cruse, J.P., Lewin, M.R. and Clark, C.G. (1981) The effects of cholic acid and bile salt binding agents on 1,2 - dimethylhydrazine-induced colon carcinogenesis in the rat. Carcinogenesis, 2, 439 - 443. Nauss, K.M., Bueche, D. and Newbeme, P.H. (1987) Effect of beef fat on DMH-induced colon tumorigenesis: influences of rat strain and nutrient composition. J. Nutr., 117, 739-747. Nigro, N.D. (1985) Animal model for colorectal cancer. In: Carcinoma of the Large Bowel and Its Precursors, pp. 161- 173. AR Liss Inc., Place. Pitot, H.C. (1990) Altered hepatic foci: their role in murine hepatocarcinogenesis. Annu. Rev. Pharmacol. Toxicol., 30, 465-500. Kendall, C.W., Jane.&, S.A., Friday, D. and Rao, A.V. (1992) Dietary cholesterol enhances preneoplastic aberrant crypt formation and alters cell proliferation in the murine colon treated with azoxymethane. Nutr. Cancer, 17, 107 - 114. Shivapurkar, N., Tang, Z.C. and Alabaster, 0. (1992) The effect of high-risk and low-risk diets on aberrant crypt and colonic tumor formation in Fischer-344 rats. Carcinogenesis, 13, 887 - 890. Periera, M.A. and Khoury, M.D. (1991) Prevention by chemopreventive agents of azoxymethane-induced foci of aberrant crypts in rat colon. Cancer Lett., 61, 27 - 33. Zhang, X.-M., Stamp, D., Minkin, S., Medline, A., Corpet, D.E., Bruce, W.R. and Archer, M.C. (1992) Promotion of aberrant crypt foci and cancer in rat colon by thermolyzed protein. J. Natl. Cancer Inst., 84, 1026 - 1030. Hardman, W.E., Cameron, I.L., Heitman, D.W. and Contreras, E. (1991) Demonstration of the need for end point validation of putative biomarkers: failure of aberrant crypt foci to predict colon cancer incidence. Cancer Res., 51, 6388 - 6392. Robblee, N.M., McLellan, E.A. and Bird, R.P. (1989). Measurement of the proliferative status of colonic epithelium as a risk marker for colon carcinogenesis: effect of bile acid and dietary fibre. Nutr. Cancer, 12, 301- 310. Weidman, W.F., Deschner, E.E., Cohen, B.I. and DeCosse, J.J. (1985) Acute effects of dietary cholic acid and methylazoxymethanol acetate on colon epithelial cell proliferation; metabolism of bile salts and neutral sterols in conventional and germfree SD rats. J. Natl. Cancer Inst., 74, 665 - 670. DeLange, R.J. and Glazer, A.N. (1990) Bile acids: antioxidams or enhancers of peroxidation depending on lipid concentration. Arch. Biochem. Biophys., 276, 19 - 25. Elitsur, Y., Bull, A.W. and Luk. G.D. (1990) Modulation of human colonic lamina propria lymphocyte proliferation: effect of bile acids and oxidized fatty acids. Dig, Dis. Sci., 35, 212 - 220. Cheah, P.Y. and Bernstein, H. (1990) Modification of DNA by bile acids: a possible factor in the etiology of colon cancer. Cancer Len., 49, 207 - 210.