Clastogenic activity of bile acids and organic acid fractions of human feces

317

Cancer Letters, 15 (1982) 317-327 Elsevier/North-Holland Scientific Publishers Ltd.

CLASTOGENIC ACTIVITY OF BILE ACIDS AND ORGANIC FRACTIONS OF HUMAN FECES

WILLIAM S. BARNES’**

ACID

and WILLIAM D. POWRIEb

aEnvironmental Carcinogenesis Unit, British Columbia Cancer Research Centre, West 10th Avenue, Vancouver, B.C. V5.Z IL3 and bDepartment of Food Science, University of British Columbia, Vancouver, B.C. V6T 1 W5 (Canada)

601 The

(Received 2 November 1981) (Accepted 9 December 1981)

SUMMARY

Chloroform extracts of fecal material from 4 subjects on normal mixed western diets were fractionated to obtain an acid fraction and a hexane extract containing neutrals and bases. The acid fraction from at least 2 of the donors induced an elevated frequency of chromosomal aberrations and exchanges in Chinese hamster ovary (CHO) cells. Since acid steroids are expected to be present in the acid fraction, 5 bile acids were assayed for clastogenic activity in CHO cells. Ursodeoxycholic acid induced chromosomal aberrations and exchanges, and this effect was enhanced by the addition of a microsomal S9 mix. However, the enhancement is probably due to physical factors rather than to enzymatic activity.

INTRODUCTION

The incidence of colon cancer varies geographically throughout the world, and consequently this disease is presumed to be environmentally related. Epidemiological studies have indicated an association between colon cancer and total dietary fat and the consumption of meat, particularly beef [ 20,211. Because of the implication of dietary fat in colon carcinogenesis, several researchers have proposed that metabolism of bile acids by fecal microflora may also play a role in cancer etiology [ 2,10-131. In order to clarify the factors contributing to colon cancer, a number of workers have investigated the mutagenic activity of extracts from human fecal material. Ether-extractable material, which was directly mutagenic

*Present address: Naylor Dana Institute for Disease Prevention, Foundation, 1 Dana Road, Valhalla, New York 10595, U.S.A. 0304-3835/82/0000-0000/$02.75 o 1982 Elsevier/North-Holland

Scientific Publishers Ltd.

American Health

318

in the Ames test, has been isolated and characterized [ 4,5,31]. Using a modified extraction procedure, Reddy et al. [23] showed that a similar fraction enhanced the mutagenic activity of 2-acetylaminofluorene and N-methyl-N’-nitro-N-nitrosoguanidine. In previous work from this laboratory, chromosomedamaging activity was found in aqueous extracts of human feces [ 251. The present work reports on clastogenic activity in bile acids and the acid fractions of saponified chloroform-methanol extracts from human fecal matter. MATERIALS

AND METHODS

Chemicals All solvents used in extractions were HPLC grade. Cholic acid (3a,7o,12c~trihydroxycholanic acid), chenodeoxycholic acid (&,7a-dihydroxy-5Pcholanic acid), ursodeoxycholic acid (3o,7fl-dihydroxy-5P-cholanic acid), deoxycholic acid (3o,l%-dihydroxy-5&cholanic acid) and lithocholic acid (3a-hydroxy-5flcholanic acid) were purchased from Sigma Chemical Co. (St Louis, MO). Stock bile acids were dissolved in glass distilled dimethylsulfoxide (DMSO) and diluted into MEM (Eagle’s Minimal Essential Medium, Grand Island Biological Co.) for treatment. Since some of the bile acids are sparingly soluble in aqueous solutions, MEM was warmed to 37°C before addition of the bile acid stock. Extraction and fractionation of human feces Fecal samples were obtained from 4 individuals on normal mixed western diets and frozen immediately in dry ice. Samples were stored in a -80°C freezer and extracted under yellow light within 3-4 days after collection. The extraction procedure is outlined in Figs. 1 and 2. These procedures are similar to and have been modified from others described previously [8,11,18,27]. Each fraction was suspended in DMSO just before dryness was reached and the last traces of organic solvent were removed by rotary evaporation. In control experiments, distilled water was used in place of fecal homogenate, with all other steps in the extraction protocol unchanged. Fractions were flushed with nitrogen and stored at -80°C. Chromosome aberration test The assay and scoring protocol used in this study is routinely used as an assay for carcinogens and other genotoxic compounds [ 251. CHO cells were grown in MEM supplemented with 10% fetal calf serum, antibiotics (streptomycin sulfate, 29.6 pg/ml; penicillin, 125 pg/ml; kanamycin, 100 pg/ ml; fungizone, 2.5 pg/ml) and sodium bicarbonate (1 mg/ml). The stock cultures were maintained in 240 ml plastic culture flasks (Falcon) at 37°C in a water-saturated CO2 incubator. Approximately 140,000 CHO cells were seeded on 22 mm2 coverslips in 3.5 cm plastic dishes (Falcon) and kept in MEM with 10% fetal calf serum at 37% for 2-3 days, at which time cells

319 Fecal sample (51-221

g)

:

I

Homogenize

I

(1 1, wet

Aqueous

in double distilled water wt./vol.)

homogenate Extract

with 2 vols. chloroform/ (1 : 1, v/v). centrifuge

methanol

at 150Xg

I

Relidue

Chloroform Water/methanol

phase Extract

phase

residue‘twice

with 50 ml chloroform

Repeat extraction

with

chloroform/methanol

Chloroform

Water/methanol

phase

phase

Chloroform

Extract

phase

1 vol. -l

with

chloroform

Pooled

r

Chloroform

1 Water/methanol

phase

phase

I

chloroform DISCARD

extracts

I

Filter with

(Whatman Na,SO,

no. 4 and no. 3 paper) dry

(10 g/lOOml)

rotary

evaporate

at 4O’C Residue Saponify ethanol dilute

15 h in 10% KOH dissolved in 90% (19 ml/g residue). After

with

water

3 times with 1.7 vols. of n-hexane

Hexanepano1 extract (neutrals, bases)

Fig. 1.

phase (acids)

saponification,

to 60% ethanol.

Extract

320 Water/ethanol

phase

(acids) Rotary

evaporate

at 40%

to about 50 ml

(to remove residual haxane and ethanol). Acidify

to pH 2.0 with concentrated

Extract

with 2.5 vols. of chloroform/methanol

(2

: 1, v/v)

I

I Chloroform

HCI.

I

ph

Water/ethanol phase Extract

twice

with

1.6 vols. of chloroform

I Water/ethanol pha a

1 DISCARD Pooled chloroform extracts Wash twice with 1 vol. double distilled water to 10 vols. extract. Na,SO, at 40%

Dry with

(10 g/l 00 ml 1. Rotary Resuspended

in DMSO

evaporate (1 g acid

residue to 1 ml DMSO) Acid fraction (acid steroids, fatty acids, other weak acids)

Fig. 2.

were 60--80% confluent. The test chemical, dissolved in DMSO, was diluted into 1 ml of MEM with 2.5% fetal calf serum. Tissue culture medium was then aspirated from the petri dishes and replaced with the test solution. Bile acids are almost insoluble in water and it was thus found necessary to use a 5% final concentration of DMSO in the treatment solution. This concentration of DMSO did not induce any additional chromosome aberrations. For tests involving S9 activation, 0.5 ml of S9 mix was added to each petri dish prior to addition of 0.5 ml of test solution. Microsomal liver preparations were obtained from Fisher male rats that had been pretreated with Aroclor 1254 [ 11. The S9 activation mix consisted of liver supematant (0.3 ml/ml mix), 0.4 M MgClz (0.02 ml/ml mix), 1.65 M KC1 (0.02 mg/ml

321

mix), 6glucosophosphate (1.3 mg/ml mix), NADP (2.55 mg/ml mix) and phosphate-buffered saline, pH 7.4 (0.62 ml/ml mix). Following a 3-h exposure time, medium was removed, the coverslips washed with MEM, and fresh MEM with 10% fetal calf serum added to the petri dishes. Colchicine (0.1 ml) (0.01% in 2.5% MEM) was added at 16 h post-exposure to the chemicals and left for 4 h. Cells were then treated with 1% sodium citrate solution for 20 min, followed immediately by fixation in ethanol/acetic acid (3 : 1) for 20 min. Air-dried slides were stained with 2% orcein in 50% acetic acid/water, dehydrated and mounted. For each sample, 120 metaphase plates were analyzed for chromosome gaps, breaks and exchanges [26]. Breaks were distinguished from gaps by a separation greater than the width of the chromatid arm and displacement of the broken ends, Exchanges, which require rejoining, included chromatid exchanges, dicentrics, and mono- and multi-radials. Data are presented as the percentage of aberrant metaphases (percentage of cells exhibiting a gap, break or exchange) and the mean exchanges per metaphase plate (total number of exchanges observed divided by number of cells scored). Exchanges are presented separately since they are considered to be less ambiguous indicators of chromosomal damage than breaks or gaps. A treatment was considered toxic when approximately 50% of the cells had detached from the surface of the coverslip. The activity of the S9 mix was confirmed in every experiment with aflatoxin B1 (5 X 10m6 M). This concentration of aflatoxin induced aberrations in at least 50% of metaphase plates when combined with S9 mix but did not elevate aberration frequencies when applied with S9 mix lacking nicotinamide adenine dinucleotide phosphate (NADP). RESULTS

The acid fractions of saponified chloroform-methanol extracts of feces from 4 human donors showed varying amounts of clastogenic activity, as shown in Table 1. With acid fractions of fecal matter from subjects A and B, frequencies of aberrant metaphases and exchanges were considerably higher than those for the control. The C and D acid fractions also brought about a slight elevation in aberrations. The addition of S9 mix did not enhance the clastogenic activities of the acid fractions. Two sources of genetic activity in human feces have been reported [4,5,25,31] . One of these is an ethersoluble mutagen, positive for the Ames Salmonella strain TAlOO, which partitions with neutral and basic compounds (our hexane extract). The other is water-soluble and clastogenic in CHO cells. In this study, clastogen(s) found in the acid fraction clearly represent a third type of chemical compound with genetic activity in fecal material. With the hexane extracts, no detectable increase in the frequency of chromosome aberrations above control values was observed either with or without the addition ,of S9 mix.

322 TABLE 1 CLASTOGENIC ACTIVITY OF THE ACID FRACTION EXTRACTS OF HUMAN FECES mg chloroform extract residue/dish

Percentage of aberrant metaphases (%)* A

B

C

5.00 4.17

1.79

Toxic 22.0 (0.170) (:::42)

1.67

-_

1.56

3.33 (0.017) -

1.47 1.25

D

Blank no. 1

6.67 (0.092) 1.67 (0.00)

0.83 (0.00) (ZO)

Blank no. 2

(ZO) 0.83 (0.00)

Toxic 7.5 (0.033)

3.57 3.13

2.50 2.08

FROM CHLOROFORM

(ZO) Toxic 11.67 (0.125) 6.67 (0.092) (i.025) 0.83 (0.0083)

0.83 (0.00)

aMean exchanges/metaphase

plate are shown in parentheses.

Since bile acids are non-mutagenic in the Ames test [ 241, it was not expected that they would account for the activity in the acid extracts. Nevertheless, when 5 of the major bile acids were assayed, it was found that cholic acid induced a slight increase in aberration frequency and ursodeoxycholic acid a much larger one. An S9 fraction was included in these tests since bile acids could conceivably be metabolized by epithelial cells as well as by the colonic microflora. As shown in Table 2, the increase observed with ursodeoxycholic acid was enhanced by the addition of S9, but the presence of S9 was not a necessary condition for the demonstration of clastogenic activity. S9 also decreased the concentration of bile acid necessary to produce chromosome damage. Experiments designed to elucidate the action of the S9 mix were performed using the standard mix containing all cofactors, the standard mix without NADP, and an S9 mix with all cofactors, in which the microsomal protein had been coagulated by heating to 80°C for 3 min. Data for this experiment are shown in Fig. 3. It can be seen that ursodeoxycholic acid

323 TABLE 2 CLASTGGENIC Bile acid

ACTIVITY

OF BILE ACIDS IN CHO CELLS

Concentration (moi/Q)

Percentage of cells with aberrant metaphases (%)a

-SQ Ursodeoxycholic acid

Cholic acid

Deoxycholic acid Lithocholic acid

Chenodeoxycholic acid DMSO control

7.0 x 5.6 4.5 2.5 2.0 1.5 1.2 5.6 x 4.5 3.6 2.9 1.2 x 9.5 x 7.5 2.0 x 1.5 1.25 1.0 1.5 x 1.2 5%

10-3

+s9

Toxic 16.67 (0.158) 0.83 (0.0083) Toxic 26.67 (0.292) 4.17 (0.05) 1.67 (0.0083)

1O-3

Toxic 2.5 (0.017)

lo-” 1o-4

Toxic 0.0 (0.00)

1o-3

Toxic 1.67 (0.00)

1o-3

Toxic 0.0 (0.00) 0 (0.00)

Toxic 8.33 0.83 Toxic 3.33 0

(0.075) (0.00) (0.033) (0.00)

Toxic 0.83 (0.0083) Toxic 1.67 (0.0167) 0 (0.00)

aFigure in parentheses is number of exchanges per metaphase plate.

induces a large number of exchanges in conjunction with all 3 S9 mixes, and while there are small differences attributable to treatment, the activity of NADP-requiring mixed function oxidases is not necessary. Furthermore, even the heat-treated S9 mix increased the toxicity and the frequency of exchanges relative to a control. DISCUSSION

Bile acids are known to possess many types of biological activity, mclud( ing transformation of tissue culture cells [16], cocarcinogenicity [9,19], single-strand DNA breakage [ 171, enhancement of benzo[a] pyrene binding to DNA [ 31, and inhibition of microsomal enzymes responsible for the activation and detoxification of chemical carcinogens [ 151. Silverman and Andrews [24] found that, of 30 bile acids tested, none were mutagenic in the Ames S&noneIla/mammalian-microsome test. The present data indicate that ursodeoxycholic acid can also induce chromosome aberrations. The dose-response curves in this study are very steep, especially for the

324

0.4 al 01 RJ L a

0.3

Ill 4-l S \

0.2 s 0 c

rm 0

0.1

lz

2

3

Concentration Ursodeoxyc

4

5

C mM holic

6

78

I

Acid

Fig. 3. Clastogenic activity of ursodeoxycholic acid upon incubation with different S9 microsomal activation systems. Concentration is plotted on a log scale. Standard ); standard S9 mix omitting NADP ( m-m); boiled S9 mix with all cofactors (a---. no S9 mix (0 --o S9 mix with all cofactors (A-A); ).

acid fractions of fecal material, with chromatid aberrations appearing at, or just below, the toxic dose. The following considerations suggest that toxicity and clastogenicity are independent phenomena, at least with regard to the agents tested in this study: (1) In control experiments performed in this laboratory, toxicity as a result of low pH or osmotic shock is not accompanied by increases in chromosome damage (M.P. Rosin, pers. comm.); (2) Deoxycholic acid, lithocholic acid and chenodeoxycholic acid do not induce chromosome aberrations up to the toxic dose; (3) Ursodeoxycholic acid induces high levels of chromosome damage even at non-toxic doses. The enhancement observed with the addition of S9 mix was most likely caused by a non-enzymatic effect, since the deletion of NADP from the mix or thermal coagulation of the proteins does not greatly alter the doseresponse. Purely physical effects of S9 have been investigated [ 281, and it was found that S9 tended to diminish mutagenicity through non-specific protein binding. Horse serum and rat muscle autolysate increase the solu-

325

bility of metallic carcinogens [30], and the serum-metal complexes have been shown to penetrate the nuclei of mouse dermal fibroblasts [29]. The possibility remains that an NADPdependent detoxification enzyme could be responsible for the slight increase in clastogenic activity of ursodeoxycholic acid when NADP is omitted. However, this can only be a partial effect since it does not explain the displacement of the dose-response curve to lower concentrations with heated S9. The structure of the bile acids does not offer any immediately apparent clue to explain their difference in activity. Chenodeoxycholic acid is inactive in the CHO chromosome aberration assay and yet is very similar to ursodeoxycholic acid. The importance of minor structural differences on genetic activity has been seen before, as, for example, the 7,8diol-9,10-epoxides of benzo[a]pyrene, where the anti-isomer is more mutagenic for CHO V79 cells than the syn-isomer [ 141. It is interesting that deoxycholic acid and lithocholic acid, the 2 bile acids against which most suspicion has been directed, were both inactive in this assay. Ursodeoxycholic acid is present in human fecal material in amounts approximately equal to chenodeoxycholic acid and cholic acid [22]. It can be expected that a large proportion of the acid steroids present in the fecal material appear in the acid fraction, but the data from these experiments are not suited to a quantitative comparison of chromosome-breaking activity and bile acid concentration. The acid fraction also contains fatty acids, other weak acids and unknown compounds which may, individually or collectively, determine the genetic activity. Separation of compounds in the acid fraction is required to determine where the clastogen(s) reside. The doses of bile acids tested in this study are relatively high, but mutagenic or clastogenic potency in short-term tests is of questionable value in predicting the degree of risk, or lack of it, in vivo. The full significance of these data is therefore not clear at the present time. However, these preliminary observations may eventually offer further insight into the role of bile acids in colon carcinogenesis. Our findings on chromosome-damaging activity associated with ursodeoxycholic acid may also have implications in the treatment of cholesterol cholelithiasis. Ursodeoxycholic acid and chenodeoxycholic acid have both been used in trial studies for the dissolution of gallstones [ 71, and larger clinical studies are in progress. Feeding of either of the bile acids to human volunteers resulted in a large increase of ursodeoxycholic acid content in the bile [6]. It might be prudent to consider carefully the long-term implications of such alterations in bile acid dynamics. REFERENCES 1 Ames, B.N., McCann, J. and Yamasaki, E. (1975) Methods for detecting carcinogens and mutagens with the Salmonella/mammalian-microsome mutagenicity test. Mutat. Res., 31,347-364. 2 Aries, V., Crowther, J.S., Drasar, B.S., Hill, M.J. and Williams, R.E.O. (1969) Bacteria and the aetiology of cancer of the large bowel. Gut, 10, 334-336.

326 3 Autrup, H., Harris, C., Fugaro, S. and Selkirk, J. (1977) Effect of various chemicals on the metabolism of benzo[a] pyrene by cultured rat colon. Chem.-Biol. Interact., l&337-347. 4 Bruce, W.R. and Dion, P.W. (1980) Studies relating to a fecal mutagen. Am. J. Clin. Nutr., 33, 2511-2512. 5 Bruce, W.R., Varghese, R., Furrer, R. and Land, P.C. (1977) A mutagen in the feces of normal humans. In: Origins of Human Cancer, Book C, pp. 1641-1646. Editors: H.H. Hiatt, J.D. Watson and J.A. Winsten. Cold Spring Harbor Laboratory, New York. 6 Federowski, T., Salen, G., Colallilo, A., Tuit, G.S., Mosbach, E.H. and Hall, J.C. (1977) Metabolism of ursodeoxycholic acid in man. Gastroenterology, 73, 11311137. 7 Goldman, M.A., Schwartz, C.C., Swell, L. and Vlahcevic, Z.R. (1979) Bile acid metabolism in health and disease. In: Progress in Liver Diseases, Vol. 6, pp. 225-241 Editors: H. Popper and F. Schaffner. Grune and Stratton, New York. 8 Grundy, S.M., Ahrens, E.H. and Miettinen, T.A. (1965) Quantitative isolation and gas-liquid chromatographic analysis of total fecal bile acids. J. Lipid Res., 6, 397410. 9 Hiasa, Y., Konishi, Y., Kamamoto, Y., Watanabe, T. and Ioto, N. (1971) Effect of lithocholic acid on DL-ethionine carcinogenesis in rat liver. Gann, 62, 239-245. 10 Hill, M.J. (1975) The role of colon anaerobes in the metabolism of bile acids and steroids, and its relation to colon cancer. Cancer, 36, 2387-2400. 11 Hill, M.J. and Aries, V.C. (1971) Faecal steroid composition and its relationship to cancer of the large bowel. J. Pathol., 104, 129-139. 12 Hill, M.J., Drasar, B.S., Aries, V., Crowther, J.S., Hawksworth, G. and Williams, R.E.O. (1971) Bacteria and the aetiology of cancer of large bowel. Lancet, i, 95-100. 13 Hill, M.J., Drasar, B.S., Williams, R.E.O., Meade, T.W., Cox, A.G., Simpson, J.E.P. and Morson, B.C. (1975) Fecal bile-acids and clostridia in patients with cancer of the large bowel. Lancet, i, 535-539. 14 Jerina, D.M., Lehr, R., Schaeffer-Ridder, M., Yagi, H., Karle, J.M., Thakker, D.R., Wood, A.W., Lu, A.Y.H., Ryan, D., West, S., Levin, W. and Conney, A.H. (1977) Bay-region epoxides of dihydrodiols: a concept explaining the mutagenic and carcino genie activity of benzo(a)pyrene and benzo(a)anthracene. In: Origins of Human Cancer, Book B, pp. 639-658. Editors: H.H. Hiatt, J.D. Watson and J.A. Winsten. Cold Spring Harbor Laboratory, New York. I5 Kawalek, J.C. (1979) The effect of bile acids on drug metabolism. Nutr. Cancer, 1, 13- 18. 16 Kelsey, M.I. and Pienta, R.J. (1979) Transformation of hamster embryo cells by cholesterol-L-epoxide and lithocholic acid. Cancer Letters, 6, 143-149. 17 Kulkarni, M.S., Heidepriem, P.M. and Yielding, K.L. (1980) Production by lithocholic acid of DNA strand breaks in L1210 cells. Cancer Res., 40, 2666-2669. 18 Miettinen, T.A., Ahrens, E.H. and Grundy, S.M. (1965) Quantitative isolation and gas-liquid chromatographic analysis of total dietary and fecal neutral steroids. J. Lipid Res., 6,411-424. 19 Narisawa, T., Magadie, N.E., Weisburger, J.H. and Wynder, E.L. (1974) Promoting effect of bile acids on colon carcinogenesis after intrarectal instillation of N-methylN-nitro-l\r-nitrosoguanidine in rats. J. Natl. Cancer Inst., 53, 1093-1095. 20 Newberne, P.M. and Zeiger, E. (1978) Nutrition, carcinogenesis and mutagenesis. In: Mutagenesis. Advances in Modern Toxicology, Vol. 5,pp. pp. 53-64. Editors: W.G. Flamm and M.A. Mehlman. Hemisphere Publ. Corp., Washington, DC. 21 Reddy, B.S., Cohen, L.A., McCoy, D., Hill, P., Weisburger, J.H. and Wynder, E.L. (1980) Nutrition and its relationship to cancer. Adv. Cancer Res., 32, 237-345. 22 Reddy, B.S., Hanson, D., Mangat, S., Mathews, L., Sbaschnig, M., Sharma, C. and Simi, B. (1980) Effect of high-fat, high-beef diet and of mode of cooking of beef

327

23

24

25 26 27 28

29

30 31

in the diet on fecal bacterial enzymes and fecal bile acids and neutral steroli. J. Nutr., 110, 1880-1887. Reddy, B.S., Sharma, C. and Wynder, E.L. (1980) Fecal factors which modify the formation of fecal co-mutagens in high and low risk population for colon cancer. Cancer Letters, 10,123-132. Silverman, J. and Andrews, A.W. (1977) Bile acids: co-mutagenic activity in the Salmonella/mammalian-microsome mutagenicity test: brief communication. J. Natl. Cancer Inst., 59,1557-1559. Stich, H.F. and Kuhnlein, U. (1979) Chromosome breaking activity of human feces and its enhancement by transition metals. Int. J. Cancer, 24, 284-287. Stich, H.F. and San, R.H.C. (1970) DNA repair and chromatid anomalies in mammalian cel1.sexposed to I-nitroquinoline l-oxide. Mutat. Res., 10, 389-404. Tu, C., Powrie, W.D. and Fennema, 0. (1970) Steroids in egg yolk. J. Food Sci., 35,601-605. Wang,, Y.Y., Talcott, R.E., Seid, D.A. and Wei, E.T. (1981) Antimutagenic properties of liver homogenates, proteins and glutathione on diesel exhaust particulates. Cancer Letters, 11, 265-275. Webb, M. and Weinzierl, S.M. (1972) Uptake of 63Ni2+from its complexes with proteins and other ligands by mouse dermal fibroblasts in vitro. Br. J. Cancer, 26, 292-298. Weinzierl, S.M. and Webb, M. (1972) Interaction of carcinogenic metals with tissue and body fluids. Br. J. Cancer, 26, 279-291. Wilkins, T.D., Lederman, M., Van Tassell, R.L., Kingston, D. and Henion, J. (1980) Characterization of a mutagenic bacterial product in feces. Am. J. Clin. Nutr., 33, 2513-2520.