Dietary Fat, Fiber, and Carcinogen Alter Fecal Diacylglycerol Composition and Mass1

Home , Essential fatty acid

[CANCER RESEARCH 55, 2293-2298, June I, 1995] Dietary Fat, Fiber, and Carcinogen Alter Fecal Diacylglycerol Composition and Mass1

Jennifer S. Pickering, Joanne R. Lupton, and Robert S. Chapkin2

Faculty of Nutrition, Molecular and Ceil Biology Croup, Texas A&M University, College Station, Texas 77843-247ÃŒ

ABSTRACT because DAG in the colonie lumen is capable of entering mucosal cells of the colon (9), where it can activate the PKC pathway (12). In Fecal diacylglycerols (DAGs) are known activators of protein kinase C addition, extracellular DAGs of the chain length found in feces can (PKC), which in turn modulates colonie epithelial cell growth programs enhance the growth of benign colonie tumors and some carcinomas, and, therefore, could play a role in the malignant transformation process. However, the effects of physiological modifiers such as diet and carcinogen while normal cells are not affected (13). These studies suggest that on fecal DAG mass and composition have not been reported. We therefore fecal intraluminal DAG may have implications in the development of designed a 2 x 2 x 2 factorial study (2 fats: corn oil and fish oil; 2 fibers: colon cancer. pectin and cellulose; with and without carcinogen). Rats were provided DAG has been detected in foods at appreciable levels, although with diets for 5 weeks. Three weeks after the second injection of only negligible amounts of dietary DAG reach the colon intact (9). azoxymethane, feces were collected from 10 rats/treatment in = 80 total) Sources of DAG in the lumen have not been elucidated, although it and analyzed for DAG mass and fatty acyl composition by combined TLC may be derived partly from bacterial action (8). As a result of these and gas chromatography. Dietary fat had a significant effect on the mol % observations, interest has focused recently on the ability of dietary fatty acyl composition of fecal DAG. Greater amounts of long chain n-3 constituents to modulate intraluminal DAG levels. For example, polyunsaturated fatty acids (20:Sn-3, 22:Sn-3, and 22:6n-3) were detected in fecal DAG of fish oil-fed animals relative to corn oil (/' < 0.001). In Reddy et al. (10) have demonstrated that the type of fiber consumed contrast, corn oil resulted in a higher mol % of 18:2n-6 relative to fish oil can modify fecal DAG mass and composition. (P < 0.016). The most salient effect of fiber was on total production Dietary fat and fiber have interactive effects in the colon (14) and (nmol/day) of DAG, which was 2.5 times higher with cellulose than pectin are capable of influencing colonie PKC isoform expression (15). supplementation. In addition, there was an effect of fiber on both mol % However, the combined effects of dietary fat and fiber on fecal DAG and concentration of 22:6n-3, with cellulose producing higher amounts mass and composition have not been examined. Therefore, the present relative to pectin (P < 0.04). A significant interaction between fat and fiber study was designed to investigate the effects of dietary fat and fiber on was observed with nmols of 17:0 excreted in 24 h, with fish oil/cellulose the mass and fatty acyl composition of fecal DAG in the rat experi producing 94.2 nmol as compared to 3.5 seen with corn oil/pectin mental colon carcinogen model. In addition, fecal diglyceride sub (P < 0.02). There was a significant interaction between fat and carcinogen classes were characterized to determine aliphatic linkage at the sn-\ on all of the DAG n-3 fatty acids, which were elevated with carcinogen/fish chemical position. oil treatment. These data show that fat, fiber, and carcinogen can modu late the fatty acyl composition and mass of fecal DAG. Since the produc tion of fecal DAG, an activator of PKC, may alter colonie mucosa! cell MATERIALS AND METHODS proliferation, our data offer insight into a mechanism by which diet may modify the risk of colon cancer development. Materials. l,2-diheneicosanoyl-.vn-glycero-phosphocholine (21:0-21:0-PC) was purchased from Avanti Polar Lipids (Alabaster, AL). Phospholipase C (Type V, Bacillus Cereus) and azoxymethane were from Sigma Chemical Co. (St. Louis, INTRODUCTION MO). Silica gel 60 plates (20 x 20 cm, 0.25-mm thickness) for TLC were from PKC'1 consists of a family of serine/threonine kinases that upon Merck (Darmstadt, Germany). Bakerbond silica gel columns were purchased from J. T. Baker, Inc. (Phillipsburg, NJ). Dextrose was from HarÃanTeklad (Madison, activation can phosphorylate proteins that regulate short- and long- WI). Cellulose, casein, DL-methionine, choline bitartrate, AIN-76 vitamin, and term cellular events, including cell proliferation and differentiation. It AIN-76 mineral mix were purchased from Bio-Serve (Frenchtown, NJ). Corn oil is now clear that PKC expression and activation are involved in the was kindly provided by Traco Labs (Seymour, IL). Vacuum-deodorized Menha regulation of colonie mucosa! proliferation (1-4). This is noteworthy den fish oil was provided by the NIH Fish Oil Test Material Program, Southeast because the elevation of colonie cell proliferation has been associated Fisheries Center (Charleston, SC). High methoxylated pectin was from Grindsted with the incidence of colon cancer (5). (Industrial Airport, KS). Animals and Diets. The animal use protocol for these experiments was DAGs link extracellular signals with intracellular responses approved by the University Animal Care Committee at Texas A&M Univer through the activation of PKC (6). Intracellular DAG produced from sity. All animals were treated in accordance with the NIH Guide for the Care inositol and choline phospholipid hydrolysis, coupled with increased and Use of Laboratory Animals (NRC 1985). Eighty male weanling Spraguc- Ca2+ levels, can activate PKC to influence cellular processes (7). It is Dawley rats (HarÃan,Houston, TX) were housed individually in cages in a interesting that several studies have recently shown that rat and human temperature- and humidity-controlled facility with a 12-h light/dark cycle. feces contain appreciable levels of DAG (8-11). This is significant After a 3-day acclimation period of consuming a nonpurified diet, rats were assigned randomly to one of 8 treatments (10 rats/treatment) in a 2 x 2 x 2 factorial design with 2 fats, 2 fibers, with or without carcinogen. The diet Received 1/4/95; accepted 4/4/95. The costs of publication of this article were defrayed in part by the payment of page compositions are shown in Table 1. The two fats were fish oil (11.5 g/100 g charges. This article must therefore be hereby marked advertisement in accordance with diet) and corn oil (15 g/100 g diet). The fish oil diet contained 3.5 g corn 18 U.S.C. Section 1734 solely to indicate this fact. 1This work was supported in part by grants from the American Institute for Cancer oil/100 g diet to ensure that essential fatty acid requirements were met (14). Corn oil and fish oil contained identical levels of tert-butyl-hydroquinone Research, the Texas A&M Interdisciplinary Research Initiatives Program, and NIH Grants CA59034 and CA61750. (0.025%) and vitamin E (1.5 mg/g a-tocopherol and 1.0 mg/g y-tocopherol). 2 To whom requests for reprints should be addressed, at 442 Kleberg Center, Texas The two fibers were pectin and cellulose (6 g/100 g diet). Each group was A&M University, College Station, TX 77843-2471. divided further into + or - carcinogen treatment. Food and water were 3 The abbreviations used are: PKC, protein kinase C; DAG, 1,2-diacyl-JTi-glycerol; AAG, l-O-alkyl-2-acyl-sn-glycerol; A'AG, l-O-alkenyl-2-acyl-jn-glycerol; ANS, available freely. anilino-1-naphthalene sulphonic acid ammonium salt; AOM, azoxymethane; PUFA, Carcinogen Administration. After 1 week of consuming semipurified polyunsaturated fatty acid. diets, the rats were given injections 2 times at 1-week intervals with 2293

dietIngredientsDextroseCaseinDL-MetCornTable 1 Modified AIN-76 added to convert any unreacted acetic anhydride to methyl acetate. Samples byweight51.0622.350.3415.03.5/11.53.911.120.226.0Percentbykcal47.6520.8631.4931.49were dried under nitrogen, and 2 ml of water and 3 ml of diethyl ether/ methanol/acetic acid (60:40:1, v/v/v) were added. After vortexing, petroleum ether was added, and the upper phase was extracted. The samples were separated into diglyceride-acetate subclasses with the use of a triple develop oil"Corn/fish oil''Mineral ment TLC system consisting of toluene development, followed by hexane/ ether (4:1, v/v) and toluene. Diglyceride subclasses were visualized with ANS. mixVitamin mixCholine Finally, the diglyceride kinase assay as described by Priess el al. (20) was bitartralePectin/cellulose" utilized as a third experimental procedure for quantitating diglyceride subclass mass. This assay uses Escherichia coli 1,2-diacyl-j/i-glycerol kinase, which Corn oildiet.h phosphorylates both l-alkyl-2-acyl-.Ã§/i-glycerol and 1,2-diacyl-sn-glycerol an Fish oil diet.Percent alogues. However, the kinase does not phosphorylate either l,3-diacyl-s;i- glycerol or l -O-alkyl-3-acyl-.sn-glycerol (21). In some experiments, the radiolabeled products of the diglyceride kinase assay were subjected to a mild azoxymethane at 15 mg/kg body weight or saline. Food intake and fecal output alkaline hydrolysis in order to separately quantitate the relative contribution of over 48 h was evaluated 1 week after the second injection. Animals were diacylglycerols (alkaline labile) and alkylacylglycerols (alkaline stable) to the weighed on arrival and weekly after their second injection. Three weeks after total radiolabled phosphatidic acid generated (21). the second injection, feces were collected and processed as described below. Statistics. The data were analyzed with the use of Statistical Analysis Fecal Sample Collection and Lipid Extraction. In order to quantitate software by three-way ANOVA. If the P values for the interactions were fecal diacylglycerols, 3 weeks after the second injection fecal samples from <0.05, means of the eight treatment groups were then separated using Dun can's Multiple Range test. If the P values were <0.05 for the effect of fat, fiber, individual animals were collected and weighed immediately after defecation. Total fecal lipids were extracted by the method of Florin-Christensen (16). A or carcinogen but not for the interactions, total means of the fat groups, fiber groups, or carcinogen groups were separated with the use of Duncan's Multiple known mass of 21:0-21:0 diacyl-j/i-glycerol internal standard was added to quantitate fecal diacylglycerol. The internal DAG standard was synthesized Range test. from 21:0â€”21:0-PC by phospholipase C treatment as described by Akoh and Chapkin (17). Briefly, 400 ng of 21:0-21:0-PC were dissolved in 2 ml RESULTS peroxide-free diethyl ether containing 0.005% butylated hydroxytoluene and 2 ml of 50 mM Tris-HCI (pH 7.4) containing 5 HIMcalcium chloride and 10 units Dietary Fatty Acids. The dietary fatly acids are shown in Table 2. of phospholipase C (Bacillus cereus). The 2-phase mixture was agitated in a Diets supplemented with corn oil contained linoleic acid (18:2n-6) as shaking water bath for 3 h at room temperature. The resultant diacylglycerols the major fatty acid constituent. With respect to polyunsaturated fatty were extracted twice with 2 ml diethyl ether. Following the addition of internal acid content, fish oil-supplemented diets contained primarily eicosa- standard, samples were homogenized in 1.0 ml of ice-cold methanol/water pentaenoic acid (20:5/1-3) and docosahexaenoic acid (22:6/1-3), and (1.0:0.9, v/v). After vortexing, 1.1 ml of chloroform was added to each sample had reduced levels of 18:2/i-6 relative to corn oil diets. and an additional 1.0 ml ice-cold methanol/water (1.0:0.9, v/v) was added. The fecal homogenates were centrifuged at 800 X g for 5 min at 4Â°C.The lower Diglyceride Subclass Characterization. Diglyceride subclass composition was determined following conversion to diglyceroac- organic phase was removed and dried under nitrogen. The lipid extracts were etates, diglycerobenzoates, or [32P]phosphatidic acid via diglyceride resuspended in chloroform and passed through chloroform-washed silica gel extraction columns (Bakerbond, Phillipsburg, NJ) with the use of a vacuum kinase. Subclasses from selected lipid extracts were separated by manifold. Fecal diacylglycerols were eluted from the column using chloro- TLC, and autoradiographs were identified by comparison to authentic form/methanol (90:10, v/v) and normalized for 48-h fecal output. standards (Fig. la). Only minor levels of lyso-phosphatidic acid Diacylglycerol Isolation. Fecal diacylglycerols were separated by TLC on (<10% of total phosphatidic acid dpm) were observed following silica gel 60 plates with the use of benzene-chloroform-methanol (80:15:3.25, alkaline hydrolysis, indicating a relative absence of l-O-alkyl-2-acyl- v/v/v) (14). The DAG band was detected under UV light after spraying with 1 s/Â¡-glyceroland l-O-alkenyl-2-.sn-glycerol in fecal lipid extracts (Fig. g/liter ANS in water (pH 7.0) and subsequently transmethylated with 3 ml of methanol/sulphuric acid (94:6, v/v). The resultant fatty acid methyl esters were extracted with the use of 3 ml of hexane and 1 ml of 0.1 mol/liter potassium Table 2 Weight percent of fatty acids in the experimental diets chloride. The fatty acid methyl ester extracts were dried under nitrogen and Fatty acids are expressed as number of carbon atoms:number and position of double bonds. Fatty acid methyl ester peak areas were quantitated by gas chromatography as resuspended in 25 fil of chloroform/methanol (2:1, v/v). The fatty acid methyl described in "Materials and Methods." esters were separated further on TLC with the use of toluene and were detected under UV light after they were sprayed with ANS. Fatty acid composition and DAG concentration were determined by gas Chromatographie analysis Fattyacid14:014:116:016:ln-717:018:018:ln-918:ln-718:2/1-618:3n-320:020:ln-920:2n-620:3/1-620:4/1-620:3/1-320:5/1-322:022:5n-322:6n-324:024:ln-9Cornoil/cellulose0.212.70.22.727.00.255.30.90.30.2Cornoil/pectin0.112.10.22.124.80.358.31.10.30.2Fishoil/cellulose7.40215.88.71.22.91(1.52.515.31.10.50.60.10.20.60.111.10.21.77.1Fishoil/pectin6.60.114.88.41.22.810.22.415.01.20.50.70.10.20.70.112.00.21.97.9 (Hewlett Packard 5890) of the DAG-derived fatty acid methyl esters as described previously (18). Characterization of Diglyceride Subclasses. Diglycerides may contain several subclasses, which are defined according to the aliphatic linkage at the .vfl-1position. These include DAG, AAG, and A'AG species. We utilized three techniques to determine the sn-l chemical linkage in fecal diglycerides. Fecal diglycerides were converted to their corresponding diglycerobenzoates and separated subsequently by TLC (17). Briefly, diglycerides were incubated with 0.2 ml of benzene containing 10 mg benzoic anhydride and 0.1 ml of benzene containing 4 mg of 4-dimethylaminopyridine at room temperature for 1 h. After incubation, 2 ml of hexane was added; excess benzoic anhydride was hydrolyzed by adding 1 ml of concentrated ammonium hydroxide while slowly agitating. Subclasses were separated by TLC with the use of benzene/hexane/ diethyl ether (50:45:4, v/v) and identified with the use of appropriate internal standards. In addition, diglyceride samples were also converted to diglyceride acetates by acetolysis to corroborate subclass identification (19). Briefly, diglyceride samples were incubated with 1 ml of acetic anhydride and 4â€”5 drops of pyridine for 75 min at 80Â°C.After cooling, 1 ml of methanol was

2294

Downloaded from cancerres.aacrjournals.org on September 27, 2021. © 1995 American Association for Cancer Research. FECAL DAG MODULATION BY DIET AND CARCINOGEN a (Tables 3c and 5b) and fatty acids from DAG excreted over 24 h (Table 5, r and d) in order to determine the effect of diet and carcinogen on DAG composition and mass. Dietary fat composition PA had a major effect on the fecal DAG mol % levels of 14:0, 18:2n-6, 20:5n-3, 22:5n-3, 22:6n-3, and 24:ln-9 (Table 3Â«).Ingeneral, animals fed fish oil had significantly elevated (P < 0.05) levels of 20:5n-3-, 22:5/;-3-, and 22:6;i-3-containing diacylglycerols. Fiber Effect. Dietary fiber significantly altered the mol % of DAG-containing 18:lÂ«-7and 22:6n-3 in pectin as compared to cel lulose animals (Table 3b). Specifically, 18:ln-7 was higher in pectin- h- C-1-P fed animals and 22:6n-3 was higher in cellulose-fed animals. Fiber composition altered the nmol of 20:5n-3 and 22:6/i-3-DAG when expressed on a per g wet weight feces (Table 4). A significant elevation in 20:5;i-3 was observed in pectin-fed animals, while 22: L-PA 6Â«-3was higher in the cellulose-fed animals. When expressed as nmol of fatty acid from DAG excreted in 24 h, only 16:0 was significantly influenced by fiber, with means higher in the cellulose diets (data not shown). In addition, the mass of fecal diacylglycerols excreted in -9 24 h was affected significantly (P < 0.05) by fiber (Table 6). Specif â€¢ m Origin ically, fecal DAG was significantly higher in animals fed cellulose- 1 2 2 containing diets (Fig. 2). Carcinogen Effect. AOM injection significantly affected (P < 0.05) the mol % levels of 18:ln-9-diacylglycerol (data not shown). The mean mol % level of 18:ln-9 in the AOM-treated PA animals was 20.5 Â±1.5 compared to saline treatment at 15.9 Â±1.5 (n = 5-10). Fat-Fiber Effect. The mol %of DAG-derived 17:0, 18:0, and 20:0 affected (P < 0.05) by a fat-fiber interaction are shown in Table 5Â«.A fat-fiber effect on 17:0 was evident when expressed as mol %, nmol/g wet weight feces, and 24-h excretion (Table 5, b and c). The highest 17:0-containing diacylglycerols were associated with the fish oil/ â€¢II )â€” C-1-P cellulose-fed animals. The nmol of 24:0 per g wet weight feces also was affected significantly by dietary fat and fiber composition (Table 5b). Diacylglycerol-derived fatty acids significantly affected by fat- fiber interaction over a 24-h excretion period were 17:0, 18:3n-3, â€¢<â€”L-PA 20:0, 22:5i!-3, 22:6Â«-3, and 24:0 (Table 5c). The n-3 fatty acids 22:5>!-3 and 22:6;i-3 were significantly higher in animals that were fed the fish oil/cellulose diet compared to fish oil/pectin, corn oil/cellulose, and corn oil/pectin. <â€” Origin Fat-Carcinogen Effect. A fat-carcinogen interaction significantly affected (P < 0.05) the nmol of 22:5n-3 per g wet weight fÃ¨ces(data

Fig. La, detection of fecal diglycerides by conversion lo |'-P]phosphatidic acid. Fecal Table 3a Effect offal on mol 7c of fully acid derived from fecal uiacylglycenil" lipids were extracted and converted to radiolabeled phosphatidic acid with the use of Â£. Coli diglyceride kinase as described in "Materials and Methods." 12Plipids were separated Fatty acid Corn oil Fish oil P value by TLC and exposed to X-ray film. Results are autoradiographs from representative 14:0 2.6 Â±0.6 5.0 Â±0.6 (1.0063 experiments repeated 3 times. Sample 1. authentic DAG standard: l-stearoyl-2-arachido- 18:2n-6 15.7 Â±2.1 8.2 Â±2.1 0.0160 noyl-.v/i-glycerol; Sample 2. fecal lipid extracts from 3 individual rats. PA. phosphatidic 0.1 Â±0.4 3.4 Â±0.4 O.IHKI1 tr" acid; C-I-P. ceramide-1-phosphate: L-PA, lyso-phosphatidic acid. b. characterization of 22:5/i-3 1.4 Â±0.2 II.IXKI! fecal diglyceride subclasses. The aliphatic chain linkage at the m-1 position was deter 22:6ii-3 0.1 Â±0.4 3.6 Â±0.4 O.INKII mined by conversion to [32P]phosphatidic acid with the use of E. Coli diglyceride kinase tr" 24:1)1-9 0.9 Â±0.3 0.0396 as described in ti. Selected samples containing radiolabeled products were subjected to alkaline hydrolysis in order to quantitate the relative levels of l-acyl-(alkaline labile). Table 3b Effect of fiher on mol % offaltv adii derived from l-alkyl-(alkaline stable), and l-alkenyl-(alkaline stable) species. Lune I, authentic un fecal diacylglycerol treated DAG standard: l-stearoyl-2-arachidonoyl-.wi-glycerol; Lune 2, alkaline hydrolysis of DAG standard; Lane 3, alkaline hydrolysis of AAG standard: l-hexadecyl-2-linoleoyl- Fatty acid Cellulose Pectin P value j;j-glycerol: Lane 4, alkaline hydrolysis of A'AG standard: 1-plasmenyl-2-acyl-.vn-glycerol 18:lÂ«-7 Â±1.2 Â±1.4 (derived from bovine heart PC): Ulne 5. fecal extract; Lune fi. fecal extract. 22:6n-32.8 2.5 Â±0.46.4 1.2 Â±0.50.0543 0.0397

Table 3c Effect offal on nmols fatty acid derivai from tiiacylglycerol per g feces lb). Greater than 90% of the subclasses were of the diacyl configu Fattyacid20:5n-322:6n-3 oil0.4 oil17.5 valueO.OW1 ration. These analyses were confirmed by TLC of fecal diglyceroben- Â±2.1 Â±2.0 zoates and diglyceroacetates (data not shown). 0.3 Â±1.73 15.5Â±1.7 (UXK)I Fat Effect. Fecal diacylglycerol composition (Tables 3-5) and 24:ln-9Corn trFish 2.8 Â±0.9P 0.0373 " Data are shown as mean Â±SE with analyses from n - 5-10. Only significant effects mass (Table 6) were affected significantly by dietary fat (P < 0.05). (P < 0.05) are shown. The data are expressed as mol % composition (Table 3, a and b, and '' tr. trace amounts [<0.1% of total fatty acids present (Table 3a) or <0.1 nmol/g feces Table 5Â«),nmols of fatty acid from DAG per g wet weight feces (Table 3c)]. 2295

Table 4 Effect offiber on nmols offairy acid derived from diacylglycerol per g feces" with saline (12.2 Â±4.3) as compared with AOM-treated animals Fatty acid Cellulose Pectin P value (3.3 Â±4.3). 20:5n-3 Â±2.2 22:6n-35.7+1.910.5 Â±1.612.2 5.3 Â±1.80.0315 0.0374 " Refer to Table 3 for legend details. DISCUSSION

not shown). The corn oil-fed animals treated with AOM had trace The results of the present study demonstrate that fecal DAG amounts of 22:5n-3, while the corn oil treatments with saline had composition and mass can be influenced by altering dietary fat and significantly higher mean values. The fish oil-fed animals treated with fiber composition. As indicated in Fig. 1, a and b, the diglycerides AOM had significantly higher levels of 22:5n-3 than did fish/saline, located in the colonie lumen were predominantly (>90%) of the corn/AOM, or corn/saline animals. Significant fat-carcinogen effects 1,2-diacyl-irc-glycerol configuration (DAG). Fecal AAG and were also observed with respect to the 24-h excretion of DAG- A'AG species were present only in minor amounts. The charac containing 20:5n-3, 22:5n-3, 22:6n-3, and 24:0 (Table 5d). In general, terization of diglyceride sn-1 aliphatic linkage is important because DAG-derived n-3 fatty acids were significantly higher in fish oil-fed ether-linked diglycerides (AAG and A'AG) are effective modula animals treated with AOM. tors of PKC, with separate and distinct calcium requirements in Fiber-Carcinogen Effect. The mol % of 18:0 was significantly comparison to DAG (22, 23). affected (P < 0.05) by fiber-carcinogen (data not shown). The saline- Dietary incorporation of fatty acids into fecal DAG was especially treated animals had higher means in both cellulose and pectin diets, evident with n-3 fatty acids. Specifically, in fish oil-fed animals, DAG while AOM animals were lower in their mol % of 18:0-DAG. The containing n-3 PUFA expressed on a mol %, nmol/g wet weight feces, nmol of 24:0 per g wet weight feces were affected significantly by a and nmol of fatty acid excreted over 24 h were increased significantly fiber-carcinogen interaction (data not shown). Pectin-fed animals had (P < 0.01) compared to corn oil animals. Previous studies have shown

diacylgfycerofFatty Table 5a Effect offat-fiber interaction on mol % offatty acid derived from fecal acid17:0 oil/cellulose0.4 oil/pectin1.1Â±1.9A oil/cellulose16.0Â±1.5B oil/pectin4.6+1.9* value0.0012 Â±1.7A 18:0 17.8Â±1.8A 12.7 + 2.0AB 9.7 Â±1.6B 13.4 + 2.0AB 0.0230 20:0Corn 2.11.4ATable Â±1.3ABCorn 0.5 + 0.0204nmol 0.7Â±1.2AFish 5.5 Â±1.4"P onFatty 5b Effect offat-fiber interaction fatty acid derivedfromfecesFish diacylglycerol per g acid17:0 oil/cellulose3.1 oil/pectin5.0 oil/cellulose70.3 oil/pectin18.2 value0.0355 Â±11.8A Â±13.2A Â±11.3B Â±13.2A 24:0Corn 9.30.0265Table Â±3.8ACorn 1.9Â±4.3AFish 2.7 Â±3.5AFish 13.5Â±4.3BP hFatty 5c Effect offat-fiber interaction on nmol offatty acid from fecal diacylglycerol excreted in 24

acid17:0 oil/cellulose6.0 oil/pectin3.5 oil/cellulose94.2 oil/pectin14.3 value0.0224 Â±15.6A Â±17.5A 14.5C Â±17.5" 18:3n-3 0.1 Â±1.4* 3.1 Â±1.4B 2.7 1.1* 0.3 + 1.4* 0.0427 20:0 15.8 Â±4.0* 1.9Â±4.5B 7.2 3.7B 15.3 + 4.5* 0.0122 22:5n-3 trA.<,0.8 ti^ 8.3 1.0* 3.1 Â±1.2A 0.0137 22:6n-3 Â±3.2A tr* 28.6 2.9B 8.0 + 3.6C 0.0032 24:0Fatty 11.4Â±3.7ATable 1.7Â±4.2Binteraction0.0196nmol 3.5 3.5BFish 11.6Â±4.2AP onCorn5d Effect offat-carcinogenCorn of fatty acidhFish from fecal diacylglycerol excreted in 24 acid20:5Â«-3 oil/saline0.6oil/AOM Corn oil/AOM23.4 oil/saline10.8 value0.0503 Â±3.5A + 3.4B Â±3.3C 22:5n-3 tr* t^ 9.0Â±1.2B 2.3 + 1.0C 0.0069 22:6n-3 tr* 0.8 Â±3.3A 25.3 Â±3.3B 11.4Â±3.2C 0.0518 24:0" 3.^4^s 3.9Bilues9.9 Â±3.9BFish 12.1 Â± 3.7Aare 3.0 Â± 0.0246 Data are shown aCornmean + SE with analyses from n = 5-10. Vt^ sharing any common capital-letter superscriptFishnot significantly different (P > 0.05).P

significantly higher levels of 24:0 per g wet weight feces when treated that dietary fish oil aids in the prevention of colon carcinogenesis

Table 6 Effect of diet and carcinogen on fecal diacylglycerol mass" Corn oil/cellulose Corn oil/pectin'' Fish oil/cellulose*" Fish oil/pectin'" Corn oil/cellulose' Corn oil/pectin' Fish oil/cellulose' Fish oil/pectin'

Total nmol fatty 653.7 Â±367.4 462.7 + 99.6 493.4 + 99.5 844.4 + 373.0 977.1 Â±487.1 425.6 Â±87.2 573.5 Â±169.4 488.2 + 68.8 acid'Vg wet feces Total nmol 327.7 Â±184.5 231.5 Â±49.8 246.8 Â±49.7 422.6 Â±186.5 487.8 Â±243.4 212.7 Â±43.6 286.9 Â±84.9 189.8Â±27.9 DAG/g wet feces Total nmol 1257.0 Â±425.3A 333.0 + 82.4BC 625.5 Â±171.6*BC 587.3 Â±240.8ABC 965.4 Â±508.2*" 276.0 Â±29.0BC 563.3 Â± 142.4Â±20.6e DAG in 24 h " Data are shown as mean Â±SEwith analyses from Â«=5-10. Fiber effects cellulose > pectin at P = 0.0192. Capital-letter superscripts are as in Table 5. * Animals treated with AOM. ' Animals treated with saline. ' Represents fatty acid derived from DAG.

2296

700 noteworthy because PKC is involved in the regulation of colonie epithelial cell proliferation (1-4). The possibility that dietary fiber

600- type could modulate the uptake of DAG by colonocytes is intriguing and requires futher investigation. â€¢o Diet composition and carcinogen exposure had no effect on the 0) luminal concentration of DAG (nmol DAG/g feces). This was unex ux pected because high luminal levels of DAG may enhance colonie cell UJ 400- proliferation (11), and cellulose generally is considered to reduce colon cancer risk. With respect to the effect of carcinogen, there was 300- an interactive effect of fat and carcinogen on the levels of 20:5n-3,

ra 22:5;j-3, and 22:6/Â¡-3-DAG. The significance of this observation re U 200- mains to be determined. In addition, the mol % composition of monoenoic 18:l;i-9-DAG species was increased in AOM-trcated rats, while 18:0-DAG was decreased. Monoenoic 18:l;i-9-DAG is derived X 100 o from the A9 desaturation of stearic acid (18:0). Previous investigators E e have demonstrated that A9 desaturase is elevated in malignant trans formed cells, thereby increasing the conversion of 18:0 to 18:1/1-9 (33, Cellulose Pectin 34). It is possible, therefore, that colonie microbial A9 desaturase Fig. 2. Etica of dietary fiber on fecal DAG mass. Columns, mean nmol DAG excreted activity is activated in AOM-injected rats relative to saline controls. in 24 h; bars, SE. For cellulose. n = 30, and for pectin, n = 22. Fiber effects: The specific effects of carcinogen on luminal bacterial enzymes cellulose > pectin at P < 0.05. requires further study. In conclusion, we demonstrate for the first time that dietary fat, (24-26), with AOM tumor incidence decreased in rats treated with fiber, and carcinogen treatment alter fecal DAG mass and composi either 20:5/i-3 (27) or 22:6n-3 (28). Consistent with the protective tion. Since the production of fecal DAG, an activator of PKC, may effect of fish oil-derived /Â¡-3PUFA, we have shown that fish oil-fed alter colonie mucosal cell proliferation, our data offer insight into a rats consistently have the lowest levels of colonie cell proliferation mechanism by which diet may modify the risk of colon cancer (14, 15). These findings suggest a potential role for lipid nutrition in development. It is essential, therefore, to determine the levels of cancer prevention. Whether fecal DAG species containing n-3 PUFA specific fecal DAG molecular species so that the mechanisms regu have unique biological properties has not been determined. This is an lating colonie PKC activation and the rates of cell proliferation, intriguing question since fecal DAG is capable of entering colono- differentiation, and apoptosis can be elucidated. cytes and activating PKC (9, 12), which in turn may mediate colonie cell proliferation (4, 15). It is interesting that the fatty acyl composi ACKNOWLEDGMENTS tion of DAG is a key determinant of its fate within intact cells and may modulate the metabolic termination of DAG-mediated intracel- We gratefully acknowledge Dr. Harold Aukerna. Chris Jolly. Meng Chao lular signaling (16, 29). Therefore, diet-induced alteration of DAG Lee. and Stella Wiese for excellent laboratory assistance. We also acknowl composition could result in the differential activation of cellular edge the generous donation of dietary corn oil by Sid Tracy (Traco Labs). phospholipases (30) and PKC (31, 32) and may selectively modulate the growth programs of colonie epithelial cells (13). REFERENCES Analysis of fecal DAG composition also revealed the presence 1. Guillem, J. G.. O'Brian, C. A.. Fitzer, C. J., FÃ¶rde,K. A.. LoGerfo, P., Treat, M., and of an uncommon, odd chain fatty acid, i.e., heptadecanoic acid Weinslein. I. B. Altered levels of protein kinase C and Ca2 ' -dependent protein (17:0). The level of 17:0-DAG was enriched highly in feces kinases in human colon carcinomas. Cancer Res., 47: 2036-2039, 1987. isolated from fish oil-fed animals (Table 5, a and b), with the 2. Craven, P. A., and DeRubertis, F. R. Suhcellular distribution of protein kinase C' in rat colonie epithelial cells with different prolifcrative activities. Cancer Res., 47: highest levels associated with fish oil/cellulose animals. Since 3434-3438, 19x7. fecal DAG is generated in part by colonie microflora (8) and 3. Chapkin. R. S.. Gao. J.. Lee, D-Y. K.. and Lupton, J. R. Dietary fibers and fats alter negligible amounts of 17:0 are found in the diet (Table 2), the rat colon protein kinase C activity: correlation to cell proliferation. J. Nutr.. 123: 649-655. 1993. enrichment of a specific population of microflora by feeding a fish 4. Davidson. L. A.. Jiang. Y. H.. Derr. J. N.. Aukema. H. M.. Lupton. J. R., and Chapkin. oil/cellulose combination may have generated 17:()-DAG. Al R. S. Protein kinase C isofornis in human and rat colonie mucosa. Arch. Biochem. though the biological properties of 17:0-DAG remain to be deter Biophys., 312: 547-553, 1994. 5. Lipkin. M. Biomarkers of increased susceptibility to gastrointestinal cancer: new mined, it is possible that this novel odd chain second messenger application to studies of cancer prevention in human subjects. Cancer Res., 4K: could play a protective role in colon carcinogenesis since DAGs 235-245, 1988. 6. Liscovitch, M., and Cantley. L. C. Lipid second messengers. Cell, 77: 329-334,1994. containing saturated fatty acyl species generally have a suppressed 7. Nakamura. S.. and Nishizuka. Y. Lipid mediators of protein kinase C activation for ability to activate PKC (7). This is based on the fact that the Kuct the intracellular signaling network. J. Biochem., 1154: 1029-1034, 1994. and Vm.Mvalues of unsaturated and saturated DAG molecular 8. Morotomi. M.. Guillem. J. G., LoGerfo, P.. and Weinstein, 1. B. Production of diacylglyccrol. an activator of protein kinase C. by human intestinal microflora. species are distinct (31, 32). Cancer Res., 50: 3595-3599, 1990. The major effect of fiber was on the total excretion (nmols/day) of 9. Morotomi. M.. LoGerfo. P., and Weinstein, I. B. Fecal excretion, uptake and metab fecal DAG, which was approximately 2.5 times higher with cellulose olism by colon mucosa of diacylglycerol in rats. Biochem. Biophys. Res. Commun.. IXI: 1028-1034. 1991. relative to pectin supplementation. An enhancement of the fecal 10. Reddy. B. S.. Simi. B.. and Englc. A. Biochemical epidemiology of colon cancer: excretion of DAG following cellulose feeding could be explained by effect of types of dietary fiber on colonie duicylglyccrols in women, (iastroenterol- ogy. 106: 883-889, 1994. either an increase in luminal DAG synthesis or a decrease in the 11. Steinbach. G.. Morotomi, M.. Nomoto. Kâ€žLupton. J. R., Weinstein. I. B.. and Holt, uptake of DAG by colonocytes. It is possible, therefore, that dietary P. R. Calcium reduces the increased fecal 1.2-.m-diacylglyccrol content in intestinal cellulose may reduce colon cancer risk in part by decreasing the bypass patients: a possible mechanism for altering colonie hyperproliferation. Cancer Res., 54: 1216-1219. 1994. mucosal uptake of luminal DAG. Such an effect potentially could 12. DeRubertis. F. R.. and Crowen. P. A. Relationship of bile salt stimulation of colonie influence colonocyte PKC-dependent signal transduction. This is epithelial phospholipid turnover and proliferativc activity: role of activation of protein 2297

kinase C. Prev. Med., 16: 572-579, 1987. 24. Deschner, E. E., Lytle, J. S., Wong, G., Reperto, J. F., and Newmark, H. L. The effect 13. Friedman, E., Isaksson, P., Rafter, J., Marian, B., Winawer, S., and Newmark, H. of dietary to-3 fatty acids (fish oil) on azoxymethanol-induced focal areas of dysplasia Fecal diglycerides as selective endogenous mitogens for premalignant and malignant and colon tumor incidence. Cancer (Phila.), 66: 2350-2356, 1990. human colonie epithelial cells. Cancer Res., 49: 544-548, 1989. 25. Reddy, B. S., and Maruyama, H. Effect of dietary fish oil on azoxymethane-induced 14. Lee, D. Y., Chapkin, R. S., and Lupton, J. R. Dietary fat and fiber modulate colonie colon carcinogenesis in male F344 rats. Cancer Res., 46: 3367-3370, 1986. cell proliferation in an interactive site-specific manner. Nutr. Cancer, 20: 107-118, 26. Reddy, B. S., and Sugie, S. Effect of levels of Â¡a-3and Â¡a-6fatty acids on azoxymeth 1993. ane-induced colon carcinogenesis in F344 rats. Cancer Res., 48: 6642-6647, 1988. 15. Davidson. L. A., Lupton, J. R., Jiang, J. H., Chang, W. C., Aukema. H. M., and 27. Minoura, T., Takata. T. Michitomo, S., Takada, H.. Yamamura, M.. Hioki. K., and Chapkin, R. S. Dietary fat and fiber alter rat colon protein kinase C isozyme Yamamoto, M. Effect of dietary eicosapentaenoic acid on azoxymethane-induced expression. J. Nutr., 125: 49-56, 1995. colon carcinogenesis in rats. Cancer Res., 48: 4790-4794, 1988. 16. Florin-Christensen, J. F., Florin-Christensen, M. F., Delfino, J. M., Stegmann, T., and 28. Takahashi, M.. Minamoto, T., Yamashita, N., Toshio, K., Yazawa, K., and Hiroyase, Rasmussen, H. Metabolic fate of plasma membrane diacylglycerols in NIH 3T3 E. Effect of docosahexaenoic acid on azoxymethane-induced colon carcinogenesis in fibroblasts. J. Biol. Chem., 267.- 14783-14789, 1992. rats. Cancer Lett., 83: 177-184, 1994. 17. Akoh, C. C., and Chapkin, R. S. Composition of mouse peritoneal macrophage 29. Florin-Christensen, J. F., Florin-Christensen. M. F., Delfino, J. M., and Rasmussen, phospholipid molecular species. Lipids, 25: 613-617, 1990. H. New patterns of diacylglycerol metabolism in intact cells. Biochem. J. 289: 18. Chapkin, R. S., Haberstroh, Bâ€žLiu, T., and Holub, B. J. Characterization of the 783-788, 1993. individual phospholipids and their fatty acids in serum and high-density lipoprotein of 30. Dawson, R. M. C., Irvine, R. F., Bray, J., and Quinn, P. J. Long-chain unsaturated the renal patient on long-term maintenance hemodialysis. J. Lab. Clin. Med., 101: diacylglycerols cause a perturbation in the structure of phospholipid bilayers render 726-735, 1983. ing them susceptible to phospholipase attack. Biochem. Biophys. Res. Commun., 19. Aukema, H. M., Yamaguchi, T., Takahashi, H., Celi, B., and Holub, B. J. Abnormal 125: 836-842, 1984. lipid and fatty acid compositions of kidneys from mice with polycyotic kidney 31. Go, M., Sekiguchi, K., Nomura, H., Kikkawa, U., and Nishizuka, Y. Further studies disease. Lipids, 27: 429-435, 1992. on the specificity of diacylglycerol for protein kinase C activation. Biochem. Biophys. 20. Preiss, J. E., Loomis, C. R., Bell, R. M., and Niedel, J. E. Quantitative measurement Res. Commun., 144: 598-605, 1987. of Â«1-1,2-diacylglycerols. Methods Enzymol., 141: 294-300, 1987. 32. Bell, M.. and Sargent, J. R. Effects of the fatty acid composition of phosphatidylserine 21. Kennerly, D. A., Parker, C. W., and Sullivan, T. J. Use of diacylglycerol kinase to and diacylglycerol on the in vitro activity of protein kinase C from rat spleen: quantitatc picomole levels of 1,2-diacylglycerol. Anal. Biochem., 98: 123-131, 1979. Influcences of (n-3) and (n-6) polyunsaturated fatty acids. Comp. Biochem. Physiol. 22. Ford, D. A., Miyake, R., Glaser, P. E., and Gross, R. W. Activation of protein kinase B, 86: 227-232, 1987. C by naturally occurring ether-linked diglycerides. J. Biol. Chem., 264: 33. Zoeller, R. A., and Wood. R. Analysis of the stearoyl-CoA desaturase system in the 13818-13824, 1989. Morris hepaloma 7288C and 7288CTC. Lipids, 19: 488-491, 1984. 23. Daniel. L. W., Small, G. W., Schmitt, J. D., Marasco, C. J., Ishaq, K., and Piantasdosi, 34. Zoeller, R. A., and Wood, R. The importance of the stearoyl-CoA desaturase system C. Alkyl-linked diglycerides inhibit protein kinase C activation by diacylglycerols. in octadecenoate metabolism in the Morris hepatoma 7288C. Biochim. Biophys. Acta, Biochem. Biophys. Res. Commun., 151: 291-297, 1988. 845: 380-388, 1985.

2298

Jennifer S. Pickering, Joanne R. Lupton and Robert S. Chapkin

Cancer Res 1995;55:2293-2298.

Updated version Access the most recent version of this article at: http://cancerres.aacrjournals.org/content/55/11/2293

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/55/11/2293. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.