[CANCER RESEARCH 55, 2293-2298, June I, 1995]

Dietary Fat, Fiber, and Carcinogen Alter Fecal Diacylglycerol Compositionand 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 77843-247Ì

ABSTRACT

Fecal diacylglycerols (DAGs) are known activators of protein kinase C(PKC), which in turn modulates colonie epithelial cell growth programsand, therefore, could play a role in the malignant transformation process.However, the effects of physiological modifiers such as diet and carcinogenon fecal DAG mass and composition have not been reported. We thereforedesigned a 2 x 2 x 2 factorial study (2 fats: corn oil and fish oil; 2 fibers:pectin and cellulose; with and without carcinogen). Rats were providedwith diets for 5 weeks. Three weeks after the second injection ofazoxymethane, feces were collected from 10 rats/treatment in = 80 total)

and analyzed for DAG mass and fatty acyl composition by combined TLCand gas chromatography. Dietary fat had a significant effect on the mol %fatty acyl composition of fecal DAG. Greater amounts of long chain n-3polyunsaturated fatty acids (20:Sn-3, 22:Sn-3, and 22:6n-3) were detectedin fecal DAG of fish oil-fed animals relative to corn oil ( /' < 0.001). In

contrast, corn oil resulted in a higher mol % of 18:2n-6 relative to fish oil

(P < 0.016). The most salient effect of fiber was on total production(nmol/day) of DAG, which was 2.5 times higher with cellulose than pectinsupplementation. In addition, there was an effect of fiber on both mol %and concentration of 22:6n-3, with cellulose producing higher amounts

relative to pectin (P < 0.04). A significant interaction between fat and fiberwas observed with nmols of 17:0 excreted in 24 h, with fish oil/celluloseproducing 94.2 nmol as compared to 3.5 seen with corn oil/pectin(P < 0.02). There was a significant interaction between fat and carcinogenon all of the DAG n-3 fatty acids, which were elevated with carcinogen/fish

oil treatment. These data show that fat, fiber, and carcinogen can modulate the fatty acyl composition and mass of fecal DAG. Since the production of fecal DAG, an activator of PKC, may alter colonie mucosa! cellproliferation, our data offer insight into a mechanism by which diet maymodify the risk of colon cancer development.

INTRODUCTION

PKC'1 consists of a family of serine/threonine kinases that upon

activation can phosphorylate proteins that regulate short- and long-

term cellular events, including cell proliferation and differentiation. Itis now clear that PKC expression and activation are involved in theregulation of colonie mucosa! proliferation (1-4). This is noteworthy

because the elevation of colonie cell proliferation has been associatedwith the incidence of colon cancer (5).

DAGs link extracellular signals with intracellular responsesthrough the activation of PKC (6). Intracellular DAG produced frominositol and choline phospholipid hydrolysis, coupled with increasedCa2+ levels, can activate PKC to influence cellular processes (7). It is

interesting that several studies have recently shown that rat and humanfeces contain appreciable levels of DAG (8-11). This is significant

Received 1/4/95; accepted 4/4/95.The costs of publication of this article were defrayed in part by the payment of page

charges. This article must therefore be hereby marked advertisement in accordance with18 U.S.C. Section 1734 solely to indicate this fact.

1This work was supported in part by grants from the American Institute for Cancer

Research, the Texas A&M Interdisciplinary Research Initiatives Program, and NIH GrantsCA59034 and CA61750.

2 To whom requests for reprints should be addressed, at 442 Kleberg Center, Texas

A&M University, College Station, TX 77843-2471.3 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,

anilino-1-naphthalene sulphonic acid ammonium salt; AOM, azoxymethane; PUFA,

polyunsaturated fatty acid.

because DAG in the colonie lumen is capable of entering mucosalcells of the colon (9), where it can activate the PKC pathway (12). Inaddition, extracellular DAGs of the chain length found in feces canenhance the growth of benign colonie tumors and some carcinomas,while normal cells are not affected (13). These studies suggest thatfecal intraluminal DAG may have implications in the development ofcolon cancer.

DAG has been detected in foods at appreciable levels, althoughonly negligible amounts of dietary DAG reach the colon intact (9).Sources of DAG in the lumen have not been elucidated, although itmay be derived partly from bacterial action (8). As a result of theseobservations, interest has focused recently on the ability of dietaryconstituents to modulate intraluminal DAG levels. For example,Reddy et al. (10) have demonstrated that the type of fiber consumedcan modify fecal DAG mass and composition.

Dietary fat and fiber have interactive effects in the colon (14) andare capable of influencing colonie PKC isoform expression (15).However, the combined effects of dietary fat and fiber on fecal DAGmass and composition have not been examined. Therefore, the presentstudy was designed to investigate the effects of dietary fat and fiber onthe mass and fatty acyl composition of fecal DAG in the rat experimental colon carcinogen model. In addition, fecal diglyceride subclasses were characterized to determine aliphatic linkage at the sn-\

chemical position.

MATERIALS AND METHODS

Materials. l,2-diheneicosanoyl-.vn-glycero-phosphocholine (21:0-21:0-PC)

was purchased from Avanti Polar Lipids (Alabaster, AL). Phospholipase C (TypeV, Bacillus Cereus) and azoxymethane were from Sigma Chemical Co. (St. Louis,MO). Silica gel 60 plates (20 x 20 cm, 0.25-mm thickness) for TLC were from

Merck (Darmstadt, Germany). Bakerbond silica gel columns were purchased fromJ. T. Baker, Inc. (Phillipsburg, NJ). Dextrose was from HaríanTeklad (Madison,WI). Cellulose, casein, DL-methionine, choline bitartrate, AIN-76 vitamin, andAIN-76 mineral mix were purchased from Bio-Serve (Frenchtown, NJ). Corn oil

was kindly provided by Traco Labs (Seymour, IL). Vacuum-deodorized Menha

den fish oil was provided by the NIH Fish Oil Test Material Program, SoutheastFisheries Center (Charleston, SC). High methoxylated pectin was from Grindsted(Industrial Airport, KS).

Animals and Diets. The animal use protocol for these experiments wasapproved by the University Animal Care Committee at Texas A&M University. All animals were treated in accordance with the NIH Guide for the Careand Use of Laboratory Animals (NRC 1985). Eighty male weanling Spraguc-

Dawley rats (Harían, Houston, TX) were housed individually in cages in atemperature- and humidity-controlled facility with a 12-h light/dark cycle.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 2factorial design with 2 fats, 2 fibers, with or without carcinogen. The dietcompositions are shown in Table 1. The two fats were fish oil (11.5 g/100 gdiet) and corn oil (15 g/100 g diet). The fish oil diet contained 3.5 g cornoil/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(0.025%) and vitamin E (1.5 mg/g a-tocopherol and 1.0 mg/g y-tocopherol).

The two fibers were pectin and cellulose (6 g/100 g diet). Each group wasdivided further into + or - carcinogen treatment. Food and water were

available freely.Carcinogen Administration. After 1 week of consuming semipurified

diets, the rats were given injections 2 times at 1-week intervals with

2293

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

FECAL DAG MODULATION BY DIET AND CARCINOGEN

Table 1 Modified AIN-76dietIngredientsDextroseCaseinDL-MetCorn

oil"Corn/fishoil''Mineral

mixVitaminmixCholine

bitartralePectin/cellulose"

Corn oildiet.hFish oil diet.Percent

byweight51.0622.350.3415.03.5/11.53.911.120.226.0Percentbykcal47.6520.8631.4931.49

azoxymethane at 15 mg/kg body weight or saline. Food intake and fecal outputover 48 h was evaluated 1 week after the second injection. Animals wereweighed on arrival and weekly after their second injection. Three weeks afterthe second injection, feces were collected and processed as described below.

Fecal Sample Collection and Lipid Extraction. In order to quantitatefecal diacylglycerols, 3 weeks after the second injection fecal samples fromindividual animals were collected and weighed immediately after defecation.Total fecal lipids were extracted by the method of Florin-Christensen (16). Aknown mass of 21:0-21:0 diacyl-j/i-glycerol internal standard was added to

quantitate fecal diacylglycerol. The internal DAG standard was synthesizedfrom 21:0—21:0-PC by phospholipase C treatment as described by Akoh andChapkin (17). Briefly, 400 ng of 21:0-21:0-PC were dissolved in 2 mlperoxide-free diethyl ether containing 0.005% butylated hydroxytoluene and 2ml of 50 mM Tris-HCI (pH 7.4) containing 5 HIMcalcium chloride and 10 unitsof phospholipase C (Bacillus cereus). The 2-phase mixture was agitated in a

shaking water bath for 3 h at room temperature. The resultant diacylglycerolswere extracted twice with 2 ml diethyl ether. Following the addition of internalstandard, samples were homogenized in 1.0 ml of ice-cold methanol/water

(1.0:0.9, v/v). After vortexing, 1.1 ml of chloroform was added to each sampleand an additional 1.0 ml ice-cold methanol/water (1.0:0.9, v/v) was added. Thefecal homogenates were centrifuged at 800 X g for 5 min at 4°C.The lower

organic phase was removed and dried under nitrogen. The lipid extracts wereresuspended in chloroform and passed through chloroform-washed silica gel

extraction columns (Bakerbond, Phillipsburg, NJ) with the use of a vacuummanifold. Fecal diacylglycerols were eluted from the column using chloro-form/methanol (90:10, v/v) and normalized for 48-h fecal output.

Diacylglycerol Isolation. Fecal diacylglycerols were separated by TLC onsilica gel 60 plates with the use of benzene-chloroform-methanol (80:15:3.25,

v/v/v) (14). The DAG band was detected under UV light after spraying with 1g/liter ANS in water (pH 7.0) and subsequently transmethylated with 3 ml ofmethanol/sulphuric acid (94:6, v/v). The resultant fatty acid methyl esters wereextracted with the use of 3 ml of hexane and 1 ml of 0.1 mol/liter potassiumchloride. The fatty acid methyl ester extracts were dried under nitrogen andresuspended in 25 fil of chloroform/methanol (2:1, v/v). The fatty acid methylesters were separated further on TLC with the use of toluene and were detectedunder UV light after they were sprayed with ANS. Fatty acid composition andDAG concentration were determined by gas Chromatographie analysis(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 andseparated subsequently by TLC (17). Briefly, diglycerides were incubated with0.2 ml of benzene containing 10 mg benzoic anhydride and 0.1 ml of benzenecontaining 4 mg of 4-dimethylaminopyridine at room temperature for 1 h.

After incubation, 2 ml of hexane was added; excess benzoic anhydride washydrolyzed by adding 1 ml of concentrated ammonium hydroxide while slowlyagitating. 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 internalstandards. In addition, diglyceride samples were also converted to diglycerideacetates by acetolysis to corroborate subclass identification (19). Briefly,diglyceride samples were incubated with 1 ml of acetic anhydride and 4—5drops of pyridine for 75 min at 80°C.After cooling, 1 ml of methanol was

added to convert any unreacted acetic anhydride to methyl acetate. Sampleswere 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, petroleumether was added, and the upper phase was extracted. The samples wereseparated into diglyceride-acetate subclasses with the use of a triple develop

ment TLC system consisting of toluene development, followed by hexane/ether (4:1, v/v) and toluene. Diglyceride subclasses were visualized with ANS.Finally, the diglyceride kinase assay as described by Priess el al. (20) wasutilized as a third experimental procedure for quantitating diglyceride subclassmass. This assay uses Escherichia coli 1,2-diacyl-j/i-glycerol kinase, whichphosphorylates both l-alkyl-2-acyl-.ç/i-glycerol and 1,2-diacyl-sn-glycerol analogues. 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 radio-

labeled products of the diglyceride kinase assay were subjected to a mildalkaline hydrolysis in order to separately quantitate the relative contribution ofdiacylglycerols (alkaline labile) and alkylacylglycerols (alkaline stable) to thetotal radiolabled phosphatidic acid generated (21).

Statistics. The data were analyzed with the use of Statistical Analysissoftware by three-way ANOVA. If the P values for the interactions were

<0.05, means of the eight treatment groups were then separated using Duncan's Multiple Range test. If the P values were <0.05 for the effect of fat, fiber,

or carcinogen but not for the interactions, total means of the fat groups, fibergroups, or carcinogen groups were separated with the use of Duncan's Multiple

Range test.

RESULTS

Dietary Fatty Acids. The dietary fatly acids are shown in Table 2.Diets supplemented with corn oil contained linoleic acid (18:2n-6) as

the major fatty acid constituent. With respect to polyunsaturated fattyacid content, fish oil-supplemented diets contained primarily eicosa-pentaenoic acid (20:5/1-3) and docosahexaenoic acid (22:6/1-3), andhad reduced levels of 18:2/i-6 relative to corn oil diets.

Diglyceride Subclass Characterization. Diglyceride subclasscomposition was determined following conversion to diglyceroac-etates, diglycerobenzoates, or [32P]phosphatidic acid via diglyceride

kinase. Subclasses from selected lipid extracts were separated byTLC, and autoradiographs were identified by comparison to authenticstandards (Fig. la). Only minor levels of lyso-phosphatidic acid

(<10% of total phosphatidic acid dpm) were observed followingalkaline hydrolysis, indicating a relative absence of l-O-alkyl-2-acyl-s/¡-glyceroland l-O-alkenyl-2-.sn-glycerol in fecal lipid extracts (Fig.

Table 2 Weight percent of fatty acids in the experimental diets

Fatty acids are expressed as number of carbon atoms:number and position of doublebonds. Fatty acid methyl ester peak areas were quantitated by gas chromatography asdescribed in "Materials and Methods."

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

2294

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

FECAL DAG MODULATION BY DIET AND CARCINOGEN

a

•1

-92

m2

•II

(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 andcarcinogen 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 3«).In general, animalsfed 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 % ofDAG-containing 18:l«-7 and 22:6n-3 in pectin as compared to cellulose animals (Table 3b). Specifically, 18:ln-7 was higher in pectin-fed animals and 22:6n-3 was higher in cellulose-fed animals. Fibercomposition 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 significantelevation in 20:5;i-3 was observed in pectin-fed animals, while 22:6«-3was higher in the cellulose-fed animals. When expressed as nmol

of fatty acid from DAG excreted in 24 h, only 16:0 was significantlyinfluenced by fiber, with means higher in the cellulose diets (data notshown). In addition, the mass of fecal diacylglycerols excreted in24 h was affected significantly (P < 0.05) by fiber (Table 6). Specifically, fecal DAG was significantly higher in animals fed cellulose-

containing diets (Fig. 2).Carcinogen Effect. AOM injection significantly affected

(P < 0.05) the mol % levels of 18:ln-9-diacylglycerol (data notshown). The mean mol % level of 18:ln-9 in the AOM-treated

animals was 20.5 ±1.5 compared to saline treatment at 15.9 ±1.5(n = 5-10).

Fat-Fiber Effect. The mol % of DAG-derived 17:0, 18:0, and 20:0affected (P < 0.05) by a fat-fiber interaction are shown in Table 5«.Afat-fiber effect on 17:0 was evident when expressed as mol %, nmol/gwet weight feces, and 24-h excretion (Table 5, b and c). The highest17:0-containing diacylglycerols were associated with the fish oil/

)— 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 (Table5b). Diacylglycerol-derived fatty acids significantly affected by fat-fiber interaction over a 24-h excretion period were 17:0, 18:3n-3,

•<—L-PA 20:0, 22:5i!-3, 22:6«-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.

<— Origin Fat-Carcinogen Effect. A fat-carcinogen interaction significantly

affected (P < 0.05) the nmol of 22:5n-3 per g wet weight fèces(data

h- C-1-P

L-PA

Origin

PA

Fig. La, detection of fecal diglycerides by conversion lo |'-P]phosphatidic acid. Fecal

lipids were extracted and converted to radiolabeled phosphatidic acid with the use of £.Coli diglyceride kinase as described in "Materials and Methods." 12Plipids were separated

by TLC and exposed to X-ray film. Results are autoradiographs from representativeexperiments repeated 3 times. Sample 1. authentic DAG standard: l-stearoyl-2-arachido-noyl-.v/i-glycerol; Sample 2. fecal lipid extracts from 3 individual rats. PA. phosphatidicacid; C-I-P. ceramide-1-phosphate: L-PA, lyso-phosphatidic acid. b. characterization offecal diglyceride subclasses. The aliphatic chain linkage at the m-1 position was determined by conversion to [32P]phosphatidic acid with the use of E. Coli diglyceride kinase

as described in ti. Selected samples containing radiolabeled products were subjected toalkaline hydrolysis in order to quantitate the relative levels of l-acyl-(alkaline labile).l-alkyl-(alkaline stable), and l-alkenyl-(alkaline stable) species. Lune I, authentic untreated DAG standard: l-stearoyl-2-arachidonoyl-.wi-glycerol; Lune 2, alkaline hydrolysisof DAG standard; Lane 3, alkaline hydrolysis of AAG standard: l-hexadecyl-2-linoleoyl-j;j-glycerol: Lane 4, alkaline hydrolysis of A'AG standard: 1-plasmenyl-2-acyl-.vn-glycerol

(derived from bovine heart PC): Ulne 5. fecal extract; Lune fi. fecal extract.

lb). Greater than 90% of the subclasses were of the diacyl configuration. These analyses were confirmed by TLC of fecal diglyceroben-

zoates and diglyceroacetates (data not shown).Fat Effect. Fecal diacylglycerol composition (Tables 3-5) and

mass (Table 6) were affected significantly by dietary fat (P < 0.05).The data are expressed as mol % composition (Table 3, a and b, andTable 5«),nmols of fatty acid from DAG per g wet weight feces

Table 3a Effect offal on mol 7c of fully acid derived from fecal uiacylglycenil"

Fatty acid Corn oil Fish oil P value

14:018:2n-6

22:5/i-322:6ii-324:1)1-9

2.6 ±0.615.7 ±2.10.1 ±0.4

tr"

0.1 ±0.4tr"

5.0 ±0.68.2 ±2.13.4 ±0.41.4 ±0.23.6 ±0.40.9 ±0.3

(1.0063

0.0160

O.IHKI1

II.IXKI!

O.INKII

0.0396

Table 3b Effect of fihe r on mol % offaltv adii derived fromfecal diacylglycerol

Fatty acid Cellulose Pectin P value

18:l«-722:6n-32.8

±1.22.5 ±0.46.4

±1.41.2 ±0.50.0543 0.0397

Table 3c Effect offal on nmols fatty acid derivai from tiiacylglycerolper g feces

Fattyacid20:5n-322:6n-3

24:ln-9Corn

oil0.4

±2.10.3 ±1.73

trFish

oil17.5

±2.015.5±1.72.8 ±0.9P

valueO.OW1

(UXK)I0.0373

" Data are shown as mean ±SE with analyses from n - 5-10. Only significant effects

(P < 0.05) are shown.'' tr. trace amounts [<0.1% of total fatty acids present (Table 3a) or <0.1 nmol/g feces

(Table 3c)].

2295

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

FECAL DAG MODULATION BY DIET AND CARCINOGEN

Table 4 Effect offiber on nmols of fairy acid derived from diacylglycerol per g feces"

Fatty acid Cellulose Pectin P value

20:5n-322:6n-35.7+1.910.5 ±1.612.2

±2.25.3 ±1.80.0315 0.0374

" Refer to Table 3 for legend details.

not shown). The corn oil-fed animals treated with AOM had traceamounts of 22:5n-3, while the corn oil treatments with saline hadsignificantly higher mean values. The fish oil-fed animals treated withAOM had significantly higher levels of 22:5n-3 than did fish/saline,corn/AOM, or corn/saline animals. Significant fat-carcinogen effectswere also observed with respect to the 24-h excretion of DAG-containing 20:5n-3, 22:5n-3, 22:6n-3, and 24:0 (Table 5d). In general,DAG-derived n-3 fatty acids were significantly higher in fish oil-fed

animals treated with AOM.Fiber-Carcinogen Effect. The mol % of 18:0 was significantly

affected (P < 0.05) by fiber-carcinogen (data not shown). The saline-

treated animals had higher means in both cellulose and pectin diets,while AOM animals were lower in their mol % of 18:0-DAG. The

nmol of 24:0 per g wet weight feces were affected significantly by afiber-carcinogen interaction (data not shown). Pectin-fed animals had

with saline (12.2 ±4.3) as compared with AOM-treated animals

(3.3 ±4.3).

DISCUSSION

The results of the present study demonstrate that fecal DAGcomposition and mass can be influenced by altering dietary fat andfiber composition. As indicated in Fig. 1, a and b, the diglycerideslocated in the colonie lumen were predominantly (>90%) of the1,2-diacyl-irc-glycerol configuration (DAG). Fecal AAG andA'AG species were present only in minor amounts. The charac

terization of diglyceride sn-1 aliphatic linkage is important becauseether-linked diglycerides (AAG and A'AG) are effective modula

tors of PKC, with separate and distinct calcium requirements incomparison to DAG (22, 23).

Dietary incorporation of fatty acids into fecal DAG was especiallyevident with n-3 fatty acids. Specifically, in fish oil-fed animals, DAGcontaining n-3 PUFA expressed on a mol %, nmol/g wet weight feces,

and nmol of fatty acid excreted over 24 h were increased significantly(P < 0.01) compared to corn oil animals. Previous studies have shown

Table 5a Effect offat-fiber interaction on mol % of fatty acid derived from fecaldiacylgfycerofFatty

acid17:0

18:020:0Corn

oil/cellulose0.4±1.7A

17.8±1.8A2.1 ±1.3ABCorn

oil/pectin1.1±1.9A

12.7 + 2.0AB0.5 +1.4ATable

5b Effect offat-fiber interactiononFatty

acid17:0

24:0Corn

oil/cellulose3.1±11.8A

9.3 ±3.8ACorn

oil/pectin5.0±13.2A

1.9±4.3AFish

oil/cellulose16.0±1.5B

9.7 ±1.6B0.7±1.2AFish

oil/pectin4.6+1.9*

13.4 + 2.0AB5.5 ±1.4"P

value0.0012

0.02300.0204nmol

fatty acid derived from diacylglycerol per gfecesFish

oil/cellulose70.3±11.3B

2.7 ±3.5AFish

oil/pectin18.2±13.2A

13.5±4.3BP

value0.0355

0.0265Table

5c Effect of fat-fiber interaction on nmol offatty acid from fecal diacylglycerol excreted in 24hFatty

acid17:0

18:3n-320:022:5n-322:6n-324:0Fatty

acid20:5«-3

22:5n-322:6n-3

24:0"Data are shown aCorn

oil/cellulose6.0±15.6A

0.1 ±1.4*15.8 ±4.0*trA.<,0.8

±3.2A11.4±3.7ATable

5d Effect of fat-carcinogenCorn

oil/pectin3.5±17.5A

3.1 ±1.4B1.9±4.5B

ti^tr*

1.7±4.2Binteraction

onCorn

oil/AOM Cornoil/saline0.6±3.5Atr*tr*

3.^4^smean + SE with analyses from n = 5-10. Vt^

t^0.8 ±3.3A9.9 ±3.9BFish

oil/cellulose94.214.5C

2.7 1.1*7.2 3.7B8.3 1.0*

28.6 2.9B3.5 3.5BFish

oil/pectin14.3±17.5"

0.3 + 1.4*15.3 + 4.5*3.1 ±1.2A8.0 + 3.6C

11.6±4.2AP

value0.0224

0.04270.01220.01370.0032

0.0196nmol

of fatty acid from fecal diacylglycerol excreted in 24hFish

oil/AOM23.4+ 3.4B

9.0±1.2B25.3 ±3.3B12.1 ±3.9Bilues

sharing any common capital-letter superscriptFish

oil/saline10.8±3.3C

2.3 + 1.0C11.4±3.2C3.0 ±3.7Aare

not significantly different (P > 0.05).P

value0.0503

0.00690.05180.0246

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'

425.6 ±87.2 573.5 ±169.4

212.7 ±43.6 286.9 ±84.9

Total nmol fatty 653.7 ±367.4 462.7 + 99.6 493.4 + 99.5 844.4 + 373.0 977.1 ±487.1acid'Vg

wet fecesTotal nmol 327.7 ±184.5 231.5 ±49.8 246.8 ±49.7 422.6 ±186.5 487.8 ±243.4

DAG/gwet feces

Total nmol 1257.0 ±425.3A 333.0 + 82.4BC 625.5 ±171.6*BC 587.3 ±240.8ABC 965.4 ±508.2*" 276.0 ±29.0BC 563.3 ±

DAG in 24 h" Data are shown as mean ±SE with analyses from «= 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

488.2 + 68.8

189.8 ±27.9

142.4 ±20.6e

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

FECAL DAG MODULATION BY DIET AND CARCINOGEN

700

600-

•o0)

uxUJ 400-

raU

300-

200-

X 100oEe

Cellulose Pectin

Fig. 2. Etica of dietary fiber on fecal DAG mass. Columns, mean nmol DAG excretedin 24 h; bars, SE. For cellulose. n = 30, and for pectin, n = 22. Fiber effects:cellulose > pectin at P < 0.05.

(24-26), with AOM tumor incidence decreased in rats treated witheither 20:5/i-3 (27) or 22:6n-3 (28). Consistent with the protectiveeffect of fish oil-derived /¡-3PUFA, we have shown that fish oil-fed

rats consistently have the lowest levels of colonie cell proliferation(14, 15). These findings suggest a potential role for lipid nutrition incancer prevention. Whether fecal DAG species containing n-3 PUFA

have unique biological properties has not been determined. This is anintriguing question since fecal DAG is capable of entering colono-

cytes and activating PKC (9, 12), which in turn may mediate coloniecell proliferation (4, 15). It is interesting that the fatty acyl composition of DAG is a key determinant of its fate within intact cells andmay modulate the metabolic termination of DAG-mediated intracel-lular signaling (16, 29). Therefore, diet-induced alteration of DAG

composition could result in the differential activation of cellularphospholipases (30) and PKC (31, 32) and may selectively modulatethe growth programs of colonie epithelial cells (13).

Analysis of fecal DAG composition also revealed the presenceof an uncommon, odd chain fatty acid, i.e., heptadecanoic acid(17:0). The level of 17:0-DAG was enriched highly in fecesisolated from fish oil-fed animals (Table 5, a and b), with the

highest levels associated with fish oil/cellulose animals. Sincefecal DAG is generated in part by colonie microflora (8) andnegligible amounts of 17:0 are found in the diet (Table 2), theenrichment of a specific population of microflora by feeding a fishoil/cellulose combination may have generated 17:()-DAG. Although the biological properties of 17:0-DAG remain to be deter

mined, it is possible that this novel odd chain second messengercould play a protective role in colon carcinogenesis since DAGscontaining saturated fatty acyl species generally have a suppressedability to activate PKC (7). This is based on the fact that the Kuctand Vm.Mvalues of unsaturated and saturated DAG molecularspecies are distinct (31, 32).

The major effect of fiber was on the total excretion (nmols/day) offecal DAG, which was approximately 2.5 times higher with celluloserelative to pectin supplementation. An enhancement of the fecalexcretion of DAG following cellulose feeding could be explained byeither an increase in luminal DAG synthesis or a decrease in theuptake of DAG by colonocytes. It is possible, therefore, that dietarycellulose may reduce colon cancer risk in part by decreasing themucosal uptake of luminal DAG. Such an effect potentially couldinfluence colonocyte PKC-dependent signal transduction. This is

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

type could modulate the uptake of DAG by colonocytes is intriguingand requires futher investigation.

Diet composition and carcinogen exposure had no effect on theluminal concentration of DAG (nmol DAG/g feces). This was unexpected because high luminal levels of DAG may enhance colonie cellproliferation (11), and cellulose generally is considered to reducecolon cancer risk. With respect to the effect of carcinogen, there wasan interactive effect of fat and carcinogen on the levels of 20:5n-3,22:5;j-3, and 22:6/¡-3-DAG. The significance of this observation re

mains to be determined. In addition, the mol % composition ofmonoenoic 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

from the A9 desaturation of stearic acid (18:0). Previous investigatorshave demonstrated that A9 desaturase is elevated in malignant transformed cells, thereby increasing the conversion of 18:0 to 18:1/1-9 (33,

34). It is possible, therefore, that colonie microbial A9 desaturaseactivity is activated in AOM-injected rats relative to saline controls.

The specific effects of carcinogen on luminal bacterial enzymesrequires further study.

In conclusion, we demonstrate for the first time that dietary fat,fiber, and carcinogen treatment alter fecal DAG mass and composition. Since the production of fecal DAG, an activator of PKC, mayalter colonie mucosal cell proliferation, our data offer insight into amechanism by which diet may modify the risk of colon cancerdevelopment. It is essential, therefore, to determine the levels ofspecific fecal DAG molecular species so that the mechanisms regulating colonie PKC activation and the rates of cell proliferation,differentiation, and apoptosis can be elucidated.

ACKNOWLEDGMENTS

We gratefully acknowledge Dr. Harold Aukerna. Chris Jolly. Meng ChaoLee. and Stella Wiese for excellent laboratory assistance. We also acknowledge the generous donation of dietary corn oil by Sid Tracy (Traco Labs).

REFERENCES1. Guillem, J. G.. O'Brian, C. A.. Fitzer, C. J., Förde,K. A.. LoGerfo, P., Treat, M., and

Weinslein. I. B. Altered levels of protein kinase C and Ca2 ' -dependent protein

kinases in human colon carcinomas. Cancer Res., 47: 2036-2039, 1987.2. Craven, P. A., and DeRubertis, F. R. Suhcellular distribution of protein kinase C' in

rat colonie epithelial cells with different prolifcrative activities. Cancer Res., 47:3434-3438, 19x7.

3. Chapkin. R. S.. Gao. J.. Lee, D-Y. K.. and Lupton, J. R. Dietary fibers and fats alter

rat colon protein kinase C activity: correlation to cell proliferation. J. Nutr.. 123:649-655. 1993.

4. Davidson. L. A.. Jiang. Y. H.. Derr. J. N.. Aukema. H. M.. Lupton. J. R., and Chapkin.R. S. Protein kinase C isofornis in human and rat colonie mucosa. Arch. Biochem.Biophys., 312: 547-553, 1994.

5. Lipkin. M. Biomarkers of increased susceptibility to gastrointestinal cancer: newapplication to studies of cancer prevention in human subjects. Cancer Res., 4K:235-245, 1988.

6. Liscovitch, M., and Cantley. L. C. Lipid second messengers. Cell, 77: 329-334,1994.7. Nakamura. S.. and Nishizuka. Y. Lipid mediators of protein kinase C activation for

the intracellular signaling network. J. Biochem., 1154: 1029-1034, 1994.8. Morotomi. M.. Guillem. J. G., LoGerfo, P.. and Weinstein, 1. B. Production of

diacylglyccrol. an activator of protein kinase C. by human intestinal microflora.Cancer Res., 50: 3595-3599, 1990.

9. Morotomi. M.. LoGerfo. P., and Weinstein, I. B. Fecal excretion, uptake and metabolism by colon mucosa of diacylglycerol in rats. Biochem. Biophys. Res. Commun..IXI: 1028-1034. 1991.

10. Reddy. B. S.. Simi. B.. and Englc. A. Biochemical epidemiology of colon cancer:effect of types of dietary fiber on colonie duicylglyccrols in women, (iastroenterol-ogy. 106: 883-889, 1994.

11. Steinbach. G.. Morotomi, M.. Nomoto. K„Lupton. J. R., Weinstein. I. B.. and Holt,P. R. Calcium reduces the increased fecal 1.2-.m-diacylglyccrol content in intestinalbypass patients: a possible mechanism for altering colonie hyperproliferation. CancerRes., 54: 1216-1219. 1994.

12. DeRubertis. F. R.. and Crowen. P. A. Relationship of bile salt stimulation of colonieepithelial phospholipid turnover and proliferativc activity: role of activation of protein

2297

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

FECAL DAG MODULATION BY DIET AND CARCINOGEN

kinase C. Prev. Med., 16: 572-579, 1987.

13. Friedman, E., Isaksson, P., Rafter, J., Marian, B., Winawer, S., and Newmark, H.Fecal diglycerides as selective endogenous mitogens for premalignant and malignanthuman colonie epithelial cells. Cancer Res., 49: 544-548, 1989.

14. Lee, D. Y., Chapkin, R. S., and Lupton, J. R. Dietary fat and fiber modulate coloniecell proliferation in an interactive site-specific manner. Nutr. Cancer, 20: 107-118,

1993.15. Davidson. L. A., Lupton, J. R., Jiang, J. H., Chang, W. C., Aukema. H. M., and

Chapkin, R. S. Dietary fat and fiber alter rat colon protein kinase C isozymeexpression. J. Nutr., 125: 49-56, 1995.

16. Florin-Christensen, J. F., Florin-Christensen, M. F., Delfino, J. M., Stegmann, T., and

Rasmussen, H. Metabolic fate of plasma membrane diacylglycerols in NIH 3T3fibroblasts. J. Biol. Chem., 267.- 14783-14789, 1992.

17. Akoh, C. C., and Chapkin, R. S. Composition of mouse peritoneal macrophagephospholipid molecular species. Lipids, 25: 613-617, 1990.

18. Chapkin, R. S., Haberstroh, B„Liu, T., and Holub, B. J. Characterization of theindividual phospholipids and their fatty acids in serum and high-density lipoprotein ofthe renal patient on long-term maintenance hemodialysis. J. Lab. Clin. Med., 101:726-735, 1983.

19. Aukema, H. M., Yamaguchi, T., Takahashi, H., Celi, B., and Holub, B. J. Abnormallipid and fatty acid compositions of kidneys from mice with polycyotic kidneydisease. Lipids, 27: 429-435, 1992.

20. Preiss, J. E., Loomis, C. R., Bell, R. M., and Niedel, J. E. Quantitative measurementof «1-1,2-diacylglycerols. Methods Enzymol., 141: 294-300, 1987.

21. Kennerly, D. A., Parker, C. W., and Sullivan, T. J. Use of diacylglycerol kinase toquantitatc picomole levels of 1,2-diacylglycerol. Anal. Biochem., 98: 123-131, 1979.

22. Ford, D. A., Miyake, R., Glaser, P. E., and Gross, R. W. Activation of protein kinaseC by naturally occurring ether-linked diglycerides. J. Biol. Chem., 264:13818-13824, 1989.

23. Daniel. L. W., Small, G. W., Schmitt, J. D., Marasco, C. J., Ishaq, K., and Piantasdosi,C. Alkyl-linked diglycerides inhibit protein kinase C activation by diacylglycerols.Biochem. Biophys. Res. Commun., 151: 291-297, 1988.

24. Deschner, E. E., Lytle, J. S., Wong, G., Reperto, J. F., and Newmark, H. L. The effectof dietary to-3 fatty acids (fish oil) on azoxymethanol-induced focal areas of dysplasiaand colon tumor incidence. Cancer (Phila.), 66: 2350-2356, 1990.

25. Reddy, B. S., and Maruyama, H. Effect of dietary fish oil on azoxymethane-inducedcolon carcinogenesis in male F344 rats. Cancer Res., 46: 3367-3370, 1986.

26. Reddy, B. S., and Sugie, S. Effect of levels of ¡a-3and ¡a-6fatty acids on azoxymethane-induced colon carcinogenesis in F344 rats. Cancer Res., 48: 6642-6647, 1988.

27. Minoura, T., Takata. T. Michitomo, S., Takada, H.. Yamamura, M.. Hioki. K., andYamamoto, M. Effect of dietary eicosapentaenoic acid on azoxymethane-inducedcolon carcinogenesis in rats. Cancer Res., 48: 4790-4794, 1988.

28. Takahashi, M.. Minamoto, T., Yamashita, N., Toshio, K., Yazawa, K., and Hiroyase,E. Effect of docosahexaenoic acid on azoxymethane-induced colon carcinogenesis inrats. Cancer Lett., 83: 177-184, 1994.

29. Florin-Christensen, J. F., Florin-Christensen. M. F., Delfino, J. M., and Rasmussen,H. New patterns of diacylglycerol metabolism in intact cells. Biochem. J. 289:783-788, 1993.

30. Dawson, R. M. C., Irvine, R. F., Bray, J., and Quinn, P. J. Long-chain unsaturated

diacylglycerols cause a perturbation in the structure of phospholipid bilayers rendering them susceptible to phospholipase attack. Biochem. Biophys. Res. Commun.,125: 836-842, 1984.

31. Go, M., Sekiguchi, K., Nomura, H., Kikkawa, U., and Nishizuka, Y. Further studieson the specificity of diacylglycerol for protein kinase C activation. Biochem. Biophys.Res. Commun., 144: 598-605, 1987.

32. Bell, M.. and Sargent, J. R. Effects of the fatty acid composition of phosphatidylserineand diacylglycerol on the in vitro activity of protein kinase C from rat spleen:Influcences of (n-3) and (n-6) polyunsaturated fatty acids. Comp. Biochem. Physiol.B, 86: 227-232, 1987.

33. Zoeller, R. A., and Wood. R. Analysis of the stearoyl-CoA desaturase system in theMorris hepaloma 7288C and 7288CTC. Lipids, 19: 488-491, 1984.

34. Zoeller, R. A., and Wood, R. The importance of the stearoyl-CoA desaturase system

in octadecenoate metabolism in the Morris hepatoma 7288C. Biochim. Biophys. Acta,845: 380-388, 1985.

2298

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from

1995;55:2293-2298. Cancer Res   Jennifer S. Pickering, Joanne R. Lupton and Robert S. Chapkin  Composition and MassDietary Fat, Fiber, and Carcinogen Alter Fecal Diacylglycerol

  Updated version

  http://cancerres.aacrjournals.org/content/55/11/2293

Access the most recent version of this article at:

   

   

   

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

  Subscriptions

Reprints and

  .pubs@aacr.orgDepartment at

To order reprints of this article or to subscribe to the journal, contact the AACR Publications

  Permissions

  Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://cancerres.aacrjournals.org/content/55/11/2293To request permission to re-use all or part of this article, use this link

Research. on August 29, 2021. © 1995 American Association for Cancercancerres.aacrjournals.org Downloaded from