Carbohydrate and Fiber

Effects of Diets Rich in Fermentable Carbohydrates on Plasma Lipoprotein Levels and on Lipoprotein Catabolism in Rats1 ANDRZEJMAZÜR, CHRISTIAN RËMËSY,ELYETT GUEUX, MARIE-ANNE CHRISTIAN DEMIGNÊ

LEVRAT AND

Laboratoire des Maladies Métaboliques,Institut National de la Recherche Agronomique, Ceyrat, France

Theix, 63122

Plant fiber consumption in humans is known to be hypolipidemic and is of particular interest in the preven tion and treatment of cardiovascular disease. The effect of plant fibers (cellulose, hemicellulose, lignin, algal polysaccharides, pectins) on cholesterolemia has been reported (1-7), but these effects may be variable, depend ing on the type offiber. Water-soluble fibers (pectin, gum or soybean polysaccharide) have been described as espe cially effective cholesterol-lowering agents. Poorly di gestible starch (high amylose cornstarch or uncooked potato starch) behaves physiologically like easily fer mented dietary fiber and could have a beneficial choles terol-lowering action (8-10). The mechanisms whereby fiber exerts its metabolic effects have been the subject of intensive investigations,- among the various mecha nisms that have been proposed to explain the hypocholesterolemic effect are impaired cholesterol ab sorption, increased neutral steroid losses, decreased postprandial insulin and glucose rises, and decreased hepatic cholesterol synthesis by short-chain fatty acid metabolites (3-5, 7, 11).Conversely, an increase in lipo protein cholesterol catabolism may be a potential pro tective mechanism of dietary fiber in atherosclerosis, but the experimental evidence for such a process is still scarce. In the present experiment, the levels of various poly saccharides were selected to reproduce the consumption of diets rich in plant fiber and to maximize the effect on plasma lipids. We report the influence of a diet rich in fermentable carbohydrates on lipid and lipoprotein lev els and metabolism in the rat.

ABSTRACT This study was conducted to examine the effects of a combination of several dietary fibers (5% guar gum, 5% apple pectin, 15% wheat bran, 22% soybean fiber) and crude potato starch (23%) on plasma lipids and lipoproteins and on liver lipid con centration in rats fed a diet containing either 15% lard or 5% oil with or without dietary cholesterol/cholic acid. Male Wistar rats ate the test diets for 3 wk. The incorporation of fiber and crude potato starch into the diet resulted in a significant enlargement of the cecum; it also increased the concentration of volatile fatty acids and the pool of acetate, propionate and butyrate. Feeding this fermentable carbohydrate decreased plasma cholesterol and triglycéridelevels in rats given a low fat diet and prevented the expected rise in plasma cholesterol and triglycéridesin rats fed cholesterol/ cholic acid or lard. Further studies of high density lipoprotein (HDL) composition, 3-hydroxy-3-methylglutaryl coenzyme A reducÃ-ase (HMG-CoA reducÃ-ase) activity and l-labeled human low density lipoprotein (LDL) turnover were done in the group fed the low fat diet without added cholesterol/cholic acid. The study of the HDL fraction in rats fed a diet rich in fermentable carbohydrates demonstrated a decrease in the HDL, subpopulation and in the proportion of apolipoprotein E. Plasma clearance of intravenously injected I25I-labeled LDL was faster in rats fed this diet than in rats fed the fiber-free diet. In the liver, cholesterol and triglycéridelevels were depressed whereas the activity of HMG-CoA reducÃ-ase was increased. In short, this study suggests that a complex fermentable carbo hydrate formulation has a very efficient cholesterollowering action and affects various mechanisms of cholesterol homeostasis. J. Nutr. 120:1037-1045, 1990. INDEXING KEY WORDS: •fermentable carbohydrates •lipoproteins •plasma clearance •volatile fatty acids •rat

'This research was supported in part by a grant from the Founda tion Française pour la Nutrition.

0022-3166/90 $3.00 ©1990 American Institute of Nutrition. Received 21 December 1988. Accepted 8 March 1990.

1037 Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

1038

MAZUR ET AL. TABLE1 Composition of experimental fiber-free(FF)and fermentable carbohydrate-rich (FC)diets' Oil diets2 FF(-C)

Component

FC(-C)

Lard diets FF (+C)

FC (+C)

FF

FC

g/100 g food CaseinWheat starchCrude starchSoybean potato fiber3GuargumApple pectinWheat bran4Arachide oil5Lard6Cholesterol added7Vitamin/mineral mix8187000000500718023225515500718690000050171802222.55155017186000000015071801322551501507 'Fermentable carbohydrate-rich diet contains 33.5% dietary fiber. 2Diets without (-C) or with cholesterol (+C). 3Sovital (François,Saint Maur, France) contains 75% dietary fiber (88/12 water-insoluble/soluble

components), 12% protein, 8% moisture, 2%

fat and 3% ash. 4Wheat bran (François,Saint Maur, France) contains 47% fiber and 16% protein. 5Arachide oil contains 19% saturated fatty acids. 6Lard contains 1% cholesterol. 'Cholesterol (Cooperation Pharmaceutique Française,Melun, France) with addition of 0.2% cholic acid (Merck, Darmstadt, W. Germany). 'Vitamins and minerals supplied, in mg/kg (except as noted) of diet: thiamin, 20; riboflavin, 15; pyridoxine, 10; nicotinamide, 100; calcium pantothenate, 70; folie acid, 5; biotin, 0.3; cyanocobalamin, 0.05; retinyl palmitate, 1.5; DL-a-tocopheryl acetate, 125; cholecalciferol, 0.15; menadione, 1.5; ascobic acid, 800; p-aminobenzoic acid, 50; myo-inositol, 100; choline, 1.36 & CaHPO4,15 & K2HPO4, 2.5 g; KC1, 5 g,-NaCl, 5 g; MgClz, 2.5 g; Fe2O3, 2.5; MnSO4, 125; CuSO4 5H2O, 25; CoSO4-7H2O, 0.2; ZnSO4-7H2O, 100; KI, O.4.

MATERIALS AND METHODS Animals and diets. Male Wistar rats (IFFA-CREDO, 1'Arbresle, France) weighing -190 g were fed for 3 wk one of six diets,- the diets were either fiber-free or they were rich in fermentable carbohydrates (combination of several dietary fibers and crude potato starch) and con tained 15% lard or 5% oil with or without cholesterolcholic acid. The composition of these diets is shown in Table 1. Rats were housed two per cage in wire-bot tomed cages in a temperature-controlled room (22°C) with the dark period from 2000 to 0800 h. Food and water were provided ad libitum. At the end of the experimental period, rats were sampled (at 0900 h) after maximal food consumption during the dark period. The animals were anesthetized with sodium pentobarbital (40 mg/kg) and maintained at 37°C.After laparotomy, blood was drawn by abdom inal aorta puncture into syringes containing EDTA (1 mg/mL). The liver was excised, and portions from the right lobe were sampled for analysis of triglycéride and cholesterol concentrations. The cecum was removed and weighed, and ~1 g of cecal content was drawn into microfuge tubes. Liver or cecal content samples were immediately plunged into liquid N2 for about 20 s, then stored frozen at -20°Cuntil the assays were performed. Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

Volatile fatty acid analysis. Volatile fatty acids (VFAp were measured by gas-liquid chromatography directly on aliquots of supernatants after the centrifugation of cecal contents at 8000 x g for 5 min at 4°C,as previously described (12). Upoprotein isolation and analysis. Plasma lipoproteins were separated into various density classes by sequential preparative ultracentrifugation (13). Ultracentrifugation was performed at 20°Cin a Beckman Model L-5-50Bultracentrifuge (Beckman Instruments, Palo Alto, CA) with a 50 titanium rotor. Before the lipoproteins were separated, NaN3 (0.02%), merthiolate (0.005%) and Na2-EDTA (0.04%) were added. The plasma samples of two animals were pooled for lipoprotein separation. Samples were overlayered with 0.15 mol/L NaCl-0.01% EDTA (d = 1.006 g/mL), and chylomicrons were recovered following two centrifugations for 30 min at 12,000 x g. Very low density lipoprotein (VLDL)was isolated under the same density conditions for 18 h at 110,000 x g. For isolation of low density

Abbreviations: apo, apolipoprotein; HDL, high density lipopro tein; HMG-CoA, 3-hydroxy-3-methylglutaryl coenzyme A; LDL, low density lipoprotein; VFA, volatile fatty acids,- VLDL, very low density lipoprotein.

FERMENTABLE CARBOHYDRATES AND LIPOPROTEINS )HYD

lipoprotein (LDL, d = 1.006-1.050 g/mL) and high den sity lipoprotein (HDL, d = 1.050-1.21 g/mL), the density was adjusted with solid KBr,and the centrifugation was done at 110,000 x g for 20 h and 24 h, respectively. For lipid and apolipoprotein analysis, fractions were then washed by a further period of ultracentrifugation at the same density. Isolated lipoprotein fractions (except chylomicrons and VLDL) were dialyzed against 0.02 mol/L phosphate buffer (pH 7.4) containing 0.15 M NaCl, 0.01% Na2-EDTA and 0.02% NaN3. The isola tion of the fraction in the density range of 1.050-1.21 g/mL was thus carried out to further appreciate the contribution of HDLi to overall modifications of HDL. This density range is larger than that based on current lipoprotein isolation (d = 1.063-1.21 g/mL), which would exclude that portion of HDL in the rat. Lipoprotein metabolism in rats fed a low fat diet. To determine the influence of a diet rich in fermentable carbohydrates on lipoprotein metabolism, we studied the detailed HDL composition and LDL catabolism in rats fed low fat diets without cholesterol/cholic acid. The molecular size distribution of lipoprotein particles in the HDL fraction isolated by ultracentrifugation was estimated by nondenaturing gradient polyacrylamide gel electrophoresis on Pharmacia precast PAA 4-30% gels (Pharmacia, Uppsala, Sweden) (14).Aliquots of iso lated lipoprotein samples were subjected to electropho resis at 125 V for 17 h in a Tris-borate buffer (pH 8.35). Reference protein standards from a high-molecularweight electrophoresis calibration kit (Pharmacia) were coelectrophoresed with the samples. Gels were fixed in 10% sulfosalicylic acid for 1 h, stained for 4 h in 0.04% Coomassie Brilliant Blue R 250 in perchloric acid (3.5%), and destained in acetic acid (5%). The HDLi subpopulation corresponding to 25.5-14.0 nm and the HDLi subpopulation corresponding to 14.0-8.4 nm were assigned by inspection of the densitometric tracing ac cording to the method of Lowe et al. (15). Sodium dodecyl sulfate-polyacrylamide gel electro phoresis (16) was carried out for apolipoprotein charac terization of ultracentrifugally isolated HDL. The gels after electrophoresis were stained with 0.1% Coomassie Brilliant Blue in methanol/acetic acid/H2O (25/10/65) and destained with methanol/acetic acid/H2O (10/7.5/82.5). Bands were identified by their apparent molecular weights using protein standards run simulta neously (Pharmacia). The gels were then scanned in a Vernon densitometer. Corrections for differences in dye uptake between apolipoprotein bands were not made. The determination of plasma clearance of LDL was conducted using human LDL. Human LDL was chosen in our study as a source of lipoproteins containing apolipoprotein (apo)B because of a significant contami nation of rat LDL preparation by apo E associated with HDL. In rats, LDL is a minor fraction and is difficult to purify because of overlap with the density range of the remnants of triglyceride-rich lipoproteins and HDL. Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

1039

Human serum (generously donated by Centre de Transfusion Sanguine, HôpitalSaint Jacques, Clermont Ferrand) was isolated by low speed centrifugation from the blood of healthy normolipemic male volunteers who fasted overnight. LDL (d = 1.020-1.060 g/mL) was isolated by ultracentrifugation of serum and washed by a further period of ultracentrifugation at 1.060 g/mL density. The isolated fraction was without detectable apo E on sodium dodecyl sulfate-polyacrylamide gel. Radiolabeling was performed using an iodine monochloride method with iodine-125 (carrier free, pH 7-11, Amersham, Buckinghamshire, England) as described by Shepherd et al. (17) to a specific activity ranging from 270 to 670 dpm/ng protein. Rats under ether anesthesia were then injected (via the jugular vein) with a dose of prefiltered (0.45 urn filter, Schleicher & Schuell, Dassel, West Germany), labeled LDL (200 ug proteins, 13 dpm/ng). At intervals during 24 h the rats were anesthe tized with ether, and blood samples (-200 uL) were collected from the contralateral jugular vein and placed in tubes containing 10 units of heparin. Plasma was obtained by centrifugation in an Eppendorf microfuge. 125Iradioactivity was determined in 100-uL samples in a gamma-counter (GAMMAmatic Kontron, Zürich, Switzerland). The results were expressed as a fraction of the 125Iradioactivity in the plasma sample taken 2-3 min after injection. The fractional catabolic rate was calcu lated from the plasma radioactivity decay curve as de scribed for LDL by Sniderman et al. (18). Activity of 3-hydroxy-3-methylglutaryl coenzyme A reducÃ-ase(HMG-CoA reducÃ-ase)in low fat-fed rats. Liver portions for determination of microsomal HMGCoA reductase (EC 1.1.1.34) activity were quickly re moved and homogenized. Liver microsome preparation and assay of microsomal HMG-CoA reductase were carried out as described by Beg et al. (19). Isolated microsomes in the presence of NaF were assayed for ex pressed HMG-CoA reductase activity. Total activity was determined after incubation of the microsomes with partially purified phosphoprotein phosphatase (20). [3-14C]HMG-CoA(Amersham, England) was used as the substrate. Separation of HMG and mevalonolactone was achieved using the chloroform extraction procedure (21) with DL-[2-3H]mevalonolactonate(Amersham, England) as the internal standard to correct for the incomplete recovery. Chemical analysis. Triglycérides(Biotrol, Paris, France), cholesterol and phospholipids (BioMérieux, Charbonnières-les-Bains, France) were determined in plasma and lipoprotein fractions by enzymatic proce dures. The protein concentration of isolated lipopro teins was determined by a modified Lowry method (22). Bovine serum albumin, fraction V, was used as a stan dard. Liver samples were homogenized, and lipids were extracted with chloroform/methanol (2/1, v/v) accord ing to the method described by Folch et al. (23).Triglyc érides in the lipid residue were saponified by 0.5 mol/L

1040

MAZUR ET AL. TABLE 2 Final body weight, cecal development and fermentation in the cecwn of rats fed fiber-free (FF)or fermentable carbohydrate-rich (FC) diets1 Pool in cecal contents

DietFinal

body wt wtgCecal Cecal

diets2FF(-C)FCI-C)FF Oil 4'263± 3'267± ±6«260 (+C)FC ± (+C)ANOVALard 4'NS265

pHTotal

VFAmmol/LAcetatePropionate\imolButyrate

0.2a14.0 ± 1.2b2.6 ± 0.3'13.5 ± LO11F2.5 ±

0.03'5.60 ± 0.07b7.20 ± 0.02a5.60 ± 0.10bF7.30 ±

5«160 ± 6"107± 4'158± 7bF103 ±

81800± SO6105 ± 5'780± 25bF98 ±

3'478± 21b42 ± 3'465± 20bF38 ±

2'193± 12b14 ± 2'180±

0.213.1 ± ±1.3*7.20

0.025.80 ± ±0.10*101

5152± ±8*109

6750± ±31*40

2461 ± ±18*13

1190± ±14*

l&F13 ±

diets3FFFC272 5260 ± ±42.7

'Values are means ±SEMof 12 determinations. 2Results with oil diets with (+C) and without cholesterol (-C) were analyzed by two-way ANOVA to determine the effect of cholesterol (C) and the effect of fiber (F); a p value of < 0.05 was considered statistically significant for main treatment effects and interaction. When the ANOVA indicated significant differences, group means were compared using least significant difference at the p < 0.05 level. Means in a column with different superscript letters are significantly different; NS - not significant. 3Results with the two lard diets were compared using a Student's i test to determine the effect of adding fiber to a lard-containing diet; *p < 0.05.

KOH-ethanol at 70°Cfor 30 min, then 0.15 mol/L MgSO4was added to neutralize the mixture. After centrifugation (2000 x g, 5 min, at 20°C)the supernatant was assayed for glycerol (24). Cholesterol content was measured in the lipid residue with an enzymatic proce dure after the addition of Triton X-100 (25).Results are expressed as milligrams triglycérideor cholesterol per gram liver (wet wt). A polyvalent control serum (Biotrol 33-Plus, Biotrol, Paris, France) was treated in parallel to samples and served as a control of accuracy of results in lipid analysis. Statistics. Values are given as means ±SEM.Data were analyzed by Student's t test or by two-way analysis of variance (ANOVA). Where necessary to achieve ho mogeneity of variances, the data were subjected to log arithmic transformation. When the ANOVA indicated significant differences (p < 0.05), differences between individual treatment groups were determined by least significant difference at the p < 0.05 level of signifi cance.

RESULTS

Effect of diets on animals and their digestive tracts. As shown in Table 2, the body weight of the rats was not significantly influenced by the diets, but part of the difference in body weight could be masked by the pres ence of material in the digestive tract and the cecum weight. In the fiber-free diet groups fed the digestible wheat starch, the cecum was poorly developed. In con Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

trast, incorporation of fiber and potato starch in the diet resulted in a significant enlargement of cecum (cecum wall and cecum content weight) up to a total weight of 14 g. Compared to rats fed the fiber-free diets, rats fed diets rich in fermentable carbohydrates presented a relatively low pH in cecal content and an increased concentration of VFA. The cecal pool of acetate, propionate and butyrate was markedly increased. Lipid-lowering effects of diets rich in fermentable carbohydrates. The effects of the dietary treatments on plasma and liver lipids and on plasma lipoprotein levels are shown in Tables 3 and 4, respectively. The replace ment of wheat starch in the low fat diet by various fiber components and poorly digestible starch caused a dra matic drop in plasma and liver lipids. The cholesterol and triglycéridelevels in the plasma lipoprotein frac tions were markedly decreased. The addition of cholesterol and cholic acid to the low fat diet led to a substantial increase in plasma choles terol and in liver triglycéridesand cholesterol. The increase in plasma cholesterol was characterized by elevated VLDL and LDL and by a decrease in HDL cholesterol levels. In rats fed diets supplemented with cholesterol/cholic acid, fiber feeding decreased the lev els of plasma and liver lipids and of plasma cholesterol. Cholesterol and triglycéride levels in plasma lipoprotein fractions were decreased in the groups fed diets rich in fermentable carbohydrates, except for HDL levels, which increased with a fiber-rich cholesterol-contain ing diet compared to the same diet without fiber.

1041

FERMENTABLE CARBOHYDRATES AND LIPOPROTEINS TABLES Plasma and liver lipid concentrations of rats fed a fiber-free (FF)or a fermentable carbohydrate-rich (FC)diet1

Cholesterolmg/

Phospholipidsmg/lOOmLLiverTriglycérides Cholesterol

DietPlasmaTriglycérides

g wetwt 138.8 31.6 115.8 46.0

± 14.1" ± 3.2b ± 8.8» ± 4.2"

F, FxC

ANOVA

70.0 37.8 195.4 80.9

±2.8a ±1.5b ±6.1C ±5.3a

179.0 ± 16.1a 80.4 ± 3.& 157.0 ± 5.9" 107.2 ± 2.9e

F, C, F x C

17.1 11.7 44.1 24.3

±0.8a ±0.6b ±2.8C ±2.1d

F, C, F x C

F, FxC

3.4 2.3 17.7 12.9

±0.1a ±0.1a ±0.7" ±1.0e

F, C, F x C

dietsFFFC179.353.4±±8.24.3*67.2 Lard ±47.7 ±1.91.7*183.8

±121.9 ±11.35.7*15.2

±11.0 ±1.61.1*433.7±0.1±

0.1*

'Values are means ± SEMof 12 determinations. Statistical analyses and abbreviations are the same as described in Table 2.

Rats fed a diet containing lard, compared to those fed oil diets (without cholesterol/cholic acid), developed a marked hypertriglyceridemia, characterized by elevated chylomicron triglycérides,and a tendency for higher liver cholesterol. It must be noted that these modifica tions were not accompanied by changes in plasma cho lesterol level. Fiber feeding of animals receiving lard decreased the plasma and liver lipid concentrations. Triglycéride and cholesterol were reduced in all lipoprotein fractions examined. As a general rule, similar changes in plasma and liver lipids were observed with diets rich in fermentable

carbohydrates: Cholesterol and triglycéride concentra tions were markedly decreased in comparison to the concentrations in the respective fiber-free diet groups. Lipoprotein metabolism and HMG-CoA reducÃ-ase activity in low fat-fed rats. A diet rich in fermentable carbohydrates significantly affected the composition of HDL isolated by ultracentrifugation compared to the composition associated with a fiber-free diet. The pro portion of phospholipids, of apo E and of the HDLi subpopulation were decreased, and the proportion of protein constituent and of apo A-I were increased (Table5).

TABLE 4

Effect of fiber-free (FF)and fermentable carbohydrate-rich (FC)diets on triglycérideand cholesterol concentrations in various plasma lipoprotein fractions1 Triglycérides Diet

Chylomicrons

Cholesterol VLDL

VLDL

LDL

HDL

mg/100 ml plasma 25.3 6.1 30.0 7.3 ANOVA

±2.9a ±0.8b ±2.6a ±0.8b

65.3 9.0 50.0 18.0

±3.7a ±1.2b ±3.1e ±1.5*

F,FxC

4.4 1.5 95.3 22.0

±0.4a ±0.2a ±5.4b ±1.7°

F, C, F x C

4.6 3.4 43.9 18.5

±0.5a ±0.1a ±4.8b ±1.8e

F, C, F x C

49.2 27.7 22.7 29.0

±1.9a ±0.5b ±0.7e ±0.6b

F, C, F x C

dietsFFFC90.520.1± Lard 2.7± 1.4*54.514.3±±1.76.10.4*

3.1±0.1± 0.1*8.85.9±

0.7± 0.4*41.134.1±0.5± 1.0*

'Values are means ±SEMof six determinations. Statistical analyses and abbreviations are the same as described in Table 2. VLDL - very low density lipoprotein, LDL - low density lipoprotein and HDL - high density lipoprotein. Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

MAZURETAL.

1042

1.0 TABLES Effect on high density lipoprotein (HDL) composition of a low fat (5% oil, without cholesterol) diet rich in fermentable carbohydrates1

FF

Component

in

•8

Diet2 FC

I 2

0.5

wt%

Triglycérides Cholesterol Phospholipids Protein

0.9 26.5 31.8 40.8

1.0 25.8 26.9 46.3

±0.1 ±0.9 ±1.0 ±0.8

±0.1 ±0.9 ±0.8 *3 ±1.4*

% HDL protein* apo A-IV apo E apo A-I apoC's

14.9 19.8 56.2 9.1

±0.8 ±0.5 ±0.8 ±0.4

14.0 ±0.8 8.8 ±0.6* 69.1 ±1.0* 8.1 ±0.6

.§ 0 CO

< (O

o. Z

% of HDL area4 HDL!

7.2 ±0.7*

16.1 ±0.8

'Values are means ±SEMof six determinations. HDL refers to the fraction isolated at d - 1.050-1.21 g/mL. 2Diets were fiber-free (FF) or fermentable carbohydrate-rich (FC). 3An asterisk indicates a significant difference from the fiber-free diet group (p < 0.05) as assessed by Student's i test. *The gels after electrophoresis were stained with Coomassie Bril liant Blue then scanned in a densitometer as % of total dye uptake.

§ 0.1

3

and the data are expressed

0.05 L

Decay curves for human labeled LDL injected into rats are presented in Figure 1. The disappearance of 115I-labeledLDL from plasma in rats fed a diet rich in fermentable carbohydrates was faster than in control rats. The calculated fractional catabolic rates were 0.066 ±0.004 h'1 and 0.104 ±0.006 h'1 (p < 0.05) for the fiberfree and the fiber-rich group, respectively. The total and expressed HMG-CoA reducÃ-aseactivi ties were markedly increased in rats fed a diet rich in fermentable carbohydrates in comparison to those fed a fiber-free diet; the activities were 136± 2 vs. 69 ± 5 (total) and 33 ±2 vs. 17 ±2 (expressed) pmol mevalonic acid formed/(mgprotein-min), respectively (n =4, p < 0.05 for both forms).

DISCUSSION The hypolipidemic effects of foods rich in water-sol uble fibers have been frequently reported in humans and animals (1-7). In our study a maximal hypolipidemic action was sought by the utilization of a mixture of various dietary fibers because large amounts of gelforming fibers are poorly tolerated in rats (26). Thus, a Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

6 TIME

12 (hr)

FIGURE 1 Semilogarithmic decay curves for plasma 125I after injection of human 125I-labeledLDL into rats fed low fat diets that were fiber-free (•)or rich in fermentable carbohy drates (O). Each data point represents the mean value for seven animals; SEM did not exceed 5% for any data point. The fractional catabolic rates (see Materials and Methods) calcu lated from the curves were 0.066 ±0.004 rr1 for the group fed the fiber-free diet and 0.104 ±0.006 rr1 for the group fed the diet rich in fermentable carbohydrates (p < 0.05).

mixture of pectin, guar gum, soybean polysaccharide and wheat bran was used throughout. To enhance the cecal fermentation, crude potato starch [incompletely broken down in the small intestine (8, 10)] was also included in this high fiber diet. In keeping with earlier studies, such a diet induced a considerable enlargement of the cecum and of the cecal supply of VFA (26-28). The present work shows that feeding fermentable carbohydrates counteracted the abnormal lipoprotein profile induced by feeding lard or cholesterol/cholic

FERMENTABLE CARBOHYDRATES AND LIPOPROTEINS

acid. A similar action of dietary fiber was observed by others for lipoproteins separated by selective precipita tion; in particular, Chen and Anderson (29) made this observation with pectin, guar gum and oat bran, and Lairon et al. (30), with wheat germ. In a recent study, conducted on lipoproteins isolated by ultracentrifugation Ney et al. (31) demonstrated that oat fiber and processed oat product prevent the abnormal distribution of HDL particle size in rats fed cholesterol and cholic acid. Conversely, we observed a marked hypotriglyceridemic effect with dietary fiber. Anderson and Chen (3, 29) have observed that many studies have failed to demon strate a hypotriglyceridemic effect of dietary fiber, prob ably because of differences in experimental conditions. In our study the rats were fed during the dark period, and plasma was sampled after the light cycle began. Accord ing to potential mechanisms of fiber action in the intes tine (3-5, 7, 11), the hypotriglyceridemic effect may be related to an alteration in the rate of fat and glucose absorption, resulting in a reduced postprandial triglyc éride level in chylomicrons and VLDL.The lipid-lowering activity of this diet was also noticed in the liver, in agreement with previous findings demonstrating that various fiber components, especially soluble fibers, re duce serum cholesterol levels (1-7). However, the cho lesterol-lowering effects of dietary fiber frequently are not consistent in plasma and in the liver. In fact, the cholesterol-lowering effects of fiber diets depend on the type and the quantity of fiber as well as on the time of adaptation. Additionally, failure to modify serum and tissue cholesterol levels by dietary manipulations in rats may reflect the natural resistance of this species to alterations in blood cholesterol concentration (32).The present diet rich in fermentable carbohydrates appears very efficient for lowering cholesterol. Further, we examined the effects of a high ferment able carbohydrate supply on HDL composition and LDL metabolism. Because of marked alterations in the com position of plasma lipoproteins in rats fed cholesterol and high fat diets (33-35), our investigations were con ducted with rats receiving a low fat diet. In contrast to the case in humans, in rats, most of the plasma choles terol is associated with HDL, which is particularly responsive to diet. In fact, rat HDL is heterogeneous and has been separated into two or three major subclasses compositionally distinct with regard to the apo E con tent, lipid and protein composition and particle size (36-38). Apo E is a major component of large HDL, referred to as HDLi. This apolipoprotein could play a major role in the reverse transport of cholesterol as a recognition signal for the hepatic uptake and removal of cholesterol-enriched HDL particles (39). The results show that a diet rich in fermentable carbohydrates markedly reduced the proportion of the HDLi subpopu lation in HDL isolated by ultracentrifugation, which was consistent with the decrease in apo E content and Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

1043

the modifications in the chemical composition. A de crease in HDLi could be explained by its reduced forma tion and/or increased removal. Reduced formation could result from a decreased supply of the surface components of triglyceride-rich lipoproteins for HDLi formation (40). Low levels of triglyceride-rich lipopro teins were actually observed, which supports the above hypothesis. Modifications in HDL composition in rats fed various dietary fibers rich in water-soluble or insol uble components were recently reported by Schneeman et al. (41).The observed effects appear to depend on the source of the fiber. For example, HDL fractions isolated from rats fed wheat bran and guar gum differ signifi cantly in their proportion of apo A-I, apo E and apo C's in HDL. As in our study, using the diet rich in ferment able carbohydrates, feeding a soluble fiber tends to de crease the apo E level in HDL. In vivo, the liver is the major site for the removal of plasma lipoprotein cholesterol. At least two processes of cell receptor-dependent uptake can be distinguished: The LDL receptor (apo B, E receptor) binds lipoproteins possessing either apo Bor apo E (LDLand HDL contain ing apo E),and the apo Ereceptor binds apo E-containing particles (chylomicron remnants and HDL with apo E) (42).Although the distinction between the LDLreceptor and the chylomicron remnant receptor in rat liver mem branes has been questioned (43),it seems that these two receptors are regulated independently (42, 43). In con trast to the apo B, E receptor (readily responsive to metabolic conditions), apo E receptor expression is rel atively constitutive (42). Several studies indicate that the apo E-containing HDL may participate in choles terol homeostasis by insuring the return of excess extrahepatic cholesterol to the liver (39, 42). To assess the importance of the LDL(apo B,E)receptor in the decrease in plasma cholesterol, the decay of labeled human LDL has been investigated. This assay provides a cumulative determination of LDL uptake by the tissue, chiefly the liver. It may be debated whether human LDL reflect exactly the fate of rat LDL (44),-nevertheless, human LDL is catabolized by the rat liver through a receptormediating pathway, at a rate similar to that observed for rat LDL (45, 46). Lipoproteins constitute an important source of he patic and biliary cholesterol. Another source of hepatic cholesterol is cholesterogenesis. It has been demon strated that 10-20% of bile cholesterol is derived from de novo synthesis. The liver can adapt its rate of choles terol biosynthesis to accommodate cholesterol needs. Cholesterogenesis involves HMG-CoA reductase as the rate-limiting enzyme (47). In our study, HMG-CoA re ductase total and expressed activities were increased in rats fed a diet rich in fermentable carbohydrates, but their ratio was comparable (about 25% of expressed form) in both groups. Whether such a low percentage of expressed activity is physiologically relevant has been disputed (47); nevertheless, the unmodified expressed/

1044

MAZURETAL.

total form ratio suggests a lack of effect of fermentable carbohydrates on the short-term regulation of HMGCoA redactase. This is in agreement with the findings of Brown et al. (48), demonstrating that the changes in HMG-CoA reducÃ-aseactivity in rats subjected to longterm manipulations are not attributable to changes in the degree of phosphorylation. Yet, it has been reported that large amounts of propionate could decrease the activity of HMG-CoA reducÃ-ase(49-51). It thus seems that the effect of the reduction in cholesterol availabil ity is predominant and results in the activation of cholesterogenesis in spite of a possible regulatory effect of VFA or some of their metabolites. In this view, a stimulation of hepatic cholesterogenesis following an increase in biliary cholesterol output has recently been described in rats fed a high bean diet (52). In conclusion, the present results demonstrate the lipid-lowering effect of a diet rich in fermentable carbo hydrates; this effect is expressed both in normolipemic rats and in conditions of hvpercholesterolemia and hypertriglyceridemia. Detailed studies conducted in rats fed a low fat diet show that this hypocholesterolemic action of fermentable carbohydrates is associated with a decrease in liver cholesterol concentration and the accelerated clearance of plasma lipoproteins. Paradoxi cally, the decrease in plasma and liver cholesterol elic ited an increase in HMG-CoA reducÃ-aseactivity. This suggesls lhal ihis diel affecls various mechanisms of cholesierol homeoslasis, perhaps because of leakage of choleslerol and insufficienl compensalory choleslerogenesis. Because VFAabsorplion is very greal in rals fed a diel rich in fermeniable carbohydrates, further re search will be required on the relalionship belween VFA metabolism and lipogenesis in ihe various tissues.

LITERATURE CITED 1. KRITCHEVSKY, D. (1978) Fiber, lipids, and atherosclerosis. Am. f. Clin. ÑutÃ-.31:S65-S74. 2. SPILLER,G. A., SHIPLEY,E. A. & BLAKE,J. A. (1978| Recent prog ress in dietary fiber (plantix) in human nutrition. CRC Crii. Rev. Food Sci. Nutr. 10:31-90. 3. ANDERSON,J. W. & CHEN, W. J. L. (1979) Plant fiber. Carbohy drate and lipid metabolism. Am. ]. Clin. Nutr. 32: 346-363. 4. VAHOUNY,G. V. (1982) Dietary fiber, lipid metabolism, and atherosclerosis. Fed. PTOC.41: 2801-2806. 5. STORY,J. A. (1985) Dietary fiber and lipid metabolism. Proc. Soc. Exp. Biol. Med. 180: 447-452. 6. SHUTLER,S. M., WALKER,A. F. & Low, A. G. (1987) The choles terol-lowering effects of legumes H: Effects of fibre, sterols, saponins and isoflavones. Hum. Nutr. Food Sci. Nutr. 41F: 87-102. 7. ULLRICH,I. H. (1987) Evaluation of high-fiber diet in hyperlipidemia: a review. /. Am. Coll. Nutr. 6: 19-25. 8. DEMIGNE,C. & REMESY,C. (1982) Influence of unrefined potato starch on cecal fermentations and volatile fatty acid absorption in rats. /. Nutr. 112: 2227-2234. 9. SAQUET,E., LEPRINCE,C. & RIOTTOT,M. (1983) Effect of amylomaize starch on cholesterol and bile acid metabolisms in germfree (axenic) and conventional (holoxenic) rats. Reprod. Nutr Dév. 23: 783-792. Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

10. ASP, N.-G., BJÖRCK, I., HOLM, J., NYMAN,M. & SILJESTRÖM, M. (1987) Enzyme resistant starch fractions and dietary fibre. Scand. J. Gastroenterol. 22: 29-32. 11. VAHOUNY,G. V. (1982) Dietary fibers and intestinal absorption of lipids. In: Dietary Fiber in Health and Disease (Vahouny, G. V. & Kritchevsky, D., eds.), pp. 203-227, Plenum Publishing, New York, NY. 12. REMESY,C. & DEMIGNE,C. (1976) Partition and absorption of volatile fatty acids in plasma after ethanolic extraction. Ann. Rech. Vet. 7: 39-55. 13. HAVEL,R. J., EDER,A. H. & BRAGDON,J. M. (1955) The distribu tion and chemical composition of ultracentrifugally separated lipoproteins in human serum. /. Clin. Invest. 34: 1345-1363. 14. BLANCHE, P. J., GONG,E. L., FORTE,T. M. & NICHOLS,A V. (1981 ) Characterization of human high density lipoproteins by gradient gel electrophoresis. Biochim. Biophys. Acta 665: 408-419. 15. LOWE,K. E., PELKEY,S., WILLIAMS, M. A. & NICHOLS,A. V. (1988) Effect of essential fatty acid deficiency on the size and distribution of rat plasma HDL. Lipids 23: 106-114. 16. WEBER,K. & OSBORN,M. (1969) The reliability of molecular weight determinations by dodecylsulfate-polyacrylamide gel electrophoresis. /. Biol. Chem. 244: 4406-4412. 17. SHEPHERD,J., BEDFORD,D. K. &. MORGAN,H. G. (1976) Radioiodination of human low density lipoprotein: a comparison of four methods. Clin. Chim. Acta 66: 97-109. 18. SNIDERMAN, A. D., CAREW,T. E., CHANDLER, J. G. & STEINBERG, D. E. (1974) Paradoxical increase in rate of catabolism of low-den sity lipoproteins after hepatectomy. Science 183: 526-528. 19. BEG, Z. H., STONK, J. A. & BREWER,H. B., JR. (1984) In vivo modulation of rat liver 3-hydroxy-3-methylglutaryl-coenzyme A reducÃ-ase,reducÃ-asekinase, and reducÃ-asekinase kinase by mevalonolacione. Proc. Nati. Acad. Sci. USA 81: 7293-7297. 20. BRANDT,H., CAPULONG,Z. L. & LEE,E. Y. C. (1975) Purification and properties of rabbit liver phosphorylase phosphatase. /. Biol. Chem. 250: 8038-8044. 21. ACKERMAN, M. E., REDD,W. L., TORMANEN,C. D., HARDGRAVE, J. E. & SCALLEN, T. J. (1977) The quantitalive assay of 3-hydroxy3-melhylgluiaryl coenzyme A reducÃ-ase: comparison of ihinlayer chromalographic assay wilh a rapid chloroform extraction assay. /. Lipid Res. 18: 408-413. 22. MARKWELL, M. A. K., HASS,S. M., BÅ’BER, L. L. & TOLBERT,N. E. (1978) A modification of the Lowry procedure to simplify protein determination in membrane and lipoprolein samples. Anal. Biochem. 87: 206-210. 23. FOLCH,J., LEES,M. & SLOANE-STANLEY, G. H. (1957) A simple method for the isolation and purification of total lipids from animal tissues. /. Biol. Chem. 226: 497-506. 24. BUGOLO,G. & DAVID,H. (1973) Quantitative determinai ion of serum üiglycerides by the use of enzymes. Clin. Chem. 19: 476-482. 25. DEHOFF,J. L., DAVIDSON,L. M. & KRITCHEVSKY, D. (1978) An enzymatic assay for determining free and loial choleslerol in tissue. Clin. Chem. 24: 433-435. 26. DEMIGNE,C. & REMESY,C. (1985) Stimulation of absorption of volatile fatly acids and minerals in ihe cecum of rals adapted to a very high fiber diet. /. Nutr. 115: 53-60. 27. DEMIGNE,C., YACOUB,C. & REMESY, C. (1986) Effects of absorp lion of large amounts of volatile fatly acids on ral liver meiabolism. /. Nutr. 116: 77-86. 28. TULUNG,B., REMESY, C. & DEMIGNE,C. (1987) Specific effects of guar gum or gum arabic on adaptation of cecal digeslion lo high fiber dieis in ihe rat. /. Nutr. 117: 1556-1561. 29. CHEN,W. J. L. & ANDERSON,f. W. (1979) Effects of plani fiber in decreasing plasma total cholesterol and increasing high-density lipoprolein choleslerol. Proc. Soc. Exp. Biol. Med. 162: 310-313. 30. LAIRON,D., LACOMBE, C., BOREL,P., CORRAZE,G., NIBBELINK, M., CHAUTAN,M., CHANUSSOT,F. & LAFONT,H. (1987) Beneficial effecis of wheal germ on circulating lipoproieins and lissue lipids in rals fed a high fai, choleslerol-coniaining diel. /. Nutr. 117:

FERMENTABLE CARBOHYDRATES AND LIPOPROTEINS 838-845. 31. NEY,D., LASEKAN, J. B. & SHINNICK,E L. (1988) Soluble oat fiber tends to normalize lipoprotein composition in cholesterol-fed rats./. Nutz. 118: 1455-1462. 32. STORY,J. A., TEPPER,S. A. & KRTTCHEVSKY, D. (1974) Influence of synthetic conjugates of cholic acid on cholesterolemia in rats. /. Nutr. 104: 1185-1188. 33. LASSER,N. L., ROHEM, P. S., EDELSTEIN, D. & EDER,H. A. (1973) Serum lipoproteins of normal and cholesterol-fed rats. /. Lipid Res. 14: 1-8. 34. MAHLEY,R. W. & HOLCOMBE,K. S. (1977) Alterations of the plasma lipoproteins and apoproteins following cholesterol feed ing in the rat. /. Lipid Res. 18: 314-^324. 35. DELAMATRE, J. G. & ROHEIM,P. S. (1981) Effect of cholesterol feeding on apo B and apo E concentrations and distributions in euthyroid and hypothyroid rats. ]. Lipid Res. 22: 297-306. 36. LUSK,L. T., WALKER,L. F., DUBÅ’N,L. H. & GETZ, G. S. (1979) Isolation and characterization of high-density lipoprotein HDLi from rat plasma by gradient centrifugation. Biochem. J. 183: 83-90. 37. OSCHRY,Y. & EISENBERG, S. (1982) Rat plasma lipoproteins: reevaluation of a lipoprotein system in an animal devoid of cholesteryl ester transfer activity. /. Lipid Res. 23: 1099-1106. 38. LEE,C. C. & Koo, S. I. (1988) Separation of three compositionally distinct subclasses of rat high density lipoproteins by heparin-affinity chromatography. Atherosclerosis 70: 205-215. 39. MAHLEY,R. W. (1988) Apolipoprotein E: cholesterol transport protein with expanding role in cell biology. Science 240:622-630. 40. EISENBERG, S. (1980) Plasma lipoprotein conversions: the origin of low-density and high-density lipoproteins. Ann. N. Y.Acad. Sci. 348: 30-47. 41.

SCHNEEMAN,

B. O.,

ClMMARUSTI,

J., COHEN,

W., DOWNES,

L. &

LEFEVRE, M. (1984) Composition of high density lipoproteins in rats fed various dietary fibers, f. Nutr. 114: 1320-1326. 42. WEISGRABER, K. H., INNERARITY, T. L., RALL,S. C., [R, &.MAHLEY, R. W. (1985) Receptor interactions controlling lipoprotein me tabolism. Can. J. Biochem. Cell Biol. 63: 898-905. 43. WADE,D. P., KNIGHT,B. L. & SOUTAR,A. K. (1986) Binding of low-density lipoprotein and chylomicron remnants to hepatic low-density lipoprotein receptor of dogs, rats and rabbits demon

Downloaded from https://academic.oup.com/jn/article-abstract/120/9/1037/4738614 by guest on 25 January 2018

1045

strated by ligand blotting. Failure to detect a distinct chylomicron-remnant-binding protein by ligand blotting. Eur. f. Biochem. 159: 333-340. 44. KOELZ,H. R., SHERRILL, B. C., TURLEY,S. D. & DIETSCHY,J. M. (1982) Correlation of low and high density lipoprotein binding in vivo with rates of lipoprotein degradation in the rat. A compar ison of lipoproteins of rat and human origin. /. Biol. Chem. 257: 8061-8072. 45. CAREW,T. E., PITTMAN,R. C. & STEINBERG, D. (1982) Tissue sites of degradation of native and reductively methylated [14C]sucroselabeled low density lipoprotein in rats. Contribution of receptordependent and receptor-independent pathways. /. Biol. Chem. 257: 8001-8008. 46. RUDLING,M. (1987) Role of the liver for the receptor-mediated catabolism of low-density lipoprotein in the 17ot-ethinylesiradioltreatedrat. Biochim. Biophys. Acta 919: 175-180. 47. BEG,Z. H., STONIK,J. A. & BREWER, H. B. (1987) Modulation of the enzymic activity of 3-hydroxy-3-methylglutaryl-coenzyme A reducÃ-aseby multiple kinase systems involving reversible phosphorylation: a review. Metabolism 36: 900-917. 48. BROWN,M. S., GOLDSTEIN, J. L. & DÅ’TSCHY, J. M. (1979) Active and inactive forms of 3-hydroxy-3-methylglutaryl-coenzyme A reducÃ-ase in the liver of the rat. Comparison with the rate of cholesterol synthesis in different physiological states. /. Biol. Chem. 254: 5144-5149. 49. IDE,T, OKAMATSU, H. & SUGANO,M. (1978) Regulation by di etary fats 3-hydroxy-3-methylglutaryl-coenzyme A reducÃ-ase in rat liver. /. Nutr. 108: 601-612. 50. ANONYMOUS(1987) Propionate and cholesterol homeostasis in animals. Nutr. Rev. 45: 188-190. 51. ILLMAN,R. J., TOPPING,D. L., MclNTOSH, G. H., TRIMBLE,R. P., STORER,G. B., TAYLOR,M. N. & CHENG, B-Q. (1988) Hypocholesierolaemic effects of dielary propionate: studies in whole animals and perfused ral liver. Ann. Nutr. Metab. 32: 97-107. 52. RICOTTI,A., MARZOLO,M. P., ULLOA,N., GONZALEZ,O. & NERVI, E (1989) Effeci of bean intake on biliary lipid secretion and on hepatic choleslerol melabolism in ihe ral. /. Lipid Res. 30: 10411048.