Dietary Oligofructose Lowers Triglycerides, Phospholipids and Cholesterol in Serum

and Very Low Density Lipoproteins of Rats Maria Fiordaliso a, Nadine Kok a Jean-Pierre Desager b, Fabienne Goethals a,

Dominique Deboyser a, Marcel Roberfroid a and Natha|ie Delzenne a'* aunit~ de Biochimie Toxicologique et Canc6rologique, D6partement des Sciences Pharmaceutiques

and bLaboratoire de Pharmacoth6rapie, Facult~ de M6decine,

Universit6 Catholique de Louvain, B-1200 Brussels, Belgium

ABSTRACT: The present study was aimed at answering the question why feeding rats an oligofructose (OFS) supplemented diet could cause a significant reduction in plasma lipid levels. Daily administration of a 10% (w/w) OFS-containing diet to nor- molipidemic male rats resulted in adecrease in plasma triglyc- erides, phospholipids and cholesterol. The triglyceride-lower- ing effect was observed after one week and lasted for at least 16 wk and was associated with a reduction in plasma very low density lipoproteins, indicating that the hypolipidemic effect of OFS may be due tochanges in liver lipid metabolism. We there- fore tested whether OFS feeding modified the capacity of the liver to synthesize triglycerides from free fatty acids. Hepato- cytes isolated from livers of control and OFS-fed rats were incu- bated in the presence of [1-14C]palmitate, and both intracellu- lar and extracellular [t4C]triglyceride formation were quanti- fied. We found that chronic feeding of an OFS-supplemented diet to rats significantly reduced the capacity of isolated hepa- tocytes to synthesize triglycerides from palmitate. The results suggest that, like other soluble dietary fibers, OF--5 significantly alters liver lipid metabolism, resulting over time in a s~gnificant reduction in plasma triglyceride, phospholipid and cholesterol levels. Lipids 30, 163-167 (1995).

Oligofructose (OFS) is an o l igomerof B-D-fructose contain- ing two to nine monomeric units which are linked by g (2-1) glycosidic bridges. OFS is resistant to hydrolysis by pancre- atic amylase and saccharidases (1,2). OFS thus belongs to the class of nondigestible oligosaccharides (NDO) which share many properties with fermentable dietary fibers (3). Being nondigestible, OFS reaches the colon intact where it is fer-

mented mostly by Bifidobacterium (4). This process produces energy for bacterial proliferation, short-chain fatty acids (SCFA) such as acetate, propionate and butyrate, as well as lactate, which all feed the host. Mostly gaseous by-products (CO2, CH 4 and I-I2) are excreted (5).

In an earlier study, we had shown that feeding rats a diet containing 20% OFS for 30 d led to a significant decrease in the concentration o f both serum and liwer trigtyceride levels (6). This suggested that the capacity of OFS-fed rats to ester- ify and secrete fatty acids as very low density lipoproteins (VLDL) may be reduced when compared to isocalorically fed control rats.

Similar effects of nondigestible, but fermentable, carbohy- drates (including the synthetic fructooligosaccharides that are chemically similar to OFS) on serum triglycerides have been reported (7). It has been hypothesized that some SCFA, in particularpropionateand possibly acetate, could be the medi- ators of this systemic modulation of the metabolism of ,triglycerides and of other lipids due to dietary fiber. However, the biochemical mechani~sm of this modulation remained largely .undefined and may also vary for different food ingre- dients. Moreover, it may not be limited to the synthesis of triglycerides or the secretion of VLDL.

We have therefore followed the time course of the hy- polipidemic effect o f O F S under conditions that would ex- clude any changes in body weights. Furthermore, the lipid compositions of serum lipoproteins during OFS feeding were compared to those <ffcontrol rats. Finally, an ex vivo protocol was developed to study the influence of OFS feeding on the capacity of hepatocytes isolatedfrom these rats to incorpo- rate [1-~4C]palmitate into triglycerides.

*To whoqn correspondence shouldbe addressed at Unit6 BCTC, UCL 7369. Avenue Mounier. 73 B-1200 Brussels, Belgium.

Abbreviations: ANOVA. analysis of variance; BSA. bovine serum albumin: HDL, high-density lipoprotein: |DL. intermediate density l~poprotein; LDL, low-density lipoprotein: NDO. rtondigestible oligosaccharide: OFS. oligofructose: SCFA. short-chain fatty acids: SEM. staudard error of the mean; VLDL-very low density lipoprotein.

MATERIALS AND METHODS

Chemicals. OFS (purity >_98%) was supplied as Raftilose| by Raffinerie Tidemontoise/nc. (Tienen, Bel- gium). It contains at least 95% (w/w) fructooligosaccharides with a mean degree of polymerization of 4.8. Collagenase and Dulbecco's medium were purchased from Boehringer-

Copyright �9 1995 by AOCS Press 163 LIPIDS, Vol. 30, nol 2 (1995)

164 M. FIORDALISO ETAL.

Mannheim (Mannheim, Germany) and Flow Laboratories (Irvine, Scotland), respectively. Bovine serum albumin (BSA) (fraction V) was from the Sigma Chemical Company (St Louis, MO); [1-14C]palmitic acid (710 laCi/mmol) was from New England Nuclear Research Products (Boston, MA); Carbo Sorb and Permafluor scintillation mixtures were from Beckman (Los Angeles, CA). All other chemicals were of the purest grade available.

Animals and diets. Male Wistar rats ICO-WY IOPS from Iffa Credo (Les Oncins, France) initially weighing about 100 g were housed individually in grill-bottom metal wire cages in a room maintained under a 12 h light/dark cycle at 21 _+ 3~ The animals were acclimatized for one week prior to the experiment. All animals had free access to food and water.

The powdered basal diet A04 was purchased from UAR (Villemoisson-Sur-Orge, France). It contained 58% (w/w) carbohydrates, including starch 33% (w/w) and fiber (3% w/w), 17% (w/w) proteins, 12% (w/w) water, 4% (w/w) cel- lulose, 3% (w/w) lipids and 5% (w/w) minerals. Vitamin A and D3 levels were 7500 and 1500 UI/kg, respectively.

The diets used for the experiments were: (i) the control diet, corresponding to the A04 diet (caloric value, 2.9 Kcal/g); (ii) the sucrose-supplemented diet (5% wt/wt) prepared by adding 5 g of dry saccharose to 95 g of AO4 diet (caloric value, 2.9 Kcal/g); (iii) the OFS-enriched diet (10% w/w) pre- pared by adding 10 g of dry OFS to 90 g of AO4 diet (caloric value, 2.8 Kcal/g, assuming that the OFS caloric value is 2 Kcal/g; Ref. 8).

Animal experiments were performed consistent with the rules for humane care according to Belgian State Directives.

Experimental protocols. Experiment 1 compared food in- take and body weight of rats fed control and OFS-containing (10% w/w) diets. However, as the OFS supplement not only diluted the diet but also slightly modified its caloric content, a diet was prepared which was supplemented with 5% w/w sucrose. This assumed that OFS had <50% the caloric con- tent of sucrose (8). Following the acclimatization period, 15 rats were weight matched and assigned to three groups which received one of the three diets for up to 7 wk. Body weight, 24-h food intake and fecal excretion were measured twice weekly. The feces collected on week 6 were frozen, lyophilized and then used for lipid analysis. Experiment 2 used two groups of ten rats and was designed to monitor the time course (up to 12 wk) of the effect of OFS feeding on serum triglycerides, which was measured once every week and then at the end of the treatment (week 16). Experiment 3 was the ex vivo protocol in which two groups of ten rats fed either a control or an OFS-containing (10% w/w) diet for 5- 6 wk were used to prepare hepatocytes according to the fol- lowing procedure: after feeding, rats were anesthetized at 10:00 a.m.; the liver was excised; and hepatocytes were iso- lated by the collagenase perfusion technique of Berry and Friend (9) as adapted by Krack et al. (10). Hepatocytes (_>90% viability as measured by Trypan blue exclusion) were then incubated at a final concentration of 0.5 106 cells/mL in Dulbecco's medium containing 5 x 10-SM [1-14C]palmitate

(0.7 mCi/mmol) bound to BSA and prepared according to Ca- puzzi et al. (11). The reaction was carded out for 150 min in an incubation vessel with a rotating helix (40 rpm) as previ- ously described (12). At 30-min intervals, samples of the sus- pension were removed for biochemical and radiochemical analysis. Each assay was carded out in duplicate, and at least eight separate experiments were performed.

Analytical procedures. For the enzymatic measurement of serum triglycerides, post-prandial blood samples were col- lected through the caudal vein. For the separation of lipopro- teins, blood was collected from the vena cava of rats that had been fasted for 18 h. Serum was then ultracentrifuged on a density gradient (rotor SW41, 40 000 rpm, 18 h) according to Chapman et al. (13). Triglycerides, phospholipids and total and esterified cholesterol were assayed in total serum and in each lipoprotein fraction using Boehringer kits. The feces col- lected on week 6 in protocol 2 were frozen, then dried and powdered. Total lipids were extracted with chloro- form/methanol 1:1. The organic phase was evaporated under nitrogen, and the residue was solved in ethanol. Total (free and esterified) fatty acids were measured using a kit from Boehringer-Mannheim. The synthesis of triglycerides and se- cretion of lipoproteins by isolated hepatocytes were assayed radiochemically. Both the incorporation of [ 1-14C]palmitate into cellular triglycerides and the increase in [14C]lipids in the culture medium were measured as described by Deboyser et al. (12). The activity of palmitoyl coenzyme A synthetase in isolated hepatocytes was measured by the spectrophoto- metric method of Ichara and Shibasaki (14). All results of the ex vivo experiment were expressed per mg of cellular protein which was determined by the method of Lowry et al. (15).

Statistical analysis. Data are expressed as mean _+ standard error (SEM). To compare the time course of the effect of OFS feeding on body weight and food intake (Experiment 1) or the in vitro incubation of hepatocytes on [ 1-14C]palmitate incor- poration (Experiment 3), analysis of variance (ANOVA) using a two-way ANOVA program was performed, followed by a Sheffe test to analyze time/treatment interaction on food intake. Statview TM 512 + software was used. To compare the lipid composition of lipoproteins and the concentration of serum triglycerides in Experiment 2, Student's t-tests were performed. In all statistical analyses, P < 0.05 was used as the significance level to analyze OFS effects.

RESULTS

Body weight growth and food intake (Experiment 1). The rats fed the diet supplemented with OFS gained weight each day at a similar rate (variance analysis; P > 0.05) as did the con- trols and sucrose-supplement fed rats (Fig. 1A). Food con- sumption was not significantly different (variance analysis; P > 0.05) between experimental groups (Fig. 1B), despite a tendency for OFS-fed rats to eat more than did the rats of the other groups. All animals fed OFS over time appeared healthy and behaved normally.

LIPIDS, Vol. 30, no. 2 (I 995)

OLIGOFRUCTOSE LOWERS SERUM AND VLDL LIPIDS 165

r162

400 -

350

300

250

200

150

A

100 -7 - - , . . . . , . . . . , . . . . , . . . . , . . . . , . . . . , . . . . , . . . . ,

0 5 10 15 20 25 30 35 40 45

Time (d)

=.

30.0'

27.5 �9

25 .0 '

22 .5 '

17.5

. . . . ; . . . . ;0 . . . . , ' i ' " 2 ' 0 " " "~'i '" ~'0"" ~ ' s ~ ' 0 " ",'~

Time (d)

FIG. 1. Body weight (A) and food intake (B) of rats fed a control diet (i-I) or a diet supplemented with either 5% w/w sucrose (O) or 10% w/w oligofructose (OFS) (0). The values represent means • SEM of five rats in each group. Variance analysis based on a two-way analysis of vari- ance was used to test for time/treatment interaction. No statistically sig- nificant effect (P > 0.05) of OFS feeding was apparent.

Serum lipids and lipoproteins (Experiment 2). Chronic ad- ministration of OFS to the rats caused a significant (P < 0.05) reduction in serum triglyceride levels (Fig. 2). The triglyc- eride-lowering effect was significant (P < 0.05) after one week of feeding and lasted for up to 16 wk (P < 0.05). Such a decrease in serum triglycerides was unlikely to be due to changes in fecal excretion of lipids. Indeed, fecal analysis at

3.0

2 .5

2 .0

~

1.5

1.0

tt~

0 .5 �84

0 .0

D u r a t i o n o f OFS f e e d i n g (dl

FIG. 2. Time course of serum triglycerides (retool/L) of rats fed a stan- dard (I-q) or an oligofructose (OFS)-supplemented (0) diet. The values represent means + SEM of five animals in each experimental group. Variance analysis based on a two-way analysis of variance was used to test for time/treatment interaction. It showed a statistically significant ef- fect (P< 0.05) of OFS feeding.

week 6 showed no statistically significant differences in 24-h excretion of esterified fatty acids between standard and OFS- supplement fed rats (15.8 _+ 2.4 Iamol vs. 14.9 _+ 1.1 Iamol es- terified fatty acids for control and treated rats, respectively). Table 1 shows that OFS-feeding for 16 wk caused a signifi- cant (P < 0.05) reduction in serum triglycerides (25% de- crease), phospholipids (15% decrease) and total cholesterol (15% decrease). This was exclusively due to the reduction of lipids in the VLDL fraction with the composition of both LDL and HDL remaining unchanged. Moreover, the effect of OFS on VLDL was likely to be due to a reduction in the num- ber rather than a change in the composition of VLDL parti- cles. Indeed, neither the composition nor the size estimated [according to Fungwee et al. (16)], by calculating the ratio of the sum of phospholipids plus free cholesterol over the serum triglycerides and esterified cholesterol, showed a statistically significant difference (Table 2).

Effect of OFS on the synthesis and secretion of triglyc- erides in isolated hepatocytes (Experiment 2). The incorpora- tion of [ 1-14C]palmitate into cellular triglycerides, as a func- tion of incubation time, in hepatocytes isolated from OFS-fed rats was lower than in control cells (P < 0.05) (Fig. 3A). After 150 min of incubation, the incorporation of [ 1-14C]palmitate into cellular triglycerides was reduced by 40% (P < 0.05). Apart from this reduction in the cellular incorporation of [l- 14C]palmitate, the accumulation over time of [14C]triglyc-

LIPIDS, Vol. 30, no. 2 (1995)

166 M. FIORDALISO ETAL.

TABLE 1 Effects of Feeding Rats an Oligofructose-Containing Diet (10%) on Various Lipid Parameters of Serum a

Concentrat ions (mM) + SEM (n) Triglycerides Phospholipids Total cholesterol Esterified cholesterol Free cholesterol

Serum control 1.14 + 0.10 (10) 1.38 _+ 0.03 (10) 1.70 • 0.08 (10) 1.28 + 0.10 (7) 0.46 • 0.04 (7) OFS 0.87 _+ 0.08 (10) b 1.1 7 + 0.04 (010) b 1.45 • 0.05 (10) b 1.04 • 0.04 (7) b 0.45 • 0.03 (7)

VLDL Control 0.79 _+ 0.01 (10) 0.16 • 0.02 (9) 0.15 + 0.02 (10) 0.06 + 0.006 (10) 0.10 • 0.01 (10) OFS 0.48 • 0.07 (8) b 0.10 + 0.01 (9) b 0.11 + 0.01 (9) b 0.05 -+ 0.052 (9) 0.07 + 0.01 (10)

LDL Control 0.06 • 0.01 (10) 0.28 • 0.02 (9) 0.54 + 0.04 (10) 0.35 + 0.03 (7) 0.13 + 0.02 (7) OFS 0.09 • 0.01 (9) b 0.26 + 0.02 (9) 0.51 _+ 0.03 (9) 0.37 _+ 0.02 (9) 0.14 + 0.01 (9)

HDL Control 0.08 + 0.005 (10) 0.52 _+ 0.03 (10) 0.84 _+ 0.07 (10) 0.73 _+ 0.07 (10) 0.12 • 0.01 (10) OFS 0.08 • 0.005 (8) 0.45 + 0.03 (8) 0.81 + 0.07 (7) 0.69 -+ 0.06 (7) 0.12 • 0.01 (9)

aValues represent means _+ SEM; numbers in parentheses are the number of samples analyzed, each sample being from one rat. Abbreviations: OFS, oligofruc- tose; VLDL, very low density lipoprotein; LDL, low-density lipoprotein; HDL, high-density lipoprotein; n is the number of samples. bp < 0.05 for difference between control and treated rats.

TABLE 2 Composition and Size of Serum VLDL Particles in Rats Fed Standard or OFS-Supplemented Diets a~

Experimental group VLDL Controls (n = 9) OFS (n = 8)

(composit ion%) Triglycerides 68 + 1 65 • 1 Phospholipids 12 • 0.1 t3 + 0.4 Total cholesterol 6 • 0.3 7 • 0.4 Total free cholesterol 3.7 • 0.4 3.3 • 0.2 Protein 13 • 0.4 15 + 1 Size (calculated) 0.30 • 0.03 0.30 + 0.04

aValues represent means • SEM; n is the number of samples analyzed. Stu- dent's t-test was used to compare the experimental data, which showed no statistical difference. See Table I for abbreviations.

erides in the incubation medium was not statistically different (P > 0.05) under the two experimental conditions (Fig. 3B).

Palmitoyl-CoA synthetase activity. The rate of activation to acyl-CoA is limiting in the incorporation of fatty acids into cellular triglycerides (17). As adaptive changes in the activity of the enzyme involved in this activation may reflect alter- ations in the metabolic flux, we examined whether palmitoyl- CoA synthetase activity was altered after OFS feeding. How- ever, the activity was not statistically different (P > 0.05) in homogenates of hepatocytes isolated from controls and OFS- fed rats (2.5 • 0.2 and 2.5 • 0.1 mmol/min/mg protein), re- spectively.

D I S C U S S I O N

Our data show that feeding rats a diet supplemented with 10% OFS had no significant effect on body weight gain, Besides a slight but not significant increase in macro- and micronutri- ent intake, the chronic feeding of OFS significantly reduced serum triglycerides when measured in both fed and fasted rats. Moreover, in the fasted state, serum phospholipids and total and esterified cholesterol were also significantly re- duced. These data confirm and extend our previous observa- tions (6) and the findings o f others (7) that the OFS-like syn- thetic fructo-oligosaccharides called Neosugar | similarly re-

duce triglyceridemia by 30% after feeding to rats for a 6-wk period of a supplemented (10 to 20% w/w) diet. Our data also agree with previous observations that fermentable dietary fibers, such as pectins, are effective nondigestible carbohy- drates that reduce serum and liver triglycerides (18,19).

Using an ex vivo protocol, we have also shown that OFS feeding results in a reduced capacity of isolated hepatocytes to incorporate free fatty acids into cellular triglycerides. Within the period of observation, no effect on secretion of newly synthesized triglycerides occurred.

The observation that reduced serum triglyceride, phospho- lipid and cholesterol levels were mainly due to a decrease in the number of VLDL particles, together with the results of our ex vivo experiments, supports the hypothesis that OFS feeding, like that of other fermentable dietary fibers (20), sig- nificantly alters the metabolism of lipids in the liver with, as a likely consequence, a reduction in VLDL production. Such an effect has also been demonstrated for lovastatin treatment (21). However, an increase in the catabolism of VLDL by lipoprotein lipase, as observed with fibrates, cannot, as yet, be discounted.

The observation that hepatocytes isolated from OFS-fed rats had a lower capacity than those from control rats to in- corporate [1-14C]palmitate into cellular triglycerides still needs to be explained. Since the activity of palmitoyl-CoA synthetase was not changed, we postulate that the availability of acyl-CoA is modified in the hepatocytes of OFS-fed rats. Two possibilities could be explored: either the pattern of es- terification of acyl-CoA could be altered or the catabolism of acyl-CoA to ketone bodies could be stimulated. In the latter case, regulation of the key enzyme carnitine acyltransferase would likely be involved. Whatever metabolic change is re- sponsible for the hypolipidemic effect, however, one must still explain how a nondigestible carbohydrate can regulate systemic lipid metabolism if, as shown in this study, it has no significant effect on the fecal excretion of lipids. The most likely mediators of such an effect may be the SCFA and/or lactate which are produced in significant amounts by ferment- ing bacteria. These carboxylic acids are absorbed through the intestinal membrane and reach the liver. Moreover, they have already been shown to act as regulators of key hepatic meta-

LIPIDS, Vol. 30, no. 2 (1995)

OLIGOFRUCTOSE LOWERS SERUM AND VLDL LIPIDS 167

=~

14 -

12 �84 .

10 -

8 .

6

4 '

2 '

A

P<O05

i . , i

o 30 60 90 12o ~so

I n c u b a t i o n t i m e ( m m )

=E

8" B

7 "

6"

5 -

4 -

0 , r , , =

0 3 0 6 0 9 0 1 2 0 150

Incubation time (rain)

FIG. 3. Effect of oligofructose (OFS) feeding on the capacity of isolated hepatocytes to incorporate [1J4C]palmitate into cellular triglycerides (A) and to secrete [14C] very low density lipoprotein (VLDL) into the cul- ture medium (B). Hepatocytes from control (R) or OFS-fed rats (@) were incubated for 150 min in Dulbecco's medium and analyzed as de- scribed in the Materials and Methods section. Variance analysis based on a two-way analysis of variance was used to test time/treatment inter- action which showed statistically significant changes (P < 0.05) for OFS feeding for incorporation into cellular triglycerides (A) but not for VLDL release (B).

A C K N O W L E D G M E N T S

We gratefully acknowledge the Fellowships received from the Belgian Insti- tute for Applied and Agricultural Research (IRSIA) (Maria Fiordatiso), and the Fund for Scientific Development, Universit6 Catholique de Louvain, Bel- gium (FDS-UCL) (Nadine Kok and Nathalie Delzenne).

REFERENCES

1. Oku, T., Tokunaga, T., and Hosoya, N. (1984) J. Nutr. 114, 1574-1581.

2. Rumessen, J.J., Bode, S., Hamberg, O., and Hoyer, E.G. (1990) Am. J. Clin. Nutr. 52,675-681.

3. Roberfroid, M. (1993) Critical Reviews in Food Science and Nutrition 33, ]03-148.

4. Modler, H.W., McKellar, R.C., and Yaguchi, M. (1990) J. Inst. Can. Sci. Technol. Aliment .23, 29-41.

5. Wang, X., and Gibson, G~R. (1993) Appl: Bacteriol. 75, 373-380.

6. Delzenne, N., Kok, N., Fiordaliso, M., Deboyser, D.,~ Goethals, F., and Roberfroid, M. (1993) Am. J. Clin. Nutr. (suppL) 57, 820S.

7. Tokunaga, T., Oku, T., and Hosoya, N. (1986) J. Nutr. Sci. Vit- aminol. 32, 111-121.

8. Roberfroid, M., Gibson, G., and Delzenne, N. (1993) Nutr. Rev. 51, 137-146

9. Berry, M.N., and Friend, D.S. (1969 J. Cell Biol. 43, 506-520. 10. Krack, G., Gravier, O., Roberfroid, M., and Mercier, M. (1980)

Biochim. Biophys. Acta 632, 619-629. 11. Capuzzi, D.M., Kackman, R.D., Uberti, M.M., and Ree, M.A.

(1974) Biochem. Biophys. Res. Common. 60, 1499-1508. 12. Deboyser, D., Goethals, F., Krack, G., and Roberfroid, M.

(1989) Toxicol. Appl. Pharmacol. 97, 473--479. 13. Chapman, M.J., Goldstein, S., Lagrangeand, D., and Lapland,

M.P. (1981)J. LipidRes. 22, 339-358. 14. Ichara, K., and Shibasaki, Y. (1991) J. Lipid Res. 32,

1709-1712. 15. Lowry, O.H., Rosebrough, N.J., Farr, A.L., and Randall, R.L

(1951) J. Biol. Chem. 193, 265-273. 16. Fungwee, T.V., Cagen, L., Wilcox, H.G., and Heimberg, M.

(1992) J. Lipid Res. 33, 179-191. 17. Groot, P.H.E., Scholte, H.R., and Hiilsmann, W.C. (1976)Adv.

Lipid Res. 14, 75-126. 18. Tsuji, Y., Yamamada, K., Hosoya, N., and Moriuchi, S. (1986)

J. Nutr. Sci. Vitaminol. 32, 93-100. 19. Vigne, J.L., Lairon, D., Borel, P, Portugal, H., Pauli, A.M.,

Hauton, J.C., and Lafont, H. (1987) Br. J. Nutr. 58, 405-413. 20. Mazur, A., Gueux, E., Felgines, C., Bayle, D., Nassir, F.,

Demign6, C., and R~m6sy, C. (1992) Proc. Soc. Exp. Biol. Med. 199, 345-350. Kasim, S.E., Leboeuf, R.C., Khilnani, S., Tallapaka, L., Dewunda, D., and Jen, D K.L.C. (1992)J. Lipid Res. 33, 1-7. Demign6, C., Yacab, C., and R6m~sy, C. (1986) J. Nutr. 116, 77-86.

21.

22.

bolic pathways (22). Studies are in progress to test these hy- potheses and, in particular, the role of the liver in mediating the hypolipidemic effect.

[Received December 8, 1993, and in revised form October 31, 1994; Revision accepted December 20, 1994]

LIPIDS, Vol. 30, no. 2 (1995)