Biosci. Biotechnol. Biochem., 69 (5), 932–938, 2005
Effect of Partially Hydrolyzed Guar Gum (PHGG) on the Bioaccessibility of Fat and Cholesterol Mans M INEKUS,1; y Mark J ELIER,1 Jin-zhong X IAO,2 Shizuki K ONDO,2 Keiji I WATSUKI,2 Sadayuki K OKUBO,2 Martin B OS,1 Bertus D UNNEWIND,1 and Robert H AVENAAR1 1 2
TNO Quality of Life, Zeist, The Netherlands Food Research and Development Laboratory, Morinaga Milk Industry Co., Ltd., Zama 228-8583, Japan
Received November 2, 2004; Accepted January 18, 2005
The Addition of a compound that lowers the intestinal uptake of fat and cholesterol might be an interesting strategy to reduce the risk of vascular disease. Partially hydrolyzed guar gum (PHGG) has been shown to have this effect in healthy volunteers after intake of a yogurt drink with 3 to 6% PHGG. In the present study a yogurt drink with 3% sunflower oil and 4% egg yolk was tested with 3% and 6% PHGG, and compared to a control without PHGG. Experiments were performed in a multi-compartmental model of the gastrointestinal tract, equipped to study the digestion and availability for absorption (bioaccessibility) of lipids. The results show that PHGG decreases the bioaccessibility of both fat and cholesterol in a dose-dependent manner. The bioaccessibility of fat was 79:4 1:7%, 70:8 2:5% and 60:1 1:1% for the control experiments and the experiments with 3% and 6% PHGG respectively. The bioaccessibility of cholesterol was 82:2 2:0%, 75:4 1:2% and 64:0 4:3% for the control and the experiments with 3% and 6% PHGG respectively. Additional experiments indicated that PHGG reduces bioaccessibility through the depletion flocculation mechanism. Depletion flocculation antagonizes the emulsification by bile salts and thus decreases lipolytic activity, resulting in a lower bioaccessibility of fat and cholesterol. Depletion flocculation with polymers might be an interesting mechanism, not described before, to reduce fat and cholesterol absorption.
negative effects of dietary fat and cholesterol by adding a compound to the meal that is able to counteract the digestion and/or absorption of cholesterol and fat. Some dietary fibers such as Guar gum have been shown to exert such an effect.1,2) Partially hydrolyzed guar gum (PHGG) has been shown to suppress postprandial serum lipid levels in healthy and moderately hypertriglyceridemial volunteers after a high-fat high-cholesterol meal.3) The aim of this study was to determine the effect of PHGG on the bioaccessibility of dietary fat and cholesterol. Bioaccessibility is defined as the fraction of a product that is available for absorption through the gut wall. In the case of lipophilic products this means the fraction of the product that is incorporated in mixed micelles and thus made accessible for intestinal absorption. The effect of PHGG on the bioaccessibility of fat and cholesterol was tested in a dynamic model of the gastrointestinal tract,4,5) equipped with a specific membrane that separates the micellar phase from the fat phase. This system mimics the major events occurring in the lumen of the gut, thus allowing study of the effects preceding absorption through the gut wall. Validation studies with this system on the availability of fat-soluble vitamins from processed vegetables show a good relation with in vivo data.6,7)
Materials and Methods Key words:
dietary fiber; cholesterol; fat; bioaccessibility; depletion flocculation
Intake of specially saturated fat and cholesterol is associated with the occurrence of vascular disease. Reduction of fat and cholesterol intake would therefore have a positive effect on health, especially for subjects that have a predisposition for high cholesterol, high blood pressure, diabetes or obesity. Apart from reduction of dietary intake, it is also possible to reduce the y
Test system. A schematic diagram of the test system is shown in Fig. 1. The system consisted of a gastric compartment (a), a duodenal compartment (b), and a small intestinal compartment (c). Each compartment was composed of two glass jackets with a flexible wall inside. The space between the glass jacket and the wall was filled with water to heat and mix the contents. Mixing was achieved by alternately squeezing the flexible walls through changing the pressure on the
To whom correspondence should be addressed. Fax: +31-30-6944-928; E-mail:
[email protected] Abbreviations: PHGG, partially hydrolyzed guar gum; FA, fatty acids; TFA, total fatty acids; GIT, gastro intestinal tract; SIES, small intestinal electrolyte solution
Effect of Fiber on Bioaccessibility of Fat and Cholesterol
e
(i). The level in the duodenal compartment was maintained by transferring contents to the small intestinal compartment with a peristaltic valve pump. The small intestinal compartment was connected to a filtration unit (j) with a micro filtration tube (X-flow, Enschede, The Netherlands). The mixed micelles that contained the bioaccessible lipophilic fraction were able to pass this filter, while undigested fat is retained. The micellar phase was pumped with a pump (k) into a vessel (l). The pump rate was controlled to maintain the level in the small intestinal compartment.
d a
f b
d
h
i
c
g
g
d
h
Experimental condition. The basic meal consisted of a low fat yogurt drink3) with 3% (w/w) sunflower oil and 4% (w/w) egg yolk. Test yogurts were supplemented with 3% (w/w) and 6% (w/w) PHGG. A yogurt without PHGG was used as a control. PHGG was produced by a partial hydrolysis of guar gum with a fungal beta-D-endomannanase (Taiyo Kagaku, Yokkaichi, Japan). The Average molecular weight determined by HPLC was 20,000 Daltons; the viscosity of a 5% solution was approximately 10 mPas. The meals were tested in duplicate during 4 h experiments in the GI model system, simulating average digestive conditions after the intake of yogurt. For this, 200 grams of meal was introduced into the gastric compartment that contained 10 ml of gastric secretion to simulate the initial amount of gastric juice. The gastric secretion consisted of pepsin (Sigma, 600 U/ml) and Lipase (Amano, 40 U/ml) in a gastric electrolyte solution, secreted at a flow rate of 0.5 ml/min.8) Hydrochloric acid was secreted to control the pH to follow a predetermined curve (Fig. 2). This curve represents the pH profile in the stomach after intake of a yogurt meal.9) Gastric emptying was based on the gastric emptying of yogurt, with a halftime of gastric emptying of 70 min
j
k
933
l Fig. 1. Gastrointestinal Model System. a, gastric compartment; b, duodenal compartment; c, small intestinal compartment; d, pH electrodes; e, gastric secretion: f, duodenal secretion; g, Bicarbonate secretion; h, level sensors; i, peristaltic valve pumps; j, separation membrane; k, filtration pump; l, collection vessels for the micellar phase.
water. The pH in each compartment was measured with a pH electrode (d). Artificial saliva and gastric juice were secreted in the gastric compartment (e). Hydrochloric acid was secreted to follow a preset pH profile (Fig. 2). Bile and pancreatic juice were secreted in the duodenal compartment (f). Bicarbonate was secreted in the duodenal and small intestinal compartment to adjust the pH to its set point (g). The levels in the compartments were monitored with a level sensor (h). The gastric compartment was gradually emptied according to a preset curve (Fig. 2) by pumping gastric contents into the duodenal compartment with a peristaltic valve pump
14
100 90
12 10
70 60
8
pH
Percentage of intake
80
50 6
40 30
4
20 2 10 0 0
60
120
180
0 240
Time (min) Fig. 2. Gastric Emptying Curve (broken line) and Gastric pH Profile (continuous line) Used to Control Gastric Emptying and pH in the Gastrointestinal Model System.
934
M. MINEKUS et al. 10)
(Fig. 2). The duodenal secretion consisted of fresh porcine bile at a flow rate of 1 ml/min, a 7% pancreatin solution at a flow rate of 0.5 ml/min, and small intestinal electrolyte solution (SIES: NaCl, 5 g/l; KCl, 0.6 g/l; CaCl2 , 0.25 g/l) at a flow-rate of 0.5 ml/min. For the first hour, bile solution was secreted also instead of the electrolyte solution to simulate the emptying of the bile bladder. The pH in both the duodenal and small intestinal compartment was controlled with a 1 M Sodium bicarbonate solution to set-points of 5.8 and 7.0 respectively. Both compartments were filled with secretion prior to the experiment to simulate initial intestinal contents. During the experiments, the total filtrate was collected to determine the bioaccessible fraction of fat and cholesterol. Residual material from the gastric and duodenal compartments was pooled to determine the fractions of fat and cholesterol that were not exposed to filtration. The residual material from the small intestinal compartment was collected to determine the fraction that was not filtrated and thus can be regarded as not bioaccessible. All samples were analyzed for total fatty acids (TFA) and cholesterol and compared to the TFA and cholesterol introduced in the system, to determine the recovery. Apart from the TFA and cholesterol in the meal, the amount of TFA and cholesterol in the secretion, especially the bile, was taken into account. TFA were analyzed after saponification to obtain a homogenous sample with all fatty acids liberated from the glycerol. Cholesterol is not affected by this saponification. After acidification, the free fatty acids and cholesterol were extracted into an organic solvent and then methylated using borium-tri-fluoride/methanol. The methyl esters obtained this way were analyzed by GC-FID. Quantification was done using C17:0 as an internal standard. Calculations. The amount of compound that empties from the gastric compartment might show some variation due to floating of lipids. This calls for a correction of compound that has not been exposed to small intestinal digestion. Therefore, the bioaccessibility is expressed as a percentage of the amount recovered from the small intestinal compartment, according to the following formula: Bioaccessibility ¼ ½F=ðF þ RÞ 100% Where F is the amount of measured compound in the micellar fraction and R is the amount of compound in the residual material from the small intestinal compartment. The recovery of compound was calculated by expressing the total amount of compound in the residues and in the filtrate as a percentage of the compound introduced into the system by the meal and the secreted liquids. Statistics. Comparison between variations was performed by a Student’s one tailed t-test, using the pooled standard error of the three variations in duplicate. A
significant difference between the means was considered to be present when P < 0:05. Mechanistic experiments. Studies were performed to get more insight in the possible mechanism whereby PHGG interferes with fat digestion. These studies focused on the surface activity of PHGG and the effect of PHGG on emulsion and digestive mixtures. Surface rheological measurements. Surface rheological experiments were performed to study the action op PHGG at interfaces in the presence of milk proteins and egg yolk. Yogurt with and without 6% PHGG was warmed to 37 C. Cooked egg yolk was mashed and added to a concentration of 3.5%, where after the mixtures were stirred for 30 min. The resulting mixtures were equal to the basic meal described above, except for the absence of sunflower oil. The surface active material present in the mixtures was material available for emulsifying the sunflower oil into yogurt. The mixtures were then 2 and 200 times diluted with SIES. The rigidity of the adsorbed layer at the surface of these mixtures was measured in a ring trough at 22 C.11) This adsorbed layer represents the layer of material adsorbed to the sunflower oil droplets. The ring trough is a special Langmuir trough with a cylinder barrier (a ground-glass cylinder, area 70.9 cm2 ). In this trough, the enclosed surface is expanded and compressed by moving the cylinder up and down in a sinusoidal way. The surface tension, which varies due to the variation in surface area, is continuously measured using a glass Wilhelmy plate. The applied frequency and deformation were 0.1 Hz and 4% respectively. Effect of PHGG on egg yolk and bile emulsions. Egg yolk and Bile solutions were prepared in duplicate according to the procedures described below. A 1% egg yolk solution was prepared by mixing cooked egg yolk with SIES at 40 C. Two egg yolk solutions were prepared. The solutions were adjusted to pH 4.5, to mimic the pH in the yogurt drink, and filtrated through a 0.2 mm filter to remove larger particles and to obtain a homogenous solution. To one solution, 6% PHGG was added. Bile was warmed to 40 C and filtered through a 0.2 mm filter to obtain a clear solution. Two 2.5% bile in SIES solutions were prepared, one as a control and one with 6% PHGG. Both solutions were set at pH of 7 by adding 0.01 M HCl, a normal pH for small intestinal contents. Both the egg yolk and bile solutions were used for the preparation of a 3% sunflower oil emulsion. Emulsions were prepared by stirring with a propeller-like blade for 30 min at 1000 rpm. During mixing, the temperature was kept at 37 C. Thereafter, samples were transferred into similar test tubes for examination. Effect of protein. Two emulsions were prepared by mixing 3% sunflower oil in skimmed milk diluted with SIES to 0.5% protein at 37 C with a propeller-like blade for 30 min at 1000 rpm. The milk was diluted to obtain a
Effect of Fiber on Bioaccessibility of Fat and Cholesterol Table 1. Consecutive Treatments of a Mixture of 0.5% Skimmed Milk and 3% Sunflower Oil with and without 6% PHGG to Mimic Different Conditions in the Gastrointestinal Tract treatment
mimicked condition
pH 4.5
meal pH
pH 2
gastric pH
150 U pepsin, for 1 h
gastric protein digestion
pH 7
small intestinal pH
2.5% bile
emulsification
more critical situation with regard to emulsion stability. To one mixture, 6% PHGG was added, and the pH of both mixtures was adjusted to 4.5 to mimic the acidic conditions in the yogurt. Under gentle stirring with a magnetic stirrer at 37 C, the mixtures were exposed to different treatments to mimic the sequential conditions in the gastrointestinal tract (Table 1). Samples were transferred into similar test tubes after each treatment for examination.
Results Experiments in the GI system The total input of fat was 12.25 g, of which 9 g originated from the meal and 3.25 g from the bile. The total input of cholesterol was calculated to be 700 mg, 120 mg from the meal and 580 mg from the bile. The average recovery of fat from all experiments was 78:6 8% and the average recovery of cholesterol was 76:1 11:4%. The bioaccessibility of fat from the control experiments without PHGG was 79:4 1:7% (Fig. 3). The experiments with 3% and 6% showed a fat bioaccessi-
935
bility of 70:8 2:5% and 60:1 1:1% respectively. The bioaccessibility of cholesterol was 82:2 2:0% for the control. The experiments with 3% and 6% showed a cholesterol bioaccessibility of 75:4 1:2% and 64:0 4:3% respectively. Mechanistic experiments Surface tension and dilatation modules The surface rheological experiments showed no difference between the yogurt drink with 6% PHGG and the control without PHGG in both the surface tension and the modulus (Fig. 4). Effect of PHGG on emulsions Macroscopic differences were observed between emulsions with and without the PHGG (Fig. 5). Therefore, the results are described macroscopically and presented schematically (Fig. 6). Basically, three phases could be distinguished: a sedimentation phase, a bulk phase, and a floating cream layer. Bile and egg yolk solutions The emulsion of sunflower oil with bile showed a clear de-emulsifying effect of PHGG (Fig. 5). The control, without PHGG, gave a turbid water phase with a stable cream layer on top. In the presence of PHGG, a cream layer with large fat droplets and free fat was formed. The emulsion with egg yolk showed a similar effect. Effects of protein, pH and pepsin digestion The emulsions of oil in diluted skimmed milk with PHGG showed a more turbid bulk phase than in the experiment without PHGG at pH 4.5, with a similar amount of sediment (Fig. 6). The cream layer with PHGG was smaller than the cream layer without PHGG. At pH 2, the sediment increased both with and without PHGG, but the bulk became clearer with PHGG. This
100 90 80 70
%
60 50 40 30 20 10 0 0%
3%
6%
% PHGG Fig. 3. Average Bioaccessibility of Fatty Acids (white bar) and Cholesterol (black bar) from Duplicate Experiments with Drink Yogurt without (control), with 3% and 6% PHGG (n ¼ 2, error bars present range). All differences between variations were significant (P < 0:05), except for the difference between the cholesterol bioaccessibility with 0% and 3% PHGG (P ¼ 0:057).
936
M. MINEKUS et al. 70
A 60
mN/m 50
40
30 0
50
100
150
200
Time (min)
50
B
40
mN/m 30 20 10 0
0
50
100
150
200
Time (min) Fig. 4. Surface Tension (A) and Surface Dilatation Modulus (B) of Yogurt Drink with PHGG (circle) and without PHGG (triangle) at a 2 Dilution (closed markers) and a 200 Dilution (open markers). Results of single experiments.
Oil + Yogurt
pH 4.5
pH 2
Fig. 5. Picture of Duplicate Experiments with a Mixture of 3% Sunflower Oil and 2.5% Bile (tubes marked B), and Duplicate Experiments with a Mixture of 3% Sunflower Oil, 2.5% Bile and 6% PHGG (tubes marked G).
+ pepsin
+ pH7
+ bile
- PHGG
+ PHGG
Fig. 6. Schematic Presentation of the Behavior of Sunflower Oil in Yogurt Mixture after Different Treatments That Mimic Subsequent Conditions in the Stomach and Small Intestine, with and without PHGG. The schematic shows the partitioning of the oil and protein fractions over a sediment phase, a bulk phase, and a creaming phase.
Effect of Fiber on Bioaccessibility of Fat and Cholesterol
was in contrast to the emulsion without PHGG in which the turbidity of the bulk phase increased. After the addition of pepsin, the sediment disappeared, while the cream layer became fluffier and showed a distinct oil layer in both the mixtures with and without PHGG. Peptic digestion resulted in a clearer bulk phase of the mixture with PHGG, with some aggregates. After adjusting both variations to pH 7, the bulk layer of the mixture with PHGG became almost completely clear, with some aggregates, while the cream layer changed from fluffy to a more compact layer. Addition of bile increased the turbidity of the bulk phase with PHGG. The cream layer with PHGG showed more oil droplets than in the control experiment without PHGG.
Discussion The results show a similar effect of PHGG on the bioaccessibility of fat and cholesterol. Bioaccessibility showed a decrease depending on the PHGG dose. This is in agreement with results obtained in human volunteers.3) To obtain more insight into the mechanism whereby PHGG reduces fat bioaccessibility, physical experiments were performed. These experiments showed that there is no adsorption of PHGG to the surface of the fat globules, either directly or via an interaction with surface active material. The experiments with bile and egg-yolk emulsions show that PHGG antagonizes emulsification, resulting in a cream layer with more and larger fat droplets than in the control. A de-emulsifying effect by guar gums has also been observed by other researchers,12) who found a positive relation between the viscosity of the fiber solutions and fat-droplet size. Although an effect of viscosity is difficult to rule out, we hypothesize another mode of action. PHGG shows no surface activity and has a mean molecular weight of app. 20 kD. This typically meets the profile of compounds that are able to cause depletion flocculation.13) This phenomenon was first described by Asakura and Oosawa.14) It occurs with polymers that show no surface activity and thus do not adsorb to the surface of fat droplets (Fig. 7). This results
937
in a zone around the fat droplet with a lower concentration of polymer (a depletion zone) as compared to the bulk solution. Under these conditions, it is energetically favorable for the droplets to overlap their depletion zone and drive the solvent towards the bulk phase to dilute the polymer outside the depletion zone. This phenomenon causes the fat droplets to flocculate and coalesce. We have shown that the presence of protein has an effect on depletion flocculation, depending on pH and digestive state. Protein can adsorb to the surface of the fat droplets and thus prevent the droplets from approaching each other and from flocculating.15) At pH 4.5, the PHGG appears to have a stabilizing effect, possibly due to the increased viscosity with PHGG or interactions between PHGG and the protein. At pH 2, the protein forms a thicker layer of sediment and, after the addition of pepsin, it is digested into small peptides. Digestion of the proteins had a big impact on the stability of the emulsion, resulting in a layer of oil floating on top of the mixtures with and without PHGG. This can be explained by the fact that the digested protein is less effective at the surface of the fat droplets at preventing flocculation. The presence of PHGG showed a thicker layer of free oil than in the control, indicating an additional effect of depletion flocculation, which forces oil droplets together and to coalesce. After adjusting the pH to 7, the effect of PHGG on the mixture is quite remarkable. With PHGG, the oil almost completely separated from the nearly clear bulk phase. The addition of bile resulted in a repartitioning of the oil between the bulk phase and the cream layer, due to the emulsifying action of the bile. With PHGG, large droplets of oil were present in the cream layer, indicating a depletion flocculation effect similar to the experiments with bile emulsions (Fig. 5). It is wellknown that emulsification by bile is an important prerequisite for efficient fat digestion. Counteraction of bile activity might therefore be an explanation for the reduced fat and cholesterol bioaccessibility observed with PHGG in the meal. So far, the effects of dietary fiber on lipid absorption have been brought into relation with increased viscosity and binding of lipids. Depletion flocculation with polymers might be an interesting mechanism not described before, to reduce fat and cholesterol absorption.
References 1)
bio polymer
2) fat droplet depletion zone Fig. 7. Schematic Presentation of the Depletion Flocculation Mechanism.13,16)
3)
Chen, W., and Anderson, J. W., Effects of guar gum and wheat bran on lipid metabolism of rats. J. Nutr., 109, 1028–1034 (1979). Ide, T., Moriuchi, H., and Nihimoto, K., Hypolipidemic effects of guar gum and its enzyme hydrolysate in rats fed highly saturated fat diets. Ann. Nutr. Metab., 35, 33– 44 (1991). Kondo, S., Xiao, J. Z., Takahashi, N., Miyaji, K., Iwatsuki, K., and Kokubo, S., Suppressive effects of dietary fiber in yogurt on the postprandial serum lipid levels in healthy adult male volunteers. Biosci. Bio-
938
4)
5)
6)
7)
8)
9)
10)
M. MINEKUS et al.
technol. Biochem., 68, 1135–1138 (2004). Minekus, M., Marteau, P., Havenaar, R., and Huis in ’t Veld, J. H. J., A multi-compartmental dynamic computer-controlled model simulating the stomach and small intestine. Alternatives to Laboratory Animals (ATLA), 23, 197–209 (1995). Minekus, M., Development and validation of a dynamic model of the gastrointestinal tract. Ph.D. thesis, University of Utrecht, Elinkwijk b.v., Utrecht, The Netherlands (1998). Minekus, M., Lelieveld, J., and van den Berg, H., A dynamic model of the stomach and small intestine to study the bio-accessibility of carotenoids from vegetables and the effect of processing. In ‘‘Abstract Book of the Bioavailability Congress’’, Interlaken, Switzerland, p. 111 (2001). Southon, S., Model systems in vitro and in vivo, for predicting the bioavailability of lipid soluble components in food. FAIR CT97-3100 (Final Report), 30-40001 (2001). Marteau, P., Minekus, M., Havenaar, R., and Huis in ’t Veld, J. H. J., Survival of lactic acid bacteria in a dynamic model of the stomach and small intestine: validation and the effects of bile. J. Dairy Sci., 80, 1031– 1037 (1997). Conway, P. L., Gorbach, S. L., and Goldin, B. R., Survival of lactic acid bacteria in the human stomach and adhesion to intestinal cells. J. Dairy Sci., 70, 1 (1987). Marteau, P., Pochart, P., Mahe´, S., Crine, L., Huneau,
11)
12)
13)
14)
15)
16)
J. F., Tome´, D., and Rambaud, J. C., Gastric emptying but not oroceacal transit time differs between milk and yoghurt in lactose digesters. Gastroenterology, 100, A535 (1991). Kokelaar, J. J., Prins, A., and de Gee, M., A new method for measuring the surface dilational modulus of a liquid. J. Colloid Interface Sci., 146, 507–511 (1991). Pasquier, B., Armand, M., Guyon, F., Castellain, C., and Lairon, D., Soluble fibers alter emulsification of dietary fat in vitro. In ‘‘Mechanisms of Action of Dietary Fiber on Lipid and Cholesterol Metabolism’’, ed. Lairon, D., Proceedings of a cost 92 workshop, Le Rouet-Marseille, France (1993). Robins, M. M., and Hibberd, D. J., Emulsion flocculation and creaming. In ‘‘Modern Aspects of Emulsion Science’’, ed Binks, B. P., Royal Society of Chemistry, England, Chap. 4 (1998). Asakura, S., and Oosawa, F., On the interaction between two bodies immersed in a solution of macromolecules. J. Chem. Phys., 22, 1255–1256 (1954). Koczo, K., Nikolov, A., Wasan, D., Borwankar, R., and Gonsalves, A., Layering of sodium caseinate submicelles in thin liquid films: A new stability mechanism for food dispursions. J. Colloid Interface Sci., 178, 694–702 (1996). Blijdenstein, T., Microstructure, rheology and demixing in emulsions flocculated by polysaccharides. Ph.D. Thesis, Wageningen, Centre for Food Sciences (WCFS), Wageningen, The Netherlands (2003).
