![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
Institute of Food Research, Norwich Research Park, Colney, Norwich NR4 7UA, United Kingdom
ABSTRACT
INTRODUCTION
METHODS
RESULTS
DISCUSSION
FOOTNOTES
LITERATURE CITED
ABSTRACT 
The creation of a finely dispersed lipid emulsion is essential for efficient hydrolysis of dietary triglycerides. The effectiveness of emulsification within the stomach depends upon the shear force generated by gastric motility and the concentration of emulsifiers present in the gastric contents. Other dietary constituents can modify these factors, and previous studies have suggested that the presence of soluble non-starch polysaccharides (NSP) during digestion might increase the average size of intraluminal emulsion droplets. In the present study, we developed a new technique for the isolation and analysis of intraluminal lipid emulsions by optical diffraction analysis. The method was applied to rats fed powdered semipurified diets that were free of all NSP or supplemented with insoluble cellulose, guar gum, or NSP derived from apple, carrot or rolled oats. Cellulose had no significant effect on emulsion size, and there was no evidence that the average sizes of lipid droplets in the gastric fundus or antrum were higher than control values in rats fed diets supplemented with any source of soluble NSP. In the groups fed oats and cooked carrot NSP, the mean droplet diameters approached half the values for diets free of NSP or containing insoluble cellulose. The difference between rats fed NSP from cooked carrot and those fed cellulose was significant in the proximal stomach (P < 0.05), and that between rats fed raw oats and rats fed cellulose was significant in the distal stomach (P < 0.05). Soluble dietary fiber does not inhibit lipid or cholesterol absorption via any inhibition of lipid emulsification.
KEY WORDS: lipids · dietary fiber · non-starch polysaccharide · digestion · ratsINTRODUCTION
Dietary triglycerides are hydrolyzed to monoacylglycerols and fatty acids by lipases that are activated by the lipid-water interface generated during intraluminal emulsification of dietary lipids (Carey 1983
). Emulsification begins during mastication and continues in the gastric antrum, where shear forces are relatively high (Prove and Ehrlein 1982). The principal emulsifiers present during this processes are phospholipids and the fatty acids produced by the action of gastric lipase. It is probable that most of the fatty acids partition into the core of the emulsion droplets, but a small fraction are present at the interface (Carey 1983). In laboratory rodents, between 10 and 30% of dietary fat digestion occurs within the acid environment of the stomach (Armand et al. 1994). Efficient lipid digestion requires the formation of a relatively stable and finely dispersed emulsion with an interfacial composition favorable to the anchoring of lipases; indeed, studies in vitro have confirmed that the rate of lipolysis is directly proportional to the available surface area (Borel et al. 1994). Using model emulsions, we have previously shown that the phospholipids exist as liquid crystalline outer layer, with a thickness of 60- 70 µm (Fillery-Travis et al. 1995). This stabilizes the emulsion against aggregation and coalescence of the droplets and allows a large droplet surface area to be maintained.
Recent reports have shown that different sources of non-starch polysaccharide (NSP)4 can modify postprandial lipemia in humans, but the mechanisms are poorly understood (Cara et al. 1992, Dubois et al. 1992 and 1993). Pasquier et al. (1996) investigated the effect of polysaccharides on lipid emulsification in vitro and proposed that viscous NSP may increase the average droplet size of emulsified lipid in vivo and hence slow the rate of triglyceride hydrolysis. In the present study, we have explored this hypothesis by measuring the effects of NSP on the characteristics of emulsified lipid in the gastric lumen of rats in vivo. It is only recently that attempts have been made to characterize the emulsion droplet size distribution formed in the stomach, and there are few, if any, previous data on the effect of other dietary constituents on lipid emulsification. Because of the diverse physicochemical properties of the NSP composing dietary fiber, the effects of any particular food are not predictable from a single analytical value for NSP content measured in vitro. One strategy to overcome this limitation is to subdivide the total NSP content of a food into soluble and insoluble fractions, but even so it is unlikely that the many soluble NSP components derived from the whole variety of sources that compose a complex diet all have similar physiological effects (Johnson 1993). To take account of this physicochemical diversity, we fed semipurified diets containing foods rich in NSP, such as carrot, apple and oats, or purified isolates of NSP derived from them. Gastrointestinal contents from these animals, and from controls fed insoluble cellulose or fiber-free (FF) diets, were recovered for analysis.
To characterize and quantify the behavior of intraluminal lipid droplets under physiological conditions, we developed a new technique for the isolation and analysis of the intraluminal lipid emulsion by optical diffraction analysis. Optical diffraction has considerable advantages over microscopy as a method of characterizing lipid droplet distributions because of its ability to access droplet sizes in the range of 0.5-500 µm. For this method to be valid, however, the sample must be free of non-lipid particles of similar size. We used optical microscopy to confirm that no observable coalescence occurred during the separation of starch and other particles from the emulsion.
METHODS
. Uronic acids were quantified colorimetrically by the method of Blumenkrantz and Asboe-Hansen (1973). Apple and carrot tissues contained small but important quantities of starch that interfered with the analysis of cell wall glucose. Estimation of total levels of NSP would have required the removal of the starch. However, in the cooked tissue, the starch was gelatinized, and cooling resulted in retrogradation, making the starch very difficult to remove without heat treatments and further solubilizing cell wall components (Gooneratne et al. 1994). Heat treatment of fruit and vegetable tissues results in an increase in the solubility of, predominantly, the pectic polysaccharides from the cell wall (A. Ng and K. Waldron, unpublished data; Greve et al. 1994). Therefore, in the present work, the total NSP was estimated from levels of uronic (galacturonic) acid, arabinose, galactose and rhamnose.
Diets.
All experimental and control diets were semipurified powdered diets containing sucrose, casein and olive oil as the main ingredients. Insoluble cellulose (Solka Flok; Johnson Jorgensen Wettre, London, UK) and guar gum (Sigma, Poole, UK) were used as reference sources of insoluble and soluble NSP, respectively, and incorporated into the diets at a level of 5 g/100 g (Table 1). Apples, carrots and oats were chosen as representative sources of soluble NSP in the human diet and obtained locally. The batches were subdivided randomly into cooked and raw subsamples and then freeze-dried and ground for incorporation into the semipurified rat diets. Isolated fractions of NSP were prepared from cooked carrot and cooked cabbage leaves as AIR. The plant tissues were homogenized in boiling alcohol (85%, v/v) for 1 min, boiled for a further 5 min and rehomogenized. The insoluble residue was recovered by filtration, re-extracted once in boiling 85% alcohol, once in absolute alcohol and once in cold acetone and finally air-dried. The NSP sources were incorporated into powdered rat diets so as to provide the quantities of total NSP and soluble NSP shown in Table 2. The values shown for the oat diets were calculated from the literature (Lund et al. 1989). The other ingredients were adjusted to provide approximately constant levels of protein and lipids (Table 2), using values obtained from food composition tables (McCance and Widdowson 1978).
|
Table 1. Composition of diets |
|
Table 2. Protein, fat and non-starch polysaccharides (NSP) in diets |
) of similar concentration. Significant errors were obtained when measuring droplets under two divisions (4.6 µm) in diameter, and therefore only droplets larger than 4.6 µm were measured.
Optical diffraction analysis of intraluminal emulsions.
Samples were thawed, thoroughly mixed and diluted to a volume of approximately 15 mL in saline (9 g/L NaCl, pH 2), sonicated for 20 min at 11,000 g at ~0°C, centrifuged for 2 h (5°C) and removed to an ice bath. The lipid layer formed at the top of the tube was removed and the pH readjusted. Investigation of the pellet and the immediately adjacent supernatant for lipid, using Nile blue stain, showed the separation to be effectively completed, and no evidence of lipase activity was detected. The lipid layer was then diluted to obtain a final lipid concentration of ~0.03% (wt/wt) and placed in the sample holder of an optical diffraction particle sizer (Malvern Mastersizer; Malvern Instruments, Malvern, UK) having a measuring range of 0.1-1000 µm. Some modification of the apparatus was necessary, i.e., bypass of the small sample handling unit, to facilitate measurement of small sample volumes (10 mL), but all the adjustments were validated with the hexadecane emulsion. All measurements were made in duplicate using lenses appropriate to the size range.
Zeta potential measurement.
The microelectrophoretic mobilities of the emulsion droplets within the lipid layer collected for the FF and cabbage AIR diets were measured in a Zetasizer 3 (Malvern Instruments) with an AZ4 capillary electrophoresis cell. The continuous phase employed as diluent for each emulsion was saline (9 g/L NaCl, pH 2). The corresponding zeta potentials were calculated from the mobility measurements using the appropriate models (Walstra 1983).
Statistical analysis.
Frequency distributions describing the range of droplet sizes within a sample were obtained by optical diffraction analysis. The mean values for such distributions were calculated for individual rats, and comparisons across the experimental groups were performed using one-way ANOVA. Intergroup comparisons were performed using Tukey's test for all comparisons, or Dunnett's test when a single control group (insoluble cellulose) was compared with all other groups (Hayter 1984). Means were accepted as significantly different at a probability level of less than 0.05. All calculations were performed using the Minitab statistical package (Minitab Inc, State College, PA). Values in the text are means ± SEM.
RESULTS
). In the present study, the emulsions obtained from gastric digesta were also birefringent under crossed polarizers and showed a degree of stability that suggests that a liquid crystalline interface was also present at the droplet surface.
). Hence this has become the distribution function most commonly employed to describe droplet size data. Examination of the size data obtained in this study and illustrated in Figure 1 showed a symmetrical distribution obtained when the data were described by a log-normal function. There is a close relationship between the normal and log-normal distributions inasmuch as in the latter the arithmetic mean diameter and the SD are replaced by their log equivalents. Thus the data can be described using the geometric mean and SD as summary statistics. The use of this function allows the distribution of sizes to be calculated as a function of total surface area of the emulsion, total weight or volume of lipid, or total number of lipid droplets. When comparing the distribution of droplet sizes obtained with the optical diffraction method with that obtained by the droplet counting method using optical microscopy, it is clear that the distributions based on number of droplets (D1,0) should be used in the comparison. However, when comparing the relative efficiency of droplet breakup within the emulsification process, the issue of interest is the amount of lipid incorporated within each droplet size class, and hence the surface mean (D3,2) diameters were employed.
Comparison of microscopical and optical diffraction methods. The effect of sample preparation on the emulsions prepared for diffraction analysis was assessed by comparing samples from rats fed the FF diet or a diet enriched with alcohol-insoluble fiber derived from cabbage. The gastric contents were analyzed both by light microscopy and by optical diffraction. The measurement range of the optical diffraction technique was much larger than that accessible by optical microscopy. Therefore to achieve a direct comparison the analysis range of the Mastersizer was restricted to droplets larger than 4.6 µm, so as to mimic the measurements performed using optical microscopy. The results of the two analyses are compared in Table 3, which gives the droplet number means, D(1,0), obtained by both methods.
|
Table 3. Comparison of lipid emulsion droplet sizes, as number means D (1,0), in material obtained from the proximal and distal stomach of rats fed two test diets as measured by light microscopy and light diffraction1 |
2.9 ± 1.6 mV) were not significantly different from those for rats fed cabbage AIR (1.5 ± 0.2 mV).
Food consumption and gastric emptying in rats fed different sources of non-starch polysaccharides.
Rats were successfully trained to consume their daily energy requirements at a single meal. The rats fed the FF control diet in the first experiment consumed 13.08 ± 0.71 g of food and those in the second experiment consumed 16.28 ± 0.96 g (P < 0.05). The quantities of foods consumed by the NSP groups, together with the proportions of the meals consumed that remained in the stomach at the time of sampling, are given in Table 4.
|
Table 4. Diet consumed by rats and proportion of the meal remaining in the stomach 6 h after consuming the fiber-free and fiber-containing diets1 |
DISCUSSION
, who recovered human gastric contents and reported that most lipid was present as emulsified droplets in the size range of 20-40 µm. However, contrary to the conclusions of Pasquier et al. (1996), which were based on in vitro studies, we did not obtain evidence that soluble NSP derived from complex food sources caused any increase in the average droplet size during gastric emulsification. Indeed, in rats fed raw oats, which are an established source of increased viscosity in the alimentary tract of rats (Lund et al. 1989), the mean droplet size in the distal stomach was significantly lower than in the controls fed insoluble cellulose. This discrepancy between our work and that of Pasquier et al. (1996) probably reflects the limitations of in vitro techniques for modeling gastric conditions in vivo.
). The antrum is therefore the most likely site of emulsification, and from the geometry and periodicity of the contractions it would seem probable that turbulent flow will occur under these conditions. Droplet disintegration probably takes place within eddies formed by the turbulence. The current models of emulsification in turbulent flow are still incomplete (Walstra 1983), but they do suggest that the viscosity of the continuous phase is not the primary determinant of droplet size, as long as it is not so high as to prevent isotropic turbulence. Empirically, the maximum droplet size has been found to be a function of interfacial tension at the droplet surface, energy input, and to a lesser extent the density of the continuous phase (Walstra 1985). If this simple model holds in this instance, then the viscosity of the continuous phase would not be expected to influence emulsion droplet size unless there was a corresponding change in the energy input into the system, for example, if the mode and frequency of the peristaltic contractions were influenced by the viscosity of the chyme. Several previous studies have described significant changes to the amplitude and duration of gastric contractions in response to the consistency of the chyme, but there has been no quantification of the change in shear and energy input (Kumar and Gustavsson 1988). However, the discrepancy between our results and the findings of Pasquier et al. (1996) may be due to the fact that, unlike simple in vitro systems, the stomach is able to respond to increased viscosity by increasing its energy input through a modulation of antral contractions, which could then lead to a smaller mean droplet diameter.
suggests that surfactant is available in excess. There are some food gums that are known to stabilize an oil-water interface, gum arabica being one example (Koseki et al. 1989), but this behavior is usually a consequence of the protein components of the gum. Our previous in vitro study of the effects of polysaccharides on emulsification did not suggest any direct interaction of the polysaccharide with the emulsion interface (Fillery-Travis et al. 1993b). However, the presence of soluble polysaccharides may increase the availability of endogenous surfactant, perhaps by enhancing gastric secretions.
, Blackburn et al. 1984, Johnson and Gee 1981). Viscous NSP are also known to reduce plasma cholesterol concentrations in experimental animals and humans. It has been proposed that this effect might be due to inhibition of fat or cholesterol absorption, but the present results suggest that any such effect of soluble fiber is not mediated by a reduction in the efficiency of gastric emulsification of lipid. A reduction in the average lipid droplet size implies both a stabilization of the emulsion and an increase in its surface area, immediately prior to its passage into the duodenum. These conditions would favor a relatively high rate of lipolysis and hence perhaps a faster rate of lipid digestion and assimilation from diets containing soluble NSP. There are some previously published findings that might provide circumstantial evidence for such a mechanism. For example, Cara et al. (1992) described evidence for a slightly higher rate of triglyceride uptake after consumption of oat bran, which is a rich source of soluble NSP, compared with wheat bran, which does not contain substantial amounts of pectin or beta glucan. Furthermore, Redard et al. (1990) reported that, at least in female subjects, consumption of test meals containing relatively small quantities of oat bran and guar gum led to significantly higher postprandial plasma triglyceride concentrations compared with a low fiber control meal. These results can be explained if some soluble polysaccharides are able to enhance the rate of lipid hydrolysis by reducing average emulsion size.
FOOTNOTES
Manuscript received 2 December 1996. Initial reviews completed 3 January 1997. Revision accepted 22 July 1997.
LITERATURE CITED
-phosphatidylcholine and sodium taurocholate.
Biophys. Chem.
1995;
54:253-260
effect of soluble non-starch polysaccharides In: Food Colloids and Biopolymers: Stability and Mechanical Properties (Dickinson, E. & Walstra, P., eds.), pp. 280-284. Royal Society of Chemistry, Cambridge, UKa useful concept? In: Plant Polymeric Carbohydrates (Meuser, F., Manners, D. J. & Seibel, S., eds.), pp. 147-153. Royal Society of Chemistry, Cambridge, UKinvolvement of cell-wall phenolics.
J. Sci. Food Agric.
1995;
68:337-346This article has been cited by other articles:
|
|
 |
|
  A. J. Edwards, C. H. Nguyen, C.-S. You, J. E. Swanson, C. Emenhiser, and R. S. Parker {alpha}- and {beta}-Carotene from a Commercial Carrot Puree Are More Bioavailable to Humans than from Boiled-Mashed Carrots, as Determined Using an Extrinsic Stable Isotope Reference Method J. Nutr., February 1, 2002; 132(2): 159 - 167. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
