David Oakenfull

A mechanism for the hypocholesterolaemic activity of saponins

1. Saponins are steroid or triterpene glycosides which occur in a number of important food plants, including such staples as soya beans (Glycine max) and chickpeas (Cicer arietinum). They are known to be hypocholesterolaemic. 2. Some saponins form an insoluble complex with cholesterol which prevents its absorption from the small intestine. Others cause an increase in the faecal excretion of bile acids, an indirect route for elimination of cholesterol. 3. We have investigated the effects of different saponins on absorption of the bile salt sodium cholate from perfused loops of small intestine, in vivo, in the rat. Purified saponins from soapwort (Saponaria officinalis), soya beans and quillaia (Quillaia saponaria) reduced the rate of absorption of the bile salt; soya-bean and soapwort saponins substantially so but quillaia saponin to a much lesser extent. 4. These results were explained by the formation of large mixed micelles by bile acid and saponin molecules in aqueous solution. These aggregates can have molecular weights in excess of 10(6) daltons, consequently the bile acid molecules incorporated in them are not available for absorption. 5. Control of plasma cholesterol and nutrient absorption through dietary saponins could provide substantial health and nutritional benefits in humans.

Saponins in food-A review

Abstract Saponins are surface active sterol or triterpene glycosides. They occur in a large number and a wide variety of plants but only about 28 of these are regularly used as food by man. The more commonly eaten of these are soybeans, chick peas, peanuts and spinach. Many different saponins occur, even within a single plant species. Although they can be readily isolated from plant materials by solvent extraction, it is difficult to isolate individual saponins from the crude mixture and this has rarely been achieved. The presence of saponins in plant extracts is readily indicated by their haemolytic activity and ability to form stable foams in aqueous solution, but for unambiguous identification, it is essential to use thin layer chromatography, with as wide as possible a range of spray reagents. Quantitative analysis of saponins is also best achieved by thin layer chromatography. Older methods, such as that based on haemolytic activity, are not reliable. Full chemical structures of some of the food saponins are known, but for the most part information is at present limited to the structure of their aglycones and the identity of the sugars. Although saponins have antibiotic activity and are toxic to fish and insects, they appear to be practically non-toxic to man, remaining within the gastrointestinal tract. Dietary saponins, either isolated or as saponin-containing food plants, lower plasma cholesterol levels in several mammalian species. They are therefore probably important in human diets to reduce the risk of coronary heart disease.

Effects of saponins on bile acids and plasma lipids in the rat.

1. The effects of feeding isolated saponins on plasma lipid concentrations and on concentrations of biliary and faecal bile acids and neutral sterols were studied in the rat. 2. The animals were given one of four diets, i.e. a standard low-cholesterol synthetic diet, the diet + 10 g saponins/kg, the diet + 10 g cholesterol/kg, the diet + 10 g cholesterol + 10 g saponins/kg. 3. Saponins partially reversed the hypercholesterolaemia caused by the high-cholesterol diet and increased both the rate of bile acid secretion and the faecal excretion of bile acids and neutral sterols. The proportionate contribution of the primary bile acids (particularly chenodeoxycholic) to faecal excretion was also increased by saponins. 4. The results are discussed in relation to the hypothesis that saponins act by inducing the adsorption of bile acids by dietary fibre.

Relationship between structure and function of dietary fibre: a comparative study of the effects of three galactomannans on cholesterol metabolism in the rat

Male adult rats were fed on diets containing 80 g/kg galactomannans with different galactose (G): mannose (M) ratios/kg. The galactomannans were compared with purified cellulose (Solkaflok) and the animal were also fed on a basal diet free from fibre. All diets contained cholesterol (10 g/kg) and sodium cholate (2 g/kg). The three galactomannans were fenugreek gum (1G:1M), guar gum (1G:2M) and locust-bean gum (1G:4M). In comparison with the fibre-free and Solkaflok diets, all three galactomannans lowered the concentrations of cholesterol in both liver and blood plasma. The galactomannans also decreased the rate of hepatic synthesis of cholesterol. Dietary galactomannans increased caecal volatile fatty acids, particularly propionic, increased the weight of the caecum and its contents and increased the amount of water in the faeces. The increase in propionic acid production was significantly related to a decrease in caecal pH, but not to changes in plasma cholesterol or hepatic cholesterol synthesis. These effects were significantly influenced by chemical composition and structure of the galactomannan; they were most evident when the proportion of galactose in the galactomannan was highest (i.e. fenugreek gum). The three galactomannans also differed markedly in their effects on the viscosity of the digesta, but the galactomannan which gave the highest viscosity was least effective in lowering plasma cholesterol. A separate experiment with perfused loops of small intestine in vivo showed that the most effective galactomannan, fenugreek gum, had no direct effect on cholesterol absorption.

A viscous fibre (methylcellulose) lowers blood glucose and plasma triacylglycerols and increases liver glycogen independently of volatile fatty acid production in the rat

1. Adult male rats were maintained on diets containing 80 g methylcellulose/kg of low (25 cP), medium (400 cP) and high (1500 cP) viscosity. 2. After 10 d, the viscosity of stomach and caecal contents was found to have increased in proportion to that of the dietary fibre. Concentrations of volatile fatty acids in caecal digesta were lowest with the high-viscosity fibre but acetate was the major acid present with all three diets. Acetate was the only acid found in significant quantities in hepatic portal venous plasma and concentrations of this acid were unaffected by diet. 3. Concentrations of glucose in arterial blood were low with the medium- and high-viscosity diets while the content of liver glycogen was high. These effects of fibre were not directly on glucose absorption as the intestines were net removers of the hexose at the time of sampling. 4. Hepatic lipogenesis and plasma triacylglycerol concentrations were both higher in rats fed on the low-viscosity fibre. Plasma cholesterol concentrations, hepatic cholesterol synthesis and faecal bile acid excretion were not altered by dietary fibre viscosity. 5. We conclude that the effects of dietary fibre on carbohydrate absorption and storage and fatty acid synthesis are a function of the viscosity of the fibre in solution, high viscosity slowing the digestion and absorption of nutrients in the small intestine. Large-bowel microbial fermentation is not of direct significance to these events. In contrast, effects of fibre polysaccharides on sterol metabolism seem not to be related to their rheological properties.

Stabilization of gelatin gels by sugars and polyols

Abstract Gelatin gels are known to be stabilized by sugars and polyols. We have made a quantitative study of this effect with the aim of determining the molecular mechanisms involved. The sugars and polyols added were sucrose, glucose, fructose, sorbitol, glycerol and ethylene glycol, at concentrations within the range 0 – 500 g/kg. With the exception of ethylene glycol, their effects were almost identical — at 400 g/kg of additive the shear modulus of the gels increased by a factor of 2.5 and the rate of gelation increased by a factor of 2. Ethylene glycol produced smaller changes. Sucrose at 500 g/kg increased the melting temperature by 3.9°C. The increased rate of gelation could be simply explained as due to increased melting temperature. The junction zones in gelatin gels are regions of collagen triple helix structure and added sugar or polyol increased the proportion of protein reverting to this conformation. The contribution per amino acid residue to the free energy of formation of junction zones changed from –0.34 kJ/mol in water to –3.44 kJ/mol with sucrose at 500 g/kg. The junction zones became smaller (averaging 15 turns of triple helix in water and 0.8 turns in sucrose at 500 g/kg) but at the same time more numerous. This results in a more extensive gel network and consequently greater rigidity. Stabilization appears to be due to preferential hydration of the unfolded polypeptide in the presence of sugar or polyol — rather than a direct effect of the additive on the intermolecular forces stabilizing the collagen triple helix.

Effects of t-butanol on micelle formation by a series of long chain alkyltrimethylamonnium bromides

Abstract Micelles of decyl-, dodecyl-, and hexadecyltrimethylammonium bromide (C 10 TAB, C 12 TAB, C 16 TAB) were studied at 298°K in 0.5 mole dm −3 sodium bromide and in the presence of various concentrations of t -butanol. (1) A fluorescent probe, N -phenylnaphthylamine, was used to estimate critical micelle concentrations (CMC) and the polarity of the hydrocarbon core of the micelles. (2) The ultraviolet absorption spectrum of solubilized benzene was used to estimate the polarity of the headgroup region. (3) Viscosity measurements were used to derive information about micelle shapes and solvent penetration. (4) Low-angle laser light scattering was used to determine aggregation numbers. Added t -butanol lowers the CMC of C 10 TAB but increases that of C 16 TAB. When mole fraction of t -butanol reaches 0.03 the micelles have a looser structure with greaer solvent penetration than the classical detergent micelle. Micelles are not formed when the mole fraction of t -butanol exceeds 0.03. t -Butanol penetrates the hydrocarbon core of the micelles and is adsorbed at the head group region. The optimum surface area per head group is consequently increased with the result that the micelles are smaller in the presence of t -butanol than in water alone and they are spherical rather than ellipsoidal.

Hydrophobic interaction in aqueous organic mixed solvents

The standard free energy of hydrophobic interaction between alkyl chains has been measured in aqueous solutions of t-butanol, dioxan, dimethyl sulphoxide and dimethyl formamide.The results make it clear that hydrophobic interaction (in the sense of a pair-wise interaction between hydrophobic groups otherwise surrounded by water) is not simply a partial reversal of solution of the non-polar molecules in water. Hydrophobic interaction should therefore be distinguished from phenomena, such as the formation of detergent micelles or phospholipid bilayers, in which non-polar molecules are completely removed from water. Only in the latter case is “reversal of solution” a correct description.The experimental method was measurement of conductance of long chain alkyltrimethylammonium carboxylate solutions. Ion pair association constants, and hence free energies of ion pair formation were calculated from these data. Then by systematically varying the alkyl chain length it was possible to separate the free energy of hydrophobic interaction from the electrostatic interaction between the charged groups.Moderate concentrations of t-butanol, dimethyl sulphoxide and dimethyl formamide strengthened hydrophobic interactions. All concentrations of dioxan weakened hydrophobic interactions. Maximum hydrophobic interaction occurred with those solute concentrations which maximised the three dimensional hydrogen bonded structure of water.The results explain some of the odd effects of alcohols and other structure making solutes on the stability of proteins.

Constraints of molecular packing on the size and stability of microemulsion droplets

When considerations of molecular pakcing are combined with a simple thermodynamic treatment, it is possible to calculate many of the characteristics of a microemulsion from the molecular dimensions of the detergent. The only parameters required are the length and volume of the non-polar tail of the detergent and the apparent molar volume in water of the head-group, all of which can readily be estimated. It is shown that a cosurfactant is required when the structure of the detergent is such that the polar head-groups surrounding the microemulsion droplet would otherwise pack too closely together. Equations are derived which give (1) the size of the droplets as a function of the molar ratio of water to detergent, (2) the maximum droplet size that can be attained (the solubilisation capacity) and (3) indicate the size distribution. These equations give excellent agreement with experimental data.