Effects of some purified saponins on transmural potential difference in mammalian small intestine

Toxic. in Vitro Vol. 3, No. 2, pp. 85-90, 1989 Printed in Great Britain. All rights reserved

0887-2333/89 $3.00+0.00 Copyright © 1989 Pergamon Press plc

EFFECTS OF SOME PURIFIED SAPONINS ON TRANSMURAL POTENTIAL DIFFERENCE IN MAMMALIAN SMALL INTESTINE J. M. GEE, K. R. PRICE, C. L. RIDOUT, I. T. JOHNSONand G. R. FENWICK Agricultural and Food Research Council Institute of Food Research, Norwich Laboratory, Colney Lane, Norwich, Norfolk NR4 7UA, England (Received 26 May 1988; revisions received 28 September 1988)

Abstract--The effect of a range of saponins, commonly present in foods or dietary supplements, on the potential difference (PD) across the mucosa of the rat small intestine/n vitro has been examined. Saponins from Gypsophila, guar, alfalfa, Quillaja, clover and liquorice together with glycoalkaloids from the potato and tomato were examined. The typical response was an immediate reduction in PD, although there was considerable variation in the response to particular compounds. Amongst the factors affecting the nature and magnitude of the de-polarizing effect were pH, solubility and the chemical form of the saponin. In agreement with the findings of others, glycyrrhizic acid, isolated from liquorice root, was found to exhibit a protective effect against the activity of a more potent saponin. The observations are discussed in the light of the known physiological activities of plant saponins and the regular, or excessive, consumption of certain foods or dietary supplements.

The present paper reports more detailed findings on the effects of a range of structurally-diverse saponins isolated from plants used as foods and dietary supplements. Factors affecting the observed gut permeabilizing behaviour of the saponins are examined, and the possible consequences of the presence of these compounds in the human diet are discussed.

INTRODUCTION

Saponins are widespread in the plant kingdom, being found in more than one hundred different families (Bader and Hiller, 1987). Their occurrence in foods and animal feedstuffs has recently been reviewed (Price et al., 1987) and their biological activities have been considered. Although most saponins share common surface-active properties because of their lipophobic monosaccharide or oligosaccharide and lipophilic aglycone moieties, they include a great diversity of chemical structures that, in turn, dictate the biological and chemical properties of the compound (Adler and Hiller, 1985; Bader and Hiller, 1987; Hiller and Voigt, 1977). There have been many investigations which have purported to show the biological effects of saponins and which have been rightly criticized for their lack of information about the nature, origin or purity of the saponins used (Malinow, 1984; Price et al., 1987). In addition, claims as to the behaviour of individual saponins in in vivo and in vitro assays have been conflicting. We have described the permeabilizing effect of a number of isolated saponins, using an in vitro assay procedure (Johnson et al., 1986). This work demonstrated that the monodesmosidic soyasaponins (i.e. those possessing a single carbohydrate chain) had a weak--but detectable--permeabilizing effect on the rat intestine, although this was much less than those produced by either ~-tomatine or a mixture of Gypsophila saponins. In subsequent studies the effects of Gypsophila and soya saponins have been examined in rive (Gee and Johnson, 1988; Southon et al., 1987 and 1988), and the mean daily intake of saponins in the UK diet has been determined (Ridout et al., 1968).

MATERIALS AND METHODS

Materials

The physiological studies were conducted using commercially available glycyrrhizin and ~t-tomatine (Sigma Chemical Co. Ltd, Poole, Dorset), and the glycoalkaloids at-solanine and ~t-chaconine, which were obtained from potato (Solanum tuberosum) sprouts as reported by Coxon et al. (1979). The isolation and purification of Gypsophila, alfalfa and clover saponins has been described by Southon et al. (1988) and Jurzysta et al. (1988). Crude Quillaja saponin was kindly supplied by Dr D. G. Oakenfull (Food Research Laboratory, CSIRO, Australia) and guar saponin was isolated by the method of Curl et al. (1986). Transmural potential difference measurements

Male Wistar rats (200-250 g) were fed a commercial pelleted non-purified diet (RM3 Expanded, Special Diet Services Ltd, Witham, Essex) and water ad lib. prior to sacrifice by an ip injection of sodium barbiturate (Euthetal, May & Baker Ltd, Dagenham, Essex) and cervical dislocation. The manufacturer's specification indicated that the diet was unlikely to contain saponins. The entire small intestine of the rat was removed, rinsed with Krebs bicarbonate Ringer solution, and the proximal 40-cm portion was everted. Following removal of the proximal 10% portion of the small intestine, three 5-cm segments of

Abbreviation: PD ffi potential difference. T.LV.

3/~-A

85

86

J, M. G~E eta/.

1

_

~

,

1

xyl (1-- 3) ~ glu (1-. 4) gal (1~) ....Q glu(1 --2) '~

R R :'

OH

(~- L- rhe (1 - - 2 ) ~ _ D_glll (1~)_

~-O-glu(1--3)

R =: (~- L- rha(l~ 2) ~ ~_D_glu(l~) - L - r ha (1 ~4 ) •

0

8 ~

0

"OR2

0,~ _4

RlOV ~

gluA(1--2)glAu

(1_) -'0

~'~

R~ - (x-L-rha(1--2)~_D_glu(l~2)_~_D_gluA(l~)_ -L -rha (1--4)

Rz"~-D-glu(1--2)-D-glu(1--)__5H O ~ .ooo

~)OH

v OH

e

H O ~ 0

OH

Fig. 1. Structures of saponins and glycoalkaloids. (1) ~-Tomatine. (2) ~-Solanine. (3) ~-Chaconine. (4) Olyvyrrhizin. (5) Medicagenic acid: aglycone of alfalfa saponins, (6) Oypsogcnic acid: aglycone of Gypsophila saponins. (7) Soyasapogenol B: aglycone of clover saponins. (8) Guar saponin. (9) Quillalc acid: aglycone of Quillaja saponins.

Effect of saponins on rat small intestine

87 11.1. (.v.)

jejunum were prepared. These were ligatured at one end, tied over glass cannulae, filled with Krebs Ringer solution and suspended in tubes filled with Krebs Ringer containing D-glucose (28 mM). The steady-state glucose transfer potential was monitored by means of silicone rubber salt bridges containing KCl/agar, which led by way of calomel half cells to a modified Keighley 177 digital multimeter (Keighley Electronics, Reading, Berks.). Once the tissue had achieved a stable potential, the cannulated sacs were transferred to either an identical control tube or to tubes containing solutions of the saponin in Krebs Ringer and glucose (28raM). Changes in the potential difference (PD) of the control and test sacs were monitored at intervals using a chart recorder. Allowance was made for any changes in PD of the control sac. The mean value for corrected change in PD (mV) at each time point (5 animals) was then plotted against time. Where a significant decline in PD occurred, the maximum rate of change was estimated by linear regression of the points lying on the straight portion of the curve.

lO

RESULTS

Responses of everted sacs to individual saponins and effects of concentration The response of the rat intestine to a variety of structurally-differing saponins and glycoalkaloids

Fig. 3. Change in rat small intestine transmural potential difference over time during exposure to saponins (corrected for variations in control) in the presence of 4mM: ~-Solanine (F'l); Quillaja saponin (0); ~-chaconine (C)) clover saponin (~7); guar saponin (A); alfalfa saponin if-l). Values are means + SEM for five observations.

I 2

4 ~m

6 8 (,m)

10

Fig. 2. Change in tat small intestine transmural potential difference over time during exposure to saponins (correcugt for variations in control) in the presence of 0.5n~: • -Solanine ([]); Qufllaja saponin (0); ~-chaconine (O); clover saponin (~7); guar mponin (A); alfalfa saponin ([3). Values are means :t: SEM for five observations.

~Fig. 1) at concentrations of 0.5 and 4mM are shown in Figs 2 & 3, respectively. The experiments were conducted at pH 7.2, with the exception of those with the glycoalkaloids ~,-solanine and ~-chaconine, in which solubility requirements dictated that the studies be carried out at pH 4.5. Reduction of the pH did not significantly affect the viability of the preparation within the timescale employed. It should be noted that there were large differences in the effects of the different saponins, both in terms of the shape and range of the response curves. Despite being structurally similar (differing only in the nature of their trisaccharide moieties) ~-chaconine and ~-solanine exhibited considerable differences in behaviour at the lower concentration, but behaved similarly at the higher concentration. The alfalfa saponin showed the greatest response of the saponins under test at both concentrations. In general, an increase in concentration led to an increased rate of response, but this was not true for the clover saponin, to which the responses were virtually negligible at both concentrations. The variation in the maximum rate of fall of PD in the presence of seven different saponins at the two molarities is illustrated in Fig. 4.

Effect of pH The responses to ~-tomatine, Gypsophila saponin and a chemically treated ('de-esterified') Gypsophila

88

J . M . GEE et al.

J_ C"

|

...t.

1

2

3

4

5

~

7

0.5ram

1

2

3

4

5

6

7

4mM

Fig, 4. Maximum rate of change in rat small intestine transmural potential difference in the presence of a range of saponins. (1) Clover saponin; (2) glycyrrhizin; (3) QuiUaja saponin; (4) guar saponin; (5) ~t-chaconine; (6) ~-solanine; (7) alfalfa saponin. Values are means + SEM for five observations. saponin at two different pHs are shown in Fig. 5. The largest difference, for a-tomatine, is probably attributable to the relative insolubility of this basic compound at the neutral pH. Gypsophila saponin contains both ether- and ester-linked glycosides, the latter being removed in the de-esterified product, which consequently possesses a free carboxylic-acid grouping. The differences observed between the original and modified compounds are thus a reflection of the removal of the ester-linked glycoside S

! Ji

"

'

°

altering the charge of the molecule. Recently Oleszek et al. (1988) have noted differences in the response of the fungus Trichoderma virides to a range of esterified and de-esterified alfalfa saponins. Effect of glycyrrhizin on Gypsophila saponin The protective action of the major saponin in liquorice root, glycyrrizin, against saponin-induced erythrocyte haemolysis (Segal et al., 1977), and the current interest in the anti-turnout (Tsuda and Okanoto, 1986) and anti-viral (Ito et al., 1987) properties of this compound prompted the examination of the effect of glycyrrhizin against Gypsophilainduced permeabilization of the rat small intestine. The effect of glycyrrhizin alone at concentrations of 4 and 10 mM (at pH 7.2) is shown in Fig. 6. The effect of adding glycyrrhizin to Gypsophila saponin is shown in Fig. 7. This experiment confirmed the potency of the latter saponin and showed a dosedependent protective effect associated with glycyrrhizin. DISCUSSION

0

0

1

2

3

4

S

6

?

cmm#r,~n (mU)

Fig. 5. The effect of pH and degree of esterification of saponins on the rate of change in rat small intestine transmural potential difference. ~.Tomatine pH 4.5 (&); ,,-tomatine pH 7.0 (&); Oypsophila saponin pH 4.5 (0); Gypsophila saponin pH 7.0 (O); de-~tefified Oypsophila saponin pH 4.5 (B); de-esterificd Gypsophila saponin pH 7.0 (I-1). All curves were drawn by eye. Values are means + SEM for five observations (omitted where smaller than symbol).

The experimental findings reported here clearly demonstrate that individual saponins elicit widelydiffering responses in the small gut of the rat and that these are significantly affected by pH, concentration, chemical form and the presence of other materials in the solution under analysis. This serves to emphasize the importance of clearly controlling (and describing) the experimental conditions and the nature and purity of the sample when carrying out biological investigations on saponins. The saponins included in this study embrace the general structures of most of those found in

Effect of saponins on rat small intestine

i 'l (~)

Fig. 6. Change in rat small intestine transmural potential difference over time in Krebs bicarbonate Ringer control (O); Krebs bicarbonate Ringer+ glycyrrhizic acid (4 raM) (0); Krebs bicarbonate Ringer + glycyrrhizicacid (10 rn~t) (A). Values are means + SEM for five observations. foodstuffs (Price et al., 1987), with the exception of those possessing a steroidal agiycone. From evidence to date it is clear that the chemical structure of a saponin plays a significant part in determining the nature, and extent, of the response of the gut. The present studies confirm the preliminary findings that the basic glycoalkaloids in potato and tomato and the complex bisdesmosides from Gypsophila, Quillaja and alfalfa are most potent, whilst the soyasaponins show only weak activity. In this context, recent work (Ridout et al., 1988) has shown soyasaponins to be the predominant type of saponin consumed in the

(mE) Fig. 7, Change in rat small intestine transmural potential difference over time in Krebs bicarbonate Ringer control (O); Gypsophila saponin (lmM) (A); Gypsophila saponin+glycyrrhizic acid (20m~t) (&); Oypsophila saponin + glycyrrhizicacid (40 raM) (0). Values are means + SEM for five observations.

89

UK (mean daily intake of tiae population as a whole about 15 rag/person, arising from baked beans, lentils and peas; mean daily intake of vegetarian group, 110 mg/person, mainly from soya, baked beans and lentils). However, in ethnic groups the importance of saponins with a more potent permeabilizing activity increases; thus the mean daily intake of saponins in a group of UK male Asian vegetarians was 214mg/person, the majority of which was derived from the guar bean (Cyamposis tetragonoloba), the saponins of which have a considerably greater permeabilizing activity than those of soya, as the present work has shown. Mention must also be made of the effect of the glycoalkaloids, since although their intake is generally considered to be low, current trends towards the consumption of potato skins (as jacket or new boiled potatoes or 'healthy' potato snacks) may result in particular groups of the overall population consuming large amounts of these very active compounds. Potato glycoalkaloids and potato sprouts fed to hamsters have been shown to produce severe gastric and mucosal necrosis (Baker and Keeler, 1987). In previous studies it has been shown that the reduction in transmural PD induced by saponins is associated with increased uptake of both passively permeable sugars and larger compounds such as polyethylene glycol (Johnson et al., 1986). At the same time, the ability of the mucosa to accumulate actively transported organic species is lost. It is probable that these changes result from diminished integrity of the brush-border membrane, brought about by an interaction between saponins and membrane sterols. Erythrocyte ghosts treated with haemolytic saponins acquire ring-shaped ultrastructural lesions. Freezefracture studies suggest the presence of a central aqueous pore about 80 A in diameter. According to the model of Seeman (1974), the hydrophobic moieties of the saponin molecules combined with membrane cholesterol to form the perimeter of a stable ring-shaped structure in the plane of the membrane. However, a slightly different structural scheme has been proposed by Tagaki et al. (1982). The effects of saponins on the intestinal mucosa demonstrated in vitro are probably more extreme than those that will occur in the intact intestine. The biological consequences of ingesting membranolytic saponins are therefore not yet clear. Earlier authors were aware that saponins could cause 'corrosion' of the intestinal mucosa that would lead to enhanced uptake and hence cause systemic toxicity (Ewart, 1931). Recent studies have shown that saponins can increase the uptake of rabies vaccine by mice, and thereby enhance the immunological response (Maharaj et aL, 1986). This effect may also be of practical importance in relation to allergens in foods. Lower concentrations of saponins may cause increased mucosal-cell exfoliation, but the normal processes of cell replacement probably have the capacity to compensate for such losses without significant damage to the villi (Johnson et al., 1986). It is interesting to speculate, however, that differences in mucosal morphology seen in apparently healthy human beings from northern European and tropical environments (Creamer, 1974) might be due to

J. M. GEE et al.

90

differences in their dietary intake of membranolytic food components. The use of diets rich in saponins has been advocated as a natural means o f reducing plasma cholesterol and hence gaining protection against atherosclerosis (Price et al., 1987). Liquorice, fenugreek, ginseng and quinoa, all rich in saponins, are being increasingly marketed as health foods and tonics. However, the above effects and others such as the well-documented pseudoaldosteronism associated with liquorice ingestion or giycyrrhizin medication (Stewart et al., 1987) indicate that such products should be treated with caution. Acknowledgements--The authors are grateful to Dr D. B. Oakenfull for the Quillaja saponin, Dr M Jurzysta for the alfalfa and clover saponins and to the Ministry of Agriculture, Fisheries and Food for funding this work. They also wish to thank Miss J. C. Brown for technical assistance and Mrs J. Cooke and Mr P. Muddel for care of the animals.

REFERENCES

Adler C. and Hiller K. (1985). Bisdesmosidische Triterpensaponine. Pharmazie 40, 676-693. Bader G. and Hiller K. (1987). Neue Ergebnisse zur Struktur und Wirkungsweise yon Triterpensaponinen, Pharmaz/e 47, 577-597. Baker D. and Keeler R. (1987), Lesions of potato sprout and extracted potato sprout alkaloid toxicity in Syrian hamsters. CIin. Toxicoi. 25, 199-208. Coxon D. T., Price K. R. and Jones P. G. (1979). A simplified method for the determination of total giycoalkaloids in potato tubers. J. Sci. FdAgric. 30, 1043-1049. Creamer B. (Editor) (1974). In The Smalllntestine. pp. 1-23. Heinemann, London. Curl C. L., Price K. R. and Fenwick G. R. (1986). Isolation and structural elucidation of a triterpenoid saponin from Guar, Cyamopsis tetragonoloba. Phytochemistry 25, 2675-=2676. Ewart A. J. (1931). The poisonous action of ingested saponins. Australian Council for Scientific and Industrial Research, Bulletin No. 50. Gee J. M. and Johnson I. T. (1988). Interactions between haemolytic saponins, bile salts and small intestinal mucosa in the rat. J. Nutr. In press. Hiller K. and Voigt G. (1977). Neue Ergebnisse in der Erforschung der Triterpensaponine. Pharmazie 32, 365-393. Ito M., Nakashima H., Baba M., Pauwels R., De Clercq E., Shigeta S. and Yamamoto N. (1987). Inhibitory effect of giycyrrhizin on the /n vitro infectivity and cytopathic

activity of the human immunodeficiency virus [HIV (HTLV-III/LAV)]. Antiviral Res. 7, 127-137. Johnson I. T., Gee J. M., Price K. R., Curl C. L. and Fenwick G. R. (1986). Influence of saponins on gut permeability and active nutrient permeabifity /n vitro. J. Nutr. 116, 2270-2277. Jurzysta M., Price K. R., Ridout C. L. and Fenwick G. R. (1988). The structures of four triterpenoid saponins isolated from the seed of Trifolium incarnatum. Planta meal. In press. Maharaj I., Froh K. J. and Campbell J. B. (1986). Immune response of mice to inactivated rabies vaccine administered orally: potentiation by Quillaja saponin. Can. J. Microbiol. 32, 414--420. Malinow M. R. (1984). Saponius and cholesterol metabolism. Atherosclerosis 50, 117-I 18. Oleszek W., Price K. R. and Fenwick G. R. (1988). The sensitivity of Trichoderma viride to medicagenic acid, its natural giucosides (saponins) and derivatives. Acta bot. po/. In press. Price K. R., Johnson I. T. and Fenwick G. R. (1987). The chemistry and biological significance of saponins in foods and feedingstuffs. CRC Crit. Rev. Fd Sci. Nutr. 26, 27-135. Ridout C. L., Price K. R., Wharf B., Johnson I. T. and Fenwick G. R. (1988). Preliminary communication--UK mean daily intake of saponins, intestine-permeabolising factors in legumes. Hum. Nutr. Fd Sci. Nutr. 42F, 111-116. Sceman P. (1974). Ultrastructure of membrane lesions in immune lysis, osmotic lysis and drug-induced lysis. Fedn Proc. Fedn Am. Socs exp. Biol. 33, 2116-2124. Segai R., Milo-Goldzweig I., Kaplan G. and Weisenberg E. (1977). The protective action of glycyrrhizin against saponin toxicity. Biochem. Pharmac. 26, 643-645. Southon S., Johnson I. T., Gee J. M. and Price K. R. (1987). The effect of Gypsophila saponins in the diet on mineral status and plasma cholesterol concentration in the rat. Br. J. Nutr. 59, 49-55. Southon S., Wright A. J. A., Price K. R., Fairweather-Tait S. J. and Fenwick G. R. (1988). The effect of three types of saponin on iron and zinc absorption from a single meal in the rat. Br. J. Nutr. In press. Stewart P. M., Wallace A. M., Valentino R , Butt D., Shackleton C. H. L. and Edwards C. R. W. (1987). Mineralocorticoid activity of liquorice: 1l-beta-hydroxysteroid dehydrogenase deficiency comes of age. Lancet ii, 821-824. Tagaki S., Otsuka H., Akiyama T. and Sankawa U. (1982). Digitonin-cholesterol complex formation: effects of varying the length of the side-chain. Chem. Pharm. Bull., Tokyo 30, 3485-3492. Tsuda H. and Okamoto H. (1986). Elimination of metabolic cooperation by giycyrrhetinic acid, an antitumor promotor, in cultured Chinese hamster cells. Carcinogenesis 7, 1805-1807.