SUGARBEET SAPONINS AND ACID BEVERAGE FLOC

Earl J. R oberts, M argaret A. Clarke* and Mary An Godshall, Sugar Processing Research Institute, Inc., 1100 R obert E . Lee Blvd., New Orleans, LA, USA 70124

ABSTRACT

Saponin from sugarbeet has been reported, for many years, as the causative agent of acid beverage floc. Recent work on isolation procedures for saponins from sugarbeet have shown that some procedures extract only a part of saponin (oleanolic acid) and not the whole molecule. Further work on comparison with isolation of complete saponin is reported. Effects of saponin and oleanolic acid alone and in combination with other possible floc-causing compounds, on appearance of floc in beverages are reported. Possibilit ies for causes of acid beverage floc are discussed.

INTRODUCTION

Saponins

Saponins are a class of compounds widely distributed in the plant kingdom in legumes, roots, shrubs and bushes, in varying degrees of concentration. Various saponins have been used as soaps because of their surface active properties - hence their name. Saponin-containing plants, or their extracts, have been used in herbal medicine, in treatment of various complaints including liver and cholesterol related diseases (2), and as anti-fungal agents (3).

Saponins fall into two classes: triterpene-based and steroid-based. To the base aglycone are attached sugar group(s) and glucuronic acid, which define the molecule as a saponin.

Sugarbeet (Beta vulgaris) is known to contain at least three triterpene-based saponin structures, all glucuronic acid glycosides of oleanolic acid (4), as shown in Figure 1.

(2) COOH 0

o l!O~O OI l

OH

HO~OH H I ,O~O~OH

o

HO~-~)t(O~H 0, ><~7-_.I/o" ~o~ Ot<

(J)

Figure 1. Three saponins of sugarbeet.

* Author to whom correspondence should be addressed .

39

Two additional compounds, seco-glycosides of saponins, isolated fro m beet leaves and roots, have recently been identified (5). Up to six beet saponins have been postulated.

Saponins are reported to be found in sugarbeet at levels of 0.01% to 0.2% of beet (3,6,7, 8), and at less than 100 ppm, generally less then 20 ppm, in white sugar. Saponins are most densely concentrated just under the sugarbeet skin, where they function as plant defense compounds, against disease and against frost damage, and are located in cell membranes (3). They are most highly concentrated in small beets grown in warm climates.

Recent work by these authors (1) comparing isolation systems and their products indicates that material reported as "saponin" in sugars and process streams may in fact be oleanolic acid. However, since oleanolic acid is derived from saponins, their presence is still indicated.

Acid beverage floc

Acid beverage fl oc, which can form in sugar-sweetened carbonated soft drinks after several days standing, has been ascribed to both beet and cane sugars. In general, any haze or turbidity in a soft drink is referred to as "floc", but there are specific characteristics which define acid beverage floc, most notably that it disappears upon shaking. Beet and cane floes can appear as turbidity or as "cotton ball floc" . Beet sugar floc can be more granular in appearance and less fluffy than cane sugar floc, in our experience. Beet sugar floc has, for many years, been ascribed to saponins (7, 8, 9), but in our tests, the addition of amounts of isolated saponin added at levels resembling those in sugar to trial sugar floc tests do not necessarily produce fl oc. The literature supports this: Eis (9) says that "separated floc" (by which was meant all material precipitable at pH2) can "produce effervescence and flocculation when suffi cient neutral solution of the floc is added to carbonated beverages". In the authors' experience, "sufficient" is far above the levels of saponin reported in white sugars « 1 to 30 ppm). "Sufficient" levels are above several hundred ppm. I t is therefore of interest to isolate sugarbeet saponins for further study of their effect on acid beverage floc formation: the evidence for the responsibility of sugarbeet saponins for this phenomenon may be circumstantial.

Acid beverage floc from cane sugars has been found, in the general case, to have two causative factors : a polysaccharide with glucuronic acid groups and a protein. (There is least one specific regional acid beverage flocs caused by specific regional infection.) In the general case, the polysaccharide is derived from plant cell wall material. The protein may come from the cane plant, or may be residual from enzyme addition. The glucuronic acid and the primary amine residues become oppositely charged at beverage pH, and, through charge attraction, come together to form a coacervate as the basis for a floc network. Suspended solids in the solutions (e.g. colloidal material) and high molecular weight solubles (e.g. dextran, starch) can come out of solution and enhance the appearance of a floc that has already formed .

METHODS AND MATERIALS

EXTRACTION OF SAPONINS

Beet peelings, obtained from fresh sugarbeet in the S.P.R.!. labs, were subjected to several methods ofextraction. The resulting extracts were compared by thin layer chromatography and by GC-MS .

40

1. Method ofRother, aqueous extraction (10).

Fresh beet peelings (5 .5 kg) were covered with water in a blender and divided into small pieces. The slurry was heated to 90°, and filtered on fabric. The residue was suspended in water, heated, and filtered again on fabric. The pH ofthe filtrate was adjusted to 1.5 with HCl, heated to 90°C for one hour, and allowed to settle overnight. After settling, the supernatant liquid was decanted. The residue was mixed with filter aid and filtered; that residue was washed with water, adjusted to pH 1.5 with HC 1, and allowed to air dry. The dried residue was extracted in a Soxhlet extractor with ethanol, the ethanol solution concentrated, and poured into water at pH 1.5 . The precipitate was dissolved in hot ethanol and again precipitated· by pouring into pH 1.5 water. The precipitate was filtered off on hardened paper, dissolved in water, and evaporated to dryness at low temperature; yield 3.0 g of brown material.

Analysis of this material by TLC (as described below) showed oleanolic acid (the aglycone, or sapogenin) and nothing corresponding to saponins. Mass spectroscopy analysis confirmed the presence of oleanolic acid. Apparently the harsh acidic treatment hydrolyzed the saponins, leaving only oleanolic acid in the isolation.

2. Method of Ridout et aI, aqueous extraction (4).

On another experiment 1734 g ofbeet peel was ground in a blender. The slurry was filtered on fabric and the residue mashed with water. The pH of the filtrate was adjusted to 1.5 with HC 1, heated to 85°C for 15 minutes, cooled overnight, and filtered on fabric coated with filter aid. The filtrate was returned to the filter twice more and the residue was washed with warm IN HC 1. All filtrates were discarded. The filter was then washed with warm 2N NaOH solution until the filt rate was clear. The filtrate was placed in a large beaker and HC 1 was added to reduce the pH to 1.5 . The precipitate was collected on fabric coated with filter aid as before, washed with IN HCI , and the filtrate discarded. The filter was then washed with warm 2N NaOH. The filtrate was acidified to pH 1.5 with HC1, filtered through Whatman 542 paper, washed with water, and extracted with 500 ml ofwarm ethanol. The filtrate was evaporated to dryness, then taken up in water and freeze dried, yielding 2.0 g of brown material. TLC analysis showed oleanolic acid but no saponin.

3. Method ofRidout et ai, methanol extraction (4).

Freeze dried beet peel (650 g) was crumbled into small pieces and extracted in a Soxhlet extractor with methanol. The methanol was evaporated under reduced pressure, residue dissolved in water, and extracted severaJ times with I-butanol .

The butanol was evaporated and the residue was dissolved in water and dialyzed against flowing tap water in a 12,000 MW cut off bag for 24 hours. The material remaining in the bag was filtered, concentrated, and freeze dried. Yield, 5.6 g of cream colored material. This materiaJ was subjected to thin-layer chromatography and mass spectrometry, as below; it appeared to contain saponins.

41

kaoI , ZUIl, . ' J1 ........ ,: OlOoNOlC.O

l~UOOO ! )0)

)00000

)00000

I

M:l$S spectrum or authenlic ole.lnolic .lcid ThIS. Mateh qU.lIHy with bed ped extr.lct it 99-/_.

$U., 1230 12 1.0i4 minI: SAPONHI.D

" 1

000000

..cooo II

.00000 "0000

=1

)0)

4COO

3500

'000

2500

:=1 "I' 20001500 150000 • 320

'" 1000 '00000 I

550000 1 _ 15 .1 , ,\ 11•7 79 500

oj e27o L..!.Jll..l ' t., 25s1 153 lUll .81, 510 558 585 o"",-" 50 · 100 15Q loo 250 100 350 400 .50 600 550 eco

mll- >

M.l11 sp~clUm sl10winc OIe:r.DOlic ::acid in .qu~ous tltract or bed peel.

Sci" 122! (21.072 min): 580.71.0

7J I

129

204

~ S8 7 17 .

' Il4 I 255 301 378417449482 617· J~ . ':' , I I II 51'558 '16J'~Jl~~ . I I d '. ,[ k ' ,f r\ I '~l t!" ,l1.1 d 1" ' 01,, _

50 100 150 200 250 300 350 400 ,450 500 550 600 650

Mus spectrum or methlnolic eltnct or beet pe~~ sbowio&: Ibsence oC ole.nolic lcid.

Figure 2. Mass spectrometric analysis of sugarbeet extracts.

Thin layer chromatographX-Qf isolates from aqueous extraction.

Isolates prepared by the traditional aqueous extraction methods (1 and 2), with repeated extractions at pH 1.5 and washing with base, showed only oleanolic acid in the final dried extract, and no saponins. Oleanolic acid identification on thin layer chromatography (solvent system: chloroform; methanol; water 65:35:10), made visible by 2N H2S04, or anisaldehyde spray, was confirmed by gas chromatography-mass spectrometry identification, as shown in Figure 2.

Method 1, of Rother (9), using aqueous extraction and low pH, yielded 3g (0.05% on beet peel) brown solids; Method 2, of Ridout et al (5), yielded 2 g (0.12% on beet peel), of brown material. Method 3 ofRidout et al (4), similar to that ofIreland (2) using methanol extraction and not including low pH treatment, yielded 5.6 g (0.8% on beet peel) of cream colored material. Thin layer chromatography of the methanol extracted material showed five major components, two of which traveled with an authentic saponin (probably soybean) obtained from Sigma Chemical Co. It should be noted that "saponin extracts" supplied to S.P.R.I., Inc., by several sugar companies (sponsoring companies of S.P.R.I., Inc.) appeared to consist mainly of oleanolic acid.

42

- - --- ------

Gas chromatography-mass spectrometry (GC-MS)

Samples containing oleanolic acid were converted to the trimethylsilyl derivative (TMS) using Pierce Tri-Sil® in pyridine solution. Gas chromatography (GC) was performed on a Hewlett Packard 5890. GC conditions were: 250 0 C for 10 min; increase temperature 50 C per min to 310 0 C for 10 min. The column was a 30 m x 0.25 mm fused silica with 0.25,u film thickness of 5% phenyl methyl silicone. Oleanolic acid eluted at 21.1 1 minutes under these conditions. Mass spectrometry (MS) was conducted with a Hewlett Packard 5972 mass selective detector ..

Charged species at beverage pH

Moving boundary electrophoresis on oleanolic acid was conducted in sucrose solution at pH3, adjusted with phosphoric acid. Oleanolic acid moved towards the anode. The oleanolic acid is thereby shown to have a negative charge at pH3 in sucrose solution (simulating acid carbonated beverage conditions).

RESULTS AND DISCUSSION

COMPOSITION AND STRUCTURE OF SAPONIN BEET ISOLATES

Saponins are known to exist in variety in anyone plant - a single structure is not common. Variations in the sugar moiety structure and linkage position are observed. Sugarbeet saponins are no exception. The three forms shown in Figure 1 all have as their base unit oleanolic acid, a carboxylic acid triterpene.

It is evident from comparison of the aqueous methods of extraction with the methanolic method that the sugarbeet saponins are indeed present in the whole root, in the peel, just under the skin. It is also evident, from chromatographic and mass spectroscopic data, that the saponins extracted by aqueous methods have been hydrolyzed by the strong acid treatment so that only the aglycone (or sapogenin), oleanolic acid, remains.

This observation throws some doubt on earlier work, all of which isolated saponins by aqueous extraction with strong acid treatment. Were these earlier results the product of oleanolic acid only? Earlier workers did not have the benefits ofGC-MS, but had to rely on colorimetric tests, which may give a false positive for saponins when oleanolic acid is present.

'Recent work (4) found saponins by aqueous extraction, not in extracts of beet roots, but only in extracts from beet molasses, where the compounds may be expected to concentrate. Subsequent investigations by the same workers (5) found saponins in methanol extracts of sugarbeet roots and leaves.

BEVERAGE FLOC

Floc tests (50 Brix, phosphoric acid to pH 2) were run on non-floccing sugars with the addition of varying amounts ofbeet extract, Of commercial saponin (not from beet), or oleanolic acid . Saponin

43

and oleanolic acid were also tried in combination with protein (gelatin and ex-amylase were used) . The methanolic extract of beet root formed a floc, as did the combination of oleanolic acid and protein.

The observation that saponins apparently are hydrolyzed during the acid extraction raises a basic question about reactions in and causes of floc formation. The assumption was that floc material was acid insoluble, and therefore the aqueous extraction method at low pH was developed. But do saponins themselves, if in white sugar and therefore in beverage, become hydrolyzed at beverage pH (about 2 - 2.5)? In that case oleanoljc acid and not saponin would be responsible for floc fo rmation.

In studies reported here, the only floc former (floc former when used alone, without other reagent) from beet was the methanolic extract, i.e., the extract that contains unhydrolyzed saponins. So it would appear that whole saponin - or at least whole when it enters the beverage - does form beverage floc . This methanolic extract contained many other compounds from beet in addition to saponins.

The obselVations on hydrolysis explains why Eis (9) found it necessary to add back a relatively large quantity of isolated floc material before obselVing flocculation; probably sufficient was added back to form a haze rather than a true floc. The isolated floc material here appears to have been largely oleanolic acid, not saponins.

The authors have, in past work ( 11 , 12), obselVed that isolated beet sugar floc (from beverage), identified as beet sugar-sourced by the high level of raffinose present, contained beet cell wall polysaccharide with galacturonic acid residues, and protein. The polysaccharide (given the trival name Indigenous Beet Polysaccharide, IBP) (11 , 12) is comparable to the cell wall sugarcane polysaccharide, containing glucuroruc acid groups, that can cause acid beverage fl oc when in combination with a protein. At beverage pH, the acid groups become negatively charged, the protein groups become positively charged; charge attraction brings the molecules together to form fi rst a coacelVate and then a flocculating network that entraps colloidal and suspended material to form a visible floc .

The authors propose that a similar mechanism can be responsible for beet sugar floc : a carboxylic-acid containing molecule, saponin, oleanolic acid, or cell wall polysaccharide, becomes negatively charged (carboxylate ion formation) at low pH; an amino-group containing molecule (protein, peptide, or other) becomes positively charged. The two come together from charge attraction to initiate a floc network. This explanation accounts for the obselVance of floc without saponin present because another negatively charged molecule can be participating (perhaps oleanolic acid) . Both saponin and oleanolic acid contain a glucuronic acid group. Beet cell wall polysaccharides contain galacturonic acid groups. This mechanism accounts for the presence of saponin without floc, if insufficient protein or positively charged amino group is present. It also accounts for the presence of floc without saponin, if oleanolic acid or IBP is forming a coacelVate with protein.

SAPONIN TESTS

All the postulates and observations in the literature involve validity of saponin tests. The traditional tests must be re-evaluated using instrumental analysis to distinguish between saponin and oleanolic

44

acid. Literature discussion on the most frequently used test, the antimony pentachloride test, points out that its color reaction is not specific for triterpenes (7). It is obviously then not specific for saponin versus oleanolic acid.

CONCLUSIONS

Comparison of traditional aqueous acid extracts of "saponin" from sugarbeet substrates with methanol extraction has shown that aqueous extraction yields only the aglycone of beet saponins, oleanolic acid. Chromatographic and mass spectrometric evidence support this.

A re-examination of the relationship between saponin and oleanolic acid and "saponin" - caused problems, acid beverage floc and foaming, is required.

It is proposed that floc formation has two causative factors: presence of a molecule that is negatively charged at beverage pH (saponin, oleanolic acid, or cell wall polysaccharide) and a molecule that is positively charged at low pH (protein or peptide); the two molecules come together in solution through charge attraction to form a coacervate that develops into a floc network.

REFERENCES

1. Roberts, E. J., M . A Clarke, M. A Godshall and L. A Edye. (1996). Saponins from sugarbeet. P roc. Conf. Sugar process. Res. pp. 492-500.

2. Ire]and, P . A, S. Z. Dziedzic and M. W. Kearsley. (1986). Saponin content of soya and some commercial soya products by means of high performance liquid chromatography of the sapogenins. J. Sci. Food Agric. 37: 694-698.

3. Hallanoro, H ., J. Ahvenainen and L. Ramm-Schmidt. (1990). Saponin, a cause offoaming problems in beet sugar production and use. Proc. Conf. Sugar Proc. Res. , pp. 174-203.

4. Ridout, C. L., K. R . Price, G. Parkin, M . G. D ijoux and C. Lavaud. (1994). Saponins from sugarbeet and the floc problem. J. Agric. Food Chern. 42: 273-282.

5. Massiot, G. , M. G. Dijoux, C. Levaud, L. LeMen-Olivier, J. D . Connolly and D . M. Sheeley. (1994). ~-glycosides ofoleanolic acid from Beta vulgariS. Phytochem. 37: 1667-1670.

6. Scruweck, H, G. Steinle and E. Fischer. (1991). Determination ofsapogenins in sugar factory products and processes. Proc. c.I.T. S., pp. 441-455.

7. Van der Poel, P . W., M. L. A Verhaart and N . H M . de Visser. (1 966). The course of the non­sugars from thick juice to white sugar. Int'l Sugar Journal, 66: 317-319: 349-352.

8. Carruthers, A, J. F . T. Oldfi eld, J. V. Dutton, L. C. Manning and M . Shore (1 961). Juice components in relation to sugar quality. British Sugar pIc. Internal report .

45

9. Eis, F. G., L. W. Clark, R. A McGinnis and P . W. Alston. (1952) . Floc in carbonated beverages. Ind. and Eng. Chemistry, pp. 2844-2848 .

10. Rother, H. (1962). Saponin precipitations in sugar sweetened, clear soft drinks. Zucker, pp. 196-199.

11. Roberts, E. 1. and M. A Clarke. (1991). Internal Reports, Sugar Processing Research Institute, Inc.

12. Clarke, M. A, 1. M. de Bruijn, E. 1. Roberts, M. A Godshall, R. S. Blanco and X. M. Miranda. (1992). Polysaccharides of beet and cane sugar: A progress report. Proc. Conf Sugar Proc. Res. pp.353-37l.

ACKNOWLEDGEMENTS

Grateful thanks are expressed to the sponsoring companies of Sugar Processing Research Institute, Inc. who supplied "saponin" extracts.

46