d4nalyst, December, 1973, Vol. 98, pp. 857-862 857 A Spectrofluorimetric Procedure for the Assay of Carrageenan BY R. B. CUNDALL, G. 0. PHILLIPS AND D. P. ROWLANDS (Department of Chemistry, Uniuersity of Nottingham, Nottingham, NG7 2RD) (Defiartment of Chemistry and Applied Chemistry, University of Salford, Salford, 1%?5 4 WT) A spectrofluorimetric procedure for the determination of carrageenan in the presence of salts, many polysaccharides, and proteins has been de- veloped. It involves the binding of acridine orange to the polyanion and observation of the decrease in green fluorescence from free monomer dye in solution. The method offers many advantages over the available chemical assays, and is a simple, accurate and rapid technique for determining carrageenan in a conventional hydrocolloid stabiliser mixture.Its successful application, after minor modifications, in the presence of protein and salts indicates its direct applicability to commercial milk products. CARRAGEENAN is the broad designation given to closely similar sulphated polysaccharides derived from certain types of seaweeds. This material, which is widely used as a stabiliser in ice cream and allied frozen milk products, is generally added in concentrations between 50p.p.m. and 1 per cent. Its quantitative determination in the presence of other carbo- hydrates (which assist the carrageenan in improving texture) and proteins is important, and there is a need for procedures for the rapid quality control of commercial samples. Several methods are already available for the quantitative determination of carra- geenan ,1-3 most of which involve removal of interfering cations and isolation of the carrageenan as an insoluble salt. These methods are tedious and time consuming.Stone, Childers and Bradley4 have determined the ester sulphate content of various plant polysaccharides by the interaction of anionic sites with cationic dyes. On becoming bound to polyelectrolytes the dyes may show a shift in their absorption spectra to shorter wavelengths ; this phenomenon, known as metachromasia, is due mainly to dye aggregation on the polymer molecules, although other factors also influence the intera~tion.5,~ Anionic groups in several polyanions7s8 can be accurately titrated on this basis. When bound, some cationic dyes, for example acridine orange or proflavine, either become non-fluorescent or show a fluorescence emission that is shifted to longer wavelengths and can easily be separated from the fluorescence of the unbound dye.The relative binding affinities towards a cationic dye with respect to the type of anionic site present increases in the order C02- < C02- + 0.S03- < C02- + 0.S03- + NHS03- < 0.S03-.9p1@ For the sulphated polysaccharide carrageenan, concentrations of metallic ions up to 1 M do not interfere with dye binding. A suitable dye for use in the determination of carrageenan is acridine orange. It binds strongly to polyanionsll and has been used extensively in histological staining and studies of metachromasia. The free dye is strongly fluorescent and on binding to anionic palysac- charides the green fluorescence is shifted to the red.It is the fluorescence of the unbound dye (Abr. 530nm) which is used in the following assay method. Metal ions can release the bound dye by competing for anionic sites. EXPERIMENTAL APPARATUS- Fluorimetric measurements were made with either an Aminco-Bowman instrument or a fully corrected spectrofluorimeter constructed at Nottingham University.12 The right-angle viewing mode was used with both instruments. @ SAC and the authors.858 CUND-ILL, PHILLIPS AND ROWL-Ih'DS -4 SPECTROFLUORIMETRIC [-4ilndyS/, 1'01. 98 REAGENTS- A specimen of A-carrageenan was also supplied by Dr. D. A. Rees, Unilever Ltd. . Other stabilisers used were gum arabic powder (BDH Chemicals Ltd.), guar gum MM (Meers Corporation), locust bean gum (Meers Corporation), pectinate as apple pectin, 250 grade (BDH Chemicals Ltd.), gum tragacanth (Griffin and George Ltd.) and the sodium salt of carboxymethylcellulose (Hercules Inc.).Acridine orange was purified by the method of Lamm and Ne~ille.1~ The extinction coefficient at a given concentration was calculated from the formula given by Stor,e and Bradley.s De-ionised water was used and all other reagents were of the purest grade obtain- able. Enzymes used were trypsin (Seravac) , subtilisin Carlsberg (Novo Industre, Copenhagen) and a-chymotrypsin (Seravac). A-, K- and L-Carrageenan samples were supplied by Copenhagen Pectin Co. METHOD Solutions of the different carrageenan-containing samples were prepared by dissolving amounts of sample that gave about 50 mg of carrageenan per 50 ml of solution.The standard solutions were stable for up to 3 days when stored a t 5 "C. Determinations were made by two different procedures. In one, 1 ml of diluted ( x 10) polyanion stock solution was added t o 5 ml of 1.54 x lo-* M solution of acridine orange and the mixture was immediately diluted to 50 ml and gently shaken. No precipitation occurred and a series of solutions was prepared by using increasing amounts of polyanion. In another procedure carrageenan solution was added in 0.01-ml portions to standard dye solution by using an Agla micrometer syringe. Both methods gave reproducible and consistent results. The absorption spectrum was measured in the range 400 to 500 nm and the fluorescent emission intensity was determined relative to a sample of unbound dye.In the fluorescence measurements the dye was excited on the short wavelength edge of the monomer absorption at 400 nm in order to achieve a low optical density, of about 0.06. Under these conditions, with right-angle viewing, the intensity of fluorescence is directly related to the concentration of free monomer dye. Fluorescence is viewed at 540 nm, rather than at the maximum inten- sity of 530 nm, so as to minimise inner filter effects from self-absorption of the fluorescence. For the experiments on the effect of the presence of other stabilisers or additives, the additive was added to the carrageenan before mixing the latter with the acridine orange solution. Bovine milk - carrageenan samples (5 g) were digested with 5 mg of trypsin, 5 mg of a-chymotrypsin and 5 mg of subtilisin Carlsberg at 27 "C and pH 6.5 to 7 for at least 6 hours before titration.Carboxymethylcellulose also binds acridine orange but less strongly than carrageenan. In the presence of carboxymethylcellulose determinations of carrageenan could be made in acidic solutions in which ionisation of the carboxyl group is suppressed,2 the optimum pH for this purpose being between 3.0 and 3.5. RESULTS Fig. 1 shows that increasing amounts of carrageenan decrease the absorption of mono- iiieric dye at 492 nm and a new band due to bound dye aggregate appears at about 462 nm. As the polymer concentration increases to a 1 : 1 equivalence of anionic sites to dye the spectrum becomes principally due to bound dye.It is, therefore, possible to plot the decrease in monomer absorption against carrageenan concentration and obtain a curve that indicates the number of sites on the polysaccharide. Sharper end-points were obtained by using spectrofluorimetric data, and typical results are shown in Fig. 2. The end-point is accurately determined by locating the point of inter- section of the two linear sections of the curve. The second section of the curves corresponds l o the concentration of dye that remains free in solution in the presence of excess of polyanion and is usually less than 5 per cent. of the total dye present. The effect of addition of calcium ions on the procedure for the determination is shown in Fig. 3. End-points were satisfactory when the polyanion was present in up to 1 0 - l ~ calcium chloride solution.High concentrations of calcium chloride removed increasing amounts of acridine orange from the carrageenan.December, 19731 PROCEDURE; FOR THE ASSAY OF CARRAGEENAN 0.8 0.6 Q) 8 m e a 0.4 0.2 0 400 420 440 460 480 500 520 540 Wavelength/nm Fig. 1. The effect of K-carrageenan on the absorption spectrum of acridjne orange. Dye concentration 1.55 x M. 1, No additive; 2, 75; 3,150;4,225; 5, 300; and 6, 375 pg in 50 ml 859 The different samples of carrageenan were titrated in the presence of six-fold excesses of five other commonly used stabilisers and a twelve-fold excess of carboxymethylcellulose. The results obtained are shown in Table I as the equivalent mass (relative molecular mass per unit dye-binding site) of the carrageenan sample.The proportions of additive used are only typical values, for example carboxymethylcellulose can be added up to a 200-fold excess and the carrageenan still successfully determined. Only one of the additives, gum arabic, decreased the sharpness of the end-point. Preliminary dialysis gave improved results with Carrageenan solution addedlml Fig. 2. Spectrofluorimetric titration of &-carrageenan - locust bean gum (1 : 6 wz/m), acridine orange 1-45 x M. Graph of de- crease in dye fluorescence intensity with in- creasing c-carrageenan 0 1.0 2.0 Ratio, site dye Fig. 3. The effect of calcium ions on the spectrofluorimetric titration of &-carrageenan with acridine orange, 0, 1 0 - a ~ CaC1,; 0, 1 0 - 1 ~ CaCl,; and x, 1 M CaCl,860 CUKDALL, PHILLIPS AND KOWLA4K11S SPHCTROPLCOKIILIETRIC [i!!na~ySt, VOl. 98 all of the carrageenan samples.In all instances the recovery is 100 per cent. within the limits of experimental error. TABLE I EQUIVilLENT MASSES OF VARIOUS CARRAGEENAN SAMPLES WITH EXCE. <ss OF ADDITIVES AS DETERMINED BY ACRIDINE ORANGE TITRATION Carrageenan solution, 0.05 per cent. m/m Additive Locust bean . . . . . . Pectinate . . . . . . Guar gum . . . . . . Gum arabic (dialysed) . . Gum tragacanth . . . . Mixure of above 5 additives Carboxymethylcellulose . . Bovine milk (enzyme digested) Mass of additive Mass of carrageenan 6 6 6 6 6 25 12 - Equivalent mass of carrageenan sample+ 7- Lambda 264 (1) 267 (1) 266 (1) 269 (2) 265 (1) 280 (6) 269 (2) - A Kappa 390 (1) 392 (1) 388 (1) 400 (2.5) 392 (1) 407 (4) 395 (1.5) __ -1 Iota 285 (2) 288 (1) 292 (1) 292 (1) 302 (4) 293 (1) 303 (4) 294 (2) * Figures given in parentheses are the mean percentage deviations of the values determined Fig.4 shows the improvement in the determination made by enzyme digestion of the milk proteins before titration; the lower curve is for an undigested milk - &-carrageenan (0-05 per cent.) sample and the upper curve for a sample that has been digested with enzyme at 27 "C for 6 hours. Results obtained with milk and digested milk samples showed that acridine orange binds to milk proteins at pH 6.5, but enzymic treatment limits the interaction with protein to an extent such that it does not interfere with the stronger binding to carra- geenan. No attempts were made to remove the polypeptides by complete digestion and dialysis.from known concentrations of carrageenan. Fig. 4. Bovine milk - 1-carrageenan titration with acridine orange. Acridine orange, 1.5 X 1 0 - 5 ~ and carrageenan, 0.05 per cent. 0, undigested milk - carrageenan cqmplex; and @, enzyme digested DISCUSSION The calculated equivalent masses determined by our method are of the order expected from the structure of the polymer and are in good agreement with the values obtained from elemental analyses of all the samples (Table 11). In order to use the method as an assay procedure it is necessary to know the equivalent mass per anionic site for each carrageenanDecember, 19731 PROCEDURE FOR THE ASSAY O F CARRAGEENAN 861 sample, which value is directly available from the titration procedure described.Other workers4S7J have obtained similar results, particularly with acridine orange, on a variety of strongly binding polyanions. Weakly binding polyanions, such as sodium carboxymethyl- cellulose, cannot be titrated in the manner described because of the large amount of unbound dye present at the equivalence p ~ i n t , ~ , ~ which is a useful advantage when carrageenan must be determined in the presence of carboxylated polysaccharides. With carrageenan and other strong binders there are only small amounts of unbound dye, as shown by the results given in Fig. 2. TABLE I1 EQUIVALENT MASSES PER REPEATING POLYSACCHARIDE UNIT FOR VARIOUS CARRAGEENAN SAMPLES Equivalent mass per sulphate group r - 3 Carrageenan sample From sulphate content By acridine orange titration Lambda (i) .. * . 286 298 Lambda (ii) .. .. 258 266 Kappa .. .. .. 370 390 Iota .. .. .. 296 290 Fluorimetric measurements have important advantages over those of absorption. Stone, Childers and Bradley4 in their absorption studies on sulphated polysaccharides did not always obtain good end-points, and we have also found that when salt was present in any of the samples only spectrofluorimetry gave satisfactory results. This effect is due to the changes in the shape of the absorption spectrum of the bound dye, which arise from variation in ionic strength. The analysis of fluorescence from monomer dye is simpler and no manipulation of the results is necessary in determining the equivalence point. In addition, it is possible, by using fluorimetry, to analyse samples containing as little as 5 p.p.m.of carrageenan even in the presence of other additives and cations. It has also been established that the method is applicable to systems involving carrageenan in the presence of other stabilisers, both neutral and charged, and proteins. The fluorescence method can also be used for determining the sulphate group content of different carrageenan samples. With protein-free samples the titration procedures can be carried out in triplicate in less than 1 hour. The results show that the method is suitable for the determination of carrageenan in enzyme-digested samples that contained protein, no removal of cations or clarification treatment being necessary for titration with the dye solution.The method is also applicable to other strongly binding polyanions such as dextran sulphate and polystyrene sulphonate. For stabilising a number of milk products the mixture of hydrocolloids studied by us is used, namely, carrageenan, carboxymethylcellulose in the presence of one or more of the neutral gums, mar, tragacanth, locust bean, etc.14 Fractionation of such mixtures is complex and is generally required before particular stabilisers can be individually determined.15 As carrageenan is a primary material for controlling stabilisation, its ready determination in a product such as, for example, infant milk, is of considerable practical value. The method we have described is simple, accurate and rapid for determining carrageenan in a conventional stabiliser mixture.The success of the method, when slightly modified even in the presence of protein, salts and other materials that may be present in the final commercial product, shows that the procedure could also prove of value in these circumstances. We thank the Copenhagen Pectin Co. and Dr. D. A. Rees (Unilever Research) for the well characterised samples of K-, L- and h-carrageenan. REFERENCES 1. 2. 3. 4. 5. 6. 7. Hansen, P. M. T., and Whitney, E. McL., J . Dairy Sci., 1960, 43, 175. Graham, H. D., J . Fd Sci., 1968, 33, 390. - , J . Dairy Sci., 1972, 55, 1675. Stone, A. L., Childers, L. G., and Bradley, D. F., Biopolymers, 1963, 1, 111. Phillips, G. O., in Balazs, E. A., Editor, “Chemistry and Molecular Biology of the Intercellular Matrix,” Volume 11, Academic Press, London and New York, 1970, p. 1033. Moore, J. S., Phillips, G. O., Power, D. M., and Davies, J. V., J . Cheun. SOG. ( A ) , 1970, 1155. Vitagliano, V., Costantino, L., and Zagari, A., J . Phys. Chem., 1973, 77, 204.862 8. 9. 10. 11. 12. 13. 14. 15. CUNDALL, PHILLIPS AND ROWLANDS Stone, A. L., and Bradley, D. F., J . Amer. Chem. Sm., 1961, 83, 3627. Davies, J. V., Dodgson, K. S., Moore, J. S., and Phillips, G. O., Biochem. J., 1969, 113, 465. Scott, J. E., and Willett, I. H., Nature, Lond., 1966, 209, 985. Lewis, C., Cundall, R. B., Llewellyn, P., and Phillips, G. O., J . Phys. Chem., 1970, 74, 4172. Cundall, R. B., and Evans, G. B. ,J. Scient. Instrum. ( J . Phys., E ) , 1968, 1, 305. Lamm, M. E., and Neville, D. M., J . Phys. Chem., 1965, 69, 3872. “Natural Plant Hydrocolloids,” Adv. Chem. Ser., 1954, No. 11. Morley, R. G., Phillips, G. O., Power, D. M., and Morgan, R. E., Analyst, 1972, 97, 315. Received May 8th 1973 Accepted June 2nd, 1973