178 Analyst, March, 1974, Vol. 99, ++, 178-183 Determination of iota- Carrageenan with 2-Thiobarbituric Acid BY W. ANDERSON AND W. BOWTLE (Department of Pharmaceutical Technology, University of Strathclyde, Glasgow G1 1X W ) A colorimetric method for the determination of iota-carrageenan is des- cribed, based on the measurement of the yellow colour developed by the reaction of the 3,6-anhydro-~-galactose residue with 2-thiobarbituric acid and hydrochloric acid. The method has a sensitivity equivalent to 0.01 mg ml-1 of carrageenan and can also be used to determine carrageenan in blood and urine in concentrations of 0.02 and 0.01 mg ml-l, respectively. The method has advantages of sensitivity, relative specificity, a small sample-volume requirement and convenience over previously published methods for the determination of sulphated polysaccharides, thus providing a means of ex- tending quantitative biological investigations of this group of substances.THE determination of sulphated polysaccharides still presents difficulty and is normally subject to appreciable error, especially when their determination in biological material is required. Several methods have been suggested for the determination of sulphated poly- saccharides; the toluidine blue1 and azure A2 methods can be used, in which the excess of cationic dye, after completion of a metachromatic reaction between the dye and the macro- anionic sulphated polysaccharide, is determined. These methods lack specificity and require carefully standardised conditions, particularly of salt concentration, rendering them unsuitable for routine, accurate determinations in complex media in which the salt concentration varies, when mixtures of macro-anions occur and when dialysis is not always possible.The determination of sulphated polysaccharides from the amount of the sulphate ester or their identification by infrared spectroscopy in media other than simple solutions requires prior removal of the sulphated polysaccharide by precipitation with a quaternary ammonium compound from the m i x t ~ r e , ~ followed by isolation and characterisation of the sulphated polysaccharide prepared from the macro-anion - quaternary ammonium complex and there- after determination with one of the methods referred to.1,2. Such methods have obvious dis- advantages for the routine determination of sulphated polysaccharides in boilogical media in which they are likely to occur in small concentrations.Some of the difficulties can be resolved for certain sulphated polysaccharides by adapting the resorcinol m e t h ~ d , ~ , ~ in which 5-hydroxymethyl-2-furaldehyde, produced from fructose or 3,6-anhydro-~-galactose that has been liberated from the polysaccharide, reacts with resorcinol in the presence of l,l-dieth~xyethane~,~ to form a coloured reaction product, which can then be determined. However, we have found that this method presents certain difficulties when applied to the determination of sulphated polysaccharides containing 3,6-anhydro- D-galactose, not only in biological fluids but even in aqueous solution. The continuing need for a rapid, convenient method for the determination of sulphated polysaccharides containing 3,6-anhydro-u-galactose has been intensified by a recent report on the intestinal toxicity of sulphated polysaccharides after oral administration to experi- mental animals8 and the demonstration that gastrointestinal absorption of certain members of this group of substances with low relative molecular mass can O C C U ~ , ~ ~ ~ ~ together with their traditional and apparently continuing use in foodstuffs, beverages and medicines.In order to provide a more accurate and sensitive method for the determination of sulphated poly- saccharides containing 3,6-anhydro-~-galactose, the use of 2-thiobarbituric acid, previously suggested as a reagent for the quantitative determination of fructosides,ll has been investi- gated, and is now reported.REAGENTS AND MATERIALS- Sul+hated +olysaccharide containing 3,6-anhydro-~-galactose-Both native and degraded iota-carrageenans were used. The native carrageenan was obtained by extraction with hot 0 SAC and the authors.ANDERSON AND BOWTLE 179 water from the red seaweed Eucheuma spinosum ; degraded carrageenan was prepared from native carrageenan by treatment with 0.001 N hydrochloric acid followed by neutralisa- tion with 0.001 N sodium hydroxide solution (after a predetermined degree of depolymerisa- tion, monitored viscosimetrically, had occurred), precipitation with ethanol, dialysis and freeze-drying. Results for these carrageenans are summarised in Table I. Lambda-carra- geenans were obtained by extraction from Chondrus crispus and Gigartina 9istiZlata in the usual manner.For purposes of spectrophotometric calibration crystalline methyl-3,6-anhydro-~-galacto- pyranoside was used. TABLE I PROPERTIES OF SULPHATED POLYSACCHARIDES CONTAINING 3,6-ANHYDRO-D-GALACTOSE Sulphate ester 3,6-Anhydro-~- Mass average content (-OS0,Na) galactose content?, relative molecular per cent. per cent. mass: Native r-carrageenan . . .. .. 37.7 18.70 f 0.21 800 000 Degraded r-carrageenan . . .. 36.1 21-01 f 0.21 25 000 chloride and a gravimetric finish. * Determined by complete acid hydrolysis with 10 N hydrochloric acid followed by addition of barium t Determined by 2-thiobarbituric acid method described herein. 1 Determined by light scattering. 2-Thiobarbituric acid, 1, l-diethoxyethwe and resorcinol-Laboratory-reagent grade.Charcoal-This was Norit A (Sigma), washed with 10 per cent. acetic acid followed by Sugars-The sugars used were D-glucose (AnalaR) , L( +)-rhamnose, D-XylOSe, D-arabinose, Heparin. Chondroitin sulphate. Blood-The blood was obtained by exsanguination from the ox, rat and guinea-pig; 4 parts of fresh blood were collected into 1 part of 3-8 per cent. m/V trisodium citrate solution. The plasma was used after centrifugation and removal of particulate sediment. Urine-Freshly voided normal human urine was used. distilled water and heated to redness in air for 15 minutes. L-fucose, D-galactose , D-mannose, D-fructose (glucose-free) and D-ribose. EXPERIMENTAL COLOUR DEVELOPMENT WITH 2-THIOBARBITURIC ACID- (a) Determinations of monosaccharides and polysaccharides in a series of 2-ml aqueous solutions (the concentrations of which are noted below) were carried out by adding 2 ml of 2 x 10-2 M thiobarbituric acid solution and 2 ml of 10 N hydrochloric acid to each, heating them for 6 minutes at 100 "C, followed by cooling for 2 minutes in an iced water bath and measuring the absorbance at 432nm within 30 minutes.Concentrations of the test sub- stances used were: carrageenans, 1 to 40 mg per 100 ml; heparin and chondroitin sulphate, 10 to 2000 mg per 100 ml; methyl-3,6-anhydro-~-galactose, 0.025 to 0.25 x M; fructose, 0.1 t o 0-25 x l o - 3 ~ ; and other monosaccharides, (b) The determination of degraded carrageenan in blood and urine was carried out as follows.Blood-A l-ml volume of plasma was shaken for 60 minutes at 37 "C with 8 ml of saturated potassium chloride solution so as to dissociate the carrageenan - protein complexes. Protein was precipitated by adding 0.3 ml of 30 per cent. trichloroacetic acid and removed by centri- fugation. The carrageenan in 2 ml of the supernatant liquid was determined by adding 1 ml of 3 x M thiobarbituric acid followed by 1.5 ml of 10 N hydrochloric acid and the method conducted as described in (a) above. The final concentrations of the reagents in the reaction mixture were the same as in the procedure described in (a). Urine-A l-g amount of charcoal was added to 1 O m l of fresh urine and shaken for 30 minutes. After filtration, 0.5 ml of 4 N sodium hydroxide solution was added to 2 ml of the filtrate and the mixture was then placed in a bath at 100 "C for 10 minutes, cooled in an iced water bath, 0-2 ml of 10 N hydrochloric acid added and a 2-ml aliquot treated as in (a) above.to 5 x l o - 3 ~ .180 ANDERSON AND BOWTLE : DETERMINATION OF IOTA-CARRAGEENAN [Analyst, Vol. 99 MODIFIED RESORCINOL METHOD- This method, intended for the determination of fructose and 3,6-anhydro-~-galactose, was carried out as described,"' except that the 1, l-diethoxyethane concentration was varied. 1.6 2 1.2 a -e s1 2 0.8 0.4 RESULTS 8-THIOBARBITURIC ACID METHOD- Colour development-The spectra obtained with the products of the interaction of Z-thio- barbituric acid with iota-carrageenan, methyl-3,6-anhydro-~-galactose and fructose are shown in Fig. 1, which shows a single peak at 432 nm in each instance.Beer's law is obeyed over the ranges of direct application and graphs of the absorbance at 432 nm versus the concentra- tion of the substance in aqueous solution yield the equations shown in Table 11. As would be expected, the lambda-carrageenans from Chondrus crisp~s and Gigartina pistillata con- tained less 3,6-anhydro-~-galactose than iota-carrageenan, which was verified by infrared studies. Colour stability-The first-order rate constants for loss of absorbance at 432 nm for thiobarbituric acid and degraded carrageenan (taken as being representative) in dark and light conditions were 12.4 x lov4 and 29.8 x lo4 min-l, respectively, at 36-5 "C and 06_x and 1.1 x min-l, respectively, at 20 "C. TABLE I1 EQUATIONS FOR COLOUR DEVELOPMENT BETWEEN 2-THIOBARBITURIC ACID AND CARRAGEENANS, METHYL-3,6-ANHYDRO-D-GALACTOSE AND FRUCTOSE Degraded &-carrageenan .. .. .. .. y = 0.0715~ Native L-carrageenan .. .. .. . . y = 0.0633~ Carrageenan from C. crispus . . .. .. y = 0.0463~ XCarrageenan from G. pistillata . . .. .. y = 0.0185~ Degraded X-carrageenan from G. $istillata . . y = 0.0132~ Methyl-3,6-anhydro-~-galactose . . .. .. y = 0.3390~ Fructose . . .. .. .. .. . . y = 0.3110~ y = absorbance a t 432nm; and x = concentration/mg per 100 ml. XCarrageenan from C. crispus . . .. .. y = 0.0121X Specificity-The ratios of the colour intensity developed at 432nm as a result of the reaction between 2-thiobarbituric acid and various sugars relative to that developed by equi- molar amounts of methyl-3,6-anhydro-~-galactose and carrageenan are shown in Table 111.March, 19741 WITH 2-THIOBARBITURIC ACID 181 TABLE I11 SPECIFICITY OF 2-THIOBARBITURIC ACID ASSAY FOR CARRAGEENAN CONTAINING 3,6-ANHYDRO-D-GALACTOSE Ratio r Sugar Fructose .. .. .. .. Ribose . . .. .. .. Fucose . . .. .. .. Mannose . . .. .. .. Rhamnose . . . . .. Xplose . . + . .. .. Arabinose .. .. .. Galactose .. .. .. Glucose . . .. .. .. A* 80.0 2.0 2.0 1-8 1.8 1.2 0.6 0.6 0.6 - Bt 197.0 5.0 5.0 4.0 4-0 2.5 1.4 1-4 1.4 absorbance (432nm)hsing lpmol of sugar absorbance (432nm) using lpmol of methyl-3,6-anhydro-~-galactose absorbance (432nm) using lpmol of sugar absorbance (432nm) using lpmol of carrageenan (repeating unit) * Ratio A = 100 x t Ratio B = 100 x The minimum theoretical relative molecular mass of the repeating carrabiose unit in carra- geenan when sodium is the micro-cation is 405 (calculated on the structures shown by Rees12) and this value was used in calculating the equivalent molar concentrations.Chondroitin sulphate and heparin at a concentration of 200 mg per 100 ml showed absorbances at 432 nm of 0-055 and 0.017, respectively. Determination of degraded carrageenan in blood and urine-Pre-treatments of blood and urine, as described in ( b ) above, required to reduce interference in the assay did not completely remove interfering substances ; blank samples containing no carrageenan showed absorbance values of 0.11 to 0-17 (urine) and 0.05 to 0.08 (ox and guinea-pig blood). However, after accounting for blank values, assays of carrageenan in blood (2 to 100 mg per 100 ml) and urine (1 to 40 mg per 100 ml) yielded graphs that indicated conformity to Beer’s law over the ranges of direct application and the equations in Table IV were derived from the average values obtained. TABLE IV EQUATIONS FOR COLOUR DEVELOPMENT BETWEEN 2-THIOBARBITURIC ACID AND DEGRADED L-CARRAGEENAN IN URINE AND BLOOD Degraded i-carvageenapz in- Urine .. .. .. y = 0.0403~ Control . . .. .. y = 0.0473~ Ox blood .. . . y = 0.0110~ + 0.046 y = absorbance a t 432 nm (corrected for blank values in blood and urine); and x = concentration of degraded L-carrageenan/mg per 100 ml. Blood and urine controls were aqueous solutions of degraded &-carrageenan, treated in a similar manner to plasma and charcoal-treated urine test samples.Control . . .. .. y = 0*0108X Reproducibility-In aqueous solution, the coefficient of variation of the absorbance at 432 nm (for a fifteen-sample series) was not more than 6.6 per cent. at degraded carrageenan concentrations of 1 to 2 mg per 100 ml, and not more than 5-5 per cent. at higher concen- trations. In blood, the coefficient of variation of the absorbance at 432 nm was not more than 6 per cent. at a degraded carrageenan concentration of 20 mg per 100 ml of blood (five-sample series). The reproducibility of the results for urine was less than that in aqueous solution or in blood because of the difficulty of removing all interfering substances from the samples. Coefficients of variation a t degraded carrageenan concentrations in the range from 3 to 45 mg per 100 ml of urine were 5 to 11 per cent.In a six-sample series containing eight to twelve different concentrations of degraded carrageenan, the correlation coefficients between the absorbance at 432 nm and concentration were not less than 0.994.182 ANDERSON AND BOWTLE : DETERMINATION OF IOTA-CARRAGEENAN [A?%a,!$si!, VOl. 99 RESORCINOL METHOD- While the results of Yaphe and Arsenault’ were, in general, confirmed for fructose and for 3,6-anhydro-~-galactose, it was found that the development of the measured absorbance peak at 555 to 558 nm was time dependent and that it occurred after formation of another peak, at 515 to 520 nm, which decreased as the size of the measured peak increased. Further, significant deviations from Beer’s law at 555 to 558 nm occurred at the upper concentration ranges for degraded iota-carrageenan, 16 to 96 mg per 100 ml, although these deviations could be decreased by adjusting the heating time for colour development from 10 to 15 minutes and increasing the 1,l-diethoxyethane concentration from 2-56 to 12-9 pmol per 100 ml of reagent.Further modifications to the 1 ,l-diethoxyethane concentration and heating time were required for carrageenan in urine and unacceptably high blank values could not be overcome when determining carrageenan in blood. Also, for fructose, 3,6-anhydro-~-galactose and carrageenan, a single concentration of 1,l-diethoxyethane did not yield maximum colour development at 555 to 558 nm over the entire range of concentrations investigated. Excess or insufficient 1,l-diethoxyethane caused distortion of the spectra and lowering of the absorb- ance peak.DISCUSSION The use of 2-thiobarbituric acid provides the basis for a method for the determination of 3,6-anhydro-~-galactose and polysaccharides that contain this sugar. Such determinations have previously involved the use of resorcinol reagents, which are believed to form coloured condensation products with 5-hydroxymethyl-2-furaldehyde or its derivative, laevulinic acid, resulting from hydrolysis of the polysaccharide and ring rearrangement of the liberated 3,6-an- hydro-D-galactose. We have found, however, that the resorcinol method is unsuitable for routine direct application in the determination of carrageenan containing 3,6-anhydro-~-galac- tose both in aqueous solution and in biological fluids because of spectral distortion effects that result from the varying 1,l-diethoxyethane requirements for optimal colour development of a variety of test substance concentrations and also because of high blank values.Recoveries of carrageenan from urine were varied (140 per cent. at 10 mg per 100 ml to 70 per cent. a t 120mg per l O O m l ) and the method was found to be virtually inapplicable to assays in blood samples. The proposed method, involving the use of 2-thiobarbituric acid, avoids the difficulties associated with the resorcinol method because measurements are made with reference to a single-peak spectrum and Beer’s law is obeyed over the entire ranges of direct application for both 3,6-anhydro-~-galactose and sulphated polysaccharides containing this sugar, in aqueous solution, blood and urine.The method shows high specificity for 3,6-anhydro-~-galactose, for carrageenan that contains this sugar and for fructose (Table 111). The low absorbances yielded by a number of mono- saccharides are comparable with those found for resorcinol,7 indicating a corresponding lack of interference in the 2-thiobarbituric acid method. In calculating the relative molecular mass of the carrabiose unit used for comparison of these sugars, the minimum theoretical value was used and thus the results given in Table I11 show the maximum interference in the assay that could be caused by these sugars. In addition, colour development in the Z-thio- barbituric acid method by the sulphated polysaccharides heparin and chondroitin sulphate is less than 0.5 per cent. of that developed by corresponding amounts of carrageenan con- taining 3,6-anhydro-~-galactose. Neither of these polysaccharides contains 3,6-anhydro- D-galactose but both might easily cause confusion in the determination of any one sulphated polysaccharide in biological fluids when other methods of assay are used.Fructose develops a strong colour with thiobarbituric acid and might be expected to cause interference, especially in blood and urine. However, when using the pre-treatments described, the interference due to normal levels of all substances is reduced to a value that is widely accepted as being unavoidable in a colorimetric assay of biological materials. In any experiment in which fructose may occur together with carrageenan, treatment of the sample by heating it with 4 N sodium hydroxide solution prior to assay with 2-thiobarbituric acid will destroy any fructose present without affecting the assay.Chloral hydrate has recently been reported13 as developing a colour when treated with thiobarbituric acid under mild, non-acidic conditions (pH 9.5). We have found, however,March, 19741 WITH 2-THIOBARBITURIC ACID 183 that this substance does not interfere in the assay of carrageenan with thiobarbituric acid. Methods that have been used for the determination of carrageenan, which involve the use of cationic dyes, for example toluidine blue, are less suited to the determination of small amounts as toluidine blue is a general reagent for macro-anions in acidic solution, which leads to low specificity and susceptibility to interference from other macro-anions, when such methods are applied to complex materials.Methods that involve the precipitation and isolation of the sulphated polysaccharide prior to infrared examination and determination with toluidine blue,14 apart from incorporating a positive identification step rendered necessary by the non-specificity of the test, are tedious, tend to be subject to substantial quantitative error and require relatively large sample volumes (20 ml of blood; 10 ml of urine) that are im- practical when the experiment calls for small laboratory animals, which may be required to survive the experiment. The need for a rapid, reliable method for the determination of carrageenan containing 3,6-anhydro-~-galactose in the biological fluids of large numbers of small animals has recently intensified, not only as a result of the unknown metabolism of carrageenans and recognition of their toxicity, but also because they find continued use in biological investigationsg~10 and in foodstuffs. The method described, involving the use of 2-thiobarbituric acid, is suitable for such investigations, which have hitherto been hindered by the lack of a suitable assay technique. 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. REFERENCES MacIntosh, F. C., Biochem. J., 1941, 35, 776. Jaques, L. B., Monkhouse, F. C., and Stewart, M., J . Physiol. (Lond.), 1949, 109, 41. Scott, J. E., Meth. Biochem. Analysis, 1960, 8, 145. O'Neill, A. N., J . Amer. Chem. Soc., 1955, 77, 2837. Yaphe, W., Analyt. Chem., 1960, 10, 1327. Arsenault, G. P., and Yaphe, W., Analyt. Biochem., 1965, 13, 133. Yaphe, W., and Arsenault, G. P., Ibid., 1965, 13, 143. Watt, J., and Marcus, R., Gut, 1971, 12, 164. Anderson, W., and Soman, P. D., J . Pharm. Pharmac., 1967, 19, 520. Anderson, W., and Bowtle, W. J., Ibid., 1972, 24, 166P. Percheron, F., C. r. Hebd. Se'anc. Acad. Sci., Paris, 1962, 255, 2521. Rees, D. A., A h . Carbohyd. Chem., 1969, 24, 267. Kamat, S. S., Barve, V. P., and Mahal, H. S., Analyst, 1972, 97, 877. Beattie, I. A., Blakemore, W. R., Dewar, E. T., and Warwick, M. H., Fd Cosmet. Toxicol., 1970, Received July 6th, 1973 Accepted October 4th, 1973 8, 257.