88 JOHNSTOX, OGSTON AND STANIER: A MODIFICATION OF THE ELSOS p o l . 76 A Modification of the Elson and Morgan Method for the Estimation olf Glucosamine BY J. P. JOHNSTON, A. G. OGSTlON AND J. E. STANIER An investigation of the stages of the Elson and Morgan method for estimating glucosamine has led to modifications of the details of its stages. It has been found that 1Opg or more of free glucosamine or combined glucosamine of hyaluronic acid can be estimated with an accuracy of 3 per cent. IN spite of the satisfactory accuracy and sensitivity of the colorimetric method for the estimation of glucosamine described by Elson and IMorgan,l a number of other workers have re-investigated and modified it: chief of these (using less than 3 mg of glucosamine) have been Palmer, Smyth and Meyer,2 Sarensen3 and B l i ~ , ~ while others have modified it for use with larger quantities.The accuracies claimed range from 2 per cent. to 10 per cent. Ogston and Stanier5 found the version of Palmer, !';myth and Meyer insufficiently accurate, in spite of minor modifications. In order to test the last method further, we made a number of parallel determinations of free glucosamine, taken from the same batch of material; we found differences of up to 6 per cent. between estimations on samples of the same solution after treatment in bulk with acetylacetone, and even larger differences between estimations on lots separately treated with acetylacetone. Investigation of the stages of the method was therefore carried out, beginning with the final stage and working backwards.From this, it appeared that considerable errors might arise from the processes of transfer and dilution, from loss of acetylacetone during heating in open tubes and from variations in the temperatures and durations of the several stages. We therefore modified the method so as to eliminate the transfer of material from one vessel to another as far as possible. At. the only remaining transfer, the whole sample is transferred and this has also the effect of reducing the amount of glucosamine that may be estimated to as little as 1Opg. Under the conditions chosen, the optimal times for some stages have been found to differ from those given by other authors. The procedure is given below, followed by comment on its stages. PROCEDURE Hydrolysis (when the glucosamine is combined in hyaluronic acid)-Seal 1 ml bf solution, containing 10 to 80 pg of glucosamine, and 1 ml of 8 N hydrochloric acid in a Pyrex tube 12 cm long and of 0.8 cm internal diameter.Mix the solution thoroughly and totally immerse the tube in a boiling water bath for 4 hours. Removal of hydrochloric acid-Cool the tube, centrifuge to remove traces of solution from the upper end and open it. Remove the hydrochloric acid and evaporate the contents of the tube to dryness by placing the tube over solid sodium hydroxide in a desiccator evacuated by an oil pump. Treatment with acetylacetone-Add 1 ml of distilled water to the tube to dissolve the contents, and then add 0.5 ml of freshly prepared acetylacetone solution (0.2 ml of acetyl- acetone in 10 ml of 0-5 N sodium carbonate).Re-seal the tube and totally immerse it in a boiling water bath for 60 minutes, then cool it in ice, centrifuge and open it. With free glucosamine, the estimation begins with the treatment with acetylacetone. Treatment with Ehrlich's reagent-Transfer the contents of the tube quantitatively by Pasteur pipette to a 5-ml graduated flask. Wash the tube three times with a total of 2.9 ml of aldehyde-free ethanol and add the washings to the flask. Stopper the flask and mix the contents; warm the flask to 37" C by immersion in a thermostat for 3 minutes. Then add 0.5 ml of Ehrlich's reagent (0.8 g of 9-dimethylaminobenzaldehyde, A.R., in 30 ml of aldehyde- free ethanol plus 30ml of concentrated hydrochloric acid) and sufficient ethanol to make the solution up to 5ml.After mixing, leave the flask in the thermostat for 60 minutes, then remove it and measure the extinction at 535 mp with a spectrophotometer. Comparison is made with standard glucosamine introduced at the acetylacetone treatment stage. COMMENTS ON PRWEDURE Hydrolysis-The yield of estimated glucosamine (Fig. 1) from ox synovial-fluid mucin rose to a maximum after 4 hours and fell slowly again with longer hydrolysis to 88 per cent.Feb., 19511 89 at 22 hours. If the hydrolysis and destruction are both unimolecular processes, the maximum yield at 4 hours would be 97 per cent. of the total glucosamine. No significant loss of pure glucosamine occurred under the same conditions up to 32 hours (Fig. 1); this allows the glucosamine standard to be introduced a t the stage of the treatment with acetylacetone.The change of reactivity of glucosamine reported by Blix4 to follow heating with 4 N hydrochloric acid was not observed. Evaporation of the hydrochloric acid in vacuo (Hadidian and Piriee) avoids the need for transfer and neutralisation. AND MORGAN METHOD FOR THE ESTIMATION OF GLUCOSAMINE L I 1 I l l 1 1 4 0 4 8 12 16 20 24 28 32 Duration of Hydrolysis, hours ' 10 20 30 40 SO 60 70 80 90 100 I10 I I I I I I 1 1 1 1 . Acetylation time. minutes Fig. 1. Effect of duration of hydrolysis Fig. 2. Effect of time of heating on extinction obtained with free glucosamine with acetylacetone on final colour develop- and with mucin. Ordinate: Elcrn a t ment. Ordinate: Elcrn a t 535 mp, as per- 535 mp as percentage of value obtained centage of value obtained after 30 minutes a t reference time.0 Extinction given by free glucosamine, as percentage of value a t start of hydrolysis. 0 Extinction given by much glucosamine as percentage of value obtained a t 4 hours Treatment with acetylacetone-This is an intrinsically unsatisfactory reaction, since both the acetylacetone and the glucosamine undergo some destruction' during its course. Blix4 has shown that a greater yield is obtainable by using larger amounts of acetylacetone and sodium carbonate and that, under his conditions, the yield is least sensitive to variations in the amounts of these reagents. We have used the amounts recommended by Palmer, Smyth and Meyer2; although these give a lower yield, the sensitivity of the reaction is sufficient, and we are satisfied that the ordinary errors of measuring the reagents will not introduce an error of more than 1 per cent.into the final estimate. We found it to be essential to carry out this reaction in sealed tubes, totally immersed at 100" C. Under these conditions, the maximum yield is at 60 minutes (Fig. 2) ; Blix4 recommends this period, but other authors have used shorter periods (Slzrrensen3 20 minutes; Palmer, Smyth and Meyer2 15 minutes; Ogston and Stanier5 30 minutes). Treatment with Ehrlich's reagent-Differences of up to 10 per cent. were found if the reagents were mixed at 0" C or room temperature and then brought to 37" C. Bringing to 37" C before mixing reduced this difference to 1.9 per cent. (standard deviation of eight determinations). A cczcracy-Fourteen parallel samples of glucosamine hydrochloride (each 30 pg of glucosamine) showed a standard deviation (single estimate) of 1.9 per cent. of the mean value, and seven parallel samples of ox synovial mucin (each containing 30 pg of glucosamine) a standard deviation of 3 per cent. One of us (J. E. S.) is indebted to the Medical Research Council for a Training S tudentship. REFERENCES 1. 2. 3. 4. Blix, G., Acta, Chem. Scand., 1948, 2, 467. 5. 6. 7. Elson, L. A., and Morgan, W. T. J., Biochenz. J . , 1933, 27, 1824. Palmer, J. W., Smyth, E. M., and Meyer, K., J . Biol. Chem., 1937, 119, 491. Slarensen, M., C.R. Lab. Carlsberg, 1938, 22, 487. Ogston, A. G., and Stanier, J. E., Biochem. J., 1950, 46, 364. Hadidian, Z., and Pirie, N. W., Ibid., 1948, 42, 260. Tracey, M. V., personal communication. DEPARTMENT OF BIOCHEMISTRY OXFORD June, 1950