March, 19511 ANALYTICAL METHODS COMMITTEE i87 Analytical Methods Committee REPORT PREPARED BY THE ANEURINE PANEL OF THE SUB-COMMITTEE ON VITAMIN ESTIMATIONS The Chemical Assay of Aneurine in Foodstuffs THE Analytical Methods Committee has received from its Sub-committee on Vitamin Estimations the following Report based on the work of its Aneurine Panel. The Report has been approved by the Analytical Methods Committee and its publication authorised by the Council. INTRODUCTION The following workers assisted in the preparation of this Report: E. R. Dawson (Chairman), F. Wokes (Secretary), D. C. M. Adamson, R. G. Booth, W. F. Elvidge, G. E. Foster, J. Greenbaum, R. A. C. Isbell, J. King, W. Martin, F. W. Norris, T. L. Parkinson, S. A. Price, H. N. Ridyard and S. Y . Thompson. Assistance in some of the investigations was also given by D.J. Finney, J. Houston, B. C. P. Jansen, G. F. Lothian, H. Monk, F. Clermont Scott and Miss S. Weiner. The Report is concerned with the chemical assay of aneurine in certain specified classes of foodstuffs, namely: cereals and cereal products; malt and malted products; yeast and yeast products; meat and meat extracts. This Report deals with the determination of aneurine by the thiochrome method. In this method the aneurine is oxidised by alkaline ferricyanide to thiochrome, which fluoresces in ultra-violet light. Under standard conditions the net fluorescence of the thiochrome is directly proportional to its concentration over a given working range. The method originated with Jansen,l and has undergone numerous modifications.Many of these are described in “Methods of Vitamin Assay,”2 which was found by the Panel to be specially helpful in devising the fluorimetric method described below. This method was based on much experimental work too lengthy to describe in detail. MAIN POINTS STUDIED Extraction-The Panel experimented with digestion with pepsin, papain, “Taka-diastase” and “Clarase,” combined with extraction with diluted hydrochloric acid, sulphuric acid or sodium acetate solution. Purification of the extract-Washing with isobutanol was found to be less effective than adsorption on base-exchange silicate for removal of interfering substances, as shown by the fluorescence of the blanks and the percentage recoveries of added aneurine. The recovery128 ANALYTICAL METHODS COMMITTEE : THE CHEMICAL v o l .76 could vary widely for a given sample according to the level of addition of ane~rine.~ This led the Panel to adopt a cautious attitude to the use of recovery methods to correct for the presence of interfering substances. Oxidation of aneurine to thiochrome-The use of a premixed alkaline ferricyanide solution gave more uniform results than separate additions of alkali and ferricyanide. Water- saturated isobutanol was found more satisfactory than dry isobutanol for extracting the thiochrome from the aqueous reaction mixture. A limit has been suggested for the amount of non-specific fluorescence in the isobutanol. Measurement of juorescence and calculation of results-Calculation of the net fluorescence is preferred to the use of a calibration curve.This calculation facilitated measurement of the fluorescence of the blanks, and was found to be effective with fluorimeters of the null-point type, at present mainly used in this country, as well as with fluorimeters of the direct-reading type. RESULTS OBTAINED BY THIS RECOMMENDED METHOD Table I summarises results obtained by the collaborating laboratories on samples of malt extract, wheat germ, wheat flour, yeast extract and dried yeast. As samples of malted barley and of malt and oil had previously been found to give satisfactory results with the assay methods applied to malt extract, they were not included-in the series. All the laboratories agreed in finding no significant amount of aneurine in the meat extract sample, which is therefore not included in Table I.Eiecause of the difficulties that the Panel had encountered when assaying yeast preparations, three samples of dried yeast differing widely in potency were examined. Table I1 contains results obtained for the samples by biological and microbiological methods. The results obtained by the proposed method showed a degree of agreement between different laboratories and with biological and inicrobiological results that seemed to justify the Panel recommending it as a method suitable for use by analysts to determine aneurine in the foodstuffs tested by the Panel. Further trials of the method are needed before it can be finally put forward as an official method of the Society. The Honorary Secretary of the Analytical Methods Committee (Dr.D. C. Garratt) will be pleased to receive comments and criticisms from any analyst using the method. The results in Tables I and I1 were obtained in the following laboratories: Boots Pure Drug Co., Nottingham; Cereals Research Station, St. Albans; Distillers Co., Ltd., Research Department, Epsom; the laboratories of D. W. Kent-Jones and A. J. Amos, London, W.5; Glaxo Laboratories, Greenford ; Government Laboratory, London, W.C. 1 ; Lyons Laboratories, London ; National Institute for Research in Dairying, Shinfield ; Nederlandsch Institut voor Volksvoeding, Amsterdam ; Ovaltine Research Laboratories, King's Langley ; Roche Products, Welwyn Garden City; Vitamins Ltd., Research Laboratories, London; Wellcome Chemical Works, Dartford. TABLE I COLLABORATIVE RESULTS BY THE RECOMMENDED METHOD AneuIine found, pg per g Malt Wheat flour Yeast f A \ Laboratory extract germ (85%) extract Sample 1 Sample 2 Sample 3 A 2.7 22 55 29 17 45 D 4.0 21 3.8 54 20 47 E 24 14 54 F 15 G 3.9 21 52 27 I 32 3.7 21 3-8 61 28 3.9 20 47 27 19 57 L M 3.5 23 3.8 39 N 3.6 37 47 Means 3-62 21.3 3.75 61.3 29.1 16.8 60 ~- A r -7 Wheat Dried yeast 16 k The data of Table I have been examined by Dr.E. C. Wood with a view to assessing the precision of the method. As the amount 'of aneurine in the various samples differed soMarch, 19511 ASSAY O F ANEURINE I N FOODSTUFFS 129 widely, it was first necessary to convert all the individual results into percentages of the mean result for each sample in turn. This has been done in Table IA, omitting references to particular laboratories.TABLE IA COLLABORATIVE RESULTS EXPRESSED AS PERCENTAGES OF THE CORRESPONDING MEAN RESULT Malt extract 74-7 110.6 107.9 102-3 107-9 90.8 Wheat germ 103.1 98-5 98.5 98-5 93.7 107.8 Wheat Yeast flour extract 101.3 107-1 101-3 105.2 101-3 101-3 96-0 118.8 91.6 76.0 7 Sample 1 99.5 82.4 92.6 109.8 96-1 92.6 127.0 Dried yeast A -l Sample 2 Sample 3 101.0 93.3 118.8 97.4 83-2 11 1.9 89.1 97.4 95.1 112.9 It is apparent that the agreement between laboratories is much better for some materials tested than for others. The coefficient of variation, which is the standard error expressed as a percentage of the mean, is in fact 4-9 per cent. for the two wheat products together, as compared with 13.5 per cent. for the four yeast products together.The appropriate test shows a high probability that this implies a real difference in the precision of the method when applied to these two classes of substance. The mean coefficient of variation for all the samples together is 11-9 per cent., though it is doubtful whether this pooling of all the results is justified. It is safer to say that when a sample of wheat flour or germ is being examined in different laboratories, 19 out of 20 of the results obtained should be found in the long run to lie within &lo per cent. of the mean for that particular sample; but if a yeast product is being examined the corresponding range will be &27 per cent. Agreement between replicate results in the same laboratory will of course be closer than this; the data available provide no evidence for assessing the “within laboratory” precision.TABLE I1 RESULTS BY BIOLOGICAL AND MICROBIOLOGICAL METHODS Aneurine found, pg per g Labora- tory Method used K Ratgrowth N Bradycardia P Ratgrowth F Yeast fermentation J Lactobacillus fermenturn Mean value of the colla- borative fluorimetric results (from Table I) r h -I Wheat Dried yeast A Malt Wheat flour Yeast f 7 extract germ (85%) extract Sample 1 Sample 2 Sample 3 20 3-9 21 3.3 45 27 38 17 64 4.1 23 68 3.62 21-3 3-75 51.3 29.1 16.8 50 RECOMMENDED METHOD APPARATUS- Base-exchange tubes-These are to be made of glass, the upper part being not less than 15 cm long and 0-8 to 1.0 cm in internal diameter and the lower part being narrow-bore tubing of suitable length. The lower end may be fitted with a tap or other method of con- trolling the rate of flow. A reservoir to contain at least 30 ml may be attached to the upper part.Oxidation vessels-These may be stoppered measuring cylinders, stoppered bottles, separating funnels or large boiling tubes of about 100ml capacity. Fluorescent grease must not be used to lubricate taps; glycerin or silicone may be used. Fluorimeter-This may be of the direct reading type, measuring in deflections, or the null- point type, measuring in densities.* The exciting radiation must be within the range 300 1 *Some null-point type instruments read also in “transmission.” Here transmission = density‘130 ANALYTICAL METHODS COMMITTEE : THE CHEMICAL vol. 76 to 400 mp; it is most conveniently obtained at suitable intensity by the use of a high-pressure mercury-vapour lamp, type MB, in conjunction with a primary filter.A secondary filter* transmitting mainly light between 400 and 450 mp is placed between the fluorescent solution and the photo-cell. REAGENTS- All chemicals should be of analytical reagenk quality or the purest otherwise obtainable. EthanoZ-Redistilled in all-glass apparatus. Sodium hydroxide, 15 per cent. solution-1)issolve 15g of sodium hydroxide in water and dilute to 100ml. Potassium ferricyanide, 1 per cent. solution-Dissolve 1 g of potassium ferricyanide in water and dilute to 100 ml. This reagent has been found to be stable for at least a week i f kept cool and in the dark, preferably in a brown bottle. Alkaline potassium ferricyanide solution-Dilute 3 ml of 1 per cent.potassium ferricyanide $0 100 ml with cool 15 per cent. sodium hydroxj.de solution. Hydrochloric acid , approximately 0.2 N soldution-Dilute 17 ml of concentrated hydro- chloric acid to 1 litre with water. Sodium acetate, 2.5 M solution-Dissolve 205 g of anhydrous sodium acetate (CH,COONa) or 340 g of CH3COONa.3H,0 in water and dilute to 1 litre. isoButyl alcohol (water-saturated)-Steam distil commercial isobutyl alcohol in an all-glass apparatus. The fluorescence of the distillate should be not more than that of a 1 in 100 .dilution in approximately 0.1 N sulphuric acid of the quinine standard (see p. 131), that is, of a diluted solution containing 0.01 pg of quinine sulphate per ml. Enzyme solution-Prepare a fresh solution daily from a suitable source of ph0sphatase.t Suspend, with thorough shaking, 6 g of the enzyme preparation in 2-5M sodium acetate solution and dilute to 100 ml with additional sodium acetate solution.Each batch of material used as a source of phosphatase should be tested for aneurine by this fluorimetric method and the necessary correction applied. Potassium chloride, 25 per cent. solution-Dissolve 250 g of potassium chloride in distilled water and dilute to 1 litre. The reagent is stable indefinitely. Acid potassizcm chloride solution-Dilute 0.5 ml of concentrated hydrochloric acid to 1 litre with 25 per cent. potassium chloride solution. The reagent is stable indefinitely. Base-exchange silicate-This consists of an. artificial zeolite in the form of a granular powder of 60 to 90 mesh size, tested for its suitability for adsorbing and eluting aneurine under the given conditions (at least 90 per cent."recovery" of aneurine should be obtained under the given conditions) .$ Activate the base-exchange silicate as follovvs: Place a convenient quantity (100 to 500 g) of the base-exchange silicate in a suitable beaker, add sufficient hot 3 per cent. acetic acid solution to cover the material and maintain the temperature a t about 100" C for 10 to 15 minutes, stirring frequently. Allow the mixture to settle and decant the supernatant liquid. Repeat the washing three times with hot 25 per cent. potassium chloride solution, and finally wash with boiling water until the last washing gives no reaction for chloride. Dry the material at approximately 100" C and store in a well-closed container.Stock aneurine soZution, 100 pg per ml-Prepare this from the British Standard Prepara- tion, or a sub-standard of equal purity. Dissolve a weighed amount of aneurine hydrochloride equivalent to 50 mg of the International Standard in sufficient 0.2 N hydrochloric acid to make 500 ml. This solution is stable for several months if stored in a refrigerator (i.e., below Standard aneurine solution-Dilute 5 ml clf stock aneurine solution, warmed to room temperature, to 100ml with water. Transfer 10ml of this dilution to a flask containing 200 ml of approximately 0.1 N sulphuric acid and 126 ml of sodium acetate solution and dilute to 250 ml with distilled water. The final concentration of aneurine is 0.2 pg per ml. This solution is stable for at least a week if stored in the refrigerator.Prepare just before use. 15" C). * Chance 0x1, 1-5 to 2.0 mm thick, has been found suitable for the primary filter, and Chance OB2 t Taka-diastase (diluted with lactose) (Parke Davis & Co., London) and Clarase (Takamine Laboratories, (blue), 1-6 to 2.0 mm thick, for the secondary filter. Clifton, N. J., U.S.A.), have been found suitable. Decalso F, supplied by the Permutit Co., Ltd., London, has been found suitable.March, 19511 ASSAY OF ANEURINE IN FOODSTUFFS 131 Stock quinine sulphate solution (100 %per mZ)-Dissolve 0.025 g of quinine sulphate, B.P., in sufficient 0.1 N sulphuric acid to make 250 ml. This solution is stable for twelve months if stored in a dark brown bottle at a temperature below 5” C.Qzhine standard (1 pg per mZ)-Dilute 10 ml of stock quinine sulphate solution to 1 litre with 0.1 N sulphuric acid. This solution is stable for three months if stored in a brown bottle at a temperature below 5” C. Any solution that has been exposed to ultra-violet light in the fluorimeter should be discarded. Acetic acid, 3 per cent. solution-Dilute 30 ml of acetic acid to 1 litre with distilled water. Bromocresol green indicator-Triturate 100 mg of bromocresol green with 7.2 ml of 0-06 N sodium hydroxide and dilute with sufficient water, free from carbon dioxide, to make 200 ml. Nitrogen gas in cylinders-If desired, an air current can be used instead. PROCEDURE EXTRACTION- (a) Sampling-The material to be assayed, if solid, should pass a No. 30 B.S.sieve or a finer sieve, and should be well mixed just before withdrawal of the sample, to ensure homogeneity. If liquid, the material should be well stirred before sampling. (b) Accurately weigh or pipette into a large boiling tube a sample (not more than 6 g) estimated to contain not more than 50 pg of aneurine. Add 65 ml of approximately 0.1 N hydrochloric acid or sulphuric acid. Digest the sample for 30 minutes in a bath of boiling water, with frequent mixing. The liquid must remain at a pH below 4.5 during the digestion. If a t the end of the digestion it is not distinctly acid to the bromocresol green indicator, the extract should be discarded and a further quantity of the sample extracted with more concentrated acid. (c) Cool the extract to below 50” C and adjust the pH to between 4 and 4.5 by addition of 2-5 M sodium acetate solution, using bromocresol green as external indicator.Add 5 ml of freshly prepared enzyme suspension, mix, and incubate at 45” to 50” C for 3 hours, or a t 37” C overnight with addition of a drop of sulphur-free toluene. (d) Cool to room temperature, centrifuge the mixture until the supernatant liquid is clear and quantitatively transfer the supernatant liquid to a 100-ml volumetric flask. Wash the residue by centrifuging successively with 10 ml, 10 ml and 5 ml of 0.1 N hydrochloric acid or sulphuric acid, Add the washings to the supernatant liquid and dilute the whole to 100 ml with water. (a) Plug the bottom of an adsorption column with glass wool, which should be lightly packed, and fill the column with 6 g of activated base-exchange silicate suspended in water.Allow the water to drain almost entirely, but leave enough to cover the base-exchange silicate, and pour in 5 ml of 3 per cent. acetic acid. Allow to drain as before. (b) Transfer 25ml of the “original extract” to the column by means of a pipette. Discard the filtrate that has percolated through the column. Wash the column with three successive portions, about 10 ml each, of boiling water and discard the washings. (c) After washing the column, pour through 10 ml of almost boiling acid potassium chloride solution from a supply kept hot in boiling water. Collect the eluate in a stoppered 25-ml graduated cylinder. Add a second 10-ml portion when all of the first portion has entered the base-exchange silicate and collect the eluate in the same cylinder.When this second portion has drained through, cool the eluate to room temperature, dilute to 25 ml with acid potassium chloride solution and mix well. This is the “sample eluate.” This is the “original extract.” PURIFICATION- OXIDATION TO THIOCHROME- source of ultra-violet light must be avoided. In this and all subsequent stages undue exposure of the solutions to direct daylighi or other (a) Pipette 5 ml of sample eluate into each of two oxidation vessels. (b) Start a stream of nitrogen or air bubbling through the solution in vessel number 1, add 5 ml of alkaline potassium ferricyanide solution and then add 25 ml* of water- saturated isobutyl alcohol, the current of nitrogen or air still being continued. Shake (or continue vigorous bubbling) for 90 seconds.* See footnote on p. 132.ANALYTICAL METHODS COMMITTEE : THE CHEMICAL [vol. 76 Start a stream of nitrogen or air bubbling through the solution in vessel number 2, then add 5ml of 15 per cent. sodium hydroxide solution, followed by 25 ml* of water-saturated isobutyl alcohol, then continue as in step (b). This is the “unknown blank.” ( d ) Repeat steps ( a ) , (b) and (c) with 5 ml of standard aneurine solution in place of the sample eluate. CAUTION. To avoid changes in experimental conditions the oxidation of all solutions used Similar precautions must be 132 (c) The solution from step (c) is the “standard blank.” ht a given assay should be carried out in irnmedide succession. taken in the reading of their JEzcorescence.SEPARATION OF THIOCHROME SOLUTION AND MEASUREMENT OF ITS FLUORESCENCE- ( a ) After the solutions have stood for a few minutes to allow complete separation, add 1 ml of ethanol to the upper layer in each vessel and stir the upper layer gently until it is clear, taking care to avoid disturbing the aqueous layer. (b) Take off each upper layer into a cuvette and measure its fluorescence against that of the quinine standard, if a null-point fluorimeter is being used, or as direct deflec- tions if a deflection instrument is being used. The blank should exhibit only faint fluorescence. “RECOVERY” EXPERIMENT- Repeat the above procedure, including the steps of extraction, purification, conversion to thiochrome, separation of thiochrome solution and measurement of fluorescence, with a “recovery” experiment, made by adding, to another portion of the sample that is the same weight as that previously taken, a volume of the stock aneurine solution containing an amount of aneurine similar to the amount expected in that weight of sample.CALCULATION- convert all densities into antilogs and take reciprocals of these.? If the fluorescence has been measured on an, instrument provided with a density scale, Let U = reciprocal for unknown; U, = reciprocal for unknown blank; S = reciprocal for standard; S, = reciprocal for standard blank; V = volume of original solution put through base-exchange silicate. If the fluorescence intensities have been measured as deflections, these are used instead of the reciprocals of antilogs.Then the aneurine content of the sample in pg per g = U-Uu, 1 25 100 S - S, 5 V g of sample taken x - x - x - The factor 1/5 converts the reading to pg per ml instead of pg per 5-ml aliquot. Since the final volume of eluate is 25 ml, the factor 25/V corrects for volume changes during adsorption and elution. If the suggested 25 ml is adsorbed, this factor becomes unity. Note-This calculation assumes that the fluorescence of the thiochrome solution in the unknown is If it is not, the assay should be repeated using a smaller less than that of the quinine standard. amount of the material. USE OF RECOVERY EXPERIMENT DATA- Calculate the percentage recovery of the ad.ded aneurine from the following fomula- x 100 A, = aneurine content of sample in pg per g calculated as above; A, = aneurine content of sample with added aneurine calculated as above; A, = pg of aneurine added to each gram of sample. A, - A, A* where- * If 25 ml of water-saturated isobutyl alcohol doe:; not provide sufficient solution for conveniently filling the particular cuvettes in use, this volume can be increased to some precise higher figure, e.g., 30 ml, which must then be maintained throughout all experiments. t If the instrument reads “transmissions,” use those readings instead of reciprocals of antilog density; U, UB, S and SB are then the readings themselves. -March, 19511 ASSAY OF ANEURINE IN FOODSTUFFS 133 This percentage recovery provides an indication of the effect of disturbing factors, including the quenching effect of impurities, but should not be relied upon to make a satis- factory correction for interfering factors. In general, if the percentage recovery falls below 80, the result of the assay should be considered unsatisfactory. REFERENCES Jansen, B. C. P., Bec. Truv. Ckim. Puys-Bus, 1936, 55, 1046. “Methods of Vitamin Assay,” Association of Vitamin Chemists, Inc., New York, 1947. Ridyard, H. N., Analyst, 1950, 75, 634. 1. 2. 3.