November, 1979 COMMUNICATIONS 1097 Communications Material for publication as a Communication must be o n a n urgent matter and be of obvious scientijic importance. Rapidity of publication i s enhanced i f diagrams are omitted, but tables and formulae can be included. Communications should not be simple claims for priority: this facility for rapid publication i s intended for brief descriptions of work that has progressed to a stage at which it is likely to be valuable to workers faced with similar Problems. A fuller paper may be offered subsequently, if justijied by later work. Manuscripts are not subjected to the usual examination by referees and inclusion of a Communication i s at the Editor’s discretion. An Improvement in the Determination of Available Lysine in Carbohydrate-rich Samples Keywords : Lysine determination ; carbohydrate-rich samples ; 2,4,6-trini- trobenzenesulphonic acid The chemical determination of available lysine in carbohydrate-rich materials is seriously affected by the formation of what Booth1 called the “humin” artifact produced during the hydrochloric acid hydrolysis of dinitrophenylated protein following reaction of the sample with l-fluoro-2,4- dinitrobenzene.2 Low and variable recoveries of edinitrophenyllysine added to carbohydrate feeds were obtained.Similar difficulties were reported by Hall and co-workers3~* in their use of 2,4,6-trinitrobenzenesulphonic acid (TNBS) as the protein coupling reagent. The presence of carbohydrate during the hydrolysis causes the production of a dark brown compound, which, depending on the amount of carbohydrate, may form a heavy black precipitate. Hall et al.3 described this interference as a “caramelisation” of the carbohydrate causing the formation of carbon particles which may adsorb the etrinitrophenyllysine (E-TNP-lysine) and concomitantly imparts a colour measurable at 415 nm, together with the yellow c-TNP-lysine.Carpenter (with Booth)5 has also drawn attention to the adverse effects of starches in the determination of available lysine. This communication describes an improvement in the standard procedure3 employed in this laboratory for the determination of available lysine with TNBS, which it is believed removes most of the difficulty with carbohydrate samples. It consists essentially of what is probably a chlorination to bleach the interfering “humins,” at the same time changing the chemical structure of the TNP products to make the application of the method more specific for lysine and more versatile for a wider range of samples.Crown Copyright.1098 COMMUNICATIONS Analyst, Vol. 104 Experimental Determination of Available Lysine The procedures employed for the preparation of the samples and for the determination of available lysine using TNBS were as described in previous papers3p4 i.e., 0.5-ml aliquots were taken of sample suspensions in 0.1% m/V of agar, containing 10, 20 and 40 mg ml-l of sample, ground in a porcelain ball mill to pass a 200-mesh BS sieve. Samples of 5 mg mass were reacted with TNBS for 30 min at 30 "C (short coupling) and lysine was determined as rep~rted.~ Samples of 5, 10 and 20 mg mass were reacted with TNBS for 75 min at 40 "C (long ~oupling).~ A 4-ml aliquot ( ~ 2 - 8 m g sample mass) of the 10ml of hydrolysate was diluted to 7ml with water and extracted twice with 5ml of diethyl ether, the extracts being discarded.The aqueous solution containing 6-TNP-lysine was treated with 0.2- 2.0 ml of 5% V/V sodium hypochlorite solution (10-14y0 m/V available chlorine), mixed and the volume made up to 10 ml with water. After centrifuging at 3000 rev min-l for 10 min the absorbance of the supernatant solution was measured at 415 nm in an optical cell of 10-mm light path. i. ii. Determination of total lysine Total lysine was determined in hydrochloric acid hydrolysates (0.3 g sample mass refluxed at 100 "C for 16 h with 200 ml of 6 M hydrochloric acid) by the method of Moore et al.6 using a Locarte bench-model amino acid analyser.Lysine was eluted from a 6-cm column (for basic amino acids) with a sodium citrate buffer of pH 5.28 and reacted with ninhydrin. Results In an attempt to apply TNBS to the determination of available lysine in carbohydrate-rich materials, Hall et aL4 proposed reacting a 5-mg sample for 30 min at 30 "C and obtained good correlation with the total lysine values for a number of plant materials, but showed that for some types of sample, an "available" lysine figure could be measured that was actually greater than the total lysine content. The problem was highlighted when samples of pea husk meal (Pisurn sativurn) "charred" so severely during hydrolysis that the measurement of the e-TNP-lysine was impracticable.In order to overcome the interference due to the highly coloured products of hydrolysis, it was decided to study the effects of bleaching and reducing agents on the hydrolysed sample. Dilute solutions of ascorbic acid, sodium sulphite, hydrogen peroxide and sodium hypochlorite were added to the hydrolysate after ether extraction of the TNP-amino compounds. Only sodium hypochlorite had any decolorising action but this was extremely effective, reducing the interfering absorbance by 70-90% or more. The introduction of the hypochlorite resulted in the absorbance for the control or method blank (i.e., the sample treated with hydrochloric acid before the addition of TNBS and then heated and extracted with diethyl ether as for the deter- mination) from most carbohydrate samples being less than the absorbance due to €-TNP-lysine.Before bleaching, the reverse situation pertained, the method blank having a much greater absorbance than that due to e-TNP-lysine, and this was analytically undesirable. Table I shows TABLE I EFFECT OF HYPOCHLORITE ON ABSORBANCE OF E-TNP-LYSINE AND Results are absorbance values at 415 nm. PEA HUSK MEAL HYDROLYSATE Treatment with 5% V / V hypochlorite/ml 7 L * Sample 0 0.2 0.5 1 .o 2.0 DL-Lysine, 100 pg . . Method control* . . .. . . .. Pea husk meal, 2 mg . . . . .. Control . . .. .. . . .. Pea husk meal, 4 mg . . . . .. Control . . .. . . .. .. Pea husk meal, 8 mg . . . . .. Control . . * . .. .. .. (corrected for method control) 0.412 0.010 0.428 0.393 0.694 0.560 1.080 0.865 0.393 0.011 0.166 0.091 0.274 0.169 0.627 0.362 0.394 0.014 0.134 0.082 0.241 0.127 0.461 0.250 0.405 0.009 0.107 0.055 0.213 0.103 0.422 0.208 0.405 0.013 0.109 0.059 0.213 0.094 0.360 0.141 * Sample and reagents subjected to the analytical procedure but with 6 M HCl acidification followed by addition of TNBS before incubation for 75 min.November, 1979 COMMUNICATIONS 1099 the effect of various amounts of hypochlorite on the absorbance using the pea husk meal and it can be seen that the absorbance difference between the test and the control became constant for 0.5-2.0 ml of hypochlorite and was stoicheiometric within experimental limits for 2-8 mg sample mass.Absorbance values for standard samples of c-TNP-lysine and -taurine were hardly affected.A comparison of the available lysine values of 14 carbohydrate-rich samples obtained by the two procedures (long and short TNBS couplings) indicated that the anomalous situation of the “available” lysine being higher than the total lysine content had been resolved ; without exception the hypochlorite-treated reactions resulted in lower available lysine values than were measured from unbleached reactions and correlated closer with the total lysine levels (Table 11). TABLE I1 TOTAL AND AVAILABLE LYSINE IN CARBOHYDRATE-RICH SAMPLES Lysine, yo mlm A I 7 Available Sample Cassava . . . . Barley . . .. Field beans . . Flaked maize . . Maize cobs . . Oats . . .. Pea meal . . .. Potato waste . . Poultry mash . . Rice bran . . . . Turkey crumbs .. Pig meal . . . . Tapioca . . . . Wheatings. . . . . I . . .. . . . . . . .. . . .. . . . . . . . . . . Total “Bleached ’’ * “Unb1eached”t . . 0.08 0.08 0.28 . . 0.36 0.21 0.36 . . 1.21 0.95 1.14 . . 0.26 0.25 0.42 . . 0.21 0.19 0.48 . . 0.34 0.27 0.40 . . 0.50 0.25 0.27 . . 0.48 0.44 0.53 . . 0.23 0.21 0.21 . . 0.53 0.47 0.52 . . 0.44 0.37 0.45 . . 0.02 0.02 0.21 . . 1.34 1.25 1.55 . . 0.36 0.35 0.61 * Long coupling, treated with 1 ml of 5% V/V hypochlorite. f Short ~oupling.~ Discussion As reported earlier,4 the determination of true nutritionally available lysine in animal feed- stuffs has been far from simple. Experience has shown that although 2,4,6-trinitrobenzene- sulphonic acid has proved very useful as the protein coupling reagent, it is necesary to be aware of its limitations.Apart from the interference of carbohydrates (which applies equally to the use of l-fluoro-2,4-dinitrobenzene) it has been shown in this laboratory that several naturally occuring amines such as agmatine, spermine, spermidine and taurine (unpublished findings), and also ornithine, hydroxylysine, galactosamine and gluc~samine,~ react with TNBS as does lysine and can be measured as lysine. During this work it was additionally observed that the action of the hypochlorite, which may well be that of a chlorination, changed the physical properties of the E-TNP-amino compounds formed from synthetic reagents, and from animal and plant samples. Thus, TNP-lysine and some of the TNP derivatives of the compounds already mentioned became soluble in diethyl ether ; TNP-agmatine and TNP-taurine remained in the aqueous phase whilst TNP-galactosamine and TNP-glucosamine were converted into colourless products.I t seems, therefore, that by selective ether extraction of the “chlorinated” c-TNP-lysine, the hypochlorite treatment of the TNP reaction should lend itself to a more accurate differential determination of available lysine in the presence of these interferences. Such a step would be invaluable in the analysis of fish meals that contain variable and significant amounts of taurine. Indeed, some evidence in this respect has already been demonstrated. Using the standard TNBS methodJ3 the apparent available lysine of a sample of krill (a small shrimp of the Euphausiacea) was 6.01% m/m (8.62 g per 16 g of N) but the total lysine and total taurine levels, measured with an amino acid analyser, were 4.05% m/m (5.81 g per 16 g of N) and 2.29% m/m, respectively. When the hypochlorite-treated c-TNP-lysine was further extracted into diethyl ether and the residual absorbance, assumed to be due to TNP-taurine left in the1100 COMMUNICATIONS Analyst, Vol.104 aqueous phase, was measured against a TNP-taurine standard, the values for available lysine and taurine were 3.45% m/m (4.95 g per 16 g of N) and 2.08% m/m, respectively. The corresponding results for an ordinary feeding fish meal were as follows: apparent available lysine, 4.68% m/m (6.68 g per 16 g of N) ; total lysine, 4.29% m/m (6.12 g per 16 g of N) ; “true” available lysine, 3.15% m/m (4.49 g per 16 g of N); total taurine, 0.93% w/m; and taurine measured spectrophotometrically, 1.33% m/m. It is believed that this further modification of the use of TNBS makes the determination of available lysine much more reliable and applicable to a wider range of materials. I t should be a significant advantage in evaluating protein quality for what is regarded as the most important amino acid for growth purposes. It is hoped to submit a full report of this work in a later publication. References 1. 2. 3. 4. 5. 6. Booth, V. H., J . Sci. Fd Agric., 1971, 22, 658. Carpenter, K. J., Biochem. J . , 1960, 77, 604. Hall, R. J., Trinder, N., and Givens, D. I., Analyst, 1973, 98, 673. Hall, R. J., Trinder, N., and Wood, M. R., Analyst, 1975, 100, 68. Carpenter, K. J. (with Booth, V. H.), Nutr. Ab:itr. Rev., 1973, 43, 423. Moore, S., Spackman, D. H., and Stein, W. H., Analyt. Chem., 1958, 30, 1185. Received August 31st, 1979 Ministry of Agriculture, Fisheries and Food, Agricultural Development and A dvisory Services, Analytical Chemistry Department, Government Buildings, Kenton Bar, Newcastle upon Tyne, NEI 2YA R. J. Hall K. Henderson