The Determination of Lactic Acid in Microgram QuantitiesBY R. P. HULLIN AND R. L. NOBLEDepartment of Biochemi8try, Univer8ity of Leeds(Received 13 February 1953)The conversion of lactic acid to acetaldehyde by hot concentrated sulphuric acid was first reported by Deniges (1909) and its use as. a basis for the determination of lactic acid was described by Mendel &Goldscheider (1925). Eegriwe (1933) described a reaction between p-hydroxydiphenyl and acetaldehyde to yield a violet colour and this colour reaction was first adopted as a sensitive means of estimatinglactic acid by Miller & Muntz (1938).The method most widely employed for determining lactic acid on a microscale is that ofBarker & Sunmmerson (1941). In the course of work in this laboratory requiring the estimation of very smallamounts of lactic acid, usually in the presence of pyruvate, reproducible results could not be achieved with this technique; further, the final -optical densities were not directly proportional to -the amounts of lactic acid present.This paper contains the results of our investigations whereby several modifications of Barker & Siummerson's procedure were made. With these changes, 1-8 ,ug. of lactic acid could be determinedwith an accuracy of ± 2%. When interfering pyruvic acid is present, the procedure involves a dilution so that the method in this case is applicable to 2-10 pig. lactic acid.METHODSReagentMp-Hydroxydiphenyl (1.5 g.) is dissolved in 10 ml. of 5% (w/v) NaOH and diluted to 100 ml. with water.Standard lactic acid solution: 0-2133 g. of pure, dry lithium lactate (Hillig, 1937), was dissolved in about 100 ml.of water, 1 ml. of concentrated H2SO4 (A.R.) added and the solution made up to 1 1. with water.Recommended methodThe protein-free solution (2 ml.) containing 10-80 pg. of lactic acid, is pipetted into a 150 x 25 mm. test tube which contains 1 ml. of 20% (w/v) CUS04, 5H20 and the final volume is made up to 10 ml. with water. Approximately 1 g. of solid Ca(OH)2 is added and, after thoroughly mixingand allowing to stand for 30 min. or more, the solution is centrifuged.

If the original 2 ml. of protein-free solution contains 100-60 ptg. of pyruvic acid, 6 ml. of the supernatant solution from the first treatment with copper-lime is pipetted into a second tube containing 0-6 ml. of 20% (w/v) CUS04, 5H20 and 0-6 g. of solid Ca(OH), added. This procedure isthen repeated a third time with 3 ml. of supernatant solution,0-3 ml. of 20% CUS04, 5H20 and 0-3 g. of Ca(OH)2,whereby the amount of pyruvic acid present is reduced to a minimum and the colour interference due to it becomes a small constant value which may be deducted from thefinal optical density.1 ml. of the supernatant solution from the third copperlime treatment is transferred to another 150 x 25 mm. testtube, held in the arm of a mechanical shaker with its lower end immersed in an ice-water mixture. CuS04, 5H20(0.05 ml. of 12% (w/v) solution) is

added followed by 6 ml.of conc. H2SO4 (A.R.) dropwise from a burette with vigorous shaking, the tap of the burette being lubricated with theconcentrated acid. After complete addition, the contents are poured and allowed to drain into a small ground-glass stoppered Pyrex tube which is then heated for 30 min. in a water bath maintained at 60+ 10. The tube is allowed to cool to 10-15°, the stopper removed, 0-1 ml. of thep-hydroxydiphenyl reagent added and the precipitated p-hydroxydiphenyl thoroughly dispersed in the H2S04. Incubation of the tube for 20 min. at 28-30° to ensure maximum colour development is then followed by 90 sec. ina boiling-water bath to destroy excess p-hydroxydiphenyl.After this treatment the tube is immediately cooled in icewater and the optical density ofthe resulting violet-coloured solution is determined in circular tubes of 1 cm. diameter with the Unicam diffraction-grating spectrophotometer at a wavelength peak of 560 m,u. against a reagent blank prepared by taking distilled water through the whole of theabove procedure.The calibration curve obtained by this method (E =,Ag.lactic acid x 0 059) for 0-8 ,ug. of lactic acid is a straight line.DEVELOPMENT OF METHODThe effect of varying the experimental conditions at different stages of the method is described in Table 1. The absorption curve of the acetaldehyde-p-hydroxydiphenyl complex shows a sharp maximum at 560 m,t. as illustrated by the following optical densities obtained from 6,ug. Of lactic acid in the conditions described on p. 289 at wavelengthsbetween 520 and 600 mp.: 520, 0-210; 540, 0-295; 550, 0.330; 560, 0 345; 570, 0-330; 580, 0-305; 600, 0-185.The calibration curve, which is strictly reproducible, indicatesthat the coloured complex obeys the Lambert-Beer Law in the concentration range of 1-8,ug. of lactic acid.Above 8 pg. the optical density does not increase in direct proportionality with the amount of lactic acid but falls off slightly. If coloured solutions, obtained from larger quantitiesof lactic acid (up to 10,ug.), are diluted with H2SO4:water (6:1, v/v) the optical densities of the diluted solutions are directly proportional to the amounts of lactic acid originally present.Speciflcity of the methodThe work for which the method has been used required the determination of lactic acid in the presence of acetoin, butane-2:3-diol, diacetyl and pyruvate. None of the first three of these compounds gives any colour when the method is appliedto solutions containing 60.ug. ofit and quantitativerecoveries of lactic acid are always obtained in the presence of such amounts. However, pyruvic acid, which is commonly present with lactic acid in media from enzymic studies,seriously interferes with the reproducibility of the method unless steps are taken to remove it. This is illustrated by the results collected in Table 2 for pyruvic acid samples withoutcopper-lime treatment. One copper-lime treatment does not yield reproducible results, although it clearly removes a considerable amount ofpyruvic acid. The repetition of the copper-lime procedure seemed a likely way of completely eliminating the interference due to pyruvate and the effect of three successive treatments is also illustrated by the results in Table 2. The third treatment reduces the optical density of the interferingpyruvic acid to a small constant value. Using this technique, recoveries oflactic acid ranging from 98 to 102% may be obtained in mixtures containing 2-10lg. of lactic acid and 10-60kg. of pyruvic acid/ml. of final solution; Table 1. Effects of varying experimental conditions('Colour' means optical density at 560 myt.)Stage OxidationOxidation

OxidationOxidationIncubation with p-hydroxydiphenylreagentDestruction of excessp-hydroxydiphenylCondition variedHeating at 1000 in opentubeStoppered and unstopperedtubesTime of heating at 600Concn. of Cu2+ addedTime at 28-30OTime of heating at 100°Recommended procedureStoppered tubes30 min.0-05 mhl. of 12% (w/v)CUS04, 5H2020 min.90 sec.Effect of other conditions The longer the heating, the less colour obtained Unstoppered tubes give about 15% less colour 24-65 min. gives maximum colour; <24 min., less colour 12-32%, same colour; <12%,less colour 20-50 min., same colour; <20 or > 50 min., less colour30-120 sec., identical colour Table 2. The effect of successive copper-lime treatments on the colour obtained with pyruvic acid samples ((I) 9 ml. of standard solutions (containing 200, 300, 400, 500 and 600,g. of pyruvic acid) + 1 ml. of 20% CUS04, 5H2O+ 1 g. Ca(OH)2. (II) 6 ml. of supernatant solution from (I) +0-6 ml. 20% CU$04, 5H20 +0-6 g. Ca(OH)2. (III) 3 ml. ofsupernatant solution from (II) +0-3 ml. 20% CuSO4, 5H20 +0-3 g. Ca(OH)2. 1 ml. of supematant solution taken for analysis in each case.)Optical densities (I)After one copperlimetreatment0-0700-1050-1100-1250-130(II)After two copperlimetreatments0-02000300-0500-0500-055(III)After three copperlimetreatments0-0170-0200-023

0-0220-020Pyruvicacid conen.

(jug./ml. infinal solution)2030405060Withouttreatment0-2000-2750-380-430-52290

Table 3. Recoveriea of lactic acidfrom mixture8 of pyruvic and lactic acid8Final solution-A Optical density Recovery ofLactic acid Pyruvic acid of mixture Optical density lactic acid(/Ag./ml.) (I.tg./ml.) (Obs.) (Corr.*) (%)2 60 0*140 0*120 1014 60 0*265 0.245 1026 60 0*368 0*348 988 60 0'490 0.470 996 20 0*385 0*365 1036 30 0*375 0*355 1006 40 0*380 0*360 1016 50 0*368 0-348 986 60 0 375 0 355 100* Observed value minu8 0.02 which is the average optical density produced by remaiuing pyruvic acid (20-60FIg./ml.final solution).these recoveries are corrected for the small constant opticaldensity due to interfering pyruvic acid after three copperlimetreatments. The results are shown in Table 3.Similar recoveries are obtained when known quantities oflactic acid are added to tissue and enzyme extracts anddeproteinization carried out with 10% (w/v) trichloroaceticacid.DISCUSSIONThe conditions under which the oxidation of lacticacid to acetaldehyde is carried out, in the first stageof the method of estimation, are of critical importance.When this oxidation is carried out in opentubes, the acetaldehyde recoverable as the finalcoloured complex varies greatly with the time ofheating employed. These findings are contrary tothose of Barker & Sunmmerson (1941) who state thatidentical results are obtained with heating periods

of 3-10 min. Mendel & Goldscheider (1925), whosemethod of lactic acid estimation used 1:2-dimethoxybenzeneinstead of p-hydroxydiphenyl to givea red compound, also claimed that heating in opentubes for 4-8 min. during the sulphuric acid oxidationstage produced no differences in the opticaldensity of the resulting coloured complex. However,Miller & Muntz (1938) did recommend the useof stoppered tubes, but gave no reason for this.Our results confirm those of Barker & Summersonregarding the influence of copper on the lactic acidoxidation but we have found it necessary to usea larger amount (0-05 ml. of 12% CUS04, 5H20instead of 0'05 ml. of a 4% solution) to achieve thefull increase in sensitivity. This modification may berequired owing to the greater recoveries of lacticacid as acetaldehyde using stoppered tubes.The modification ofmethod necessary to estimatelactic acid accurately in the presence ofpyruvic acidwas suggested by the work of Van Slyke (1917).Miller & Muntz (1938) have suggested heating thesulphuric acid oxidation mixture for 15 min.instead of 5 min. in order to remove any interferingpyruvate.SUMMARY1. Lactic acid has been determined in the range0-8 ,ug./ml. with an accuracy of ± 2%.2. The method, which can be employed over a2-10 jtg./ml. range in the presence of 10-60 ,ug./ml.of pyruvic acid, is based on that of Barker &Summerson (1941), but certain modifications areincorporated to achieve reproducibility.We wish to thank the Medical Research Coun,cil forfinancial amistance, and for a grant to one of us (R. L. N.).REFERENCESBarker, S. B. & Summerson, W. H. (1941). J. biol. Chem. 138,535.Denig6s, G. (1909). Bull. Soc. chim. Fr. 5, 647.Eegriwe, E. (1933). Z. anal. Chem. 95, 323.

Hiflig, F. (1937). J. A8s. off. agric. Chem., Wash., 20, 130.Mendel, B. & Goldscheider, I. (1925). Biochem. Z. 164,163.Miller, B. F. & Muntz, J. A. (1938). J. biol. Chem. 126,413.Van Slyke, D. D. (1917). J. biol. Chem. 32, 455.19-2

292 NOTES AND COMMENTcells. I. The bleaching process. Biochim. -, AND R. F. SCAGEL. 1958. Diurnal studyBiophys. Act., 29: 359-368. of phytoplankton pigments. An in situ studySTRICKLAND, J. D. H., AND T. R. PARSONS. 1960. in East Sound, Washington. J. Mar. Res.,A manual of sea water analysis. Bull. Fish. 17 : 567-583.Res. Bd. Canada, No. 1.25. 185 pp.YENTSCH. C. S.. AND T. H. RYTHER. 1957. Short C. D. MCALLISTER

term variations in phytoplankton chlorophylland their significance. Limnol. Oceanogr., Fisheries Research Board of Canada,2: 140-142. Nanaimo, B. C.ESTIMATION OF LACTIC ACID IN SEA WATER SOLUTIONS AND HOMOGENATESLactic acid is commonly estimated by the method of Barker and Summerson ( 1941) ; after oxidation in strong sulphuric acid solution to acetaldehyde the latter is coupled with p-hydroxydiphenyl in the presence of cupric ions to yield a purple compound. In biological material protein must be first precipitated, usually with trichloracetic acid, after which other interfering substances are removed by treatment with copper sulphate and calcium hydroxide. It has recently been pointed out ( Barnes, Finlayson, and Piatigorsky 1963) that the method is not entirely satisfactory in strong salt solutions and even less so in sea water of normal salinity(+ 32%0). In the investigation of anaerobic processes in marine animals the estimationof lactic acid in sea water media is often of considerable importance and a modificationof the Barker and Summerson method has, therefore, been developed which isapplicable in such media.Reagents, apparatus, and methods (i ) Copper sulphate: B.D.H., Analytical Reagent grade,(ii) Calcium hydroxide: B.D.H., finely precipitated,(iii) Sulphuric acid: B.D.H., Analytical Reagent grade, nitrogen-free,(iv) p-hydroxydiphenyl: a commercial product recrystallized several times,(v) Thallous sulphate: B.D.H., Analytical Reagent grade,(vi) Trichloracetic acid: B.D.H., Analytical Reagent grade,(vii) Lactic acid standards: pure recrystallized lithium lactate.The reaction was carried out with the slight modifications recommended in Umbreit,Burris, and Stauffer ( 1945; p 104), the transmittances being measured at 560rnp on a Unicam spectrophotometer. Concentrations are given in pg/ml of the solution,1 ml of which is eventually taken for oxidation and development of the color.When testing the method on sea water homogenates of biological material, proteinwas precipitated with 10% trichloracetic acid and other interfering materialsby the standard copper sulphate-calcium hydroxide treatment.Effect of salt solutions and sea water Figure 1 gives a number of calibration curves in distilled water, in sodium chloride

solutions, and in sea water, using the unmodified technique, It is evident that, contrary to what had been anticipated, sodium chloride at a strength equivalent to sea water does not give similar values; it had previously been thought that the formation of hydrochloric acid when sulphuricacid is added at the oxidation stage was responsible for the discrepancy. The values in sodium chloride solutions are clearly less reliable and for this the formation of hydrochloric acid may be responsible. The reaction is slightly less sensitive in chloride solutions; in sea water it ismuch less sensitive and the calibration curves and estimations of unknowns are not satisfactory. Absorption curves in the various media showed that in all the colored product is identical, and the lower sensitivity in sea water would therefore seem to be due to inhibition of its develNOTES

AND COMMENT 293 Lithium lactate /uq per ml. FIG. 1. Calibration curves for lactic acid estimationin various media: (a) Distilled water, (b) 3.5% sodium chloride solution, (c) 0.01% potassium bromide solution, (d) 50% sea water, (e) sea water. opment or its destruction. In view of the wide variety of substances tested for interference by Barker and Summerson and a knowledge of the composition of sea water it seemed that only the bromide content of the latter was likely to be responsible for the observed results, possibly as a result of the liberation of hydrobromicacid. Figure 1 shows the effect of adding bromide alone to distilled water, in an equivalent quantity to that in sea water. The similarity to the calibration curve in sea water alone is striking and together with the fact that removal of bromide (see below) gives values equivalent to those indistilled water strongly supports this view. Since the object was only to develop a satisfactory method for the estimation of lactic acid in sea water, the reasons for the interference have not been further investigated. Removal of bromide (and some chloride) by thallous sulphateKrogh and Keys (1934) found that thallous sulphate was a useful reagent with which to remove chloride from sea water prior to the estimation of its organic carbon content. The solubility of thallous bromide is only about one-sixth that of the chloride and addition of thallous sulphatemight be expected to serve the present purpose, provided that the presence of thallium ions does not interfere with any other part of the estimation. Table 1 shows the effect of adding various quantities of a saturated solution of thallous sulphate to sea water containing lactate.There is a suggestion in Table 1 that 0.5 ml of saturated thallous sulphate gives somewhat low returns at the highest lactic acid content; otherwise the method seems adequate to prevent interference by substances in the sea water and to suppress the variability associated with solutions containing chloride. Estimation of lactic acid in sea water In view of the above results the following procedure was adopted for the estimation in solutions of sea water when proteinand other materials are only likely to be present in trace amounts; such is often the case when any excretion of lactic acid into the surrounding sea water takes place after animals are exposed to anaerobic conditions.To 4 ml of the solution add 2 ml of a saturated solution of thallous sulphate. Centrifuge, take 1 ml of the supernatant, and proceed as described in the Barker- TABLE 1. Effect of the addition of thalloussulphate on lactic acid estimation. 2 ml ikctic acidsolution: variable quantity precipitant: resultspg/ml in original lactate solutionSaturatedthalloussulphatesolution(ml)Lactic acid found(LG.)

0.0 1.1 3.2 5.4 9.6 9.60.5 1.1 3.2 5.7 8.6 8.71.0 1.2 3.1 5.5 9.6 9.62.0 1.1 3.2 5.5 9.5 8.4294 NOTES AND COMMENT

Summerson method as indicated in Umbreit,Burris and Stauffer ( 1945).Estimation when body tissuespresent in sea water are If, as often happens, body tissues containing lactic acid must be homogenized in sea water then the proteins must be removed with trichloracetic acid and othersubstances by the copper sulphate-calcium hydroxide treatment, before the additionof thallous sulphate. The following method was shown to be satisfactory by adding

lactic acid to a tissue homogenate and estimating the return. To 4 ml of the homogenate add 1 ml of 10% trichloracetic acid (TCA); stir, centrifuge, and transfer the supernatant to a lo-ml standard flask; wash the precipitate with dilute TCA and add the washings to the supernatant. Neutralize the solution and make up to 10 ml. Withdraw 4 ml of the neutral solution and add 2 ml of asaturated solution of thallous sulphate. Centrifuge. Draw off 3 ml of the supernatant and add 0.5 ml of 20% copper sulphate and 0.5 g calcium hydroxide. Allow to stand half an hour with shaking. Centrifuge. Withdraw 1 ml and proceed as in the Barker-Summerson method. Carry calibrationthrough whole procedure.Some typical results are given in Table 2.TABLE 2. Estimation of lactic acid in tissuehomogenates in sea water. For method, see text.Values are reported as pg/ml lactic added andfoundLactic acidfoundha)

1.0 1.12.2 2.23.1 3.24.3 4.34.9 4.85.4 5.46.5 6.47.4 7.5REFERENCESBARNES, H., D. M. FINLAYSON, AND PIATIGORSKY.

The effect of desiccation and anaerobic conditionson the behaviour, survival, and generalmetabolism of three common cirrepedes. (Znpress. )KROGH, A., AND A. B. KEYS. 1934. Methodsfor the determination of dissolved organiccarbon and nitrogen in sea water. Biol. Bull.,67: 132-144.UMBREXT, W. W., R. H. BURRIS, AND J. F.STAUFFER. 1945. Manometric techniquesand related methods for the study of tissuemetabolism. Burgess Publ. Co., Minn. 203PP.H. BARNESAND

The Marine Station, D. M. FINLAYSON

Millport, Scotland.

The Estimation of Lactic Acid using Ceric SulphateBY S. R. ELSDEN AND Q. H. GIBSONAgricultural Research Council Unit of Microbiology, Department of Microbiology;and Department of Phy8iology, Univer8ity of Sheffield(Received 6 March 1954)Gordon & Quastel (1939) showed that ceric sulphateoxidizes lactic acid according to the followingequation:CH3,. CHOH. COOH + 2Ce4+CH3. CHO + C02 + 2H+ + 2Ce3+,and they made this reaction the basis of a methodfor the estimation of lactic acid. The reaction wascarried out at 500 and the acetaldehyde so formedwas removed by a stream of nitrogen, trapped in

bisulphite and estimated iodometrically in the usualway. The advantages claimed for this procedurewere simplicity and the fact that very few compoundsgave rise to volatile aldehydes under theexperimental conditions recommended by theauthors. Glucose in particular did not interfereand a preliminary treatment of glucose-containingtest solutions with the copper-lime reagent was notconsidered necessary.In its essentials the method devised by Long(1946) was similar to that of Gordon & Quastel(1939). He showed, however, that ceric sulphateoxidizes acetaldehyde to acetic acid; but in spite ofthis he was able to obtain good yields of acetaldehydefrom lactic acid by aerating very rapidly, thusremoving the acetaldehyde almost as fast as it wasformed. The aeration rate recommended was500 ml. air/min. as compared with the three to fourbubbles nitrogen/sec. advised by Gordon &Quastel.Winnick (1942) carried out the reaction inConway vessels with bisulphite in the centralchamber to trap the acetaldehyde and incubated for2 hr. at 50°. His method was modified by Conway(1950), who carried out the oxidation in stopperedtest tubes. After 30 min. incubation at 370, hetransferred 1 ml. samples of the reaction mixtureto diffusion vessels where the acetaldehyde wastrapped in bisulphite placed in the centre chamber.In view of the widely divergent conditions recommendedby the various authors, it seemed desirable,before using any of the ceric sulphate methods, toexamine the various steps in the procedure further.The results of these investigations are given below.EXPERIMENTAL AND RESULTSFactor8 influencing the oxidation oflactic acid by ceric 8ulphateEffect of ceric sulphate concentration and temperature on thevelocity of the reaction. Since carbon dioxide is one of theproducts ofthe reaction, Warburg manometers were used tostudy the effect of both temperature and concentration ofceric sulphate on the velocity of the reaction. The reactionwas carried out in N sulphuric acid and 0*2 ml. lithiumlactate solution in N sulphuric acid was added from the sideESTIMATION OF LACTIC ACIDbulb when thermal equilibrium had been attained. In Fig. 1are plotted the half times for the reaction at 22, 40 and 53°.It will be seen that the concentration of ceric sulphate hasa pronounced effect and that at 530 the half time is less than5 min. even with the most dilute solution of ceric sulphatetested.140 r_120 _100

E 80% 60I

4020n

400lS * I I--1-2 1-4 1-6 1-8 2-0 2-2 2-4 2-6Log. m-moles ceric sulphate/l.Fig. 1. Effect of temperature and ceric sulphate concentrationon the velocity of oxidation of lactic acid. Experimentswere carried out in Warburg manometers. 11-1 -

moles lithium lactate in 0-2 ml. N-H2SO4 were added fromside bulb at zero time; main compartment contained2-8 ml. ceric sulphate in sulphuric acid; gas phase, air.100190IV

4,.

E4,1-2 1-4 1-6 1-8 2-0 2-2 2-4 2-6 2-8Log. half-time (min.)Fig. 2. Effect of temperature and ceric sulphate concentrationon the velocity of oxidation of acetaldehyde. Experimentswere carried out in an apparatus similar to that ofLong (1946). Volume of reactants, 10 ml.; each tubecontained 26-2p.moles acetaldehyde; concentration ofsulphuric acid, N. Reaction was stopped by the addition ofexcess of ferrous sulphate and residual acetaldehydeaerated into bisulphite and estimated iodometrically.Oxidation of aedehyde by ceric Buiphate. Acetaldehydeand ceric sulphate were incubated in Long's apparatus atvarious temperatures and, at suitable time intervals, theresidual ceric sulphate was reduced to the cerous form byaddition of excess of ferrous sulphate. The acetaldehyderemaining was then transferred by aeration to bisulphitesolution and estimated iodometrically. The test solution

contained 26-2 ltmoles acetaldehyde, and the total volumeincluding the ceric sulphate solution was 10 ml. The finalconcentration of sulphuric acid was normal in all cases. Theresulte are given in Fig. 2 in which the log of the half-time ofacetaldehyde destruction is plotted against temperaturefor three different concentrations of ceric sulphate. It willbe seen that the rate of oxidation of acetaldehyde isgreater the higher the temperature and the greater theceric sulphate concentration.Removal of acetaldehyde by aeration. The effects of temperatureand velocity of aeration on the rate of removal ofacetaldehyde were next examined. These experiments wereconducted in an apparatus of the form and dimensionsrecommended by Long (1946). The volume of the reactionmixture was 10 ml. containing 20jsmoles of acetaldehyde inN sulphuric acid. Air was blown through the mixture andthe acetaldehyde was trapped in bisulphite. The rate of airflow was measured with a flow meter. After suitable intervalsof time, aeration was stopped and the acetaldehyde trappedin the bisulphite determined iodometrically. The experimentswere carried out with flows of 200-1000 ml. air/mi.

2-4.-2-2-

2-0 -

1-8 -

1-6 -

14'-1X2-

0-8 --\~\0-6 %-O~.0-40-2I I I I I 1,20 30 40 50 60 70 80 90Temperature (0)Fig. 3. Effect of rate of air flow and temperature on therate of removal of acetaldehyde. Experiments werecarried out in apparatus similar to that of Long (1946).Each apparatus contained 20pmoles acetaldehyde in10 ml. N sulphuric acid. Acetaldehyde was trapped inbisulphite and estimated iodometrically.VI I

VoI. 58 155S. R. ELSDEN AND Q. H. GIBSONTable 1. Recovery of lactic acid by aeration methodEach reaction tube contained a total volume of 10 ml. mixture which was normal with respect to H2SO4 and contained500juequiv. ceric sulphate; T = 600, aeration rate = 500 ml./min., t = 45 min.Expt. no. ... ... ... ... ...Zinc lactate added (as pmoles lactic acid)No. of blanksNo. of estimationsMean blank value (,tmoles lactic acid)Mean recovery corrected for blanks(,umoles lactic acid)Rangeand at temperatures from 20 to 80°. The results are givFig. 3. Over the range studied the rate of removacetaldehyde was found to be approximately proportto (air flow)0 75 and to the vapour pressure of acetaldewhich would exist over pure acetaldehyde at the tenature of the experiment.Working conditions for the aeration method. The resu]the above experiments show that, for a given air flovcreasing either the ceric sulphate concentration ortemperature will diminish the recovery of acetaldelThese experiments have enabled us to define woiconditions which permit a good recovery of acetaldefrom lactic acid. We recommend an initial concentraticeric sulphate of 0-05 N, a temperature of 50-60°, an aerrate of 500-600 ml. air/min./reaction tube and an aeritime of at least 45 min. Representative recoveriesthese conditions are given in Table 1.Steam distillation method

Although the aeration method gives satisfactorycoveries, the blanks were substantial, due to the contantion of the laboratory air with carbonyl compoundstherefore examined alternative ways of carrying oulreaction. The results described above show that to olthe greatest rates of formation and the smallest lossacetaldehyde, it is desirable to work at as high a temperEas is compatible with the maximum usable rate of aeriand the minimum concentration of ceric sulphate. Sdistillation seemed to offer the best possibility, andliminary experiments in which the reaction was carriesin the apparatus of Markham (1942) indicated that atthe reaction was complete in a very short time and thaacetaldehyde could be distilled over into bisulphite in a10 ml. of distillate. However, losses were encounteredthese we considered to be due to the fact that all thesulphate had to be added at once, with the result that, atemperature of the mixture, the amount of ceric suljpresent was sufficient to oxidize significant amounacetaldehyde. We therefore designed an apparatus in Mit was possible to add the ceric sulphate dropwise whilssteam was passing. The apparatus, which has prentirely satisfactory, is shown in Fig. 4. The folloreagents are required: (i) 0-05N ceric sulphate in N sulp]acid prepared from a stock solution of 0-5N ceric sulphaN sulphuric acid. The stock ceric sulphate is standarwith ferrous ammonium sulphate. We have foundsulphate (pure), Hopkin & Williams, to be satisfac(ii) 1ON sulphuric acid. (iii) 0-5% (w/v) sodium bisulsolution. (iv) 0-1N, 0-O1N and 0-005N iodine solution.15-65241-135-39211-3240-8711-2322-441-3722-24112-9240-97109-65-00-5-69 10-88-11-91 21-11-22-40 107-9-110-7,en inal ofional 25' ml.,hyde 20-aper- 15

Its of 10

v, inityde~~~yde. ~~~~~~~~~dDoauvbisletyspu

king

Steam condensehyde Sta inton of B40 jonFig. 4. Apparatus for the estimation of lactic acid by thesteam-distillation method.The sample containing lactic acid, which preferablyshould not exceed 5 ml., is measured into the reaction flaskand sufficient of the lON sulphuric acid added to make theconcentration of this acid normal. The reaction flask is thenattached to the apparatus. The receiving tube which contains2 ml. of the 0.5% (w/v) sodium bisulphite is attachedin such a way that the tip of the condenser dips below thesurface of the bisulphite solution. The micro-burnerheating the reaction flask is now turned on and the flameadjusted so that the solution just boils. The steam isthenturned on and the steam flow adjusted so that 15-20 ml./min. of distillate is collected (this rapid distillation rate isessential and necessitates the use of an efficient doublesurface condenser); the 0-05N ceric sulphate is then rundropwise into the reaction flask at such a rate that each dropis decolorized before the next drop goes in. When a permanentyellow colour is obtained, indicating that an excessof ceric sulphate is present, further ceric sulphate up to atotal of 5 ml. is rapidly added. The only critical feature of the156 I954ESTIMATION OF LACTIC ACIDTable 2. Recovery of lactic acid by the steam distillation methodIn Expt. 1, 0-002574N-12 was used for final titration; blank titration 0 08 ml. 0-002574N-12, and the results correctedfor this amount. In remaining experiments 0 01287N-I2 used and the blank was too small to be measured.Expt. no....No. of estimations

Lactic acid taken (jpmoles)Mean lactic acid recovered (,umoles)S.E.

Range162-172-14±0-022-08-2-162610-8510-64±0-0510-60-10-733621-7021-13

±0-0321-07-21-144632-5531-82+0-1031-61-31-87estimation is the dropwise addition of ceric sulphate. When15 ml. of distillate have been collected the receiver islowered and distillation is continued until 20 ml. has beencollected. The steam is then discontinued, the microburnerturned out and the receiver placed in an ice-waterbath to cool. While it is cooling, the reaction flask is disconnectedand thoroughly washed out with distilled water.The still head is also carefully washed to remove traces ofceric sulphate solution, which, if left, could bring about thepremature oxidation of part of the lactic acid in the nextsample to be analysed.The acetaldehyde is then estimated in the usual way, usingsolid NaHCO3to destroy the aldehyde bisulphite compound.Like Friedemann & Graeser (1933) we have found that it isimportant to cool the mixture thoroughly before the finaltitration; the temperature should be 4-5'. The results of aseries of recovery experiments are given in Table 2.Under the conditions of this method, glucose gives rise tocarbonyl compounds, and we have found it essential totreat the test solution with the copper-lime reagent; for thispurpose we use 1 ml. 20% (w/v) CUS04, 5H20/10 ml. testsolution and a spatula full of solid Ca(OH)2. This reagentwill also effectively deproteinize bacterial suspensions andrumen liquor.Trichloroacetic acid cannot be used, since under theconditions of the estimation it is oxidized with formation ofsuch large volumes ofgas that frothing is uncontrollable. Thevarious methods for the precipitation of proteins whichinvolve the use of tungstic acid have however provedentirely satisfactory. We have also found that perchloricacid (Neuberg, Strauss & Lipkin, 1944) is very satisfactorywhere an acid precipitation is required. If perchloric acidtreatment is followed by treatment with the copper-limereagent we have found it convenient to add the coppersolution followed by 2N-KOH until a slight trace of cuprichydroxide is formed, at which point the lime is added. Inthis way the perchloric acid is removed as the insolublepotassium salt. The removal of perchloric acid is notessential.Protein hydrolysates, such as are used in bacteriologicalmedia cause some interference, but no carbonyl compoundswere produced from ethanol, from citric acid or from malicacid.The results of an experiment in which the lactic acid,formed from malic acid by Lactobacillus arabinomsu underthe conditions described by Nossal (1951), was estimatedboth manometrically and by the steam distillation methodafter treatment of the mixture with copper-lime, are givenin Table 3. Table 4 shows the results obtained when a sampleof blood was analysed in triplicate by the steam distillation

method.Table 3. Estimation of lactic acid formed from DLmalicacid by the action of the malic decarboxylaseof Lactobacillus arabinosusThe malic decarboxylase converts L-malic acid quantitativelyinto lactic acid and CO2, and the amount of CO2formed is thus a measure of the amount of lactic acid. Inthis experiment a solution of DL-malic acid was treatedwith the decarboxylase preparation according to theprocedure of Nossal (1951), and at the end of the experimentthe cup contents were transferred quantitatively to ameasuring cylinder, treated with the copper-lime reagent,made up to 10 ml. and centrifuged; samples of the supernatantsolution were used for the estimation of lactic acidby the steam distillation method. Strength of iodine,0-002754N.Manometric methodSteam distillation method(a)(b)CO2(,moles)9-37Lactic acid(,umoles)9-379-358-92Table 4. Estimation of lactic acid in defibrinatedsheep's blood by the steam distillation methodThree 2 ml. samples of defibrinated sheep's blood weretaken and 7 ml. of distilled water added to each followed by2 ml. 15% (wfv) perchloric acid. Precipitates were centrifugedoff and supernatants filtered through cotton woolinto graduated centrifuge tubes; 0-5 ml. 20% (w/v)CuSO4, 5H20 was added and 40% (w/v) KOH addeddropwise until a slight permanent blue precipitate obtained,a spatula full of solid Ca(OH)2 (approx. 1 g.)added, the contents of the tubes were mixed and thevolumes noted. After 30 min. tubes were centrifuged and5 ml. samples taken for analysis.Lactic acidSample ,umoles/100 ml. blood123193197200DISCUSSIONOur observations on the reaction between cericsulphate and lactic acid confirm and amplify thosepublished by Long (1946). Significant amounts ofthe acetaldehyde produced are oxidized to aceticacid if the former is not rapidly removed from thereaction mixture. The rate of destruction ofacetaldehyde is related to the concentration of cericVoI. 58 157

158 S. R. ELSDEN AND Q. H. GIBSON I954sulphate in the reaction mixture and to the temperature.Consequently, to obtain maximum recoveryof acetaldehyde, the concentration of ceric sulphateshould be such as to ensure only a slight excess atthe end of the reaction. The rate of aeration shouldbe the maximum possible. From our results itwould appear that the concentration of ceric sulphaterecommended by Long (1946) is somewhat toohigh.In view of our findings it is difficult to understandhow a quantitative conversion of lactic acid toacetaldehyde can be obtained by the procedure ofWinnick (1942). This author recommends a mixtureof 3 ml. sample and 1 ml. saturated ceric sulphateand the reaction is carried out in Conway vessels andthus relies upon diffusion to remove the acetaldehyde.Since the purity of his ceric sulphate is notstated we can only guess at the final concentration,but it would seem to be in the region of 0 2N.Likewise the modification of Winnick's proceduredescribed by Conway (1950) is open to seriouscriticism in the light of the observations of bothLong (1946) and ourselves. In this method thereaction is carried out in stoppered test tubes, underconditions somewhat similar to those used both byLong and ourselves to study the oxidation ofacetaldehyde by ceric sulphate.It is possible to remove the acetaldehyde formedfrom lactic acid, almost quantitatively, either byaeration or by steam distillation, and we haveexamined both procedures. We prefer the steamdistillation method for the following reasons. Theblanks are negligible save when using 0-00257Niodine; but even here they are small and areequivalent to about 011,mole lactic acid. Themethod is extremely rapid, one distillation takinglittle more than a minute to perform. The end pointis sharper because the final volume in which thetitration is carried out is only 30 ml. and this isparticularly important when small amounts oflactic acid are to be estimated. The advantages ofthis ceric sulphate method over the permanganatemethods, described by Friedemann, Cotonio &Shaffer (1927), Friedemann & Kendall (1929), andFriedemann & Graeser (1933) are speed and thesmall volume in which the titration is carried out.SUMMARY1. The effect of ceric sulphate concentration andtemperature on the oxidation of lactic acid by cericsulphate have been investigated.2. The effects of temperature and rate of aerationon the removal of acetaldehyde from solutions have

been investigated.3. Conditions for the estimation of lactic acidwith ceric sulphate using the aeration method aredescribed.4. A new method for the estimation of lacticacid using ceric sulphate and the removal ofacetaldehyde by steam distillation is described.The authors wish to thank Mrs M. Martin for technicalassistance and Mr Hatfield for making and helping with thedesign of the apparatus.REFERENCESConway, E. J. (1950). Microdiffusion Analyysis and VolumetricError, 3rd ed. p. 241. London: Crosby Lockwoodand Son Ltd.Friedemann, T. E., Cotonio, M. & Shaffer, P. A. (1927).J. biol. Chem. 73, 335.Friedemann, T. E. & Graeser, J. B. (1933). J. biol. Chem.100, 291.Friedemann, T. E. & Kendall, A. I. (1929). J. biol. Chem. 82,23.Gordon, J. J. & Quastel, J. H. (1939). Biochem. J. 33, 1332.Long, C. (1946). Biochem. J. 40, 27.Markham, R. (1942). Biochem. J. 36, 790.Neuberg, C., Strauss, E. & Lipkin, L. E. (1944). Arch.Biochem. 4, 101.Nossal, P. M. (1951). Biochem. J. 50, 349.Winnick, T. (1942). J. biol. Chem. 142, 451.

Active Transport of Sodium Ions from the Yeast CellBY E. J. CONWAY, H. RYAN AND ENID CARTONDepartment of Biochemi8try, Univer8ity College, Dublin(Received 12 November 1953)It has been shown (Conway & Hingerty, 1948) thatsodium ions which had entered muscle extensivelyin rats on a potassium-free diet were activelyextruded on changing the animals to a potassiumrichdiet. The mean half period of extrusion, evenwith the plasma potassium above the normal levelwithin 24 hr. after the change, was about 3 days.Very recently, Desmedt (1953) has described a morerapid net excretion of sodium ions from the isolatedfrog sartorius in which sodium ions had enteredextensively after immersion in a potassium-freeRinger solution in the manner described bySteinbach (1951). To demonstrate the excretion,companion muscles which were similarly treated

A RAPID COLORIMETRIC METHOD FOR THE DETERMINATION

OF LACTIC ACID IN MILKBURDET IIEINEMANNChief Chemist, Producers Creamery Company, Springfield, Mo.~[any methods for determining lactic acid in biological fluids have beenreported. Several have been applied successfully to dairy products. Mostof these methods, however, are too time consuming to be used regularly inthe control laboratory. A search was made, therefore, for a method whichwas both rapid and reliable: A method using a simple procedure for theremoval of interfering substances; a simple oxidation procedure; and asimple means of estimating the quantity of lactic acid present.The method Mendel and Goldscbeider (5) used for determining the lacticacid in blood seemed to be the most promising, but when applied to milkgave unreliable results. Many variations were tried, one of which is reportedbelow. The method for removing interfering substances is almostidentical with that described by Troy and Sharp (7). Their work on precipitationprocedures was repeated to determine the most satisfactory techniqueto be used in conjunction with the eolorimetric method, but no changewas found necesssary.PROCEDUREWeigh 5 grams of milk in a 50 ml. volumetric flask. Add 30 ml. ofcold water and mix. Dilute a portion of a 25 per cent solution of coppersulphate 1 : 3 and add dropwise with agitation 2-8 drops of the dilute solutionto the water and milk mixture until precipitation occurs. Warm theflask in hot water until the contents reach 45 ° C. Add 6 ml. of 25 per centcopper sulphate solution and mix thoroughly by rotation. Hold the contentsof the flask at 45 ° to 47 ° C. for 8 to 10 minutes. Next loosen any curdparticles which may adhere to the flask with a bent wire. Add 6 ml. of calciumhydroxide suspension. This suspension is prepared by slaking 300grams of the best quality calcium oxide using 1400 ml. water and workingthrough a 24 mesh screen after 30 minutes standing. Before mixing thelime with the other constituents in the flask, water should be added to bringthe contents to exactly 50 ml. Then mix until the contents are homogeneousand allow to stand at 45 ° to 47 ° c. for 10 minutes. Cool to 25 ° C. and filter,discarding the first few ml. of filtrate.Place 1.0 ml. of the filtrate in a thoroughly clean and dry test tube.Hold the test tube in ice water and allow 5.0 mI. of cold concentrated sulphuricacid (arsenic free) to run down the side wall of the test tube. Mixgently while cooling. Place mixture in boiling water for exactly 5 minutes~Received for publication March 28, I940.969970 BURDET HEINEMANN

Cool in ice water for three. Add 4 drops of a 0.1 per cent solution of veratrolein water and mix. Hold in ice water for 60 minutes and compare colors withartificial or natural standards.Natural standards may be prepared according to Hil!ig (3). Dissolve106.6 rag. of purified lithium lactate in 100 ml. of distilled water. 1.0 ml.of this solution contains the equivalent of 0.1 per cent lactic acid. Consequently,to prepare a 0.01 lactic acid standard, 5.00 ml. of this solution isdiluted to exactly 50.0 ml. and mixed. One ml. of this dilution is thenplaced in a test tube and 5 ml. of concentrated sulphuri¢ acid added. Themixture is held in boiling water 5 minutes and cooled in ice water 3 minutes.Four drops of a 0.1 per cent solution of veratrole is then added and the colorobserved after 60 minutes. Additional standards may be made by dilutingthe original lithium lactate solution as required.Artificial standards which are fairly satisfactory may be prepared accordingto the method of Nordb5 (6). Mix 25.0 ml. of a 0.01 per cent solutionof fuchsin in water and 4.75 ml. of a 0.01 per cent solution of Tropaolin000. (The mixture must be diluted immediately for a precipitate forms onstanding.) One tenth ml. of this mixture diluted to 50.0 ml. of solution is

equivalent to approximately 0.005 per cent lactic acid; 0.25 ml. diluted to 50ml. is equivalent to approximately 0.01 per cent and ~ 0.9 ml. diluted to 50ml. is equivalent to 0.02 per cent. NordbS's standards are only suitable forconcentrations below 0.03 per cent. Above this percentage, the yellow colorincreases in intensity. Consequently, for a 0.05 per cent lactic acid standard,it is necessary to add 0.4 ml. of .01 per cent Tropaolin and 1.8 ml. of.01 per cent fuchsin to 47.8 ml. of water; for a 0.07 per cent standard, 1.0ml. of Tropaolin and 2.0 ml. of fuchsin to 47.0 ml. of water; and for a 0.10per cent standard, 2.0 ml. of Tropaolin and 2.0 ml. of fuchsin to 46.0 ml. ofwater. These artificial standards should be checked against natural standardsbefore using.DISCUSSION

According to the authors, the original test which was applied to blood isaccurate to -+- 0.004 per cent between 0.001 and 0.05 per cent (5) and 1 partin 400,000 lactic acid could be detected. Mendel (4) later made two improvementswhich he stated made possible the detection of one part in1,000,000. I-Iowever, when applied to milk, the original test was sensitiveto 0.01 per cent and had a limited degree of accuracy (± 0.04 per cent).With the modification outlined above, the test is sensitive to 0.005 per centand accurate to ± 0.004 per cent between the ranges Of 0.005 and 0.12 percent lactic acid providing the procedure is closely followed.Mendel (4) has pointed out that small changes in the concentration ofthe sulphuric acid may lead to inaccurate results. A difference of 2 to 3per cent in the concentration may result in as much as a 15 to 20 per cent

DETERMINING LACTIC ACID IN MILK 971error. It is wise, therefore, to check frequently the color of the artificialstandards. In this respect, it should be noted that the individual analyst

should try several different quantities of sulphuric acid (i.e., 4.5, 4.75, 5.0ml. etc.,) in the preparation of natural standards in order to ascertain theexact amount to add to the 1.0 ml. filtrate to obtain the greatest degree ofcolor development for each batch of sulphuric acid. Derviz (1) describes amethod for purifying and storing sulphuric acid so as to maintain a constantconcentration.1Vfendel (4) makes the recommendation that after the mixture of acid andfiltrate has been held in boiling water five minutes, it be cooled 2 minutes inice water, veratrole added, and set at 25 ° C. After 20 minutes, the colormay be compared with standards similarly treated. This method may beemployed whenever results of great accuracy are not desired. ~Mendel and Goldscheider (5) used 0.1 ml. of a 0.125 per cent veratrolcin 99.8 per cent alcohol. Other investigators suggested 0.05 ml. of a 1 : 800solution in 96 per cent alcohol (6), 0.1 ml. of 20 per cent veratrole solutionin glacial acetic acid (1) and 0.1 ml. of 20 per cent in absolute alcohol (2).These were all tried, but none gave as satisfactory results as 4 drops of a0.1 per cent solution in water. Various solutions of guaicol, hydroquinone,and Schiffs reagent were also tried and considered unsatisfactory.Although the original method called for a 4 minute heating period, 0.5ml. filtrate and 3.0 ml. of conc. tt2SO, it was found that 5 minute heatingperiod, 1.0 ml. filtrate and 5.0 ml. of concentrated sulphuric acid gave moreuniform results.The same substances interfere with this method as those Troy and Sharp(7) found to interfere with theirs. Formaldehyde, overneutralization,rancidity, sucrose, and products resulting from the heating of milk at ornear the boiling point for one half hour or longer resulted in high values.Mendel and Goldscheider state, however, that acetone, ~-oxybutyric acid,acetic acid, urea, uric acid, creatin, creatinin, glycocol, alanine, and propionicacid, do not interfere with the color development.Traces of organic matter interfere with proper color development and

consequently the test tubes must be thoroughly clean and kept stopperedas much as possible during the determination.When greater rapidity of testing is required, the precipitate of proteins,lactose and fat, may be centrifuged. In this case, cream test bottles (18gram body, 9 gram neck, graduated in 1 per cent between 0 and 55 per cent)were selected from stock and graduated to contain 50.0 ml. at 20 ° C. Themilk was weighed in these bottles which, after precipitation and cooling,were placed in a Babcock tester and centrifuged 4 minutes. The centrifugatewas then decanted through a fluted filter, the first few ml. of filtratebeing discarded.

972 BURDET HEINEMANNSUMMARYA rapid colorimetric method for the quantitative estimation of lacticacid in milk is described. The steps involved are precipitation of fat, lactoseand protein with copper sulphate and calcium hydroxide, filtering,adding concentrated sulphuric acid (arsenic free) to the filtrate, heating,cooling, and adding veratrole. A red color develops on standing which is inproportion to the amount of lactic acid present.The test is sensitive to about 0.005 per cent and is accurate to ___ 0.004per cent between the ranges of 0.005 and 0.12 per cent lactic acid.:REFERENCES(1) DERWZ, G.V. The colorimetric lactic acid determination according to Mendel-Goldscheider.Zhur. Exptl. Biol. Med. 12: 147-50. 1929. (Original not seen.See Chem. Abstr. 24- 1880. 1930.)(2) FUCHS, HANS J. Einige Verbesserungen der kolorimetrischen Milchsfiurcbestimmungnach Mendel und Goldscheider. Biochem. Ztschr. 217: 405-8. 1930.(3) HILLIG, FRED. The colorimctric determination of lactic acid in milk and milk products.J.A.O.A.C. 20: 130-40. 1937.(4) MENDEL, BRUNO. Zur Methode der k01orimetrischen Milchs~urebestimmnng. Biochem.Ztschr. 202: 390. 1928.(5) MENDEL, BRUNO AND GOLDSCYIEIDER, INGEBORG. Eine kolorimetrische Mikromethodezur quantitativen Bestimmung der Milchsiiure im Blut. Biochem. Ztschr. 164:163-74. 1925.(6) NORDBh, I:~AGNVALD. Zur Methode der Milchs~urebestimmung nach Mendel und Goldscheider.Biochem. Ztschr. 271: 213-215. ]934.(7) TROY, It. C. AND SHARP, P. F. Quantitative determination of lactic acid in dairyproducts. Corner Univ. Ag. Exp. Sta. Memoir 179. 1935.