IN the estimation of sugars in complex mixtures, such as occur in

Transcription

1 THE NEW PHYTOLOGIST VOL. XXXV, No. i 27 February, 1936 THE ESTIMATION OF SMALL QUANTITIES OF FERMENTABLE SUGARS BY CARBON DIOXIDE PRODUCTION BY W. O. JAMES AND A. L. JAMES Department of Botany, Oxford (With 2 figures in the text) IN the estimation of sugars in complex mixtures, such as occur in plant extracts, fermentation methods have been found indispensable and are used by all workers. The procedure commonly adopted is to ferment away the sugar mixture completely and to estimate the loss of ability to reduce alkaline cupric copper, ferricyanide, etc. The residual reduction is regarded as being due to non-sugars and is deducted from the initial reduction determined before fermentation to correct the estimate of fermentable sugars present. Such a method is satisfactory allowing two assumptions. In the first place it must be assumed that the yeast destroys only sugars among the suitable reducing substances present, e.g. there must not be appreciable quantities of fermentable phosphoric esters, glyceraldehyde, etc., in the mixture. In the second place the yeast must not give rise during the removal of the sugars to substances capable of reducing copper or ferricyanide. It is usually safe with plant extracts to assume that the sugars present will greatly outweigh the other fermentable substances, and errors due to the first assumption are not likely to be important. The second assumption may be more troublesome, particularly if the "unfermentable reducing power" is an appreciable fraction say 20 per cent, or more of the whole. There is usually no means of telling how much of this fraction was present at the start and how much has arisen during the period of fermentation. In the course of some experiments on barley embryos this difficulty was met in an acute form. After 3 hours' fermentation with bakers' yeast (in an PHYT. XXXV. I I

2 2 W. O. JAMES AND A. L. JAMES air stream to avoid the formation of alcohol) the reducing power of the extracts, so far from disappearing completely, was actually considerably greater than at the start. The following results were obtained for the reduction of alkaline ferricyanide after the method of Hagedorn and Jensen. TABLE I. Reducing power of extract from ioo embryos, ex. thiosulphate in the Hagedorn-Jensen titration Before After fermentation fermentation Increase N/ioo Such results can only be due to the formation of suitable reducing substances by the yeast in excess of any sugars simultaneously destroyed. In the present instance there is a strong possibility that the substance formed is acetyl methyl carbinol ("acetoin"), CH3CHOH.CO.CH3. It is already known to be formed by yeast, especially when strongly aerated, as is necessary in the above experiments to avoid the formation of alcohol (Elion, 1927). It reduces Fehling's solution strongly. An attempt was made to detect the appearance of new carbonyl groups, which would necessarily happen if acetoin was formed, by measuring the bisulphite binding capacity of the solution before and after fermentation with the following result. TABLE II. Mg. bisulphite bound by extract from 100 embryos, ex. N/ioo iodine bisulphite bound Before fermentation 4-85 After fermentation 9-37 Increase 4-52 This increase if due to acetoin is more than would be required to account for the simultaneous increase of reducing power, as is to be expected if some sugar is being destroyed at the same time. Negative results with the "dimedon" test (Simon and Neuberg, 1932) and Schiff's reagent also suggested that the bisulphite compound might be formed by a ketone rather than an aldehyde. As a specific test for the presence of acetoin the usual method of oxidation to diacetyl, distillation of the diacetyl and subsequent formation of the nickel dimethylglyoxime (Kluyver, Donker and Visser 't Hooft, 1925) was employed; this test is very sensitive. A shght reaction indicated by a blue coloration was obtained with the extract before fermentation, agreeing with the result of Lemoigne and Mongouillon (1930) who have previously shown that acetoin is

3 Estimation of Fermentable Sugars by Carbon Dioxide 3 formed during the germination of barley grains. With the extracts after fermentation a much stronger reaction (a red precipitate) was developed. Table I shows that one or more substances able to reduce actively in alkaline solutions were formed during the fermentation of the extracts. This makes the combination of reduction and fermentation methods vmsatisfactory with these and perhaps other plant materials. There seems to be no alternative reagent to yeast for picking out fermentable sugars, but the difficulty might be avoided by estimating the carbon dioxide evolved during the fermentation instead of the change of reducing power. Several advantages might be expected in theory, (i) The results would be independent of other reducing substances present; (2) there would be no objection to the formation of alcohol in moderate quantities, and hence the use of aerobic conditions and the consequent risk of intrusive oxidative processes could be avoided; (3) acetoin even if formed does not influence the carbon dioxide output since decarboxylation occurs at an earlier stage of the reactions, the following equations having a high degree of probability, as representing the changes brought about by yeast: CH3CO.COOH -^CHgCHO + CO2, 2CH3CHO ^CHjCHOH. CO. CH3. It has also been shown directly that the presence of acetoin does not influence the first reaction (Neuberg and Kobel, 1925). The following section gives the details of a method we have worked out for small quantities of sugars based on these principles. MATERIALS USED (1) Yeast. A fresh supply of pressed bakers' yeast is obtained daily or a stock may be kept for a few days in an ice-chest. 10 gm. are weighed out, washed with distilled water and thrown down in a centrifuge; a single washing is sufficient. After decanting off the washings fresh water is added, the yeast stirred into a suspension and made up in a measuring flask to 100 c.c. (2) N/100 HCN. This is prepared by weighing out pure potassium cyanide, dissolving and adding 50 per cent, hydrochloric acid until a small piece of litmus paper included in the solution becomes purple red. The solution is then made up to the required volume and may be kept as stock more or less indefinitely. (3) Nitrogen. Commercial compressed nitrogen is used. This contains about i per cent, of oxygen. To remove this is troublesome.

4 4 W. O. JAMES AND A. L. JAMES and it would also be difficult to avoid the introduction of traces of oxygen from the outside air during manipulation. These precautions become unnecessary when fermentation is carried out by the technique described below. (4) Baryta solution. This is made up at approximately iv/150 by the usual methods. (5) Standardised hydrochloric acid. A stock of approximately normal hydrochloric acid is prepared and diluted two hundred times. The roughly N/200 acid is standardised accurately as follows. An accurate Njioo sodium carbonate solution is prepared by weighing out oven-dried (or ignited) anhydrous sodium carbonate (we used B.D.H. " analar"). 5 c.c. are measured accurately into a boiling tube and a little brom thymol blue added as indicator. The solution is raised to boiling point over a small flame and the hydrochloric acid run in cautiously from a microburette. The carbon dioxide evolved is boiled off and further hydrochloric acid run in until the colour after boiling indicates ph 7-0. When the end-point is approached the indicator colour is tested against a standard in a comparator. If necessary further additions of acid are then made. Owing to the nature and dilution of the reagents accurate titration to pli 7-0 is essential. Once the strength of the acid has been determined in this way, the baryta solution can be titrated against it, and it is more convenient to standardise subsequent batches of acid against the known baryta, provided that this is kept secure from atmospheric carbon dioxide. (6) I per cent, alcoholic phenolphthalein as indicator. THE FERMENTATION Before the estimation proper the yeast is shaken in a bath (about 50 oscillations per minute) at 35 C. for i hour, i c.c. aliquots of the yeast suspension (p. 3) s 100 mg. of yeast are transferred to fermentation tubes of the form shown in Fig. i. 3 c.c. of distilled water and i c.c. iv/ioo HCN are added, making a total of 5 c.c. in the tube. The preliminary starvation of i hour serves to reduce the rate of spontaneous carbon dioxide formation (" autofermentation ") to a half or less of its initial value. It is naturally desirable that the amount of carbon dioxide due to this cause should be small relative to the amount produced from the added sugar. Another advantage is the adjustment of the suspension to the standard temperature before the introduction of the sugar. When the yeast suspension has been starved for an hour it is

5 Estimation of Fermentable Sugars by Carbon Dioxide 5 ready for use. The tube contaitiing it is removed from the bath (p. 4) and exhausted by attaching the outlet.b to a filter pump via the exhaustion chamber. Fig. 2, E. This removes the carbon dioxide now present in the fermentation tube, and the last traces are swept out by closing the clip e and allowing nitrogen to enter from the aspirator by opening the clip a, until the pressure in E returns to the Fig. I. atmospheric. E should be about c.c. capacity ( = 7 or 8 times that of F). The tube is now detached from E and nitrogen allowed to flow gently through it while i c.c. Njioo HCN is run in through B followed by a known volume, 2 or 3 c.c, of the sugar solution and distilled water to make a total volume of 10 c.c. including the 5 c.c. already in the tube. Care must be taken that the solutions are introduced cleanly into the cavity of the fermentation tube, and for this a burette with a lengthened narrow jet is very useful. The

6 W. O. JAMES AND A. L. JAMES From baryta reservoir' From nitrogen supply To filter pump Fig. 2.

7 Estimation of Fermentable Sugars by Carbon Dioxide 7 burette is then withdrawn slowly, the clips a and b immediately closed and the fermentation tube put back into the water bath at 35 C. and shaken for 2 hours. The total volume of liquid in the tube should now be 10 c.c. and will be made up as follows: 1 c.c. yeast suspension ( = 100 mg. yeast). 2 c.c. TV/ioo HCN (final concentration = A/^/5oo HCN). X c.c. sugcir solution. 10 (A; 4-3) c.c. distilled water. When a number of experiments are to be performed consecutively it is wise to have min. intervals between starting one tube and the next. METHOD OF TITEATION The carbon dioxide formed in the fermentation tube is estimated by precipitation with baryta and titration of the residual alkali; the estimation is carried out in a carbon dioxide free atmosphere in the apparatus shown in Fig. 2. The fermentation tube is connected up as shown and the chamber E exhausted while the clip b is still shut. In setting up B a few drops of phenolphthalein are included and 25 c.c. of baryta solution are run in from the automatic pipette (G) while exhaustion is in progress. Clip e is then shut and the carbon dioxide evolved in F is drawn over by opening b and then a so that carbon dioxide free nitrogen flows into E through the liquid in the fermentation tube (F). The whole of the carbon dioxide in F including that in solution is thus swept over into E (see p. 8). The whole apparatus is mounted on a board hinged to a stand at the top and free at the bottom. This is necessary so that the baryta solution in E can now be well shaken up to absorb the carbon dioxide above it. A simple mechanical shaking device is desirable. Five minutes' fairly vigorous shaking is enough. The residual baryta is now titrated with the standard acid run in from the burette C. The carbon dioxide value is obtained by deducting the acid equivalent of the residual baryta so obtained from the corresponding value for the full amount of baryta. From this the equivalent weight of sugar may be simply calculated relating carbon dioxide to hexose sugar by the fermentation equation I c.c. A^/200 HCl = mg. CeHijOg.

8 8 W. O. JAMES AND A. L, JAMES TESTS OF THE METHOD Estimation of carbon dioxide. The accuracy of the carrying over and absorption of fhe carbon dioxide was tested as follows. 2 c.c. of standard Njio sodium carbonate solution were put into a fermentation tube. An excess of phosphoric (or sulphuric) acid was put into a small tube which was gently inserted. The stopper was then put in and the carbon dioxide liberated from the carbonate by spilling the strong acid into it. When the reaction was complete the carbon dioxide was drawn over and shaken up with the baryta in the usual way. TABLE III CO2 in 2 c.c. Na2CO3 calculated 4-40 mg.,, estimated 15 min. shaking 4'38,, 15., 4-43 >.,, '4i ,, The error of the estimation in every case is less than i per cent. Correction for carbon dioxide of autofermentation. Estimations of the carbon dioxide given off without the addition of external sugars were carried out with several batches of yeast at various times. The mean value of these was used as a deduction from all experimental values. This procedure is less laborious than carrying out a control value with every experiment. Applying the latter method did not lead to any increase of accuracy; rather the reverse indeed, since the individual values recorded for autofermentation in the earlier experiments were less regular than those with added sugars. With practice the irregularity disappeared. The values obtained (calculated as equivalent hexose) were 0-46, 0^04, 0-26, 0-24, 0-34, 0-26, Average 0-27 mg. The use of HCN. According to Meyerhof (1925) HCN in concentrations between A^/500 and 7V/iooo almost entirely suppresses oxidative effects in bakers' yeast without markedly retarding the rate of fermentation. This applies with the normal atmospheric percentage of oxygen. Using A^/300 HCN Dixon and EUiott (1929) found that inhibition in the presence of sugars was 92 per cent., but only about 85 per cent, for autofermentation even with the cyanide 10 times as strong. In our own experiments using iv/500 HCN the COg yield in air was only 82 per cent, of the theoretical, but with i per cent, oxygen (commercial "nitrogen ") the yield in a companion experiment rose to 98-8 per cent. Experiments with varying concentrations of HCN gave the following results (Table IV).

9 Estimation of Fermentahle Sugars by Carbon Dioxide 9 TABLE IV. Fermentation in nitrogen containing i per cent, of oxygen. Yields of carbon dioxide as percentages of the theoretical, mg. sugar supplied HCN N/iooo N/joo. N/^oo N/ Approximate percentages only are given for concentrations of HCN other than N/500, as but few autofermentation readings were taken. The results show that vnth quantities up to 5 mg. of sugar good estimations may be made using A^/500 HCN. At the lower concentrations, e.g. with about i mg. sugar, the error is still small but naturally increases as a percentage. The most satisfactory range lies between 3 and 5 mg., where the percentage error is about i. With quantities greater than 5 mg. low results are recorded, and longer fermentation times or more yeast would be required. Example of an estimation. The fermentation tube contained: 1 c.c. yeast suspension (=100 mg. yeast). 2 c.c. Njioo HCN. 4 c.c. glucose solution ( = 3-84 mg. glucose). 3 c.c. distilled water. 10 c.c. Titration: 25 c.c. baryta = 47-2 c.c. HCl (o-o6o N). After absorption s c.c. HCl..". CO2 s c.c. HCl. = X mg. glucose mg. glucose mg. glucose: autofermentation correction mg. glucose fermented = per cent. Estimation of sucrose and maltose. These disaccharides are fermented by bakers' yeast, but using the technique described above very low values were obtained. To obtain complete fermentation more yeast would therefore be required. This would inevitably raise the autofermentation value to a considerable percentage of the whole CO2 output. It is, therefore, considered better to hydrolyse the disaccharide first and estimate it in the hexose form.

10 10 W. O. JAMES AND A. L. JAMES REFERENCES DixoN, M. and ELLIOTT, K. A. (1929). The effect of cyanide on the respiration of animal tissues. Biochem. J. 23, 812. ELION, L. (1927). Die Acetoinbildung bei alkoholischer Zuckergarung. Chem. Z. 98, 2 (i), KLUYVER, A. J., DONKKR, H. J. R. and VISSER 'T HOOFT, F. (1925)- Uber die Bildung von Acetylmethylcarbinol und 2-3 Butylenglykol im Stoffwechsel der Hefe. Biochem. Z. 161, 361. LEMOIGNE, M. and MONGOUILLON. P. (1930). Presence de l'acetylmethylcarbinol et du 2-3-butyl6neglycol chez les plantes superieures. Formation au cours de la germination. C.R. Acad. Sci., Paris, 190, MEYERHOF, O. (1925). tlber den Einfluss des Sauerstoffs auf die alkoholischer Garung der Hefe. Biochem. Z. 162, 43. NEUBERG, C. and KOBEL, M. (1925). Uber das physiologische Verhalten des Acetoins. Biochem. Z. 160, 250. SIMON, E. and NEUBERG, C, {1932). Aldehyde and ketone. In Klein's Handbuch der Pflanzenanalyse, 2, 5, 261.

11