ENZYMATIC QUANTITATIVE DETERMINATION OF HEXOSES, SINGLY AND IN MIXTURES WITH THEIR OLIGOSACCHARIDES

Quantitative determinations of a number of naturally ocurring D-hexoses and their oligo saccharides can be accomplished by combinations of enzymatic methods without prior separa tion of the sugars. T hese spectrophotometric methods employ readily-available enzymes and &ommercial reagents. The procedures are particularly useful for the quantitative analysis of fructofuranosides, such as raffinose or stachyose, in the presence of sucrose, glucose and/or fructose; maltose or lactose mixed with galactose and/or glucose; melibiose in mixtures with raffinose, and other combinations.


INTRODUCTION
Many chemical and enzymatic methods have been described for qualitative and quantitative determination of naturally-occurring sugars (Aminoff et al., 1970;HoUigan, 1971;Holligan and Drew, 1971; Lewis and Smith, 1967). The latest employ gas-liquid chromatography (GL C). Since GLC is not yet widely available ' to botanists, this paper presents modified enzymatic methods for the rapid quantitative determination of mixtures of oligosaccharides and their component monosaccharides. These methods, which employ readily-available enzymes and commerical reagents, have been used in studies of carbohydrate physiology of orchid seedlings (Ernst, 1967;Ernst, Arditti and Healey, 1972). MATERIALS AND METHODS in saturated anunonium sulphate and a prepared galactose oxidase reagent, Galactostat, were purchased from Worthington Biochemical Corporation, Freehold, New Jersey.
Spectrophotometry. Spectrophotometric measurements were conducted in a Beckman DB-G instrument at 340 nm and I cm-light path. Mixing was by several gentle inversions. The cuvette chamber was maintained at 25° C to prevent bubble formation. Calculations were based on the micromolar extinction coefficient of 6.22 cm 2 for NADH 2 and NADPH 2 (Horecker and Kornberg, 1948).
Replication. All hydrolyses and determinations were replicated four times. The results reported are averages of these.

Fructose
Fructose was determined by converting it mto glucose-6-phosphate with PGI (Schmidt, 1961) added to the glucose reagent. This enzyme mixture is referred to as the 'fructose reagent'. Thus, following glucose determination, 5 µl PGI (about 9 U) were added to each 3 ml of the glucose assay reagent containing the sugar sample. Stable absorbance values and quantitative recoveries of known amounts of sample were obtained after 15 minutes at 25° C ( Fig. 1) when fructose was alone or in mixture with glucose, sucrose or raffinose.

Raffinose
Hydrolysis of raffinose was carried out under conditions recommended for inversion of sucrose (Bergmeyer and Klotzsch, 1965). Aqueous solutions of the trisaccharide, Raffinose. Points representing recovery were obtained following r-hour incubation with invertase. Reaction rates were determined with 0.1 M solutions. Fig. 3. Stachyose. Points representing recovery were obtained following r-hour incubation with invertase. Reaction rates were determined with o.r M solutions. Fig. 4. Maltose. Points representing recovery we!'e obtained following 1-hour incubation with cx-glucosidase. Reaction rates were determined with 0.01 M solutions. Fig. 5. Lactose. Points representing recovery were obtained following 75-minutes incubation with P-galactosidase in glucose reagent. Reaction rates were determined with 10 µ1 of 0.025 M solutions in 3 ml glucose reagent. Fig. 6. Melibiose ( 4 ) and raffinose (e ). Points representing recovery were obtained following t-hour incubation with Galactostat reagent. ranging in molarity from 0.025 to 0.1, were incubated with 5.2 times their volumes of 0.1 M sodium acetate buffer (pH 4.6) containing 0.77 mg/ml invertase, at 37° C, for I hour. Although melibiose samples (0.1 M) were included to check for possible melibiase activity, the hydrolysed samples yielded only melibiose and fructose by TLC and glucose was also shown to be absent by enzymatic analysis. Hydrolysis was complete in 20 minutes and recovery corresponded to 96.4-98.7% (Fig. 2).
The quantity of raffinose (measured as fructose) contained in 20 µI of hydrolysed sample added to the fructose reagent is derived as follows, where O.D. Fru is observed change in optical density, total volume of reagents in cuvette = 3 ml and micromolar extinction coefficient for NADPH 2 = 6.22:

Sucrose/raffinose mixtures
Hydrolysis of sucrose and raffinose mixtures under the conditions outlined above will result in a hydrolysate containing glucose (from sucrose) and fructose (from sucrose and raffinose). Fructose values in excess of double the optical density obtained for glucose represent hydrolysed raffinose. Therefore with O.D. Glc = observed change in optical density with glucose reagent, and other factors as above, the following calculations apply: Aqueous and methanol-water solutions of 0.025-0.075 M sucrose and 0.075-0.025 M raffinose, as well as o.x M sucrose only and 0.1 M raffinose only, were hydrolysed as for raffinose. TLC of samples taken after l hour showed absence of sucrose and raffinose and presence of melibiose, glucose and fructose. The recovery of sucrose was 97.2-xor.7% and that for raffinose 95.7--<)8.7% when analysed with the glucose and fructose reagents. Presence of methanol had no adverse effect on this method (but see maltase assay below).

Stachyose
Hydrolysis of stachyose was carried out as for rafiinose. Hydrolysis was slower than for raffinose and the optical density obtained with the fructose reagent levelled off after about 40 minutes (Fig. 3). Recoveries ranged from 93.5--<)6.7% (Fig. 3).

Sucrose/stachyose mixtures
Hydrolysis of mixed aqueous sucrose and stachyose soiutions was carried out as for sucrose/raffinose mixtures. TLC of samples taken after hydrolysis showed absence of sucrose and stachyose, but the presence of manninotriose, glucose and fructose. In assays of mixtures in various proportions as for sucrose and raffinose, recovery ranged from 98':7 to 104.6% for sucrose and 94.2 to 95.7% for stachyose Calculations for sucrose are those given for sucrose/rafiinose mixtures, whereas stachyose recovery was calculated as follows: . Maltose was measured as glucose following hydrolysis. Aqueous maltose solutions (0.025-0.1 M) and maltose dissolved in methanol-water were hydrolysed with maltase as follows: 50 µl maltose solution, 400 µl, o. r M sodium acetate buffer (pH 6.o ), and 250 µl maltase suspension (20 U) were incubated at 37° C for r hour. Samples of glucose only (o.r M) dissolved in water or methanol-water were included as controls. Unreacted maltose samples showed the presence of 0.09% glucose (by enzymatic procedure) as impurity, and minor amounts of a higher oligosaccharide (probably maltotriose) by TLC ( Table l ). On completion of hydrolysis, TLC of aqueous solutions showed the presence of glucose only. With methanol-water solutions, a second spot appeared with a higher RF value than glucose. Since a-methyl-D-glucoside (meGlc) can be synthesized by action of maltase from yeast where methanol is present (Bourquelot, Herisse and Bridel, 1913) we tested for its presence chromatographically. The data show that meGlc was present in the hydrolysate ( Table l). Hydrolysis of aqueous maltose solutions resulted in stable optical density values after 40 minutes (Fig. 4). Methanol-water solutions gave a recovery of only 67.5% with maltose and of 69.2% with glucose.

Lactose
Hydrolysis and estimation of lactose was by an adaptation of a previous method (Reithel and Venkataraman, 1956). To each 3 ml of glucose reagent were added 10 µl of 0.0125-0.25 M aqueous lactose solutions, free of glucose. Following the addition of 44 units of P-galactosidase (lactase) in 20 µl, lactose was hydrolysed and measured as glucose over a period of 75 minutes. Recoveries ranged between 96.5 and 100.3% (Fig. 5).
Lactose content(%) in the sample prior to hydrolysis 0.869 x 0.D. Glc x 342.3 180.16 In this as well as all previous examples involving oligosaccharides, the presence of glucose and/or fructose (free or as phosphate) must be determined prior to hydrolysis. Calculations for oligosaccharides should be corrected accordingly.  et al., 1962). Employing Galactostat, we obtained standard curves with melibiose, lactose, raffinose and stachyose. These were of different slopes at equivalent concentrations. Hence melibiose can be measured directly with the Galactostat reagent if other galactosyl containing sugars or galactose are absent. The same is true for raffinose stachyose and lactose.

Melibiose/raffinose mixtures
Melibiose present in mixtures with raffinose was determined as follows. Raffinose was hydrolysed with invertase, yielding equimolar amounts of fructose and melibiose. Total melibiose was then determined with the Galactostat reagent. Correction was made for melibiose derived from raffinose on the basis of fructose present in the hydrolysate (see raffinose determination). Thus, 0.1 M raffinose treated with invertase for l hour as above, was diluted with water to a concentration of 2.5 mM. Aliquots of 25-150 µl from this solution were each added to 1.5 ml Galactostat reagent and diluted to 3 ml with water. After incubation for l hour at 37° C, the reaction was stopped with 0.2 ml 0.5 M EDTA-Na 4 (Sempere, Gancedo and Asensio, 1965) and the optical density read at 425 nm. Melibiose treated in like manner was used as control (Fig. 6).

DISCUSSION
Various techniques involving the enzymes glucose oxidase, invertase and melibiase in conjunction with determination of carbohydrate by reducing power, anthrone, manometry or optical rotation have been previously developed for assay of glucose, fructose, sucrose, melibiose and raffinose in various combinations (Eddy and Mapson, 1951;Potter and Williams, 1958;Bottger and Steinmetzer, 1959;Johnson et al., 1964 andDe Whalley, 1965). Problems include: (1) the necessity to remove interfering substances, e.g. ascorbic acid with ascorbic acid oxidase; (2) colour development by the anthrone method, which requires the inclusion of standard solutions and depends on temperature and duration of reaction; (3) the determination of optical rotation which requires special apparatus, and measurements may be low where large amounts of optically active species are present (Bergmeyer and Klotzsch, 1965); and (4) melibiase is not commercially available.
The methods described here for these five sugars have important advantages over these techniques. They are simpler, accurate and, because of the enzymes involved, entirely specific. Except for melibiose, they do not require a calibration curve for each analysis and utilize equipment generally found in biological laboratories. Raffinose can also be measured with Galactostat but this reagent will also react with galactose and other galactosyl-glycosides. The fructose reagent can also be used to determine stachyose after treatment with invertase (Fig. 3).
The method of quantitative analysis of sucrose/raffinose or sucrose/stachyose mixtures should also be suitable for the determination of other galactosyl sucroses with a fructofuranosyl group detachable by invertase, such as verbascose (Bourquelot and Bride!, 1910) and ajugose. A difficulty may be the decreasing rate of hydrolysis with increasing molecular weight (Adams, Richtmyer and Hudson, 1943;Courtois, 1958). Although mixtures of glucose, fructose, sucrose and raffinose occur commonly in plant materials, stachyose or verbascose are less widespread (Bailey, 1965) but such mixtures could be measured quantitatively without separation by our methods.
Pan, Nicholson and Kolachov (1953) reported the determination of mixtures of glucose, maltose and other oligosaccharides fermentable by yeast, but substantial differences in the degree of fermentation, based upon type of yeast employed, have been shown by Kempf and Lindemann (1954) who could not confirm the quantitative fermentation of glucose in 90 minutes and that of maltose in 1 50 minutes with commercial baker's yeast. Quantitative analysis of glucose and maltose in mixture is readily achieved by our method. Although maltotriose would not be distinguished from maltose, this is not an important shortcoming because maltotriose does not occur naturally in plant tissues (Pazur, 1970).
Lactose has been detected in such plant tissues as the fruit of Achras sapota (Venkataraman and Reithel, 1958) and pollen of Forsythia (Kuhn and Low, 1949). Chromatographically purified cx-galactosidase was employed by us to avoid the interferences encountered with less pure preparations. The Galactostat reagent can be used for lactose determination only if galactose or other sugars with a terminal galactosyl group are absent. The Galactose UV-test reagent permits the measurement of this sugar in the presence of other galactosides. However, analogues of D-galactose such as L-arabinose and D-fucose will also react (Finch et al., 1969;Hu and Grant, 1968