Communication THE ANALYSIS OF MADUROSE, AN ACTINOMYCETE WHOLE-CELL SUGAR, BY HPLC AFTER ENZYMATIC TREATMENT AKIRA YOKOTA AND TORU HASEGAWA

Transcription

1 J. Gen. Appl. Microbiol., 34, (1988) Short Communication THE ANALYSIS OF MADUROSE, AN ACTINOMYCETE WHOLE-CELL SUGAR, BY HPLC AFTER ENZYMATIC TREATMENT AKIRA YOKOTA AND TORU HASEGAWA Institute for Fermentation, Osaka, 17-85, Juso-honmachi 2-chome, Yodogawa-ku, Osaka 532, Japan (Received July 21, 1988) Since the early work of LECHEVALIER and LECHEVALIER (1), whole-cell sugar analysis has become a widely used technique to classify and identify actinomycetes. The whole-cell sugar analysis method can be used to recognize four types of wholecell sugar pattern in aerobic actinomycetes. In these patterns, arabinose, xylose, galactose, and madurose (3-0-methyl-D-galactose) are diagnostic sugars. All the sugars can be identified with conventional chromatographic techniques such as paper chromatography (PC) and thin-layer chromatography (TLC). However, madurose, a diagnostically important sugar, is often present only in very small amounts (2, 3), and is frequently difficult to identify. PC and TLC have been used to rapidly analyze the whole-cell sugar composition qualitatively (4-7). Another established technique, applied only rarely to such sugar analysis, is to convert the sugars into volatile derivatives and then analyze them by combined gas-liquid chromatography/mass spectrometry (GC/MS) (8, 9). However, this system is expensive. Recently, high-performance liquid chromatography (HPLC) has been applied to the analysis of sugars (10). Among the methods reported, a system using an anion exchange column and fluorescence detector (11) is the most suitable to analyze a mixture of small amounts of sugars. However, the application of HPLC to analyze madurose has not yet been described (12). Here, we offer an improved method involving the use of enzymes and HPLC to simply and reliably identify madurose, a taxonomically important whole-cell sugar. Address reprint requests to: Dr. Akira Yokota, Institute for Fermentation, Osaka, honmachi 2-chome, Yodogawa-ku, Osaka 532, Japan , Juso- 445

2 446 YOKOTA and HASEGAWA VOL. 34 Table 1. Madurose content in the whole-cell hydrolysate of various actinomycete strains. The strains used in this study are given in Table 1. The cells were grown in a medium containing 1 % yeast extract and 1 % D-glucose (ph 7.0) at 28 C for 4 days with shaking. The biomass was washed with distilled water and lyophilized. The dried cells (100 mg) were hydrolyzed with 4 N HC1 at 100 C for 4 hr in a screwcapped test tube. After filtration, the hydrolysate was concentrated in vacuo. The residue was dissolved in distilled water and neutralized with NaOH; water was added to make 1 ml. D-Mannose and D-glucose in the hydrolysate were converted into their phosphate esters with hexokinase (grade II, Oriental Yeast Co., Ltd., Japan, EC ) (13) in the reaction mixture A, which was composed of 50 pl of the hydrolysate, SOul of 50 mm ethanolamine buffer (ph 7.6) containing 1.4 mm mercaptoethanol and 7 mm MgClz, 50 pi of 50 mm ATP, and 100 pg (12.8 units) of hexokinase dissolved in 20p1 of water (total 200,1). To confirm the presence of madurose, the hydrolysate was then treated together with D-galactose oxidase (Sigma Co., St. Louis, U.S.A., EC ) (14) and catalase (Sigma Co., EC ) in the reaction mixture B (200 Ri), which contained 170 p1 of reaction mixture A, 200pg (32 units) of D-galactose oxidase dissolved in 30p1 of water, and 70 pg (140 units) of catalase dissolved in 2 p1 of water (total 202 R1). Both reaction mixtures were incubated at 37 C for 20 hr in the presence of 1 drop of chloroform. The madurose used in the experiment was prepared by Dr. T. Kusaka from a hydrolysate of Actinomadura kijaniata IFO according to the method of LECHEVALIER and GERBER (15). Based on the data of optical rotation, 1 H-NMR, 13C-NMR and GC MS, the purified sugar was confirmed to be 3-0-methyl-Dgalactose (madurose) (T. Kusaka, unpublished results). The optimized HPLC separation conditions followed the method described by MIKAMI and IsHIDA (11). The HPLC system consisted of a Shimadzu Model LC-5A pump, 25 pl Rheodyne Model 7125 loop injector, a Shim-pack ISA column (4.0 x 250 mm, Shimadzu) containing an anion exchanger with a Shimadzu Model RF-530 spectrofluorometer at an exciting wavelength of 320 nm and

3 1988 Chemotaxonomy of Aerobic Actinomycetes 447 emission wavelength of 430 nm, and a Shimadzu Chromatopac C-R1B integrator. The column was heated to 65 C, and the column effluent (flow rate of 0.6 ml/min) and detection reagent, a mixture of 100 arginine and 3 % borate delivered by another Model LC-5A pump at a flow rate of 0.5 ml/min, were led to to a Shimadzu Model CRB-3A chemical reaction bath and heated to 140 JC. The solvent system used was a step wise elution with borate buffer at concentration of 0.2 M (ph 8.5, 10 min), 0.3 M (ph 9.0, 10 min) and 0.4 M (ph 9.0, 20 min). As the retention time of madurose and D-mannose were almost the same (27 min) in this HPLC system, it was necessary to eliminate D-mannose from the hydrolysate to estimate the madurose content; we applied hexokinase to achieve this. To verify the reliability and accuracy of the enzymatic elimination procedure, mixtures of authentic sugars were used as a model system. Typical elution patterns Fig. 1. Analysis of a mixture of D-mannose, D-galactose and D-glucose. a, no treatment; b, treated with hexokinase; c, treated with D-galactose oxidase. Peak designations: 1, D-mannose; 2, D-galactose; 3, D-glucose. Fig. 2. Analysis of a mixture of madurose and L-rhamnose. a, no treatment; b, treated with hexokinase; c, treated with D-galactose oxidase. Peak designations: 1, L-rhamnose; 2, D-ribose; 3, madurose.

4 448 YOKOTA and HASEGAWA VOL. 34 of the sugars in the mixtures before and after enzymatic treatment are shown in Figs. 1 and 2. D-Mannose and D-glucose completely disappeared after the hexokinase treatment (Fig. 1b); however, D-galactose and madurose were not affected by this treatment (Figs. 1 b and 2b). The presence of a relatively high concentration of ATP (12.5 mm) was necessary to completely phosphorylate hexoses in the reaction mixture. We used D-galactose oxidase to identify madurose. This enzyme is known to oxidize madurose as well as D-galactose (15). As shown in Figs. lc and 2c, D- galactose and madurose were completely oxidized by D-galactose oxidase and disappeared from the chromatogram on HPLC. Thus, it was confirmed that the phosphorylation of D-mannose and the oxidation of madurose were smoothly catalyzed by the respective enzymes. Figure 3 shows the HPLC chromatogram of the whole-cell hydrolysate of Microtetraspora fusca IFO The hydrolysate contained D-glucose, D- galactose, and possibly, an equal amount of D-mannose and madurose (Fig. 3a). The chromatogram of the sample treated with hexokinase clearly shows the presence of madurose in the hydrolysate (Fig. 3b). Peak 1 was confirmed to be the peak of madurose by its retention time and its disappearance after incubation with D-galactose oxidase (Fig. 3c). Examples with quantitative data of madurose in some actinomycete strains are shown in Table 1. The madurose content in actinomycetes seems to vary sig- Fig. 3. Analysis of whole-cell hydrolysate of Microtetraspora fusca IFO a, no treatment; b, treated with hexokinase; c, treated with hexokinase and D- galactose oxidase. Peak identities: 1, madurose; 2, D-mannose; 3, D-galactose; 4, D-glucose.

5 1988 Chemotaxonomy of Aerobic Actinomycetes 449 nificantly, depending on the strain. Such quantitative estimations of madurose in many other actinomycete strains of our culture collection are in progress in our laboratory. STANECK and RoBERTS (5) and HASEGAWA et al. (6) have reported a rapid analytical method for whole-cell sugars using TLC. In their method, medurose was distinguished from other sugars by the differences in Rf value and color reaction on the chromatogram. However, in the strains with trace amounts of madurose, it is difficult to distinguish differences in color. The procedure described here involves two steps of enzyme reaction. D- Mannose, which shows the same retention time as madurose on HPLC and therefore, disturbs the estimation of madurose, was phosphorylated with commercially available hexokinase. Madurose was identified from its retention time and from the disappearance of the corresponding peak on HPLC after treating the enzyme reaction mixture of the first step with commercially available D-galactose oxidase. Thus, the procedure of HPLC after enzymatic treatment appears to be effective to distinguish madurose from D-mannose in the whole-cell hydrolysates, and also to rapidly identify madurose. Our method can be easily used to analyze strains with low madurose levels, so it is well adapted to the needs of individual laboratories for clinical, ecologic and taxonomic studies of actinomycetes. We are indebted to Dr. T. Iijima, Director of this Institute, for his encouragement and discussions, and also to Dr. T. Kusaka, Integrated Technology Laboratories of Takeda Chemical Industries Ltd., for supplying madurose. REFERENCES 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12) 13) 14) 15) H. A. LECHEVALIER and M. P. LECHEVALIER, Ann. Inst. Pasteur, 108, 662 (1965). A. FISCHER, R. M. KROPPENSTADT, and E. STACKEBRANDT, J. Gen. Microbiol., 129, 3433 (1983). S. MIYADOH, S. AMANO, H. TOHYAMA, and T. SHOMURA, Int. J. Syst. Bacteriol., 37, 342 (1987). B. BECKER, M. P. LECHEVALIER, R. E. GORDON, and H. A. LECHEVALIER, Appl. Microbiol., 12, 421 (1964). J. L. STANECK and G. D. ROBERTS, Appl. Microbiol., 28, 226 (1974). T. HASEGAWA, M. TAKIZAWA, and S. TANIDA, J. Gen. Appi. Microbiol., 29, 319 (1983). L. P. TEREKHOVA, T. P. PREOBRAZHENSKOYA, and 0. A. GALTENKO, Actinomycetes, 19, 73 (1986). K. H. AAMLID and S. MORGENLIE, Carbohydr. Res., 124, 1 (1983). G. J. GERWIG, J. P. KAMERLING, and J. F. G. VLIEGENTHART, Carbohydr. Res., 127, 245 (1984). R. W. FRET, In Chemical Derivatization in Analytical Chemistry. Vol. 1. Chromatography, ed. by R. W. FREI and J. F. LAWRENCE, Plenum Press, New York (1981), p. 233, 241, 259. H. MrKAMI and Y. ISHIDA, Bunseki Kagaku, 32, E207 (1983). D. P. LABEDA, J. Ind. Microbiol., Suppl. No. 2, 28, 115 (1987). R. A. DARROW and S. P. CoLOWICK, Methods Enzymol., 5, 226 (1962). G. AVIGAD, D. AMARAL, C. ASENSIO, and B. L. HORECKER, J. Biol. Chem., 237, 2736 (1962). M. P. LECHEVALIER and N. N. GERBER, Carbohydr. Res., 13, 451 (1970).