BIOMEDICAL CHROMATOGRAPIIY, VOL. 8, 199-201 (1994)

Separation of Reducing Monosaccharides by Capillary Zone Electrophoresis

Li Qi* and Yichuan Zhu New Technology Application Institute of Beijing, Beijing 100035. China

The separation of reducing monosaccharides derived from glycosidoproteins and glycolipids by capillary zone electrophoresis (CZE) is dependent on the pH and concentration of the borate buffer. Five saccharides were completely separated in a fused silica capillary tube (50 pm i.d., 65 cm) containing 50 mM borate buffer (pH 10.5) as carrier, with high resolution, at an applied potential of 20 kV after the reducing saccharides were derivatized with 1-naphthylamine. On-column UV (254 nm) monitoring allowed quantitation of these saccharides at least in the concentration range of 10-100 mM in reaction solution. This method was applied to the determination of the monosaccharides composition of various carbohydrate materials to demonstrate its usefulness.

INTRODUCTION

During the last few years, major progress in theory, instrumentation and applications of capillary electro- phoresis (CE) have taken place (Chen et af., 1993; Huang et al., 1993; Schure and Lenhoff, 1993; Carson et al., 1993; Zhao et al . , 1993). Capillary zone electro- phoresis (CZE) has become a very important technique for the separation of peptides, proteins and related compounds for several reasons. The main ones are: (1) it is applicable to small sample volumes and complex and multicomponent samples; (2) it is capable of deve- loping extraordinarily high efficiency because of ‘plug flow’ and elimination of mass-transfer factors since a second phase is absent; (3) it consumes very little reagent; (4) it provides a rapid separation; ( 5 ) it usually involves a mild separation condition. These and other advantages can also be realized when CZE is used to separate classes of smaller-sized analytes or those with higher mobilities, including inorganic anions and cations.

In many cases CZE has demonstrated separation far superior to that of other chromatographic methods. However, there have been no papers dealing with the analysis of carbohydrates.

Since carbohydrates are usually neutral, CZE is not directly applicable. However, they can be converted to anionic borate complexes in a carrier buffer containing borate ion after they were derivatized with 1- naphthylamine (Figs 1 and 2). These anionic complexes then can be analysed by CZE.

* Author to whom correspondence should bc addressed.

CH,OH

HO Q0,< OH +

MATERIALS AND METHODS

Chemical. All chemicals and carbohydrate samples were of the highest quality commercially available and used without further purification. Borate solutions were prepared by dis- solving in triply distilled water and adjusting the pH to the indicated values with HCI.

Capillary zone electrophoresis. A home made CZE apparatus was used. In this work, however, the constant high-voltage power supply (30 kV) was run. A capillary tube (SO pm i.d., 6Scm) of fused silica was obtained from Yong Nian Glass Factory (He Bei, China). The samples were detected at 2S4nm with a modified UV detector (model 8823 from Beijing, China, on-column).

Precolumn derivatization of reucing monosaccharides with 1-naphthylamine. This was performed essentially as described by Wang et al. (1988). The reagent solution was freshly prepared before derivatization by dissolving 35 mg of sodium cyanoborohydride and 100 mg 1-naphthylamine in 0.45 mL of a methanolic solution containing 40kL acetic acid. One hundred pg of reducing monosaccharide or a mixture of reducing monosaccharides was dissolved in 10 pL of triply distilled water, and 40 pL of the reagent solution was added to make the concentration of the saccharide between 10 mM

and 100 mM. The saccharide was dissolved by heating for 1 h at 80°C after it was gently vortex-mixed. To each of the carbohydrate solutions was added 0.5 mL of trichlormethane and 0.5mL of triply distilled water after cooling to room temperature. The aqueous solution was introduced into the capillary tube for CZE.

CH,OH L H O H

Figure 1. Reaction scheme for t h e derivatization of carbohydrate with I-naphthylamine.

CCC 0269-3879/94/040 199-03 0 1994 by John Wiley & Sons, Ltd

Rrcrioed 22 March 1994 Accepted (revised) 4 May IY94

200 L. QI AND Y . ZHU

r R 1 -

t

Figure 2. Formation of anionic complex of carbohydrate-1 - naphthylamine in carrier buffer containing borate ions. R, -CH(OH(CH(OH)CH,OH; R', -CH,NH-naphthyl.

RESULTS AND DISCUSSION

Relationship between electrophoretic mobility and carbohyd- rate structure. When a high voltage is applied between the ends of a capillary tube of fused silica filled with an electrolyte solution (carrier), the surface of the inner wall of the tube is negatively charged and the boundary monolayer of the carrier in contact with the inner wall becomes positively charged. This layer moves to the cathode together with the medium of the carrier, to generate electro-osmotic flow. If a sample is introduced to the tube from the anodic end, all components of the sample, ionic or nonionic, are driven to the cathode at a uniform velocity, V,, . Anionic components are pulled back at the time by electrostatic force. This effect is electrophoresis, and its magnitude depends on the charge and the size of the ion. Electrophoretic mobility, u,,, can be calculated by the equation (Cheng et al . , 1990),

u,,=Ltv-l(t,'-t-'),

where L , I , V and t are the length of the capillary tube, the distance between the anodic end and the detector, the applied voltage, and the retention time, respecti- vely. The symbol to is the time required for electro- osmotic flow to travel distance I and can be obtained from the peak of methanol. Therefore, the apparent velocity, V,, - V,, , also differs from component to component, and as a result the components are separ-

+ t i + + + + + +

BOE i

Veo Cathode

/'P A hF.-

0 0 \ /

1 0-

+ * + + + + + i t

Figure 3. Principle of the separation of the borate complexes of carbohydrate-I-naphthylamines by CZE.

1 l r n l " l 15 I U 5 0

Figure 4. Separation of carbohydrate-I-naphthylamines der- ived from various reducing monosaccharides. Carrier, 50 mM borate buffer, pH 10.5; applied potential, 20 kV; detection, UV absorption at 254 nm. Peak assignment of parent saccharides: Reag, reagent; 1. rhamnose; 2, ribose; 3, glucose; 4, arabinose; 5, galactose; IS (internal standard), cinnamic acid.

4 J-

Figure 5. Analysis of the component monosaccharides in: (a) arabic gum; (b) ribose fermentation liquor. The analytical con- ditions were the same as those described in Fig. 4. Peak assignment was also as in Fig. 4.

SEPARATION OF REDUCING MONOSACCHARIDES BY CZE 20 I

ated. The mechanism of the carbohydrate separation is illustrated in Fig. 3 .

The results showed that the retention time or the electrophoretic mobility of rhamnose, xylose and glu- cose is smaller than that of ribose, arabinose and galactose. It indicates that the pairs having trans- oriented hydroxyl groups at C-3/C-4 (rhamnose, xylose and glucose) gave values of relative uCp smaller than those of the pairs having cis-oriented hydroxyl groups (ribose, arabinose and galactose). Thus, the orientation of the vicinal hydroxyl groups at C-3/C-4 is in good agreement with the theory that cis-oriented hydroxyl groups in carbohydrates preferentially form borate complexes, compared to trans-disposed hydroxyl groups.

Separation of carbohydrates. Since the wavelength for maximum absorption of the 1-naphthylamine is at 254 nm, the monosaccharide-1-naphthylamines could be sensitively detected by a UV detector.

Carbohydrates were charged with appropriate pro- cedure ion CZE. Except under high pH(>12.0) con- dition, the saccharides can form charged complexes with borate or some metal ions such as Zn++ and Cu'+ etc. Further studies show that the monosaccharide-l- naphthylamine complexes with borate can generate high resolution at optimum conditions. To optimize the separation, the effects of applied voltage, the pH, composition and Concentration of the buffer were investigated in detail. For a given capillary, the maxi- mum concentration of borate or boric acid was 150 mM and the voltage applied should be less than 25 kV. In general the buffer composed of borate is better than that of boric acid, because the latter requires a large amount of NaOH to adjust the pH so that the electro- phoretic current increases dramatically.

The relationship between pH and resolution ( R s ) is quite complex. When borate is used, the optimum pH lies in the range of 10.30-10.65. However, the Rs values of all carbohydrates were continuously increased at pH 10.5 depending on the increasing of the buffer concentration. Baseline noise became significant above 150 mM. Therefore, 150 mM was the optimum buffer

conccntration. Figure 4 depicts the electropherogram of the reaction

mixture of 5 reducing monosaccharides, together with cinnamic acid as an internal standard, obtained under the optimized conditions. Note that the reagent was not migrated under the optimized conditions. Note that the reagent was not migrated at all by electrophoresis and gave the earliest sharp peak, well separated from the peaks of 1-naphthylamines. A large excess of the rea- gent did not interfere with the analysis of reducing monosaccharides. The underivatized portions of the monosaccharide samples and by-products (possibly alditols) were indifferent to the analysis, because they gave no peaks.

Quantification. All the selected saccharides gave calib- ration curves of relative peak area to 5 m ~ cinnamic acid, passing through the origin and linear at least in the range 10-100 mM after treatment by the derivatiza- tion procedure. The relative standard deviations (n = 5 ) or the reproducibility for the relative peak area of ribose at 20,40,60 and 80 mM were 2.28,3.09,2.90 and 1.17%, respectively. The corelation coefficient r for ribose and glucose calibration curves were 0.993 and 0.991, respectively. By using calibration curves, these saccharides could be accurately quantified, although the yields of the derivatives were not quantitative.

Application to the analysis of component monosaccharides. The component monosaccharides in various carbohyd- rate materials, including oligosaccharides, oligosac- charide glycosides and polysaccharides, such as lactose (containing glucose and galactose), melibiose (contain- ing glucose and galactose), rutin (containing rhamnose and galactose), digitonin (containing xylose, glucose and galactose), arabic gum (containing rhamnose, ara- binose and galactose), ribose fermentation liquor (con- taining glucose and ribose), were determined by the established method. The electropherograms are shown in Fig. 5 . It is therefore concluded that CZE is a proniising tool for the analysis of monosaccharides compared with high-performance liquid chromatogra- phy and HPAEC.

REFERENCES

Carson, S., Cohen, A. S., Belenkii,A., Ruiz-Martinez, M. C., Berka, J. and Karger, B. L. (1993). Anal. Chem. 65, 3219.

Chen, S. and Pietrzyk, D. J. (1992). Anal. Chem. 65, 2770. Cheng, Y . eral. (1990). Anal. Chem. 62, 496. Huang, T.-L., Tsai, P., Wu, C.-T. and Lee, C. S. (1993). Anal.

Chem. 65, 2887. Schure, M. R. and Lenhoff. A. M. (1993). Anal. Chem. 65,3024. Wang, W., Chen, Q. and Chai, W. (1988). Shengwu HuaxueZazhi

Zhao, Z., Malik. A. and Lee, M. L. (1993). Anal. Chem. 65, 2747. 4, 116