The present invention relates generally to a new derivatisation reagent useful in the quantitative determination of organic molecules containing an aldehyde and/or keto group, such as sugars, by reversed phase HPLC.

Compositional analysis of the oligosaccharide moieties of glycoconjugates is of fundamental importance in structural studies on these compounds. Over the past decade, several methods have been developed for the analysis of saccharides in the mid-picomole to low nanomole range. Historically, gas chromatography methods have been used to improve sensitivity(l) with derivatised sugars. In HPLC methods, post-column derivatization has been employed for analysis of saccharides and a number of fluorescent reagents have been examined(2-5). Most recently, anion exchange chromatography with amperometric detection has been used for the analysis of underivatised carbohydrates(6).

There have been numerous research studies on the pre-column labelling of saccharides because in this mode, a high yield of derivatives can be achieved. Detection procedures involving benzoylation(7, 8), p-bromobenzoylation(9), p-nitrobenzoylation(lθ) and dimethylphenylsilylation(ll) are based on the introduction of chromophores ϊia hydroxyl groups present in the sugars. DNS-hydrazine(12-14), DABS- hydrazine(15) and Dobsyl-hydrazine(lβ) have been used to produce fluorophore or chromophore-labelled saccharides. These derivatives are based on the reaction of the hydrazine group of the derivatization reagents with the carbonyl group on the sugar. The reaction specificity of the hydrazide reagents with the aldehyde or keto group of the sugar should be more controllable than derivatisation -___ the hydroxyl group. Nevertheless, with these earlier reagents(14) two chromatographically resolvable peaks are produced from a single sugar. Most recently, l-phenyl-3-methyl-5-pyrazolone, as a labelling

reagent for electrochemical detection (17), has been developed for determining saccharides.

In work leading up to the present invention, stable sugar derivatives were sought with the characteristics of high detection sensitivity, ease of resolution and quantitation and broad applicability in sugar analysis, composition determination and sequence evaluation. As a result a new fluorescence labelling reagent is provided and its synthesis, reaction with sugars and other organic aldehydes and ketones and the separation and determination of these derivatives is herein described.

Accordingly, one aspect of the present invention is directed to a fluorescent labelling reagent capable of reacting with an aldehyde or keto group of an organic molecule which reagent comprises a tricyclic compound having a hydrazine moiety depending from the central ring thereof. The present invention is particularly applicable to sugars.

In a preferred embodiment, the tricyclic compound is a derivative of fluorenylmethyl and more preferably 9-fluorenylmethyl. In a most preferred embodiment, the fluorescent labelling reagent is 9-fluorenyloxycarbonyl hydrazine having the structural formula (I):

CH ₂ OCONH-NH ₂ Fmoc-hydrazine

Hereinafter this compound will be referred to as "Fmoc-hydrazine". Furthermore, reference herein to "Fmoc-hydrazine" is not to be construed as a limitation to the ambit of the present invention which extends to the tricyclic- hydrazine compounds stated above. Accordingly, reference to "Fmoc- hydrazine" or "9-fluorenyloxycarbonyl hydrazine" herein is considered to encompass all fluorescent labelling reagents contemplated by the present invention including any derivatives or functional analogues of Fmoc-hydrazine.

One skilled in the art will immediately recognise that one or more hydrogen atoms of the tricyclic portion of Fmoc-hydrazine may be capable of substitution without substantially affecting the fluorescent labelling activity of the compound and which will not depart from the scope of the present invention. By way of example only, substituents contemplated herein include, alkyl (e.g. methyl, ethyl), substituted alkyl, alkenyl, substituted alkenyl, acyl, dienyl, arylalkyl, arylalkenyl, aryl, substituted aryl, heterocyclic, substituted heterocyclic, cycloalkyl, substituted cycloalkyl, halo (e.g. Cl, Br, I, F), haloalkyl, nitro, hydroxy, thiol, sulfonyl, carboxy, alkoxy, aryloxy and alkylaryloxy and the like as would be apparent to one skilled in the art. By alkyl, substituted alkyl, alkenyl and substituted alkenyl and the like is meant to encompass straight and branched molecules, lower (Cj - C_) and higher (more than Cg) derivatives. The term "substituted" includes all the substituents setforth above.

The derivatives of Fmoc-hydrazine contemplated by the present invention are represented by the structural formula (II):

CH ₂ OCONH-NH ₂

wherein each of R^, R2 and R3 comprises the substituents set forth above with the proviso that R^, R2 and R3 are not simultaneously hydrogen.

The reagent Fmoc-hydrazine of the present invention is convenientiy synthesized by the reaction of hydrazine hydrate with 9-fluorenylmethyl chloroformate (Fmoc-Cl) dissolved in acetonitrile or dichloromethane. It should be noted, however, that the present invention is not limited to Fmoc- hydrazine prepared only by this method and variations in conditions and starting materials may be made without departing from the scope of the present invention. For example, Fmoc-Cl may be substituted by any of the groups listed above and, hence, form the corresponding substituted Fmoc- hydrazine the only requirement being that the resulting substituted Fmoc- hydrazine be able to derivatise sugars as herein described

Accordingly, the present invention contemplates Fmoc-hydrazine and any functional biochemical and/or chemical derivatives thereof. Hence, reference herein to "Fmoc-hydrazine" is meant to include above-mentioned substituted derivatives. The present invention also provides Fmoc-hydrazine either in solid of liquid form packaged for sale in a suitable container. There is no absolute requirement that the Fmoc-hydrazine be pure or even substantially pure. Suitable Fmoc-hydrazine can be obtained by reacting Fmoc-Cl and hydrazine without further purification.

In accordance with the present invention, Fmoc-hydrazine provides an improved means for derivatising organic aldehydes and ketones such as sugars. The present invention, therefore, contemplates a method for derivatising an organic aldehyde and/or ketone prior to reversed phase HPLC which method comprises contacting said organic aldehyde and/or ketone with a fluorescence labelling effective amount of a tricyclic compound having a hydrazine moiety depending from a central ring of said tricyclic compound for a time and under conditions sufficient for said organic aldehyde or ketonesugar to be derivatised. Preferably, the tricyclic compound is a derivative of fluorenylmethyl and more

preferably 9-fluorenylmethyl. In a most preferred embodiment, the fluorescent labelling reagent is Fmoc-hydrazine including its derivatives as outlined above and the organic molecule is a sugar.

By "sugar" is meant any polyhydroxylated aldehyde or ketone and includes trioses, tetroses, pentoses, hexoses and septoses as well as higher sugars. The term "sugar" also includes derivatives thereof including a sugar in combination with other molecules (i.e. glycoconjugates) such as in glycoproteins, lipopolysaccharides and the like. "Sugar" and "saccharide" are used herein interchangeably and include disaccharide, trisaccharide and polysaccharide. The present invention is not limited to any particular stereo isomer of a sugar. Furthermore, phosphosugars, sulphated sugars, amino sugars (such as sialic acid) and the like may also be used provided that the resulting derivatives allow a reducing sugar to be used. The method of the present invention, therefore, provides a highly sensitive and reliable method for the analysis of sugars at the sub-picomole to low picomole level. Furthermore, the present invention extends to the determination of organic molecules containing an aldehyde and/or keto group such as glycoconjugates, aliphatic and/or aromatic aldehydes and/or ketones, and ketoacid and/or analogues thereof. For reasons of brevity, all such molecules are encompassed by the term "sugar".

The principle product of the derivatisation procedure is a sugar-hydrazone derivative and generally a sugar-tricyclic compound-hydrazone derivative. Preferably, the product is a sugar-Fmoc-hydrazone derivative. The present invention, therefore, provides sugar-Fmoc-hydrazone derivatives including any substituted Fmoc-hydrazone and/or sugar derivatives as outlined above. Conveniently, the sugar-Fmoc-hydrazone derivative is packaged for sale in solid or liquid form in a suitable container. Such derivatives are useful, for example, as standards in reversed phase HPLC analysis of sugars.

The present invention also provides a kit useful for derivatising sugars and/or in the quantitative determination of sugars comprising a first compartment adapted to containing Fmoc-hydrazine and a second compartment adapted to contain one or more sugar-Fmoc-hydrazone derivatives for use as a standard. The kit may contain additional compartments adapted to contain or receive one or more sugars for labelling or the sugar(s) to be labelled may be added directly to the Fmoc-hydrazine compartment. Further compartments may also be provided to contain appropriate buffers, diluents and/or other reagents required in the derivatisation process. The kit contemplated herein would most conveniently be used in conjunction with a suitable HPLC facility and additional apparatus may also be required such as a water bath.

The present invention is further described by reference to the following non- limiting Figures and Examples:

In the Figures:

Figure 1 is a diagrammatic representation showing the synthesis of Fmoc- hydrazine and its reaction with a carbonyl compound.

Figure 2 is a graphical representation of a chromatogram of Fmoc-hydrazine unpurified after synthesis, chromatographic conditions: column Zorbax ODS 150x4.6 mm, mobile phase 30% acetonitrile 0.08M acetic acid, isocratic elution flow rate l.Oml/min.

Figure 3 is a graphical representation showing the effect of excitation wavelength and emission wavelength on fluorescence intensity, O Mannose; x xylose.

Figure 4 is a graphical representation showing the effects of molecular ratio of monosaccharide to Fmoc-hydrazine on derivatisation reaction at 65°C for 120 min.

Figure 5 is a graphical representation showing the effect of concentration of acetic acid on derivatisation of Fmoc-hydrazine with sugars. O Mannose; x xylose.

Figure 6 is a graphical representation depicting the effect of reaction time of derivatisation of Fmoc-hydrazine with sugars. 1. Mannose; 2. Ribose; 3. Xylose; 4. Glucose; 5. Galactose; 6. Fructose; 7. Maltose. Reactions at 65°C.

Figure 7 is a graphical representation showing (A) the chromatogram of sugar Fmoc-hydrazones, sample injected after derivatization. Chromatographic conditions: column Zorbax ODS 150 x 4.6mm; mobile phase A 20% acetonitrile 0.08M acetic acid, B 50% acetonitrile 0.08M acetic acid gradient elution for 30 min flow rate 1.0 ml/min. 1. Mannose, 2. Fructose, 3. Xylose. (B) The chromatogram of sugar Fmoc-hydrazones, stored for two weeks at 4°C (conditions as for (A)).

Figure 8 is a graphical representation showing RP-HPLC of sugar Fmoc- hydrazones. The separation was achieved by gradient elution. Conditions: column Zorbax C - 8 150 x 4.6 mm mobile phase A 27-5% acetonitrile 0.08 M acetic acid, B 30% acetonitrile 0.08 M acetic acid gradient elution for 30 min flow rate 1.0 ml/min. 1. Lactose, 2. Maltose, 3. Mannose, 4. Galactose, 5. Fructose, 6. Ribose, 7. Xylose.

Figure 9 is a graphical representation of calibration graphs for reducing sugars. Man, mannose; Rib, ribose; Xyl, xylose; Lac, lactose; Fru, Fructose.

Figure 10 is a graphical representation of an analysis of constituent monosaccharides of calf serum fetuin (A) and ovalbumin (B). Sample amount injected, 7 ng each as protein. Solid and dotted lines represent the results obtained for the hydrolysates for 5h and 8h at 100°C, respectively. (A) peak.l containing mannose 0.7 and 0.94 pmol; peak 2 containing galactose 1.06 and 1.46 pmol. (B) peak 1 containing mannose 0.91 and 1.3 pmol; peak 2

containing galactose 0.26 and 0.48 pmol. Eluant, 30% acetonitrile in 0.08 M acetic acid. Other chromatographic conditions were the same as those described in Figure 8.

Figure 11 is a graphical representation depicting calibration graphs for mannose (Man) and galactose (Gal).

EXAMPLE 1 MATERIALS AND METHODS

Chemicals and Samples

Fmoc-Cl(9-fluorenylmethyl chloroformate) and carbohydrates were purchased from Sigma Chemical Co., hydrazine hydrate from BDH Chemicals Australia Pty Ltd, acetic acid (A.R.) from AJAX Chemicals, ethanol from MAY & BAKER Australian Pty Ltd, acetonitrile and methyl alcohol (Chrom AR HPLC) were obtained from Mallinckrodt Australia Pty Ltd. Ovalbumin (chicken egg) and calf serum fetuin were obtained from Sigma Chemical Co.

Synthesis of Fmoc-hydrazine The strategy for the synthesis of Fmoc-hydrazine is shown in Figure 1.

Fmoc-Cl(100 mg) was dissolved in 25 ml of acetonitrile and added dropwise with stirring to 1 ml of hydrazine hydrate. The reaction mixture was stirred for 30 min at room temperature and concentrated by rotary evaporation. The white flaky Fmoc-hydrazine was recovered in 93% yield. The chromatographic result for the crude product, melting at 172-174°C, is shown in Figure 2. This product can be directly used for the derivatization of saccharides without further purification. The product can be further purified by recrystallization from ethanol or acetonitrile as white needles, m.p 173-175°C. Anal, calcd: for Fmoc-NHNH ₂ :C, 70.87; H, 5.51; 0, 12.60; N, 11.02. Found: C, 69.33; H, 5.54; O, 12.65: N, 11.85.

In an alternative procedure for synthesis Fmoc-hydrazine, FMOC-C1 (Igr) was dissolved in 200 ml dichloromethane and added dropwise with stirring to 1.5 ml hydrazine hydrate (7.5 times excess of hydrazine hydrate). The reaction mixture was stirred for 30 minutes at room temperature and concentrated by rotary evaporation. Recrystallized from ethanol, gave white needles m.p. 174- 174.5° in 65% yield. The NMR and mass spectra indicated that the expected product had been synthesised.

Derivatization procedure with Fmoc-hydrazine. The following procedure is representative of the protocol. To 10 ul of an ethanol solution containing 0.05 picomole-50 nanomole of the saccharide was added 110 μl of ethanol containing 0.1% acetic acid, followed, with mixing, by 100 μl of 0.4%(w/v) Fmoc-hydrazine in acetonitrile. The mixture was then put into a 5 ml Teflon tube with screw cap, and heated at 65°C for 180 min in a water bath and then cooled to room temperature. The derivatised sugar mixture was diluted with ethanol and injected directly onto the HPLC column.

Hydrolysis of glycoproteins.

Acid hydrolysis of proteins (fetuin and ovalbumin) was carried out as follows. Samples were dissolved with 200 μl of water in screw cap teflon tubes and 200 μl of 4M TFA was added. The samples were hydrolysed in a boiling water bath for 5-8h. The tubes were cooled and the samples were dried by nitrogen, dissolved in ethanol and subjected to derivatisation as described above.

HPLC conditions.

The apparatus used for HPLC was DuPont Instruments 8800 system (ternary chromatographic pump; series 8800 gradient controller; column oven). Detection of the saccharide derivatives was performed with Perkin-Elmer LS-5 Luminescence Spectrometer. The wavelengths of excitation and emission were 270 and 320 nm, respectively.

The separation of monosaccharide derivatives was performed at ambient temperature and at a flow-rate of 1.0 ml/min using DuPont Chromatographic Bioseries C-18, C-8 and Phenyl column (5μm, 150 x 4.6 mm). For gradient runs, the mobile phase (A) was 27.5% of aqueous acetonitrile containing 0.08 M acetic acid and (B) was 30% of aqueous acetonitrile containing 0.08 M acetic acid. Isocratic elution was 30% of aqueous acetonitrile containing 0.08 M acetic acid.

EXAMPLE 2

ANALYSIS OF FMOC-HYDRAZINE AND DERIVATISATION OF

SACCHARIDES

Characterisation of Fmoc-hydrazine. Fmoc-hydrazine was synthesized as described in Example 1. The reversed phase HPLC analysis of the crude product is shown in Figure 2. The new compound was also characterised by its organic elemental analysis and mass spectra. The data of organic elemental analysis are given in Example 1, whilst mass spectroscopy indicated (m/e, relative mass) of 255(molecular ion. MH+). On the basis of these data, the compound synthesized is characterised as 9- fluorenyloxycarbonylhydrazine as depicted in Figure 1.

The determination of the optimal excitation and emission wavelengths of the sugar Fmoc-hydrazone were examined with chromatographic conditions used for separation of sugars. The results are shown in Figure 3, the optimal wavelength was 270 nm for excitation and 320 nm for emission.

Optimisation of derivatization conditions.

The effects of molar ratio of Fmoc-hydrazine to sugar on derivatization yield Using the previously described derivatisation procedure (see Example 1), the effects of the molar ratio of Fmoc-hydrazine to sugar on derivatisation yield was examined. The results are shown in Figure 4 and it can be seen that the

influence of the molar ratio ranging between 2.5:1 to 20:1 on the derivative yield were different for each sugar. For example, the yields of mannose and xylose exhibited almost no change in this range of mole ratio. However, the derivative yields of ribose, lactose and galactose were enhanced by increasing the mole ratio. The results indicate that ten-fold excess of the Fmoc-hydrazine reagent is enough for quantitative analysis of aldoses and other sugars.

The derivative yield of fructose increased slowly with the increasing mole ratio. The reactivity of sugars with Fmoc-hydrazine is also shown from Figure 4. The order is mannose = xylose > ribose > lactose > galactose > fructose. These results with mannose and xylose at 2.5-fold excess of reagent, in addition to reflecting reactivity, also showed that during the process of the reaction, the Fmoc-hydrazine is quite stable. In fact, the reagent can be stored in the dark at 4°C for at least fourteen weeks with no loss of potency.

The effects of the concentration of acetic acid on derivatisation yield. The concentration of acetic acid in the reaction mixture was found to affect the rate of hydrazone formation. As shown in Figure 5, the optimum concentration of acetic acid was around 0.5%. When the concentration of acetic acid was reduced to 0.05%, the reaction was nearly stopped; over 1% of acetic acid the recovery for hexose (mannose) was decreased. The use of trichloroacetic acid and trifluoroacetic acid was examined for the derivatisation reaction, but no satisfactory results were obtained.

The effects of reaction temperature and time on derivatisation yield.

The optimum temperature for hydrazone formation was found to be 65°C, at lower temperature the reaction was slower, while at higher temperatures it was difficult to maintain a constant concentration of acetic acid. The time at which the reaction reached completion at 65°C was different for the various sugars. From Figure 6, it can be seen that the complete derivatization for most of the sugars examined was achieved at around 180 min. When extending the reaction time to 240 min, the recovery of the derivatised sugars remained

constant.

These results show that the sugar-Fmoc-hydrazones were stable under these reaction conditions. These results are also different from other hydrazones, e.g., DNS-hydrazine, when used to label sugars, where the recovery of the derivatised product decreased with extended reaction times after reaching the yield maximum. Under the experimental conditions employed in these studies, the high stability of sugar-Fmoc-hydrazones was particularly evident. For example, a mixture of the mannose, fructose, and xylose-Fmoc-hydrazone has been stored in the derivatised solution at 4°C for two weeks with no change in the composition (Figure 7(A) and (B)). This characteristic of the sugar-Fmoc- hydrazone is of benefit to quantitative analysis by HPLC and is a feature not evident with other derivatisation procedures.

Reversed phase HPLC of sugar Fmoc-hydrazones

Separation of the sugar-Fmoc-hydrazones was investigated by using various reversed phase stationary phases. The results are shown in Table 1.

TABLE 1

Elution of sugars (capacity factors, k') from different columns at same elution conditions

k'-l: gradient elution, A: 27.5% CH ₃ CN, 0.08 M acetic acid for 30 min. B:

30% CH ₃ CN, 0.08M acetic acid k'-2: isocratic elution 30% CH ₃ CN, 0.08 M acetic acid.

The elution order of sugar derivatives was different for C-8, C-18 and Phenyl columns under the same elution conditions. Disaccharides were the earliest to elute on all of the columns examined due to their more hydrophilic properties. The elution order of the lactose and maltose derivatives on the alkyl bonded phases was reversed on the phenyl bonded phase, they can not be separated on C-18 bonded phase but were resolved on the C-8 phase. The selectivity of the sugar derivatives examined is better on the C-8 column than on the other two columns. The separation of the sugar derivatives is shown in Figure 8.

The effects of mobile phase were also examined. While there was no difference in the selectivity of sugar derivatives observed with acetonitrile and methanol, acetonitrile showed higher efficiencies for analysis than methanol. Addition of 0.08 M acetic acid to the mobile phase resulted in a marked decrease in the retention of the sugar derivatives and the separation was also improved both in terms of selectivity and peak shape. Phosphate and ammonium acetate 0.08 M has no influence on the separation.

Previous studies have reported the presence of two derivative products for fructose. In the present study, all of the sugar derivatives examined showed a single peak including fructose. Furthermore, some earlier derivatisation procedures result in the formation of several by-products which are difficult to separate from the sugar derivatives. There were very few derivatisation by¬ products with the present process and all eluted after the sugar derivatives. Unreacted Fmoc-hydrazine and breakdown products were easily removed from the column, so that the derivatisation reaction sample can be injected directly onto the HPLC system without any pretreatment.

Fluorescent response The fluorescence response of ten sugars, normalised to fucose in terms of peak area, are listed in Table 2.

TABLE 2 RELATIVE RESPONSE FACTORS OF SUGAR FMOC-HYDRAZONES

(Fucose - 1.0)

Of the hexoses, the Fmoc-hydrazone derivative of fructose was found to exhibit the lowest response.

Calibration graphs and sensitivity

The linearity of the detector response was investigated by injection of progressive dilutions of the sugar Fmoc-hydrazones. As shown in Figure 9, the response was linear over the range of lOpmol - llOpmol. Further studies have shown that this linear response can be extended out to lOfmol.

The detection limits were determined with a signal-to-noise ratio of 3 and the results are shown in Table 3. A comparison of detection sensitivity among the recently published methods for sugars is given in Table 4.

TABLE 3 DETECTION LIMITS OF REDUCING SUGARS

Sugars Detection limit of sugars (picomole)

Mannose 0.05

Ribose 0.09

Glucose 0.1 Galactose 0.07 Fructose 0.04 Xylose 0.01 Maltose 0.01 Lactose 0.2 Fucose 0.2 Arabinose 0.3

TABLE 4

SENSΓΓIVΓΠES OF RECENT PUBLISHED METHODS FOR SUGAR

DETERMINATIONS

Method Sensitivity Reference

GC, silylation Auto Anal. UV with

Cyanoacetamide HPLC, Dns-hydrazine Flu* HPLC, Dns-hydrazine Flu HPLC, per-p-bromobenzoylation

UV 254 nm HPLC, Phenyldimethylsilylation UV 260-254 nm

HPLC, 2-cyanoacetamide

Electrochem. detection HPLC, Dns-hydrazine Flu HPLC, Microbore column Moving-wire FID**

HPLC, Ethylenediamine 1 pmol (20)

Electrochem. detection HPLC, Dabsyl-hydrazine HPLC, DABS-hydrazine Flu HPLC, Pulsed amperometric detector HPLC, Gal-OD/POD reactor

Flu HPLC, l-phenyl-3methyl-5- pyrazolone UV 254 nm 100 find electrochem. detection

Fluorescence detector

** Flame ionisation detector

Analysis of the component monosaccharides of glycoproteins

In order to test the accuracy and the sensitivity of the analytical method, the sugar composition of two well-characterised glycoproteins was determined. Fetuin and ovalbumin were hydrolysed and the released sugars were derivatised with Fmoc-hydrazine. Figures 10 (A) and (B) show the HPLC chromatograms of Fmoc-derivatives obtained from the hydrolysates of fetuin and ovalbumin, respectively.

Figures 10 (A) and (B) indicate that under the experimental conditions, using hydrolysis of the sample under temperature conditions of a boiling water bath, the sugar release with 2 M TFA for 5h was not completed. When extending the time of hydrolysis the recovery of sugars was increased. The results also proved that this method can be used for the determination of component monosaccharides of glycoproteins in the sub-picomole to low picomole level. The sugar composition was calculated from the hydrolysis sample and the calibration curve of standard sugars derivatised under the same conditions (Figure 11).

The results of determination of the sugar composition of fetuin and ovalbumin are shown in Table 5. The relative derivation of quantitative determination was within ± 15% for monosaccharides of hydrolysates of glycoproteins in sub- picomole to low picomole range.

TABLE 5

DETERMINATION OF COMPONENT MONOSACCHARIDES

IN GLYCOPROTEINS

Glycoprotein Hydrolysis Content (w/w%)* (source) 100°C/h Mannose

Fetuin 5 (calf serum) 8

Ovalbumin 5

(chicken egg) 8

Numbers in parentheses are published values and references

REFERENCES:

1. Mawhinney, T.P., Feather, M.S., Barbeor, G.J., and Martinez, LR. ΔnaL Bioc fiin.1QL- 112, 1980.

2. Honda, S., Tamura, T., Suzuki, S., Takehi, M., Kakehi, K., and Ganno, S. Anal. Binchem. 134: 483, 1983.

3. Schlaback, T.D., and Robinson, J. J. Chromatography 282: 169, 1983.

4. Honda, S., Takahashi, M., Kakehi, K., and Ganno, S. Anal. Biochem. 112; 130, 1981.

5. Kiba, N., Shitara, K., and Furusawa, M. J. Chromatography 463: 183, 1989.

6. Horda, M.R., Townsend, R.R. and Lee, Y.C. Anal. Biochem. _2__ 54, 1988.

7. White, CA., Kennedy, J.F., and Golding, B.T. Carhohydr. Res. _ξ_ 1, 1979.

8. Ullman, M.D., and McCluer, R.H. Methods in En2ymolngy 12ft 117, 1987.

9. Golik, J., Kiu, H.W., Dinovi, M., Furukawa, J., and Nokonishi, K. Carhohydr. Res. US; 135, 1983.

10. Dua, V.K., and Bush, CA. J. Chromatography 244; 128, 1982.

11. White, CA., Vass, S.W., Kennedy, J.F., and Large, D.G. JL Chromatography 264: 99, 1983.

12. Eggert, F.M. and Jones, M. J. Chromatography 222; 123, 1985.

13. Takeda, M., Maeda, M. and Tsuji, A.J. J. Chromatography 244; 347, 1982.

14. Mapper, K., and Johnson, L. J. Chromatography 256: 27, 1983.

15. Muramoto, K., Goto, R., and Kamiya, H. Anal. Biochem. 162; 435, 1987.

16. Lin, Jen-Kun and Wu, Shan-Shou Anal. Chem. 52; 1320, 1987.

17. Honda, S., Akao, E., Suzuki, S., Okuda, M., Kakehi, K., and Nakamura, J. Anal. Biochem. lfiQ; 351, 1989.

18. Honda, S., Konishi, T. and Suzuki, S. J. Chromatography 222; 245, 1984.

19. Veening, H., Tock, P.P.H., Kraak, J.C. and Poppe H. l_ Chromatography 252; 345, 1986.

20. Honda, S., Enami, K., Konishi, T., Suzuki, S. and Kakehi, K. L. Chromatography 261; 321, 1986.