ANALYTICAL
BIOCHEMISTRY
188,432-435
(1990)
Analysis of lnositol by High-Performance Liquid Chromatography WeiTong
Wang,
Jiri
Safar,*
and David
Zopf
BioCarb Inc., Parkuiew Building, Suite 100,300 Professional Drive, Gaithersburg, *Laboratory of Central Nervous System Studies, National Institute of Neurological National Institutes of Health, Bethesda, Maryland 20892
Received
January
8,199O
A high-performance liquid chromatographic (HPLC) method has been developed for identification and quantification of inositol isomers and monosaccharides in inositol-containing glycans. The method, which can determine 10 pmol of inositol, utilizes an Aminex HPX87C column packed with an 8% crosslinked cation-exchange resin in the calcium form eluted with deionized water at 5O’C. NaOH solution is added to the column effluent through a postcolumn tee to increase the pH (pH > 11.6) before entering a pulsed amperometric detector which is highly sensitive for polyhydroxylated compounds. Samples in which inositol is linked to sugar through a glycosidic bond are hydrolyzed with 5.5 N trifluoroacetic acid, lOO”C, 4 h, and then reduced with NaBH4. Samples in which inositol is linked via a phosphate ester are hydrolyzed with 6 N HCl, llO’C, 24 h. This method has been applied to the analysis of inositol in the hamster prion proteins (PrP) PrP2,-30, and PrPS”. 0 1990 Academic Press, Inc.
Inositol is a ubiquitous component of living cells and body fluids, serving in a variety of functional and structural roles. For example, D-rnyo-inositol 1,4,5triphosphate has been recognized as an intracellular second messenger (1). In addition, both myo- and chiro-inositols occur as components of the glycosylphosphatidylinositol anchor that secures some glycoproteins to cell membranes (2-6). Several inositol-containing oligosaccharides are excreted in human urine (7,8). In the course of analyzing several inositol-containing glycans from different biological sources we have noted vastly different rates of release of inositols during acid hydrolysis, depending upon whether the inositol moiety is linked via glycosidic or phosphate ester bonds: inositols linked via glycosidic bonds can be released under conditions usually employed for monosaccharide com432
Maryland 20879, and Disorders and Stroke,
positional analysis, such as 5-6 N TFA, lOO”C, 4 h (9); in contrast, inositols linked via phosphate ester bonds are completely released only under considerably stronger acid hydrolysis conditions usually employed for amino acid compositional analysis, such as 6 N HCl, llO”C, 24 h. In the present work, we have investigated hydrolytic conditions to effect acid-catalyzed release of inositol from both glycosidic linkages and phosphate ester linkages and we have explored chromatographic conditions for HPLC separation of free inositol isomers from monosaccharides. Utility of the analytical method described is illustrated for prion proteins (PrP),’ components of the agent thought to be responsible for transmission of scrapie (4). MATERIALS
AND METHODS
Inositol. myo-Inositol was purchased from CalBioChem (La Jolla, CA). scyllo-, D-chiro-, and neo-inositol were gifts from Dr. L. Anderson, University of Wisconsin, from chemical synthesis. myo-inositol l-phosphate (I- 1-P) was from Serva. Scrapie prion protein (F’rP). Hamster PrP27-30was purified according to Prusiner’s procedure with minor modifications (10). Hamster PrPSc was purified as described (11). Sample hydrolysis and reduction. In a 5ml reaction vial (Pierce Chemical), a sample containing 0.1-l nmol inositol was hydrolyzed with 200 ~11of 5.5 N TFA, lOO”C, 4 h. The hydrolysate was dried under a stream of nitrogen gas at 37°C. The residue was reduced by adding 100 ~1 of 50 mM freshly made NaOH solution containing 2.5 mg NaBH,. The reaction vial was kept at room temperature overnight. One drop of glacial acetic acid was added 1 Abbreviations used: PrP, Prion Protein; PAD, pulsed amperometric detector; I-l-P,
TFA, trifluoroacetic acid; myo-inositol l-phosphate. 0003-2697/90
$3.00
Copyright 0 1990 by Academic Press, Inc. All rights of reproduction in any form reserved.
CHROMATOGRAPHIC
ANALYSIS
to stop the reduction. Methanol (0.5 ml) and one drop of glacial acetic acid were added and the sample was dried under a stream of nitrogen gas at 37°C. The residue was twice evaporated to dryness from methanol plus acetic acid and twice from methanol alone. The residue was then dissolved in 0.5 ml deionized water and was applied on a mixed bed ion-exchange column (2 X 0.4 cm i.d. AGSOW X 8 [H’] form plus AG3 X 4A [OH-] form) and eluted with deionized water. The lyophilized eluent was redissolved in 100 ~1 deionized water and was ready for HPLC injection. A solvent blank sample was treated in the same way. This hydrolysis and reduction method was applied only to samples in which inositol was linked to sugar through a glycosidic bond. HPLC analysis. Dionex 4000i HPLC system equipped with Bio-Rad Aminex HPX-87C column (Ca2+ form 300 X 8.7 mm i.d.) was used. The column was mounted in a water jacket for temperature control. Deionized water was used as solvent at a flow rate of 0.6 ml/min. A 0.1 M NaOH solution was added to the column effluent at 0.2 ml/min through a postcolumn mixing tee to increase the pH of the eluent (pH > 11.6) before entering the PAD. The PAD operation parameters were set as follows: E, = 0.1 V, T, = 300 ms; E2 = 0.6 V, T2 = 120 ms; E, = -0.8 V, T3 = 300 ms. The chromatogram was monitored using a Spectra-Physics SP4270 integrator. Hydrolysis of inositol 1 -phosphate. In a 5-ml reaction vial, 0.1-l nmol of I-l-P was subjected to hydrolysis with 6 N HCl, llO”C, 24 h or up to 70 h in time-course studies. The hydrolysate was dried under a gentle stream of nitrogen gas at 37°C. The residue was dissolved in 100 ~1 deionized water, and was ready for HPLC injection. This procedure was applied to the inositol analysis in PrP27-30, and PrP*. RESULTS
AND
DISCUSSION
Inositol has nine isomers shown below:
q-%&OH OH
epi-lnositol
e/lo-lnositol
no-TGL
Ho*
OH
neo-lnositol
OH
myo-lnositol
muco-lnositol
n”x5& 0”
scyllo-lnositol
D-chiro-lnositol
OH
L-chiro-lnositol
OF
433
INOSITOL TABLE
1
Retention Time of Monosaccharidesand Inositols Retentiontime 50°C
85°C
13.64 14.16 11.16 19.58 14.73 18.37 26.16 24.62 19.82 25.50 25.10 12.24 ((u + 0)
13.21 13.28 11.32 11.22 13.93 16.20 20.11 19.09 16.68 20.51 18.94 11.94
11.55 Q3) 13.20 (cu) 13.37 (a + 0)
11.61
25°C myo-Inositol chiro-Inositol scyllo-Inositol neo-Inositol 2-Deoxyribositol 2-Deoxygalactitol Sorbitol Galactitol Mannitol Fucitol Perseitol Mannose
14.25 15.30 11.20 22.91 15.55 21.34 35.04 33.44 24.03 31.58 35.20 11.95 13.87 11.76 13.79 13.23 15.79 10.01 11.62 18.11” 20.46 21.82 14.00
Galactose Fucose Glucose Glucosamine Galactosamine Air D Column
temperature:
(a) ((3) (0) (ti) (cu) (p) (0) (n) (0)” (a)’
(min)
10.02 (p) 11.22 (a) 15.37
12.92 10.45 16.00 12.88 -
29°C.
The most widely distributed inositol in nature is myoinositol. Of the optically active isomers, D-(+)and L(-)-chiro-inositols are next most abundant, followed in order by muco- and neo-inositols. The remaining epi-, allo-, and cis-inositol isomers are known only as chemically synthesized compounds. Inositols currently are most commonly analyzed by GC or GC/MS as acetylated or trimethylsilylated derivatives, usually requiring 0.5 nmol of sample (6,12,13). Separation of D- and Lchiro-inositol derivatives has been reported (14). Recently developed HPLC methods employing pelicular anion-exchange resins at high pH to separate partially dissociated polyhydroxylated compounds have proven highly efficient for analysis of monosaccharides; however, under chromatographic conditions primarily designed for analysis of monosaccharides, myo-inositol elutes too close to the void volume to permit its reliable identification and quantification (15). HPLC separation of perdeuterated inositol isomers in deionized water has been achieved using a cation-exchange column (Bio-Rad Aminex HPX-87C, Ca2+ form) (16), but the retention behavior of monosaccharides under the same conditions has not been reported. The retention times of monosaccharides and inositols on HPX-87C column temperatures at 25, 50, and 85°C from the present study are shown in Table 1. The four
434
WANG,
SAFAR,
AND
ZOPF
TABLE
2
Relative Molar Response Factor
(myo-Inositol/Mannitol)
1’
0 0
I 10
20 Tine
30
26
1.332 1.320 1.322 1.330 1.338 1.349 1.332 + 0.010
.%+SD
1.332 1.330 1.329 1.349
1.336 + 0.008
(tin)
FIG. 1. HPLC separation of inositol isomers. The chromatographic conditions were as described under Materials and Methods. The column temperature was 50°C. The peaks are: (1) scyllo-inositol, (2) myoinositol, (3) chiro-inositol, and (4) neo-inositol.
natural inositols were well separated at all three temperatures (see Fig. 1). As little as 10 pmol inositol can be detected using a PAD. However, the retention of inositols overlaps with several free monosaccharides, especially fucose. Efforts to improve separation by lowering column temperature were ineffective because at temperatures below 2O”C, cy- and ,&anomers of each monosaccharide begin to separate, further complicating the analysis. The retention times for (Y- and /3-anomers were assigned according to the rule of geometrical arrangement for complex formation among hydroxyl groups of sugars and Ca2+ ions on the column (16). Figure 2 shows chromatographic separation of a reduced mixture of rnyu-inositol, chiro-inositol, mannose, galactose, glucose, fucose, galactosamine, and glucosamine. The monosaccharide alditols tested were more
I I
a Starting b Starting
with with
1 nmol 1 nmol
myo-inositol myo-inositol
and 10 nmol and 10 nmol
mannose. mannitol.
retained on the column than the native monosaccharides over the same temperature range, facilitating inosito1 analysis. The cluster of galactitol, sorbitol, and fucito1 elutes several minutes after the inositols. There are no peaks corresponding to the amino sugars as workup of the hydrolysate (see Materials and Methods) includes passage over an ion-exchange column that removes molecules containing positively charged amino groups. Mannitol elutes separately from, but with similar retention to, the inositols. Thus, when samples are to be analyzed both for monosaccharide composition (9) and for inositols, any mannose present in the monosaccharide analysis (or added as internal standard) can be used conveniently as a reference for calculation of the molar ratios among monosaccharides and inositol. Relative molar response factors for myo-inositol to mannitol are shown in Table 2. In column 1 of Table 2, the starting compounds were myo-inositol and free mannose and in column 2, myo-inositol and mannitol. The close agreement for these two analyses indicates that the reduction step is essentially quantitative (>99.7%). The chemical bond between inositol and phosphate is much more stable than the linkage between inositol and
3 n
2
Ii
, I
0
I
10
T
20
30
I
40
Tin-e (ti)
FIG. 2. HPLC separation of inositols from monosaccharide alditols. The chromatographic conditions are the same as in Fig. 1. The peaks are: (1) myo-inositol (2) chiro-inositol, (3) mannitol, (4) sorbitol, (5) fucitol, and (6) galactitol.
30
45
SO
Time (h)
FIG. 3. The time course of hydrolytic release of myo-inositol from inositol l-phosphate in 6 N HCl at 1lO’C (0, curve l), and the stability curve for myo-inositol in the presence of monosaccharides under the same hydrolytic conditions (0, curve 2).
CHROMATOGRAPHIC
ANALYSIS
OF
435
INOSITOL
sugar. A model compound I-l-P (1 nmol) was chosen for a time-course study of acid hydrolysis. No detectable inositol was released under hydrolytic conditions commonly used for monosaccharide analysis, i.e., 5-6 N TFA, lOO”C, 4 h. Release of free inositol required much stronger acid hydrolysis conditions, such as those commonly used for amino acid compositional analysis. Figure 3, curve 1, shows the time course of hydrolysis of I1-P in 6 N HCl, at 110°C. After 24 h, the inositol released reached a maximum (ca. 90%). The stability of myo-inosit01 under the same conditions of acid hydrolysis in a mixture containing fucose, galactose, glucose, mannose, galactosamine, and glucosamine (each 1 nmol) is shown in curve 2. All sugars were totally destroyed (disappeared from chromatograms); however, the myo-inositol was quantitatively recovered. myo-Inositol was identified previously as a component of PrPs using GCjMS but its molar fraction was not reported (4). On the basis of the above experiments, the inositol content in hamster PrP with molecular range 27-30K (PrP2& and 33-35K (PrPSc), which are prepared by different methods, was analyzed by HPLC. myo-Inositol was the only inositol isomer found in both preparations. The molar ratios of myo-inositol to the intact glycoprotein molecules are 0.95 (PrP2& and 0.78 (PrPSc), respectively.
REFERENCES
ACKNOWLEDGMENTS
15. Olechno, J. O., Carter, S. R., Edwards, (1987) Amer. Biotech. Lab. Sept/Oct.
W. T., and Gillen,
16. Sasaki,
G. (1988)
The authors thank paring the manuscript.
Ms. Carol
Culwell
for excellent
assistance
in pre-
1. Berridge,
M. J. (1987)
Anna
Reu. Biochem.
56,159-193.
2. Low, M. G. (1987) B&hem. J. 244,1-13. M. A. J., and Williams, A. F. (1988) 3. Ferguson, them.
Annu.
Reu. Bio-
67,285-320.
D. R., Hsiao, K., and Prusiner, S. B. (1987) 4. Stahl, N., Borchelt, CellB1,229-240. 5. Futerman, A. H., Low, M. G., Ackerman, K. E., Sherman, W. R., and Silman, I. (1985) Biochem. Biophys. Res. Comma 129,312317. 6. Low, M. G., Futerman, A. H., Ackerman, K. E., Sherman, W. R., and Silman, I. (1987) Biochem. J. 241,615-619. 7. Lundblad, A. (1978) in Methods in Enzymology (Ginsburg, V., Ed.) Vol.
50, Academic
Press,
San Diego,
CA.
8. Wang,
W., Matsuura, F., Nunez, H., LeDonne, B., and Sweeley, C. C. (1984) Glycoconjugates
9. Wang,
W., andzopf,
10. Prusiner, Bendheim, 349-358.
D. (1989)
Carbohydr.
N. C., Jr., Baltzer, J.
Res.
S. B., McKinley, M. P., Bowman, P. E., Groth, D. F., and Glenner,
1,17-35. 189,1-11.
K. A., Bolton, D. C., G. G. (1983) Cell 36,
11. Safar, J., Wang, W.,Padgett, M. P., Ceroni, M., Piccardo,P., D., Gibbs, C. J., Jr., and Gajdusek, D. C., submitted. 12. Wells, them.
W. W., Pittman,
T. A., and Wells,
H. J. (1965)
Ad.
Zopf, Bio-
10.450-458.
13. Sherman, W. R., Eilers, Mass Spectrom. 3,829-840.
N. C., and
Goodwin,
S. L. (1970)
Org.
14. Larner, J., Huang, L. C., Schwartz, Charles, F. W., Oswald, A. S., Shen, T.-Y., Kinter, M., Tang, G., and Zeller, K. (1988) Biochem. Biophys. Res. Commun. 151,1414-1420.
183,1-9.
K., Hicks,
K. B., and Nagahashi,
Ca~bohydr.
D. G. Res.