A Practical Guide to the Methylation Analysis of Carbohydrates

4. Gas Liquid Chromatography of Partially Methylated Alditol Acetates

GLC offers the best method for the separation and quantification of the methylated sugars obtained on hydrolysis of a methylated polysaccharide or glycoconjugate. The field has been reviewed 45 46 . A prerequisite is that the methylated sugars are converted into volatile derivatives. Several types of derivatives have been used, e.g. methyl glycosides 47 , acetates 48 , alditol acetates 2 3 and aldononitrile acetates 49 . A serious disadvantage of derivatives preserving the anomeric center is that a single methylated sugar may give rise to at least two, possibly four peaks on GLC, viz α-and β-pyranosides and α- and β-furanosides. One advantage of using alditol acetates is that each aldose derivative will give only one peak on GLC. Ketose derivatives, on the other hand, will give two peaks. Another advantage of partially methylated alditol acetates is that quantification can be performed without the use of response factors 50 except for the analysis of mixtures containing N--acetamido sugars. The symmetry introduced by converting sugars into alditols is avoided by using sodium borodeuteride as the reducing agent.
The best separation of partially methylated alditol acetates are obtained on medium-polar columns such as ECNSS-M 2 (a nitrile-silicon polyester copolymer) or OV-225 3 (a silicon polymer containing methyl, phenyl and nitrile groups). Good separations may also be obtained on OS-138. OV-225 is usually preferred in the authors' laboratory for routine work since it combines good separation properties with thermal stability
The stationary phase is coated on an inert carrier, usually of diatomaceous earth. In the authors' laboratory a 3% coating on Gas Chrom Q (100/120 mesh) is glass-columns [180 cm x 2.0 mm (I.D.)] has been commonly used. The columns are fitted in a gas chromatograph equipped with flame-ionisation detector (FID). The response from the FID is linear over a broad range. The injection and detection systems are kept about 25°C higher than the oven of the chromatograph. Most separations are performed iso-thermally in the temperature range 170-200°C. Relative retetion times (T-values) have been determined by interpolation using two internal standards (e.g. 1,5-di-O-acetyl-2,3,4,6-tetra-O-methyl-D-glucitol and 1,4,5,6-tetra-O-acetyl-2,3-di-O-methyl-D-glucitol) and show a reproducibility better than ± 3% on most columns. T-Values obtained on ECNSS-M, OV-225 and OS-138 columns are given in Appendix A.
More efficient separation of sugar derivatives are obtained on capillary columns which are of two major types; wall-coated columns (WCOT) in which the stationary phase is coated directly on the column wall, and support-coated columns (SCOT) in which the stationary phase is adsorbed on a finely divided carrier. Columns of both these types are commercially available. In the authors' laboratory OV-225 SCOT and OS-138 SCOT columns (15 m x 0.5 mm; Perkin-Elmer Co) have been used. These columns, whilst giving better separation than the corresponding packed columns, in most cases give closely similar relative retention times for a particular derivative. By far the best separations have been obtained on SP-1000 WCOT columns (25 m x 0.25 mm, LKB-products, Bromma, Sweden) which afford up to 105 theoretical plates. Relative retention times obtained at 220°C on such columns are tabulated in the Appendix A. It should be noted that separations on capillary columns are usually performed using an injection "splitter" system and a "make-up" gas system to add carrier gas after the column to optimise the response in the FID.
Methylated N-acetamido sugars present special problems in GLC due to their thermal lability and comparatively low volatility. The separation of a number of methylated derivatives of 2-acetamido-2-deoxy-D-glucose and -D-galactose as alditol acetates have been reported 8 13 51 52 . It should be emphasised that the quantitative evaluation of the chromatograms may be subject to serious error due to different response factors for different N-acetamido sugar derivatives and, in particular, between N-acetamido sugar derivatives and neutral sugars. This aspect does not seem to have been fully evaluated yet.