Use of Enzymes
Several enzymes are available for releasing N-glycans. The most popular is peptide-^-glycosidase F (PNGase F) , an amidase that releases the intact glycan as the corresponding glycosylamine. This process converts the
attached asparagine to aspartic acid, effectively labeling the site of glycan attachment with a one mass unit increase in molecular weight. The released glycosylamines are relatively unstable and readily hydrolyze to the glycan. Commonly, treatment with dilute organic acid is used to speed the reaction. Without this treatment, some glycosylamine can be retained, particularly if the glycan release has been performed in the presence of ammonium- containing buffers . Various commercial preparations of PNGase F are available, most containing various additives to aid release. Some of these additives are incompatible with MS. Glycerol, for example, can be difficult to remove and can inhibit crystal formation in MALDI analysis. Dithiothreitol is sometimes present as a reagent for denaturing the protein. This compound can decompose releasing H2S, which competes with water for reacting with the released glycosylamines. This reaction adds SH rather than OH to the reducing terminus of the glycan. The nominal mass difference between these moieties is 16, which is the same as that of oxygen, giving the impression that the glycan contains one more oxygen atom than is actually the case .
PNGase F releases most N-glycans except those containing fucose a1 ^ 3- linked to the reducing-terminal GlcNAc . Proteins need to be denatured for efficient glycan release, but, even so, reactions can sometimes take longer than the typical overnight procedure. Various techniques such as microwave irradiation [92, 93] or high pressure  can be used to reduce the reaction time. Use of these methods has produced reaction times of less than 1 h. Glycans can be released in solution or by infusing the enzyme into isoelectric focusing  or SDS-PAGE gels [96-101], the latter technique being particularly useful when only small amounts of glycoprotein are available. After release, glycans are usually recovered from the surrounding solution although dissolution of the gel has also been used .
In situations where PNGase F fails to release glycans, PNGase A is usually effective. This enzyme is larger than PNGase F (75.5 as compared to 35 KDa), does not readily penetrate gels, and only releases glycans from smaller peptides . Consequently, the glycoprotein should be digested with, for example, trypsin prior to incubation . Like PNGase F, PNGase A releases the entire glycan, but another popular group of endoglycosidases, exemplified by endoH and endoS, cleaves the glycan between the two GlcNAc residues of the chitobi- ose core. Although this reaction effectively leaves the reducing terminal attached to the protein as a label for glycan attachment, the released glycan has lost the information regarding the attachment of groups such as fucose to the core GlcNAc. These endoglycosidases are usually more specific in their substrate specificity than the PNGases. EndoH, for example, only releases high-mannose and some hybrid glycans, whereas endoS from Streptococcus pyogenes releases some complex glycans but not high-mannose or hybrid structures .