Even though all animal alpha-amylases include glycosylation sequons (Asn-Xaa-Thr/Ser) in their sequences, amylases purified from natural sources are not quantitatively glycosylated. When wild-type rat pancreatic alpha-amylase, which contains two glycosylation sequons, was expressed in animal cell lines the protein displayed a very low rate of glycosylation (approx. 2%), even after Brefeldin A treatment to increase the contact with the glycosylation machinery. Site-directed mutagenesis of the first glycosylation sequon (Asn(410)-->Gln) resulted in 90% of the protein being glycosylated at the second glycosylation sequon (Asn(459)). Mutation of the second sequon completely inhibited glycosylation. In order to ascertain if the interference in the glycosylation of Asn(459) that was eliminated by the Asn(410)-->Gln mutation could be due to the position of the asparagine residue in the Cys(448)-Cys(460) disulphide bridge, these cysteine residues were mutated to serine residues. The resulting mutant was found to be 100% glycosylated. All mutants with mutations in the C-domain had specific activities identical to that of the wild-type enzyme, indicating that enzymic activity is independent of the structure and modification of the C-terminal domain. To further test the independence of the C-domain with respect to the two N-terminal domains of the protein, which harbour the catalytic site, the last seven of the ten beta\beta-strands that make up the beta-sandwich configuration of the domain were deleted. The truncated protein was not secreted from cells and all enzyme activity was destroyed. These observations show that Asn(459) is the only site that can be glycosylated in wild-type amylase, and confirm the relative independence of the C-terminal domain of alpha-amylase with respect to enzyme activity. In addition, they also establish that the C-terminal domain is absolutely essential for the correct post-translational folding of the enzyme that is responsible for its activity and allows for its secretion.