Histochemical identification of 9-O-acyl sialic acids: studies of bovine submandibular and rat sublingual gland and human colon

Summary Formalin-fixed tissue specimens containing glycoproteins with side chain O-acylated sialic acids were used to re-examine, compare and evaluate the usefulness of three methods based on the periodic acid-borohydride reduction-saponification-periodic acid-Schiff sequence (PA-Bh-KOH-PAS) for the histochemical identification of 9-O-acyl sialic acids (9-O-AcSA). Method I, modified from Vehet al. (1979), involved a comparison of the staining intensely obtained when both oxidation steps of the PA-Bh-KOH-PAS sequence were carried out with the selective oxidation technique of Volzet al. (1987) with that obtained when the initial oxidation step was carried out with 0.5m periodic acid for 4h at room temperature. Methods II and III, modified from Reidet al. (1978), involved an initial PA-Bh step under oxidation conditions that cleaved all the vicinal diols associated with neutral sugars and side chain unsubstituted and 7-O-acyl sialic acids. The Schiff staining obtained following subsequent re-oxidation with either 0.5m (method II) or 1% periodic acid (method III) for 4h at room temperature (PA-Bh-PAS procedure) identifies 9-O-AcSa.The results of this study indicate that (a) bovine submandibular gland acinar cell glycoproteins contain 9-O-AcSA as well as sialic acids which have ester substituents at C7 or C8, or which are di-(C7C8, C7C9, C8C9) or tri-(C7C8C9) substituted, (b) the side chain O-acyl sialic acids of the glycoproteins of Sprague Dawley rat sublingual gland acinar cells are entirely or almost entirely 9-O-AcSA and (c) it is likely that the majority of the human adult and foetal glycoproteins studied contain small quantities of 9-O-AcSA mixed with sialic acids which are substituted at C7 or C8 or which have two or three side chain O-acyl substituents. However, the interpretation of the results are complicated by observations that indicate that (a) treatment with 0.5m periodic acid either extracts or removes sialic acids from bovine submandibular gland glycoproteins, (b) some human colonic epithelial glycoproteins apparently contain a component other than 9-O-AcSA that oxidises slowly with periodic acid and (c) 1% periodic acid for 2h at room temperature oxidises a small but significant quantity of 9-O-AcSA, thus reducing the intensity of staining in methods II and III. It is concluded that when adequately controlled, methods I, II and III are capable of detecting 9-O-AcSA in glycoproteins containing large quantities of the sialic acid. However, these methods may not detect small quantities of 9-O-AcSA in the presence of large quantities of sialic acids which have O-acyl substitutents at positions C7 or C8 or which have two (C7C8, C7C9, C8C9) or three (C7C8C9) side chain O-acyl substituents. Thus, caution should be used when interpreting data that indicates the absence of 9-O-AcSA.