Residue dehydrothreonine

Direct proof for the existence of O-glycosidic linkages involving the hydroxyl groups of serine and threonine was provided simultaneously from three laboratories. Anderson et al. (39) reported partial losses of serine and threonine after treatment with 0.52V NaOH (or 0.452V KOH) at 4°C or room temperature for about 20 hours. Subsequent reduction with platinum showed formation of some alanine and < -aminobutyric acid. Harbon et al (40) treated ovine submaxillary glycoprotein at pH 12.8 for 45 minutes at 70 °C. The serine and threonine content decreased by 78 and 60%, respectively. Treatment of this product with 0.1M sulfite, (pH 9, 24 hours, room temperature) caused a conversion of the dehydroserine residues to cysteic acid, but had no action on the dehydrothreonine residues. This reaction has been further studied by Simpson et al. (41). 

The best procedure for most purposes was the treatment with alkali in the presence of sodium borohydride carried out on BSM (42) the dehydroserine linkages were converted to alanine. Since this procedure did not reduce all dehydrothreonine residues, the procedure was modified by a final reduction in the presence of palladium chloride and borohydride (43, 44). The reactions involved are shown in Figure 6. 

Figure 7 shows that the best conditions for elimination of side chains from BSM and OSM are probably 0.3M sodium borohydride in 0.12V NaOH at 45°C for about 10-12 hours. Over 90% of the total hexosamine and sialic acid were dialyzable under these conditions (36). The reduction of dehydroserine and dehydrothreonine residues was not studied,... 

Figure 6 shows that the dehydrothreonine residue is a derivative of 2-amino-2-butenoic acid. This residue is not easily reduced by sodium borohydride, and complete reduction requires a palladium catalyst (43, 44). Also, the peptide bond involving the amino group of the dehydrothreonine residue is not hydrolyzed by dilute acids whereas that of the dehydroserine residue is labile under the same conditions (44), as will be discussed below. 

A deactivation by the methyl group seems unlikely since compounds of the type RiR2C = C(COOH)2 (Ri and R2, H or Me) were shown by Kadin (67) to exhibit no such effect and are reducible by sodium borohydride. However, the propriety of comparing these compounds with dehydrothreonine residues is questionable. 

Polypeptide. The release of phosphogalactomannan, mannose, mannobiose and mannotriose from the polymer by alkali suggests that these saccharides are attached to a polypeptide (11). The amino acid composition of a peptidophosphogalactomannan preparation was determined (Table III). This table shows that approximately one-half of the amino acyl residues of the polypeptide are either seryl or threonyl residues and that the polypeptide has no aromatic or sulfur-containing amino acids. Treatment of the pol3nner with alkali followed by reduction of the a,3-dehydroamino-acyl residues formed resulted in a loss of all but 2 of the seryl residues and essentially all of the threonyl residues (Table IV). Furthermore, the number of alanyl residues increased from 4 to 8 and 4 residues of a-aminobutyric acid were obtained. These were derived from the reduction products of the a,3-dehydroserine (a-aminoacrylic acid) and a,3-dehydrothreonine (a-aminocrotonic acid), respectively, following 3-elimination of saccharides from seryl... 

---