On the N-methyl-L-threonine residue in stendomycin.

Sir: In a previous communication^ to this Jour nal, the isolation of N-methyl-L-threonine from acid hydrolysates of stendomycin2~4) was briefly mentioned. We wish to report here some experiments that led to the final assignment of the configuration for this amino acid with two centers of asymmetry. The specific rotation of the N-methyl amino acid from stendomycin was in good agreement with the values reported5) for N-methyl-L-threonine. This agreement, however, could not be considered as con clusive evidence for their identity, especially because the rotation of the optically active forms of N-methylallothreonine was not known from literature, where only pro perties of N-methyl-DL-allothreonine have been published6). It seemed possible that the specific rotations of the optically active N-methylallothreonines are not sufficiently different from the rotations in the normal series to permit the assignment of con figuration solely on this basis. N-Methyl L-threonine is more levorotatory than L threonine which represents an exception to the rule7) that in the L-series the N-methyl derivatives have more positive rotation than the amino acids from which they are derived. The contribution from the second center of asymmetry (the /? carbon atom) could be the reason for this discrepancy^, which cautions against the complete reliance in the assign ment of configuration on a comparison with only one of the diastereoisomers. In order to ascertain the correctness of the above assignment a sample of authentic N-methyl-L-threonine was prepared from L-threonine through the N-benzyl-, and benzylmethyl derivatives8^ A sample of N-methyl-L-allothreonine was obtained through the application of the same pro cedure to L-allothreonine which in turn was secured by the epimerization method of Elliott9). The optical rotations of the products, summarized in Table 1, show that the methylamino acid in stendomycin is indeed N-methyl-L-threonine. The nmr spectra of N-methyl-L-threoriine and N-methyl-L-allothreonine (Fig. 1) are different especially in respect to the chemical shifts of the doublets corresponding to the C-methyl groups. It is interesting to note that a mixture of equal amounts of N methylthreonine and N-methylallbthreotene, exhibits an nmr spectrum which is different from those of its components. The for mation of a molecular compound is probably the best explanation for this observation. The nmr spectrum of the natural compound is identical with that of authentic N-methyl L-threonine. In the quantitative amino acid analysis of the hydrolysate of stendomycin according to the procedure of Spackman, Stein and Moore10), N-methylthreonine appears close to the position of aspartic acid but with an extremely low ninhydrin color yield. In these chromatograms N-methylallothreonine and N-methylthreonine cannot be distin guished from each other. Since in the recordings of the amino acid analyses often a small doublet w^as observed, some doubt existed about the nature of the second material revealed at the aspartic acid posi tion. It could be a trace of aspartic acid formed as a secondary degradation product,, e.g., by the oxidation of proline, but the possibility of the presence of N-methylallo threonine (in addition to N-methyl-L-threo nine) could not be excluded. Since stendo mycin contains two D-allothreonine residues,, this problem had to be investigated. Table 1. Optical rotations of diastereomeric threonines and N-methylthreonines