In a study designed to establish the stereochemical course of the biological synthesis of tryptophan from serine and indole or indole-containing derivatives, we planned feeding experiments of serine samples stereospecifically labelled with tritium at position 3 to Aspergillus amstelodami cultures producing echinuline (3) and neoechinuline (1) (which contain a tryptophan and a dehydrotryptophan moiety, respectively). Preliminary experiments with randomly labelled tryptophan showed that all the tritium present at position 3' in the precursor is retained into echinuline (3), whereas ca. 50% is lost in the conversion into neoechinuline (1). However, since experiments with [3-3H;3-14C] serine indicated a 6% incorporation into echinuline (3), accompanied by ca. 30% tritium loss, it was necessary to establish the manner of incorporation of serine and of compounds metabolically related to it into the two metabolites (1) and (3), using 13C-labelled precursors owing to the lack of easy degradation procedures for (1) and (3). Furthermore, we expected to obtain information on the origin of the oxalamide moiety present in the dioxopiperasine ring of (1). From cultures of Aspergillus amstelodami grown for feeding purposes on synthetic medium, neoechinuline (1) was not isolated; instead, a new metabolite, cryptoechinuline A, was isolated in small amount close to a large quantity of echinuline (3). Cryptoechinuline A, C24H27N3O2, is optically inactive, has a melting point of 190-Z°, and forms yellowish crystals from benzene. Its UV spectrum [λmax (EtOH) 231, 273, 283(s), and 380 nm (ε 37000, 24100, 21400, and 14000)] resembles that of neoechinuline (1). Comparison of the 13C-NMR spectrum of cryptoechinuline A with that of neoechinuline (1) showed a close similarity in chemical shifts for both saturated and unsaturated carbon atoms. Moreover, the absence of one carbonyl signal in cryptoechinuline A's spectrum and the appearance of two additional signals in the olefinic region (at δ 102.8 and 134.8) pointed to structure (2) for cryptoechinuline A. Further support for this structural assignment came from mass spectrometry: the molecular ion had an exact mass of 389.2111 (± 0.004), corresponding to C24H27N3O2, with major fragment ions of exact masses 321.1489 (C19H19N3O2, M - C5H8), 251.1660 (C18H21N, M - C6H6N2O2), and 250.1589 (C18H20N, M - C6H7N2O2). Permanganate oxidation of cryptoechinuline A (2) led to a complex mixture of acidic compounds which, after methylation, showed the presence of the dimethyl ester of amino teraphtalic acid via GC-MS comparison with an authentic sample, supporting the proposed substitution pattern in the benzene moiety of (2). IR and 1H-NMR spectra also agreed with the proposed structure. The initial aim—determining the origin of the dehydroalanine moiety of cryptoechinuline A and the biosynthetic relationship between cryptoechinuline A and neoechinuline (1)—requires the use of 13C-labelled precursors due to Birch's observation of large randomization of labelling in echinuline (3) biosynthesized from 14C-labelled alanine. We preliminarily examined the natural abundance 13C-NMR spectrum of echinuline (3), assigning 10 signals in the saturated carbon region (δ values from internal Me4Si: C3b 54.8, C3d 50, Cpa 39.1, Cfa 34.6, C7a 31.4, Cja 29.6, Cpd 27.3, C7e 25.7, C3e 19.9, C7d 17.9). Echinuline (3) biosynthesized from 90% enriched 2-13C glycine showed a 120% intensity enhancement of the signal due to C3b, supporting the intact incorporation of glycine into the C1' and C2' carbon atoms of the tryptophan moiety of (3).