A new metabolite, 6-amino-6-carboxy-2-trimethylammoniohexanoate has been isolated from the red alga Schottera nicaeensis, and its structure was determined by spectral and chemical methods. Its distribution in about fifty species of Rhodophyceae has been investigated. Dragendorff-positive compounds isolated from or detected in red algae include glycine betaine (1), β-alanine betaine (2), γ-butyrobetaine (3), proline betaine (stachydrine) (4), 4-(2-dimethylaminoethyl)phenol (hordenine) (5), picolinic acid betaine (homarine) (6), nicotinic acid betaine (trigonelline) (6), trimethyltaurine (7), choline sulphate (8), 3-dimethylsulphoniopropionate (9), and 4-dimethylsulphonio-2 methoxybutyrate (10). In a search for compounds of this type in Mediterranean red algae, we have now found that Schottera nicaeensis (Duby) Shott. [syn. Perroglossum nicaeense (Lamouroux ex Duby) Guiry et Hollenberg] (Gigartinales) accumulates a highly polar metabolite that reacts with both Dragendorff's reagent (orange) and ninhydrin (purple). Ion-exchange chromatography of the EtOH extract of the red alga and further purification by hplc afforded chromatographically pure 1 as an off-white, hygroscopic powder, [α]²⁵D -6.4. On account of its salt-like characteristics, the mass spectrum was uninformative, and the molecular formula was established as C₁₀H₂₂O₄N₂ by combustion analysis. The ¹³C-nmr spectrum contained two carboxyl (carboxylate) resonances at 172.45 and 170.78 ppm, two α-amino acid methines at 76.45 (d, C-2) and at 53.01 (d, C-6), three methylene triplets at 29.63 (C-3), 26.05 (C-5), and 21.37 (C-4), and a three-methyl quartet at 52.29 [-N(CH₃)₃]. In the ¹H nmr, determined at 250 MHz in D₂O, the C-4 methylene protons appeared as a 2-H multiplet at δ 1.49, while those of the C-3 and C-5 methylenes gave rise to a complex 4H signal centred at δ 2.01. A 9H singlet at δ 3.19 was assigned to the -N(CH₃)₃ group. The remaining signals were a 1H double doublet at δ 3.74 (J=10 and 3.7 Hz) and a 1H triplet at δ 4.00 (J=6.2 Hz), which were assigned respectively to the C-2 and C-6 methines. This assignment was based on the shift at δ 4.05 of the first signal after protonation of the carboxylate function and comparison with the spectrum of 2,6-trimethylammonio-heptanedioate (3) (see Experimental section). Esterification of 1 with MeOH in the presence of HCl afforded the monomethylester 2 (δ 3.78, singlet, 3H, -OCH₃); comparison of its ¹H-nmr spectrum (see Experimental section) with that of the original compound showed that the C-7 carboxylic group, rather than the C-1 carboxylate group, had been esterified, because in the latter case a downfield shift for the C-2 methine resonance would have been expected. The mass spectrum of 2, which showed a small molecular ion at m/z 246 confirming the molecular formula of 1, could be interpreted taking into account that betaines show fragmentation patterns from their pyrolysis product(s) in the gas phase (11), the most important pyrolytic process involving an intermolecular transfer of a methyl group from the quaternary nitrogen to the carboxylate function. Accordingly, the mass spectrum of 2 displayed diagnostically important peaks from the corresponding transalkylated compound at m/z 215 (M⁺-OCH₃), 187 (M⁺-COOCH₃), 156 (M⁺-OCH₃-COOCH₃), 142 (187-H₂N(CH₃)₃), 128 (M⁺-2COOCH₃), 84 [CH₂=CH-CH=N(CH₃)₂], and 58 [CH₂=N(CH₃)₂]. The above data led to structure 1 for the new algal metabolite. Definite proof was obtained by exhaustive methylation of the amino group in non-epimerizing conditions according to the procedure reported by Patchett and Witkop (12). The permethylated compound was chromatographically and spectroscopically (¹H nmr) indistinguishable from a sample of 2,6-trimethylammonioheptanedioate obtained by methylation of 2,6 diaminopimelic acid. Because the semisynthetic sample was optically inactive, it was deduced that the two chiral centers in 1 have opposite configuration. The new metabolite also occurs in Gastroclonium clavatum (Roth) Ardiss. (Rhodymeniales) and Liagwa distenta (Mert.) C. Ag. (Nemalionales), while it was not detected in a number of other species of red algae (about fifty) representative of all the seven orders belonging to the Florideae (see Table 1). This restricted distribution indicates a possible taxonomic value. It is to be also noted that the closely related laminine (N,N,N-trimethyllysine), originally isolated from Laminaria angustata (13) and widespread in brown algae, until now has not been found in red algae. If one considers that 2,6-diamino pimelic acid in a number of organisms is the biosynthetic precursor of lysine, compound 1 could bear the same relationship to laminine, and its accumulation in red algae could be the consequence of the lack of the enzymic system required for decarboxylation.