From the basic amino acid fraction of the red alga Schottera nicaeensis, a previously unreported nitrogenous compound has been isolated by chromatography and its structure determined as 1-(3-amidinoureido)-4-(N-methylacetamido)butane (nicaeensin) [1] by degradation and spectroscopic measurements. An examination of the amino acid fraction from aqueous extracts of the red alga Schottera nicaeensis (Duby) Schott. (Phyllophoraceae; Gigartinales) revealed the presence of a metabolite that gave an orange color with Dragendorff's reagent but did not react with ninhydrin. This unknown compound, nicaeensin, isolated by ion-exchange chromatography and further purified by partition chromatography, was optically inactive and gave a brilliant pink color with Sakaguchi's reagent and a brown color with sodium nitroprusside/potassium ferricyanide, indicating the presence in the molecule of a monosubstituted guanidino group (1). Its ir spectrum in the region 1500-1800 cm-' showed bands at 1550, 1620, 1687, and 1725 cm-'. The molecular formula was deduced as C9H19N5O2 from elemental analysis of the picrate, C15H24N8O9 (mp 171-173'), and from the fab mass spectrum (m/z 230 [M + H]+, 252 [M + Na]+, and 322 [M + H + glycerol]+). Acid hydrolysis (6 N HCl, 48 h at 110°) afforded, along with guanidine, a basic compound (EArg 1.97) whose structure was determined as N-methyl-1,4-butanediamine [2] based on its spectral properties. Important peaks in its ms spectrum were observed at m/z 102 [M]+, 84 [CH2-CH2-CH2-CH=NH-CH3]+, 73 [M - CH2=NH2]+, and 59 [M - CH2=N-CH3]+. The 13C-nmr spectrum displayed, in addition to the N-Me signal at δ 35.69, four methylene triplets at δ 51.23 (-CH2-NH-), 41.77 (-CH2-NH2), 26.78, and 25.53 (-CH2-CH2-CH2-CH2-). In the 1H-nmr spectrum of 2 the N-Me protons occurred as a singlet at δ 2.69, the two nitrogen-bonded methylenes appeared as partially overlapped triplets at δ 3.02 and 3.05, and the remaining two methylenes gave a four-proton multiplet at δ 1.74 (W 54 Hz). Consideration of the above findings and the molecular formula suggested that the new algal metabolite was 1-(3-amidinoureido)-4-(N-methylacetamido)butane (nicaeensin) [1]. The 1H- and 13C-nmr spectra were fully consistent with this structure. The 1H-nmr spectrum in D2O showed the typical signal splitting, which is frequently observed in compounds containing an amide group in the molecule, as the result of the presence in solution of two "stable" conformers. In fact, the N-Me appeared as a pair of singlets at δ 3.01 (conformer A; rel. int. 0.57) and 2.86 (conformer B; rel. int. 0.43), the MeCO protons as two singlets at δ 2.06 (A) and 2.08 (B), the C-1 methylene protons as two triplets at δ 3.10 [J=6.0 Hz (A)] and 3.12 [J=6.0 Hz (B)]. Two additional triplets assignable to the C-4 methylene protons at δ 3.32 [J=7.5 Hz (A)] and 3.36 [J=7.5 Hz (B)], and a four-proton resonance for the methylenes at C-2 and C-3 (broad signal centered at δ 1.52; W = 76.5 Hz) completed the spectrum. Assignments of the signals for each conformer were based on the relative intensities of the pertinent peaks as well as on the results of extensive spin-decoupling experiments. In the 13C-nmr spectrum of nicaeensin, splitting was observed for the following signals: N-Me [quartets at δ 39.01 (A) and 36.20 (B)]; MeCO- [quartets at δ 23.74 (A) and 23.04 (B)]; MeCO- (singlets at δ 176.6 and 176.7); C-1 [triplets at δ 42.02 (A) and 42.11 (B)]; C-2 [triplets at δ 29.20 (A) and 29.08 (B)]; C-3 [triplets at δ 26.61 (A) and 27.46 (B)]; C-4 [triplets at δ 50.22 (A) and 53.60 (B)]; -C(=NH)- (singlets at δ 163.06 and 163.15); and -NH-CO-NH- (singlets at δ 166.43 and 166.71). 13C-1H shift correlations allowed us to assign each signal to the pertinent conformer. The eims fully agreed with the proposed structure and contained diagnostically important peaks at m/z 228 [M-H]+, 213 and 171 [sequential losses of Me and CH2=CO from the ion at m/z 228], 141 [228 - H2N-C(=NH)-NH-COH]+, 127 [M - H2N-C(=NH)-NH-CO-NH2]+, 115 [H2N-C(=NH)-NH-CO-NH=CH2]+, 100 [NH-C(=NH)-NH-CO-NH]+, 86 [H2N-C(=NH)-NH=CO]+, and 58 [H2N-C(=NH)-NH2 - H]+. Partial acid hydrolysis of nicaeensin (6 N HCl, 12 h at 110°) gave, in addition to guanidine and N-methyl-1,4-butanediamine, deacetyl nicaeensin [3] and N-methyl-N-acetyl-1,4-butanediamine [4]. Compound 3, which had an electrophoretic mobility higher than the parent molecule (EArg 1.36), gave color reactions with Dragendorff's (orange), Sakaguchi's (pink), and sodium nitroprusside/potassium ferricyanide (brown) reagents. Compound 4 did not react with these last two chromogenic reagents but gave positive reaction with ninhydrin (violet) and Dragendorff's reagent (orange). Splitting of signals, not seen in the 1H- and 13C-nmr spectra of 3 (see the Experimental section), was observed in the 1H-nmr spectrum of 4. In this spectrum the N-Me appeared as two singlets at δ 3.05 (A) and 2.90 (B), while the acetyl group gave two singlets at δ 2.10 (A) and 2.12 (B). The spectrum also contained three triplets at δ 3.38 [J=6.3 Hz, -CH2-N(Me)- (A)], 3.43 [J=6.3 Hz, -CH2-N(Me)- (B)], and 2.95 [J=7.05 Hz, -CH2-NH2 (A and B)] in addition to a complex multiplet at δ 1.62 [W = 67.5 Hz, -CH2-CH2-CH2-CH2- (A and B)]. Treatment of nicaeensin with acetylacetone in alkaline medium gave the expected 1-[3-(4,6-dimethylpyrimidin-2-yl)ureido]-4-(N-methylacetamido)butane [5] (m/z 293 [M]+). In the 1H-nmr spectrum of 5 the methyls on the pyrimidine ring appeared as a singlet at δ 2.30, while the heterocyclic proton signal was a singlet at δ 6.77. The rest of the signals (see the Experimental section) showed splitting analogous to that observed in the spectrum of 1. In the 13C-nmr spectrum the signals relative to the 4,6-dimethylpyrimidine moiety were at δ 25.69 (methyls), 117.41 (methine), 158.81 (C-2), and 171.35 (C-4 and C-6). So far, only two amidinoureido compounds, gigartinine (6) (2) and gongrine (7) (3), have been reported as natural products, both from the red alga Gymnogongrus flabelliformis (Phyllophoraceae; Gigartinales); the occurrence of 1 in a seaweed belonging to the same family could be of some interest from the chemotaxonomic point of view. Nicaeensin may be biosynthetically derived from gigartinine via decarboxylation, methylation, and acetylation. The possibility that gongrine is biogenetically related to gigartinine cannot be ruled out. Furthermore, considering that in previous work (4) S. nicaeensis has been shown to accumulate 6-amino-6-carboxy-2-trimethylammoniohexanoate [8], it is not unlikely that, in this alga, a biosynthetic relationship may exist between the latter compound and nicaeensin.