5) closely matched those obtained by LC–MS2 (Fig. 3a and Table 1). This provides strong independent evidence that the compounds listed in Table 1 are Adda-containing compounds, and is also consistent with the observation of prominent [MH−134]+ fragments (Fig. 1) in the MS2 spectra of 1–31 during LC–MS2

DNA Damage inhibitor (method A) analysis. LC–MS/MS with precursor-ion scanning for m/z 135 also readily identified microcystins which contained modifications at position-7 that render them unreactive towards thiols, such as [Mser7]-derivatives 14, 15, and 22, making this approach highly complementary to thiol derivatization with LC–MS2 (method A) when a sufficiently concentrated sample is available. A 500 mL sample of net-haul concentrate (BSA4) from Lake Victoria was extracted and selected microcystin analogues purified by standard chromatographic procedures to provide a specimen of MC-RY (9) of sufficient purity for NMR spectroscopy, as well as specimens of MC-LR (1), MC-YR (2), and MC-RR (3) for

comparison. Examination of the 1H, COSY, TOCSY, DIPSY and a series of 1D-SELTOCSY NMR spectra of MC-RY (9) in CD3OD revealed signals HDAC inhibitor (Table 2) attributable to the presence of 7 amino acids, namely Ala, Arg, erythro-β-methylaspartic acid (Masp), Tyr, 3-amino-9-methoxy-l0-phenyl-2,6,8-trimethyldeca-4,6-dienoic acid (Adda), glutamic acid (Glu), and N-methyldehydroalanine (Mdha). The chemical shifts of the majority of the proton signals arising from these groups in 9 were similar to the limited NMR Adenosine triphosphate data reported for MC-YR (2) ( Kondo et al., 1992; Namikoshi et al., 1992), however the chemical shifts of some of the protons of 9, including the Arg H-2 (4.08 ppm), and the Tyr H-2 (4.48 ppm), H-3 (2.45 and 3.38 ppm) and aryl H-5/9 signals (6.96 ppm), differed significantly from those which we observed in the same solvent for MC-YR (2) isolated from BSA4 (4.47 ppm (Arg H-2), 4.31 ppm (Tyr H-2), 3.06 and

3.12 ppm (Tyr H-3) and 7.19 ppm (Tyr H-5/9) ( Table 2)). Differences were also apparent in some of the 13C resonances of 9, as revealed in gHSQC experiments optimized for coupling constants of 140 Hz and 130 Hz, compared to those reported for the corresponding atoms of 2. For example, the Arg and Tyr C-2 resonances of 9 occurred at 57.1 and 55.1 ppm, respectively, whereas the corresponding resonances of 2 occur at 52.7 and 59.7 ppm. These differences, while consistent with the presence in 9 of the same set of 7 amino acids as in MC-YR (2), were indicative of a change in the location of the Arg and Tyr residues of 9, relative to their locations in 2. This proposal is also in accord with the consistent disposition of the Adda5, Mdha7, Ala1, Masp3 and Glu6 residues in other microcystins, such as 1 and 3, as well as with fragmentations observed during LC–MS2 analysis (Fig. 4 and Supplementary data).