Structure and Activity of Largazole, a Potent Antiproliferative Agent from the Floridian Marine Cyanobacterium Symploca sp. Kanchan Taori, ² Valerie J. Paul, ‡ and Hendrik Luesch* ,² Department of Medicinal Chemistry, UniVersity of Florida, 1600 SW Archer Road, GainesVille, Florida 32610, and Smithsonian Marine Station, 701 Seaway DriVe, Fort Pierce, Florida 34949 Received September 17, 2007; E-mail: luesch@cop.ufl.edu The identification of new pharmacophores is of paramount biomedical importance and natural products have recently been regaining attention for this endeavor. 1 This renaissance is closely tied to the successful exploitation of the marine environment which harbors unmatched biodiversity that is presumably concomitant with chemical diversity. 2 In particular, marine cyanobacteria are prolific producers of bioactive secondary metabolites, 3 many of which are modified peptides or peptide-polyketide hybrids with promising antitumor activities, such as dolastatin 10, 4 curacin A, 5 and apratoxin A. 6 As a result of our ongoing investigations to identify new drug leads from cyanobacteria in Florida, we report here the structure determination and preliminary biological characterization of a marine cyanobacterial metabolite with novel chemical scaffold and nanomolar antiproliferative activity from a cyanobacterium of the genus Symploca. Symploca species have scarcely been investigated compared to the more prevalent Lyngbya spp., yet a Palauan Symploca sp. previously yielded the clinical trial compound dolastatin 10, 4 prompting us to target this genus. A sample of Symploca sp. was collected from Key Largo, Florida Keys, and extracted with organic solvents. The resulting cytotoxic crude extract was subjected to bioassay-guided fractionation by solvent partition, silica gel chromatography, and reversed-phase HPLC to yield largazole (1) as a colorless, amorphous solid {[R] 20 D +22 (c 0.1, MeOH)}. 1 H and 13 C NMR data coupled with a [M + H] + peak at m/z 623.2397 in the HRESI/APCIMS of 1 suggested a molecular formula of C 29 H 42 N 4 O 5 S 3 (Δ +0.1 mmu, Δ +0.16 ppm). The 1 H NMR spectrum exhibited two signals characteristic for secondary amides (δ 2-NH 7.15, δ 14-NH 6.45). Further two-dimensional NMR analysis in CDCl 3 using COSY, HSQC, and HMBC data indicated that these exchangeable protons belong to valine and modified glycine residues, respectively (Table 1 and Supporting Information). The putative glycine carbonyl (δ C-13 167.9) was part of a 2,4- disubstituted thiazole unit as evidenced by HMBCs from the only aromatic methine (δ H-12 7.76, δ C-12 124.2) to C-13 and to another quaternary sp 2 carbon, C-11 (δ C 147.4). Furthermore, HMBCs from a methyl singlet (δ H-9 1.87) to carbonyl C-6 (δ C 173.5), quaternary carbon C-7 (δ C 84.4), and methylene carbon C-8 (δ C 43.3), combined with an HMBC from H-8a (δ H 4.04) to C-10 (δ C 164.6) suggested the presence of a 2-substituted thiazoline-4-methyl-4- carboxylic acid unit (C-6 to C-10). The only other HMBC to C-10 was from the thiazole proton H-12, indicating that C-10 bore the thiazole substituent. The methyl thiazoline carboxylate and the amino terminus of the valine residue were unambiguously connected via an amide linkage based on HMBC data (Table 1). The remaining signals in the 1 H NMR spectrum belonged to two spin systems, as concluded from COSY analysis (Supporting Information). One of the units was a 7-substituted 3-hydroxyhept-4-enoic acid moiety (C-15 to C-21) with E-geometry of the double bond based on a large coupling constant for 3 J H-18,H-19 of 15.6 Hz, consistent with NOESY cross-peaks between H-18 and H 2 -20. This unit was attached to the amino terminus of the glycine-derived unit as shown by HMBCs from 14-NH and H-14a/b to C-15 as well as ROESY cross-peaks between 14-NH and H-16a and H-16b. The last unit ² University of Florida. ‡ Smithsonian Marine Station. Table 1. NMR Spectral Data for Largazole (1) in CDCl3 (600 MHz) C/H no. δH (J in Hz) δC, mult. HMBC a,b 1 168.9, qC 2 4.61, dd (9.2, 3.3) 57.7, CH 1, 3, 4, 5, 6 3 2.10, m 34.2, CH 1, c 2 c 4 0.68, d (7.2) 18.9, CH3 2, 3, 5 5 0.50, d (7.2) 16.6, CH3 2, 3, 4 2-NH 7.15, d (9.2) 1, 6 c 6 173.5, qC 7 84.4, qC 8a 4.04, d (-11.4) 43.3, CH2 6, 7, 10 8b 3.27, d (-11.4) 6, 7, 9 9 1.87, br s 24.2, CH3 6, 7, 8 10 164.6, qC 11 147.4, qC 12 7.76, s 124.2, CH 10, c 11, 13 13 167.9, qC 14a 5.29, dd (-17.4, 9.6) 41.1, CH 13, 15 14b 4.27, dd (-17.4, 2.5) 13, 15 14-NH 6.45, dd (9.6, 2.5) 15 c 15 169.4, qC 16a 2.86, dd (-16.5, 10.5) 40.5, CH2 15, 17, 18 16b 2.68, dd (-16.5, 1.8) 15 17 5.66, ddd (10.5, 7.2, 1.8) 72.0, CH 18 5.51, dd (15.6, 7.2) 128.4, CH 17, 20 19 5.82, dt (15.6, 7.2) 132.7, CH 17, 20 20 2.31, br q (7.2) (2H) 32.3, CH2 18, 19, 21 21 2.90, t (7.2) (2H) 27.9, CH2 19, 20, 22 22 199.4, qC 23 2.52, t (7.5) (2H) 44.1, CH2 22, 24, 25 24 1.64, m (2H) 25.6, CH2 22, 23, 25/26 25 1.29, m (2H) 28.9, CH2 26 26 1.25, m (2H) 28.9, CH2 25, 27 27 1.26, m (2H) 31.6, CH2 28 1.28, m (2H) 22.6, CH2 29 0.87, br t (6.9) 14.0, CH3 27, 28 a Protons showing HMBC correlations to the indicated carbon. b Opti- mized for n J ) 7 Hz if not indicated otherwise. c Optimized for n J ) 3.5 Hz. Published on Web 01/19/2008 1806 9 J. AM. CHEM. SOC. 2008, 130, 1806-1807 10.1021/ja7110064 CCC: $40.75 © 2008 American Chemical Society