FK228 from Burkholderia thailandensis MSMB43

FK228 [systematic name: (1S,4S,7Z,10S,16E,21R)-7-ethylidene-4,21-di(propan-2-yl)-2-oxa-12,13-dithia-5,8,20,23-tetrazabicyclo[8.7.6]tricos-16-ene-3,6,9,19,22-pentone], C24H36N4O6S2, also known as FR901228, depsipeptide, NSC 630176, romidepsin, and marketed as Istodax by Celgene Corporation, is crystallized from ethyl acetate in P21 as compared to the absolute configuration of FK228, first crystallized from methanol in P212121 [Shigematsu et al. (1994 ▶). J. Antibiot. 47, 311–314]. A slight difference is observed between the absolute configuration of FK228 and the present structure. The molecular structure is stabilized by intramolecular N—H⋯O hydrogen bonds. In the crystal, molecules are linked via N—H⋯O hydrogen bonds.


Related literature
While exploring novel natural products from the culture broth of B. thailandensis MSMB43, the title compound FK228 was isolated in large quantity (~100 mg/L). The physicochemical properties (UV spectrum, IR spectrum, MS and NMR) of the FK228 from B. thailandensis MSMB43 are identical to those of the FK228 isolated in the early 1990s from Chromobacterium violaceum No. 968 (Ueda, Nakajima, Hori, Fujita et al., 1994;Shigematsu et al., 1994). The crystal structure of the newly isolated FK228, herein, is slightly different than that reported by (Ueda, Manda et al., 1994), due to different crystallization conditions employed. Under the different crystallization conditions presented, FK228 formed a different molecular arrangement than previously described (Fig. 1). The title crystal structure was obtained from ethyl acetate crystallizing in the P2 1 space group, while the provirus structure co-crystallized with one methanol in asymmetric unit in the P2 1 2 1 2 1 space group. The absolute configurations of all the chiral centers in each of these two structures are the same: C1(S), C3(S), c7 (S) and C22 (R). The geometric configuration of the double bonds in the Acyl and Dhb components are all determined as E and Z respectively. In comparing the these two structures, the configuration of these skeletons are the same. But there is a slight difference at the end of L-Val group. Under the different crystallization conditions, C18 and H19 had the opposite positions in these two structures because of the free rotation of the single bond (Fig. 2). Hydrogen bonds N2-H4···O6 and the weak intermolecular interactions N1-H3···O5 and N3-H16···O3 are observed in the title crystal structure.
( Burkholderia thailandensis strain MSMB43 (obtained from the US Centers for Disease Control, CDC) was routinely activated on LB agar containing 50 mg ml -1 of apramycin (Am50) at 37°C for 1 to 2 days as a master plate. A single colony was then transferred into a 1 L flask containing 300 ml of LB medium and Am50, and the culture were growing at 37°C for 24 h as seed culture. For batch fermentation 250 ml of seed culture was transferred into a 20 L fermentor (BioFlo IV, New Brunswick Scientific Co., USA) containing 12 L of M8 medium (0.5% glucose, 0.5% peptone, 0.3% NaCl, 0.12% Na 2 HPO 4 , and 0.05% KH 2 PO 4 ; pH 7), and fermented at 30°C for 96 hr under 20 L/min aeration and 200 rpm agitation; pH was maintained at 7.0 with 1 N NaCl or 1M HCl. For fed-batch fermentation, feeding of 3 L of 10X concentrated medium between 24 hr and 48 hr was performed on top of the batch fermentation.
Recovery of the crude extract Bacteria cells were removed by centrifugation of cell broth at 3,800 rpm for 15 min. The supernatant was applied on a 2 L column (Φ 8.0 × 40 cm) packed with a mixture of Diaion HP-20 (Sigma-Aldrich, USA) resin and Amberlite XAD16 (Sigma-Aldrich, USA), which has been equilibrated in water at 5 ml/min. The resin was dried and then extracted with ethyl acetate for three times. The ethyl acetate extracts were combined and concentrated to dryness in vacuo at 35°C.

Isolation and purification of the title compound
The ethyl acetate extract was subjected to a two step isolation and purification protocol. Briefly, the ethyl acetate extract was mixed with 50 g silica gel (230-400 mesh, Whatman Purasil, USA). The mixture of silica gel was dried overnight and then applied to a 120 -g silica gel column (300 ml), which has been equilibrated with hexane. The column was eluted sequentially with 1 L hexane, 1 L hexane: ethyl acetate (3:1, v/v), 1 L hexane: ethyl acetate (1:1, v/v), 1 L ethyl acetate, 1 L ethyl acetate: acetone (1:1, v/v), and finally 2 L of acetone. The obtained fractions were applied on a flash chromatography equipped with a silica gel universal column (Yamazen Corporation, 23 ×123 mm, 16 g) connected to an injection column (Yamazen Corporation, 20 × 65 mm, 14 g). The column was eluted by a mixture of chloroform and acetone with an increasing of polarity according to the following concentrations of acetone, 1%, 5%, 10%, 15%, 20%, 25%, 30%, 35%, 65%, and 100%. Fractions by 15% and 20% acetone were concentrated into dryness in vacuo and dissolved in acetonitrile. The acetonitrile solution was applied on a preparative HPLC system equipped with an Agilent Prep-C18 column (21.2 × 250 mm, 10 µm, PN 410910-102, USA). The mobile phase was composed of acetonitrile and water (purified by NANO pure Diamond Life Science ultrapure water system). After loading the sample, the column was eluted by a linear gradient aqueous acetonitrile from 40% to 55% within 0-25 min, FK228 was collected at 21-25 min.
The flow rate was 8 ml/min. The UV spectrum was monitored at 210 nm. The FK228 solution was concentrated into dryness in vacuo on a rotary evaporator at 35 °C. The dried FK228 was dissolved in ethyl acetate and the crystals were obtained at room temperature after 5-7 days.

Refinement
All hydrogen atoms attached to the carbon atoms were placed in geometrically idealized positions (C-H = 0.98, 0.99 and 1.00 Å on the primary, secondary and tertiary aliphatic C atoms respectively, 0.95 Å on aromatic C). The H atoms were refined as riding, with isotropic displacement coefficients of Uiso(H) = 1.5Ueq(C) for methyl groups or 1.2Ueq(C) otherwise. The hydrogen atoms attached to N and O were located in difference maps and refined independently with

Figure 1
A molecular structure of FK228 with displacement ellipsoids shown at the 30% probability level. All hydrogen atoms were omitted for clarity except for H19.    Refinement. Refinement of F 2 against ALL reflections. The weighted R-factor wR and goodness of fit S are based on F 2 , conventional R-factors R are based on F, with F set to zero for negative F 2 . The threshold expression of F 2 > σ(F 2 ) is used only for calculating R-factors(gt) etc. and is not relevant to the choice of reflections for refinement. R-factors based on F 2 are statistically about twice as large as those based on F, and R-factors based on ALL data will be even larger.