The Discovery and Structure‐Activity Evaluation of (+)‐Floyocidin B and Synthetic Analogs

Abstract Tuberculosis represents one of the ten most common courses of death worldwide and the emergence of multidrug‐resistant M. tuberculosis makes the discovery of novel anti‐tuberculosis active structures an urgent priority. Here, we show that (+)‐floyocidin B representing the first example of a novel dihydroisoquinoline class of fungus‐derived natural products, displays promising antitubercular hit properties. (+)‐Floyocidin B was identified by activity‐guided extract screening and its structure was unambiguously determined by total synthesis. The absolute configuration was deduced from a key synthesis intermediate by single crystal X‐ray diffraction analysis. A hit series was generated by the isolation of further natural congeners and the synthesis of analogs of (+)‐floyocidin B. Extensive biological and physicochemical profiling of this series revealed first structure‐activity relationships and set the basis for further optimization and development of this novel antitubercular scaffold.


Experimental Procedures
For the avicennone C-floyocidin B hybrids, the synthetic steps are described in details for the series with the same absolute stereochemistry of the epoxide as (+)-floyocidin B (4) and the corresponding stereoisomers are listed afterwards.
After 3 h 30 min stirring at room temperature, saturated aqueous NH4Cl (20 mL) was added. The mixture was extracted with ethyl acetate (2 x 100 mL) and the combined organic layers were washed with saturated aqueous NaCl (100 mL), dried over MgSO4, filtered, and concentrated under reduced pressure. The crude product was purified by flash column chromatography (n-heptane/ethyl acetate 10:1 to 2:1) to give 29 (4.42 g, 14.1 mmol, 93% over two steps) as colorless solid.
Then DIPEA (10.4 mL, 59.7 mmol, 5.01 eq.) was added and the solution was stirred for 30 min at −78 °C and for further 30 min without cooling bath. A mixture of toluene/saturated aqueous NH4Cl (3:1, 240 mL) was added to the reaction mixture. The aqueous layer was separated and the organic layer was washed again with saturated aqueous NH4Cl (3 x 80 mL), with saturated aqueous NaHCO3 (80 mL), and with saturated aqueous NaCl (80 mL). The organic layer was dried over MgSO4, filtered, and concentrated under reduced pressure to yield aldehyde 30 (11.9 mmol, quant.) as beige solid which was used in the next stage without further purification. For full characterization a small portion of the crude product was purified by flash column chromatography (0-5% ethyl acetate in n-heptane).
After stirring at room temperature for 2 h, the TLC showed some remaining starting material, so a second portion of 1-bromobutane (0.600 mL, 5.60 mmol, 0.20 eq.) was added. The reaction mixture was stirred at room temperature for further 3 h, until the TLC indicated a complete conversion. H2O (50 mL) was added and the mixture was extracted twice with ethyl acetate (150 mL and 100 mL). The combined organic layers were washed with saturated aqueous NaCl (50 mL), dried over MgSO4, filtered, and concentrated under reduced pressure.
The residue was dissolved in DCM (100 mL) and m-CPBA (≥ 70%, 21.8 g, 88.3 mmol, 3.04 eq.) was added carefully at 0 °C. The suspension was allowed to stir overnight at room temperature and then saturated aqueous Na2S2O3-Lsg.
(60 mL) was added. The mixture was extracted with Et2O (2 x 300 mL) and the combined organic layers were washed with 1 M aqueous NaOH (2 x 100 mL) and saturated aqueous NaCl (150 mL), dried over MgSO4, filtered, and concentrated under reduced pressure to yield 31 (6.92 g, 26.0 mmol, 93% over two steps) as slightly yellow oil, which was used in the next step without further purification. The reaction was carried out in moisture-free glassware under argon atmosphere. To a solution of 31 (2.30 g, 8.63 mmol, 1.10 eq.) in anhydrous THF (20 mL) KHMDS (0.7 M in toluene, 13.8 mL, 9.50 mmol, 1.20 eq.) was added dropwise at −55 °C. After 1 h 10 min reaction time, aldehyde 30 (2.44 g, 7.85 mmol, 1.00 eq.) in anhydrous THF (12 mL) was added.
The reaction mixture was stirred for 1 h at −55 °C and was then diluted with saturated aqueous NH4Cl (80 mL). The mixture was extracted with ethyl acetate (2 x 150 mL) and the combined organic layers were washed with saturated aqueous NaCl (100 mL), dried over MgSO4, filtered, and concentrated under reduced pressure. The crude product was purified by flash SI-4 column chromatography (n-heptane/ethyl acetate 10:1 to 9:1) to give 32 (2.07 g, 5.90 mmol, 75%, E/Z = 94:6) as colorless oil.
Prior to the described synthesis of the Julia-Kocieński-olefination, a base screening was performed on 0.650 mmol scale to evaluate isolated yield and E/Z-ratios (see Table S1). KHMDS was chosen due to the observed high E-selectivity.
Toluene (40 mL) was added and the reaction mixture was concentrated in vacuo.
To complete the acetate migration, the product mixture was dissolved in anhydrous DCM (15 mL) and DBU (0.199 mL,  2 A stagnation of the acyl migration was observed after stirring overnight and the reaction was therefore stopped. The two regioisomers could not be separated by chromatography. The minor isomer was finally removed on a later stage after TIPS-protection and saponification by flash chromatography.

2(1aH)-one (35):
The halogen-lithium exchange was carried out in a moisture-free glassware under argon atmosphere. To a solution of nitrile SI10 (0.080 g, 0.15 mmol, 1.00 eq.) in anhydrous THF (15 mL), prediluted n-BuLi solution (0.14 mmol, 1.00 eq.) was added dropwise at −100°C. 7 The obtained orange solution was stirred for 10 min at −100 ºC and for 5 min without cooling bath. For hydrolysis of the intermediate imine, aqueous citric acid (10%, 80 mL) was added and the mixture was extracted with ethyl acetate (100 mL and 50 mL). The combined organic layers were washed with saturated aqueous NaHCO3 (80 mL) as well as with saturated aqueous NaCl (80 mL), dried over MgSO4, filtered, and concentrated under reduced pressure. The crude product was purified by flash column chromatography (0-3% ethyl acetate in n-heptane), to give TIPS-protected 35 (0.059 g, 0.13 mmol) which was directly deprotected in the next step.
In a 50 mL falcon tube, glacial acetic acid (0.011 mL, 0.18 mmol, 1.40 eq.) followed by TBAF in THF (1 M in THF, 0.152 mL, 1.52 mmol, 1.20 eq.) were added to a solution of TIPS-protected 35 in THF (5 mL). After 1 h, a second portion of TBAF (0.40 eq.) and glacial acetic acid (0.60 eq.) and after further 1 h, a third portion of TBAF (0.50 eq.) and glacial acetic acid (0.70 eq.) were added to complete conversion of the starting material which was monitored by TLC. The reaction mixture was allowed to stir overnight at room temperature and was then loaded on silica and filtered through a small silica column (25% ethyl acetate in n-heptane). The product containing fractions were collected, concentrated in vacuo and the residue was purified via flash column chromatography (0-35% ethyl acetate in n-heptane) to give 35 (0.027 g, 0.085 mmol, 58% over two steps, E/Z ≈ 89:11) as colorless oil.

Determination of logD7.4
LogD values at pH 7.4 were determined by a standardized HPLC method derived from by Genieser et al. 8 The calculation of the logD value for measured compounds is performed by comparison of the retention times with standard compounds of known distribution coefficients between 1-octanol and water at pH 7.4.

Microsomal stability
Pooled liver microsomes are purchased.   The permeability coefficient (Papp) for each compound is calculated from the following equation:

Cellular permeability assay utilizing Caco2 cell line
Where dQ/dt is the rate of permeation of the drug across the cells, C0 is the donor compartment concentration at time zero and A is the area of the cell monolayer. C0 is obtained from analysis of the dosing solution.
For bi-directional experiments, an efflux ratio (ER) is calculated from mean A-B and B-A data. This is derived from:

Cytotoxicity on THP1 and HepG2 cells
After 4 hours of HepG2 cell line (provided by ATCC) plating at 2.500 cells /well in 384-w plates, compounds were added to the cells. After an incubation at 37 °C for 40 h, the CellTiterGlo assay (provided by Promega) was performed.
Luminescence measurement is related to the cell viability. Compound activity is determined by using the control wells with no treatment as control (100% of growth). The assay was performed identically with THP1 cells. All incubations were performed in duplicate.

Biosynthetic hypothesis of floyocidins
The structure of the floyocidins indicates a polyketide synthase (PKS)-dependent biosynthesis. 9 Hence, as proposed for (+)-ambuic acid (1) and its derivatives, 10 the backbone structure can be build up from one acetyl-CoA (SI12) and six malonyl-CoA (SI13) building blocks (Scheme S1). The nascent linear polyketide chain SI14 undergoes an intramolecular cyclization that can be catalyzed by a cyclase, resulting in the six-membered ring present in (+)-floyocidin A (3). In contrast to (+)-ambuic acid (1), an additional E-double bond in the hepta-1,3-dien-1-yl chain is located, indicating ketoreduction and dehydration taking place at the second acetate unit, yielding this C=C double bond. The prenyl side chain is presumably acetyl-CoA-derived, too. It can be deduced from dimethylallyl pyrophosphate (DMAPP, SI16), which is an

SI-19
For the formation of the dihydroisoquinolinone scaffold of (+)-floyocidin B (4), a nitrogen atom has to be introduced. Hence, the putative precursor SI23 could be accessible by a selective oxidation of the terminal hydroxyl group and followed by an amidation-cyclization cascade. Alternatively, an amidase could add an ammonia equivalent to the hepta-1,3-dien-1-yl side chain, which then could be oxidatively cyclized. Final oxidative aromatization would complete the proposed biosynthesis of (+)-floyocidin B (4). SI-20

Crystallographic data collection and refinement details
Diffraction data were collected at low temperatures (100K) using φ-and ω-scans on a BRUKER D8 Venture System equipped with dual IµS microfocus sources, a PHOTON100 detector and an OXFORD CRYOSYSTEMS 700 low temperature system. Mo-Kα radiation with a wavelength of 0.71073 Å and a collimating Quazar multilayer mirror were used.
Semi-empirical absorption corrections from equivalents were applied using SADABS. 11 The structure was solved by direct methods using SHELXT 12 and refined against F² on all data by full-matrix least squares using SHELXL. 13 All non-hydrogen atoms were refined anisotropically O-H hydrogen atoms were located in the Fourier difference map and set to ideal distances and C-H hydrogen atoms were positioned at geometrically calculated positions and refined using a riding model.
The isotropic displacement parameters of all hydrogen atoms were fixed to 1.2x or 1.5x (CH3 and OH hydrogens) the Ueq value of the atoms they are linked to. The crystallographic data were deposited with the Cambridge Crystallographic Database as 2075537 and can be obtained free of charge. 14 The structure of 17 was solved in the orthorhombic space group C2221. The asymmetric unit contains one molecule of 17, exhibiting full molecule disorder. The disorder was modelled with the help of same distances restraints, similarity restraints on anisotropic displacement parameters, restraints to a common plane 15 and advanced rigid bond restraints. 16 Some disordered atoms with close coordinates were set to the same anisotropic displacement parameters. The disorder ratio was allowed to refine freely and converged to 0.520(2). The absolute structure was confirmed by classical Flack x (0.011(8)) and the Parsons parameter (0.014(2)). 17 Figure S2. Thermal ellipsoid plot of the molecular structure of 17. Thermal ellipsoid probability set to 50%, only most occupied disorder part shown.