Structure Revision and Protein Tyrosine Phosphatase Inhibitory Activity of Drazepinone

From the marine-derived fungus Penicillium sumatrense (Trichocomaceae), a pair of enantiomers [(+)-1 and (−)-1] were isolated with identical 1D NMR data to drazepinone, which was originally reported to have a trisubstituted naphthofuroazepinone skeleton. In this study, we confirmed the structures of the two enantiomers as drazepinone and revised their structures by detailed analysis of extensive 2D NMR data and a comparison of the calculated 13C chemical shifts, ECD, VCD, and ORD spectra with those of the experiment ones. (+)-1 and (−)-1 were evaluated for their PTP inhibitory activity in vitro. (−)-1 showed selective PTP inhibitory activity against PTP1B and TCPTP with IC50 values of 1.56 and 12.5 μg/mL, respectively.


Introduction
Without the assistance of single-crystal X-ray diffraction analysis, many complex natural products that exhibited important biological activities and played a major role in drug discovery were assigned to erroneous structures or could not be unambiguously established [1]. Drazepinone (a previously proposed structure as a trisubstituted tetrahydronaphthofuroazepinone core; Figure 1), a specialized metabolite that is a type of alkaloid, was obtained for the first time in 2005 from a pathogenic fungal strain of Drechslera siccans from seeds of Lolium perenne [2]. In biological activity studies, drazepinone has been proved to have broad-spectrum herbicidal properties with low zootoxic activity at a concentration of 2.0 µg/µL, suggesting that it could be applied as an environmentally friendly and safe herbicide [2]. During the course of our ongoing research on bioactive natural products from marine-derived fungi [3,4], an extract of the fungal strain Penicillium sumatrense (Trichocomaceae) showed inhibitory activity against protein tyrosine phosphatase (PTP). Our chemical investigations of this fungus using a bioassay-guided method led to the isolation of compounds (+)-1 and (−)-1, comprising a racemate with identical 1D NMR data to drazepinone. However, we found that some 2D NMR data of drazepinone ( 1 H-1 H COSY and key HMBC) were incorrectly interpreted, prompting us to carefully analyze the NMR spectroscopic data of (+)-1 and (−)-1, and to revise the structure of drazepinone ( Figure 1).
By the rapid development of modern strategies and methods for structural elucidation, DFT calculations of 13 C chemical shifts and chiral spectra proved to be critical in several profile structure revisions and have greatly facilitated the reliable determination of natural products with undescribed structures [5,6]. In the present work, we thus tried to definitively By the rapid development of modern strategies and methods for structural elucidation, DFT calculations of 13 C chemical shifts and chiral spectra proved to be critical in several profile structure revisions and have greatly facilitated the reliable determination of natural products with undescribed structures [5,6]. In the present work, we thus tried to definitively determine the structure of drazepinone by detailed analysis of 2D NMR data and quantum-mechanics-based computational studies of 13 C chemical shifts, electronic circular dichroism (ECD), vibrational circular dichroism (VCD), and optical rotatory dispersion (ORD) of (+)-1 and (−)-1. Here we report a definitive revision of the structures of (+)-1 and (−)-1 and also describe their PTP inhibitory activity.

Structural Elucidation and Revision
The fungal strain penicillium sumatrense was cultivated using a rice medium. Purification of the EtOAc extract of the fungus by HPLC yielded 1. It is worth noting that the optical rotation of 1 was near zero, indicating its racemic nature. Finally, the mixture of (±)-1 was separated using a Chiralpak IB column to yield (+)-1 and (−)-1.
Compounds (+)-1 and (−)-1 were isolated as white amorphous substances with the molecular formula C19H19NO2 on the basis of their HRESIMS data at m/z 294.1464 [M + H]+ ion. The 1 H and 13 C NMR data of (+)-1 and (−)-1 (Tables 1 and S1) displayed three methyls, nine methines including seven aromatic carbons, and seven non-protonated carbons. These findings were in agreement with 1D NMR and MS data reported for drazepinone [2]. However, in the interpretation of the 1 H-1 H COSY and key HMBC data for (+)-1 and (−)-1, some important correlations were inconsistent with the proposed structure of drazepinone. For example, when the HMBC correlations for the proposed structure were interpreted, strong HMBC correlation from H-6 (δH 7.43) to C-2 (δC 162.6), an unreasonable six-bond correlation, could not be well explained ( Figure 2). Similarly, the HMBC correlations from H-14 (δH 1.29) to C-3 (δC 150.5), and from H-15 (δH 1.65) to C-4 (δC 126.3), which represent uncommon four-bond and five-bond correlations, respectively, were not taken into consideration for the proposed structure of drazepinone. To further confirm the above deduction, 13 C NMR chemical shift calculation based on the gauge-independent atomic orbital (GIAO) [7] was performed for the proposed structure of drazepinone at the level of B3LYP/6-311G+(d,p). Unfortunately, a very low correlation coefficient (R 2 ) of 0.9492 was given for the proposed structure ( Figure 3). Moreover, this procedure gave a mean deviation (|∆δ|mean) of 6.1 ppm and a maximum deviation (|∆δ|max) of 29.9 ppm between the experimental and calculated (corrected) 13 C chemical shifts for the wrong structural assignment (Figure 3). Thus, the structures of (+)-1 and (−)-1 should be reassigned.

Structural Elucidation and Revision
The fungal strain Penicillium sumatrense was cultivated using a rice medium. Purification of the EtOAc extract of the fungus by HPLC yielded 1. It is worth noting that the optical rotation of 1 was near zero, indicating its racemic nature. Finally, the mixture of (±)-1 was separated using a Chiralpak IB column to yield (+)-1 and (−)-1.
Compounds (+)-1 and (−)-1 were isolated as white amorphous substances with the molecular formula C 19 H 19 NO 2 on the basis of their HRESIMS data at m/z 294.1464 [M + H]+ ion. The 1 H and 13 C NMR data of (+)-1 and (−)-1 (Tables 1 and S1) displayed three methyls, nine methines including seven aromatic carbons, and seven non-protonated carbons. These findings were in agreement with 1D NMR and MS data reported for drazepinone [2]. However, in the interpretation of the 1 H-1 H COSY and key HMBC data for (+)-1 and (−)-1, some important correlations were inconsistent with the proposed structure of drazepinone. For example, when the HMBC correlations for the proposed structure were interpreted, strong HMBC correlation from H-6 (δ H 7.43) to C-2 (δ C 162.6), an unreasonable six-bond correlation, could not be well explained ( Figure 2). Similarly, the HMBC correlations from H-14 (δ H 1.29) to C-3 (δ C 150.5), and from H-15 (δ H 1.65) to C-4 (δ C 126.3), which represent uncommon four-bond and five-bond correlations, respectively, were not taken into consideration for the proposed structure of drazepinone. To further confirm the above deduction, 13 C NMR chemical shift calculation based on the gauge-independent atomic orbital (GIAO) [7] was performed for the proposed structure of drazepinone at the level of B3LYP/6-311G+(d,p). Unfortunately, a very low correlation coefficient (R 2 ) of 0.9492 was given for the proposed structure ( Figure 3). Moreover, this procedure gave a mean deviation (|∆δ| mean ) of 6.1 ppm and a maximum deviation (|∆δ| max ) of 29.9 ppm between the experimental and calculated (corrected) 13 C chemical shifts for the wrong structural assignment ( Figure 3). Thus, the structures of (+)-1 and (−)-1 should be reassigned.

General Experimental Procedures
ORD data were measured using a JASCO P-2000 spectrometer in CH 3 OH. UV and ECD spectra were performed using a Perkin-Elmer model 241 spectrophotometer and a JASCO J-715 CD spectrometer, respectively. VCD spectra, including the corresponding IR spectra, were acquired using a BioTools ChiralIR-2X spectrophotometer. NMR data with TMS as an internal standard were measured using Bruker Avance-III 600 MHz NMR spectrometer. HRESIMS data were obtained from a Bruker apex-ultra 7.0T spectrometer. Semipreparative HPLC, using Waters (XBridge OBD, 5 µm, 10 × 250 mm) and Daicel (Chiralpak IB, 5 µm, 10 × 250 mm) columns, was carried out on a Shimadzu LC-20AT system with a SPD-M20A photodiode array detector. Column chromatography was performed on Silica gel 200-300 and 300-400 mesh, and Sephadex LH-20 18−110 µm.

Enzyme Inhibitory Activity Assay
PTP inhibitory activity of (+)-1 and (−)-1 against PTP1B, TCPTP, SHP1, and CD45 was tested in the same way as that described in the literature [19]. Na 3 VO 4 was used as the positive control.

Molecular Docking
The complex crystal structure of PTP1B-inhibitor (PDB: 5T19) [20] or TCPTP-inhibitor (PDB: 2FJM) [21] was used as the starting model for molecular docking employing Discovery Studio 2017 software. Before docking calculations, conformational searches of compounds (+)-1 and (−)-1 were performed with GMMX conformer calculation (force field: MMFF94; energy window: 5.0 kcal/mol) was performed in GaussView 6.0. Then, the top 2 conformations with the lowest energy were optimized at the B3LYP/6-31G(d) basis set using density functional theory (DFT) in Gaussian 16. To simulate real conditions, the solvent effects of water were studied using the solvation model based on density (SMD). For the protein, protein preparation processes were carried out, such as removing water molecules, adding hydrogen atoms, and supplementing amino acid residues. The flexible docking protocol, which allows for some receptor flexibility during docking of flexible ligands [22], was used in this study employing CHARMm in Discovery Studio 2017 software. The receptor binding sites were determined from the location of ligand in complex PDB: 5T19 or 2FJM.

Conclusions
In conclusion, based on combinational analysis of 2D NMR data, 13 C NMR chemical shift calculation, and computational studies of ECD, VCD, and ORD data, the structure of drazepinone was revised. Furthermore, experimental and molecular docking of the inhibitory effect between the revised (+)-1 and (−)-1 against PTPs were investigated. The complexity of intriguing structures and the significance of bioactivity for (+)-1 and (−)-1 may encourage further investigations on the chemistry and activity of this cluster of metabolites.