Oxalactam A, a Novel Macrolactam with Potent Anti-Rhizoctonia solani Activity from the Endophytic Fungus Penicillium oxalicum

A novel macrolactam named oxalactam A (1), three known dipeptides (2–4) as well as other known alkaloids (5–7) were obtained from the endophytic fungus Penicillium oxalicum, which was derived from the tuber of Icacina trichantha (Icacinaceae). All chemical structures were established based on spectroscopic data, chemical methods, ECD calculations, and 13C-DP4+ analysis. Among them, oxalactam A (1) is a 16-membered polyenic macrolactam bearing a new skeleton of 2,9-dimethyl-azacyclohexadecane core and exhibited potent anti-Rhizoctonia solani activity with a MIC value of 10 μg/mL in vitro. The plausible biosynthetic pathway of 1 was also proposed via the alanyl protecting mechanism. Notably, three dipeptides (2–4) were first identified from the endophytic fungus P. oxalicum and the NMR data of cyclo(L-Trp-L-Glu) (2) was reported for the first time. In addition, the binding interactions between compound 1 and the sterol 14α-demethylase enzyme (CYP51) were studied by molecular docking and dynamics technologies, and the results revealed that the 16-membered polyenic macrolactam could be a promising CYP51 inhibitor to develop as a new anti-Rhizoctonia solani fungicide.


Introduction
Rice sheath blight (RSB), also called "snakeskin disease", "mosaic footstalk", and "rotten footstalk", is one of the most devastating rice diseases caused by the necrotrophic pathogen Rhizoctonia solani [1,2]. The RSB incidence has witnessed a sharp increase owing to the unrestricted nitrogen fertilizers and semi-dwarf high-yield cultivars in recent decades [3]. In China, RSB was already the second most serious disease in rice, which could lead to a yield loss of 10~30% with 15~20 million hm 2 every year [4,5], even up to 50% in the Yangtze river valley in epidemic years [6,7]. Currently, chemical fungicides are major approaches to preventing RSB and related diseases with increased human health risks, environmental pollution, and resistant phytopathogens [8]. Therefore, natural anti-R. solani agents, such as strobilurins or QoI from Strobilurus tenacellus (wild mushroom) [8], have engaged worldwide attention owing to their environmental friendliness, high selectivity, and new mechanism of action.
Recently, macrolactam scaffolds are usually regarded as precursor components with high efficacy including antibiotics, antifungals, anticancer as well as immunosuppressants [9]. A typical macrocyclic glycopeptide antibiotic, vancomycin, has been applied to treat Gram-positive bacterial infections including methicillin-resistant Staphylococcus aureus and penicillin-resistant Streptococcus pneumonia in the clinic for a long time [9]. Previous investigations indicated that 59% of approved small molecules bearing a macrolactam core are natural products or related derivatives [10]. Therefore, macrolactam components might be a promising lead scaffold for pharmaceutical purposes [11]. However, there is no report on their anti-R. solani activity in agricultural applications.
Icacina trichantha Oliv. (Icacinaceae), distributed in West and Central Africa [12], is an excellent plant resource that has the concomitant function of both medicine and foodstuff to relieve constipation, food poisoning, and malaria [13]. Due to the tropical rainforest climate near the equator in Nigeria, I. trichantha might be a potential plant source to seek various bioactive endophytes producing interesting secondary metabolites. Until now, no report on the endophytic fungus derived from Icacina species and related bioactivity has been documented. Previously, our group found that Penicillium oxalicum from the tuber of I. trichantha could significantly reduce the growth of R. solani strain with a 96.83% of inhibition rate (Alternaria alternata, 83.33%; Fusarium solani, 50.67%; Colletotrichum gloeosporioides, 76.04%) by the plate confrontation test ( Figure S2, Table S1). Therefore, the further anti-R. solani activity-guided fractionation resulted in the isolation of one novel macrolactam (1), three dipeptides (2-4) as well as other alkaloids (5-7) ( Figure 1). Notably, oxalactam A (1) represents the first example of a 16-membered polyenic macrolactam bearing a 2,9-dimethylazacyclohexadecane core, and this is the first report of the NMR data for cyclo(L-Trp-L-Glu) (2). Additionally, three dipeptides (2-4) were first identified from the endophytic fungus P. oxalicum. Among them, oxalactam A (1) exhibited potent anti-R. solani activity with a MIC value of 10 µg/mL by the mycelium growth rate method in vitro. Herein, the isolation, structural elucidation, anti-R. solani activity, molecular docking, and molecular dynamics simulation of these isolates were elaborated.    Figure 2) from one methyl singlet H 3 -18 to C-8/C-9/C-10 suggested the direct connections of C-8, C-10, and C-18 toward C-9. The connection between C-2 and C-3 was verified by the key HMBC correlation from H-3 to C-2. The HMBC correlation from the H-1 to C-17 established the connections of one glucopyranosyl group to the oxygenated primary carbon C-17. Furthermore, H-16 (δ H 3.97)/C-16 (δ C 54.7) were different compared to a typical oxygenated secondary carbon, thus, a nitrogen bridge exists between C-2 and C-16 according to the molecular formula and resulting degrees of saturation of 1. Notably, all three trans double bonds ∆ 4,5 , ∆ 13,14 , and ∆ 9,10 could be established by the coupling constants of However, due to the flexibility of the 16-membered ring, the relative configuration of oxalactam A (1) could not only be constructed from the NOESY data and require a special assignment. The stereochemistry of C-3 and C-15 were confirmed according to the DP4+ probability for theoretical 13    To confirm the assigned structure of 1, (3R*,15S*,16S*)-1/(3R*,15R*,16S*)-1/(3S*,15S*, 16S*)-1/(3S*,15R*,16S*)-1 were further analyzed by the artificial neural networks (ANNs) method [15]. The NMR data of four possible isomers were recalculated at the OPT/HF/ 3-21G and GIAO/PCM/mPW1PW91/6-31G(d,p) levels in methanol, the ANNs calculation results classified four possible isomers of 1 into category 1 (correct) based on the analysis of 18 parameters ( Figures S3-S6), and the ratio of category 1 (correct) to category 2 (incorrect) was calculated as 0. Oxalactam A (1) could be proposed as a 3-aminobutyrate unit in the starter position based on the β-amino acid incorporation pathway in polyketide macrolactam biosynthesis reported previously [16]. Therefore, the absolute configuration of C-16 in 1 could be assigned as S. Moreover, the coupling constant of H-1 (J = 7.8 Hz) assigned the β-glucopyranosyl linkage in 1, and the D-glucose was proved by the comparison of the HPLC retention times of the monosaccharide derivative of hydrolysate of 1 (t R = 21.924 min) with those of the standards D-glucose (t R = 22.179 min) and L-glucose (t R = 19.962 min) ( Figures S28-S30). Further, the absolute configuration of 1 was verified by the electronic circular dichroism (ECD) method with time-dependent density functional theory [17]. The positive Cotton effect (222.4 nm) of 1 was consistent with that (224.2 nm) in the theoretically calculated ECD spectrum of (3R,15S,16S)-1 ( Figure 4). Thus, compound 1 was constructed as depicted and given a trivial name oxalactam A. Oxalactam A (1) was the first example of a 16-membered polyenic macrolactam bearing a 2,9-dimethyl-azacyclohexadecane scaffold, and its biosynthetic pathway was also similar to those of vicenistatin from Streptomyces halstedii [18,19]. As shown in Figure 5, oxalactam A (1) biosynthesis originated from L-glutamate, which could convert into 3-aminobutyrate by VinH/VinI [20]. VinN, a adenylation enzyme, could recognize 3-aminobutyrate and transfer it to 3-aminobutyrate-VinL, which was aminoacylated with L-alanine under the catalysis of VinM to generate a dipeptidyl-VinL. The dipeptidyl group could be tied to the ACP domain of the modular PKS VinP1 by VinK. Then, modular PKSs VinP1-P4 elongated the polyketide chain with a terminal alanyl group, which could be removed by VinJ before intermediate IV generation by VinP4 thioesterase domain. Finally, the hydroxylase and glycosyl-transferase could successfully catalyze IV to yield oxalactam A (1). This alanyl protecting mechanism was also detected in the biosynthetic pathways of some other antibiotics including butirosin [21,22] and desertomycin [23].

Anti-Rhizoctonia solani Activity
The previous investigation indicated that P. oxalicum is a promising fungal agent to prevent plenty of plant diseases [29], including Phytophthora root rot of azalea [30], Pythium seed rot [31], and tomato Fusarium and Verticillium wilts [32]. These applications have aroused our interest to search for active secondary metabolites from the endophytic fungus P. oxalicum as anti-R. solani drugs to relieve rice sheath blight. Consequently, all isolates 1-7 were evaluated for their anti-R. solani activity using the mycelium growth rate method in vitro. Among these alkaloids, compound 1 was the most active component with a 29.30% of inhibition rate at the concentration of 100 µM (MIC = 10 µg/mL), whereas compounds 2-7 did not show potent anti-R. solani activity under the same condition (Table 2, Figures S27 and S28). Up to now, this is the first 16-membered polyenic macrolactam with anti-R. solani activity for agricultural tools.

Assessment of Binding Affinity of 1 and CYP51
The 14α-demethylase enzyme (CYP51, PDB ID: 3GW9) [34] is an essential component of the fungal cell membrane, which played an important role in the fungal-specific ergosterol biosynthesis pathway. Azole antifungals (hexaconazole for example) could compete with the CYP51 substrate by binding to the heme iron in the active site in a ligand-binding pocket, leading to ergosterol depletion, membrane fluidity reduction, and lipid layer destruction [35]. To analyze the binding mode of compound 1 to CYP51 enzyme, molecular docking and dynamics studies have been performed using Discovery Studio 2019. The result illustrated that 18-methyl of 1 showed favorable steric interactions with LUE356, MET460, and VAL461 residues (Figure 6), and MET460 might cause a misfolded fungus-specific loop that affects the binding efficiency of the cognate NADPH-cytochrome P450 reductase [36]. Moreover, the 1,2,4-triazole and 1,3-dichlorobenzene moieties of hexaconazole were observed to generate interactions with the ALA291, CYS422, and ILE423 (Figure 7), and the ALA291 substitution is vital for the microbial resistance to triazole drugs including voriconazole and fluconazole [37]. Notably, it showed that the higher negative binding energy value of 1 (−202.0196 kcal/mol) compared to that of hexaconazole (−105.4279 kcal/mol) indicated more favorable binding interactions between compound 1 and the CYP51 enzyme (Table 3) [33], which could explain the reason for the superiority of compound 1 over hexaconazole.   The steric interactions between CYP51 enzyme and oxalactam A (1)/hexaconazole were further estimated by molecular dynamics simulation according to RMSD (Root-meansquare deviation), RMSF (Root-mean-square fluctuation), and the total energy ( Figure 8). Accordingly, RMSD values suggested that both complexes could reach an equilibrium state at 0.6-0.9 nm of the vibration amplitude within 200 ps. Additionally, the active and binding sites (THR295-LEU359/GLY414-VAL461) of CYP51-1 have a slightly higher RMSF than those of the CYP51-hexaconazole, indicating the CYP51-hexaconazole complex possesses better thermodynamic stability. Notably, the CYP51-1 complex showed lower total energy values than that of CYP51-hexaconazole with 55 kcal/mol per picosecond, which demonstrated that the binding affinity of CYP51-1 was much stronger than that of CYP51-hexaconazole. These results indicated that compound 1 could directly bind to the high-affinity catalytic areas of the CYP51 enzyme and yield stable enzyme-ligand complexes in the saline condition.  The steric interactions between CYP51 enzyme and oxalactam A (1)/hexaconazole were further estimated by molecular dynamics simulation according to RMSD (Root-mean-square deviation), RMSF (Root-mean-square fluctuation), and the total energy ( Figure 8). Accordingly, RMSD values suggested that both complexes could reach an equilibrium state at 0.6−0.9 nm of the vibration amplitude within 200 ps. Additionally, the active and binding sites (THR295−LEU359/GLY414−VAL461) of CYP51−1 have a slightly higher RMSF than those of the CYP51−hexaconazole, indicating the CYP51−hexaconazole complex possesses better thermodynamic stability. Notably, the CYP51−1 complex showed lower total energy values than that of CYP51−hexaconazole with 55 kcal/mol per picosecond, which demonstrated that the binding affinity of CYP51−1 was much stronger than that of CYP51−hexaconazole. These results indicated that compound 1 could directly bind to the high-affinity catalytic areas of the CYP51 enzyme and yield stable enzyme−ligand complexes in the saline condition.

General Experimental Procedures
A UV-2600 spectrophotometer (Shimadzu, Kyoto, Japan) was applied to measure the UV absorption spectra. An MCP 150 digital polarimeter (Anton Paar, Graz, Austria) was utilized to gain the optical rotations. A J-810 spectrometer (JASCO, Kyoto, Japan) could be exerted to acquire CD spectra in methanol at room temperature. The HR-ESI-MS

General Experimental Procedures
A UV-2600 spectrophotometer (Shimadzu, Kyoto, Japan) was applied to measure the UV absorption spectra. An MCP 150 digital polarimeter (Anton Paar, Graz, Austria) was utilized to gain the optical rotations. A J-810 spectrometer (JASCO, Kyoto, Japan) could be exerted to acquire CD spectra in methanol at room temperature. The HR-ESI-MS data could be collected by a Waters SYNAPT G2-Si Q-TOF mass spectrometer (Waters, Milford, MA, USA). The NMR spectra were recorded on a Bruker AV-500 spectrometer (Bruker, Karlsruhe, Germany). The semi-preparative RP-HPLC included a dual-wavelength detector, a Shimadzu LC-20AR instrument, and an Ultimate XB-C 18 column (10 × 250 mm, 5 µm, Welch).

Strain Material
The test endophytic fungus was obtained from the fresh tubers of the West African plant I. trichantha, the Orba village in Nsukka of the Enugu State, Nigeria, in June 2011. It was identified as P. oxalicum based on its morphological characteristics and 16S rRNA gene sequence by Shanghai Shenggong Bioengineering Co., Ltd. A voucher sample (ZRZ20210301) was stored in the Key Laboratory of prescriptions of Nanjing University of traditional Chinese medicine.

Fermentation and Isolation
The potato dextrose agar (PDA) was used to culture the P. oxalicum strain at room temperature for 72 h. Next, the PDA was cut into small pieces and then inoculated into a mass of sterilized Erlenmeyer flasks containing 10 g glucose, 4 g soluble starch, 2.5 g peptone, 1 g sodium chloride, 1 g calcium carbonate, 0.25 g magnesium sulfate, 1 g yeast extract, 0.25 g dipotassium hydrogen phosphate, and 500 mL distilled water. All flasks were incubated at room temperature for 72 h. The culture broths (82 L) and the smashed mycelium were extracted by ethanol three times, then combined and concentrated under reduced pressure to yield a crude extract, which was partitioned between ethyl acetate and water to obtain the EtOAc-soluble extract (42.58 g). This part was loaded on a 200~300 mesh silica gel eluted with CH 2 Cl 2 -MeOH (100:0 to 0:100, v/v) to furnish eight fractions (Fr.1~Fr.8).

Enzymatic Hydrolysis
Oxalactam A (1) (2 mg) and β-cellulase (0.5 mg) were dissolved in an aqueous solution (2 mL) at 50 • C for 4 h. The reaction solution was extracted by EtOAc five times, 5 mL each. The EtOAc extract liquor was combined and evaporated, and analyzed by the TLC and HPLC methods. Finally, the aglycone of compound 1 was purified by a semipreparative HPLC with an Ultimate XB-C 18 column.

Sugar Identification
The Enzymatic Hydrolysis mixtures were analyzed using HPLC with an Ultimate XB-C 18 column (CH 3

ECD and NMR Calculation Methods
HyperChem 8.0 program was carried out to search for the most stable stereoisomers of compound 1 with molecular mechanics MMFF94s. Gaussian 16 software was applied to optimize geometries to screen minimum-energy conformers at the OPT/B3LYP/6-31+G(d) (OPT/HF/3-21G for ANNs analysis) level. The optimized conformers were submitted to the ECD calculation program with the TD/B3LYP/6-311+G(d) level in methanol. The overall ECD curves were summed up by the Boltzmann distribution averaging of all conformers with SpecDis 1.71 software [39]. Additionally, the 13 C NMR shifts calculation of all stable conformers of 1 was conducted with the GIAO/PCM/mPW1PW91/6-311G(d,p) (GIAO/PCM/mPW1PW91/6-31G(d,p) for ANNs analysis) level in methanol. The DP4+ probability and linear regression were obtained by analyzing overall theoretical NMR data to verify the relative configurations of C-3/C-15 in 1 [14].

The Plate Confrontation Test
The fungal blocks of Rhizoctonia solani, Alternaria alternata, Fusarium solani, and Colletotrichum gloeosporioides (each diameter 5 mm) were, respectively inoculated in the center of the PDA plate (each diameter 150 mm), and purified endophytic fungus P. oxalicum was inoculated in the surroundings. All plates were incubated at room temperature for 72 h. Each treatment was repeated three times and the sterile water was used as a blank control. The antagonistic effect was determined by the width of the inhibition belt when colonies fully covered the whole Petri dish in blank control.

Anti-Rhizoctonia solani Assays In Vitro
The sensitivity of R. solani to compounds 1-7 from the endophytic fungus P. oxalicum was measured by the mycelium growth rate method in vitro. This phytopathogenic fungus, R. solani, was provided by Beijing Baioubowei Biotechnology Co., Ltd.
A mycelial plug (1 cm in diameter) was cut from the beforehand PDA medium fully covered by R. solani colonies, which was placed with their mycelia-side down on the PDA medium in the center of each plate containing 100 µM of pure compounds 1-7. The positive control is hexaconazole (100 µM), a commercial fungicide. An equal concentration of DMSO solution served as the solvent control. A thermostatic chamber was applied to inoculate the treatment plates, and the average diameter of each R. solani colony was recorded based on the previous literature [40]. The anti-R. solani activity was evaluated in terms of the inhibition rate, which was performed in triplicate for each treatment. The inhibition ratio could be acquired by the following formula: Inhibition (%) = (A − B)/A × 100% (A: average diameter of the control group, B: average diameter of the treatment group) [41].

Molecular Docking Study
The molecular docking simulation [42] of compound 1 and hexaconazole to the sterol 14α-demethylase enzyme (CYP51, PDB ID: 3GW9) was performed by the CDOCKER protocol of Discovery Studio 2019 (BIOVIA, San Diego, CaA). The CYP51 X-ray crystal structure was downloaded from RCSB Protein Data Bank (http://www.rcsb.org, accessed on 1 June 2022). All conformations of compound 1 were searched by ChemHyper 8.0 program with the density functional method (6-31G*). The CYP51 protein was protonated and deleted water at pH 7 using Prepare Protein tool under the CHARMm forcefield [43]. Additionally, the implicit solvent model was selected as the Generalized Born with Molecular Volume method. Additionally, the binding energy of CYP51-1/hexaconazole was calculated by the Spherical Cutoff mode. All other parameters were set as default.

Molecular Dynamics Simulation
Molecular dynamics simulation [42] was also performed for CYP51-1/hexaconazole based on molecular docking results using the Standard Dynamics Cascade protocol of Discovery Studio 2019 (BIOVIA, San Diego, CA, USA). This software brings together over 30 years of peer-reviewed research and world-class in silico techniques such as molecular me-chanics, free energy calculations, and biotherapeutics developability and more into a common environment, and it could be available at the website: https://www.3ds.com/productsservices/biovia/products/molecular-modeling-simulation/biovia-discovery-studio/ (accessed on 1 January 2022).
The CYP51-1/hexaconazole complex was in the aqueous environment with the Extended Simple Point Charge water model under the CHARMm force field [44]. The entire system was equilibrated under NVT (isothermal-isochoric) and NPT (isothermal-isobaric) ensembles to maintain the stabilized pressure [45]. All hydrogen bonds were constrained during equilibration using LINC algorithms [46]. Besides, the Particle Mesh Ewald module was applied for the long-range ionic interaction [47]. Finally, the entire trajectories were saved for analysis at a frequency of 1 ps.

Conclusions
In summary, a novel macrolactam (1), three dipeptides (2-4) as well as other alkaloids (5-7) were isolated from the fermentation liquid of P. oxalicum derived from the tuber of I. trichantha. Among them, only oxalactam A (1) displayed anti-R. solani activity at 100 µM with a 29.30% of inhibition rate in vitro (MIC = 10 µg/mL). Notably, the content of 1 is more than 1.70 mg/L in the fermentation liquid of the endophytic fungus P. oxalicum. These findings provided an alternative natural resource to obtain novel anti-R. solani macrolactam leads.
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/molecules27248811/s1, Figure S1: Flow chart of extraction and isolation; Figure Table S1: Penicillium oxalicum resisted the growth of four plant pathogenic fungi strains; Table S2: Cartesian coordinate for the lowest energy conformer of 1 calculated at TD/B3LYP/6-311G(d,p) level of theory in the gas phase; Table S3: Experimental and calculated chemical shifts of 1 for DP4+ probability statistical analysis.