The Discovery of Acremochlorins O-R from an Acremonium sp. through Integrated Genomic and Molecular Networking

The fermentation of a soil-derived fungus Acremonium sp. led to the isolation of thirteen ascochlorin congeners through integrated genomic and Global Natural Product Social (GNPS) molecular networking. Among the isolated compounds, we identified two unusual bicyclic types, acremochlorins O (1) and P (2), as well as two linear types, acremochlorin Q (3) and R (4). Compounds 1 and 2 contain an unusual benzopyran moiety and are diastereoisomers of each other, the first reported for the ascochlorins. Additionally, we elucidated the structure of 5, a 4-chloro-5-methylbenzene-1,3-diol with a linear farnesyl side chain, and confirmed the presence of eight known ascochlorin analogs (6–13). The structures were determined by the detailed interpretation of 1D and 2D NMR spectroscopy, MS, and ECD calculations. Compounds 3 and 9 showed potent antibacterial activity against Staphylococcus aureus and Bacillus cereus, with MIC values ranging from 2 to 16 μg/mL.


Introduction
Ascochlorins (ASCs) and their congeners are primarily produced by filamentous fungi and represent a unique class of polyketide-terpenoid hybrid natural products.They are generally characterized by the presence of an orsellinic acid unit combined with a sesquiterpene (C 15 ) moiety [1].They have attracted extensive attention from pharmacologists.Owing to their distinctive structural diversity, they have been reported to exhibit a range of promising biological activities including antitumor [2], anti-inflammatory [3], antimicrobial [4,5], and anti-trypanosome [6,7].Recent studies indicated that they showed potent hDHODH inhibitory activity, which means they have been involved in the treatment of cancers.
A recent review summarized their structure isolation and identification, biosynthesis, and biological activity in 2023.To date, about 71 ASCs have been reported from filamentous fungi and are classified into three main chemotypes, the linear type, the monocyclic type, and the bicyclic type, which account for about 30%, 65%, and 0.6% of all the ASCs, respectively [8].In addition, the gene cluster and biosynthesis of the representative products of ASCs, such as ascofuranone and ascochlorin, have been characterized, which are conducive to the discovery of more ASC congeners by genomic mining [1].
The filamentous fungus Acremonium sp. was isolated from soil collected on the University of Utah campus, UT, USA, and was shown to produce a rare class of 15-residue peptaibols [9].In that study, the 38 Mbp genome of the strain was also reported to harbor 44 putative biosynthetic gene clusters, including 1 predicted for ASCs' biosynthesis.As no ASCs had been previously reported from the target strain, we implemented genomic and GNPS molecular networking to determine if this orphan biosynthetic cluster was indeed responsible for the production of ASCs.Guided by UV absorption and MS data, we identified four undescribed ASCs (1)(2)(3)(4), a newly natural product (5), and eight known ASC analogs (6)(7)(8)(9)(10)(11)(12)(13).Herein, details of their isolation, structure elucidation, and antibiotic activities are described.

General Experimental Procedures
Optical rotations were measured on an Anton Paar MCP 5500 polarimeter equipped with a sodium lamp (589 nm) and a 25 mm microcell.A Jasco J1500 spectrometer (Jasco Inc., Tokyo, Japan) was used to obtain the electric circular dichroism (ECD) spectra.The 1D and 2D NMR experiments ( 1 H, 13 C, NOESY, COSY, HSQC, and HMBC) were performed at 300 K in CDCl 3 on a Bruker Avance Neo 600MHz spectrometer (Bruker BioSpin, GmBH) equipped with a Bruker 5 mm PI HR-BBO600s3 Probe.HR-ESIMS was utilized on an LTQ Orbitrap XL mass spectrometer or an Orbitrap Eclipse mass spectrometer (Thermo Fisher Scientific, Waltham, MA, USA) equipped with Xcalibur 4.0 software.The 200-300 mesh silica gel (Shanghai, China), ODS (12 nm, YMC*GEL), and TOYOPEARL HW-40F were employed for column chromatography.HPLC separations were conducted on an Agilent 1260 series pumping system equipped with an Agilent DAD-G7115A refractive index detector on an X-bridge C 18 column (4.6 × 250 mm, 5 µm, flow rate 1.0 mL/min).RP-HPLC separations were conducted on a Shimadzu LC-20AP series pumping system equipped with a Shimadzu SPD-20A refractive index detector and Shimadzu injector on an X-bridge C 18 column (10 × 250 mm, 5 µm, flow rate 4.0 mL/min).TLC analyses were carried out using precoated HF254 (0.20 mm thickness) plates (Nuotai, Shanghai, China); compounds were detected by 10% sulfuric acid/ethanol (Sigma-Aldrich, St. Louis, MO, USA).All MS/MS data were converted to mzXML format files by ProteoWizard 3.0 software [10].Molecular networking was performed using the GNPS data analysis workflow and the spectral clustering algorithm.The spectral networking was imported into Cytoscape (ver.3.6.1)for visualization.

Fungal Material
The fungal strain Acremonium sp.(GenBank accession number MT053262) was originally isolated from soil collected on the University of Utah campus, UT, USA, and formally characterized as an Acremonium sp. in a previous publication from our group [9].

Incubation and Extraction
Acremonium sp. was cultured on the seed medium Potato Dextrose Agar plates (PDA medium: 20 g of potato extract powder, 20 g of glucose, 18 g of agar in 1 L of tap distilled H 2 O) at 28 • C for four to five days.Subsequently, the large-scale fermentation of Acremonium sp. was performed using modified rice solid medium (80 g of rice, 3.0 g/L of NaNO 3 , and 120 mL of H 2 O).Briefly, 120 mL of rice medium was added to a 48 × 1 L Erlenmeyer flask and inoculated using two 5 × 5 mm 2 agar plugs from the PDA plates.The cultures were incubated at room temperature under static conditions, and after 30 days, they were harvested for chemical analysis.The fermented cultures were extracted three times with equal volumes of EtOAc, and the organic extracts were combined and concentrated under vacuo to provide a crude extract (87.1 g).

Isolation and Purification
Using normal silica gel chromatography, the crude extract was fractionated into nine fractions using different concentrations of petroleum ether, dichloromethane, and methanol.Fr.2 (19.1 g) eluted with 100% dichloromethane was further separated into 7 subfractions (Frs.

Antimicrobial Activities
The minimum inhibitory concentrations (MICs) were determined in 96-well plates using the microdilution method to screen compounds 1-9 and 11 for bioactivity.To prepare the inoculum for susceptibility testing, bacteria were streaked independently onto LB agar plates and incubated overnight at 37 • C. Individual colonies were then isolated and transferred to 50 mL of LB liquid medium and incubated at 37 • C for 4-6 h.The culture density was adjusted with LB liquid medium so that a concentration of 5 × 10 6 cfu/mL was achieved.Compounds 1-9 and 11 were tested for their individual activity against S. aureus, MRSA, MRCNS, B. Subtilis, and B. cereus using chloramphenicol as a positive control (64 µg/mL dissolved in DMSO).Briefly, 1-9 and 11 were dissolved in DMSO to generate 128 mg/mL stock solutions.The stock solutions were then serially diluted with LB liquid medium to afford working concentrations of 128 to 2 µg/mL.More specifically, to a 96-well microtiter plate, 2 µL stock solutions mixed with 98 µL of LB liquid medium was added to well A1.From this mixture, 50 µL was transferred to well A2 and mixed with 50 µL of fresh LB media.This process was repeated across the 96-well plate, and 50 µL of the appropriate bacterial cultures was then added to each well.The plates were incubated at 37 • C for 16-20 h.MIC values were determined by visual inspection and verified with an OD 600 measurement using a BioTek Neo2 plate reader (Agilent, Winooski, VT, USA).The respective MIC values for 1-9 and 11 are reported in Table 2.All the pathogenic strains were clinical isolates and donated by the Marine Medicinal Biological Resources Center, Ocean University of China.Specific strain information can be found in Table S2.All assays were performed in triplicate.

Hydroxyl Radical Scavenging Activity
The Fenton reaction was used to produce hydroxyl radicals, which reacted with salicylic acid to form 2,3-dihydroxybenzoic acid with special absorption at 510 nm.
The test compound was prepared into a 200 mM solution with DMSO as the solvent.Then, 25 µL of the 200 mM sample solution, 25 µL of 9 mM FeSO 4 •7H 2 O, 25 µL of 9 mM salicylic acid, and 25 µL of 8.8 mM H 2 O 2 were added into the 96-well plate successively and mixed well.After heating in a 37 • C water bath for 30 min, it was taken out and its absorbance at 510 nm was measured using a BioTek Neo2 plate reader.

Results and Discussion
The 38 Mbp genome of Acremonium sp. was previously sequenced and assembled, and the antiSMASH analysis revealed that a gene cluster, hereby named ascw, showed high similarity (87%) at the amino acid level to the characterized asc-1 gene cluster (Figures 1 and S1).A more detailed bioinformatic analysis of the ascw gene cluster revealed that all eight of the genes encoding enzymatic machinery responsible for ASCs' assembly in A. egyptiacum were present in ascw, suggesting the ability of Acremonium sp. to produce ASCs [1].To evaluate whether the strain significantly produced ASCs or not, the fungal strain was cultured in rice media (80 g of rice, 3.0 g/L of NaNO3, and 120 mL of H2O) for 30 days.The EtOAc extract was evaluated by LC-MS/MS in the positive mode, and the data were processed through GNPS (http://gnps.ucsd.edu,accessed on 22 December 2023).The obtained molecular networking featured 13 clusters and 91 nodes, with GNPS analysis uncovering a cluster of 23 nodes matching ASC compounds, which displayed typical isotopic To evaluate whether the strain significantly produced ASCs or not, the fungal strain was cultured in rice media (80 g of rice, 3.0 g/L of NaNO 3 , and 120 mL of H 2 O) for 30 days.
The EtOAc extract was evaluated by LC-MS/MS in the positive mode, and the data were processed through GNPS (http://gnps.ucsd.edu,accessed on 22 December 2023).The obtained molecular networking featured 13 clusters and 91 nodes, with GNPS analysis uncovering a cluster of 23 nodes matching ASC compounds, which displayed typical isotopic peaks for monochloride compounds in the grouped structure (Figures 2, S2 and S3).Known compounds 8 and 13 were directly identified by molecular networking.Based on the UV absorption of known compounds, further targeted isolation resulted in a total of 13 Ascochlorin (ASC) derivatives (Figure 2).To evaluate whether the strain significantly produced ASCs or not, the fungal strain was cultured in rice media (80 g of rice, 3.0 g/L of NaNO3, and 120 mL of H2O) for 30 days.The EtOAc extract was evaluated by LC-MS/MS in the positive mode, and the data were processed through GNPS (http://gnps.ucsd.edu,accessed on 22 December 2023).The obtained molecular networking featured 13 clusters and 91 nodes, with GNPS analysis uncovering a cluster of 23 nodes matching ASC compounds, which displayed typical isotopic peaks for monochloride compounds in the grouped structure (Figures 2, S2 and S3).Known compounds 8 and 13 were directly identified by molecular networking.Based on the UV absorption of known compounds, further targeted isolation resulted in a total of 13 Ascochlorin (ASC) derivatives (Figure 2).Compound 1 was obtained as a yellow oil.The HR-ESIMS exhibited a characteristic pseudomolecular ion at m/z 405.1840/407.1804 in a ratio of 3:1 ([M+H] + /[M+2+H] + , calcd for C23H30ClO4, 405.1827/407.1798),suggesting the presence of a chlorine atom in 1 and supporting the molecular formula of C23H31ClO5 containing nine degrees of unsaturation.Further analysis of the 1D NMR and HSQC data showed the presence of a hexasubstituted benzene moiety (δC 113.6, 158.9, 107.9, 156.3, 116.2, and 140.8), a non-conjugated ketone carbon (δC 213.8), one aldehyde carbon (δC/H 193.5/10.13), a double bond (δC/H 116.2/6.73 and 126.5/5.53),five methyls, four sp 3 methylenes, two sp 3 methines, and two quaternary carbons (one oxygenated carbon δC 82.0 and one sp 3 δC 43.3), indicating the existence of two additional ring systems in the structure of 1 (Table 1).1).
The comparison of NMR data between 1 and co-isolated ilicicolin C (9) revealed that both compounds share identical monochlorinated benzaldehyde and cyclohexanone moieties (Figure 3).In 1, the double bond is located between C-9 and C-10, whereas it is in position between C-10 and C-11 in 9.This was confirmed by the 1  HMBC correlations of H-9 with C-2 (δC 158.9),C-3 (δC 107.9),C-4 (δC 156.3), and Me-20 (δC/H 15.6/0.58) with C-13 (δC 30.8) allowed us to establish the connections of the monochlorinated benzaldehyde and cyclohexanone group by a single bond between C-3 and C-9.To satisfy the degree of unsaturation, the molecular formula and downfield chemical shift of C-11 (δC 82.0), a benzopyran moiety, was proposed, thus assigning the planar structure of 1.  Compound 2 was obtained as yellow oil and was determined by HR-ESIMS data to be at 405.1843/407.1807(3:1) ([M+H] + /[M+2+H] + , calcd for C23H30ClO4, 405.1827/407.1798)and to have the same molecular formula of C23H29ClO4 as 1.When isolated using HPLC, compounds 1 and 2 eluted as adjacent peaks (Figure S4).The detailed inspection of 1D and 2D NMR data revealed that compounds 1 and 2 share identical planar structures.HMBC correlations of H-9 with C-2 (δC 158.9),C-3 (δC 107.9),C-4 (δC 156.3), and Me-20 (δC/H 15.6/0.58) with C-13 (δC 30.8) allowed us to establish the connections of the monochlorinated benzaldehyde and cyclohexanone group by a single bond between C-3 and C-9.To satisfy the degree of unsaturation, the molecular formula and downfield chemical shift of C-11 (δC 82.0), a benzopyran moiety, was proposed, thus assigning the planar structure of 1.  Compound 2 was obtained as yellow oil and was determined by HR-ESIMS data to be at 405.1843/407.1807(3:1) ([M+H] + /[M+2+H] + , calcd for C23H30ClO4, 405.1827/407.1798)and to have the same molecular formula of C23H29ClO4 as 1.When isolated using HPLC, compounds 1 and 2 eluted as adjacent peaks (Figure S4).The detailed inspection of 1D and 2D NMR data revealed that compounds 1 and 2 share identical planar structures.The relative configurations of 1 and 2 were assigned by key NOESY correlations and coupling constants (Table 1 and Figure 5).The signal intensity of H-15 and H-13 (δ C/H 30.8/1.61 for 1 and 30.6/1.57for 2) increased after the irradiation of H-19, indicating a similar relative configuration of the cyclohexanone moiety in 1 and 2. The Z configuration of the ∆ 9 (10) double bond of 1 and 2 was deduced through a strong NOESY correlation between H-9 and H-10 and relatively small coupling constants (J H-9/H-10 = 10.1 Hz).The relative configuration of C-11 was not deduced.S5).Thus, 1 and 2 are isomers of each other.
The relative configurations of 1 and 2 were assigned by key NOESY correlations and coupling constants (Table 1 and Figure 5).The signal intensity of H-15 and H-13 (δC/H 30.8/1.61 for 1 and 30.6/1.57for 2) increased after the irradiation of H-19, indicating a similar relative configuration of the cyclohexanone moiety in 1 and 2. The Z configuration of the Δ 9 (10) double bond of 1 and 2 was deduced through a strong NOESY correlation between H-9 and H-10 and relatively small coupling constants (JH-9/H-10 = 10.1 Hz).The relative configuration of C-11 was not deduced.The ECD calculation and biosynthetic origin were involved in the assignment of the absolute configuration of 1 and 2. Surprisingly, the experimental ECD curves of 1 and 2 showed almost opposite cotton effects (Figure 6).We propose that the observed differences in the ECD cotton effects are primarily due to the benzopyran moiety, rather than the cyclohexanone group, as shown in a previous study, and the ECD method is suitable for the assignment of C-11 [13].Thus, theoretical ECD calculations were performed using the time-dependent density functional theory (TD-DFT) approach.As shown in Figure 6, the experimental ECD spectrum of 1 displayed a good match with the calculated spectrum of 11S, while the curve of 2 was in accord with that of the 11R.The 14S, 15R, and 19R of 1 and 2 were deduced on the basis of the enzyme-mediated formation of the cyclohexanone group [1,2] and the same relative configuration of this moiety to that of co-isolated compound 9.In fact, the literature survey revealed that all the cyclohexanone groups in ASCs share ommon stereochemistry without exception [8].Thus, the absolute configurations were finally determined to be 9Z, 11S, 14S, 15R, 19R-1 and 9Z, 11R, 14S, 15R, 19R-2, respectively, indicating that compounds 1 and 2 are diastereoisomers.The ECD calculation and biosynthetic origin were involved in the assignment of the absolute configuration of 1 and 2. Surprisingly, the experimental ECD curves of 1 and 2 showed almost opposite cotton effects (Figure 6).We propose that the observed differences in the ECD cotton effects are primarily due to the benzopyran moiety, rather than the cyclohexanone group, as shown in a previous study, and the ECD method is suitable for the assignment of C-11 [13].Thus, theoretical ECD calculations were performed using the time-dependent density functional theory (TD-DFT) approach.As shown in Figure 6, the experimental ECD spectrum of 1 displayed a good match with the calculated spectrum of 11S, while the curve of 2 was in accord with that of the 11R.The 14S, 15R, and 19R of 1 and 2 were deduced on the basis of the enzyme-mediated formation of the cyclohexanone group [1,2] and the same relative configuration of this moiety to that of co-isolated compound 9.In fact, the literature survey revealed that all the cyclohexanone groups in ASCs share ommon stereochemistry without exception [8].Thus, the absolute configurations were finally determined to be 9Z, 11S, 14S, 15R, 19R-1 and 9Z, 11R, 14S, 15R, 19R-2, respectively, indicating that compounds 1 and 2 are diastereoisomers.4).The E configuration of both ∆ 10 and ∆ 14 double bonds was assigned by NOESY correlations between H 2 -9 and Me-23, H-10 and H 2 -12, H 2 -13 and Me-22, and H-14 and H-16, respectively (Figure 5).The ECD calculation was used to address the absolute configuration of C-16; the calculated ECD curve of 3 showed positive Cotton effects at around 200-250 nm and negative Cotton effects at around 280-320 nm, coinciding well with the experimental ECD spectrum and suggesting a 16R-configuration in 3 (Figure 7).).The NMR data of 3 were highly similar to those of the co-isolated chlorocylindrocarpol (6), suggesting that 3 contained an acyclic sesquiterpene moiety (Figure 3) [14].The only difference between 3 and 6 is that the double bond group of C-16 and C-18 was substituted by a hydroxyl group and a non-conjugated ketone moiety in 3. The differences were supported by the COSY correlations of H-16/H-17 and HMBC correlations of Me-20 (δC/H 24.4/1.28)with C-18 (δC 218.1) and Me-22 (δC/H 11.3/1.63)with C-14 (δC/H 128.7/5.5),C-15 (δC 133.1),C-16 (δC/H 77.9/4.52)(Figure 4).The E configuration of both Δ 10 and Δ 14 double bonds was assigned by NOESY correlations between H2-9 and Me-23, H-10 and H2-12, H2-13 and Me-22, and H-14 and H-16, respectively (Figure 5).The ECD calculation was used to address the absolute configuration of C-16; the calculated ECD curve of 3 showed positive Cotton effects at around 200-250 nm and negative Cotton effects at around 280-320 nm, coinciding well with the experimental ECD spectrum and suggesting a 16R-configuration in 3 (Figure 7).   ) and H-18 (δC/H 79.6/4.99).Additionally, the acetyl group was determined to be attached to OH-18 through HMBC correlations between H-1′ (δC/H 21.2/2.07)and H-18 with C-2′ (δC 170.8).Both Δ 10 and Δ 14 double bonds were also assigned as E configuration by NOESY correlations (Figure 5).Attempts to obtain crystals for further analysis were unsuccessful.

Figure 6 .
Figure 6.Experimental and calculated ECD spectra for 1 and 2.

Figure 6 .
Figure 6.Experimental and calculated ECD spectra for 1 and 2.

Figure 7 .
Figure 7. Experimental and calculated ECD spectra of compound 3. Compound 4 was obtained as a yellow oil.Analysis of the HR-ESIMS data showed a characteristic pseudomolecular ion indicative of a monochloroinated compound at m/z 465.2039/467.2012 in a ratio of 3:1 ([M-H 2 O+H] + /[M+2-H 2 O+H] + ) and gave the molecular formula of C 25 H 35 ClO 7 .The high similarity of NMR spectroscopic data of 4 to compound 3 suggested that both structures were closely related.The difference between compound 3 and 4 was the O-acetylation of the ketone moiety at C-18 in 3, which was confirmed by COSY correlations between H-16, H-17 and H-18, and the key HMBC between Me-21 (δ C/H 25.5/1.23)and H-18 (δ C/H 79.6/4.99).Additionally, the acetyl group was determined to be attached to OH-18 through HMBC correlations between H-1 ′ (δ C/H 21.2/2.07)and H-18 with C-2 ′ (δ C 170.8).Both ∆ 10 and ∆ 14 double bonds were also assigned as E configuration by NOESY correlations (Figure5).Attempts to obtain crystals for further analysis were unsuccessful.