Bioactive Indole Diketopiperazine Alkaloids from the Marine Endophytic Fungus Aspergillus sp. YJ191021

Six new prenylated indole diketopiperazine alkaloids, asperthrins A–F (1–6), along with eight known analogues (7–14), were isolated from the marine-derived endophytic fungus Aspergillus sp. YJ191021. Their planar structures and absolute configurations were elucidated by HR-ESI-MS, 1D/2D NMR data, and time-dependent density functional theory (TDDFT)/ECD calculation. The isolated compounds were assayed for their inhibition against three agricultural pathogenic fungi, four fish pathogenic bacteria, and two agricultural pathogenic bacteria. Compound 1 exhibited moderate antifungal and antibacterial activities against Vibrio anguillarum, Xanthomonas oryzae pv. Oryzicola, and Rhizoctonia solani with minimal inhibitory concentration (MIC) values of 8, 12.5, and 25 μg/mL, respectively. Furthermore, 1 displayed notable anti-inflammatory activity with IC50 value of 1.46 ± 0.21 μM in Propionibacterium acnes induced human monocyte cell line (THP-1).


Introduction
Endophytic fungi refer to microorganisms that spend their entire or part of their life cycle in plant tissues, animals, and environments without causing any obvious infection or visible disease to the host [1]. Endophytic fungi are prolific microbial resources for the production ability of many biologically active secondary metabolites, which can help the host to resist pathogenic microorganisms [2]. Various endophytic fungi have drawn substantial attention due to their potential to produce chemically diverse and biologically active secondary metabolites with anti-cancer, anti-microbial, anti-viral, and insecticidal activities [3][4][5][6]. In our continuous searching for novel bioactive secondary metabolites from marine endophytic fungi [7][8][9], the Aspergillus sp. YJ191021 attracted our attention, not only for the characteristic indole diketopiperazine ultraviolet (UV) absorptions of the crude extracts, but also for their potent antimicrobial activities against agricultural pathogenic fungi.
Diketopiperazine alkaloids are valued not only for their properties and functions in fungal self-biology, but also for niche establishment to defend abiotic and biotic stress in nature. They are cyclodipeptides formed by condensation of two amino acids under the control of NRPS genes [10], especially those isolated from the genera Aspergillus and Penicillium [11]. Among them, those derived from tryptophan and proline are the most popular types in the current study, especially in structural diversity, chemical synthesis, and pharmacological activity [12][13][14]. Besides, the substitution of the isopentenyl group enriches the variability of their structures. Prenylated indole alkaloids have been reported to show a wide array of biological activities including antimicrobial, insecticidal, and cytotoxic activities [14,15]. The fascinating structural and biological properties of prenylated indole alkaloids make it possible for them to be developed into our armor and weaponry: Natural agrochemicals and drugs.

Results and Discussion
The ROESY spectrum ( Figure 3 and Figure S6) exhibited correlations between 19-NH (δ H 8.82) and H-23 (δ H 1.22), between H-21 (δ H 2.31) and H-24 (δ H 1.55), supporting that H-21 and H-24 were co-facial and assigned as α-oriented whereas H-23 is β-oriented, respectively. Additionally, the absence of a cross peak between H-21 (δ H 2.31) and 19-NH (δ H 8.82) indicated that the relative configuration between N13-C17 and C21-22 was anti [21]. Williams reported that the Cotton effect at λ = 200-250 nm arising from an n-π* transition of the diketopiperazine moiety is diagnostic of the bicyclo[2.2.2]diazaoctane ring system [21,22]. The negative Cotton effect at 225 nm in ECD spectrum ( Figure 4A and Figure S8), which was opposite to that of 6-epi-avrainvillamide [18], empirically indicated that the absolute configurations of C-11 and C-17 in 1 were 11R, and 17R. Combined with the analysis of the ROESY spectrum, the absolute configuration of C-21 was assigned as 21S.
To further verify the aforementioned absolute configuration deduction of 1, the calculated ECD spectrum was conducted. The absolute configurations of 11R, 17R, and 21S were determined for the well match between the calculated and the experimental ECD spectra ( Figure 4A). The ROESY spectrum (Figures 3 and S6) exhibited correlations between 19-NH (δH 8.82) and H-23 (δH 1.22), between H-21 (δH 2.31) and H-24 (δH 1.55), supporting that H-21 and H-24 were co-facial and assigned as α-oriented whereas H-23 is β-oriented, respectively. Additionally, the absence of a cross peak between H-21 (δH 2.31) and 19-NH (δH 8.82) indicated that the relative configuration between N13-C17 and C21-22 was anti [21]. Williams reported that the Cotton effect at λ = 200-250 nm arising from an n-π* transition of the diketopiperazine moiety is diagnostic of the bicyclo[2.2.2]diazaoctane ring system [21,22]. The negative Cotton effect at 225 nm in ECD spectrum ( Figures 4A and S8), which was opposite to that of 6-epi-avrainvillamide [18], empirically indicated that the absolute configurations of C-11 and C-17 in 1 were 11R, and 17R. Combined with the analysis of the ROESY spectrum, the absolute configuration of C-21 was assigned as 21S. To further verify the aforementioned absolute configuration deduction of 1, the calculated ECD spectrum was conducted. The absolute configurations of 11R, 17R, and 21S were determined for the well match between the calculated and the experimental ECD spectra ( Figure 4A).
Asperthrin D (4) was isolated as white powders. The molecular formula was determined as C27H31N3O6 by the (+)-HRESIMS data from the [M + Na] + ion at m/z 516.2106 Asperthrin B (2) was obtained as white powders. The molecular formula C 26 H 29 N 3 O 5 , which was determined by the [M + Na] + ion at m/z 486.2002 (calcd. for C 26 H 29 N 3 O 5 Na, 486.1999) from the HR-ESI-MS and 13 C NMR data was 18 amu higher than the molecular mass of 1, implying the presence of an additional hydroxy group in its structure. A careful comparison of the 13 C NMR data of 2 with those of 1 ( Table 2) showed significant upfield shifts of C-3 (δ C 75.8) and C-10 (δ C 36.0), indicating a saturation of the double bond between C-3 and C-10. The HMBC correlations from H-4 (δ H 7.32) to C-3 (δ C 75.8) (Figure 2 and Figure S14) confirmed that the hydroxyl group was attached to C-3. The ROESY correlations (Figure 3 and Figure S15) from 19-NH (δ H 7.52) to H-23 (δ H 1.34) and H-10a (δ H 2.64) to 3-OH (δ H 6.39)/19-NH (δ H 7.52)/H-23 (δ H 1.34) indicated these protons were co-facial and β-oriented. Accordingly, the ROESY correlations between H-21 (δ H 2.13) and H-24 (δ H 1.54) revealed that H-21 and H-24 were α-oriented. The absolute configurations of C-3, C-11, C-17, and C-21 in 2 were assigned as 3R, 11R, 17R, and 21S based on the negative Cotton effect at 225 nm in ECD spectra ( Figure 4B and Figure S17) and calculated ECD spectra ( Figure 4B). 5.75)/H-23 (δH 0.52) indicated that these protons located on the β-orientation of the cyclopentane ring. Besides, the ROESY correlations from H-21 (δH 2.68) to H-4 (δH 6.89) indicated that the cyclopentane ring was orthogonal to the plane of the indoxyl ring. Based on the positive Cotton effect at 225 nm and the calculated ECD spectra results ( Figures 4A  and S44), the absolute configurations of C-3, C-10, C-11, C-17, and C-21 in 5 were assigned as 3R, 10S, 11R, 17S, 21R.   13 C NMR data, the molecular formula was determined as C 27 H 31 N 3 O 6 , which was 30 amu more than 2. Comparation of the 1 H NMR spectrum ( Figure S19) of 3 with that of 2 indicated that there was a methoxy group in 3. The HMBC correlation from OMe-10 (δ H 3.03) to C-10 (δ C 76.9) (Figure 2 and Figure S23) suggested that the methoxy group was attached to C-10. The ROESY correlations (Figure 3 and Figure S24) between 19-NH (δ H 7.74) and OMe-10 (δ H 3.03)/H-23 (δ H 1.31) indicated that these protons were co-facial and assigned as β-oriented. The ROESY correlations between 3-OH (δ H 6.31) and H-10 (δ H 4.72)/H-21 (δ H 3.08), and between H-21 (δ H 3.08) to H-24 (δ H 1.32), indicated that they were in the α-orientation. Based on the negative Cotton effect at 225 nm in ECD spectra ( Figure 4C and Figure S26) and the well match result between experimental and calculated ECD spectra ( Figure 4C), the absolute configurations of C-3, C-10, C-11, C-17, and C-21 in 3 were assigned as 3S, 10S, 11S, 17R, and 21S.
Asperthrin D (4) was isolated as white powders. The molecular formula was determined as C 27 H 31 N 3 O 6 by the (+)-HRESIMS data from the [M + Na] + ion at m/z 516.2106 (calcd. for C 27 H 31 N 3 O 6 Na, 516.2105) as same as that of 3. Detailed analyses of the 1D NMR and 2D NMR spectra (Figures S28-S32) indicated that the planar structure of 4 was the same as that of 3. However, the chemical shifts of H-10 (δ H 4.12) and OMe-10 (δ H 3.31) in 4 differed from those of 3, implying that the configuration of C-10 was opposite to that of 3. The relative configuration was assigned by the ROESY correlations from 19 (Figure 3 and Figure S33). The absolute configurations of C-3, C-10, C-11 C-17, and C-21 in 4 were assigned as 3S, 10R, 11R, 17S, 21S based on the positive Cotton effect at 225 nm and the calculated ECD spectra results ( Figure 4C and Figure S35).   Figure S41) indicated the presence of an indoxyl core with a spiro-quaternary center at C-3. Detailed analyses of 1D NMR and 2D NMR spectra (Figures S37-S41) indicated that the planar structure of 5 was the same as that of 10-O-acetylsclerotiamide [23]. The ROESY correlations (Figure 3 and Figure S42) from 19-NH (δ H 8.54) to H-24 (δ H 1.12) indicated that NH-19 and H-24 were located on the α-face of the cyclopentane ring. The ROESY correlations from H-21 (δ H 2.68) to H-10 (δ H 5.75)/H-23 (δ H 0.52) indicated that these protons located on the β-orientation of the cyclopentane ring. Besides, the ROESY correlations from H-21 (δ H 2.68) to H-4 (δ H 6.89) indicated that the cyclopentane ring was orthogonal to the plane of the indoxyl ring. Based on the positive Cotton effect at 225 nm and the calculated ECD spectra results ( Figure 4A and Figure S44), the absolute configurations of C-3, C-10, C-11, C-17, and C-21 in 5 were assigned as 3R, 10S, 11R, 17S, 21R.

Fungal Material
The fungal strain A. sp. YJ191021 was isolated from a soil sample, which was collected from the intertidal zone of Zhoushan, Zhejiang, China, in June 2018. The fungal strain was identified according to their morphological characteristics and by sequencing the fungal ITS region in rDNA. The fungal strain is stored in State Key Laboratory of Bioreactor Engineering laboratory of Shanghai at −80 • C.

Fermentation, Extraction, and Isolation
The fungus was incubated on potato dextrose agar (PDA) medium at 28 • C for 3 days. Then the grown strain was inoculated to a 250 mL Erlenmeyer flask containing 50 mL of potato dextrose broth (PDB). After 2 days of fermentation, the seed cultures were added to Erlenmeyer flasks (100 × 1000 mL), each containing 100 g of dry rice and 120 mL of distilled water, which was previously sterilized at 121 • C for 30 min. All flasks were incubated at room temperature for 30 days. After incubation, whole fermented rice medium was extracted three times using ethyl acetate (EtOAc), and then solvents were concentrated under reduced pressure to give a crude extract (193.4 g). Next, the crude extract was subjected to a macroporous resin column eluting by a gradient EtOH-H 2 O (from 30%, 50%, 70% to 100% EtOH). The 50% fraction (33.6 g) was then separated on a silica gel column eluting with a stepwise gradient of CH 2 Cl 2 -MeOH (from 25:1 to 5:1) to yield five subfractions (A-E). Fraction D (4.3 g) was further purified by an ODS column (MeCN-H

Antimicrobial Assays
Minimum Inhibitory Concentration (MIC) assays were used to assess antimicrobial activities of all isolated compounds against two agricultural pathogenic bacteria (Xanthomonas oryzae pv. oryzae and Xanthomonas oryzae pv. oryzicola) and three agricultural fungi (Colletotrichum gloeosporioides penz, Fusarium oxysporum and Rhizoctonia solani). Furthermore, Compound 1 was tested for antibacterial activities against four fish pathogens, Edwardsiella tarda, Vibrio anguillarum, Aeromonas hydrophilia, and Vibrio parahaemolyticus. Chloromycetin was used as a positive antibacterial control and ketoconazole was used as a positive antifungal control. The experimental procedure is detailed in the Supporting Information (SI). All the experiments were performed in three independent replicates.

Anti-Inflammatory Assays
The human monocyte cell line THP-1 (Cell Bank of China Science Academy, Shanghai, China) and Propionibacterium acnes (ATCC6919, Xiangfu biotech, Shanghai, China) were used in anti-inflammatory experiments. The P. acnes in logarithmic growth phase was used to induce inflammation in THP-1 cells. MTT method was carried out for tested compounds to determine their safe concentration to THP-1 cells. Besides, antimicrobial assays were performed to exclude false anti-inflammatory activity of these compounds raised from their inhibition to P. acnes. The inhibitory activity of the test compounds on the secretion of inflammatory factor 1L-1β by THP-1 cells was assayed by ELISA experiment [24,25]. The experimental procedure is detailed in the SI. All the experiments were performed in three independent replicates.
Author Contributions: J.Y., Y.D., and Z.W. performed the isolation, purification, and identification of all compounds. L.G. and Y.J. tested the anti-inflammatory activities, and supervised the laboratory work. M.G. and X.X. edited the manuscript. F.A. supervised the laboratory work, designed the experiments, and edited the manuscript. All authors have read and agreed to the published version of the manuscript.