Secondary Metabolites from Aspergillus sparsus NBERC_28952 and Their Herbicidal Activities

Fungi have been used in the production of a wide range of biologically active metabolites, including potent herbicides. In the search for pesticides of natural origin, Aspergillus sparsus NBERC_28952, a fungal strain with herbicidal activity, was obtained. Chemical study of secondary metabolites from NBERC_28952 resulted in the isolation of three new asperugin analogues, named Aspersparin A–C (2–4), and a new azaphilone derivative, named Aspersparin D (5), together with two known compounds, Asperugin B (1) and sydonic acid (6). The structures of these compounds were elucidated based on extensive spectroscopic data and single-crystal X-ray diffraction analysis. All of the isolated compounds were evaluated for their herbicidal activities on seedlings of Echinochloa crusgalli and Amaranthus retroflexus through Petri dish bioassays. Among them, compounds 5 and 6 exhibited moderate inhibitory activities against the growth of the roots and shoots of E. crusgalli seedlings in a dose-dependent manner, while 6 showed obvious inhibitory effect on seedlings of A. retroflexus, with an inhibitory rate of 78.34% at a concentration of 200 μg/mL. These herbicidal metabolites represent a new source of compounds to control weeds.


Introduction
Weeds can seriously threaten crop yields and cause huge economic losses [1]. A lot of measures, including cultivation, mechanical operations, and the application of chemicals, have been implemented to control weeds. However, the excessive use of chemical herbicides leads to residue seepage into the environment and the increased resistance of weeds to such compounds [2]. Given the present concerns about the negative impacts of chemical herbicides on human health and the environment, it is necessary to develop safer compounds to ensure the sustainability of crop production.
Natural products have played an important role in the development of pesticides for crop protection [3]. An analysis showed that 41.8% of the pesticides listed in the registry of the Environmental Protection Agency were based on active ingredients developed with natural products [4]. Natural products might be a source of new herbicidal entities with potentially new modes of actions [5]. Fungi are considered as one of the richest sources of natural products among living organisms [6]. The phytotoxins produced by fungi are often suitable for the pathogenesis or infection of weeds [7]. Fungi of Aspergillus spp. have been shown to be excellent sources of new natural chemicals [8][9][10][11][12], some of which have shown promising herbicidal activities. For example, Asperalacid D, a new natural sesquiterpenoid from Aspergillus alabamensis, showed higher plant growth inhibitory activity on wheat some of which have shown promising herbicidal activities. For example, Asperalacid D, a new natural sesquiterpenoid from Aspergillus alabamensis, showed higher plant growth inhibitory activity on wheat root and shoot elongation than terbutryn [13]. Additionally, 8-methoxycichorine, 8-epi-methoxycichorine, and N-(4′-carboxybutyl) cichorine, three novel cichorine analogues with an isoindolinone skeleton, obtained from A. nidulans, exhibited superior phytotoxicity to cichorine on the leaves of Zea mays and Medicago polymorpha [14]. Dihydrosterigmatocystin, isolated from an alga-derived fungus, Aspergillus versicolor, caused leaf necrosis and plant wilting in Amaranthus retroflexus, with a MIC of 24.5 μM, i.e., almost four-fold stronger than that of glyphosate [15]. This indicates the potential for discovering novel herbicidal compounds from the secondary metabolites of fungi in the genus Aspergillus. During the course of screening for microbial secondary metabolites possessing bioactivities, we isolated some active compounds with new structures [16][17][18]. In an ongoing effort to discover bioactive natural products obtained from microbes, our attention was drawn to the fungus A. sparsus NBERC_28952 because of its potent herbicidal activity. In this study, we describe the isolation, structure, and herbicidal activities of the compounds obtained from NBERC_28952. To the best of our knowledge, this is the first report on this topic.
Compound 5 was isolated as yellow crystalline powder. Its molecular formula was determined to be C 22 Table 3, Figure S31) and HSQC (see Figure S32 Figure S34) of H-1 to C-3 (δ C 160.3), 4a (δ C 146.0), 8a (δ C 111.1), H-4 to C-3, 5 (δ C 103.7), 8a, and H-5 to C-4 (δ C 108.1), C-7 (δ C 87.7), and C-8a suggested the presence of an azaphilone skeleton. Moreover, the presence of a ketone aliphatic side chain was supported by the HSQC, COSY ( Figure S33) and HMBC spectra. A comparison of the 1 H-and 13 C-NMR data of 5 with those of didehydrochermesinone B [21] showed that the former contained four more CH 2 in the fatty chain. Accordingly, the planar structure of 5 was established. The absolute configuration of 5 was determined by single-crystal X-ray diffraction analysis. An Olex2 plot is shown in Figure 4, illustrating the absolute configuration of 5 to be 7R, 2 S with Flack parameter -0.07 (10) (CCDC 2219385). The structure of 5 was finally determined and the compound was named Aspersparin D.

Evaluation of Herbicidal Activities
The herbicidal activities of the isolated compounds toward seedlings of E. crusgalli and A. retroflexus were assessed at 200 µg/mL using 2,4-dichlorophenoxy acetic acid (2,4-D) as the positive control through Petri dish bioassays (Table 4). Of the tested compounds, Aspersparin D (5) and sydonic acid (6) exhibited moderate inhibitory activities against the growth of the roots and shoots of E. crusgalli seedlings (with inhibitory rates ranging from 40% to 60%). Compound 6 showed a lower inhibition rate to E. crusgalli seedlings than 5, but a much higher one to the radicle and germ of A. retroflexus (78.34%), similar to that of 2,4-D (80.70%). a All the compounds were tested at a concentration of 200 µg/mL. b Values are presented as a percentage of the mean compared to the control (mean ± SD). c 2,4-dichlorophenoxy acetic acid, used as a positive control.
Compounds 5 and 6 exhibited much stronger inhibitory activities against the growth of E. crusgalli and A. retroflexus seedlings than the other tested compounds. Therefore, further investigation of herbicidal activities of 5 and 6 was conducted. The inhibitory effects of gradient concentrations (12.5, 25, 50, 100, 200 µg/mL) of these compounds on the growth of seedlings of the two weed varieties were assayed. As shown in Figure 5, compounds 5 and 6 inhibited the growth of the seedlings of the two weeds in a dose-dependent manner. Furthermore, in the concentration range of 50-200 µg/mL, 5 and 6 displayed moderate inhibitory activities against the growth of shoots of E. crusgalli seedlings. In addition, in contrast to the positive control 2,4-D, the inhibitory effects of 5 and 6 on shoots of E. crusgalli seedlings were stronger than on the roots. Both 5 and 6 exhibited good inhibition on A. retroflexus seedlings, i.e., higher than 50% inhibition at a concentration of 50 µg/mL. Interestingly, when the concentration was below 50 µg/mL, the inhibitory effect on A. retroflexus of 5 was much lower than that of 6, but it increased more rapidly with an increase of concentration, achieving over 60% inhibition at 100 µg/mL and over 75% inhibition at 200 µg/mL. a All the compounds were tested at a concentration of 200 μg/mL. b Values are presented as a percentage of the mean compared to the control (mean ± SD). c 2,4-dichlorophenoxy acetic acid, used as a positive control.

Discussion
Compounds 1-4 were asperugin analogues which showed weak inhibitory activities against the growth of E. crusgalli and A. retroflexus seedlings at the concentration of 200 µg/mL. Compounds 1-3 were double-bond isomers with the same structure in the phenyl part, while compound 4 had a -CH 2 OCOCH 3 located at the phenyl ring. The bioassay results showed that 4 exhibited higher inhibition on the tested weeds than compounds 1-3. Therefore, the substituents on the benzene ring might play an important role in the herbicidal activities of asperugin analogues. This is the first report on the herbicidal activities of asperugin analogues. These compounds may prove to be promising in the research and development of natural or plant-derived herbicides.
Azaphilones are a large family of metabolites derived from fungi with variable structures, exhibiting a wide range of biological activities [22]; however, at present, there are relatively few reports about their herbicidal activities [23]. As reported in our previous study [24], Chaetomugilin O, an azaphilone with tetrahydrofuranone, exhibited the most potent inhibition among the tested azaphilones of seedling growth of several weeds. In our study, Asperaprin D (5) was identified as a new azaphilone derivative with a tetrahydrofuranone group. The compound showed moderate inhibitory activity against E. crusgalli and A. retroflexus seedlings. Based on a comparison of the structures of Chaetomugilin O and Asperaprin D (5), we speculate that the presence of a tetrahydrofuranone group may be an important factor for the growth-suppression activity of azaphilones, with the side chains connected to C-3 and C-7 also influencing the inhibitory effects. In any case, further studies on the structure-activity relationship and structural modifications are needed to verify this hypothesis. Previous studies have examined the antibacterial activity [25] and cytotoxic activity [26] of sydonic acid (6), a bisabolane-type sesquiterpene produced by several fungi. In the present study, sydonic acid (6) was found to have herbicidal activities, showing moderate inhibitory effects on E. crusgalli seedlings and obvious inhibitory effects on A. retroflexus seedlings. This is the first time that compounds 5 and 6 have been reported as herbicidal fungal metabolites. These findings could expand the application of these compounds in the agricultural domain and broaden our knowledge on the possible structures of natural herbicides. However, whether the compounds are useful as mycoherbicides remains to be determined by further study via pot experiments in a greenhouse and field experiments.

Cultivation, Extraction, and Isolation
A stock of A. sparsus strain NBERC_28952, previously stored at −86 • C, was streaked on a potato-dextrose agar plate and incubated at 25 • C until good growth was observed. Culture discs with an internal diameter of 5 were made with a sterile stainless-steel puncher mm. Five culture discs were inoculated into separate 500 mL flasks containing 100 mL of seed medium each. The seed culture was shaken at 120 rpm at 28 • C. After 96 h of cultivation, seed cultures (10%) were transferred to 500 mL Erlenmeyer flasks containing 100 mL medium with the following composition: 6.25 g/L malt extract, 6.25 g/L maltose, 1.0 g/L yeast extract, 0.625 g/L soybean peptone, 1.25 g/L KH 2 PO 4 , and 1.25 g/L MgSO 4 . The inoculated flasks were incubated at 28 • C for 96 h on a rotary shaker (120 rpm).
The fermented material (10.0 L) was extracted using ethyl acetate (3 × 10.0 L) following three 30 min periods of stirring [16]. The organic solvent was filtrated and then concentrated in vacuo to obtain a crude extract (3.6 g).

Statistical Analysis
Data were expressed as means ± standard deviation of mean (SD) (n = 3). Statistical analysis was performed using Microsoft excel (Microsoft Corp, Redmond, WA, USA) and GraphPad Prism ver. 5 (GraphPad Software, La Jolla, CA, USA). A statistically significant difference was considered when p < 0.05.

Conclusions
Three new Asperugin analogues (2-4) and a new azaphilone derivative (5), together with Asperugin B (1) and sydonic acid (6), were obtained from an EtOAc extract of A. sparsus NBERC_28952. The structures were elucidated based on the interpretation of extensive spectroscopic data and single-crystal X-ray diffraction analysis. The herbicidal activities of compounds 1-6 were evaluated. Aspersparin D (5) and sydonic acid (6) exhibited inhibitory activities against the growth of seedlings of E. crusgalli and A. retroflexus in a dose-dependent manner, while compound 6 showed an obvious inhibitory effect against seedlings of A. retroflexus (78.34%) at 200 µg/mL, similar to 2,4-D (80.70%). This is the first report of a chemical investigation of A. sparsus and on the herbicidal activities of the isolated compounds, which, with further study, might represent new herbicide candidates.