Introduction

Alzheimer’s disease (AD), which is commonly known as dementia, is a neurodegenerative disorder characterized by memory loss and other cognitive impairments (Jalili-Baleh et al. 2018; Friker et al. 2020). Reportedly, over 46 million people live with dementia worldwide, and the evaluated number is expected to increase to 131.5 million by 2050 (Prince et al. 2015). The rapid growth of AD has led many medicinal chemists to develop drug-discovery investigations in this field (Prince et al. 2015; Mohammadi-Khanaposhtani et al. 2015; Friker et al. 2020).

The etiology of AD is not well understood. Thus, different pathogenesis hypotheses regarding AD, including cholinergic, amyloid cascade, oxidative stress, and tau protein, have been proposed. Among them, the cholinergic hypothesis is widely accepted (Goedert and Spillantini 2006; Sonmez et al. 2017; Oh et al. 2019). According to this hypothesis, the most effective therapeutic approach for treating AD is the restoration of the acetylcholine (ACh) levels in the brain by inhibiting acetylcholinesterase (AChE). AChE inhibitors (AChEIs) can effectively improve neurotransmitter activity levels and duration by inhibiting the hydrolysis of ACh (Sonmez et al. 2017; Oh et al. 2019). Currently, the mainstay drugs used for the clinical management of AD remain AChEIs. Four AChEIs, namely, tacrine, donepezil, galantamine, and rivastigmine, have been approved by European and US agencies (Sonmez et al. 2017; Oh et al. 2019). Huperzine A (Hup A), a powerful and selective lycopodium alkaloid AChEI firstly isolated from Qian Ceng Ta (i.e., Huperzia serrata) in traditional Chinese medicine, was also approved in the 1990s in China (Fig. 1) (Ma et al. 2007). Although the existing AChEI drugs are essential for palliating AD, their clinical efficacy is limited because of their poor selectivity, low bioavailability, and adverse side effects on the peripheral nervous system and liver. Therefore, great efforts have been dedicated for investigating better and novel AChEIs from natural sources (Zaki et al. 2020).

Fig. 1
figure 1

Chemical structures of the main acetylcholinesterase inhibitors used in clinical management

Full size image

Fungi are important organisms in the production of bioactive secondary metabolites. Currently, many of the drugs in the market that possess various activities, such as antitumor, immunosuppressant, antibiotic, hypocholesterolemic, antifungal, antiparasitic, anti-inflammatory, and enzyme-inhibiting activities, were obtained from fungal metabolism (Kingston 2011; Teles and Takahashi 2013). Endophytic fungi are microorganisms that reside in the internal tissues of living plants or animals without causing apparent disease (Gupta et al. 2019). Endophytes, an eclectic group of microbes that can chemically bridge the gap between plants and microbes, have attracted the most attention because of their relatively high metabolic versatility (Zaki et al. 2020). Thus, endophytes have been demonstrated to be a rich source of bioactive metabolites with diverse structural features, and a large number of compounds with novel structures and various bioactivities are continuously being isolated from them (Gupta et al. 2019; Zaki et al. 2020). Notably, some endophytic fungi have produced identical or similar chemical compounds to those produced by their host. For example, the widely prescribed anticancer drug paclitaxel (Taxol), one of the most amazing natural products initially isolated from the Pacific yew tree Taxus brevifolia, was later found in the endophytic fungus Taxomyces andreanae and other fungal genera (Liu et al. 2016). Hup A was also found in the endophytic fungus Shiraia sp. Slf14, as well as in Cladosporium cladosporioides LF70, Penicillium sp. LDL4.4, Trichoderma harzianum L44, and other endophytic fungi (Ellman et al. 1961; Zhang et al. 2011; Cao et al. 2021).

The endophytic fungi from H. serrata are a huge untapped source of natural products including AChEI. Thus far, there are only few reports on AChEI form endophytic fungi of H. serrata (Cao et al. 2021). In our previous screening studies, a total number of 22 endophytic fungi strains, including strain FL15, showed high AChE inhibitory activity (≥ 50%) (Wang et al. 2016b). In the present study, as part of an ongoing search for Hup A-producing endophytes and new natural AChEI, the secondary metabolites from the ethyl acetate extract of fungal strain FL15 were isolated, and seven natural metabolites were purified and identified. AChE inhibition results showed that the three asterric acid derivatives exhibited highly selective and moderate AChE inhibitory activities. Then, the molecular docking analysis was carried out to reveal the binding patterns between the three asterric acid derivatives and AChE proteins. This study not only provided new precursors of the anti-AChE drug, but also contributed to the application of endophytic fungi as compounds producers in biopharmaceutical industry.

Materials and methods

General

The strain FL15 used in this study was isolated from the leaves of H. serrata, which were collected from natural populations at Lushan Botanical Garden in Jiangxi Province, Central China. This strain was subsequently deposited in the China Center for Type Culture Collection (CCTCC NO: 2019832), Wuhan, China.

1H and 13C data were acquired on a Bruker AV400 spectrometer at 400 and 100 MHz, respectively, using CDCl3, acetone-d6, and CD3OD as solvents. Chemical shifts were given in parts per million (δ) using tetramethylsilane (TMS) as an internal standard. J values were given in Hertz. Abbreviations for the 1H NMR data quoted are as follows: s, singlet; d, doublet; t, triplet; q, quartet; m, multiplet; bs, broad singlet. ESI–MS data were recorded on a Waters Xevo G2 quadrupole time-of-flight/time-of-flight (QTof/Tof) mass spectrometer (Milford, Massachusetts, USA). High-performance liquid chromatography (HPLC) was performed on a Waters W2996 chromatograph equipped with a 1525 pump using a C18 column (19 × 150 mm, 5 μm; YMC Co., Ltd.).

Chemicals and enzymes

All solvents and reagents were obtained from commercial sources, unless stated otherwise. Electric eel acetylcholinesterase (AChE, Type-VI-S, EC 3.1.1.7, 425.84 U/mg, Sigma) and equine serum butyrylcholinesterase (BuChE, E.C.3.1.1.8) were purchased from Sigma (Steinheim, Germany). Acetylthiocholine iodide (AChI), butyrylthiocholine iodide (BChI), and dithiobis nitrobenzoic acid (DTNB) were purchased from Sigma-Aldrich (Taufkirchen, Germany). The solvents used for chromatography were of HPLC grade, whereas the solvents used for extraction were of American Chemical Society grade. Silica gel (200–300 mesh, Qingdao Marine Chemical Group Co., Ltd. Qingdao, China) and Sephadex LH-20 (25–100 μm; Amersham Biosciences) were used for column chromatography. Thin-layer chromatography (TLC) was performed with glass-precoated silica gel 60 plates (0.25 mm; Merck, Darmstadt, Germany). Other chemicals were from China Medicine Shanghai Chemical Reagent Co., Ltd.

Identification of the endophytic fungus FL15

The endophytic fungus FL15 was identified by morphological observation combined with the determination of the sequences of the rDNA internal transcribed spacer (ITS), 18S rDNA, and 28S rDNA (Lai et al. 2014).

A small number of mycelia were selected from the preserved brilliant medium, and the strains were inoculated into potato dextrose agar (PDA), yeast extract sucrose agar (YES), and Czapek yeast extract agar (CYA) plate medium using the point-planting method, respectively. The strains were cultured in a 28 °C constant temperature incubator for 10–20 days, of culture, and the changes in, and characteristics of, colony morphology were regularly observed: the shape, size, texture, color, edge characteristics, and other conditions of the colony were recorded (Lai et al. 2014). After sampling, the morphology of the mycelium was observed under an optical microscope (BA300, Motic, China) after alkaline methylene blue staining and a scanning electron microscope (SEM, QUANTA-200F, FEI, The Netherlands).

The strains were inoculated into a liquid medium and cultured at 28 °C and 150 rpm for 14 days. The mycelia were filtered for molecular biological identification. Genomic DNA was extracted using the improved CTAB method (Zhang et al. 1996). The ITS, 18S, and 28S sequences of the FL15 strain were amplified. The amplified sequences were detected via 1% agar electrophoresis gel, and the gel plate was stained with ethidium bromide. After detection of the bands, the PCR products were sent to Sangon Biotech (Shanghai) Co., Ltd. for sequencing, to obtain the relevant gene sequence fragments. The fragments were uploaded to the GenBank database of the National Center for Biotechnology Information (NCBI) for Basic Local Alignment Search Tool (BLAST) comparison, and the application number was obtained. Moreover, a phylogenetic analysis was performed using the MEGA 7.0.14 ClustalX software, and an evolutionary tree was constructed to determine the classification status of the strain (Vig et al. 2021).

Fermentation and extraction of mycelia

The fungal strain FL15 was cultured on slants of PDA at 28 °C for 7 days. The mycelia from the PDA plate were harvested and grown in 150 mL of PDB medium for 5 days at 28 °C and 150 rpm. Aliquots (5 mL) of this seed culture were inoculated into 500 mL Erlenmeyer flasks that contained 100 mL of medium. The samples were incubated at 28 °C and 150 rpm for 14 days (Lai et al. 2014). The mycelium was separated from the culture broth and dried at 40 °C. Finally, the dried mycelium (980 g) was obtained. The mycelia were ultrasonically broken and extracted exhaustively with 85% alcohol (3 × 1 L) at room temperature, filtered under vacuum, and transferred to clean bottles. Ethanol was removed from the solution using a rotary evaporator at 40 °C. The water phase was extracted with ethyl acetate (EtOAc). The organic solvent was evaporated to dryness under reduced pressure, to give 12.8 g of crude extract.

Isolation and purification of compounds

The separated crude EtOAc fraction was first subjected to column chromatography (CC) by eluting the silica gel with a gradient of petroleum ether/EtOAc from 9:1 to 0:10 (v/v). Nine fractions (Fs. A to Fs. I) were obtained based on TLC. Fraction Fs. B (134 mg) dissolved in CHCl3 was chromatographed over a silica gel column with a gradient of petroleum ether/EtOAc from 9:1 to 1:1 (v/v), to afford compound 2 (24 mg). Fraction Fs. C (3.6 g) with CH3OH was heated to 60 °C and stirred until completely dissolved, and then placed in a ventilation cabinet for evaporation drying. After stewing for 24 h, crystals were formed and washed with methanol three times, to obtain compound 7 (231 mg). The residual liquid was subjected to a Sephadex LH-20 column with an isocratic elution of 1:1 CH3OH/CH2Cl2 (v/v), to obtain compound 3 (165 mg) and Fraction Fs C1. In turn, Fraction Fs C1 was dissolved in acetone and separated by silica gel CC with a gradient of petroleum ether/EtOAc from 9:1 to 1:1 (v/v), to afford compounds 1 (18 mg) and 4 (14 mg). Fraction Fs. D (274 mg) was dissolved in CHCl3 and the solution was separated by silica gel CC using a petroleum ether–EtOAc mixture (20:1 to 1:1), to give compound 5 (16 mg) and subfraction Fs. D1. Fraction Fs. D1 was dissolved in CH3OH–CHCl3 (1:1) and purified in a Sephadex LH-20 column with an isocratic elution of CH3OH–CH2Cl2 (1:1 v/v), to obtain compound 6 (21 mg). The chemical characteristics of the compounds isolated were as follows.

Acetylcholinesterase/butyrylcholinesterase inhibition activity in vitro assay

The determination of the in vitro AChE inhibition activity of the endophytic fungal extracts and compounds 1–7 was conducted according to the method of Ellman’s spectrophotometry (Ellman et al. 1961; Devidas et al. 2021). Known AChEIs, i.e., rivastigmine and huperzine A, were used as the positive controls. The assay was carried out in the 96-well microtiter plates. Briefly, a preincubation solution of 250 μL of phosphate buffer (200 mM, pH 7.7) that contained 15 μL of purified compounds/rivastigmine/HupA, 80 μL of DTNB [3.96 mg of DTNB and 1.5 mg of sodium bicarbonate dissolved in 10 mL of phosphate buffer (pH 7.7)], and 10 μL of AChE/BuChE was prepared. The mixture was incubated for 5 min at 25 °C. After preincubation, 15 μL of the substrate AChI/BChI (10.85 mg in 5 mL of phosphate buffer) was added and incubated again for 5 min. The color developed was measured in a microwell plate reader at 412 nm (Bio-Rad, Hercules, CA). Percent inhibition was calculated through the following formula: (control absorbance − sample absorbance)/control absorbance × 100.

Kinetic study of acetylcholinesterase inhibition

According to Ellman’s method, a kinetic analysis of AChE was performed (Ellman et al. 1961; Sonmez et al. 2017). The type of inhibition was deduced by determining Km and Vmax using Lineweaver–Burk reciprocal plots by plotting 1/V against 1/S at varying concentrations of the acetylthiocholine substrate (0.01–0.04 mM). The inhibitory constant, Ki, was calculated by secondary plots obtained by plotting 1/V versus different inhibitor concentrations.

Docking study

The preparation of ligand file includes three aspects: (i) draw the small molecular structure of ligand by drawing software; (ii) minimize the energies of small ligand molecules drawn; (iii) initialization of small molecule docking of ligand. In the first step, the two-dimensional structural formulae of asterric acid, methyl asterrate, and ethyl asterrate were drawn using ChemBioDraw Ultra 14.0 drawing software, and the files were saved in “.mol” format. In the second step, open the “.mol” file with ChemBio3D Ultra 14.0, minimize the energy using MMFF94 position in the software, and save the minimized molecule as a “.pdb” file for the next step. The third step is to open the “.pdb” format file processed in the previous step with AutoDockTools software, and process the small ligand molecules, including adding Gasteiger charge, detecting the root of ligand, setting the number of rotating structures, etc. All parameters here are default values. The processed ligand molecule is saved as a coordinate file in “.pdbqt” format for molecular docking backup. The “.pdbqt” format file is a file containing atomic coordinates, AutoDock atomic type, charge, and torsion bond information (Pan et al. 2019).

The AutoDock v.4 software was used in the docking simulations. The crystallographic structures of AChE (PDB: 1F8U) (Kryger et al. 2000) and its ligands were processed with AutoDock Tools (version 1.5.6, Sep_17_14) to delete water, add hydrogens, compute Gasteiger charges, and select rotatable side-chain bonds. Affinity (grid) maps of 60 × 100 × 60 points with a grid spacing of 0.375 Å were generated using the help of the program AutoGrid v.4 program included in the AutoDock 4 distribution (Singh et al. 2020; Devidas et al. 2021). AutoDock parameter settings and distance dielectric functions were used in the calculation of van der Waals and electrostatic terms, respectively (Kumar et al. 2021; Singh et al. 2021). The receptor site is semi-flexible docked with the ligand. The receptor site is maintained as rigid, while the ligand is treated as flexible (Singh et al. 2020). The docking simulation was performed using the “Lamarckian Genetic Algorithm” method and the following associated parameters: 150 individuals in a population with a maximum of 2,500,000 energy evaluations and a maximum of 27,000 generations, the maximum number of top individuals that automatically survive is usually set to 1, the rate of gene mutation in the genetic algorithm is 0–1, and the default value is 0.02, followed by 100 iterations of Solis and Wets local search. Other parameter values were kept in default. The final figures were generated using the Discovery Studio Visualizer program (Accelrys) (Meng et al. 2012a, b; Kou et al. 2021).

Results

Identification of the endophytic fungal FL15

Strain FL15 was cultured in CYA medium at 28 °C for 7 days. The colony diameter can reached 35–40 mm; its color was goose yellow; and the surface mycelium was white and fluffy, slightly convex in the middle, and had yellow droplet of exudation and a flat back (Fig. 2A). In the same culture conditions on PDA, the colony diameter was 46–50 mm; its color was yellow velvet; and the mycelium bulged in the middle, with no droplet exudation on the surface. With the extension of culture time, the color of the front of the colony deepened from yellow to dark yellow, whereas the color of the back and front remained the same (Fig. 2B). In the meanwhile, the strain FL15 was cultured on YES medium at 28 °C for 7 days. The colony diameter was 36–40 mm; its color was white velvet; and it had a flat surface, was slightly bulged in the middle, and a pale-yellow edge at the back, whereas the colony center was white (Fig. 2C). Furthermore, a scanning electron microscopy observation showed that the mycelia of this strain were slender and bamboo-like, with many branches. Clustered conidia were observed in the head, and most broom-like branches were bicycles. There were elliptical or nearly spindle-shaped densely distributed spores near the mycelia, and the walls were smooth (Fig. 2D). The mycelium had a slender septate, was branched, and the wall was smooth (Fig. 2E, F). Based on the colony and cell morphological observation, strain FL15 was recognized as a member of the genus Talaromyces. (Fig. 2).

Fig. 2
figure 2

Colony morphology and mycelium morphology of the endophytic fungus FL15. Colony characteristics of FL15 on CYA (A), PDA (B); and YES (C), respectively; microscopic morphology of FL15 (100 ×) (D); scanning microscopy-based electron micrograph morphology of FL15 (bar = 50 μm) (E); scanning electron microscopy-based micrograph morphology of FL15 (bar = 20 μm) (F)

Full size image

The length of the ITS, 18S rDNA, and 28S rDNA sequence was 589, 1708, and 940 bp, respectively. The accession number(s) for the ITS, 18S rDNA, and 28S rDNA nucleotide sequences of the strain FL15 in GenBank were MZ542471, MZ540308, and MZ540309, respectively. BLAST analysis of ITS of strain FL15 showed 99% of sequence similarity with T. aurantiacus. Based on the ITS, 18S rDNA, and 28S rDNA sequence analysis, strain FL15 was identified as belonging to the phylum Ascomycota, order Eurotiales, family Trichocomaceae, and genus Talaromyces. A phylogenetic tree was constructed using the neighbor-joining and MEGA 7.0.14. By combining the results of morphological observation analysis and those ITS, 18S rDNA, and 28S rDNA sequence phylogenetic analysis, the strain FL15 was identified as T. aurantiacus FL15 (Fig. 3).

Fig. 3
figure 3

Phylogenetic tree of strain FL15 and corresponding strains based on ITS (A), 18S rRNA (B), and 28S rRNA (C) sequences

Full size image

Structural elucidation of the purified compounds

A total of seven natural compounds were isolated and purified from the crude ethyl acetate extract of the mycelia of the endophytic fungus T. aurantiacus FL15. Their chemical structures were elucidated based on 1H nuclear magnetic resonance (NMR), 13C NMR, and electrospray ionization/mass spectrometry (ESI–MS) analyses (Supplementary Tables S1–S3), as well as via the comparison of their properties and spectral characteristics with published data. The compounds were confirmed as being asterric acid (1) (Hargreaves et al. 2002), and its derivatives, methyl asterrate (2) (Hargreaves et al. 2002), and ethyl asterrate (3) (Li et al. 2008), together with four other compounds, emodin (4) (Li et al. 2008), physcion (5) (Guo et al. 2011), chrysophanol (6) (Guo et al. 2011), and sulochrin (7) (Liu et al. 2015). The chemical structures of the compounds are depicted in Fig. 4.

Fig. 4
figure 4

Structures of compounds 1–7 isolated from FL15

Full size image

Anti-acetylcholinesterase/butyrylcholinesterase activity

The AChE/BuChE inhibitory activities of the purified compounds were evaluated using Ellman’s spectrophotometric method, with rivastigmine and Hup A as the reference compounds. The half-maximal inhibitory concentrations (IC50) of the compounds for AChE/BuChE inhibition are summarized in Table 1. The results showed that compounds 1–3 exhibited moderate AChE inhibitory activities, whereas the four other compounds displayed no AChE inhibitory activities. The tested compounds did not afford any inhibition of BuChE. Based on these results, the asterric acid derivatives could be described as being highly selective AChEIs. Moreover, compound 3 exhibited better inhibition against AChE, with an IC50 value of 20.1 µM than did compounds 2 (IC50 = 23.3 µM) and 1 (IC50 = 66.7 µM).

Table 1 Acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) inhibitory activities and log P values of the title compounds
Full size table

The ability of a compound to cross the blood–brain barrier (BBB) is essential for AD treatment. Thus, log P was an important physicochemical parameter for the evaluation or prediction of the ability of the compounds to cross the BBB. The log P values of the isolated compounds were calculated using Lipinski’s Rule of Five (http://www.scfbio-iitd.res.in/software/drugdesign/lipinski.jsp) and are shown in Table 1. The optimum log P value for central nervous system penetration is around 2 ± 0.7. Thus, the log P results indicated that the isolated compounds were sufficiently lipophilic to pass the BBB.

Kinetic study of acetylcholinesterase inhibition

The crystal structure of AChE in complex with inhibitors revealed the presence of dual binding sites: a Ser–His–Glu catalytic site (CAS) located at the bottom of the gorge, and a peripheral anionic-binding site (PAS) located at the gorge entrance (Bartus et al. 1982; Miles and Ross 2021). An enzyme kinetic study was performed to explore the AChE inhibition mode of compounds 1–3. The results obtained from in the reciprocal Lineweaver–Burk plot (Fig. 5) show increased slopes (decreased Vmax), and intercepts (higher Km) at higher inhibitor concentrations, and a mixed-type inhibition was demonstrated. Therefore, compounds 1–3 might be able to simultaneously bind to CAS and PAS, as well as the catalytic triad of AChE. The inhibitory constant, Ki (0.14 mM), was determined by plotting the slopes of the Lineweaver–Burk reciprocal plots versus the concentrations of compound 3.

Fig. 5
figure 5

Kinetic study of the inhibition mode of AChE by compound 3. Overlaid Lineweaver–Burk reciprocal plots of AChE initial velocity at increasing substrate concentration (0.01–0.04 mM) in the absence of inhibitor and the presence of different concentrations of compound 3 are shown

Full size image

Docking study

Docking studies were performed using AutoDock 4 to analyze the binding mode of the asterric acid derivatives to AChE (PDB ID: 1F8U) (Kryger et al. 2000). From the docking results, AChE interacted with compound 1 through 14 amino acid residues, namely, Asp73, Trp85, Gly119, Tyr123, Ser124, Tyr132, Glu201, Ser202, Phe292, Tyr332, Phe333, His471, and Tyr 473. Hydrogen-bond interaction with Asp73, Gly119, Tyr123, Ser124, Tyr132, Glu201, and Ser202, π–π interactions with Trp85, π–σ interactions with Tyr332, π-alkyl interactions with Phe292, Phe333, and Tyr 473, and carbon–hydrogen bonds with His471 was observed (Fig. 6.1A–C). The methyl ester group of benzene ring interacts with His471 by van der Waals force at a distance of about 3.12 Å, and with Ser202 by hydrogen bonding at a distance of about 2.12 Å. Whereas, AChE interacted with compound 2 through 15 amino acid residues, hydrogen bonds with residues Asp73, Gly119, Gly120, Tyr123, Ser124, Tyr132, Glu201, and Ser202, π–σ interactions with Tyr332, π-alkyl interactions with Phe292 and Phe333, and carbon–hydrogen bonds with Thr82, Trp85, Asn86, and His471 (Fig. 6.2A–C). Compared to the interaction between compound 1 and electric eel AChE, the ester bond site of methyl asterrate at the peripheral anion site formed hydrogen bond with Thr82 and Asn86. These two hydrogen bonds were the non-existent interaction between compound 1 and electric eel AChE, which may increase the interaction force between methyl asterrate and the peripheral anion site. Compound 3 interacted AChE through 14 amino acid residues. Compound 3 formed hydrogen bonds with residues Asp73, Gly119, Gly120, Tyr123, Ser124, Tyr132, Glu201, Ser202, π–σ interactions with residues Trp85, Tyr332, π-alkyl interactions Pro87, Phe292, and Phe333, while His471 exhibited carbon–hydrogen-bond interaction (Fig. 6.3A–C). The values of the free energy of binding (ΔG) for the ‘asterric acid–AChE’, ‘methyl asterrate–AChE’ and ‘ethyl asterrate–AChE’ interactions were − 7.89, − 9.72, and − 9.74 kcal/mol, respectively, and their corresponding IC50 values were estimated to be 66.7, 23.3, and 20.1 μM, respectively (Kryger et al. 2000; Shaikh et al. 2015). The active site of AChE was reported to be located at the bottom of a deep and narrow gorge, consisting of a number of domains. The peripheral anion site is located at the entrance of the active pocket, which comprises residues Trp85 and Trp304. The catalytic active site for AChE of electric eels composed of Ser202, Glu325, His471, Trp85, and Trp304. From the docking results in Fig. 6, Compound 1–3 could enter the active cavity of electric eelfish AChE and interact with the amino acids of the peripheral anion sites in the active cavity and the main amino acids of the catalytic active center. Ser202 and His471 interact with the catalytic active center mainly by hydrogen bond, and Trp85 and so on form π–π stacking interaction. At the bottom of the gorge, the benzene ring structure of the diphenyl ether interacted with Trp85 via π–π stacking, and the oxygen atom of the ester group created a hydrogen bond with the hydroxyl group of Tyr123 (Xu et al. 2020; Miles and Ross 2021). In summary, the asterric acid derivatives can interact with both peripheral activity sites (PAS) and catalytic activity sites (CAS) of AChE.

Fig. 6
figure 6

Diagrams the docking poses (A), interactions (B), and two-dimensional interactions (C) of asterric acid (1), methyl asterrate (2), and ethyl asterrate (3) docked to AChE (PDB: 1F8U). The dashed lines represent bonding interactions. The interacting amino acid residues are labeled. The figures were generated using the Discovery Studio Visualizer software (Accelrys)

Full size image

Discussion

AChEIs were initially isolated from plants (Su et al. 2017). However, the production of AChEIs through plant extraction processes is limited due to the lack of natural plant resources. Meanwhile, marked-available AChEIs derived from plants showed a lot of disadvantages such as low bioavailability and other abdominal side effects. Thus, exploring other alternatives of AChEIs derived from microbial sources with different niches is a must (Su et al. 2017; Zaki et al. 2020). Endophytes are a rich source of bioactive and chemically novel compounds with huge medicinal and agricultural potential. Furthermore, they can produce bioactive substances identical or similar to those of host plants (Su et al. 2017; Zaki et al. 2020). Hence, searching for a natural, cost-effective, and sustainable source of effective AChEIs from endophytes has become an attractive subject for many researchers (Su et al. 2017; Zaki et al. 2020). So far, various structural types of AChEIs, including alkaloids, terpenoids, and other compounds, were found in fungi, especially endophytic fungi (Table 2), suggesting that fungi represent valuable, novel, and alternative resources with good AChE inhibitory activity (Su et al. 2017; Zaki et al. 2020). To date, more than 300 endophytic fungal isolates from H. serrata have been isolated, of which 9 endophytic fungal strains can produce Hup A (Cao et al. 2021). Additionally, avertoxin B isolated from endophytic fungi of H. serrata showed AChE inhibitory activity (IC50, 14.9 μM) (Wang et al. 2015). Herein, we also isolated diphenyl ethers AChEIs from endophytic fungi of H. serrata.

Table 2 Various AChEIs derived from fungi
Full size table

The existed studies showed that fungal AChEIs displayed different AChE inhibitory activity with IC50 from 0.026 μM to 280 μM (Table 2). Our result showed that asterric acid derivatives (compounds 1–3) exhibited AChE inhibitory activities with IC50 values of 66.7, 23.3, and 20.1 μM, respectively. Compared to IC50 values of other AChEIs derived from fungi, the asterric acid derivatives exhibited moderate AChE inhibitory activity (Table 2). Meanwhile, asterric acid derivatives have no inhibitory activities against BuChE (Table 1), indicating that asterric acid derivatives have high selectivity.

According to the available reports, the asterric acid derivatives were firstly isolated from the fermentation broth of Aspergillus terreus in 1960 (Curtis et al. 1960). Thereafter, these metabolites were gradually discovered in fungi that belong to different genera, such as Penicillium frequentans (Mahmoodian and Stickings 1964), Oospora sulfureaochracea (Natori and Nishikawa 1962), Scytalidium spp. (Stermits et al. 1973), Pestalotiopsis spp. (Ogawa et al. 1995; Liu et al. 2009), Phoma sp. (Jayasuriya et al. 1995), Geomyces sp. (Li et al. 2008), and A. flavipes (Zhang et al. 2016), as well as in endophytic fungi, such as Neoplaconema napellum IFB-E016 from Hopea hainanensis (Wang et al. 2006), Aspergillus sp. F1 from Trewia nudiflora (Lin et al. 2009), and Pseudogymnoascus sp. from the Antarctic marine sponge Hymeniacidon spp. (Figueroa et al. 2015). To the best of our knowledge, these three asterric acid derivatives were isolated from the genus Talaromyces for the first time.

Asterric acid and its analogs have attracted considerable interest as the first non-peptide endothelin-1-binding inhibition (Ohashi et al. 1992). Thus, asterric acid, as a vascular endothelial growth factor inhibitor and antibiotic, has been commercialized as a biological reagent and used in biological and medical-related research (Tang et al. 2007). Furthermore, several asterric acid derivatives are useful for treating myocardial infarction and renal insufficiency (Curtis et al. 1960). Asterric acid and its analogs also exhibit other good or excellent medicinal activities. Dimethyl 2,3-dimethoxyosoate and 2-(2-formyl-3-hydroxy-5-methylphenoxy)-5-hydroxy-3-methoxybenzoate displayed cytotoxicity against the K562 cell line (Liu et al. 2006) and low cytotoxic activity in HepG2 and Raji cells (Fang et al. 2012), respectively. Geomycins B and C displayed substantial antifungal activity against A. fumigatus and antimicrobial activities against Gram-positive and Gram-negative bacteria, respectively (Li et al. 2008). More recently, methyl asterrate, methyldichloroasterrate, methyl 3-chloroasterric acid, monomethylosoic acid, and 2,4-dichloroasterric acid exhibited more potent inhibitory activities against α-glucosidase vs. acarbose (Wang et al. 2016a, b; Zhang et al. 2016). Herein, we showed for the first time that asterric acid and its derivatives displayed highly selective AChE inhibitory activities.

To further reveal the relationship between asterric acid derivatives and acetylcholinesterase and the reasons for the differences activities of asterric acid derivatives, the molecular docking analysis was carried out. Docking is a structure-based drug design method, which can effectively estimate the binding energy and conformation of drugs (Devidas et al. 2021). Docking results show that the three asterric acid derivative molecules could interact with PAS and CAS of eel AChE. One of the benzene rings of diphenyl ether binds through hydrogen bonding and π–π interaction with the key amino acid residues of peripheral sites, while another substituent of the benzene ring interacts with center of the catalytic activity. Among them, the hydrophobic ester bond and hydrophobic benzene ring formed a large hydrophobic pocket with hydrophobic amino acid residues such as Gly119, Gly120, Phe292, and Phe333. These interactions enable the three small molecules to bind well to the electric eel AChE, and thus compete with the substrate to inhibit the AChE activity. Pan et al (2019) reported that linarin improves the dyskinesia by inhibiting AChE. To assess whether linarin could dock with AChE and decipher mechanism of linarin as AChEI, molecular docking simulation was used. The result shows that linarin may inhibit AChE by binding to the hydrophobic active site of the alkoxy substrate including residues Phe 330 and Phe 331. This active site is different from our docking result of AChE inhibition by simultaneously binding to CAS and PAS. In comparison with the docking results of Devidas et al. 2021, our result is consistent with those of previous studies in which the same amino acid residues of AChE played an essential role in substrate binding. The IC50 values of asterric acid (1), methyl asterrate (2), and ethyl asterrate (3) were 66.7, 23.3, and 20.1 µM, respectively, which suggests that the size and variety of the esterification substituent at C-8′ on the parent nucleus may contribute to the AChE inhibitory activities of these compounds. Moreover, a kinetic analysis revealed a mixed-type AChE inhibition of diphenyl ether compounds (Kou et al. 2021). Therefore, the AChE inhibitory activities of diphenyl ether compounds are important to identify their structure–activity relationships and design compounds with higher activity based on this lead compound. Accordingly, the asterric acid scaffold could be considered in the design of a new AChEI.

Conclusions

In the present study, three natural compounds asterric acid (1), methyl asterrate (2), and ethyl asterrate (3) possessed a diphenyl ether structure were isolated from endophytic fungal T. aurantiacus FL15 of H. serrata. These compounds exhibited potent AChE inhibitory activities, with IC50 values of 66.7, 23.3, and 20.1 μM, respectively. Docking analysis results not only showed that they displayed selectivity inhibitory activities to AChE, and their action against AChE was related to esterification R groups of 8 carbon on the parent nucleus, but also demonstrated binding interactions with the PAS and CAS of the enzyme. According to the calculated log P values, all three compounds might pass the BBB. To the best of our knowledge, this study was the first to report on asterric acid derivatives that act as AChEIs. Asterric acid could be considered as a new lead scaffold to develop more potent AChEIs.