A Platform for Screening Potential Anticholinesterase Fractions and Components Obtained from Anemarrhena asphodeloides Bge for Treating Alzheimer's Disease

Alzheimer's disease (AD) is a neurodegenerative disease characterized by progressive memory loss and cognitive impairment. Cholinesterase inhibitors are widely used for the symptomatic treatment of Alzheimer's disease to enhance central cholinergic transmission. In this study, a bioactivity-oriented screening platform based on a modified Ellman's method and HPLC-QTOF MS technique was developed to rapidly screen active agents of Anemarrhena asphodeloides Bge. The 60% ethanol fraction from an ethyl acetate extract exhibited the most potential anticholinesterase activity. Fifteen steroid saponins were identified by the mass spectrum, standards and literature reports. Twenty-five compounds were isolated from the active fraction. The results showed that compounds with the C6–C3–C6 skeleton probably had both AChE and BuChE inhibitory activities. Xanthone and benzene derivatives exhibited no or little activity. Lignans showed weak BuChE inhibitory activity. The steroidal saponins demonstrated moderate or weak AChE inhibitory activity.


Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disorder of the central nervous system (CNS), characterized by deposits of aberrant proteins, namely, -amyloid (A ) and -protein, loss of synapses, death of cholinergic neurons, and oxidative stress [1,2]. The etiopathogenesis of AD still remains unknown, although the neurodegeneration leads to remarkable reduction of neurotransmitter acetylcholine at the synaptic clefts [3,4]. An effective strategy to slow down the progression of deterioration in AD patients is using anticholinesterase inhibitors.
Acetylcholinesterase (AChE) and butyrylcholinesterase (BuChE) constitute the group of cholinesterases. AChE hydrolyses acetylcholine (ACh) and is mainly associated with nerves and muscles, being typically found on the synapses. AChE inhibitors are widely used for the symptomatic treatment of AD to enhance central cholinergic transmission. BuChE hydrolyses butyrylcholine (BuCh) and is synthesized by the liver, being found in large concentration in serum [5]. In healthy brains, AChE hydrolyzes the majority of ACh while BuChE plays a secondary role. However, as AD progresses, BuChE can compensate for AChE when the activity of AChE is inhibited by AChE inhibitors. Thus, BuChE hydrolyses the already depleted levels of ACh in these patients [6]. It has been proposed that individuals with low activity of BuChE can sustain cognitive functions better than individuals with normal BuChE activity. Furthermore, BuChE inhibitors have been reported to produce a significant increase in brain extracellular AChE without triggering severe peripheral or central side effects [7].
Anemarrhena asphodeloides Bge, which belongs to the family Liliaceae, is widely distributed in China [8]. The rhizomes of Anemarrhena asphodeloides Bge have been reported 2 Evidence-Based Complementary and Alternative Medicine to have the cholinesterase inhibitory activity relevant to treatment of AD [9]. In order to find the candidate of a drug treating AD, the activities of fractions and compounds were assayed. Screening and identification of bioactive constituents in traditional Chinese medicines (TCMs) are the keys of the development of TCMs [10,11]. However, the complexity and variability of TCMs present a challenge to the identification of their structures. Now more and more attention has been attracted to bioactivity-based LC-MS/MS identification technology owing to its high efficiency and high specificity [12,13].
In this work, we used a bioactivity-oriented screening strategy, which was based on a modified Ellman's method and high performance liquid chromatography quadrupole-timeof-flight mass spectrometer (HPLC-QTOF MS) technique. The 60% ethanol fraction from an ethyl acetate extract showed the most potential anticholinesterase activity. Fifteen steroid saponins were identified by the mass spectrum, standards, and literature reports. Twenty-five compounds were isolated from the active fraction. Compounds with the C 6 -C 3 -C 6 skeleton probably had both AChE and BuChE inhibitory activities. Xanthone and benzene derivatives exhibited no or little activity. Lignans showed weak BuChE inhibitory activity. The steroidal saponins demonstrated moderate or weak AChE inhibitory activity.

General
Instrumental Equipment. HPLC system (Agilent, USA) consisted of a model G1276A pump, model G1367B Autosampler and model G1316A UV detector. The chromatograph was equipped with a reversed-phase C18 column of Grace Alltima (250 mm × 4.6 mm, 5 m). The QTOF-MS system (Bruker, Germany) with an ESI source was performed. HPLC separations were performed on a Hitachi 655-15 series pumping system equipped with a Hitachi L-2490 refractive index detector using a YMC-Park ODS-A column (250 × 10 mm I.D, S-5 m, and 12 nm). NMR spectra were performed on a Bruker ARX-300, ARX-400, and ARX-600 spectrometer using trimethylchlorosilane as the internal standard. Column chromatography was performed on a 200-300 mesh silica gel (Qingdao Marine Chemical Factory, People's Republic of China). Column chromatography was performed using YMC ODS-A gel (12 nm S-75 m, YMC Co., Ltd., Japan) and D-101 macroporous adsorption resin (Shanghai Hualing Resin Factory, People's Republic of China). TLC was performed with precoated silica gel GF 254 plates (Qingdao Marine Chemical Factory, People's Republic of China). Microplate reader (Thermofisher Scientific, Finland) was used to test the activities. Acetylthiocholine iodide, Sbutyrylthiocholine iodide, AChE, and BuChE were bought from Sigma Company.

Extraction Procedures.
Dried rhizomes of Anemarrhena asphodeloides Bge were powdered into a homogeneous size by a disintegrator and then sieved (60 mesh). The materials were extracted by three different techniques (ultrasonic, heat reflux, and cold soak techniques). Various solvents including petroleum ether, dichloromethane, ethyl acetate, acetone, methanol, 95% ethanol, 70% ethanol, 50% ethanol, 30% ethanol, and water were used for preparing active fractions. Accurate 5.0 g of the Anemarrhena asphodeloides Bge powders was weighted into ten Erlenmeyer flasks containing the mentioned ten solvents of 500 mL individually and then extracted with ultrasonic-assisted method twice (30 min each). After filtering, the filtrates were amalgamated and evaporated to dryness by rotary evaporator (50 ∘ C). For the heat reflux, the powders of 5.0 g were soaked in different solvents of 50 mL each for 30 min. Then, the heated reflux extraction experiments were conducted in water bath (90 ∘ C) for 2 h. The solutions were filtrated when they were still hot. Extractions were carried out for three times and the filtrates were evaporated to dryness. Lastly, 5.0 g of the powders was accurately weighed and soaked in 500 mL solvents overnight and then evaporated the filtrates to dryness.

AChE and BuChE Inhibitory Assay.
Cholinesterase inhibitory activity was evaluated using the modified method of Ellman. For AChE inhibitory assay, the reaction mixture consisted of 50 L of 100 mM PBS, pH 8.0, 25 L of 15 mM acetylthiocholine iodide, and 25 L of agents (1 mg/mL for extracts, 0.1 mg/mL for compounds) in a 96-well plate. The mixture was preincubated in 4 ∘ C for 10 min, and after that, 25 L of AchE (0.226 U/mL) and 125 L of 3 mM DTNB were added. The reactions were kept in 37 ∘ C for 20 min and then scanned at 412 nm with a microplate reader. For BuChE inhibitory assay, the same procedures were followed except for the use of substrate and enzyme, and S-butyrylthiocholine iodide and BuchE were used, respectively [14]. All of the experiments were repeated at least three times.

Structure Characterization and Identification of Active
Fraction. The ethyl acetate extract dealt with by the ultrasonic method was applied to a D-101macroporous resin column and eluted with ethanol and water to give 0%, 20%, 40%, 60%, 80%, and 95% ethanol fractions. The constituents of active fraction were assayed by HPLC-QTOF-MS. HPLC system (Agilent, USA) consisted of a model G1276A pump, a model G1367B Autosampler, and a model G1316A UV detector. The chromatograph was equipped with a reversed-phase C18 column of Grace Alltima (250 mm × 4.6 mm, 5 m) eluted with a gradient mobile phase. Mobile phases were water with 0.1% of formic acid (A) and acetonitrile (B). The gradient used was as follows: 0 min, 5% B; 0-10 min, 5% to 15% B; 10-18 min, 15% to 20% B; 18-23 min, 20% to 23% B; 23-30 min, 23% to 25% B; 30-48 min, 25% to 30% B; 48-55 min, 30% to 50% B; 55-80 min, 50% to 100% B. The injection volume of sample was 5 L. The flow rate was 0.8 mL min −1 and the column temperature was ambient temperature. The QTOF-MS system (Bruker, Germany) with an ESI source was performed in positive mode. The parameters of ESI-MS were

Statistical
Analysis. Data were expressed as means ± S8EM and analyzed statistically by one-way ANOVA, followed by post hoc (LSD) test. The results were considered statistically significant at value < 0.05.

The Activities of the Extracts.
Screening the extracts of Anemarrhena asphodeloides Bge dealing with three different techniques and ten kinds of solvents showed that all the extracts demonstrated no AchE inhibitory activity but exhibited BuChE inhibitory activity. The extract dealing with dichloromethane, ethyl acetate, and acetone had much more inhibition ratio than the others. The ultrasonic method and the cold soak method were better than the heat reflux method. It was found that Anemarrhena asphodeloides Bge extracted with ethyl acetate by the ultrasonic technique had the most potential anticholinesterase activity. The results on the effects of the tested extracts on AChE and BuChE inhibitory activities were summarized in Table 1.

3.2.
Screening the Active Fraction. The extract dealing with ethyl acetate by the ultrasonic method was applied to a D-101 macroporous resin column and eluted with ethanol and water to give 0%, 20%, 40%, 60%, 80%, and 95% ethanol fractions. Then they were tested on AChE and BuChE inhibitory activities. It indicated that the extract was partitioned into six fractions and active compounds were concentrated into 60% ethanol fraction. The results were summarized in Figure 2. Donepezil was used as positive drug with the inhibition ratio of 98.2% and 79.8%.

Structure Characterization and Identification of the Active
Fraction. The 20%, 40%, 60%, 80%, and 95% ethanol fractions were analyzed by HPLC-QTOF-MS technique. The water fraction was not tested for its inactivity. The total ion chromatograms of the five fractions were shown in Figure 3. As summarized in Table 2, a total of fifteen saponins from the 60% ethanol fraction were identified and tentatively characterized.
To obtain the information about precursor ions and characteristic fragment ions of the compounds, six authentic standards were injected into the LC-MS system. The fragmentation patterns for these authentic standards were discussed in detail below. MS spectra of the authentic standards were shown in Figure 4.
Peak 3 was definitely identified as timosaponin BII in comparison with an authentic standard. The precursor ion [M + Na] + at / 943.4886 was the base peak in ESI experiments. Peaks 4 and 5 showed the same formula and similar fragment ions as peak 3. Based on the chromatography behavior of steroidal saponin on a C 18 column, peak 4 was tentatively identified as the C25R stereoisomer of timosaponin BII. While peak 5 was identified as 25Sofficinalisnin-I or its isomer [8].
Peak 6, peak 7, and peak 8 were identified as timosaponin BIII, timosaponin G, and Anemarrhena saponin I in comparison with authentic standards.

Discussion
The anticholinesterase activity of the agents was tested in the presence of a known concentration. It was one of the approaches to study the effect of treating Alzheimer's disease. The anticholinesterase inhibition ratio of the agents was expressed as percentage inhibition ratio based on the optical density. The conditions of AChE and BuChE inhibitory assay were confirmed by a series of literature reports and experiments.
To determine which fraction has anticholinesterase activity, the rhizome of Anemarrhena asphodeloides was fractionated further by a series of extractions with different solvents and techniques. The findings revealed that more bioactive components would be extracted by dichloromethane, ethyl acetate, and acetone. The extracts dealt with by the ultrasonic and cold soak method had higher inhibition ratio than the extracts dealt with by the heat reflux method. It is proposed that the molecular structures of the active constituents may be destroyed at a high temperature.  The effect of ethyl acetate extract extracted by the ultrasonic method on inhibiting activity of cholinesterase was better than those dealt with by other solvents and methods. The ethyl acetate crude extract of Anemarrhena asphodeloides may also contain many compounds with differing polarities, so it was fractionated further by a D-101 macroporous resin column. The bioactive compounds of the 60% ethanol fraction were analyzed by HPLC-QTOF-MS system.
The type of compounds in the 60% ethanol fraction could be deduced by the MS data. If the ions [M + Na] + and [M + H − H 2 O] + were detected and the [M+H] + ion was not observed, the saponin would be a furostanol saponin with the C22 hydroxyl group. If the ions at / 417, 255, and 273 were detected, there should be no substituent attached to the sarsasaponin. If the ions at / 433 and 415 were observed, there should be one more hydroxyl group substituent on the aglycone. The retention time of saponins would be affected by the number, classes, and connection orders of sugar moieties and the number and position of hydroxyl group substituent at the aglycone. What is more, the retention time of steroidal saponins with the 25R configuration was longer than those with the 25S configuration on a C18 column.
Twenty-five compounds were isolated from the active fraction. The structure-activity relationships were discussed based on the results of cholinesterase inhibitory activity. It showed that the compounds with the C 6 -C 3 -C 6 skeleton probably had both AChE and BuChE inhibitory activities. Compound 3 showed potential BuChE inhibitory activitity with the inhibition ratio of 75.8%. Xanthone and benzene derivatives exhibited no or little activity. Lignans showed weak BuChE inhibitory activity. The steroidal saponins had moderate or weak AChE inhibitory activity. It was supposed that adding OH groups to the steroidal saponins may increase the effect a little. Although some of the individual compounds exhibited potential in vitro activity, they were less effective than the 60% ethanol fraction and this might be due to a synergistic action among the active components or maybe the most active constituents have not been isolated by us.
Steroid saponins were considered as potentially drug preventing and treating neurodegenerative diseases, such as Alzheimer's disease. However, we found that compounds with the C 6 -C 3 -C 6 skeleton probably had both AChE and BuChE inhibitory activities, especially compound 3. The activity in vivo needs to be tested in the future.

Conclusion
AChE and BuChE inhibitory activities of Anemarrhena asphodeloides Bge extracted with three different techniques