Bioassay-Guided Isolation of Anti-Alzheimer Active Components from the Aerial Parts of Hedyotis diffusa and Simultaneous Analysis for Marker Compounds

Previous studies have reported that Hedyotis diffusa Willdenow extract shows various biological activities on cerebropathia, such as neuroprotection and short-term memory enhancement. However, there has been a lack of studies on the inhibitory activity on neurodegenerative diseases such as Alzheimer’s disease (AD) through enzyme assays of H. diffusa. Therefore, H. diffusa extract and fractions were evaluated for their inhibitory effects through assays of enzymes related to AD, including acetylcholinesterase (AChE) and butyrylcholinesterase (BChE), and β-site amyloid precursor protein cleaving enzyme 1 (BACE1), and on the formation of advanced glycation end-product (AGE). In this study, ten bioactive compounds, including nine iridoid glycosides 1–9 and one flavonol glycoside 10, were isolated from the ethyl acetate and n-butanol fractions of H. diffusa using a bioassay-guided approach. Compound 10 was the strongest inhibitor of cholinesterase, BACE1, and the formation of AGEs of all isolated compounds, while compound 5 had the lowest inhibitory activity. Compounds 3, 6, and 9 exhibited better inhibitory activity than other compounds on AChE, and two pairs of diastereomeric iridoid glycoside structures (compounds 4, 8, and 6, 7) showed higher inhibitory activity than others on BChE. In the BACE1 inhibitory assay, compounds 1–3 were good inhibitors, and compound 10 showed higher inhibitory activity than quercetin, the positive control. Moreover, compounds 1 and 3 were stronger inhibitors of the formation of AGE than aminoguanidine (AMG), the positive control. In conclusion, this study is significant since it demonstrated that the potential inhibitory activity of H. diffusa on enzymes related to AD and showed the potential use for further study as a natural medicine for AD treatment on the basis of the bioactive components isolated from H. diffusa.


Sample
.92 ± 11.24 *** Data are expressed as the mean ± SD (n = 3); a IC 50 was calculated from the least-squares regression line of the logarithmic concentrations plotted against the residual activity; b Berberine was used as a positive control of ChE inhibitory activity; c Quercetin was used as a positive control of BACE1 inhibitory activity; d AMG was used as a positive control for the inhibition of the formation of AGE; e ND: not detected; * states a significant difference from control; * p < 0.05, ** p < 0.005, *** p < 0.001.
The H. diffusa extract remarkably inhibited AChE and BChE activity (IC 50 of 102.01 ± 4.90 and 99.79 ± 4.52 µg/mL, respectively). In the ChE inhibitory assay, the EA and BuOH fractions exhibited the highest inhibitory activity (IC 50 values of 25.98 ± 3.07 and 1.15 ± 0.32 µg/mL, respectively). Similarly, in the BACE1 inhibitory assay, the EA and BuOH fractions were found to show inhibitory activity much stronger than that of the H. diffusa extract and other fractions (IC 50 of 14.84 ± 1.24 and 26.92 ± 3.48 µg/mL, respectively). In addition, the EA fraction (IC 50 of 99.32 ± 1.31 µg/mL) showed the strongest inhibitory activity against the formation of AGE followed by the BuOH fraction (IC 50 of 109.27 ± 5.76 µg/mL).
In summary, the H. diffusa extract was an effective inhibitor of AChE, BChE, BACE1, and the formation of AGE. In all assays, the EA and BuOH fractions showed stronger inhibitory activity than the other fractions. In contrast, the DCM and water fractions possessed slight or no potential inhibitory activity.

Inhibitory Activities of Compounds 1-10
Isolated from H. diffusa against ChE (AChE and BChE), BACE1, and the Formation of AGE Compounds 3, 6, and 9 were much stronger AChE inhibitors than other compounds, except compound 10. Moreover, this study revealed the following relationships between the iridoid glycoside structure and AChE inhibitory activity: (1) the iridoid glycoside with methyl ferulate on C-6 (compound 3) was much more active than those with a methyl trans-p-coumarate and methyl p-coumarate substitutent on C-6 (compounds 1 and 2); (2) iridoid glycosides with a substituted hydroxyl group at the C-6 position (compounds 4 and 8), which are diastereomers, showed mild activity with IC 50 values of 172.26 ± 20.55 and 157.68 ± 13.18 µM, respectively; (3) iridoid glycosides with a substituted methoxy group at the C-6 position (compounds 6 and 7), which are diastereomers, had completely opposite activity. Although compound 6 significantly inhibited AChE, with an IC 50 value of 81.06 ± 5.58 µM, compound 7 did not inhibit AChE. (4) The iridoid glycoside with a substituted acetyl group on C-10 (compound 9), which had an IC 50 value of 68.34 ± 5.11 µM, exhibited the strongest inhibitory activity among the iridoid glycoside compounds.
With respect to BChE inhibitory activity, two kinds of diastereomers among the iridoid glycosides (compounds 4, 8 and 6, 7) showed higher inhibitory activity than others. Furthermore, among compounds 1-3, derivatives of methyl p-coumarate at the C-6 position (compound 2) was more active than the derivatives of methyl trans-p-coumarate and methyl ferulate (compounds 1 and 3).
Although compound 1, 2, 4 and 8 showed mild activity on AChE, these compounds had effective inhibitory activities on BChE.
In the BACE1 inhibitory assay, compound 10 was more effective than quercetin (positive control) and had the highest inhibitory activity against BACE1. Among the isolated compounds, only the iridoid glycosides with methyl trans-p-coumarate, methyl p-coumarate, and methyl ferulate at the C-6 position (compounds 1, 2, and 3) showed considerable inhibitory activity, in that order. Other compounds, except compounds 1-3, had low inhibitory activity against BACE1 among the iridoid glycoside compounds.
Compounds 1 and 3 were stronger inhibitors of the formation of AGE than AMG (positive control). This study suggested the following structural features for inhibition of AGE formation by iridoid glycosides: iridoid glycosides with derivatives of methyl trans-p-coumarate, methyl p-coumarate, and methyl ferulate (compounds 1-3) were much stronger than other compounds, except compound 10. Compound 4-9 showed low or no activity.
In particular, compound 5 had considerably low inhibitory activity against AChE, BChE, BACE1, and the formation of AGE with IC 50 values of 258.81 ± 7.48, >500, >500, and >1000 µM, respectively. The flavonol glycoside compound (compound 10) showed significantly high inhibitory activities against AChE, BChE, BACE1, and the formation of AGE with IC 50 values of 46.22 ± 1.59, 13.77 ± 0.37, 4.49 ± 1.86, and 2.71 ± 0.06 µM, respectively. The results are summarized in Table 4. Table 4. IC 50 of the compounds 1-10 for cholinesterase (acetylcholinesterase (AChE) and butyrylcholinesterase (BChE)), β-site amyloid precursor protein cleaving enzyme 1 (BACE1), and the formation of advanced glycation end-product (AGE) inhibition. Data are expressed as the mean ± SD (n = 3); a IC 50 was calculated from the least-squares regression line of the logarithmic concentrations plotted against the residual activity; b Berberine was used as a positive control of ChE inhibitory activity; c Quercetin was used as a positive control of BACE1 inhibitory activity; d AMG was used as a positive control of the inhibition of the formation of AGE; e ND: not detected; * states a significant difference from control; * p < 0.05, ** p < 0.005, *** p < 0.001.

Simultaneous Quantitative HPLC Analysis of Six Bioactive Components in H. diffusa and H. corymbosa
An HPLC analysis of H. diffusa and H. corymbosa extracts was performed for the quantitative evaluation of the bioactive components ( Figure 2). After screening the collected samples of H. diffusa, compounds 1 and 9 were identified as the first and second major compounds, respectively, of H. diffusa extract. The six bioactive components (1, 3, 4, 8, 9, and 10) from H. diffusa exhibited considerably strong inhibitory activity in AD assays. To optimize the extraction efficiency, samples were extracted by altering the extraction solvent, solvent ratio, and time (Table 5). Among these different extraction times and solvent compositions, the sample extracted after 90 min and using 70% methanol (as solvent) contained the highest amount of the six marker compounds. Furthermore, comparing the two chromatograms of H. diffusa and H. corymbosa, the major components, compounds 1 and 3, of the former were not present in the latter (Figure 4). To distinguish between H. diffusa and H. corymbosa, we discovered, using the developed simultaneous analysis method, that compounds 1 (E-6-O-p-coumaroyl scandoside methyl ester) and 3 (E-6-O-feruloyl scandoside methyl ester) can be potential biomarkers. The quantity of all compounds, except compound 8 (Scandoside methyl ester), was much higher in H. diffusa than in H. corymbosa (Figure 4). In H. diffusa extract, the quantities of the major bioactive compounds 1, 3, and 10 were much higher than those in H. corymbosa. These results suggested that simultaneous quantitative HPLC analysis of these six bioactive compounds can be used to obtain quality control standards of H. diffusa.

Extraction, Fractionation, and Isolation of H. diffusa
The aerial parts of H. diffusa (7.7 kg) were dried and powdered, and then extracted in methanol (20 L × 3) at room temperature. The total filtrate was concentrated to dryness at 50 • C to yield the MeOH extract (2256.32 g). It was suspended in distilled water and then partitioned sequentially in Hx, DCM, EA, and n-BuOH. The results yielded Hx (50.72 g), DCM (56.07 g), EA (10.59 g), n-BuOH (45.19 g), and water (86.50 g) fractions. Among these five fractions, the EA and n-BuOH fractions were found to be most potent in the four anti-Alzheimer disease model assays. Therefore, open column chromatography of these active fractions was performed repeatedly, and five compounds were obtained in each of the EA and n-BuOH fractions.

NMR
1D nuclear magnetic resonance (NMR) spectra were analyzed at 600 MHz ( 1 H-NMR) and 150 MHz ( 13 C-NMR) using a JNM-ECZ600R spectrometer (JEOL, Tokyo, Japan). Samples were dissolved in deuterated methanol (CD 3 OD) and dimethyl sulfoxide (DMSO-d 6 ). Chemical shifts are presented as ppm (parts per million) on the δ scale and coupling constants (J) are presented in Hertz.

HPLC Analysis
To analyze the six bioactive components, such as E-6-O-p-coumaroyl scandoside methyl ester (1) , from H. diffusa, the Waters Kromasil C 18 column (4.6 × 250 mm, 5 µm) was used to analyze the major compounds isolated from H. diffusa. The mobile phase system consisted of 0.1% acetic acid in water (solvent A) and acetonitrile (solvent B) at a flow rate of 0.8 mL/min. The linear gradient elution was implemented with the following elution program: 0-40 min, 10-30% B; 40-50 min, 30-60% B. All eluents were filtered in a 0.45 µm PVDF syringe filter. The sample injection volume was 10 µL, and UV 254 nm was selected as the optimal wavelength for detecting the compounds. For preparation of extract stock solutions, plant powders were sonicated with 70% MeOH for 90 min and dried under vacuum by using a rotary evaporator at 50 • C. After then, they were dissolved in MeOH to a concentration of 10,000 ppm. Standard compound stock solutions were also dissolved in MeOH. Prior to injection, all analyzed stock solutions were strained using a 0.45 µm PVDF syringe filter. The standard calibration curve was constructed using five different concentrations. The linear relationship between peak area and concentration is described in Table 6. The concentrations of the six major components were calculated using regression equations based on the calibration curves. In the regression equation y = ax + b, x refers to the concentration of the compound (µg/mL), y the peak area; r 2 : the correlation of the equation; Rt: retention time; LOD: limit of detection; LOQ: limit of quantification.

Measurement of ChE Enzyme Assay
The inhibitory activities of the test samples against ChE (AChE and BChE) were evaluated using the spectrophotometric method developed in a previous study [38]. ATCh and BTCh were used as substrates to measure the inhibitory activities of AChE and BChE, respectively. The assay mixture contained 0.1 M sodium phosphate buffer (pH 7.8), 0.3 U/mL AChE or BChE, 0.5 mM DTNB, 0.6 mM ATCh or BTCh, and the tested sample solution that were mixed and incubated for 15 min at room temperature. All tested samples and positive control (berberine) were dissolved in 10% analytical grade DMSO at five different final concentrations (10-500 µg/mL for extracts and fractions or 10-500 µM for isolated compounds). Reactions started on addition of 10 µL of DTNB and 10 µL of either ATCh or BTCh. The ChE inhibitory activity was monitored based on the formation of the yellow 5-thio-2-nitrobenzoate anion at 412 nm for 15 min that was due to the reaction of DTNB and thiocholine released from ATCh or BTCh. All reactions were measured in 96-well microplates and tested in triplicate. The percentage of inhibition (%) was estimated using the following formula: {(Ac − As)/Ac} × 100, where Ac is the enzyme activity without the test sample and As is the enzyme activity with the test sample.

Measurement of BACE1 Enzyme Assay
The BACE1 enzyme assay was performed in accordance with the manufacturer's recommended protocol, with minor modifications. Briefly, the assay mixture contained 1.0 U/mL of BACE1, 50 mM sodium acetate buffer (pH 4.5), the substrate (750 nM Rh-EVNLDAEFK-Quencher in 50 mM ammonium bicarbonate), and samples. All samples and positive control (quercetin) were dissolved in 10% analytical grade DMSO at five different final concentrations. The reaction mixture was incubated for 60 min at 25 • C in the dark. Prior to measurement, stop solution (2.5 M sodium acetate) was added to the assay mixture. The BACE1 enzyme assay was determined by measuring the proteolysis of two fluorophores (Rh-EVNLDAEFK-Quencher) to form a fluorescent donor (Rh-EVNL) that increased in fluorescence wavelengths at 530-545 nm (excitation) and 570-590 nm (emission), respectively. All reactions were measured in black 96-well microplates and tested in triplicate. The percentage of inhibition (%) was obtained using the following formula: [1 − (S 60 − S 0 )/(C 60 − C 0 )] × 100, where C 60 is the fluorescence of the control after 60 min of incubation, C 0 is the initial fluorescence of the control, S 60 is the fluorescence of the tested sample after 60 min of incubation, and S 0 is the initial fluorescence of the tested sample. BACE1 inhibitory activity assay of each sample was presented in terms of IC 50 , as calculated from the log dose inhibition curve.

Measurement of Inhibition of Formation of the AGE
The inhibitory activity of the formation of AGE was measured with a spectrophotometric method developed previously, with slight modifications [39]. Briefly, the assay mixture contained 50 mM phosphate buffer (pH 7.4) with 0.02% sodium azide, 0.4 M fructose and glucose, bovine serum albumin (10 mg/mL), and the sample. Next, the assay mixture was incubated at 60 • C for 2 days. After incubation, 200 µL of the reaction product was measured at excitation and emission wavelengths of 350 and 450 nm, respectively. All samples and positive control (aminoguanidine) were dissolved in 10% analytical-grade DMSO at five different final concentrations. All reactions were measured in black 96-well microplates and tested in triplicate. The percentage of inhibition (%) of the formation of AGE was estimated using the following formula: {(Ac − As)/Ac} × 100, where Ac is the fluorescence of the control and As is the fluorescence of the sample. Inhibition of the formation of AGE of each sample was presented in terms of IC 50 , as calculated from the log dose-inhibition curve.

Statistical Analysis
All assays were conducted in triplicates. Data are presented as the mean ± standard deviation (SD) and determined using one-way analysis of variance to evaluate differences between the positive control and treatment sample groups. The data were considered statistically significant at p < 0.05.

Conclusions
In this study, we determined the inhibitory potential of MeOH extract and fractions, isolated from H. diffusa, against ChE (AChE and BChE), BACE1, and the formation of AGE. The EA and BuOH fractions exhibited the strongest inhibitory activities, hence, we isolated ten major bioactive compounds, nine iridoid glycosides (1-9) and one flavonol glycoside (10), from these fractions, in accordance with bioassay-guided isolation. They were identified using 1 H-NMR, 13 C-NMR, and ESI/LTQ-Orbitrap-HRMS techniques. Then, compounds 1-10 from H. diffusa were evaluated for their anti-AD potencies, by conducting inhibitory assays of AChE, BChE, BACE1, and the formation of AGE. Since compounds 1, 3, 4, 8, 9, and 10 had potential anti-AD inhibitory activities, we developed and validated a method that quantified and analyzed these compounds. Compound α (asperuloside) had the lowest anti-AD activity in all assays. Previous studies have reported that compound β (quercetin-3-O-sophoroside) does not show effective neuroinflammation inhibitory activity for the treatment of AD [40]. Therefore, among the eight biomarkers (compounds 1, 3, 4, 8, 9, 10, α, and β) separated by simultaneous HPLC analysis, compounds α and β that were identified as asperuloside and quercetin-3-O-sophoroside, were not selected from the chromatograms. Using this developed simultaneous analysis method, these six marker compounds (compounds 1, 3, 4, 8, 9, and 10) were successfully quantified in collected H. diffusa and H. corymbosa samples under optimized and efficient solvent extraction conditions. Among the samples, the H. corymbosa Lamark sample could be mistook for H. diffusa (which is also a genus of flowering plant in the family Rubiaceae), owing to their similar external appearance. This study suggested that compounds 1 (E-6-O-p-coumaroyl scandoside methyl ester) and 3 (E-6-O-feruloyl scandoside methyl ester) could be identified as biomarkers to distinguish between H. diffusa and H. corymbosa using the developed HPLC analytical method. This study also proved that H. diffusa is a novel remedy for AD and could be a potential drug. In addition, the developed analytical HPLC method could be applied to various fields for quality control of H. diffusa. It is necessary to conduct further research on H. diffusa to confirm the results obtained in this study, including prospective clinical trials, and investigate the medicinal effects of the isolated compounds. Furthermore, the compounds isolated from H. diffusa may be valuable therapeutic agents for other diseases.

Conflicts of Interest:
The authors declare no conflict of interest.