Inhibition of Amyloid Beta Aggregation and Deposition of Cistanche tubulosa Aqueous Extract

Cistanche tubulosa aqueous extract (CTE) is already used as a botanical prescription drug for treating dementia in China. Our previous studies reported that phenylethanoid glycosides of CTE have anti-Alzheimer’s disease (AD) activity by inhibiting amyloid β peptide (Aβ) aggregation and deposition. However, recent studies considered that the phenylethanoid glycosides may be metabolized by intestinal bacteria, because all analysis results showed that the bioavailability of phenylethanoid glycosides is extremely low. In this study we demonstrate how iron chelation plays a crucial role in the Aβ aggregation and deposition inhibition mechanism of phenylethanoid glycosides of CTE. In addition, we further proved phenylethanoid glycosides (1–3) could reach brain. Active CTE component and action mechanism confirmation will be a great help for product quality control and bioavailability studies in the future. At the same time, we provide a new analysis method useful in determining phenylethanoid glycosides (1–3) in plants, foods, blood, and tissues for chemical fingerprint and pharmacokinetic research.


Introduction
Dementia is one of the most common chronic aging diseases. In 2015 the World Alzheimer Report estimated that about 46.8 million people suffered from dementia, and the number is expected to be 74.7 million in 2030 and 131.5 million in 2050 [1]. The global dementia population will increase year by year and become out of control in the future. Global healthcare expenditures to treat dementia were almost 604 billion US dollars in 2010 and the amount is expected to be 1000 billion in 2030 [2]. Dementia patients have greater risk of accidental death [3] and therefore require more medical care. This is a heavy burden for dementia patient familial caregivers. The impact of dementia on caregivers, family and society can produce great physical, psychological, life and economic stress. There are two major forms of dementia, Alzheimer's disease (AD) and vascular dementia [4,5]. AD is the most common dementia and is an irreversible and progressive neurodegenerative disorder [4][5][6]. AD was the sixth leading cause of death in the United States in 2015 [6].
Recent studies support that the neuron toxicity induced by amyloid β peptide (Aβ) (plaques) and protein tau (tangles) aggregation are closely related to AD pathogenesis. The accumulation of Aβ

Analysis of AD-Like Rat Brain Tissues
After 14 days of CTE oral administration (200 mg/kg/day), AD-like rat brain tissues were collected and divided into two parts, hippocampus and striatum, and were both homogenized [19]. Because the hippocampus and striatum homogenates from each rat were too scant to be loaded into solid phase extraction and to avoid excessive experimental errors, the hippocampus and striatum homogenates from four individual rats were mixed into an analyzed sample for UPLC/MS/MS using the solid phase extraction method ( Figure 3A,B). The phenylethanoid glycosides (1-3) were observed in the hippocampus ( Figure 3A) and striatum ( Figure 3B). There was no significant peak in blank brain tissues ( Figure 3C). In the hippocampus the echinacoside content was 11.97 ± 0.34 ng/mL, acteoside content was 1.25 ± 0.16 ng/mL, and isoacteoside content was 1.38 ± 0.08 ng/mL. In the striatum the echinacoside content was 22.60 ± 1.69 ng/mL, acteoside content was 2.03 ± 0.61 ng/mL, and isoacteoside content was 4.90 ± 0.64 ng/mL. As shown in Figure 4, the CTE would pass through the blood brain barrier according to the significant detectable amounts of 1-3 in the hippocampus and striatum. The content of 1 is much higher than 2 and 3 in CTE by UPLC analysis ( Figure S2). Therefore, it is reasonable to observe that 1 is the highest in brain tissue after oral CTE administration.

Analysis of AD-Like Rat Brain Tissues
After 14 days of CTE oral administration (200 mg/kg/day), AD-like rat brain tissues were collected and divided into two parts, hippocampus and striatum, and were both homogenized [19]. Because the hippocampus and striatum homogenates from each rat were too scant to be loaded into solid phase extraction and to avoid excessive experimental errors, the hippocampus and striatum homogenates from four individual rats were mixed into an analyzed sample for UPLC/MS/MS using the solid phase extraction method ( Figure 3A,B). The phenylethanoid glycosides (1-3) were observed in the hippocampus ( Figure 3A) and striatum ( Figure 3B). There was no significant peak in blank brain tissues ( Figure 3C). In the hippocampus the echinacoside content was 11.97 ± 0.34 ng/mL, acteoside content was 1.25 ± 0.16 ng/mL, and isoacteoside content was 1.38 ± 0.08 ng/mL. In the striatum the echinacoside content was 22.60 ± 1.69 ng/mL, acteoside content was 2.03 ± 0.61 ng/mL, and isoacteoside content was 4.90 ± 0.64 ng/mL. As shown in Figure 4, the CTE would pass through the blood brain barrier according to the significant detectable amounts of 1-3 in the hippocampus and striatum. The content of 1 is much higher than 2 and 3 in CTE by UPLC analysis ( Figure S2). Therefore, it is reasonable to observe that 1 is the highest in brain tissue after oral CTE administration.

The Phenylethanoid Glycosides (1-3) Metal Chelating Activity.
In the literature, the determination of the free-from compounds and their metal complex were performed using HPLC [34,35]. Measure of chelation activity were analyzed by UPLC in this study. The 50 μg/mL solutions of each the phenylethanoid glycosides (1-3) were added into a 10 μg/mL solution of copper (Cu), calcium (Ca), magnesium (Mg), zinc (Zn), iron (Fe) or rat serum respectively. Each compound with each metal solution or serum was analyzed by UPLC. The phenylethanoid glycosides (1-3) exhibted metal chelating activity with iron and serum ( Figure 5, Supplementary Materials Figure S1). Conc. (ng/ml) Figure 4. The phenylethanoid glycosides (1-3) were detectable in rat brain tissues [19] using UPLC/MS/MS. The data represented mean ± S.D., n = 3 for each group.

The Phenylethanoid Glycosides (1-3) Metal Chelating Activity.
In the literature, the determination of the free-from compounds and their metal complex were performed using HPLC [34,35]. Measure of chelation activity were analyzed by UPLC in this study. The 50 µg/mL solutions of each the phenylethanoid glycosides (1-3) were added into a 10 µg/mL solution of copper (Cu), calcium (Ca), magnesium (Mg), zinc (Zn), iron (Fe) or rat serum respectively. Each compound with each metal solution or serum was analyzed by UPLC. The phenylethanoid glycosides (1-3) exhibted metal chelating activity with iron and serum ( Figure 5, Supplementary Materials Figure S1).

The Phenylethanoid Glycosides (1-3) Metal Chelating Activity.
In the literature, the determination of the free-from compounds and their metal complex were performed using HPLC [34,35]. Measure of chelation activity were analyzed by UPLC in this study. The 50 μg/mL solutions of each the phenylethanoid glycosides (1-3) were added into a 10 μg/mL solution of copper (Cu), calcium (Ca), magnesium (Mg), zinc (Zn), iron (Fe) or rat serum respectively. Each compound with each metal solution or serum was analyzed by UPLC. The phenylethanoid glycosides (1-3) exhibted metal chelating activity with iron and serum ( Figure 5, Supplementary Materials Figure S1).

Anslysis of Echinacoside (1) in Rat Serum in Vivo
In order to understand the distribuion of echinacoside in rat serum, the serum was collected from the initial time to 720 min after echinacoside (1) oral administration (100 mg/kg) and analyzed by UPLC. Figure 6 showed the concentration versus time profiles of echinacoside in rat serum. copper (Cu), calcium (Ca), magnesium (Mg), zinc (Zn), iron (Fe), or rat serum by UPLC analysis. The data represented mean ± S.D., n = 3. *** p < 0.001 compared with compound alone.

Anslysis of Echinacoside (1) in Rat Serum in Vivo.
In order to understand the distribuion of echinacoside in rat serum, the serum was collected from the initial time to 720 min after echinacoside (1) oral administration (100 mg/kg) and analyzed by UPLC. Figure 6 showed the concentration versus time profiles of echinacoside in rat serum.

Discussion
Dementia, especially AD, is an irreversible aging and chronic disease that leads to a Gordian knot of questions. This insurmountable disease leads to serious economic issues with a huge and increasing financial burden. Worldwide, AD researchers take lots of effort to investigate new anti-AD drugs, but failed in several large clinical trials targeting Aβ in 2012 [9]. Except for directly targeting Aβ, metal ion homeostasis in the brain is considered the key reason related to Aβ aggregation and deposition which leads to AD formation [9,10]. Therefore, metal ions play an essential role in the pathogenesis of AD, and the metal chelation hypothesis has become an important research direction [9,[13][14][15].
According to our previous studies [18,19,24,25], CTE displayed potential anti-dementia activity. Even in human clinical studies, after 1 year of treatment with CTE capsules for moderate AD patients, the Alzheimer's Disease Assessment Scale-cognitive subscale (ADAS-cog) score would not show significant deterioration compared to before treatment [18]. CTE kept the ADAS-cog score of AD patients stable, but other prescription drugs did not. The only choice of prescription drugs licensed for AD treatment is acetylcholinesterase inhibitors (AChEIs) such as donepezil and galanthamine that improve AD symptoms in the short term, but deterioration occurs after 1 year of treatment [36]. Based on clinical studies, CTE has the opportunity to be a potential treatment for anti-AD agent development. Based on our previous reports [24,25], echinacoside (1), acteoside (2), and isoacteoside (3) would protect neurons from damage by Aβ, decrease Aβ oligomerization in vitro, and significantly ameliorate cognitive dysfunction induced by Aβ in vivo. The in vitro study indicated the active doses of 1-3 were up to 50 μg/mL [24,25]. But previous studies considered that the oral bioavailability of 1 and 2 were very low (0.83% for 1 and 0.12% for 2) in rats [29,30]. Compound 1 could not even be identified in human serum after oral echinacea tablet administration [37]. Current studies show that 1-3 had poor membrane permeability and absorption in intestinal cells [38] and most of the 1 would be metabolized in gastrointestinal ducts [39]. Compound 2 content in rat serum was only about 4.5 μg/mL after 15 min of intravenous injection with a dose of 10 mg/kg of acteoside [30]. However, this study shows that 1-3 would be detected in the hippocampus and striatum of rat brain tissues after CTE oral administration (Figures 3 and 4). In addition, Figure 6 showed that

Discussion
Dementia, especially AD, is an irreversible aging and chronic disease that leads to a Gordian knot of questions. This insurmountable disease leads to serious economic issues with a huge and increasing financial burden. Worldwide, AD researchers take lots of effort to investigate new anti-AD drugs, but failed in several large clinical trials targeting Aβ in 2012 [9]. Except for directly targeting Aβ, metal ion homeostasis in the brain is considered the key reason related to Aβ aggregation and deposition which leads to AD formation [9,10]. Therefore, metal ions play an essential role in the pathogenesis of AD, and the metal chelation hypothesis has become an important research direction [9,[13][14][15].
According to our previous studies [18,19,24,25], CTE displayed potential anti-dementia activity. Even in human clinical studies, after 1 year of treatment with CTE capsules for moderate AD patients, the Alzheimer's Disease Assessment Scale-cognitive subscale (ADAS-cog) score would not show significant deterioration compared to before treatment [18]. CTE kept the ADAS-cog score of AD patients stable, but other prescription drugs did not. The only choice of prescription drugs licensed for AD treatment is acetylcholinesterase inhibitors (AChEIs) such as donepezil and galanthamine that improve AD symptoms in the short term, but deterioration occurs after 1 year of treatment [36]. Based on clinical studies, CTE has the opportunity to be a potential treatment for anti-AD agent development. Based on our previous reports [24,25], echinacoside (1), acteoside (2), and isoacteoside (3) would protect neurons from damage by Aβ, decrease Aβ oligomerization in vitro, and significantly ameliorate cognitive dysfunction induced by Aβ in vivo. The in vitro study indicated the active doses of 1-3 were up to 50 µg/mL [24,25]. But previous studies considered that the oral bioavailability of 1 and 2 were very low (0.83% for 1 and 0.12% for 2) in rats [29,30]. Compound 1 could not even be identified in human serum after oral echinacea tablet administration [37]. Current studies show that 1-3 had poor membrane permeability and absorption in intestinal cells [38] and most of the 1 would be metabolized in gastrointestinal ducts [39]. Compound 2 content in rat serum was only about 4.5 µg/mL after 15 min of intravenous injection with a dose of 10 mg/kg of acteoside [30]. However, this study shows that 1-3 would be detected in the hippocampus and striatum of rat brain tissues after CTE oral administration (Figures 3 and 4). In addition, Figure 6 showed that compound 1 content in rat serum increased about 5 times from the 15 min to 720 min after compound 1 oral administration (100 mg/kg). We considered phenylethanoid glycosides of CTE would pass through the blood brain barrier and there would be chelation action between CTE and the metal. Chelation is a reversible chemical reaction. Iron is the essential element in serum and the brain [10,40]. Figure 5 shows that three phenylethanoid glycosides would have metal chelating activity with iron and serum. Iron and serum obviously change the peak retention time area of three phenylethanoid glycosides. Iron chelation may be the crucial reason leading to misidentification of the very low bioavailability of phenylethanoid glycosides. Therefore, the metal chelating activity should be considered in blood serum and brain analysis of the three CTE phenylethanoid glycosides to recover the real bioavailability and pharmacokinetics. In addition, the results indicated three phenylethanoid glycosides (1-3) would pass through the blood brain barrier and arrived at brain tissues through body circulation. This study developed a most rapid and sensitive method for 1-3 analysis using UPLC/MS/MS and provided powerful evidence to prove that phenylethanoid glycosides (1-3) are the major bioactive constituents of CTE.

Isolation and Purification
The dried C. tubulosa stem was ground into powder, and then extracted five times with 75% EtOH. After solvent evaporation under reduced pressure, the crude extract was subjected to Macroporous resin AB-8 column chromatography with H 2 O/EtOH gradient solvent systems from 20% EtOH up to 100% EtOH. According to the thin layer chromatography, four fractions (Fr.1~Fr.4) were collected for further separation. Fr. 2 was subjected to preparative high performance liquid chromatography (HPLC) on a COSMOSIL ® 5C18-AR-II column (250 mm × 20 mm i.d., 5 µm) using 18% acetonitrile as the mobile phase system. The flow rate was 15 mL/min. Three major peaks of interest were selectively collected. The fractions containing the targeted compounds were further condensed to dryness and produced 1, 2, and 3, respectively.

Sample Preparation
Rat brain tissue used 50% methanol for homogenization (1 g of tissue for 5 mL of methanol). The homogenate centrifuged at 10,000 rpm at 4 • C for 10 min. The supernatant (50 µL) with 10 µL IS (100 µg/mL) added pure water to 1 mL. The mixture was carefully loaded onto the Oasis HLB (hydrophilic-lipophilic balanced) cartridge (Waters, Milford, MA, USA), which had already been eluted with 1mL of methanol and then equilibrated with 1 mL of water under vacuum conditions. After 1 mL of 10% methanol was used for washing the cartridge, 1 mL of methanol was used to elute to obtain the analyzed sample which was transferred into another 1.5 mL Eppendorf tube with 10 µL of 1% ascorbic acid solution and dried by evaporation. Finally, the residue from the sample analyzed was dissolved in 200 µL of 50% methanol for analysis.

Preparation of the Calibration Standard Curve
Compounds 1-3 were used to prepare calibration standards respectively added to brain tissue homogenate and hesperidin as internal standard to obtain the final calibration standard concentrations which were 1.0, 5.0, 10.0, 50.0, and 100.0 ng/mL.

Validation
The calibration standard curve was obtained using the linearity test using five different concentrations of 1-3, in the ranges 1.0, 5.0, 10.0, 50.0, and 100.0 ng/mL, respectively, and was regarded as linear with over 0.99 of coefficient of determination (R-squared). The lower limit of detection (LLOQ), defined as ± 20% of the concentration of a test sample, produced a signal peak rather than noise. The credibility and accuracy of the analysis method was evaluated using a recovery test at 10 ng/mL of 1-3.

Metal Chelating Activity Assay
UPLC was used to analyze the metal chelating effect using the 50 µg/mL solution of CTE phenylethanoid glycosides 1-3 containing 10 µg/mL of metal solution of copper, calcium, magnesium, zinc, iron or rat serum. Each compound with each metal solution or serum was analyzed by UPLC. The metal chelating effect would change the retention time for 1-3 in liquid chromatography. The percentage of metal chelating effect was calculated as below and the data represented mean ± S.D. (n = 3).

Conclusions
This study demonstrated that 1-3 are the active components of CTE for anti-AD activity. Iron chelation has become the new concept for designing a new generation of drugs for the treatment of AD [41]. CTE is a potential botanical anti-AD Chinese medicine targeting iron chelation induced Aβ aggregation and deposition. In addition, we further proved that 1-3 would reach the brain though blood-brain barrier (BBB). Active components and underlying mechanism confirmation of CTE will greatly help quality control and the bioavailability of product studies in the future. At the same time, the advanced and efficient analysis method presented will be useful in determining 1-3 in plants, foods, blood, and brain tissues for chemical fingerprint and pharmacokinetic research.  Figure S1. Measure of chelation activity for the phenylethanoid glycosides (1-3) by UPLC analysis. Figure S2. The UPLC chromatogram of CTE. Figure S2. The UPLC chromatogram of CTE. Echinacoside (1), acteoside (2), isoacteoside (3), and tubuloside A (4) were the components of CTE. Figure S3. The 1 H NMR (500 MHz, CDCl 3 ) data of 1. Figure S4. The 13 C NMR (125 MHz, CDCl 3 ) data of 1. Figure S5. The ESI-MS spectrum of 1. Figure S6. The 1 H NMR (500 MHz, DMSO-d6) data of 2. Figure S7. The 13 C NMR (125 MHz, DMSO-d6) data of 2. Figure S8. The ESI-MS spectrum of 2. Figure  S9. The 1 H NMR (500 MHz, DMSO-d6) data of 3. Figure S10. The 13 C NMR (125 MHz, DMSO-d6) data of 3. Figure S11.