Therapeutic Effects and Metabolic Spectrum of Traditional Chinese Medicine Hengqing II Prescription on Alzheimer's Disease

Alzheimer's disease (AD) seriously damages elders' social and daily abilities around the world. Traditional Chinese medicine (TCM), a rich drug resource bank, could help research AD. In order to explore the role of TCM in AD treatment, 86 AD patients were recruited from the hospital, then treated with Hengqing II prescription and donepezil hydrochloride. The cognitive and serum lipid levels were investigated before and after treatment. The patient's urine was collected after three months of treatment. Metabolites in the urine samples were extracted with methanol and detected on the UHPLC-MS platform. Results proved that Hengqing II can improve cognitive levels and reduce the levels of Hcy, D-D, FIB, Apo B, TC, and LDL-C compared with donepezil hydrochloride (P < 0.05). The results of multivariate statistical analysis revealed that the metabolism of HQII was significantly different compared with Control groups. A total of 66 differential metabolites were found in this comparison (50 were down-regulated and 16 were up-regulated). Four amino acid pathways and one linoleic acid pathway were found through these metabolites. After receiver operating characteristic analysis, it was suggested that palmitic acid, palmitoleic acid, linoleic acid, oleic acid, SAH, and methionine can be used as biomarkers for treating AD, while the effects of daidzein, genistein, and naringenin on the treatment of AD need to be further studied.


Introduction
Alzheimer's disease (AD) is a degenerative disease in the central nervous system and is mainly characterized by cognitive decline [1][2][3]. It has become a major public concern due to symptoms of AD being difficult to diagnose at the early stage [4]. It is estimated that by 2050, the number of patients with AD in the world will exceed 130 million [5].
us, it is urgent to strengthen the research on the diagnosis and treatment of AD. However, the pathogenesis of AD is not fully explained and the development of drugs for treating AD is still full of challenges. At present, the discovery of new biomarkers is of great help in the treatment of AD. us, many researchers and resources are invested in the research on biomarkers of AD every year [6][7][8]. We noticed that numerous studies focus on biomarkers of AD patients' blood and brain, while there are relatively few studies researching biomarkers related to AD patients' urine [9,10]. e changes in metabolites in the urine of AD patients aroused our interest. Whether new biomarkers can be found in AD patients' urine and whether there are significant differences between metabolites in urine and metabolites in blood and brain need to be confirmed.
Traditional Chinese medicine (TCM) is one of the most influential alternative medicines in the world. Faced with the challenges of treating AD, TCM has gained importance in the prevention and treatment of AD, with a bright future [11][12][13][14]. To date, there is no theoretical basis for using TCM prescription to treat AD. According to our clinical practice, TCM Hengqing II prescription has satisfactory effects on the treatment of AD. However, due to many plant components and complex chemical composition in Hengqing II, we do not know, which component plays a major role in the treatment of AD. Meanwhile, as a common method for diagnosing AD, scale tests will be affected by the subjective of AD patients, thus resulting in inaccurate data. erefore, studying the metabolic changes of AD patients after taking Hengqing II can help clarify the main components of Hengqing II and provide objective evidence for the diagnosis of AD as well. To date, metabolite detection has become more efficient and reliable [15,16], and it is a powerful approach for studying AD pathogenic processes and biomarkers. erefore, to gain a deeper understanding of AD pathology and provide a data basis for TCM diagnosis of AD, in this study, the effects of Hengqing II on the metabolic spectrum were explored by UHPLC-MS, and reliable biomarkers were discovered.

Recruitment and Grouping of AD Patients and Urine
Collection. From December 2016 to December 2018, 86 patients suffering from AD were treated in the Shanghai Jiao Tong University Affiliated Sixth People's Hospital, and they were recruited as the research population. is study was approved by the hospital ethics committee (Approval No. 2019-12). e AD patients had obvious mental symptoms and cognitive impairment and met the diagnostic criteria of AD in the Chinese Classification and Diagnostic Criteria of Mental Disorders (3rd Edition). According to TCM syndrome differentiation, 58 patients were in the KDPS group and 28 patients were in the non-KDPS group. Urine samples were collected to detect metabolic differences between the two groups. After that, the KDPS group was equally divided into two groups by the random number table method. One group took Hengqing II formula once a day, no more than five times a week (HQII group), and another group took donepezil hydrochloride tablets (5 mg/tablet) simultaneously (Control group). Hengqing II consists of 11 kinds of TCM (Table 1). e herbs were extracted using boiling water for 2 h. e extract was filtered, evaporated on a rotary vacuum evaporator, and lyophilized using a freeze-dryer. Most AD patients recruited were accompanied by other diseases such as hypertension and diabetes. erefore, in order to eliminate such interference in the treatment process, relevant patients were provided corresponding drug treatment. Urine samples were collected from the HQII group and Control group after three-month treatment. Before detection, urine samples were stored in an environment of minus 80°C.

Scale Tests of AD Patients Taking Hengqing II and Control
Group. e scale tests of AD patients after taking Hengqing II and donepezil hydrochloride tablets were measured before therapy, 1 month, and 3 months of therapy. Montreal Cognitive Assessment (MoCA) was used to evaluate the cognitive impairment of patients. e scale has 30 points. If the score is higher, the mental state will be better. e cognitive function of patients was evaluated by Alzheimer's Disease Assessment Scale-Cognitive Subscale (ADAS-Cog), with a score range of 0-75 points. If the ADAS-Cog score is higher, the cognitive ability of patients will be worse. e Activities of Daily Living (ADL) scale was used to evaluate the ability of daily living of patients. e score range was 14-64 points. If the ADL score is higher, the patients' ability of daily living will be worse. e Personal and Social Performance (PSP) scale was used to evaluate the behavior ability of patients. 71-100 points represent slight difficulty, 31-70 points indicate ability defect, and 0-30 points denote low function and need support or monitoring. Cohen Mansfield Aging Inventory (CMAI) was used to evaluate the frequency of patients' aggressive behavior. If the CMAI score is higher, the behavior will be more aggressive.

Detection of Serum Biochemical Indexes in AD Patients
Taking Hengqing II Formula and Control Group. Venous blood (8 mL) was taken from HQII and Control groups (fasting 12 h) before therapy, 1 month, and 3 months of therapy, and serum was separated. e contents of homocysteine (Hcy), apolipoprotein A1 (Apo A1), apolipoprotein B (Apo B), total cholesterol (TC), low-density lipoprotein cholesterol (LDL-C), and high-density lipoprotein cholesterol (HDL-C) were measured by HITACHI7600 automatic biochemical analyzer. e levels of plasma D-D and fibrinogen (FIB) were detected by the turbidimetric inhibition immune method with SYSMEX CS5100 System.

Metabolites Were Extracted from Urine Samples.
e samples were thawed slowly on ice. After that, 100 μL urine per sample was transferred into a 1.5 ml centrifuge tube and 300 μL extract (pure methanol) was added. With vortex for 30 s and ultrasonic treatment at 4°C for 30 min, the centrifuge tubes stood at −20°C for 1 hour and centrifuged at 15000 rpm for 10 min. Subsequently, 200 μL supernatant was transferred into the injection vial and a 5 μL internal standard (1 mg/ml 2-chlorophenylalanine) was added for UHPLC-MS detection. e supernatant was mixed in all tubes and took 500 μL and 12.5 μL internal standards were used as quality control (QC) samples. All samples were stored in an environment of 4°C for no more than 12 h waiting for on-board detection.

Operating Parameters of Ultra-High Performance Liquid
Chromatograph-Mass Spectrometer. Ultra-high performance liquid chromatography (AB SCIEX LC-30A) equipped with an XBridge amide column (4.6 mm × 250 mm, 3.5 μm) was used to separate materials in urine, and the injection volume was 4 μL. e composition of liquid phase A was pure water, with 10 mM ammonium formate and 0.1% (v/v) formic acid. Liquid phase B is chromatographic grade acetonitrile. During elution, the flow rate of the mobile phase was 0.8 mL/ min and the column temperature was 40°C. e gradient changes of mobile phase B were 95% (0.5 minutes), 95%-65% (0.5-10 minutes), 65%-40% (10-14 minutes), 40% (14-16 minutes), 40%-95% (16-16.1 minutes), and 95% (16.1-20 minutes). e mass spectrometer (AB SCIEX TripleTOF5600+) equipped with electron spray ionization (ESI) was used to ionize molecules. e voltage of ESI was 4200 V and the temperature was 350°C. Nitrogen was used as sheath gas and auxiliary gas of ESI, and the flow rates were 35 Arb and 15 Arb, respectively. All data were collected using a data-dependent acquisition (DDA) model with m/z ranging from 100 to 1000. e isolation span was 3 Da and the activation time was 30 ms. e high-resolution Fourier transform approach was used. MS1's resolution was 60000, whereas MS2's resolution was 15000.

UHPLC-MS Results
Processing. e format obtained from the UHPLC-MS platform was a raw data file. e file was converted to ABF format and MS-DIAL software was used to extract peaks, contrast peaks, filter peaks, fill gaps, and identify materials. e mass threshold of peak extraction was 5 ppm, the retention time threshold was 0.2 min, and the signal-to-noise ratio threshold was 3. e compounds were identified using a data-dependent acquisition pattern with 60 matching factor score thresholds. After processing, a table containing metabolite name, molecular weight, m/z, retention time, and peak area was obtained. According to the material information on the PubChem and Human Metabolome Database (HMDB) websites, another manual identification was conducted to ensure reliable metabolites. PCA, OPLS-DA analysis, and permutation tests were performed on the peak area using SMICA-P 14.0 software to analyze the metabolic differences between the two groups [17]. PCA projects multidimensional data to two or three dimensions through variance, so that it can quickly observe whether there are differences between groups [18,19]. Compared with PCA, OPLS-DA includes orthogonal conversion, which filters out the data irrelevant to the model classification [20]. A p-value less than 0.05 and a VIP value greater than 1 were used as the conditions for screening differential metabolites. Metaboanalyst website was used for pathway analysis and ROC analysis, and SPSS 13.0 statistical software was the tool for data processing.

Effects of Taking Hengqing II on the Scale Tests of AD
Patients. After taking Hengqing II and donepezil hydrochloride tablets, the MoCA and PSP scores of the HQII group after 3 months improved more significantly than those of the Control group after 3 months. e ADAS-Cog, ADL, and CMAI scores of the HQII group after 3 months were obviously lower than those of the Control group after 3 months (Table 2), which showed that Hengqing II was better than donepezil hydrochloride in treating AD after taking at the same time.

Effects of Taking Hengqing II on Serum Biochemical
Indexes in AD Patients. After treatment with Hengqing II and donepezil hydrochloride, the contents of Hcy, D-D, FIB, Apo B, TC, and LDL-C in the HQII group after 3 months were lower than in the Control group after 3 months of treatment. e levels of Apo A1 and HDL-C in the HQII group after 3 months of treatment were significantly higher than those in the Control group (Table 3). ese results revealed that Hengqing II can reduce the levels of Hcy, D-D, FIB, Apo B, TC, and LDL-C in AD patients and the effects were better than donepezil hydrochloride tablets. Combined with the previous results of Hengqing II in improving the cognitive ability of AD patients, it was speculated that the Hengqing II formula can regulate the metabolites of AD patients, especially lipids. erefore, a follow-up metabolomics difference study was conducted.

PCA and OPLS-DA Analysis of Metabolites' Data.
It can be observed that the KDPS group was significantly different from the non-KDPS group, and the HQII group was significantly different from the Control group (Figures 1(a) and 1(b)). In Figures 1(c) and 1(d), the HQII and KDPS samples could be distinguished from the Control and non-KDPS individuals. It was also found that the separation effects of KDPS on non-KDPS are not as good as that of the HQII group and Control group in the center of a confidence interval, which showed the metabolic differences between HQII and Control groups were greater than the previous comparison. In permutation test plots (Figures 1(e) and 1(f )), the OPLS-DA models exhibited a well fit, which demonstrated that the urine metabolomics signature differs ese metabolomics profiles of KDPS and non-KDPS suggested that AD categorization based on TCM syndrome distinction was dependable and practicable. e significant differences between the HQII group and Control group also confirmed the effective impacts of HQII on the scale tests and serum indicators. Subsequently, we focused on analyzing the effects of Hengqing II on urine metabolites in AD patients.

Differential Metabolites between the HQII and Control
Group. After observing significant differences in metabolic profiles between the HQII group and the Control group, all metabolites found were screened. e p value of the student's test and VIP value in OPLS-DA analysis were used as screening conditions. VIP is variably important in the projection. If the VIP value is greater, the contribution of the variable to the differences between the two groups will be greater. When the p value was less than 0.05 and VIP was greater than 1, there were 66 differential metabolites in the HQII group compared with the Control group (Table 4). Among these different substances, amino acids, long-chain fatty acids, aromatic compounds, purine derivatives, sugar derivatives, pyrimidine derivatives, and small molecular organic acids were found. In addition, three flavonoids appeared in differential metabolites. Flavonoids may come from plant components in the Hengqing II prescription, which requires our in-depth analysis.
Heatmap was used to visualize the relative contents of metabolites in urine samples ( Figure 2). It can be clearly observed that a large number of metabolites in the HQII group (left side of the figure) showed blue down-regulation compared with the Control group (right side of the figure).
at meant that the relative contents of almost differential materials of AD patients can be reversed after taking Hengqing II. In order to understand the change range of metabolites more intuitively, variance analysis on these metabolites was conducted, which showed that the Hengqing II treatment significantly reversed the relative abundance of 50 urinary metabolites ( Figure 3). It was that Hengqing II prescription may regulate AD by a variety of metabolites. e results of the heatmap and histogram could explain the enhancement of cognitive capacities in AD patients after treatment.

ROC Analysis of Differential Metabolites.
To determine whether these differential metabolites can be used as biomarkers for Hengqing II against AD and to exclude false-positive results, ROC analysis (receiver operating characteristic curve) on differential metabolites were performed. In ROC analysis, AUC (area under curve) value was used as a standard to measure the reliability of metabolites. AUC equals 1, which indicates that the prediction is perfect, AUC between 0.85 and 0.95 implies that the prediction effect is very well, 0.7-0.85 means that the prediction effect is general, 0.5-0.7 refers to a poor effect, 0.5 stands for that the model is meaningless, and less than 0.5 indicates that the prediction is worse than random. Among the differential metabolites we found, palmitic acid had the highest AUC value, and the AUC values of 14 differential metabolites were between 0.85   Figure 4). It was also noted that the AUC values of fatty acids were generally high and they may be biomarkers.

Pathway Analysis of Differential
Metabolites. In order to analyze the relationship between these differential metabolites and the physiological processes, Metaboanalyst 5.0 website was used to locate these differential metabolites on the metabolic pathway. In the HQII and Control groups, the  Evidence-Based Complementary and Alternative Medicine pathways for linoleic acid metabolism, arginine biosynthesis, alanine, aspartate, glutamate metabolism, D-glutamine and D-glutamate metabolism, glycine, serine, and threonine metabolism were discovered ( Figure 5).

Discussion
Firstly, according to the obvious differences in metabolic profile between KDPS and non-KDPS group, it is feasible and meaningful to classify AD through KDPS. At present, there are few studies on the TCM syndrome differentiation of dementia. Liu studied the clinical manifestation of Kangxin capsule in the therapy of senile vascular dementia with kidney deficiency and blood stasis in 2007 [21]. Tan revealed the therapeutic biomarkers of processed Aconitum Carmichaeli Debx in treating hydrocortisone-induced Kidney-Yang deficiency syndrome rats [22]. As we all know, this paper was a study on the TCM classification of AD. ese results of the multivariate statistical analysis provided a data basis for the therapy of AD by TCM. However, this study did not analyze the changes in metabolites in AD patients after TCM typing, which is worthy of further research.
In the analysis of the Hengqing II group and the Control group, it was found that Hengqing II significantly improved the clinical manifestation of patients. All indexes in the scale tests have been improved. Meanwhile, Hengqing II also  decreased the contents of Hcy, D-D, FIB, Apo B, TC, and LDL-C in serum, which showed the importance of lipid changes in the pathogenesis of AD. According to the papers consulted, fatty acids play a complex role in the pathogenesis of AD, and they are closely related to intestinal microbiota and neuroinflammatory response [23]. In the metabolite   Evidence-Based Complementary and Alternative Medicine analysis, palmitic acid, palmitoleic acid, oleic acid, and linoleic acid were significantly down-regulated and their VIP value was higher than other metabolites, which indicated that they contributed more differences. In the analysis of the metabolic pathway, the linoleic acid metabolism pathway was found and the impact value was much higher than other pathways. In ROC analysis, the AUC values of palmitic acid, palmitoleic acid, linoleic acid, and oleic acid performed very well. e AUC of palmitic acid even reached a maximum. us, palmitic acid, palmitoleic acid, linoleic acid, and oleic acid are regarded as biomarkers for curing AD. Many studies on the correlation between fatty acids and AD also accord with the abovementioned views. Palmitic acid can induce neuroinflammatory, and result in neuronal damage [24][25][26]. However, as an unsaturated fatty acid, linoleic acid has a different performance from palmitic acid in the nervous system. Linoleic acid has protective effects on oxidative stress, neuroinflammation, and memory damage [27,28]. e significant down-regulation of linoleic acid and palmitic acid in this study may be closely related to the taking of Hengqing II. erefore, we believe that fatty acids are of great significance in the pathogenesis of AD.
Many amino acids and their derivatives in the screening of differential metabolites and four amino acid metabolic pathways were also found. ese results confirmed that amino acids and their derivatives have a great impact on the   Evidence-Based Complementary and Alternative Medicine 9 pathology of AD patients. ere are many studies on the effects of branched-chain amino acids (BCAA) on AD, and there is a certain correlation with lipids [29,30]. In our study, leucine was also remarkably down-regulated, which was consistent with the abovementioned studies. Unfortunately, the AUC value of leucine was not enough. In addition, it was also found that SAH (S-Adenosyl-Lhomocysteine) and methionine were significantly downregulated after treatment, which was highly correlated with AD [31][32][33]. We also noted that the amino acid with the highest AUC was the dimer of alanine. Lin held the view that higher D-alanine levels are associated with more behavioral symptoms of AD [34]. Combined with ROC analysis, it was suggested that methionine and SAH can be used as biomarkers for treating AD. ree kinds of flavonoids were found in differential metabolites, and they are likely to come from a plant component in Hengqing II. It is well known that flavonoids can resist oxidation and improve cell activity. Many studies have proved that flavonoids have antioxidant and anti-inflammatory effects on AD [35][36][37]. After taking Hengqing II, the relative contents of daidzein, genistein, and naringenin increased significantly, which may help patients reverse the pathology of AD. Ko confirmed that 6,7,4′-Trihydroxyisoflavone (a major metabolite of daidzein) could enhance learning and memory by the cholinergic system and activate the p-CREB/BDNF signaling pathway in mice [38]. Daidzein can also promote hippocampal neurogenesis in middle-aged mice [39]. On the other hand, genistein can alleviate cognitive impairment and correct the behavior of rats by exerting antioxidant and anti-inflammatory effects [40,41]. Similarly, naringenin can improve the learning and memory ability of AD rats [42,43]. However, since the AUC value of naringenin was between 0.5 and 0.7, the reliability of the prediction was low. e AUC values of daidzein and genistein were between 0.7 and 0.8, which are general for the prediction. erefore, whether daidzein, genistein, and naringenin can be used as biomarkers for treating AD, still needs more in-depth and detailed studies. In a word, after taking Hengqing II, the anti-inflammatory and anti-oxidation effects of flavonoids alleviate neuron damage, thus improving the cognitive ability of AD patients. With the recovery of cognitive function, the content of AD biomarkers (palmitic acid, methionine, and SAH) decreased. Of course, these explanations still require more experimental data to confirm.

Conclusion
e metabolic profiles of HQII and Control groups differed considerably. 66 distinct metabolites were discovered (50 were down-regulated and 16 were up-regulated). After being treated with Hengqing II prescription, some amino and fatty acids were decreased significantly. ree kinds of flavonoids were significantly up-regulated. After ROC analysis, palmitic acid, palmitoleic acid, linoleic acid, oleic acid, methionine, and SAH can be used as biomarkers for treating AD by Hengqing II. ese metabolites are conducive for us to understand the disease process of AD and pave the way to successfully treating the condition.

Data Availability
e data used to support the findings of this study are included in the article.

Conflicts of Interest
e authors declare that there are no conflicts of interest.

Authors' Contributions
Shengxi Meng investigated the study, was responsible for formal analysis, wrote original draft, and reviewed and edited the manuscript. Shaopeng Li proposed the methodology and was resposible for data curation. Huize Chen proposed the methodology and was responsible for data curation. Chujun Deng investigated the study and was responsible for data curation. Zeyu Meng proposed the methodology. Yimo Wang proposed the methodology. Shengxi Meng, Shaopeng Li, Huize Chen Chujun Deng, and Zeyu Meng contributed equally to this work.