Abstract

   Neuroinflammation is a significant driving force in the pathogenesis
   and progression of central nervous system (CNS) disorders.
   Polymethoxylated flavones (PMFs), the key lipid-soluble constituents in
   Citri Reticulatae Pericarpium (CRP), exhibit excellent blood–brain
   barrier permeability and anti-inflammatory properties, holding
   therapeutic potential for CNS disorders. However, the specific
   bioactive components and therapeutic effects of PMFs in treating CNS
   disorders are not well understood. This study employed a comprehensive
   sequential metabolism approach to elucidate the dynamic
   biotransformation of PMFs in vivo and identified seven brain-targeting
   components. Subsequently, network pharmacology and experimental
   validation were utilized to explore the potential mechanisms of PMFs.
   The results suggested that PMFs have potential therapeutic value for
   Alzheimer’s disease (AD)-like mice, with the inhibition of
   neuroinflammation likely being a key mechanism of their anti-AD
   effects. Notably, sinensetin, tangeretin, nobiletin, and
   3,5,6,7,8,3′,4′-heptamethoxyflavone were identified as potent
   neuroinflammatory inhibitors. This research elucidated the chemical and
   therapeutic foundations of PMFs, indicating their potential as
   treatments or nutritional supplements for AD prevention and treatment.
   Moreover, the integrated triad approach of sequential metabolism,
   network pharmacology, and experimental validation may serve as a
   promising strategy for screening bioactive compounds in herbs or
   functional foods, as well as for elucidating their therapeutic
   mechanisms.

   Keywords: Citri Reticulatae Pericarpium, polymethoxylated flavones,
   anti-neuroinflammation, Alzheimer’s disease, sequential metabolism
   method, network pharmacology

1. Introduction

   Disorders of the central nervous system (CNS), such as Alzheimer’s
   disease (AD), Parkinson’s disease (PD), multiple sclerosis (MS),
   cerebral ischemic diseases, and brain cancer, are major causes of
   disability and rank as a high risk factor of death worldwide
   [[48]1,[49]2]. Despite many therapeutic agents being developed, there
   is still a shortage of effective therapies for these diseases. An
   increasing number of recent studies indicate that the inflammatory
   response leads to neuronal death through multiple pathways and serves
   as a significant driving force in CNS disorders [[50]3]. In particular,
   for neurodegenerative diseases such as AD, PD, and MS, inflammation
   within the CNS is considered to be one of the pathogenic mechanisms of
   these conditions. As the disease progresses, persistent deposition of
   β-amyloid protein (Aβ) and dopaminergic deficits, microglial cells, and
   astrocytes become activated in CNS, triggering the production of
   various neuroinflammatory-related signaling factors, including
   cytokines, chemokines, and cell adhesion molecules [[51]4,[52]5]. With
   the amplification of the inflammatory cascade, brain homeostasis and
   the blood–brain barrier (BBB) are disrupted, further exacerbating
   disease progression [[53]6,[54]7]. Therefore, alleviating
   neuroinflammation is one of the important therapeutic strategies for
   ameliorating neurological diseases.

   Herbs and plant-derived natural compounds have been crucial sources for
   drug discovery and development. They are characterized by their
   multi-pathway and multi-target therapeutic effects, as well as having
   high safety profiles suitable for long-term consumption [[55]8]. Citri
   Reticulatae Pericarpium (CRP), the dried and aged fruit peel of Citrus
   reticulata Blanco and its cultivated forms, exhibits rich
   pharmacological properties, such as oxidative stress resistance,
   inflammation inhibition, anti-asthmatic effects, and neuroprotection
   [[56]9]. For centuries, it has been widely used as a condiment, herbal
   tea, and therapeutic agent in Eastern and Southeastern Asia. As a
   condiment, CRP is often braised or fried at high temperatures. As an
   herbal tea, it is brewed in hot water. As a therapeutic agent, CRP is
   usually taken after decoction. These processing methods optimize the
   release of liposoluble components, promoting their absorption by the
   body. Polymethoxylated flavones (PMFs), the key liposoluble
   constituents in CRP, exhibit diverse biological activities
   ([57]Supplementary Table S1). Notably, their excellent BBB permeability
   and potent anti-inflammatory properties [[58]10,[59]11,[60]12,[61]13]
   suggest potential therapeutic benefits for the management of CNS
   disorders. In particular, as part of the daily diet, CRP is regularly
   consumed through tea and culinary practices in many regions, leading to
   daily and prolonged intake of PMFs. This consistent intake may confer
   potential benefits for the prevention, early intervention, and ongoing
   management of CNS diseases. However, the ameliorative effects of PMFs
   from CRP on CNS diseases, along with their pharmacological
   constituents, have yet to be thoroughly investigated.

   PMFs in CRP exhibit multi-component characteristics, leading to complex
   pharmacokinetic and pharmacodynamic profiles [[62]14]. Consequently, it
   is crucial to investigate the in vivo chemical and metabolic profiles
   of PMFs. Additionally, assessing the BBB permeability of candidate
   agents is an important criterion in the development of CNS medications
   or nutrients. However, this criterion is often overlooked in many
   studies [[63]15,[64]16]. Previous studies have not provided a complete
   understanding of the dynamic biotransformation process of PMFs in vivo,
   and the BBB permeability has been underexamined [[65]17,[66]18]. In
   this study, we employed a comprehensive sequential metabolism approach
   utilizing intestinal perfusion with venous sampling (IPVS) [[67]19], an
   in situ closed-loop technique [[68]20], and UPLC-Q Exactive-Orbitrap
   HRMS technology as its core analytical platform. This method simulated
   the dynamic transformation of orally administered multi-component
   substances within the body, detailing the process of their absorption
   and metabolism through the gastrointestinal tract and liver, followed
   by entry into the systemic circulation and further transport to the
   diseased site. This integrated strategy delineated the dynamic
   biotransformation process of PMFs and precisely screened the potential
   pharmacological ingredients for treating CNS disorders.

   Network pharmacology leverages leveraging high-throughput omics data
   analysis, computer simulation, and network database search methods to
   reveal the network relationships of drug–gene and drug–target
   interactions with disease. Network pharmacology is characterized by its
   holistic, systematic, and comprehensive nature, which highly aligns
   with the multi-component, multi-target, and holistic features of
   Traditional Chinese Medicine. Currently, network pharmacology has been
   widely applied in the prediction of the molecular mechanisms,
   therapeutic efficacy, and adverse reactions of Chinese herbal medicines
   [[69]21,[70]22]. During this study, network pharmacology was employed
   to predict the pharmacological and molecular mechanisms of PMFs for
   treating CNS diseases.

   In summary, we established a comprehensive sequential metabolism
   approach to elucidate the dynamic biotransformation of PMFs in CRP and
   to further analyze the components that enter the brain. Subsequently,
   network pharmacology was employed to predict the potential
   pharmacological and molecular mechanisms of PMFs, which were then
   validated in animal and cell models. Finally, an in vitro screening
   assay was conducted to screen neuroinflammatory inhibitors from PMFs.
   This study was designed to provide a detailed chemical and metabolic
   profile, elucidating the potential bioactive ingredients of PMFs for
   the amelioration of CNS diseases. Additionally, it sought to delineate
   the role of PMFs in mitigating CNS disorders via anti-inflammatory
   pathways and to pinpoint potential neuroinflammatory inhibitors,
   providing a preliminary research foundation for the development of PMFs
   as therapeutic agents or dietary supplements. Furthermore, this
   integrated strategy could effectively be extended to elucidate the
   efficacy of herbs or functional foods and to screen for active
   components that ameliorate CNS disorders. The flowchart of our study is
   illustrated in [71]Figure 1.

Figure 1.

   [72]Figure 1
   [73]Open in a new tab

   Flowchart of the study design.

2. Results

2.1. Chemical and Metabolic Profiles of PMFs in CRP

   A comprehensive sequential metabolism method was utilized to mimic the
   biotransformation in vivo of multiple components following oral
   administration in this research. The components in samples were
   detected and identified using UPLC-HRMS, including the PMFs extract,
   the incubation solution of PMFs with simulated gastric juice, plasma
   samples from the mesenteric vein (for intestinal wall/flora
   metabolism), femoral vein (for hepatic metabolism), and abdominal
   aorta, as well as cerebrospinal fluid (CSF) and brain tissue samples. A
   total of 18 prototype components were identified in the extract of PMFs
   from CRP ([74]Table 1, and [75]Supplementary Figure S1A)
   [[76]23,[77]24,[78]25,[79]26]. Nobiletin (NOB, peak 9) was selected as
   the representative component to elucidate the characteristic
   fragmentation patterns of PMFs ([80]Supplementary Figure S2).
   Furthermore, based on the identification results of the PMFs extract,
   the components of various biological samples were further analyzed. All
   18 components were detected in simulated gastric juice (SGJ) group and
   the mesenteric blood (MB) from intestinal wall/flora metabolism groups.
   The femoral vein blood (FVB) group yielded seventeen components, the
   abdominal aorta (AA) group had fifteen, the brain tissue (BT) group
   contained fourteen, and the CSF group showed seven ([81]Table 1,
   [82]Supplementary Figure S1B–H). To deeply investigate the
   biotransformation process of PMFs in vivo, we used NOB as a model to
   conduct the metabolite analysis and deduce the metabolic pathways
   [[83]27,[84]28]. As shown in [85]Table 2, a total of 12 metabolites
   were identified: seven components in the MB from intestinal wall
   metabolism group, five in the MB from intestinal flora metabolism
   group, six in the FVB group, three in the AA group, and two in the BT
   group.

Table 1.

   Identification of prototypical compounds from polymethoxylated flavones
   (PMFs) in Citri Reticulatae Pericarpium (CRP) and the absorbed
   components in different biological samples.
   No. t[R] /min Observed Mass Error
   /ppm MS/MS Fragments Molecular
   Formula Proposed Compounds PMFs SGJ MB-W MB-F FVB AA CSF BT Ref.
   1 12.20 [M + H]^+
   375.1434 −1.065 211.0604, 196.0369, 168.0415, 150.0314 C[20]H[22]O[7]
   5,6,7,3′,4′-pentamethoxyflavanone Y Y Y Y Y N N N [[86]23]
   2 13.07 [M + H]^+
   373.1282 0.136 358.1048, 343.0814, 315.0865, 181.0132, 153.0183
   C[20]H[20]O[7] Isosinensetin Y Y Y Y Y Y Y Y [[87]24]
   3 13.72 [M + H]^+
   403.1389 0.106 388.1147, 373.0921, 327.0861 C[21]H[22]O[8]
   5,7,8,3′,4′,5′-hexamethoxyflavone Y Y Y Y Y Y N Y [[88]23]
   4 13.79 [M + H]^+
   389.1234 0.658 374.1000, 359.0761, 328.0958, 313.0702 C[20]H[20]O[8]
   3′(4′)-monohydroxy-5,6,7,8,4′(5,6,7,8,3′)-pentamethoxyflavone Y Y Y Y Y
   Y N Y [[89]23]
   5 14.17 [M + H]^+
   373.1281 −0.266 358.1042, 343.0814, 312.0993, 329.1021, 357.0967
   C[20]H[20]O[7] Sinensetin * Y Y Y Y Y Y Y Y *
   6 14.30 [M + H]^+
   375.1441 0.641 211.0604, 196.0369 C[20]H[22]O[7]
   5,7,8,3′,4′-Pentametho-xyflavanone Y Y Y Y Y Y N Y [[90]23]
   7 14.51 [M + H]^+
   389.1233 0.427 374.1002, 359.0761, 341.0656, 316.0576 C[20]H[20]O[8]
   Monohydroxy-pentamethoxyflavone Y Y Y Y Y Y N N [[91]25]
   8 14.78 [M + H]^+
   403.1387 0.016 388.1135, 373.0919, 327.0866 C[21]H[22]O[8]
   5,6,7,3′,4′,5′-hexamethoxyflavone Y Y Y Y Y Y Y Y [[92]23]
   9 15.09 [M + H]^+
   403.1387 −0.074 388.1151, 373.0919, 355.0811, 327.0863 C[21]H[22]O[8]
   Nobiletin * Y Y Y Y Y Y Y Y *
   10 15.43 [M + H]^+
   343.1176 −0.014 328.0937, 313.0706, 285.0758, 181.0133, 153.0184,
   133.0649 C[19]H[18]O[6] 5,6,7,4′-tetramethoxyflavone Y Y Y Y Y Y Y Y
   [[93]23]
   11 15.67 [M + H]^+
   433.1494 0.234 418.1245, 403.1024, 388.0774, 360.0845, 385.0923,
   345.0604 C[22]H[24]O[9] 3,5,6,7,8,3′,4′-heptemethoxyflavone * Y Y Y Y Y
   Y Y Y *
   12 15.70 [M + H]^+
   389.1231 −0.036 374.1004, 359.0761, 356.0887, 341.0656, 197.0883,
   169.0132 C[20]H[20]O[8] Monohydroxy-pentamethoxyflavone Y Y Y Y Y Y N Y
   [[94]25]
   13 16.28 [M + H]^+
   373.1282 0.136 358.1043, 343.0813, 325.0714, 300.0622, 297.0757
   C[20]H[20]O[7] Tangeretin * Y Y Y Y Y Y Y Y *
   14 16.38 [M + H]^+
   419.1330 −1.572 404.1095, 389.0865, 371.0760, 361.0908 C[21]H[22]O[9]
   5-monohydroxy-6,7,8,3′,4′,5′-hexamethoxyflavone Y Y Y Y N N N N
   [[95]23]
   15 16.55 [M + H]^+
   359.1124 −0.277 344.0887, 329.0660, 311.0545, 301.0695, 163.0755,
   138.9975 C[19]H[18]O[7] Monohydroxy-tetramethoxyflavone Y Y Y Y Y Y N Y
   [[96]25]
   16 17.03 [M + H]^+
   403.1388 0.076 388.1156, 373.0919, 327.0872 C[21]H[22]O[8]
   Hexamethoxyflavone Y Y Y Y Y Y N Y [[97]23]
   17 17.41 [M + H]^+
   389.1231 −0.036 374. 1000, 359. 0761,
   341. 0655, 331. 0811,
   197. 0082 C[20]H[20]O[8] 5-hydroxy-6,7,8,3′,4′-pentamethoxyflavone Y Y
   Y Y Y Y N Y [[98]26]
   18 17.92 [M + H]^+
   419.1338 0.313 404.1100, 389.0868, 371.0758 C[21]H[22]O[9]
   Monohydroxy-hexamethoxyflavone Y Y Y Y Y N N N [[99]25]
   [100]Open in a new tab

   Note: t[R]: retention time; *: compounds identified by reference
   standards; Y: Detected; N: Not Detected; PMFs: PMFs Extract; SGJ:
   simulated gastric juice; MB-W: mesenteric blood from intestinal wall
   metabolism group; MB-F: mesenteric blood from intestinal flora
   metabolism group; FVB: femoral venous blood; AA: abdominal aorta; CSF:
   cerebrospinal fluid; BT: brain tissue.

Table 2.

   Identification of metabolites of nobiletin in different biological
   samples.
   No. t[R]
   /min Observed Mass Error
   /ppm Molecular
   Formula MS/MS Fragments SGJ MB-
   W MB-
   F FVB AA CSF BT Metabolites Structure of
   Prediction
   M1 8.78 [M + H]^+ 595.1661 0.661 C[27]H[30]O[15] 419.1342, 389.0872,
   404.1107, 371.0768, 595.1672, 403.1023, 346.0687, 165.0548 N Y N Y N N
   N Hydroxylation + Glucuronidation C[15]H[12]O[2]+6OCH[3]+OH+GluA-12H
   M2 9.44 [M + H]^+ 595.1663 0.964 C[27]H[30]O[15] 419.1342, 389.0870,
   404.1104, 371.0764, 346.0688, 343.0444, 374.0628, 403.1026 N Y N N N N
   N Hydroxylation + Glucuronidation C[15]H[12]O[2]+6OCH[3]+OH+GluA-12H
   M3 9.54 [M + H]^+ 537.1243 0.835 C[24]H[24]O[14] 211.0604, 361.1285,
   196.0369, 183.0291, 360.1165, 167.0339, 330.0693, 177.0547 N N Y N N N
   N Demethylation + Glucuronidation C[15]H[10]O[2]+3OCH[3]+3OH+GluA-8H
   M4 9.68 [M + H]^+ 565.1556 0.704 C[26]H[28]O[14] 389.1234, 359.0763,
   257.081, 565.1561, 85.0291, 565.372, 390.127 N N Y Y N N N
   Demethylation + Glucuronidation C[15]H[10]O[2]+5OCH[3]+OH+GluA-8H
   M5 10.92 [M + H]^+ 551.1401 1.104 C[25]H[26]O[14] 375.108, 345.0609,
   342.0739, 360.0839, 376.1107, 359.0764, 85.0291, 270.0887 N Y N N N N N
   Demethylation + Glucuronidation C[15]H[10]O[2]+4OCH[3]+2OH+GluA-8H
   M6 10.98 [M + H]^+ 565.1559 1.341 C[26]H[28]O[14] 389.1236, 331.0816,
   356.0894, 313.0708,
   374.0999, 390.12732, 359.0766, 328.0945 N Y N N N N N Demethylation +
   Glucuronidation C[15]H[10]O[2]+5OCH[3]+OH+GluA-8H
   M7 12.30 [M + H]^+ 389.1234 0.658 C[20]H[20]O[8] 389.1237, 359.0766,
   169.0134, 331.0817, 341.066, 390.1272, 316.0580, 374.1003 N Y Y Y Y N N
   Demethylation + Hydroxylation C[15]H[10]O[2]+5OCH[3]+OH-6H
   M8 15.88 [M + H]^+ 389.1231 −0.113 C[20]H[20]O[8] 389.1234, 331.0814,
   356.0893, 313.0709, 359.0765, 341.0660, 328.0945, 390.1264 N Y Y N Y N
   Y Demethylation + Hydroxylation C[15]H[10]O[2]+5OCH[3]+OH-6H
   M9 15.91 [M + H]^+ 375.1074 0.336 C[19]H[18]O[8] 375.1077, 345.0607,
   356.0892, 317.0657, 313.0707, 331.0813, 360.0830, 374.0966 N N Y N N N
   N Demethylation C[15]H[10]O[2]+4OCH[3]+2OH-6H
   M10 17.38 [M + H]^+ 389.1231 −0.113 C[20]H[20]O[8] 389.1235, 359.0764,
   341.0661, 197.0083, 169.0134, 163.0756, 390.1267, 316.0581 N N N Y Y N
   Y Demethylation + Hydroxylation C[15]H[10]O[2]+5OCH[3]+OH-6H
   M11 17.50 [M − H]^−
   467.0661 3.943 C[20]H[20]O[11]S 387.1093, 372.0857, 357.0623, 342.0385,
   467.0659, 327.0153, 299.0203, 388.1125 N N N Y N N N Demethylation +
   Sulfation C[15]H[10]O[2]+5OCH[3]+OH+SO[3]-6H
   M12 20.78 [M − H]^− 453.0505 4.065 C[19]H[18]O[11]S 373.0933, 343.0463,
   358.0699, 328.023, 374.0968, 453.0513, 300.0279, 359.0729 N Y N Y N N N
   Demethylation + Sulfation C[15]H[10]O[2]+4OCH[3]+2OH+SO[3]-6H
   [101]Open in a new tab

   Note: t[R]: retention time; Y: Detected; N: Not Detected; SGJ:
   simulated gastric juice; MB-W: mesenteric blood from intestinal wall
   metabolism group; MB-F: mesenteric blood from intestinal flora
   metabolism group; FVB: femoral venous blood; AA: abdominal aorta; CSF:
   cerebrospinal fluid; BT: brain tissue.

   Overall, this study provided a detailed chemical and metabolic
   profiles, elucidating the potential bioactive ingredients of PMFs for
   the amelioration of CNS diseases.

2.2. Network Pharmacology Analysis

   To conduct precise research on the relationship between PMFs and CNS
   disorders, seven PMFs identified in both brain tissue and CSF were used
   for network pharmacology analysis ([102]Table 3). After removing
   repeated targets, 148 active targets of PMFs and 20,997 CNS
   disorder-related targets were collected. After processing by Venn
   diagram analysis, 148 targets were found to be common to both PMFs and
   CNS disease ([103]Figure 2A). Cytoscape 3.8.2 software was used to
   construct a “components-targets” interaction network for describing the
   correspondence between PMFs and disease targets ([104]Figure 2B). Based
   on the overlapping targets, Gene Ontology (GO) enrichment analysis was
   utilized to annotate the functions. The findings showed that numerous
   GO terms were enriched, and [105]Figure 2C shows the top 10 significant
   terms in the biological processes (BP), molecular functions (MF), and
   cellular components (CC) categories. To delve deeper into the
   importance of the key targets, a Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathway enrichment analysis was performed. [106]Figure
   2D illustrates the top 30 pathways that were enriched by the KEGG
   pathway enrichment analysis. Based on the pathway enrichment analysis
   results, a network was constructed that included brain-targeting
   components, key targets, and significant pathways, with the objective
   of gaining a comprehensive understanding of the action mechanisms
   ([107]Figure 3). The enriched pathways suggested that PMFs may
   intervene in CNS disorders through multiple pathways, including
   metabolism (such as metabolic pathways, nitrogen metabolism, steroid
   hormone biosynthesis, ovarian steroidogenesis), infection and immunity
   (such as human papillomavirus infection, Kaposi sarcoma-associated
   herpesvirus infection, human cytomegalovirus infection), and signal
   transduction (PI3K-Akt, MAPK, Rap1 signaling pathway, FoxO signaling
   pathway, phospholipase D signaling pathway, calcium signaling pathway,
   neuroactive ligand-receptor interaction). Moreover, the results
   revealed that neurodegenerative diseases and AD were highly enriched in
   the major KEGG pathways, suggesting that PMFs might possess potential
   activity for the prevention and treatment of neurodegenerative
   diseases, particularly for AD.

Table 3.

   Chemical structures of PMFs used for network pharmacology analysis.
   Molecule Name Structure
   Isosinensetin graphic file with name molecules-30-00771-i001.jpg
   Sinensetin graphic file with name molecules-30-00771-i002.jpg
   5,6,7,3′,4′,5′-hexamethoxyflavone graphic file with name
   molecules-30-00771-i003.jpg
   Nobiletin graphic file with name molecules-30-00771-i004.jpg
   5,6,7,4′-tetramethoxyflavone graphic file with name
   molecules-30-00771-i005.jpg
   3,5,6,7,8,3′,4′-heptemethoxyflavone graphic file with name
   molecules-30-00771-i006.jpg
   Tangeretin graphic file with name molecules-30-00771-i007.jpg
   [108]Open in a new tab

Figure 2.

   [109]Figure 2
   [110]Open in a new tab

   Predicted pathways of ploymethoxylated flavones (PMFs) in treating
   central nervous system (CNS) disorders. (A) Venn diagram of overlapping
   targets. (B) Interaction network of “Components-targets” interaction
   network. (C) GO enrichment analysis (BP, CC, MF). (D) KEGG enrichment
   analysis.

Figure 3.

   [111]Figure 3
   [112]Open in a new tab

   PMFs-targets–pathways interaction network. Blue circles represent PMFs
   identified in both brain tissue and cerebrospinal fluid (CSF). Green
   rectangles denote major overlapping targets. Purple arrowhead shapes
   indicate the top 30 enriched pathways. The red square represents one of
   the enriched pathways: Alzheimer’s disease.

2.3. Effects of PMFs on the AD-like Mice

2.3.1. Behavioral Tests

   The experimental protocol is detailed in [113]Figure 1. The cognitive
   function of mice was assessed using the Y-maze spontaneous alternation
   test and the novel arm exploration test [[114]29,[115]30]. Mice in the
   sham group displayed robust physical activity, high spontaneous
   alternation rates, and frequent novel arm entries, indicative of strong
   spatial memory and exploratory abilities. In contrast, the model group
   exhibited significantly lower spontaneous alternation rates compared to
   the sham group (p < 0.001, [116]Figure 4A,C). The high-dose PMFs group
   and the positive drug control group, however, exhibited a significant
   improvement in spontaneous alternation rates compared to the model
   group (p < 0.001, and p < 0.05, respectively). During the novel arm
   exploration test, the model group exhibited a significant reduction in
   percentage of entries into the novel arm compared to the sham group (p
   < 0.05, [117]Figure 4B,D). Both the middle- and high-dose PMFs group,
   as well as the positive drug group, showed a significant increase in
   the percentage of entries into novel arm compared to the model group (p
   < 0.05, p < 0.05, and p < 0.01, respectively). Additionally, the model
   group exhibited a significant decrease in percentage of time spent in
   the novel arm (p < 0.05, [118]Figure 4B,E), i.e., the effect that was
   reversed by the middle- and high-dose PMFs groups, as well as the
   positive drug control group (p < 0.05, p < 0.05, and p < 0.001,
   respectively).

Figure 4.

   [119]Figure 4
   [120]Open in a new tab

   Y-maze test. Representative movement trajectory maps for each group of
   mice in spontaneous alternation test (A) and novel arm exploration test
   (B). (C) Spontaneous alternation rate (%). (D) Percentage of entries
   into the novel arm (%). (E) Percentage of cumulative time spent in the
   novel arm (%). ns represents no significant difference. # p < 0.05, ###
   p < 0.001, compared with the sham group; * p < 0.05, ** p < 0.01, *** p
   < 0.001, compared with the model group.

   The Open Field test was used to assess exploratory, locomotor, and
   anxiety-like behaviors in mice [[121]31]. Mice in the sham group
   displayed robust physical activity, frequent entries into the central
   area, and a prolonged cumulative duration there, indicative of strong
   exploratory behavior and a healthy mental state. In contrast, the model
   group showed significant reductions in the total distance moved, the
   frequency of entries into the central area, and the time spent
   exploring the central area compared to the sham group (p < 0.05, p <
   0.01, and p < 0.05, respectively) ([122]Figure 5). The high-dose PMFs
   and positive drug groups showed significantly improved behavioral
   abnormalities in AD-like mice, as [123]Figure 5 shows.

Figure 5.

   [124]Figure 5
   [125]Open in a new tab

   Open field test. (A) Representative movement trajectory maps. (B) Total
   distance moved (cm). (C) Frequency of entries into the center area. (D)
   Cumulative time exploring in the center area (s). ns represents no
   significant difference. # p < 0.05, ## p < 0.01, compared with the sham
   group; * p < 0.05, ** p < 0.01, compared with the model group.

   In conclusion, the intervention administration with PMFs and donepezil
   could ameliorate the abilities of learning, memory, mental state and
   physical activity of Aβ[1–42] induced AD-like mice.

2.3.2. Histological Evaluation

   The accumulation of Aβ leads to neuronal damage or death, ultimately
   causing cognitive dysfunction [[126]32]. To assess histological
   alterations in hippocampal neurons, we utilized hematoxylin and eosin
   (HE) staining. In the sham group, neurons in the hippocampal CA3 region
   appeared neatly arranged with normal morphology, and no significant
   nuclear pyknosis was observed ([127]Figure 6). In contrast, the model
   group displayed pronounced nuclear pyknosis, disarrayed arrangement,
   and enlarged neuronal gap. Treatment with PMFs and donepezil, however,
   mitigated these histological abnormalities, restoring the normal
   structure and arrangement of neurons in the hippocampal CA3 region.

Figure 6.

   [128]Figure 6
   [129]Open in a new tab

   The morphological changes of the neural cells in the hippocampal CA3
   region. Representative images of H&E staining of brain tissue from the
   different treatment groups (scale bar: 200 μm and 50 μm). The green
   arrows represent nuclear pyknosis, and the red ones represent enlarged
   neuronal gap.

2.3.3. Aβ and P-Tau Measurement

   The Aβ deposition and Tau hyper-phosphorylation are the core
   pathological hallmark of AD [[130]33]. This study evaluated changes in
   Aβ and p-Tau levels in brain tissue using the ELISA method. The levels
   of Aβ and p-Tau in the model group were significantly elevated compared
   to those in the sham group (p < 0.001 and p < 0.01, respectively)
   ([131]Figure 7A,B). Both the middle- and high-dose PMFs groups, as well
   as the positive drug group, exhibited a significant reduction in Aβ
   levels compared to the model group (p < 0.05, p < 0.01, and p < 0.05,
   respectively). Similarly, the middle-dose PMFs and positive drug groups
   significantly decreased p-Tau levels versus the model group (both p <
   0.05).

Figure 7.

   [132]Figure 7
   [133]Open in a new tab

   The levels of Aβ, p-Tau, and PSD-95 in the brain tissue from the
   different treatment groups. (A) Aβ, (B) p-Tau, (C) PSD-95. ns
   represents no significant difference. ## p < 0.01, ### p < 0.001,
   compared with the sham group; * p < 0.05, ** p < 0.01, *** p < 0.001,
   compared with the model group.

2.3.4. Postsynaptic Density Protein 95 (PSD-95) Measurement

   Synaptic plasticity impairment and synaptic loss are central events in
   early cognitive dysfunction, with PSD-95 playing a critical role in
   synaptic function. In the brains of AD patients, the accumulation of Aβ
   is closely associated with a reduction in PSD-95 levels [[134]34]. This
   study evaluated changes in PSD-95 levels in brain tissue using the
   ELISA method. The level of PSD-95 in the model group were significantly
   lower than this in the sham group (p < 0.001) ([135]Figure 7C). Both
   the low- and high-dose PMFs groups exhibited a significant increase in
   PSD-95 levels compared to the model group (p < 0.05 and p < 0.001,
   respectively).

2.3.5. Inflammatory Response in Brain Tissue

   Neuroinflammation has been recognized as a pivotal factor in the
   pathogenesis of AD. Microglia, macrophage-like innate immune cells in
   the CNS, play a crucial role in neuroinflammation [[136]35].
   Immunohistochemistry was employed to assess the quantity and morphology
   of microglia in the brain. The results illustrated that in the sham
   group, microglia predominantly displayed a resting morphology with
   small cell bodies and multiple branches ([137]Figure 8A,B). In
   contrast, the model group featured activated microglia with enlarged
   cell bodies and swollen short branches. Compared to the model group,
   PMF treatment significantly reversed microglial activation in the
   hippocampal and cortical regions, while donepezil had a less pronounced
   effect. Additionally, RT-qPCR was employed to assess the expression of
   inflammatory factors. The results showed that the relative mRNA levels
   of the pro-inflammatory factors TNF-α and IL-1β in the model group
   significantly increased compared to those in the sham group (both, p <
   0.001) ([138]Figure 8C,D). PMFs effectively mitigated these changes,
   reducing the expression of TNF-α and IL-1β. Meanwhile, the gene
   expression of anti-inflammatory factors IL-10 and TGF-β in the model
   group was significantly lower compared to the sham group (p < 0.001 and
   p < 0.05, respectively) ([139]Figure 8E,F). Treatment with low- and
   high-dose PMFs significantly increased IL-10 expression compared to the
   model group (both, p < 0.05), with all PMF doses reversing the decrease
   in TGF-β expression (both, p < 0.001). By contrast, donepezil exhibited
   a weaker inhibitory effect, only significantly decreasing the
   expression of IL-1 (p < 0.01) and increasing the expression of TGF-β (p
   < 0.01).

Figure 8.

   [140]Figure 8
   [141]Open in a new tab

   The inflammatory response in the brain. Representative images of
   immunohistochemical staining for IBA-1 in hippocampal region (scale
   bar: 100 μm and 50 μm) (A) and cortical region (scale bar: 50 μm and 20
   μm). The red arrows represent activated microglia with enlarged cell
   bodies and swollen short branches. (B). Expression levels of TNF-α (C),
   IL-1β (D), IL-10 (E), and TGF-β (F) in the brains from the different
   treatment groups. ns represents no significant difference. # p < 0.05,
   ### p < 0.001, compared with the sham group; * p < 0.05, ** p < 0.01,
   *** p < 0.001, compared with the model group.

   In conclusion, PMFs exerted an anti-inflammatory effect in AD-like mice
   by mitigating microglial activation, downregulating the expression of
   pro-inflammatory cytokines (TNF-α and IL-1β) and upregulating the
   expression of anti-inflammatory cytokines (IL-10 and TGF-β).

2.3.6. BBB Functionality Assessment

   BBB dysfunction and neuroinflammation interact synergistically,
   creating a vicious cycle that accelerates the progression of AD
   [[142]36]. Therefore, we further evaluated the functionality of BBB.
   RT-qPCR assay showed that the relative mRNA level of matrix
   metalloproteinase-9 (MMP-9) in the model group was significantly higher
   than that in the sham group (p < 0.001) ([143]Figure 9A). All PMF
   intervention groups exhibited a significant reduction of MMP-9 levels
   compared to the model group (p < 0.01, p < 0.001, and p < 0.001,
   respectively). Furthermore, we examined the levels of tight junction
   protein closely related to BBB integrity. The results showed that the
   levels of Zonula Occludens-1 (ZO-1) and Claudin-5 in the model group
   were significantly reduced compared to those in the sham group (both, p
   < 0.05) ([144]Figure 9B,C). All groups treated with PMFs showed a
   significant elevation in ZO-1 levels compared to the model group (p <
   0.01, p < 0.001, and p < 0.001, respectively). Moreover, the high-dose
   PMF intervention group significantly reversed the decrease in Claudin-5
   levels observed in the model group (p < 0.001).

Figure 9.

   [145]Figure 9
   [146]Open in a new tab

   BBB Functionality Assessment. The levels of MMP-9 (A), ZO-1 (B), and
   Claudin-5 (C) in the brain tissue from the different treatment groups.
   ns represents no significant difference. # p < 0.05, ### p < 0.001,
   compared with the sham group; ** p < 0.01, *** p < 0.001, compared with
   the model group.

   In conclusion, PMFs might preserve BBB integrity in AD-like mice by
   reducing the expression of MMP-9 and increasing the expression of tight
   junction proteins (ZO-1 and Claudin-5).

2.4. Anti-Inflammatory Effect of PMF-Containing CSF

   The CSF containing PMFs was prepared as described in [147]Section
   4.4.2. To evaluate the potential cytotoxicity of samples, we
   investigated the effects of different concentrations of PMF-containing
   or blank CSF on cells viability. As [148]Supplementary Figure S3A,B
   shows, PMFs-containing or blank CSF at concentrations ranging from 10%
   to 50% did not cause significant cytotoxicity. However, cell viability
   marginally decreased at 40% and 50% blank CSF. Therefore, to ensure the
   accuracy of the results, concentrations of 10%, 20%, and 30% were
   ultimately chosen for subsequent experiments.

   Subsequently, we investigated the effect of PMF-containing CSF on the
   LPS/IFN-γ-induced inflammation in BV2 cells. First, the cell morphology
   was observed under a microscope. As [149]Figure 10A shows, BV2 cells in
   the control group, under resting conditions, exhibited small cell
   bodies with a round shape (suspended state) or an epithelial-like
   morphology with 2–3 pseudopodia (adherent state). In contrast, after
   stimulation with LPS/IFN-γ, the cells transitioned from a resting to an
   activated state, characterized by an enlarged cell body, an increase in
   the proportion of adherent cells, a decrease in pseudopodia or the
   formation of filopodia, and the manifestation of an amoeboid
   morphology. However, PMF-containing CSF effectively mitigated these
   changes, reducing the proportion of activated cells. Subsequently, we
   measured intracellular ROS levels, which were significantly increased
   by LPS/IFN-γ stimulation ([150]Figure 10B). The increase in ROS levels
   was attenuated by the PMF-containing CSF, restoring them to normal
   levels. Furthermore, the assay results indicated that the NO content
   was significantly higher in the LPS/IFN-γ group compared to the control
   group (p < 0.001) ([151]Figure 10C). Treatment with 20% and 30%
   PMF-containing CSF significantly reduced NO levels compared to both the
   LPS/IFN-γ group and the corresponding blank CSF group (p < 0.001).
   Notably, 20% and 30% blank CSF also demonstrated an inhibitory effect
   on NO production induced by LPS/IFN-γ (p < 0.01). In addition, the
   levels of pro-inflammatory cytokines TNF-α, IL-1β, and IL-6 were
   measured to evaluate the anti-inflammatory effects. The ELISA results
   showed that the content of TNF-α, IL-1β, and IL-6 in BV2 cells were
   significantly increased in response to LPS/IFN-γ stimulation (p <
   0.001). Compared to the LPS/IFN-γ group and the corresponding blank CSF
   group, treatment with 20% and 30% PMF-containing CSF significantly
   reversed the increased levels of TNF-α, IL-1β, and IL-6 ([152]Figure
   10D–F). As with NO, the groups treated with 20% and 30% blank CSF also
   significantly attenuated the LPS/IFN-γ-induced increase in IL-1β
   production (p < 0.05). Overall, the detection results indicated that
   PMFs effectively mitigate the inflammatory response induced by
   LPS/IFN-γ in BV2 microglial cells. Notably, the blank CSF also
   demonstrated an anti-inflammatory trend, in line with the findings that
   healthy CSF exerted a protective effect on the brain [[153]37].

Figure 10.

   [154]Figure 10
   [155]Open in a new tab

   Effects of PMF-containing cerebrospinal fluid on the inflammation in
   LPS/IFN-γ-induced BV2 cells. (A) Morphology of BV2 cells under various
   treatment condition (scale bar: 50 μm). The red arrows represent
   activated BV2 microglia cells. (B) Intracellular ROS levels in
   different treatment groups (scale bar: 200 μm). The levels of NO (C),
   TNF-α (D), IL-l β (E), and IL-6 (F) in the different treatment groups.
   ns represents no significant difference. ### p < 0.001, compared with
   the control group; * p < 0.05, ** p < 0.01, *** p < 0.001, compared
   with the LPS/IFN-γ-treated group; + p < 0.05, ++ p < 0.01, +++ p <
   0.001, compared with the LPS/IFN-γ+blank cerebrospinal fluid treated
   group.

2.5. Screening for Neuroinflammation Inhibitors from PMFs

   To screen for potent bioactive monomers, we further investigated the
   efficacy of four monomers that were identified in both brain tissue and
   CSF: sinensetin (SIN), tangeretin (TAN), nobiletin (NOB), and
   3,5,6,7,8,3′,4′-heptamethoxyflavone (HMF). First, we tested the effects
   of different concentrations of monomers on cell viability. The results
   indicated that, at concentrations of 20 μM or lower, all monomers had
   no significant effect on the viability of BV2 cells. Consequently,
   concentrations of 5, 10, and 20 μM were chosen for further experimental
   investigations. Subsequently, the cells were observed under an optical
   microscope to assess the changes in morphology after treatment with
   respective groups. Four monomers effectively inhibited the activation
   of microglia induced by LPS/IFN-γ ([156]Figure 11A). We proceeded to
   examine their potential capability for scavenging ROS. Four monomers
   reduced the levels of intracellular ROS compared to the LPS/IFN-γ group
   ([157]Figure 11B). Moreover, the indicators of inflammation, including
   NO, TNF-α, IL-1β, and IL-6, were also investigated. Compared with the
   LPS/IFN-γ group, four monomers reduced the content of NO, and the
   inhibitory effect was dose-dependent in the group of NOB and HMF
   ([158]Figure 11C). By contrast, TAN exhibited a weaker inhibitory
   effect, displaying no significant impact at a concentration of 5 μM.
   Additionally, four monomers inhibited the production of TNF-α in a
   dose-dependent manner ([159]Figure 11D). Similarly, four monomers
   caused the inhibition of the LPS/IFN-γ-induced increase in IL-1β and
   IL-6 production ([160]Figure 11E,F). In conclusion, both SIN, TAN, NOB,
   and HMF showed potential anti-neuroinflammatory activity.

Figure 11.

   [161]Figure 11
   [162]Open in a new tab

   Effects of four monomers on the inflammatory response in
   LPS/IFN-γ-induced BV2 cells. (A) Morphology of BV2 cells under various
   treatment condition (scale bar: 50 μm). The red arrows represent
   activated BV2 microglia cells. (B) Intracellular ROS levels in
   different treatment groups (scale bar: 200 μm). The levels of NO (C),
   TNF-α (D), IL-l β (E), and IL-6 (F) in the different treatment groups.
   ns represents no significant difference. ### p < 0.001, compared with
   the control group. * p < 0.05, ** p < 0.01, *** p < 0.001, compared
   with the LPS/IFN-γ-treated group.

3. Discussion

   Neuroinflammation has been recognized as a significant factor
   contributing to CNS diseases, exacerbating multiple pathological
   processes through inflammatory cascades, such as deposition of Aβ,
   dopaminergic deficits, disruption of brain microenvironment
   homeostasis, and BBB leakiness [[163]4,[164]5]. PMFs, the key active
   components in CRP, possess excellent anti-inflammatory activity and
   superior BBB permeability [[165]10,[166]11,[167]12,[168]13], showing
   potential for the treatment of CNS disorders. However, the active
   components and therapeutic effects of PMFs in treating CNS diseases
   remain unclear.

   In this study, a comprehensive sequential metabolism approach was
   utilized to identify the active components of PMFs. Unlike traditional
   studies of metabolic characteristics that focus only on isolated
   fragments of the overall metabolic process [[169]17,[170]18], we
   developed a holistic, sequential metabolites approach. This method
   detailed the continuous biotransformation of PMFs from oral
   administration through major metabolic sites, systemic circulation, and
   into the brain. The results indicated that 15 of the 18 prototype
   components in PMFs extract successfully traversed the stomach,
   intestinal wall, intestinal flora, and liver, reaching systemic
   circulation. This suggested that PMFs possess chemically stable
   properties, allowing most components to enter the body in their
   original forms. Furthermore, among the fifteen prototypical components
   that entered systemic circulation, fourteen permeated brain tissue and
   seven entered the CSF, demonstrating their potent ability to permeate
   the BBB. Additionally, the analysis results of NOB metabolites
   indicated that the intestinal wall, intestinal flora, and the liver are
   the primary metabolic sites for PMFs. The structural modifications of
   the metabolites indicated that the predominant metabolic pathways for
   PMFs involved demethylation, glucuronidation, hydroxylation, and
   sulfation [[171]27,[172]28]. Only the metabolites that underwent
   demethylation and hydroxylation entered the systemic circulation (M7,
   M8, M10), and among them, M8 and M10 were further transported to the
   brain tissue. Nevertheless, due to the structural homology of PMFs
   identified in CRP, a phenomenon of metabolite overlap was observed.
   Consequently, the detailed characterization of individual PMF
   metabolites requires further research. In summary, the findings
   successfully delineated the chemical profile and the map of the dynamic
   biotransformation of PMFs in CRP and effectively narrowed down the
   range of potential bioactive components in PMFs for the treatment of
   brain disorders.

   A network pharmacology approach was utilized to investigate the
   pharmacological and molecular mechanisms underlying the treatment of
   CNS disorders based on the potential active compounds in PMFs that
   penetrate the BBB. The results suggested that PMFs might ameliorate CNS
   diseases through multiple pathways, involving metabolism, infection and
   immunity, and signal transduction. Key pathways like PI3K-Akt, MAPK,
   FoxO, and nitrogen metabolism are linked to inflammation. It has been
   shown that activated microglia in AD could induce PI3K-AKT pathway
   activation, promoting NF-κB nuclear translocation. As a pivotal
   transcription factor, NF-κB upregulates the expression of
   pro-inflammatory mediators, including cytokines (e.g., TNF-α and IL-1β)
   and enzymes (e.g., COX-1, COX-2, and iNOS) [[173]38]. These mediators
   amplify neuroinflammation, ultimately contributing to
   neurodegeneration, neuronal damage, and cognitive decline [[174]39].
   Moreover, preclinical evidence suggests that PI3K inhibitors may
   attenuate neuroinflammatory responses [[175]40]. The mitogen-activated
   protein kinase (MAPK) signaling pathway plays a pivotal role in
   modulating inflammatory responses during brain aging, serving as a
   critical regulator of cellular processes, such as survival, apoptosis,
   and inflammatory reactions [[176]41]. Key members of the MAPK family,
   particularly p38 and extracellular signal-regulated kinases 1/2
   (ERK1/2), are strongly implicated in the regulation of inflammatory
   mechanisms [[177]42]. Upon activation, these kinases initiate a
   phosphorylation cascade that modulates pro-inflammatory transcription
   factors, thereby triggering and amplifying the inflammatory response
   cascade [[178]43]. Additionally, the targets of FoxOs include
   pro-inflammatory molecules like TLR2, TLR4, IL-1β, and TNF-α [[179]44].
   Moreover, NO, a key product of nitrogen metabolism, is a
   pro-inflammatory mediator in AD [[180]45]. Overall, the prediction
   results suggested that PMFs in CRP could potentially exert anti-CNS
   disorders effects by inhibiting neuroinflammation. Interestingly, the
   pathways associated with AD were significantly enriched within the
   primary KEGG pathways, suggesting that PMFs could be promising
   candidates for anti-AD treatment. AD, a prevalent neurodegenerative
   disorder, progressively impairs cognitive function, memory, and
   reasoning, ultimately resulting in the loss of self-care skills
   [[181]46]. Current studies indicate that inflammation is a pivotal
   pathological feature driving the pathogenesis and progression of AD
   [[182]47]. Inhibiting the inflammatory response may provide a viable
   approach for the prevention, early intervention, and ongoing management
   of AD [[183]48]. In subsequent experiments, we investigated the
   potential of PMFs to ameliorate AD by inhibiting neuroinflammation.

   The intracerebroventricular (ICV) injection of Aβ[1–42] to induce
   AD-like mice is a commonly used animal model in AD research, exhibiting
   behavioral and pathological features typical of AD [[184]49,[185]50].
   This study revealed that PMFs ameliorated the core symptoms of AD-like
   mice, including memory decline, anxiety-like states, neural cell damage
   in hippocampal region, and increased levels of Aβ and p-Tau in the
   brain. Moreover, the assessment of synaptic function (PSD-95) and
   BBB-related markers (MMP-9, ZO-1, and Claudin-5) revealed that PMFs
   exert neuroprotection and improve BBB integrity in AD-like mice.
   Notably, we focused on the neuroinflammatory inhibitory effects of
   PMFs, paying particular attention to microglia. As the resident immune
   cells of the brain, microglia play a pivotal role in regulating
   inflammatory responses. A moderate activation of microglia and the
   inflammatory response are essential for neuroprotection. However,
   persistent and excessive inflammation can lead to a dysregulation of
   the balance between anti-inflammatory and pro-inflammatory effects,
   triggering and driving the progression of neurological diseases
   [[186]4]. The results of animal study revealed that PMFs effectively
   suppressed microglial activation and regulated the balance of
   inflammatory factors, thereby alleviating brain inflammation. To
   further evaluate the anti-neuroinflammatory properties of PMFs, we
   utilized the LPS/IFN-γ-induced BV2 microglial cell model in subsequent
   experiments, a widely recognized system for assessing
   anti-neuroinflammatory efficacy [[187]51]. To ensure experimental
   rigor, we investigated the anti-inflammatory efficacy of PMFs by
   analyzing CSF containing PMFs, rather than the extracts alone. Cellular
   morphology and a series of recognized inflammation-related indicators
   were utilized to assess the anti-neuroinflammatory potential. We
   validated the anti-inflammatory efficacy of PMF-containing CSF,
   indicating that PMFs in CRP not only successfully entered the CSF but
   also exerted anti-inflammatory effects. In summary, the in vivo and in
   vitro experimental data supported the therapeutic potential of PMFs for
   AD, with neuroinflammation inhibition likely playing a pivotal role in
   their anti-AD mechanism.

   Based on the established efficacy of anti-AD and the mechanism of
   anti-neuroinflammation, further studies were conducted to screen for
   neuroinflammatory inhibitors among the components of PMFs. We evaluated
   the anti-inflammatory effects of SIN, TAN, NOB, and HMF, which were
   simultaneously identified in CSF and brain tissue. The experimental
   results of the neuroinflammation cell model revealed that SIN, TAN,
   NOB, and HMF exhibited satisfactory anti-neuroinflammatory properties.
   NOB inhibits neuroinflammation by preventing NF-κB activation, reducing
   the expression of NO, IL-6, and TNF-α [[188]52]. TAN exhibits
   anti-neuroinflammatory activity by suppressing Akt phosphorylation,
   decreasing mRNA expression of NO, TNF-α, IL-6, and IL-1β [[189]53]. The
   results of the current study aligned with the previously mentioned
   findings. By contrast, it is rare for the evaluation of
   anti-neuroinflammatory properties of SIN and HMF through the microglia
   to be reported. Consequently, our investigation not only offers
   critical evidence supporting the anti-neuroinflammatory capabilities of
   SIN and HMF but also lays a solid groundwork for subsequent inquiries
   into their potential therapeutic roles in ameliorating CNS disorders.

   In summary, these findings provide preliminary evidence supporting the
   potential of PMFs in CRP as a dietary supplement or therapeutic agent
   for the prevention and treatment of AD. However, further clinical
   studies are essential to validate their efficacy and practical
   application. Additionally, this study employed an integrated approach
   combining sequential metabolism, bioinformatics, and experimental
   validation, successfully elucidating the role of PMFs in ameliorating
   CNS disorders through anti-neuroinflammation, as well as identifying
   potential active components. Nevertheless, the underlying molecular
   mechanisms and the in vivo pharmacological efficacy of these active
   components require further in-depth investigation.

4. Materials and Methods

4.1. Materials

   CRP was procured from Beijing Tong Ren Tang Co., Ltd. (Beijing, China)
   and was authenticated by Professor Jingjuan Wang at Beijing University
   of Chinese Medicine (Beijing, China). MS-grade acetonitrile (≥99.9%
   purity), MS-grade methanol (≥99.9% purity), and formic acid (≥99.9%
   purity) were obtained from Fisher Scientific (Fairlawn, NJ, USA).
   Absolute ethanol of high purity (≥99.9%) was supplied by Tianjin Damao
   Chemical Reagent Factory (Tianjin, China). The compounds NOB (CAS:
   478-01-3, ≥98% purity), SIN (CAS: 2306-27-6, ≥98% purity), and TAN
   (CAS: 481-53-8, ≥98% purity) were acquired from Beijing Beiterenkang
   Biomedical Technology Co., Ltd. (Beijing, China). HMF (CAS: 1178-24-1,
   ≥98% purity) was sourced from Nanjing Dilger Medical Technology Co.,
   Ltd. (Nanjing, China). β-Amyloid (1–42) (≥99.66% purity) was acquired
   from MedChemExpress LLC (Monmouth Junction, NJ, USA).
   Lipopolysaccharide (Potency ≥ 500,000 EU/mg) was acquired from Beijing
   Lanboleide Biotechnology Co., Ltd. (Beijing, China). Recombinant Murine
   IFN-gamma (≥98% purity) was obtained from PeproTech, Inc. (Cranbury,
   NJ, USA). The IBA-1 antibody (Concentration: 700 μg/mL) was purchased
   from Proteintech Group, Inc. (Chicago, IL, USA). The horseradish
   peroxidase (HRP)-labeled goat anti-rabbit (Cat No. 5220-0336) was
   purchased from SeraCare Life Sciences (Milford, MA, USA). BV2
   microglial cells (STR identification. Cat No. CL-0493A) and their
   corresponding culture medium (Cat No. CM-0493A) were provided by Wuhan
   Pricella Biotechnology Co., Ltd. (Wuhan, China). A digestive solution
   (Cat No. C100C1) containing 0.25% pancreatin and 0.01% EDTA, as well as
   the Cell Counting Kit-8 (Cat No. C6005), were purchased from NCM
   Biotech Co., Ltd. (Suzhou, China). The NO assay kit (Cat No. S0021S)
   was procured from Shanghai Biyuntian Biotechnology Co., Ltd. (Shanghai,
   China). The ELISA Kits for TNF-α (Cat No. SEA133Mu)/IL-1β (Cat No.
   SEA563Mu)/IL-6 (Cat No. SEA079Mu) were sourced from Wuhan Cloud-Clone
   Biotechnology Co., Ltd. (Wuhan, China). The ELISA Kits for PSD-95 (Cat
   No. F9447-B) was sourced from Shanghai Kexing Trading Co., Ltd.
   (Shanghai, China). The Evo M-MLV RT Mix Kit (Cat No. AG11728),
   SteadyPure Quick RNA Extraction Kit (Cat No. AG21101), and SYBR Green
   Premix Pro TaqHS qPCR Kit (Cat No. AG11739) were purchased from
   Accurate Biotechnology Co., Ltd. (Hunan, China). The hematoxyin and
   eosin staining solutions (Cat No. B1001) were purchased from BaiQianDu
   Biotechnology Co., Ltd. (Hunan, China). The NKA-9 adsorption resin (Cat
   No. 59LSA21LX23) was obtained from Tianjin Berens Biotechnology Co.,
   Ltd. (Tianjin, China). All additional reagents employed were of
   analytical grade and were readily available commercially.

4.2. Preparation of PMFs Extract

   The CRP was milled into a powder and approximately 600 g was weighed.
   Then, the powdered CRP was sonicated twice with 60% ethanol (1:20
   solid-to-liquid ratio) at room temperature, with each session lasting
   40 min. Following the extraction, the supernatant was combined and
   concentrated to dryness. This residue was then redissolved in 30%
   ethanol to prepare a CRP solution with a concentration of 0.5 g/mL.
   Subsequently, an NKA-9 macroporous resin column chromatography was
   utilized to enrich PMFs from CRP solution (0.5 g/mL). HPLC with SIN,
   TAN, NOB, and HMF as standards, was employed to quantify the enriched
   PMFs extract. Optimal purification: Flow rate of 1 mL/min, sample
   loading volume of 1.5 g raw drug/g resin, and elution with varying
   ethanol concentrations. Sequential washing with water was followed by
   20% and 40% ethanol to eliminate interfering substances. Ultimately,
   the column chromatography was rinsed with 95% ethanol to collect the
   PMFs, which were then concentrated into a dried paste. The PMFs dried
   paste was accurately weighed and dissolved in a 0.5% sodium
   carboxymethyl cellulose solution to prepare a 50 mg/mL PMFs solution
   for animal studies. For the chemical analysis, the PMFs solution was
   diluted to 5 mg/mL. Four Reference standards were separately dissolved
   in methanol within 10 mL volumetric flasks and stored at 4 °C. The
   sample solution was filtered through a 0.22 µm filter before UPLC-HRMS
   analysis.

4.3. Animals

   Male Sprague–Dawley rats, weighing 200–250 g, and 10-week-old male
   C57BL/6J mice, were both obtained from Spfanimals Laboratory Animal
   Technology Co., Ltd. (Beijing, China). The animals were maintained
   under controlled environmental conditions with a temperature at 25 ± 2
   °C and a relative humidity of 60 ± 10%, with a 12-h light-dark cycle,
   and were provided with sufficient food and water. They were allowed a 7
   day acclimatization period before the commencement of experiments.
   Furthermore, the rats were subjected to a 12 h fast period prior to the
   initiation of experimental procedures. All animal handling and
   experimental protocols were ethically reviewed and approved by the
   Animal Ethics Committee of Beijing University of Chinese Medicine, with
   the approval number BUCM-4-2022061502-2062.

4.4. Sequential Metabolism of PMFs in CRP

4.4.1. Stability of PMFs Extract in Simulated Gastric Juice

   Simulated gastric juice is commonly used to study the stability and
   biotransformation of chemical compositions under stomach physiological
   conditions [[190]54]. Simulated gastric juice is typically composed of
   0.82 mL of hydrochloric acid and 0.5 g of pepsin (pH 1.5). A total of 1
   mL volume of the PMFs extract was added into 50 mL of simulated gastric
   fluid. The mixture was then incubated at a physiological temperature of
   37 °C for a duration of 2 h. The reactions were stopped by adjusting
   the pH to 6–7 using NaOH (0.1 M) solution to terminate the reaction.
   The supernatant was filtered through a 0.22 µm filter before UPLC-HRMS
   analysis.

4.4.2. In Vivo Metabolic Experiments

   A comprehensive set of metabolic studies were undertaken in animal
   models, focusing on the processes of intestinal wall metabolism,
   intestinal flora metabolism, and hepatic metabolism. The surgical
   techniques employed for these studies were in accordance with
   established protocols for IPVS and the in situ closed-loop technique
   [[191]55,[192]56]. Initially, in preparation for the perfusion surgery,
   a cohort of 5–7 rats were utilized as blood donors. Whole blood was
   aspirated from the abdominal aorta and maintained at 37 °C in a water
   bath. The animal was positioned in a recumbent position on the surgical
   table after anaesthetization. The supplied blood was infused into the
   recipient animal via the left external jugular vein using a peristaltic
   pump to offset surgical blood loss. The abdominal cavity was
   meticulously opened along the midline to visualize and locate the
   jejunum and its corresponding mesenteric veins. Subsequently, the PMF
   solution was infused into the jejunal segment from the proximal end and
   discharged from the distal end using a syringe pump. Blood was
   collected from the corresponding location in different experiments. In
   the intestinal wall metabolism experiment, a catheter was cannulated
   into the mesenteric vein to collect blood, and the hepatic portal vein
   was ligatured simultaneously. In the hepatic metabolism experiment,
   blood was collected through the right external femoral vein. The
   metabolic transformation of substances within the femoral vein plasma
   samples was influenced by both the intestines and liver. The in situ
   closed-loop method was utilized in the intestinal flora metabolism
   experiment. After the abdominal cavity was opened, the colon segment
   was located. The PMF solution was injected into the colon, and both
   ends of the colon were ligated to seal the solution within the segment.
   The other procedures were the same as those used in the intestinal wall
   metabolism experiment. During the experiments assessing intestinal wall
   and hepatic metabolism, the PMF solution was administered at a
   controlled flow rate of 0.2 mL/min. In all instances, blood was
   circulated at a consistent flow rate of 0.3 mL/min. The duration for
   blood sampling typically ranged from 1.5 to 2 h.

   Additionally, a separate set of metabolic experiments were carried out
   with oral drug administration, aiming to identify chemical constituents
   in the abdominal aorta blood, brain tissue, and CSF. During this
   process, PMF-containing CSF was collected. The rats were randomly
   divided into eight groups (n = 3). The experimental groups received an
   oral administration of a PMFs solution (1 g/kg) in a volume of 20
   mL/kg, while the control groups were provided with saline. Blood, CSF,
   and brain tissue samples were collected at designated time points (0.5,
   1, 1.5, and 2 h) following intragastric administration. Blood samples
   collected from the abdominal aorta represent the equilibrium state
   achieved after metabolic processing by various organs and microbial
   flora [[193]56]. To ensure uncontaminated brain tissue and CSF free
   from blood, transcardiac perfusion with saline was conducted before
   sample collection [[194]57]. After perfusion, the rat was placed in a
   prone position on the operating table. A needle was carefully inserted
   at the junction of the atlantooccipital membrane and the dura mater,
   advancing it until the tip reached the cisterna magna. CSF was then
   slowly aspirated. Subsequently, the rat was decapitated, and the whole
   brain was carefully separated from the cranial cavity.

4.4.3. Biological Sample Processing

   The plasma sample: The plasma was separated from the whole blood by
   centrifugating at 4000 rpm for 15 min. Afterward, protein precipitation
   of the plasma sample was performed using methanol, and the supernatant
   was obtained by centrifuging at 8000 rpm for 10 min. Then, the
   supernatant was desiccated under a nitrogen gas stream at a temperature
   of 40 °C. The resulting residue was subsequently dissolved in 200 µL of
   methanol for subsequent UPLC-HRMS analysis. The CSF sample: A portion
   of CSF sample was stored directly at −80 °C for subsequent
   pharmacological experiments of PMF-containing CSF. The residual CSF was
   subjected to centrifugation at 12,000× g revolutions per min for a
   duration of 10 min. The resulting supernatant was collected for
   subsequent UPLC-HRMS analysis. The brain tissue sample: The brain
   tissue was pulverized in methanol for a duration of 3 min.
   Subsequently, the supernatant was obtained by centrifuging at 10,000
   rpm for a period of 10 min, followed by drying under a nitrogen stream
   at 40 °C. The residue was dissolved in 200 µL of methanol for
   subsequent UPLC-HRMS analysis.

4.5. UPLC-Q Exactive-Orbitrap HRMS Analysis

   The chromatographic separations were conducted by a Waters (Milford,
   MA, USA) ACQUITY UPLC BEH Shield RP C[18] column (100 mm × 2.1 mm, 1.7
   µm) maintained at a constant temperature of 35 °C. The elution process
   was executed at a flow rate of 0.3 mL/min, with a sample injection
   volume of 5 µL. The mobile phase composition involved water acidified
   with 0.1% formic acid (solvent A) and acetonitrile (solvent B). The
   gradient elution profile was as follows: 0–10 min, 10–30% B; 10–30 min,
   30–95% B; 30–31 min, 95–10% B; 31–35 min, 10% B.

   For MS analysis, the conditions were optimized as follows: alternating
   (−)/(+) ESI modes, with the HESI probe temperature set at 400 °C. The
   capillary temperature was maintained at 320 °C. The spray voltages were
   adjusted to +3.5 kV for positive ion mode and −3.0 kV for negative ion
   mode. The sheath gas (N2) flow rate was set at 35 Arb, while the
   auxiliary gas flow rate was 10 Arb. Full-scan mass spectra was acquired
   across a mass range of m/z 100–1500, with collision energies of 20, 40,
   and 60 eV for fragmentation. MS/MS experiments were conducted in a
   data-dependent acquisition mode. Data acquisition and subsequent
   processing were performed using Xcalibur software (Thermo Fisher
   Scientific. Released 2017. Xcalibur software, version 4.2. Waltham, MA,
   USA).

4.6. Screening and Identification of Bioactive Compounds in PMFs

   Utilizing the Traditional Chinese Medicine System Database and Analysis
   Platform (TCMSP) ([195]https://old.tcmsp-e.com/tcmsp.php, accessed on 2
   November 2024), Herb ([196]http://herb.ac.cn/, accessed on 2 November
   2024), the Bio-informatics Analysis Tool for Molecular mechanism of
   Traditional Chinese Medicine (BATMAN-TCM)
   ([197]http://bionet.ncpsb.org.cn/batman-tcm/index.php, accessed on 2
   November 2024), TCM-Suite ([198]http://tcm-suite.aimicrobiome.cn/,
   accessed on 2 November 2024) and pertinent literature
   [[199]23,[200]24,[201]25,[202]26,[203]27,[204]28], an exhaustive search
   was conducted to identify the potential chemical constituents of CRP
   using the search terms “chemical name” OR “ingredient” = “Citri
   Reticulatae Pericarpium”. After screening the collected information,
   detailed specifics about PMFs were obtained. TUPLC-Q Exactive-Orbitrap
   HRMS was utilized to obtain mass spectrometry data for each sample,
   combined with the above information to elucidate the chemical
   constituents.

4.7. Network Pharmacology Analysis

   Targets Collection: The spatial data file pertaining to brain-targeting
   PMFs ([205]Table 3) was retrieved from the PubChem database
   ([206]https://pubchem.ncbi.nlm.nih.gov/, accessed on 18 December 2024).
   SwissTargetPrediction ([207]http://www.swisstargetprediction.ch,
   accessed on 18 December 2024) was used to obtain the therapeutic
   targets. Meanwhile, GeneCards ([208]https://www.genecards.org/,
   accessed on 18 December 2024) was applied to collect the therapeutic
   targets of CNS disorders by searching for the keyword “Neurological
   disorders” and “Central nervous system disorders” and limiting the
   species with “Homo sapiens”.

   Network Construction and Analysis: The Venny software 2.1.0
   ([209]https://bioinfogp.cnb.csic.es/tools/venny/, accessed on 18
   December 2024) was used to identify the overlapping targets between
   ligands and CNS disorders. Interaction network of “components-targets”
   was constructed using Cytoscape 3.8.2, a specialized bioinformatics
   software.

   Functional Enrichment Analysis: The intersection genes of components
   and CNS disorders were uploaded to DAVID data
   ([210]https://david.ncifcrf.gov/summary.jsp, accessed on 18 December
   2024), with “OFFICIAL_GENE_SYMBOL” as the gene identifier and “Homo
   sapiens” as the species. DAVID 6.8 GO was applied to annotate the
   targets from three aspects: Biological Process (BP), Cellular Component
   (CC), and Molecular Function (MF). To elucidate the role of target
   points in signaling pathways, KEGG pathway enrichment analysis was
   conducted. The top 10 entries (p < 0.05) from GO analysis (BP, CC, MF)
   and the top 30 pathways (p < 0.05) from KEGG pathways were selected as
   the primary gene function enrichment processes and signaling pathways.
   To predict the mechanisms by which PMFs treat CNS diseases, a network
   was constructed that included brain-targeting components, key targets,
   and significant pathways.

4.8. Investigation the Effects of PMFs on the AD-like Mice

4.8.1. Groups and Treatment

   The intracerebroventricular (ICV) injection technique was used in this
   study. Aβ[1–42] was utilized to induce AD-like mice. First, Aβ[1–42]
   was processed into its toxic oligomeric form. Briefly, 1 mg of
   lyophilize peptides was dissolved in 1 mL of hexafluoroisopropanol,
   mixed by vortexing, and rested at room temperature for 1 h to achieve
   complete dissolution. After drying at room temperature, the peptide
   membrane was obtained and resuspended in 20 μL of DMSO. The Aβ[1–42]
   oligomers were obtained by diluting a stock Aβ suspension in PBS to a
   final concentration of 2.5 μg/μL, followed by incubation at 4 °C for 48
   h.

   A total of 48 mice were used. They were randomly assigned to 6
   different groups (n = 8). Before the surgery, the mice were given
   anesthesia and then fixed in a stereotaxic apparatus of brain. At 0.5
   mm posterior to the Bregma and 1.0 mm lateral to the sagittal suture, a
   drill bit was used to carefully drill a hole in the skull. A volume of
   1.8 μL of 2.5 μg/μL Aβ[1–42] solution was slowly injected into
   bilateral ventricles at an injection depth of 2.0 mm. After the
   injection was completed, the needle was left in place for 6 min before
   being slowly withdrawn. The wound was sutured and disinfected with
   iodophor. The mice were wrapped in cotton to maintain body temperature.
   The sham group was injected with sterile PBS solution containing 5 %
   (v/v) DMSO. To prevent infection, penicillin was administered via
   intraperitoneal injection for three days postoperatively.

   The mice were divided into 6 groups: sham group, Aβ[1–42]-induced model
   group, donepezil group (1.3 mg/kg/day), low-dose PMFs group
   (25 mg/kg/day), middle-dose PMFs group (50 mg/kg/day), high-dose PMFs
   group (100 mg/kg/day). The duration of the study was divided into two
   phases: a 2 week pre-model preventive administration phase, followed by
   a 3 week post-model therapeutic administration phase. During the
   experimental period, the mice in the sham group and Aβ[1–42]-induced
   model group were administered an equivalent amount of 0.5% sodium
   carboxymethyl cellulose solution through gavage.

4.8.2. Y Maze Test

   The Y maze was used to evaluate the memory and spatial learning of the
   mice, including the spontaneous alternation test and the novel arm
   exploration test.

   Spontaneous alternation test. To assess spontaneous alternation in the
   animals, a Y maze of 30 cm × 6 cm × 20 cm was constructed (Y-Mazes,
   Beijing Zhongshi Dichuang Technology Development Co., Ltd., Beijing,
   China). One mouse was placed in one arm of the Y maze and allowed to
   explore freely in the three arms for 5 min [[211]30,[212]31]. An
   animal’s decision was deemed a correct alternation when it entered
   three different arms consecutively, whereas any deviation from this
   pattern was considered an incorrect alternation. The spontaneous
   alternation among the three arms was recorded and analyzed by an
   Ethovision XT 15 system (Noldus Ltd., Released 2020. Wageningen, The
   Netherlands). The percentage alternation was calculated: spontaneous
   alternation rate (%) = number of spontaneous alternations/(total number
   of entering the arm − 2) × 100%.

   The novel arm exploration test. The test was undertaken in 2 phases:
   the first phase was the training, during which a partition was employed
   to obstruct one of the arms, preventing the mice from entering the arm.
   This arm was designated as the novel arm. Then, the mouse was placed in
   the start arm and allowed 5 min of exploration in open arms. A total of
   4 h after the training, mice performed the second phase. In this phase,
   the partition was removed, maintaining the novel arm in an open state.
   The mouse was placed into the start arm of the maze and explored freely
   in the three arms for 5 min [[213]58]. The movement trajectory of mice
   in Y maze was recorded and analyzed by an EthoVision XT system. Novel
   arm exploration was measured by normalizing novel arm exploration to
   total exploration: (i) The percentage of entries into the novel arm =
   number of entries into novel arm/number of entries into all arms ×
   100%. (ii) The percentage of time spent in novel arm = cumulative time
   in novel arm/cumulative time in all arms × 100%.

4.8.3. Open Field Test

   To assess the ability of exploratory, locomotor, and anxiety-like
   behavior in mice, an open field of 50 cm × 50 cm × 40 cm was
   constructed (Open Field, Beijing Zhongshi Dichuang Technology
   Development Co., Ltd., Beijing, China). The mouse was gently placed in
   the center of the open field and allowed to move freely for 10 min
   [[214]59]. Total distance moved, the frequency of entries into the
   center area, and time exploring in the center area were analyzed using
   EthoVision XT system.

4.8.4. Tissue Preparation

   Animals were sacrificed after behavioral tests to collect brain tissue.
   Firstly, the mice were rendered unconscious through an intraperitoneal
   injection of anesthetics. Then, the mice were placed in a supine
   position on the operating table and perfused transcardially with
   ice-cold saline. After decapitation, brain tissue was removed and
   weighed. The left brain tissue was stored at −80 °C for biochemical
   analysis, and the right brain tissue was immersed in 4%
   paraformaldehyde for HE staining.

4.8.5. HE Staining for Histological Evaluation

   HE staining was applied to evaluate the morphological changes in the
   hippocampal region [[215]60]. The brain tissue was removed from the
   fixative, then dehydrated using a series of alcohol concentrations,
   followed by paraffin treatment and embedding. Paraffin-embedded brain
   tissue was cut into 4 μm sections using a pathological slicer (RM2016,
   Leica Biosystems, Shanghai, China), and sections were stained using a
   standard HE staining protocol. HE images of brain tissue sections were
   obtained using the imaging system (DS-U3, Nanjing Kangni
   Mechanical&Electrical Co., Ltd., Tokyo, Japan). The histological
   evaluation was conducted by an individual who was unaware of the
   experimental groups.

4.8.6. Activated Microglial Evaluation

   Immunohistochemistry was applied to evaluate the quantity and
   morphology of microglia in hippocampal and cortical region [[216]29].
   The fixed brain tissues underwent dehydration, embedding, and
   sectioning before being incubated with the IBA-1 antibody. Then
   secondary antibodies were incubated. Images of brain tissue sections
   were obtained using the imaging system (DS-U3, Nanjing Kangni
   Mechanical&Electrical Co., Ltd., Tokyo, Japan).

4.9. Cell Culture and Treatment

   BV2 microglial cells were cultured in a 37 °C, 5% CO[2] cellular
   incubator. Cells were cultured to approximately 80% confluence, and
   then passaged for subsequent experiments. After cells reached the
   logarithmic growth phase, they were stimulated with LPS (1 µg/mL)/IFN-γ
   (2.5 ng/mL) for 24 h to establish a cell model of analog
   neuroinflammation. Broadly, the experimental groups included the
   control group, LPS/IFN-γ alone group, and LPS/IFN-γ combined with
   different components group. In the LPS/IFN-γ combined with different
   components (blank CSF, PMFs-containing CSF, PMFs monomers) treated
   groups, BV2 cells were first incubated with designated component for 2
   h. After preincubation, LPS/IFN-γ were added to the culture medium, and
   the cells were then co-incubated with LPS/IFN-γ in the presence of
   designated component for an additional 24 h.

4.10. Cell Viability Assay

   BV2 cells were moved into 96-well plates at a seeding density of 1 ×
   10^5 cells/mL, with each well receiving 0.1 mL of cell suspension.
   Following cell adhesion, the designated component was added to the
   culture medium, and the cells were then co-incubated with the
   designated component in the cell culture incubator for 24 h.
   Subsequently, cell viability was measured according to the Cell
   Counting Kit-8’s instructions.

4.11. NO and ROS Measurement

   The supernatant from the cell culture was collected and centrifuged at
   3000 rpm for a duration of 5 min to eliminate dead cells and cellular
   debris. The concentration of NO within the cell culture medium was
   measured using the Griess method.

   The cell culture solution was removed from each well. Subsequently, 1
   mL of fresh solution containing a fluorescent probe for ROS was added,
   and the plates were incubated for an additional 0.5 h in the incubator.
   The cells were subsequently observed and imaged using a fluorescence
   microscope (U-HGLGPS, Olympus Corporation, Tokyo, Japan).

4.12. ELISA Analyses

   Brain samples were ground with PBS solution, then centrifuged at 5000
   rpm at 4 °C for 10 min, and the brain extract was collected. The
   supernatant from the cell culture was collected and centrifuged at 3000
   rpm for a duration of 5 min to eliminate dead cells and cellular
   debris. The concentrations of PSD-95, Aβ, P-Tau, TNF-α, IL-1β, and IL-6
   in the cell culture medium or brain extract were quantified using ELISA
   assays, in accordance with the manufacturer’s protocols.

4.13. RT-qPCR Analyses

   Tissues were homogenized, and RNA was extracted using the SteadyPure
   Quick RNA Extraction Kit, followed by reverse transcription to cDNA
   with the Evo M-MLV RT Mix Kit. The cDNA was then combined with
   gene-specific primers and the SYBR Green Premix Pro TaqHS qPCR Kit for
   qPCR reactions. The amplification conditions for qPCR included an
   initial denaturation at 95 °C for 30 s, followed by 40 cycles of 95 °C
   for 3 s (denaturation) and 60 °C for 20 s (annealing/extension). Gene
   expression was normalized to GAPDH and calculated using the 2^−ΔΔCt
   method. All RT-qPCR analyses were performed in triplicate on the FQD
   96A RealTime Fluorescence Quantitative PCR System (QuantStudio 7 Flex,
   Applied Biosystems, Inc., Carlsbad, CA, USA). The primer sequences used
   in these analyses are provided in [217]Table 4.

Table 4.

   PCR primers for detecting mRNA expression in this study.
   Gene Symbol    Forward Primer         Reverse Primer
      TNF-α    CCCTCACACTCACAAACCAC   ACAAGGTACAACCCATCGGC
      IL-1β    CACAGCTCTGGAGATGGTGA   CTTTCAAGCTTGGGCACTTC
      IL-10    GGTTGCCAAGCCTTATCGGA  GACACCTTGGTCTTGGAGCTTA
      TGF-β    CCTCGAGACAGGCCATTTGT   GCCAGCTGACTGCTTTTCTG
      MMP-9    TCTAGGCCCAGAGGTAACCC   AGTCGAATCTCCAGACACGC
    Claudin-5  GTTAAGGCACGGGTAGCACT   TACTTCTGTGACACCGGCAC
      ZO-1     CTCAAGTTCCTGAAGCCCGT   GCAAAAGACCAACCGTCAGG
      GAPDH    AGGTCGGTGTGAACGGATTTG  GGGGTCGTTGATGGCAACA
   [218]Open in a new tab

4.14. Statistical Analysis

   Statistical analysis was conducted employing GraphPad Prism software,
   version 9.5.0. (GraphPad Software, Inc., Released 2023. La Jolla, CA,
   USA). Data derived from three separate replicate experiments were
   collectively analyzed, and the findings were presented as the mean ±
   standard deviation (SD). The data from multiple groups were evaluated
   through one-way analysis of variance (ANOVA), with subsequent testing
   via the Sidak multiple comparisons post hoc analysis. Statistical
   significance was determined at p < 0.05.

5. Conclusions

   This study developed a comprehensive sequential metabolism approach
   that emulated the process of continuous biotransformation of bioactive
   constituents after oral administration. This approach innovatively
   characterized the chemical and metabolic profiles of PMFs in CRP and
   identified seven brain-targeting components as potential
   pharmacological ingredients for treating CNS disorders. Following the
   identification of potential constituents, network pharmacology analysis
   was utilized to predict the pharmacological and molecular mechanisms.
   The results indicated that pathways associated with AD and inflammation
   were highly enriched, suggesting that PMFs may exert beneficial effects
   in ameliorating AD by regulating the inflammatory response. Subsequent
   in vivo and in vitro experimental validation suggested that PMFs hold
   potential therapeutic value for AD, with the inhibition of
   neuroinflammation likely being a key mechanism of their anti-AD
   effects. The integration of bioinformatics predictions with
   experimental validation significantly enhanced the efficiency and
   accuracy of research in elucidating therapeutic action. Additionally,
   four monomers, SIN, TAN, NOB, and HMF, were identified as potent
   inhibitors of neuroinflammation. These findings elucidated the chemical
   and functional underpinnings of PMFs in CRP, suggesting that PMFs have
   potential as therapeutic agents or dietary supplements in the
   amelioration of AD.

Acknowledgments