Abstract

Objective

   Through transcriptomic and metabolomic analyses, this study examined
   the role of high-fiber diet in obesity complicated by diabetes and
   neurodegenerative symptoms.

Method

   The expression matrix of high-fiber-diet-related metabolites, blood
   methylation profile associated with pre-symptomatic dementia in elderly
   patients with type 2 diabetes mellitus (T2DM), and high-throughput
   single-cell sequencing data of hippocampal samples from patients with
   Alzheimer's disease (AD) were retrieved from the Gene Expression
   Omnibus (GEO) database and through a literature search. Data were
   analyzed using principal component analysis (PCA) after quality control
   and data filtering to identify different cell clusters and candidate
   markers. A protein–protein interaction network was mapped using the
   STRING database. To further investigate the interaction among
   high-fiber-diet-related metabolites, methylation-related DEGs related
   to T2DM, and single-cell marker genes related to AD, AutoDock was used
   for semi-flexible molecular docking.

Result

   Based on GEO database data and previous studies, 24 marker genes
   associated with high-fiber diet, T2DM, and AD were identified. Top 10
   core genes include SYNE1, ANK2, SPEG, PDZD2, KALRN, PTPRM, PTPRK, BIN1,
   DOCK9, and NPNT, and their functions are primarily related to
   autophagy. According to molecular docking analysis, acetamidobenzoic
   acid, the most substantially altered metabolic marker associated with a
   high-fiber diet, had the strongest binding affinity for SPEG.

Conclusion

   By targeting the SPEG protein in the hippocampus, acetamidobenzoic
   acid, a metabolite associated with high-fiber diet, may improve
   diabetic and neurodegenerative diseases in obese people.

   Keywords: high-fiber diet, neurodegenerative diseases, obesity,
   neurovascular, SPEG, AD

1. Introduction

   The incidence of obesity is increasing annually worldwide. According to
   the recent data published in NEJM, the incidence of obesity has been
   increasing at a high rate since the 1980s, with the incidence rate
   being 12 and 5% among adults and children, respectively. The number of
   individuals with obesity is highest in China ([29]1, [30]2). Obesity
   and type 2 diabetes mellitus (T2DM) significantly increase the
   incidence of neurodegenerative diseases such as depression, dementia,
   stroke, and memory loss ([31]3–[32]8). Chronic systemic inflammation
   throughout the body is a common feature of obesity and diabetes and may
   be present in the central nervous system, suggesting an important
   relationship among obesity, diabetes, and neurodegenerative diseases
   ([33]9). Therefore, examining the regulation of energy balance in
   obesity and identifying biomarkers are major research directions at
   present.

   The central nervous system plays an important role in regulating energy
   balance in the body, energy metabolism is also linked to the health of
   the central nervous system ([34]10–[35]14). As the main appetite
   control center of the hypothalamus, the arcuate nucleus region (ARC) is
   one of the most studied neural circuits for energy balance ([36]10). As
   the sensors of peripheral nutrients and hormones, AgRP and POMC neurons
   in the ARC are considered key neurons involved in sensing the global
   energy status of an organism and playing an important role in diet and
   weight regulation ([37]15–[38]18). In a state of energy surplus, POMC
   neurons release neuropeptides such as α-MSH, which have
   appetite-suppressing effects, to reduce the energy intake of the body,
   thereby maintaining body weight ([39]17). There are network loops in
   the hypothalamus that regulate feeding and are precisely interconnected
   ([40]19). The hippocampus is part of the limbic nervous system. In
   addition to its relevance to cognition and learning, it has received
   increasing attention in the study of feeding and digestion; processing
   visceral sensory information and participating in the regulation of
   energy balance mainly through connections with the hypothalamus,
   amygdala and medulla ([41]20, [42]21). This phenomenon may be
   attributed to the regulation of the hypothalamus via the
   hippocampal–hypothalamic neural loop. Direct neural projections from
   the ventral pole of hippocampal CA1 to hypothalamic loci are involved
   in the control of food intake ([43]22). In a study, significant
   alterations in feeding behavior were observed in rats with a damaged
   hippocampus. The effects of nesfatin-1 on GD-responsive neurons in the
   ventral medial nucleus of the hypothalamus were significantly reduced
   after the hippocampal CA1 region of rats was electrically damaged,
   thereby affecting gastric motility ([44]21). However, factors affecting
   hippocampal function are intricate. Previous studies have found that
   many hormone receptors are related to feeding and energy regulation in
   the hippocampus, such as ghrelin, nesfatin-1, and insulin ([45]21,
   [46]23). The hypothalamic–pituitary–adrenal (HPA) axis and its
   neuroendocrine hormones can mediate stressful effects in the
   hippocampus ([47]24). In addition, various neurovascular markers can
   influence hippocampal function ([48]25). Therefore, we hypothesized
   that modulation of hippocampal function can improve the energy
   homeostasis function of the hypothalamus by improving the
   hippocampal–hypothalamic neural circuit, thus suppressing obesity.

   An increase in dietary fiber intake is associated with a reduced risk
   of obesity, and dietary fiber also plays a beneficial role in
   obesity-related metabolic diseases ([49]26, [50]27). High-fiber and
   low-glycaemic-index diets with conventional T2DM treatment can improve
   the disorder of glucolipid metabolism and have certain hypoglycaemic
   effects in elderly patients with T2DM ([51]28, [52]29). Consumption of
   whole grains may prevent the development of T2DM ([53]30, [54]31).
   Consumption of almonds increases dietary fiber intake, which is
   beneficial for obesity, glycaemic control and lipid profile, probably
   owing to the presence of fiber, which promotes an antidiabetic
   microbiome by increasing the amount of short-chain fatty acids
   ([55]32–[56]34). Moreover, high-fiber diet is essentially a low-calorie
   diet and consuming foods rich in fiber increases satiety and hence
   reduces caloric intake ([57]35). In addition, a high fiber intake is
   associated with a reduced risk of Alzheimer's disease (AD) ([58]36).
   Metabolites associated with high-fiber diet are blood markers that
   target key genes and suppress obesity. Inulin, a type of soluble
   dietary fiber, promotes the production of glucagon-like peptide-1
   (GLP-1) in enteroendocrine cells and suppresses postprandial blood
   glucose elevation and appetite through short-chain fatty acids (SCFAs)
   produced by the intestinal microbiota ([59]37, [60]38). Kimura et al.
   reported that SCFA exerts an inhibitory effect on fat accumulation via
   GPR43 ([61]39). Therefore, high-fiber diet improves metabolic status
   and prevents obesity. However, its effects on the
   hippocampal–hypothalamic functional axis remain unknown. Previous
   studies have reported that dietary fiber plays a role in improving
   insulin resistance ([62]40). Adherence to high-fiber diet can decrease
   the plasma levels of ghrelin and GLP-1 ([63]41). Many hormone receptors
   related to feeding and energy regulation are present in the
   hippocampus, such as ghrelin, nesfatin-1 and insulin ([64]21, [65]23).
   Therefore, we speculate that metabolites from high-fiber diet might act
   on the hippocampus. In order to identify new treatment targets for
   obesity, it is important to investigate how high-fiber diet work in
   obesity and the corresponding complications.

   This study examined high-fiber diet' ability to delay the progression
   of obesity complicated by diabetes type 2 and neurodegenerative
   symptoms using integrated scRNA transcriptomic and metabolomic
   analyses, providing new insights into obesity management.

2. Materials and methods

2.1. Methylation genes in the peripheral blood of patients with T2DM–AD

   Gene methylation signature matrix data associated with pre-dementia in
   elderly T2DM patients were downloaded from the GEO database. The
   following search strategy was used: keywords, “Type 2 diabetes” and
   “Alzheimer's disease;” study subjects, “Homo sapiens;” study type,
   “Blood methylomic signatures.” Datasets were obtained using
   whole-genome RNA-expression microarrays, and human-derived whole blood
   was used for experiments. After fine data screening, the microarray
   dataset [66]GSE62003 was eventually selected (Illumina HumanMethylation
   450 BeadChip; HumanMethylation 450_15017482) ([67]42). The platform
   used for testing samples was [68]GPL13534, and the dataset contained
   the expression data of methylation-related genes from 58 patients with
   T2DM. Firstly, samples with >10% missing methylation sites were
   excluded. Then, the R package “ChAMP” was used to perform a series of
   processes: the missing values were filled in using the ChAMP ([69]43,
   [70]44). Data were extracted and screened for DEGs using the R package.
   The probe IDs were converted to standard gene symbols. Genes with P <
   0.05 and |log[2]FC| > 0.1 were selected as methylation genes in
   peripheral blood of T2DM–AD patients.

2.2. Acquisition of metabolomics data for high-fiber diets

   Metabolomic data of the high-fiber diet were obtained using the GEO
   database and a literature search, and metabolite expression matrices
   were created to screen and identify potential metabolic markers of
   serum endogenous origin based on differential expression multiplicity
   (|log[2]FC| > 1.00) and t-test (P < 0.05) results. The MetaboAnalyst
   5.0 ([71]http://www.metaboanalyst.ca/) database was further used to
   perform metabolic pathway analysis of the potential metabolic markers
   that were significantly back-regulated after the high-fiber diet
   intervention and to obtain metabolism related differentially expressed
   genes (DEGs) ([72]45).

2.3. Downloading single-cell RNA sequencing data of hippocampal tissue from
the AD patients

   The high-throughput scRNA-seq data of the hippocampal samples
   associated with AD were downloaded from the GEO database with the
   following screening criteria: (i) Alzheimer's disease; (ii) human;
   (iii) hippocampus; and (iv) single cell RNA-seq/scRNA-seq. The
   single-cell sequencing data of hippocampal samples from patients with
   AD were extracted from the GEO database using “Alzheimer's disease” as
   the search term, and the scRNA-seq dataset ([73]GSE163577) was selected
   for further analysis ([74]46). The single-cell sequencing data of nine
   patients with AD (AD group) and eight healthy individuals (control
   group) in the dataset were selected, and cells with gene counts of
   200–10,000 and mitochondrial gene proportion of < 5% were screened
   using the Seurat package as previous researches ([75]47–[76]52).
   Subsequently, the data were normalized using the “Normalizedata”
   function of the Seurat package and the global scaling normalization
   method “LogNormalize.” To remove the batch effects of cells include in
   the analysis and maximize the preservation of the gene expression data
   of these cells, the “ScaleData” function of the Seurat package was used
   to regress the variances of “nCount/nFeature_RNA” and “percent. Mt”
   ([77]Supplementary Figure 2C). Subsequently, the “RunPCA” function of
   the Seurat package was used for dimensionality reduction and t-SNE
   clustering. The identified cells were subjected to top-down clustering
   analysis and annotated according to the known human gastric tissue cell
   marker genes.

2.4. Single cell sequencing quality control and data removal

   For quality assessment, quality control processing of sequencing raw
   expression data, Limma, Seurat, Dplyr, and Magrittr packages were used.
   Seurat's R package is used to generate objects, remove poor quality
   data, and calculate the percentages of gene counts, cell counts, and
   mitochondrial sequencing counts from the matrix data. For quality
   control, set Seurat's screening criteria as follows: cells with < 3
   genes expressed and < 50 genes were rejected. It was necessary to
   remove cells with more than 5% mitochondrial genes.

2.5. Cell annotation and screening of marker genes

   In order to obtain the highest principal component in the cell
   population, further principal component analysis (PCA) was conducted
   after quality control of the sequencing data. The principal components
   were determined by P < 0.05 screening and the t-SNE clustering
   algorithm was used to select the significant components. We used the
   Seurat package for the t-SNE analysis, and all the removed data was
   classified by setting the clustering parameter of the FindClusters
   function in Seurat to 0.5. The R package by limma was used to adjust P
   < 0.05, and the expression change was greater than or equal to twice
   the (|log[2]FC| ≥ 1.00) as the criteria to filter the marker genes. In
   addition, candidate marker genes were found in different clusters of
   cells using the ggplot2 package.

2.6. GO/KEGG enrichment analysis of marker genes

   The Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes
   (KEGG) signaling pathway analysis of the marker genes obtained in the
   previous step was performed in the R language to further explore the
   potential mechanisms of marker genes in specific cell clusters in
   Alzheimer's disease ([78]53). Potential marker genes are categorized by
   Cellular Components (CC), Molecular Functions (MF), and Biological
   Processes (BP). Pathview is also used to map the corresponding
   signaling pathways. GO enrichment and KEGG pathway analyses were
   performed using the DAVID ([79]http://david.ncifcrf.gov) and Metascape
   databases, and the results were visualized using the R software
   ([80]54–[81]58). The DAVID online database was conducted on methylation
   genes associated with high-fiber diet, T2DM, and AD for GO and KEGG
   enrichment analysis. Based on the P-value of each item p < 0.05, the
   best biological process and enrichment pathway were selected.

2.7. Network involving high-fiber diet, T2DM, and AD in single cells

   To illustrate the relevant target sets of the T2DM-AD-monocyte marker
   gene methylation network, the Venn R package was used to map
   hippocampus-associated marker genes in AD to DEGs in high-fiber diets
   and differentially methylated genes in T2DM patients, respectively. In
   order to map intersecting protein interaction networks and output
   protein-interaction relationship data, the STRING database
   ([82]https://string-db.org/) was applied ([83]59–[84]63). By analyzing
   the topological structure of the protein-protein interaction network
   model with Cytohubba plugin of Cytoscape 3.7.2, the top-ranked core
   targets were selected, and their degree, closeness, and betweenness
   values were visualized ([85]64, [86]65).

2.8. Molecular docking

   The study used AutoDock 4 for semi-flexible molecular docking of
   ligands and receptors to investigate the interaction of differentially
   metabolized compounds of a high fiber diet with the T2DM-AD-monocyte
   marker methylation gene ([87]66, [88]67). Small molecule ligands are
   flexible and changeable in semi-flexible docking, while receptors are
   robust and difficult to change. By downloading the 3D SDF file from
   PubChem, modifying the structure with Chem Bio3D Ultra 14.0, and saving
   it as mol2 format, the active ingredient was first processed. Our next
   step was to download the 3D protein model of the core target from the
   PDB database, dehydrate, hydrogenate, extract the ligand, and save it
   as a PDB file using PyMOL. The files were then converted to pdbqt files
   using Auto Dock Tools-1.5.6. The docking results were visualized in
   PyMOL to map the “protein-molecule” docking interaction patterns. In
   addition, run Discovery Studio 2019 to find the docking site and
   calculate the LibDockScore. Analyze the chemical bonds formed between
   the docking model and the 2-dimensional image.

3. Results

3.1. Differential expression analysis of metabolites associated with
high-fiber diet

   Potential serum endogenous metabolic markers were identified via
   differential expression analysis (|log[2]FC| > 1) and a t-test (P <
   0.05) using data from the GEO database and previous studies ([89]68,
   [90]69). A total of nine upregulated and seven downregulated
   metabolites were identified. According to the MetaboAnalyst 5.0
   database, the three most relevant pathways (P < 0.05) were related to
   the citrate cycle (TCA cycle); beta-alanine metabolism and the
   biosynthesis of neomycin, kanamycin, and gentamicin ([91]Figure 1A). In
   addition, the enrichment analysis of potential metabolic markers
   revealed that their functions were mainly related to the Warburg
   effect, glycolysis, and beta-alanine metabolism ([92]Figure 1B).
   Furthermore, the combined multifunctional analysis of
   high-fiber-diet-related metabolic DEGs and differential metabolites
   revealed that these metabolites were mainly involved in the citrate
   cycle (TCA cycle), beta-alanine metabolism and the biosynthesis of
   neomycin, kanamycin, and gentamicin, which is consistent with the
   results of the previous pathway enrichment analysis ([93]Figure 1C). As
   a result of the above analysis, we identified differential metabolites
   associated with high fiber diets.

Figure 1.

   [94]Figure 1
   [95]Open in a new tab

   Metabolomic analysis. (A) Signaling pathways associated with
   differentially expressed metabolites associated with high-fiber diet.
   (B) Bubble map of differentially expressed metabolites associated with
   high-fiber diet. (C) Bubble map of the interaction between
   differentially expressed metabolites associated with high-fiber diet
   and DEGs.

3.2. Methylation genes in the peripheral blood of patients with T2DM–AD

   The expression profiles of methylation-related genes in the peripheral
   blood of patients with T2DM–AD were obtained. And a total of 11
   upregulated and 10 downregulated methylation-related genes were
   identified and visualized on a heat map ([96]Supplementary Figure 1A).
   The transcriptomic data of these DEGs were imported into the R software
   to construct a volcano plot ([97]Supplementary Figure 1B). The
   peripheral blood of T2DM-AD patients was analyzed for
   methylation-related genes.

3.3. Cellular distribution and characteristics in the hippocampal tissue of
patients with AD

   Hippocampal samples from patients with AD were extracted from the GEO
   database using “Alzheimer's disease” as the search term, and the
   scRNA-seq dataset ([98]GSE163577) was selected for further analysis.
   The single-cell sequencing data of nine patients with AD and eight
   healthy individuals in the dataset were selected, and cells with gene
   counts of 200–10,000 and mitochondrial gene proportion of < 5% were
   screened using the Seurat package in R ([99]Supplementary Figures 2A,
   [100]B). To remove the batch effects of cells include in the analysis
   and maximize the preservation of the gene expression data of these
   cells, the “ScaleData” function of the Seurat package was used to
   regress the variances of “nCount/nFeature_RNA” and “percent. Mt”
   ([101]Supplementary Figure 2C). Subsequently, the “RunPCA” function of
   the Seurat package was used for dimensionality reduction and t-SNE
   clustering ([102]Supplementary Figures 3A, [103]B). A total of 182,056
   cells and 27,005 associated genes were identified after quality
   control, integration and normalization of data and removal of batch
   effects from the single-cell sequencing data of patients with AD and
   healthy individuals. The identified cells were subjected to top-down
   clustering analysis and annotated according to the known human gastric
   tissue cell marker genes. A total of 27 cell types were identified
   ([104]Supplementary Figure 3B), with uniform scattered distribution and
   good integration of batch effects. The expression of marker genes in
   each cell type was specific, indicating that the cell annotation
   results were accurate ([105]Figure 2). Thus, hippocampal tissue of AD
   patients was examined with respect to its distribution and
   characteristics.

Figure 2.

   [106]Figure 2
   [107]Open in a new tab

   Annotated distribution maps of cells in tSNE map.

3.4. Screening and enrichment analysis of marker genes

   Based on the results of cell annotation, different types of cell
   subpopulations were identified using the Seurat package, and DEGs in
   different cell subpopulations in the hippocampal tissue of patients
   with AD and healthy individuals were screened using the FindMarkers
   function. A total of 12,610 DEGs were identified and visualized on a
   heat map ([108]Figure 3A). GO analysis revealed that the identified
   DEGs were mainly enriched in cellular responses to nitric oxide, muscle
   cell differentiation, and RNA polymerase II transcriptional regulation
   complex ([109]Figure 3B). KEGG analysis revealed that the genes were
   mainly enriched in pathways associated with autophagy, cellular
   senescence, AGE–RAGE signaling in diabetic complications and NF–kappa B
   signaling ([110]Figure 3C).

Figure 3.

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

   Enrichment analysis for the identification of markers genes. (A) Heat
   map demonstrating the distribution of marker genes. (B) Histogram of GO
   enrichment analysis of marker genes. (C) Histogram of KEGG functional
   analysis of marker genes.

3.5. Interaction network of high-fiber-diet-related metabolites,
methylation-related DEGs associated with T2DM–AD, and single-cell marker
genes associated with AD

   There were 24 high-fiber-diet–T2DM–AD marker genes identified by
   intersecting high-fiber-diet-related metabolites with
   methylation-related DEGs associated with T2DM ([113]Figure 4A). The
   STRING (version 11.0) database was used to construct a gene interaction
   network for these marker genes. The species selected was “Homo
   sapiens.” To identify hub genes ([114]Figures 4B–[115]D), the cytoHubba
   plug-in in Cytoscape 3.7.2 software was used to construct a
   protein–protein interaction network based on network topology,
   calculate degree, closeness, and betweenness values, and select the
   size of the values for ranking. Top hub genes included SYNE1, ANK2,
   SPEG, PDZD2, KALRN, PTPRM, PTPRK, BIN1, DOCK9, and NPNT. As a result,
   we constructed a network of interactions between high-fiber
   diet-associated metabolites, methylation-associated DEGs associated
   with T2DM-AD, and single-cell marker genes related to AD. Functional
   and pathway enrichment analyses of the 24
   high-fiber-diet–T2DM–AD-related marker genes were performed using
   Metascape. GO enrichment analysis revealed a total of 245 biological
   processes (BPs), 33 cellular components (CCs), and 26 molecular
   functions (MFs). BPs included muscle cell differentiation, muscle cell
   development, cell–cell adhesion mediated by integrin, CC assembly
   involved in morphogenesis, and nephron development. MFs included
   transmembrane receptor protein tyrosine phosphatase activity,
   transmembrane receptor protein phosphatase activity, protein
   serine/threonine kinase activity, protein tyrosine phosphatase
   activity, and cysteine-type endopeptidase activity. CCs included
   contractile fibers, cell–cell junction, sarcomere, T-tubules, and
   myofibrils ([116]Figures 5A–[117]D). According to KEGG pathway
   enrichment analysis, the marker genes were enriched in two pathways
   ([118]Figures 5E, [119]F), of which autophagy was the pathway of
   interest in this study.

Figure 4.

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

   Enrichment analysis for the identification of markers genes. (A) Venn
   diagram demonstrating high-fiber-diet–T2DM–AD-related marker genes. (B)
   Potential core target genes identified based on degree values. (C)
   Potential core target genes identified based on closeness values. (D)
   Potential core target genes identified based on between-ness values.

Figure 5.

   [122]Figure 5
   [123]Open in a new tab

   GO and KEGG enrichment analysis. (A) Histogram of BP functional
   analysis. (B) Histogram of CC functional analysis. (C) Histogram of MF
   functional analysis. (D) Histogram of GO enrichment analysis. (E)
   Histogram of KEGG enrichment analysis. (F) Autophagy signaling pathway.

3.6. Molecular docking validation

   The most significantly altered metabolic markers, including
   phosphoenolpyruvate and acetamidobenzoic acid, as well as the top three
   core proteins, SYNE1 (4DXR), ANK2 (5Y4E), and SPEG (6CY6) (the
   corresponding structural domains of the proteins are mentioned in
   parentheses) were analyzed with the AutoDock Tools (version 1.5.6)
   software. Smaller the binding energy, the stronger the binding
   capacity. [124]Table 1 shows that phosphoenolpyruvate and
   acetamidobenzoic acid had adequate binding affinities for the core
   proteins SYNE1 (4DXR), ANK2 (5Y4E), and SPEG (6CY6). Acetamidobenzoic
   acid had the strongest binding affinity for the core protein SPEG
   (6CY6), with an RMSD value of < 2.00 ([125]Figure 6). Docking the
   active molecule with the corresponding target protein in Discovery
   Studio 2019 revealed that acetamidobenzoic acid binds via hydrogen and
   hydrophobic bonds to SPEG (6CY6). Acetamidobenzoic acid forms hydrogen
   bonds with amino acid residues at position 31 (ARG) and 123 (ILE) and
   hydrophobic bonds with amino acid residues at position 159 (ALA) and
   150 (PHE) of the structural domain of SPEG (6CY6) ([126]Figure 6).
   Acetamidobenzoic acid, a metabolite associated with high-fiber diet,
   may target SPEG in the hippocampus and affect autophagy related
   pathway.

Table 1.

   Molecular docking and binding energy.
   Protein Compound Structure DS (LibDockScore) Vina (kcal·mol^−1) RMSD
   SYNE1 (4DXR) Phosphoenolpyruvate graphic file with name
   fneur-13-1026904-i0001.jpg 67.3832 −2.9 1.276
   ANK2 (5Y4E) Phosphoenolpyruvate graphic file with name
   fneur-13-1026904-i0002.jpg 44.551 −3.5 1.491
   SPEG (6CY6) Phosphoenolpyruvate graphic file with name
   fneur-13-1026904-i0003.jpg 57.2527 −4.5 1.011
   SYNE1 (4DXR) Acetamidobenzoic acid graphic file with name
   fneur-13-1026904-i0004.jpg 90.7005 −4.9 2.728
   ANK2 (5Y4E) Acetamidobenzoic acid graphic file with name
   fneur-13-1026904-i0005.jpg 56.6351 −4.5 1.450
   SPEG (6CY6) Acetamidobenzoic acid graphic file with name
   fneur-13-1026904-i0006.jpg 75.5652 −6.0 1.719
   [127]Open in a new tab

Figure 6.

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

   Molecular docking analysis. (A) SYNE1 phosphoenolpyruvate (macroscopic)
   (A1); SYNE1 phosphoenolpyruvate (microscopic) (A2); SYNE1
   phosphoenolpyruvate (A3). (B) ANK2 phosphoenolpyruvate (macroscopic)
   (B1); ANK2 phosphoenolpyruvate (microscopic) (B2); ANK2
   phosphoenolpyruvate (B3). (C) SPEG phosphoenolpyruvate (macroscopic)
   (C1); SPEG phosphoenolpyruvate (microscopic) (C2); SPEG
   phosphoenolpyruvate (C3). (D) SYNE1 acetamidobenzoic acid (macroscopic)
   (D1); SYNE1 acetamidobenzoic acid (microscopic) (D2);
   SYNE1–acetamidobenzoic acid (D3). (E) ANK2–acetamidobenzoic acid
   (macroscopic) (E1); ANK2–acetamidobenzoic acid (microscopic) (E2);
   ANK2–acetamidobenzoic acid (E3). (F) SPEG–acetamidobenzoic acid
   (macroscopic) (F1); SPEG–acetamidobenzoic acid (microscopic) (F2);
   SPEG–acetamidobenzoic acid (F3).

Discussion

   In this study, metabolites and signaling pathways associated with
   high-fiber diet were identified based on databases. The GEO database
   was used to identify 10 core high-fiber-diet–T2DM–AD-related marker
   genes, whose functions are mainly related to autophagy. The results of
   molecular docking suggested that high-fiber-diet-associated metabolites
   can stably bind to the core high-fiber diet–T2DM–AD-associated
   proteins, with the most stable binding observed between SPEG (6CY6) and
   acetamidobenzoic acid. These results suggest that high-fiber diet
   influences autophagic homeostasis in the hippocampus through binding of
   the metabolite acetamidobenzoic acid to the SPEG (6CY6) protein and
   regulation of the hippocampal–hypothalamic endocrine axis, eventually
   improving the diabetic and neurodegenerative disease states of patients
   with obesity.

   Autophagy is involved in the regulation of lipid metabolism, and its
   dysregulation in adipose tissue is associated with the development of
   metabolic diseases ([130]70, [131]71). Dysregulation of autophagy
   alters energy metabolism in hypothalamic neurons and white adipose
   tissue (WAT). Imbalance of autophagy in hypothalamic neurons can lead
   to increased caloric intake and weight gain, resulting in obesity and
   metabolic disorders ([132]72). Mitochondrial autophagy is critical for
   protecting neurons in the hippocampal CA1 region from ischaemic stress
   injury ([133]73). In addition to hippocampal CA1 neuronal deletion
   leading to cognitive impairment, direct neural projections from the
   ventral pole of hippocampal CA1 to hypothalamic loci are involved in
   the control of food intake ([134]22, [135]74). Furthermore, the
   autophagic pathway is closely related to the pathogenic mechanism
   underlying the impairment of intestinal mucosal barrier function
   ([136]75). Impairment of intestinal barrier structure and function is
   an important pathogenic process in T2DM ([137]76). Autophagy is also
   involved in the pathological process of AD through several mechanisms,
   such as the removal of misfolded proteins, and is a novel therapeutic
   target for AD ([138]77, [139]78). Therefore, autophagy is closely
   related to the pathogenesis of obesity and the associated metabolic
   diseases T2DM and neurodegenerative diseases and regulates the
   hippocampal–hypothalamic neuroendocrine axis. In the present study,
   KEGG pathway enrichment analysis strongly suggested that the function
   of high-fiber-diet–T2DM–AD-related DEGs was mainly related to
   autophagy, which is consistent with the results of previous similar
   studies ([140]79–[141]82). Therefore, metabolites associated with
   high-fiber diet may play a role in the pathological process of T2DM and
   AD among patients with obesity by affecting autophagy in the
   hippocampus and hypothalamus.

   Furthermore, this study demonstrated that the
   high-fiber-diet-associated metabolite acetamidobenzoic acid can bind to
   the SPEG (6CY6) protein in the hippocampus and affect autophagic
   homeostasis in the hippocampus, thus improving the diabetic and
   neurodegenerative disease states of individuals with obesity.
   Acetamidobenzoic acid is a derivative of benzoic acid. Benzoic acid
   derivatives are involved in promoting the activity of the
   autophagy–lysosome pathway ([142]83). Benzoic acid is involved in the
   composition of the compound Mn (III) tetrakis (4-benzoic acid)
   porphyrin chloride (MnTBAP), which reduces obesity by reducing
   adipocyte hypertrophy and adipogenesis and regulating energy balance
   and improves insulin function ([143]84, [144]85). Benzoic acid
   derivatives present in garlic shells can be combined with other
   compounds to synergistically activate the PPAR signaling pathway or
   inactivate the phospholipase D signaling pathway to exert an anti-T2DM
   effect ([145]86). In addition, the bifunctional molecule BPBA
   synthesized using benzoic acid can target Aβ and inhibit
   neuroinflammation, which plays a role in AD ([146]87). To the best of
   our knowledge, this study is the first to identify the potential role
   of acetamidobenzoic acid in the pathogenesis of obesity–T2DM–AD, thus
   laying a foundation for the subsequent development of new drugs.

   Striated muscle preferentially expressed protein kinase (SPEG) is a
   myosin light-chain kinase containing a double serine/threonine kinase
   domain and multiple immunoglobulin (Ig)-like and proline-rich regions
   involved in protein–protein interactions ([147]88). SPEG as a single
   gene can be alternatively spliced into several tissue-specific
   isoforms, including BPEG (the brain) and SPEGα (skeletal muscle) and
   SPEGβ (cardiac muscle) ([148]89–[149]91). Patients with
   neurodegenerative diseases had significantly lower levels of SPEG
   methylation, which is strongly associated with obesity. The m6A
   demethylase FTO may regulate adipocyte differentiation and adipogenesis
   by regulating the expression of proteins such as gastric starvation
   hormone, pro-adipogenic factors and peroxisome proliferator-activated
   receptor, thereby affecting the development of obesity ([150]92).
   Furthermore, FTO plays a major role in the development of T2DM, as m6A
   methyltransferases can inhibit adipogenesis, delay the onset and
   progression of obesity by inhibiting autophagosome formation, blocking
   mitotic clone expansion, and controlling adipogenic differentiation in
   mesenchymal stem cells ([151]93). And missense mutations in SPEG are
   also closely associated with the development of T2DM in GK rats
   ([152]94). Therefore, acetaminobenzoic acid may target SPEG in the
   hippocampus, thereby affecting the autophagic balance of the
   hippocampus and regulating the hippocampal-hypothalamic endocrine axis,
   improving diabetic and neurodegenerative disease states.

   To the best of our knowledge, this study is the first to report that
   the SPEG protein in the hippocampus, a peripheral blood biomarker in
   patients with obesity with concomitant T2DM and neurodegenerative
   diseases, can bind to acetamidobenzoic acid, a high-fiber-diet-related
   metabolite, and plays an important role in the development of T2DM and
   neurodegenerative diseases in patients with obesity. In addition to
   acetamidobenzoic acid and SPEG, the metabolite phosphoenolpyruvate,
   which is significantly altered by high-fiber diet, was found to have a
   strong binding affinity for the core proteins SYNE1 and ANK2,
   suggesting that high-fiber diet can benefit individuals with obesity
   through multiple targets and pathways.

   However, this study has some limitations. This study was mainly based
   on bioinformatic analysis of data extracted from databases and lacks
   relevant experimental validation. Further research is needed to
   determine exactly how acetamidobenzoic acid binds to hippocampal SPEG
   proteins.

5. Conclusion

   Acetamidobenzoic acid, a metabolite associated with high-fiber diet,
   can target SPEG (6CY6) protein in the hippocampus, thereby affecting
   autophagic homeostasis in the hippocampus, regulating the
   hippocampal–hypothalamic endocrine axis and eventually improving the
   diabetic and neurodegenerative disease states of patients with obesity.

Data availability statement

   The original contributions presented in the study are included in the
   article/[153]Supplementary material, further inquiries can be directed
   to the corresponding author.

Ethics statement

   Ethical review and approval was not required for the study on human
   participants in accordance with the local legislation and institutional
   requirements. Written informed consent from the patients/participants
   or patients/participants' legal guardian/next of kin was not required
   to participate in this study in accordance with the national
   legislation and the institutional requirements.

Author contributions

   LP: conceptualization, methodology, software, investigation, formal
   analysis, and writing—original draft. YG: data curation,
   writing—original draft, methodology, software, investigation, and
   formal analysis. FD: visualization and investigation. NL:
   conceptualization, funding acquisition, resources, supervision, and
   writing—review and editing. All authors contributed to the article and
   approved the submitted version.

Acknowledgments