Abstract Background There is increasing evidence that inflammation plays a key role in the pathophysiology of periodontitis (PT) and Alzheimer’s disease (AD), but the roles of inflammation in linking PT and AD are not clear. Our aim is to analyze the potential molecular mechanisms between these two diseases using bioinformatics and systems biology approaches. Methods To elucidate the link between PT and AD, we selected shared genes (SGs) with gene-disease-association scores of ≥ 0.1 from the Disease Gene Network (DisGeNET) database, followed by extracting the hub genes. Based on these genes, we constructed gene ontology (GO) enrichment analysis, pathway enrichment analysis, protein-protein interaction (PPI) networks, transcription factors (TFs)-gene networks, microRNAs (miRNAs)-gene regulatory networks, and gene-disease association analyses. Finally, the Drug Signatures database (DSigDB) was utilized to predict candidate molecular drugs related to hub genes. Results A total of 21 common SGs between PT and AD were obtained. Cell cytokine activity, inflammatory response, and extracellular membrane were the most important enriched items in GO analysis. Interleukin-10 Signaling, LTF Danger Signal Response Pathway, and RAGE Pathway were identified as important shared pathways. IL6, IL10, IL1B, TNF, IFNG, CXCL8, CCL2, MMP9, TLR4 were identified as hub genes. Both shared pathways and hub genes are closely related to endoplasmic reticulum (ER) stress and mitochondrial dysfunction. Importantly, glutathione, simvastatin, and dexamethasone were identified as important candidate drugs for the treatment of PT and AD. Conclusions There is a close link between PT and AD pathogenesis, which may involve in the inflammation, ER and mitochondrial function. Supplementary Information The online version contains supplementary material available at 10.1186/s12903-024-04775-9. Keywords: Periodontitis, Alzheimer’s disease, Endoplasmic reticulum stress, Inflammation, Mitochondrial dysfunction, Hub gene Introduction Periodontitis (PT), a chronic inflammatory disease, is caused by infection and abnormal aggregation of bacterial [[36]1]. Its main pathological changes involve pathological loss of periodontal ligament and alveolar bone. To date, approximately 800 species of microorganisms have been reported to be associated with PT, with the majority being Gram-positive bacteria (early colonizers), followed by Gram-negative bacteria (late colonizers) [[37]2]. These bacteria are located on the surface of the teeth, contributing to the formation of dental plaque. Biofilm non-removal will lead to biofilm maturation creates anaerobic conditions in dental plaque, which promotes the colonization of pathogenic bacteria (e.g. Porphyromonas gingivalis and Aggregatibacter actinomycetemcomitans) and the release of inflammatory mediators and toxins into the bloodstream, leading to systemic inflammation [[38]3]. Active PT is characterized by a significant presence of reactivated herpes virus (e.g. HCMV, EBV-1) [[39]4], which can lead to immune suppression and bacterial overgrowth, transforming stable PT into a progressive disease [[40]5]. Alzheimer’s disease (AD), the most prevalent form of dementia, is marked by the loss of neurons and a gradual decline in memory, language, and other cognitive functions, ultimately leading to death [[41]6]. Despite tremendous efforts by researchers and clinicians, effective drugs for the treatment of AD have not yet been developed. To date, the most popular neuropathological criteria for AD diagnosis is the aggregation of extracellular amyloid beta-peptide (AβP) and neurofibrillary tangles of hyperphosphorylated tau protein inside neurons [[42]7]. However, many clinical trials testing anti-amyloid drugs, there has been no significant improvement in cognitive abilities in patients [[43]8]. Therefore, new approaches for the prevention or treatment of AD are urgently needed. Currently, many reports support inflammation as an important pathological driver of AD and cognitive decline, and there is actually evidence supporting the connection between brain and peripheral immune system [[44]3]. For example, amyloid plaques in AD patients’ brains contain herpes simplex virus (HSV), and HSV-1 particles can directly induce Ab42 fibrillation in vitro [[45]9]. In the past decade, numerous studies have reported the relationship between PT and cognitive dysfunction (e.g. AD) [[46]10, [47]11]. Epidemiological research has identified PT as a significant risk factor for AD [[48]12]. Although many studies have pointed out that the treatment of PT may be a way to explore the prevention regimens of AD [[49]13], the specific pathogenesis of PT and AD is still unclear. It has been well acknowledged that the immune inflammatory response mediated by periodontal pathogenic bacteria is a key pathogenic factor for AD. For the process of immune inflammation, endoplasmic reticulum (ER) stress and mitochondrial dysfunction are the most important physiological processes in cells. Both immune inflammation and oxidative stress can exacerbate lipid peroxidation, protein misfolding and dysfunction [[50]14], and organelle stress such as mitochondria or ER stress [[51]15]. When the inflammatory response becomes excessive and surpasses the cell’s self-regulation ability, it can induce apoptosis of periodontal ligament cells or activate inflammatory cytokines. These cytokines can diffuse through the body fluids to the brain, causing mitochondrial dysfunction and excessive production of reactive oxygen species (ROS). The excessive ROS can also induce cellular apoptosis and ER stress, triggering a cascade amplification reaction of inflammatory cytokines [[52]16]. PT may trigger or exacerbate inflammatory conditions in the elderly population, leading to memory impairment and progression of neurodegenerative diseases such as AD [[53]17]. Besides the immune-inflammatory response triggered by PT, metabolic changes due to genetic deficits or aging can also lead to mitochondrial dynamics deficits and ER stress, adversely affecting neuronal metabolism in the brain [[54]18]. Some scholars have found that oxidative stress caused by aging makes mitochondria less flexible in responding to the constantly changing demands of neurons, which may exacerbate mitochondrial dysfunction in neurons when facing chronic diseases such as PT [[55]18]. Mitochondrial dysfunction and the resulting oxidative stress, as well as ER stress and tau protein hyperphosphorylation, can mutually reinforce each other. These abnormal metabolic processes form a vicious cycle, leading to the neurodegenerative changes of AD [[56]19]. Therefore, understanding of the molecular and cellular mechanisms between PT and AD, especially ER stress and mitochondrial dysfunction, is particularly important. This study was designed to explore the common pathogenic mechanisms between PT and AD based on bioinformatics methods. In addition, we focused on cellular damage such as ER and mitochondrial dysfunction, with an aim to identify candidate drugs that may be used for treating PT and AD. Materials and methods Data sources To explore the molecular characteristics of AD and PT, we compiled a list from the Disease Gene Network (DisGeNET) database consisting of 3,397 unique AD-related genes and 682 PT-related genes [[57]20]. Among these genes, we selected those with a gene-disease-association score greater than or equal to 0.1 (Score-gda ≥ 0.1), resulting in 21 genes shared between the two diseases (Table [58]1). These common genes involve in various functional groups, such as interleukins (e.g. IL6, IL10, and IL1B) and immune-related genes (e.g. IFNG and TNF). The study’s design roadmap is illustrated in Fig. [59]1. Table 1. Shared genes of AD and PT, gene-disease association score > 0.1 Gene Gene_id UniProt Gene_Full_Name Disease_PT Disease_id_PT Score_gda_PT Disease_AD Disease_id_AD Score_gda_AD CRP 1401 [60]P02741 C-reactive protein Periodontitis C0031099 0.4 Alzheimer’s Disease C0002395 0.1 MMP9 4318 [61]P14780 matrix metallopeptidase 9 Periodontitis C0031099 0.4 Alzheimer’s Disease C0002395 0.1 CD14 929 [62]P08571 CD14 molecule Periodontitis C0031099 0.3 Alzheimer’s Disease C0002395 0.1 IL1B 3553 [63]P01584 interleukin 1 beta Periodontitis C0031099 0.3 Alzheimer’s Disease C0002395 0.6 IL10 3586 [64]P22301 interleukin 10 Periodontitis C0031099 0.3 Alzheimer’s Disease C0002395 0.1 TNF 7124 [65]P01375 tumor necrosis factor Periodontitis C0031099 0.3 Alzheimer’s Disease C0002395 0.4 CCL2 6347 [66]P13500 C-C motif chemokine ligand 2 Periodontitis C0031099 0.28 Alzheimer’s Disease C0002395 0.1 AGER 177 [67]Q15109 advanced glycosylation end-product specific receptor Periodontitis C0031099 0.22 Alzheimer’s Disease C0002395 0.3 PPARA 5465 [68]Q07869 peroxisome proliferator activated receptor alpha Periodontitis C0031099 0.2 Alzheimer’s Disease C0002395 0.1 PLG 5340 [69]P00747 plasminogen Periodontitis C0031099 0.19 Alzheimer’s Disease C0002395 0.1 TLR4 7099 [70]O00206 toll like receptor 4 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 IL1A 3552 [71]P01583 interleukin 1 alpha Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 TLR2 7097 [72]O60603 toll like receptor 2 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 VDR 7421 [73]P11473 vitamin D receptor Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 PTGS2 5743 [74]P35354 prostaglandin-endoperoxide synthase 2 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.3 IFNG 3458 [75]P01579 interferon gamma Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 IL6 3569 [76]P05231 interleukin 6 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.4 CXCL8 3576 [77]P10145 C-X-C motif chemokine ligand 8 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 IL17A 3605 [78]Q16552 interleukin 17 A Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 MPO 4353 [79]P05164 myeloperoxidase Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.5 NKAIN2 154,215 [80]Q5VXU1 sodium/potassium transporting ATPase interacting 2 Periodontitis C0031099 0.1 Alzheimer’s Disease C0002395 0.1 [81]Open in a new tab Fig. 1. [82]Fig. 1 [83]Open in a new tab The schematic diagram of the overall workflow of this study GO and pathway enrichment analysis We carried out GO enrichment analysis across Biological Process (BP), Cellular Component (CC), and Molecular Function (MF) categories utilizing the online tool ToppGene ([84]https://toppgene.cchmc.org/). Additionally, pathway enrichment analysis for Reactome, BioPlanet, WikiPathways, and Kyoto Encyclopedia of Genes and Genomes (KEGG) was performed using Enrichr ([85]https://maayanlab.cloud/Enrichr/). Both analyses were performed with an aim to identify common functions and pathways between AD and PT, with a significance threshold setting at P < 0.05. PPI network analysis The STRING database ([86]http://string-db.org), encompassing over 14,000 species, 67 million proteins, and 20 billion interactions, was employed to analyze PPI networks. This resource facilitates the extraction of functional insights from comprehensive genomic datasets [[87]21]. We constructed a PPI network for proteins from SGs, ensuring an interaction score of greater than 0.4. Classification and functional analysis of hub genes Within the PPI network, nodes with the highest connectivity are identified as hub genes. Cytohubba ([88]http://apps.cytoscape.org/apps/cytohubba), a novel Cytoscape plugin, was utilized to pinpoint central elements within biological networks [[89]22]. Seven algorithms including MCC, Degree, EPC, MNC, Closeness, Stress, and Radiality were employed to select hub genes, followed by intersection analysis. The Gene Multiple Association Network Integration Algorit (GeneMANIA) ([90]https://genemania.org) website [[91]23], a versatile and user-friendly gene function analysis tool, was used to construct a co-expression network for the identified hub genes. Construction of TF‑gene and miRNA‑gene regulatory network TFs and miRNAs play crucial roles in regulating gene transcription, providing essential molecular insights [[92]24]. NetworkAnalyst, a comprehensive platform for statistical, visualization, and meta-analysis of gene expression data, facilitated this analysis [[93]25]. JASPAR ([94]http://jaspar.genereg.net) offers TF binding profiles for multiple species across six classification groups [[95]26]. MirTarbase aids in selecting top miRNAs and detecting biological functions relevant for hypothesis development [[96]27]. Then reliable TFs binding to hub genes were identified from the JASPAR database via NetworkAnalyst, followed by extraction of miRNAs interacting with hub genes using mirTarbase through the same platform. Gene-disease association analysis DisGeNET, a thorough platform integrating gene and variant information linked to human diseases, was utilized to explore the molecular basis of specific diseases and their comorbidities [[97]20]. Additionally, we accessed the DisGeNET database through NetworkAnalyst to investigate the associations between genes and diseases, uncovering the diseases and complications related to hub genes. Evaluation and identification of candidate drugs The Drug Signatures database (DSigDB), containing 22,527 gene sets, was employed to identify small molecules capable of downregulating hub gene expression [[98]28]. This database was accessed via the Enrichr platform ([99]https://amp.pharm.mssm.edu/Enrichr/) [[100]29]. Drug molecules were identified based on the selected hub genes. Results GO and pathway enrichment analysis The most important terms in the MF, BP, and CC categories were cytokine activity, inflammatory response, and external side of the plasma membrane, respectively (Fig. [101]2). Additionally, GO terms such as signal receptor binding and cell surface were also enriched. Fig. 2. [102]Fig. 2 [103]Open in a new tab Functional enrichment analysis of the 21 shared genes. Biological processes (BP), cellular components (CC), and molecular functions (MF) Enriched pathways of SGs between AD and PT were collected from the Reactome, BioPlanet, WikiPathways, and KEGG databases (Fig. [104]3). The important pathways included IL-10 Signaling, LTF Danger Signal Response Pathway, and RAGE Pathway. Fig. 3. [105]Fig. 3 [106]Open in a new tab Bar charts of pathway enrichment analysis for shared genes between AD and PT: (A) Reactome pathway, (B) BioPlanet pathway, (C) WikiPathways, (D) KEGG 2021 human pathway PPI network analysis We carefully studied and visualized the PPI network from STRING to predict the interactions of common SGs. PPI network consisted of 21 nodes and 156 edges, with a PPI enrichment p-value less than 1.0e-16 (Fig. [107]4A). In total, 9 overlapping hub genes were selected by 7 different algorithms (Fig. [108]4B). Fig. 4. [109]Fig. 4 [110]Open in a new tab (A) PPI visualization of shared genes between AD and PT. (B) Venn diagram showing 9 overlapping hub genes selected by 7 algorithms Classification and functional analysis of hub genes The top 10 hub genes were selected using 7 algorithms from the Cytoscape plugin cytoHubba, leading to the identification of 9 common hub genes through Venn diagram intersection: IL1B, IL6, IFNG, IL10, TNF, MMP9, TLR4, CXCL8, and CCL2 (Fig. [111]4). A comprehensive gene interaction network was constructed using the GeneMANIA database to elucidate the biological functions of these hub genes (Fig. [112]S1). The co-expression rate was 64.89%, shared protein domains were 1.25%, physical interactions were 15.60%, and co-localization was 8.09%. Determination of regulatory networks at the transcriptional level Using NetworkAnalyst, we predicted and visualized interactions between the 9 hub genes, TFs, and miRNAs (Fig. [113]5 and Fig. [114]6). The TF-gene interaction network consisted of 76 nodes, 100 edges, and 9 genes, while the miRNA-gene interaction network included 112 nodes, 146 edges, and 9 genes. Moreover, in the TF/miRNA-Hub gene network, we obtained the top 10 TFs and miRNAs with high scores. The scores were calculated using the Analyze Network tool in Cytoscape, which analyzed the network based on the connectivity of nodes to other nodes. Nodes with more connections to other nodes showed higher scores. FOXC1, YY1, and JUN were potentially important TFs (Table [115]S1). Hsa-mir-24-3p, hsa-mir-98-5p, and hsa-mir-106a-5p were potentially important miRNAs (Table [116]S2). Fig. 5. [117]Fig. 5 [118]Open in a new tab Hub gene-TF regulatory interaction network. In this network, triangular nodes represent TFs, and hexagonal nodes represent hub genes Fig. 6. [119]Fig. 6 [120]Open in a new tab Hub gene-miRNA regulatory interaction network. Triangular nodes represent miRNAs, and hexagonal nodes represent hub genes Identification of disease association Gene-disease association analysis via NetworkAnalyst revealed that myocarditis, aneurysm, tuberculosis, gastritis, allergic asthma, glomerulonephritis, nasal polyps, emphysema, and IgA nephropathy were most closely related to our hub genes based on their composite scores. These relationships are detailed in Table [121]2. Table 2. List of the top 10 diseases associated with hub genes Index Name P-value Adjusted P-value Odds Ratio Combined score 1 Arthropathy 1.085e-20 6.639e-18 178875.00 8222978.62 2 Myocarditis 1.168e-20 6.639e-18 178866.00 8209308.91 3 Aneurysm 6.651e-20 2.700e-17 178632.00 7887849.11 4 Tuberculosis, Pulmonary 1.770e-19 5.590e-17 178479.00 7706378.68 5 Gastritis 3.330e-19 7.887e-17 178371.00 7588983.45 6 Allergic asthma 7.258e-19 1.473e-16 178227.00 7444009.49 7 Glomerulonephritis 1.305e-18 1.952e-16 178110.00 7334608.60 8 Nasal Polyps 1.918e-18 2.725e-16 178029.00 7262744.35 9 Pulmonary Emphysema 2.457e-18 3.325e-16 177975.00 7216445.17 10 IGA Glomerulonephritis 3.955e-18 4.867e-16 177867.00 7127408.15 [122]Open in a new tab Identification of candidate drugs Table [123]3 showed the small molecule drugs that involved in regulating the expression of pivotal genes, which were derived from the DSigDB database of the Enrichr platform. The association of the screened drugs and the hub genes were listed based on the order of the P values. The top three included glutathione, simvastatin, and dexamethasone. Table 3. List of the top 10 suggested drugs for patients with both AD and PT Index Name P-value Adjusted P-value OR Combined score 1 glutathione CTD 00006035 2.374e-20 2.151e-17 178776.00 8078362.98 2 simvastatin CTD 00007319 3.847e-17 6.338e-15 177264.00 6699969.41 3 N-Acetyl-L-cysteine CTD 00005305 5.320e-17 6.782e-15 177165.00 6638803.28 4 dexamethasone CTD 00005779 7.454e-16 5.485e-14 176211.00 6137880.72 5 curcumin CTD 00000663 5.824e-15 2.574e-13 175248.00 5744065.86 6 ARSENIC CTD 00005442 4.482e-13 7.062e-12 172323.00 4899742.95 7 CADMIUM CTD 00005555 6.261e-12 7.319e-11 169722.00 4378264.07 8 resveratrol CTD 00002483 1.322e-10 1.064e-9 165591.00 3766683.06 9 progesterone CTD 00006624 6.649e-10 4.381e-9 162765.00 3439447.66 10 quercetin CTD 00006679 6.042e-8 2.412e-7 151578.00 2519511.56 [124]Open in a new tab Discussion Individuals with PT have an increased risk of developing AD [[125]30], while those with AD or dementia are more prone to show oral health issues due to declining cognitive abilities, making them more susceptible to chronic oral diseases such as PT, tooth loss, and mucosal lesions [[126]31]. Numerous studies have suggested that the immune-inflammatory response may be a common pathogenic mechanism linking PT and AD [[127]10, [128]12, [129]17]. Despite the brain’s immune-isolated environment, systemic inflammation leads to neurodegeneration through microglial activation and the release of pro-inflammatory molecules, thereby promoting the progression of AD [[130]32]. One study described the association between inflammatory cytokine levels in AD and PT patients, indicating that PT may be related to the onset, progression, and exacerbation of AD [[131]33]. In comparison with the healthy control group, AD patients show a large amount of lipopolysaccharides (LPS) in the brain, which is the main component of the cell membrane of the periodontal pathogens P. gingivalis [[132]34]. LPS co-localizes with AβPs (Aβ40/42) in amyloid plaques and around brain blood vessels in AD patients [[133]35]. Peripheral LPS injection in mice has been shown to activate microglia, releasing pro-inflammatory cytokines like interleukins (ILs) and tumor necrosis factor-α (TNF-α), inducing brain tissue inflammation [[134]36]. When inflammation persists, β-amyloid products accumulate, leading to the induction of more cytokine production, thus forming a feedback loop that increases inflammation and tissue damage [[135]37]. Therefore, these reports suggest that bacteria can induce local inflammatory damage, and under chronic conditions, local inflammatory damage is a triggering factor for neuroinflammation and an important contributing factor to neurodegeneration and AD [[136]38]. Although previous studies have explored the relationship between AD and PT [[137]10, [138]17], research into their common molecular mechanisms through bioinformatics methods remains largely unknown. In this study, we first explored and identified SGs and hub genes between PT and AD, which contributes to elucidate the pathogenesis of both diseases. Recent studies have found that chronic inflammation can reduce the expression of lysine acetyltransferase 6B (KAT6B, a histone acetyltransferase also known as MORF), and lead to the upregulation of the key unfolded protein response (UPR) sensor PERK, thereby causing ER stress. This in turn leads to sustained activation of the UPR in periodontal ligament stem cells [[139]39]. ER stress is not a standalone metabolic process but rather a node in many interconnected pathways [[140]40]. It senses changes in ER homeostasis and, when inflammation occurs, triggers the UPR pathway to activate genes involved in protein folding and degradation to restore homeostasis. This process activates some inflammatory cytokines (IL-4, IL-8, IL-1B, TNF-α) and the nuclear factor kappa B (NF-kB) pathway, leading to mitochondrial dysfunction and the production of downstream elements such as ROS and mtRNA. This participates in a cascade amplification reaction in the release of inflammatory cytokines [[141]41]. The toxic accumulation of ROS will induce cell apoptosis and further exacerbate ER stress. When the ER’s self-regulating capacity is exceeded, it will also induce apoptosis through modulating the C/EBP homologous protein [[142]42]. The evidence suggested interconnections between ER stress, inflammation, and oxidative stress. Therefore, further understanding the molecular mechanisms of these interconnected pathways occurring in many diseases may lead to the discovery of new therapeutic targets. We used the online tool ToppGene for GO enrichment analysis, which identified the most significant terms in the MF, BP, and CC categories were cytokine activity, inflammatory response, and external side of plasma membrane, respectively. This was consistent with our hypothesis. For the enrichment pathways for SGs between PT and AD, we screened three pathways, including IL-10 Signaling, LTF Danger Signal Response Pathway, and RAGE Pathway. IL-10 regulates the ER stress response mechanism by regulating ATF-6 recruitment to the GRP-78 gene promoter [[143]43]. Additionally, it can prevent protein misfolding and ER stress by maintaining mucin production in goblet cells. The antibacterial activity of LTF has been well established, with one mechanism being its affinity for LPS, acting as a direct bactericidal agent against Gram-negative bacteria [[144]44]. LTF could interact with cell surface receptors involved in ‘danger signal’ recognition (e.g. TLR4, CD14, and CD22) [[145]45]. RAGE is a multi-ligand receptor belonging to the cell surface molecule immunoglobulin superfamily. Upon ligand binding, RAGE increases oxidative stress, leading to a vicious cycle of sustained oxidative stress and neuroinflammation through upregulation of NF-kB [[146]46]. These findings highlight the interconnected roles of inflammatory response, ER stress, and oxidative stress in the shared pathways between PT and AD. Nine common hub genes were identified through Venn diagram intersection, including IL1B, IL6, IFNG, IL10, TNF, MMP9, TLR4, CXCL8, and CCL2. These genes are closely related to inflammatory responses, ER stress, and mitochondrial dysfunction. For example, when immune cells recognize LPS from Gram-negative bacteria through TLRs, the inflammatory signaling pathway was activated. TLR triggers intracellular cascade activation of downstream kinases (e.g. IkB and MAP kinases), and regulating TFs (e.g. NF-kB and AP-1), which induce the expression of pro-inflammatory genes like cytokines and chemokines to recruit additional immune cells [[147]47]. MMP9 is a mitochondrial-related gene associated with M0 macrophages, activated dendritic cells, and M2 macrophages, which can inhibit mitochondrial injury and oxidative stress to reduce cardiovascular calcification [[148]48]. IFNG and TNF are closely related to immunity. Activation of the immune system leads to the production of TNF and IFNG, which disrupt tight junctions, TNF may lead to cell apoptosis, and increased barrier permeability further exacerbates the inflammatory process and initiates a vicious cycle [[149]49]. It is well known that aging is a common risk factor for PT and AD, which is related to oxidative stress and metabolic disorders. Studies have found that IL-6, CXCL8, and CCL2 are present in most aging cells [[150]50]. They can lead to ER stress and mitochondrial dysfunction, stimulate chronic inflammation, and affect surrounding cells, leading to age-related diseases [[151]51]. TLR4 recognizes LPS and initiates intracellular signaling through NF-κB or JNK/SAPK pathways. Therefore, in treatment, genetic engineering or drugs can be used to block or promote hub gene signaling, regulating ER stress and mitochondrial dysfunction, thereby effectively inhibiting the occurrence of PT and AD [[152]52]. To better understand the pathological basis of PT and AD, the upstream regulatory factors of these hub genes, TFs and miRNAs, were also identified. The study indicated that FOXC1 overexpression can significantly reduce the infiltration of inflammatory cells (e.g. neutrophils and macrophages) and reduce the production of inflammatory factors such as IL-1β and TNF-α [[153]53]. It can also inhibit NF-κB activation, which further suppresses the inflammatory response [[154]54]. Has-miR-24-3p induces cell division in cancer cells by inhibiting CDKN1B RNA translation and affects pathways leading to PD [[155]55]. Overexpression of miR-24-3p can reduce basal intracellular Ca^2+ levels, while miR-24-3p silencing can lead to a decrease in cytoplasmic Ca^2+ and mitochondrial fragmentation in cortical neurons [[156]56]. Loss of miR-24-3p function affects mitochondrial function, alters ATP levels, increases fragments and production of ROS, and leads to defects in mitochondrial autophagy [[157]57]. MiR-24-3p causes its retention on the ER, leading to chronic ER stress and induction of cell death [[158]58]. While previous studies have indicated that these TFs and miRNAs have therapeutic potential [[159]59], in the future, further experiments are needed to confirm their efficacy and authenticity. Table [160]3 shows the top ten potential hub gene-associated small molecule drugs, including glutathione, simvastatin, and dexamethasone. Glutathione depletion is closely related to the development of neurodegenerative diseases such as AD, Huntington’s disease, and Parkinson’s disease [[161]60]. One of the characteristics of AD is the accumulation of TDP-43 protein in the nervous system, which can further reduce glutathione levels [[162]61]. This may be related to glutathione’s ability to degrade large amounts of ROS caused by aging. As an antioxidant, glutathione can reduce oxidative stress in cells, thereby alleviating cellular injury caused by ER stress and protecting mitochondria against the injury induced by oxidative stress [[163]61]. Also, it can maintain the correct structure of ER proteins by regulating the folding and translation processes of ER proteins, reducing the accumulation of abnormal proteins, and thereby alleviating ER stress [[164]62]. There is a certain relationship between dexamethasone and ER stress. Dexamethasone is a synthetic corticosteroid drug commonly used for anti-inflammatory, immunosuppressive, and anti-allergic treatments [[165]63]. ER stress is a cellular stress response that occurs when ER function is impaired or disturbed. Dexamethasone exerts its anti-inflammatory effects through multiple mechanisms, one of which is to affect the physiological state of cells by regulating ER stress [[166]63]. Studies have shown that dexamethasone can alleviate cell injury and inflammatory responses caused by ER stress, thereby protecting cells from stress-induced injury [[167]64]. Simvastatin has multiple effects, including stabilizing atherosclerotic plaques, immune regulation, and anti-inflammatory properties [[168]65]. It can prevent ER stress, inhibit the expression of CHOP, caspase-12, and p-JNK, and reduce ER stress-induced cell apoptosis [[169]66]. Furthermore, it can improve mitochondrial function in several aspects such as cytochrome c release, ROS production, Bcl-2/Bax ratio, and membrane potential [[170]67]. Some scholars have found that low-dose simvastatin can inhibit amyloid-β production, while higher doses may affect neuronal function, emphasizing the importance of appropriate dosage selection [[171]68]. These potential small molecule drugs, by targeting oxidative stress, ER stress, and mitochondrial dysfunction, offer promising therapeutic avenues for mitigating the pathological processes underlying PT and AD. Our study has some limitations. Firstly, the data were sourced from public databases, making input errors difficult to evaluate. Secondly, we used a single dataset in this study, which may have limitations in terms of data volume. Finally, external experimental validation is needed to verify our findings, such as the targets of hub genes, the therapeutic effects of TFs and miRNAs, and the validation of drugs, which need to be further verified in in vitro models. Conclusions We have identified SGs and hub genes associated with AD and PT, and conducted various bioinformatics analyses. AD and PT share some common pathogenic mechanisms, which may be mediated by specific key genes. This study provides new insights and methods for further research into molecular mechanisms, drug discovery, and the development of personalized treatments for patients with AD and PT. Electronic supplementary material Below is the link to the electronic supplementary material. [172]12903_2024_4775_MOESM1_ESM.docx^ (921.7KB, docx) Supplementary Material 1: Fig. S1. Hub genes interaction network. Table S1. Top 10 TFs with high scores in the TF-Hub gene network. Table S2. Top 10 miRNAs with high scores in the miRNA-Hub gene network. Acknowledgements