High level expression of OSBPL11 is associated with atherosclerosis and Alzheimer's disease

Abstract

With the global aging population, the incidence of aging-related diseases such as Alzheimer's disease (AD) and atherosclerosis (AS) is increasing. AS has also been identified as a key risk factor for AD. However, the conjoint molecular mechanisms driving these diseases remain unclear. This study first used bioinformatics analysis to analyze gene expression data in the cortex of AD patients and plaques of AS patients. We identified OSBPL11 as a gene whose expression co-increased under both conditions, and it is expressed in macrophages in AS patients and astrocytes in AD patients. Further validation was conducted through patient sample collection and an ApoE^−/− mouse model. Strikingly, our findings suggest that OSBPL11 plays a role in the development of both AS and AD, and that there is increased co-localization of OSBPL11 with macrophages and astrocytes. This discovery offers new insights into the association between AS and AD and offers a new target for the treatment AS and AD.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10753-025-02252-1.

Introduction

Atherosclerosis (AS) is a chronic vascular inflammatory disease that affects large- and medium-sized arteries throughout the body and often presents no apparent symptoms during its early stages[1, 2]. As the disease progresses, it can lead to ischemia and dysfunction of the corresponding tissues, which varies depending on the locations of the affected vessels[3, 4]. Amyloid-β peptide (Aβ) and neuronal degeneration are significant pathological changes observed in Alzheimer's disease (AD), with the inflammatory process recognized as a common mechanism contributing to the development of AD[5–7]. However, the precise mechanisms underlying its development and the role of neuroinflammation remain inadequately understood. When AS occurs within the cerebrovascular system, subsequent ischemia and hypoxia have been shown to increase the production of amyloid Aβ, which in turn can induce neurological damage. This damage influences the formation of atherosclerotic lesions through oxidative stress and endothelial dysfunction[8, 9]. These findings suggest a close association between AD and AS. Consequently, investigating the shared mechanisms between AS and AD-particularly the novel genes that play critical roles in both conditions-may generate new insights for diagnosis, prevention, and treatment.

In this study, our goal was to identify the bridging genes that connect AS and AD, thereby providing new therapeutic strategies for both conditions. To achieve this goal, we initially downloaded relevant datasets from the GEO database and conducted a bioinformatics analysis. We then collected patient samples from individuals with AS and AD and constructed corresponding mouse models. Ultimately, we validated the key genes identified through bioinformatics using various experimental approaches in both patient and mouse model samples. Our results indicate that OSBPL11 may serve as a hub gene for AS and AD, offering a novel perspective for understanding the association between these two diseases.

Materials and Methods

Data Acquisition and Combined

AD and AS-related datasets were retrieved from the Gene Expression Omnibus (GEO) database (http://www.ncbi.nlm.nih.gov/geo), which includes bulk-tissue mRNA sequence data as well as single-cell RNA-seq data. The datasets incorporated and the baseline clinical characteristics of the AS patients are detailed in Supplementary Table S1. To address batch effects across different datasets, the "ComBat" package was employed, followed by the application of the "PCA" package for detection and visualization through plotting.

Identification of DEGs and AS-AD-Related Genes

The datasets related to AD and AS were subjected to differential gene analysis using the "Limma" package in R[10]. Differentially expressed genes (DEGs, P < 0.05 and |log2(fold change)|> 1) were identified through volcano plots. Weighted gene co-expression network analysis (WGCNA)[11] was conducted utilizing the "WGCNA" package in R, where gene expression data served as input and disease classification was used as feature data to construct co-expression networks. Initially, sample clustering was performed using the hclust function, with outliers removed, and the "average" method was employed to calculate distances. An optimal soft threshold was subsequently determined to achieve an unscaled network. The dynamic tree-cutting algorithm was then applied to partition modules, and modules pertinent to information extraction were identified through Pearson correlation analysis, with modules exhibiting p values ≤ 0.05 classified as feature modules. Finally, the intersection of the feature module genes and DEGs was considered to indicate genes that were differentially associated with either AS or AD. The genes identified by stratifying the AS and AD DEGs based on upregulation or downregulation trends were regarded as key genes associated with both conditions.

Pathway Enrichment Analysis

Gene Ontology (GO) biological process and Kyoto Encyclopedia of Genes and Genomes (KEGG) annotations were performed using the DAVID website (https://david.ncifcrf.gov). Gene set enrichment analysis (GSEA) utilized data from the "GSEA" package, with the criterion for statistically significant clustering set at an adjusted P < 0.05.

Lasson Regression Analysis, Randomforest Analysis and SVM-RFE Analysis

A least absolute shrinkage and selection operator (LASSO) regression predictive model was constructed using the "cv.glmnet" function from the R package "glmnet". The LASSO analysis applied parameters α = 1 and λ = 1,000, with lambda chosen as the optimal value. Random forest analysis was performed using the "RandomForest" function in R, with the minimum error selected as the node value and the number of trees (ntree) optimized. Key genes were identified on the basis of the mean decrease in accuracy (MDA) and the mean decrease in the Gini value (MDG) of feature weights. Training datasets were fitted into an SVM classifier using the R package "e1071", where the importance of each feature was assessed by the weights or coefficients of the SVM model and ranked accordingly. By integrating these three machine learning methods, the hub genes with the highest importance were effectively identified.

Single-Cell RNA-seq Data Analysis

The dataset was initially processed into a "Seurat object" using the "Seurat" package[12]. The cells were subsequently thoroughly filtered based on strict criteria: those with fewer than 200 genes, more than 3000 total counts, or those comprising more than 20% mitochondrial genes were excluded from further analysis. The expression matrix was standardized through the application of the "NormalizeData" and "ScaleData" functions. Dimensionality reduction via principal component analysis (PCA) was then conducted using the "RunPCA" function, leading to the identification of the top 2000 variable genes for further study. To determine the composition of brain and atherosclerotic plaques, density clustering was performed utilizing the top 20 principal components. To address batch effects, the "RunHarmony" function from the Harmony package[13] was employed to harmonize single-cell data from various samples. The resulting clusters were visualized using Unified Manifold Approximation and Projection (UMAP). The marker genes for each cluster were identified using the "FindAllMarkers" function. Cell type annotation was carried out using cell markers provided by the dataset publisher for reference[14, 15]. Further exploration of cell types exhibiting high gene set scores included the extraction of all astrocyte and macrophage clusters to construct a new Seurat object. PCA was subsequently applied for additional dimensionality reduction, followed by UMAP projection. Pseudotemporal trajectory analysis was conducted independently for astrocyte and macrophage subpopulations using the "Monocle2" package[16]. Predefined marker genes were selected, and pseudotemporal trajectory analysis was executed using the "reduceDimension" function and the "DDRTree" method. Finally, the "PLOT_CELL_TRACTURE" function was utilized to visualize the cell trajectories.

Human Sample Collection

Autopsy samples were collected from cases of accidental death provided by the Forensic Identification Center of the University of South China, as well as from individuals with varying degrees of AS and AD. The data from the AD and AS patient samples used in the experiment are detailed in Supplementary Table S2. The research protocol was approved by the Ethics Committee of the University of South China (Approval No: 2024–0090).

Animals

Male ApoE^−/− mice (8 weeks old, weighing 17–20 g) were obtained from Changzhou Cavon Experimental Animal Co., Ltd. (Jiangsu, China). The mice were maintained in a controlled environment at standard room temperature, subjected to a 12:12-h light/dark cycle, and given ad libitum access to standard food and water in the SPF facility. Following a 7-day adaptation period, the mice were randomly assigned to two groups: the control group (ApoE^−/− + normal fat diet (NFD)) and the high fat diet group (ApoE^−/− + high fat diet (HFD)). After 12 weeks, all the mice were euthanized with sodium pentobarbital (150 mg/kg). All animal experiments in this study were approved by the Ethics Committee of the University of South China (Approval No: USC2023XS098).

Tissue Collection

The mice were deeply anesthetized with sodium pentobarbital (150 mg/kg) and euthanized via cervical dislocation. Aortic vessels and brain tissues were swiftly dissected, and the cerebral cortex was isolated, rapidly frozen in liquid nitrogen, and subsequently stored at −80 °C for western blotting. Concurrently, vascular and brain tissue samples from each group of mice were collected for fixation, dehydration, and embedding. The brain tissues were sliced into coronal sections with a thickness of 20 μm, whereas the vascular tissues were processed into coronal sections with a thickness of 5 μm and stored at −20 °C.

HE Staining and Masson Staining

The frozen sections were allowed to reach room temperature, followed by staining with hematoxylin for 1 min and eosin for 2 s. Masson staining, which was used to evaluate the collagen content, was performed according to the instructions provided with the Masson Trichrome Staining Kit (Solarbio, G1346), resulting in blue collagen staining.

Oil Red O Staining

A stock solution of oil red O was prepared by dissolving 300 mg of oil red O (Sigma, USA, O0625) in 100 ml of 99% isopropyl alcohol. After thawing, the frozen sections were immersed in a working solution of oil red O (30 ml of oil red O stock solution and 20 ml of distilled water) for 10 min. The sections were then restained with hematoxylin for 1 min and rinsed with distilled water for 10 min.

Immunohistochemical Staining and Immunofluorescence

Immunohistochemical staining was performed with a ready-to-use UltraSensitiveTM SP kit (MXB, KIT-9720). The primary antibodies used were Aβ (1:100, Wanleibio, WL01427) and OSBPL11 (1:1000, Affinity, DF15433), with DAB as the chromogen for slice visualization. For immunofluorescence, the samples were incubated overnight at 4 °C with a blocking solution containing 10% fetal bovine serum (FBS) to minimize nonspecific binding. The samples were subsequently incubated with OSBPL11 (1:100, Affinity, DF15433), GFAP (1:400, Cell Signaling, 3670), and CD68 (1:100, Abcam, ab233172), followed by incubation with the appropriate fluorescent secondary antibodies. Nuclei were counterstained with DAPI and observed via fluorescence microscopy.

Real-Time Quantitative PCR (qPCR)

Total RNA was extracted from the Aortic vessels and brain tissues using RNA Extraction Kit (Aidlab, RN28). RNA was reverse transcribed into cDNA using the Evo M-MLV RT Premix (Accurate Biotechnology, AG11706). qPCR was performed using the BIO-GENER™ Real-Time system. PCR products were detected by SYBR GreenPro 7agHS Premixed qPCR kit (Accurate Biotechnology, AG11701). All primers used for qPCR were designed and synthesized by Shanghai Sangong Biotechn, China. The 2^−ΔΔCt method was employed to calculate the relative expression of genes. (M-OSBPL11-F: GCGAAGTGGCATCTGGTGTGTGAG, M-OSBPL11-G: ACTCCTCTCCGTGCTCCAACAG; M-GAPDH-F: AGGTCGGTGTGAACGGATTTG, M-GAPDH-R: GGGGTCGTTGATGGCAACA).

Protein Extraction and Western Blotting

The mouse brain and vascular tissues were harvested. Proteins were extracted using a column-based animal tissue extraction kit (EpiZyme, PC201Plus), and samples were prepared accordingly. The proteins were separated by 10% SDS‒PAGE and then transferred to a PVDF membrane, which was blocked with 5% skim milk for 1 h at room temperature in a shaker or statically incubated at 4 °C overnight. The membrane was incubated with OSBPL11 (1:1000, Affinity, DF15433) and GAPDH (1:7000, Proteintech, 10494–1-AP) overnight at 4 °C. The membrane was subsequently incubated with the appropriate HRP-conjugated secondary antibody (1:7000, Proteintech, SA00001-2) for 80 min at room temperature. Bands were visualized using enhanced chemiluminescence (ECL) substrate (Abbkine, ATWD14081) for color development. GAPDH served as an internal control to normalize target protein expression.

Statistical Analysis

The images obtained were submitted to ImageJ software for quantitative analysis. The images obtained were analyzed using ImageJ software for quantitative assessment, whereas the statistical analyses were conducted via GraphPad Prism software (version 9.5.1). All bioinformatics analyses utilized R software (version 4.2.3). The Shapiro‒Wilk test was employed to assess the normality of the data distributions. Descriptive data are expressed as the means ± standard deviations. Differences between two groups were analyzed via two-tailed Student's t test, whereas differences among multiple groups were analyzed using one-way ANOVA followed by Tukey's post hoc test. A P value of < 0.05 was considered statistically significant, denoted as *P < 0.05 and **P < 0.01. Unless otherwise specified, the data are presented as the mean ± standard error of the mean (SEM).

Results

Selection and Validation of Key AS-AD Disease-Related Key Genes

The volcano plots illustrate the upregulated and downregulated DEGs in the AS and AD groups compared with the normal control group (Fig. 1A-B). Pathway enrichment analysis of these DEGs revealed significant enrichment in biological processes such as inflammation, cytokine stimulation, and cell adhesion for both disease conditions (Supplementary Fig. 1A-D). Subsequently, WGCNA was conducted on each disease dataset to construct gene modules specific to AS and AD (Supplementary Fig. 2A-B). A correlation analysis between disease characteristics and gene modules was then conducted (Supplementary Fig. 2C-D), revealing that the gray module in the AS dataset and the white and purple modules in the AD dataset were strongly associated with disease occurrence (P < 0.05) (Supplementary Fig. 2C-D). The intersection of the WGCNA module genes and DEGs revealed that 117 genes were highly correlated with AS and that 1,019 genes were correlated with AD (Supplementary Fig. 2E-F). Among these disease-associated DEGs, three upregulated genes were identified for both AS and AD patients: POM121C, RPL22, and OSBPL11, along with one downregulated gene, C9orf16 (Fig. 1C-D). Violin plots (Fig. 1E-F) illustrate the expression levels of these four genes in both the normal and diseased groups (***P < 0.001, **P < 0.01, *P < 0.05 compared with the control group). Additionally, the importance of these four genes was ranked using LASSO regression (Supplementary Fig. 2G-H), random forest analysis (Supplementary Fig. 2I-J), and support vector machine recursive feature elimination (SVM-RFE) (Fig. 1G-H). Ultimately, OSBPL11 was identified as the key gene demonstrating significant differential expression in both disease states. To validate the classification potential of OSBPL11 for these dual diseases, previously used datasets were employed as the training set, supplemented by three additional AS and AD datasets. The "pROC" package was used to compute the receiver operating characteristic (ROC) curves for OSBPL11 across these ten datasets. The findings revealed that the area under the curve (AUC) values for OSBPL11 in both the AS and AD datasets exceeded 0.7, indicating its effectiveness in distinguishing between the disease and control groups (Supplementary Fig. 3A-J).

Fig. 1.

Selection and validation of key AS-AD disease-related genes. (A-B) The volcano plot illustrates differentially expressed genes between normal blood vessels and AS plaques. (C-D) The intersection of upregulated and downregulated genes in AS and AD. (E–F) Bubble Chart of common feature genes in AS and AD datasets. (G-H) Support Vector Machine Recursive Feature Elimination (SVM-RFE) for Feature Selection identifies and filters key genes

Cellular Localization of OSBPL11 in AS and AD.

The identification of cell clusters in the AS and AD samples was conducted as described in the methods. We processed the raw scRNA-seq data, performed dimensionality reduction and clustering, and named the cell clusters on the basis of the markers provided by the original authors. In the scRNA-seq data of AS, a total of nine cell clusters were identified: smooth muscle cells (SMCs), macrophages, fibroblasts (FCs), endothelial cells (ECs), SMC-derived intermediate cells (ICSs), T cells, B cells, NK cells, and mast cells (Fig. 2A). Overall, OSBPL11 expression was significantly higher in symptomatic AS patients than in asymptomatic patients (Fig. 2B-C). Moreover, dot plots revealed that OSBPL11 expression was most pronounced in the macrophage subgroup (Fig. 2D). We separately extracted the macrophage subgroup (Supplementary Fig. 4A) and observed a more substantial increase in OSBPL11 expression in macrophages from symptomatic patients (Supplementary Fig. 4B-C), with no significant differences noted among the various subtypes of macrophages (Supplementary Fig. 4D). Pseudotemporal analysis indicated that OSBPL11 expression increased over time in subgroups approaching the late stage of AS, specifically in groups 1, 2, 7, and 8 (Supplementary Fig. 4E-G). Similarly, in the scRNA-seq data of AD patients, seven cell clusters were identified: microglia (Mic), excitatory neurons (Ex), inhibitory neurons (In), astrocytes (Ast), oligodendrocytes (Oli), oligodendrocyte progenitor cells (Opc), and endothelial cells (End) (Fig. 2E). The overall expression level of OSBPL11 was also increased in AD patients (Fig. 2F-G), with the highest enrichment observed in astrocytes (Fig. 2H). We separately extracted the astrocyte subgroup, and consistent with the results observed in macrophages in AS, OSBPL11 expression was significantly increased in astrocytes from AD patients, with no significant differences noted among the various subtypes of astrocytes (Supplementary Fig. 4H-K). Pseudotemporal analysis further revealed a continuous increase in OSBPL11 expression along the pseudotime axis in both AS and AD patients (Supplementary Fig. 4L-N).

Fig. 2.

Cellular localization of OSBPL11 in AS and AD. (A) TSNE projection shows the single-cell atlas of the human carotid atherosclerotic plaques. (B) Dot plot shows the expression of key gene OSBPL11 expression in different patients (SPT: symptomatic patients; APT: asymptomatic patients). (C-D) Feature plot and Dot plot show the expression of key gene OSBPL11 expression in each cell cluster. (E) TSNE projection shows the single-cell atlas of the human cerebral cortex. (F) Dot plot shows the expression of key gene OSBPL11 expression in different populations (Normal: Normal individuals; AD: Alzheimer's disease patients). (G-H) Feature plot and Dot plot show the expression of key gene OSBPL11 expression in each cell cluster (Dot size: fractions of expressing cells. Dot color: mean expression level in expressing cells)

Increased Expression of OSBPL11 in AS and AD Patients

To confirm whether OSBPL11 expression is elevated in AS and AD patients, we obtained samples from patients diagnosed with AS and AD. Histological examination through hematoxylin and HE staining and Masson staining revealed the presence of plaques in the arteries of AS patients, along with an increase in the number of collagen fibers (Fig. 3A). Additionally, OSBPL11 expression was increased in arterial plaques from AS patients (Fig. 3B-C). There was also observed an increase in OSBPL11 co-localization with macrophages in the plaques of AS patients (Fig. 3C). In the cortical tissues of AD patients, both Aβ and OSBPL11 expression were elevated (Fig. 3D-E). Immunofluorescence analysis further demonstrated increased co-localization of OSBPL11 with astrocytes (Fig. 3F). Thus, we preliminarily validated the upregulation of OSBPL11 in the arterial and brain tissues of AS and AD patients, as well as its increased co-localization with macrophages and astrocytes, which is in line with our bioinformatics analysis results.

Fig. 3.

Increased expression of OSBPL11 in AS and AD patients. (A) HE and Masson staining of arterial vessels in healthy control and AS patients (Scale bar: 200 μm). (B) Immunohistochemical staining for OSBPL11 expression in healthy control and AS patients (brown-yellow indicates positive staining area, from left to right magnification: 4x; 40x, **P < 0.01 compared to healthy control group, n = 8). (C) Immunofluorescence staining for OSBPL11 expression in healthy control and AS patients (Right fluorescence plot is a merge of the left single-channel fluorescence plot; Scale bar: 75 μm; DAPI in blue fluorescence indicates cell nuclei, OSBPL11 in green, CD68 in red, n = 8). (D-E) Immunohistochemical staining for Aβ and OSBPL11 expression in the cortical regions of healthy control and AD patients (brown-yellow indicates positive staining area, from left to right magnification: 4x; 40x, **P < 0.01 compared to healthy control group, n = 8). (F) Immunofluorescence staining for OSBPL11 expression in the cortical regions of healthy control and AD patients (Right fluorescence plot is a merge of the left single-channel fluorescence plot; Scale bar: 150 μm; DAPI in blue fluorescence indicates cell nuclei, OSBPL11 in green, GFAP in red, n = 8)

AS and Pathological Changes Associated with AD Occurred in ApoE^−/− Mice Fed a High Fat Diet

To investigate the associations between OSBPL11 and both AS and AD, ApoE^−/− mice were fed a high fat diet for 12 weeks, resulting in an animal model of AS. Dissection of the aortas from these mice, followed by HE staining, revealed significant plaque formation in the aortas of the ApoE^−/− + HFD group. Moreover, Oil Red O staining revealed increased lipid deposition in the aortic sinus of the ApoE^−/− + HFD group, whereas Masson staining revealed more pronounced collagen fibers in the same group (Fig. 4A). These findings suggest that AS developed in the ApoE^−/− + HFD group. Aβ is a key pathological marker of AD[17]. Additionally, Aβ plaques were deposited in the cortex of ApoE^−/− + HFD-fed mice (Fig. 4B-C). Together, these results indicate that both AS and AD pathology are induced in ApoE^−/− mice subjected to a high fat diet.

Fig. 4.

Pathological changes of AS and AD occurred in ApoE^−/− mice fed with high fat diet. (A) The schematic illustrations show aortic plaques of ApoE^−/− mice under light microscopy, ApoE^−/− mouse aorta HE, Oil red O, Masson's staining (from left to right, magnification 4x); (B) Immunofluorescence staining for Aβ expression in the ApoE^−/− brain cortex (Scale bar: 75 μm; DAPI in blue fluorescence indicates cell nuclei, Aβ in green, n = 6). (C) Immunohistochemical staining for Aβ expression in the ApoE^−/− brain cortex (brown-yellow indicates positive staining area, magnification: 4x; 40x, **P < 0.01 compared to ApoE^−/− + NFD group, n = 6)

Increased Expression of OSBPL11 in High Fat Diet Fed ApoE^−/− Mice

To assess changes in OSBPL11 expression in the ApoE^−/− mouse model, we measured OSBPL11 levels in the arterial and cortical tissues of these mice. In the ApoE^−/− + HFD group, we observed an increase in the fluorescence intensity of OSBPL11, along with increased co-localization with both astrocytes and macrophages (Fig. 5A-B). Immunohistochemistry further confirmed the increased OSBPL11 expression in the brain and vascular tissues of the ApoE^−/− + HFD group (Fig. 5C-D). Increased expression of OSBPL11 in the ApoE^−/− + HFD group at both mRNA and protein levels was detected by qPCR (Fig. 5E-F) and western blotting analysis (Fig. 5G-H). Therefore, the expression of OSBPL11 was significantly elevated in the brain and blood vessels of ApoE^−/− + HFD-fed mice compared with those of normal fat diet (ApoE^−/− + NFD)-fed mice. These findings indicate that OSBPL11 expression is increased in the context of AS and AD in this mouse model.

Fig. 5.

Increased expression of OSBPL11 in high fat diet fed ApoE^−/− mice. (A) Co-localization of OSBPL11 with macrophages in the aorta of ApoE^−/− mice (Right fluorescence plot is a merge of the left single-channel fluorescence plot; Scale bar: 75 μm, DAPI in blue fluorescence indicates cell nuclei, CD68 in red fluorescence indicates macrophages, OSBPL11 in green, n = 6). (B) Co-localization of OSBPL11 with astrocytes in the brain of ApoE^−/− mice (Scale bar: 150 μm; DAPI in blue fluorescence indicates cell nuclei, GFAP in red fluorescence indicates astrocytes, OSBPL11 in green, n = 6). (C-D) Immunohistochemical staining for OSBPL11 expression in the aorta and the brain of ApoE^−/− mice (left to right magnification: 4x; 40x, n = 6). (E–F) Real-time quantitative PCR was performed to detect the expression of OSBPL11 in the aorta and the brain of ApoE^−/− mice. (G-H) Western blotting shows the expression of OSBPL11 in the aorta and the brain of ApoE^−/− mice. (**P < 0.01 compared to ApoE^{−/− +}NFD group, n = 6)

Discussion

As the population ages, the prevalence of AS and AD is steadily increasing, highlighting the need to explore the association between these two conditions. AD is the most common form of dementia, with its prodromal phase potentially lasting several decades[18, 19]. Current research indicates that cerebral atherosclerosis (CA) involves the accumulation of cholesterol plaques within the large arterial walls of the brain, leading to reduced blood flow and metabolism[3, 20, 21]. Vascular risk factors and indicators of vascular disease have been shown to be correlated with cognitive impairment and AD, particularly among older individuals. Additionally, the presence of cerebrovascular disease exacerbates both the presence and severity of clinical symptoms associated with AD[22–24]. It is estimated that approximately 40% of dementia cases could be prevented through early lifestyle interventions targeting modifiable cardiovascular risk factors[25]. Moreover, both AS and AD are characterized by inflammation during pathogenesis[2, 6]. AS is a form of chronic vascular inflammation that is influenced by mechanisms of both innate and adaptive immunity and can either promote or inhibit AS [26]. In contrast, AD can result in neuronal degeneration during disease progression because various mediators are secreted by inflammatory cells[6]. Although AS and AD share similarities in etiology and pathogenesis, the specific genes and mechanisms involved remain unclear. Therefore, this study aims to identify commonalities between AS and AD by investigating key genes that could serve as targets for preventing the onset of both conditions.

To identify the key genes associated with AD and AS, we performed bioinformatics analysis by integrating RNA-seq and single-cell RNA-seq (scRNA-seq) data from the cortices of AD patients and plaques from AS patients. Initially, we merged four RNA-seq datasets related to AS using the Combat algorithm and conducted differential analysis with the limma package to identify DEGs in the cortices of both AS and AD patients. We subsequently utilized WGCNA to obtain the gene modules most closely associated with the disease phenotypes of AD and AS. The intersection of these gene modules with upregulated and downregulated differentially expressed genes led us to identify the genes most relevant to both diseases. Further analysis revealed four genes that were significantly associated with both AS and AD, comprising three upregulated genes and one downregulated gene. Additionally, through LASSO regression, random forest, and SVM-RFE analyses, we pinpointed OSBPL11 as a central gene crucial for differentiating between the two diseases. ROC validation using external datasets confirmed that OSBPL11 has excellent discriminatory power for both conditions. Further exploration of published scRNA-seq data revealed consistent upregulation of OSBPL11 in both AS and AD patients, confirming our RNA-seq findings. Moreover, cell type enrichment analysis revealed that in AS, OSBPL11 is predominantly expressed in macrophages, whereas in AD, its expression is concentrated in astrocytes, indicating consistency across the enriched cell types in both diseases.

The gene OSBPL11, which was identified through bioinformatics analysis, belongs to the oxysterol-binding proteins (OSBP)-related protein (ORP) family[27]. OSBP/ORPs are capable of interacting with and transferring oxysterols, cholesterol, or phospholipids, thereby influencing lipid metabolism, cell signaling, and vesicle trafficking[28, 29]. Oxysterols, which inhibit cholesterol synthesis[30], are produced through both enzymatic and nonenzymatic oxidation of cholesterol. They function by binding to liver X receptors (LXR α and LXR β), signaling an excess of cholesterol[27]. Consequently, oxysterols are often regarded as signaling compounds and are involved in a range of biological and pathological processes, including cell differentiation, apoptosis, necrosis, cellular calcium uptake, inflammation, immunosuppression, AS, gallstone development, and AD[31, 32]. Additionally, OSBPL11, an oxysterol-binding protein, is associated with these diseases. The occurrence of AS and AD is frequently associated with hyperlipidemia, suggesting that OSBPL11 may be upregulated in response to excess cholesterol[33]. This finding implies a potential connection between OSBPL11 and the pathogenesis of AS and AD. Previous studies have examined the transcriptomes of visceral fat tissue in severely obese nondiabetic men, both with and without metabolic syndrome (MetS), revealing new candidate genes related to cardiovascular disease (CVD) risk factors. OSBPL11 was found to be significantly overexpressed in the MetS group, indicating a possible association with CVD risk factors and diabetes[34]. On the basis of this evidence and our bioinformatics analysis, we hypothesize that OSBPL11 plays a crucial role in the pathogenesis of AS and AD. Our experimental findings demonstrate that OSBPL11 expression is upregulated in the arteries of AS patients and in the cerebral cortex of AD patients and that OSBPL11 is localized to macrophages in the blood vessels of AS patients and to astrocytes in AD.

Given the conservation of OSBPL11 across humans, mice, and other species, we established an AS animal model using ApoE^−/− mice. Our animal experiments revealed pathological changes associated with both AS and AD in high fat diet fed ApoE^−/− mice, alongside an increase in OSBPL11 expression. Therefore, for the first time, we identified OSBPL11 as a key gene influencing AS and AD. This study further confirmed that OSBPL11 is highly expressed in macrophages and astrocytes during the development of AS and AD, suggesting that it may influence the development of these conditions in inflammatory cells. These findings provide additional insights into the association between AS and AD.

The specific impact of OSBPL11 on AS and AD is still unclear, and no definitive conclusions have been reached regarding whether OSBPL11 primarily affects AS before leading to AD or if it influences both the blood vessels and the brain simultaneously. Furthermore, the role of OSBPL11 in inflammatory cells remains unclear. Whether the increased expression of OSBPL11 in macrophages and astrocytes activates the expression of inflammatory factors, thereby affecting the onset and progression of AS and AD, the mechanisms involved need to be explored and investigated in further cellular experiments. Moreover, the biological function of OSBPL11 also needs to be explored. It has been shown that OSBP responds to 25HC by translocating to the Golgi membrane[27]. Although OSBP has been shown to be involved in cholesterol and sphingolipid metabolism, the regulation of lipid levels is unknown. Our experiments demonstrated that OSBPL11 expression was increased under high-fat dietary conditions. Thus, our experimental results also provide evidence for an association between OSBPL11 and cholesterol. This highlights the need for more in-depth studies to elucidate the detailed molecular mechanisms linking OSBPL11 to cholesterol and inflammation with implications for AS and AD, thereby contributing valuable insights for the prevention and treatment of these conditions. In conclusion, our study presents, for the first time, evidence that OSBPL11 expression is elevated during the progression of AS and AD, activating in macrophages and astrocytes, respectively. These findings offer a new perspective on the pathogenesis of AS and AD.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

All authors contributed to the study conception and design. Material preparation, data collection and analysis were performed by [HuaYu Zhang], [Yanyu Chen], [Qian Xu], [ Xuanshuang Wu], [Naiqi He], [Zhong Ren], [Zhihan Tan], [Guixue Wang], [Qiong Xiang], and [Lushan Liu]. The first draft of the manuscript was written by [HuaYu Zhang], [Yanyu Chen] and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the Natural Science Foundation of Hunan province (2021JJ30596 to Lushan Liu).

Data Availability

No datasets were generated or analysed during the current study.

Declarations

Ethics Approval

Autopsy samples and animal studies were performed in compliance with the ARRIVE guidelines. All experimental procedures strictly adhered to the regulations set forth by the Ethics Committee of the University of South China. The approval number for autopsy samples and animal studies are 2024–0090 and USC2023XS098, respectively.

Competing Interests

The authors declare no competing interests.