Proteome Profiling of Serum Reveals Pathological Mechanisms and Biomarker Candidates for Cerebral Small Vessel Disease

Proteome Profiling of Serum Reveals Pathological Mechanisms and Biomarker Candidates for Cerebral Small Vessel Disease | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Proteome Profiling of Serum Reveals Pathological Mechanisms and Biomarker Candidates for Cerebral Small Vessel Disease Yun-chao Wang, Hang-hang Zhu, Liu-Chang He, Ya-ting Yao, Lei Zhang, and 10 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5439901/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 11 Feb, 2025 Read the published version in Translational Stroke Research → Version 1 posted 10 You are reading this latest preprint version Abstract Background Cerebral small vessel disease (CSVD) is a global brain disorder that is characterized by a series of clinical, neuroimaging, and neuropathological manifestations. However, the molecular pathophysiological mechanisms of CSVD have not been thoroughly investigated. Liquid chromatography-tandem mass spectrometry-based proteomics has broad application prospects in biomedicine. It is used to elucidate disease-related molecular processes and pathophysiological pathways, thus providing an important opportunity to explore the pathophysiological mechanisms of CSVD. Methods Serum samples were obtained from 96 participants (58 with CSVD and 38 controls) consecutively recruited from The First Affiliated Hospital of Zhengzhou University. After removing high-abundance proteins, the serum samples were analyzed using high-resolution mass spectrometry. Bioinformatics methods were used for in-depth analysis of the obtained proteomic data, and the results were verified experimentally. Results Compared with the control group, 52 proteins were differentially expressed in the sera of the CSVD group. Furthermore, analyses indicated the involvement of these differentially expressed proteins in CSVD through participation in the overactivation of complement and coagulation cascades and dysregulation of insulin-like growth factor-binding proteins. The proteomic biomarker panel identified by the machine learning model combined with clinical features is expected to facilitate the diagnosis of CSVD (AUC = 0.947,95% CI = 0.895–0.978). Conclusions The study is the most in-depth study on CSVD proteomics to date, and suggests that the overactivation of the complement cascade and the dysregulation of IGFBP on- IGF may be closely correlated with the occurrence and progression of CSVD, offering the potential to develop peripheral blood biomarkers and providing new insights into the biological basis of CSVD. cerebral small vascular disease complement cascade insulin-like growth factor proteomics Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Introduction Cerebral small vessel disease (CSVD) is a syndrome with clinical, neuroimaging, and neuropathological manifestations in the brain[ 1 ]. Computed tomography or magnetic resonance imaging (MRI) reveals CSVD as prominent in lacunar infarction (LI), white matter hyperintensities (WMH), cerebral microbleeds (CMB), MRI-visible enlarged perivascular spaces (EPVS), and atrophy[ 2 ]. As a cerebrovascular disease, CSVD is characterized by ischemic and hemorrhagic strokes, gait and balance dysfunction, and behavioral and neuropsychiatric disorders, which impose a great social burden[ 3 ]. However, the molecular pathophysiological mechanisms of CSVD have not been thoroughly investigated, and there are currently no specific interventions for this condition. It is highly likely to involve multiprotein interactions, resulting in a wide range of clinical phenotypes and neuropathological manifestations. Since proteins are the direct executors of most cellular functions and the proteome profile after extensive transcription and translation more closely resembles the final clinical phenotype[ 4 , 5 ], a comprehensive and systematic exploration of the CSVD protein profile will further promote our understanding of the pathological mechanism of CSVD, thereby contributing to the discovery of new biomarkers and therapeutic targets. Liquid chromatography/tandem mass spectrometry (LC-MS/MS)-based proteomics, can specifically identify and quantify proteins in biological or clinical samples, thus being a potentially promising tool for the discovery of key molecular alterations[ 6 – 8 ]. To this end, we collected serum samples from 58 consecutively enrolled patients with CSVD and 38 controls and developed a unique serum proteome analysis workflow for detection[ 9 , 10 ]. Through an in-depth and comprehensive analysis of the serum proteomics of patients with CSVD, we aimed to provide new insights into CSVD pathogenesis, contributing to the discovery of new targets and prevention and treatment strategies for the individualized clinical treatment of CSVD. Materials and Methods Participant recruitment A total of 96 adults were consecutively recruited from The First Affiliated Hospital of Zhengzhou University between 2018 and 2020. The inclusion criteria for the CSVD group (n = 58) included age ≥ 45 years and total image burden score of CSVD ≥ 3 points[ 11 , 12 ]. Conversely, the control group (n = 38) comprised participants enrolled during the same period without any imaging abnormalities on brain MRI. The exclusion criteria for all patients were as follows: (1) cerebral infarction lesions (lesion diameter of diffusion-weighted imaging > 20 mm), (2) acute cerebral or subarachnoid hemorrhage, (3) dementia due to confirmed neurodegenerative diseases, such as Alzheimer's or Parkinson's diseases, (4) evident white matter lesions of non-vascular origin, such as multiple sclerosis, adult white matter dysplasia, and metabolic encephalopathy, (5) mental diseases diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, and (6) intracranial infection, traumatic brain injury, or tumors. This study was approved by the Ethics Committee of The First Affiliated Hospital of Zhengzhou University. Clinical data collection We collected basic information from all participants, including demographic data, medical history, smoking history, and family history of stroke. Fasting venous blood samples were collected at baseline to detect homocysteine, total cholesterol, total triglyceride, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, fasting blood glucose (FBG) levels, and glycosylated hemoglobin (HbAlc) value. In addition, two experienced neurologists, who were blinded to the participants’ clinical data, independently evaluated the MRI images of all participants and reached a consensus according to the Standards for Reporting Vascular Changes on Neuroimaging-2[ 13 ]. Periventricular and deep WMH (PWMH and DWMH, respectively) were assessed according to the 4-point Fazekas rating scale 11 . EPVS in the centrum semiovale and basal ganglia were assessed using a validated 4-point visual rating scale (0 = no EPVS; 1 = 5–10 EPVS; 2 = 11–20 EPVS; 3 = 21–40 EPVS; and 4 = > 40 EPVS)[ 14 ]. CMB presence and number were evaluated on susceptibility-weighted imaging according to current consensus criteria and scored using a 4-point scale (0 = 0; 1 = 2–4; 2 = 5–9; and 3 = ≥ 10)[ 15 , 16 ]. Sample processing, mass spectrometry detection, and retrieval Peripheral venous blood from all participants was collected and centrifuged for 10 min at 3200 rpm at 4°C; of this, 8 µL of serum from each enrolled participant was subjected to removal of HSA and immunoglobulin using the High-Select ™ HSA/Immunoglobulin Depletion Resin (Thermo Scientific™, A36369). The remaining proteins were then precipitated with acetone and redissolved using ultrasound in 25 mM ammonium bicarbonate. Dithiothreitol was added at a final concentration of 5 mmol. Trypsinization was performed twice for 12 h and 2 h. The peptides were desalted and eluted using a C18 column. Finally, label-free unlabeled quantitative proteomic technology combined with high-performance LC-MS/MS and Thermo QE-HFX mass spectrometry was used for peptide detection and analysis. The Sequest search engine of Proteome Discoverer 2.3 software was used to search for original data generated by the Q-Exactive HF-X mass spectrometer. Data processing and statistical analysis Statistical and bioinformatic analyses were performed using R Studio and SPSS 22.0. All data analyses were performed under the guidance of professional statisticians and by following relevant guidelines and regulations[ 17 ]. All abundance values were preprocessed using the R package "edge R" to facilitate the subsequent analysis. The R package "limma" was used to identify differentially expressed proteins (DEPs) between groups. Gene ontology (GO)[ 18 ] and Reactome[ 19 ] enrichment analyses were conducted using R packages. “Weighted co-expression network analysis was implemented by R package "WGCNA"[ 20 ]. We correlated the clinical phenotype data with the protein expression data in the reconstructed sample clustering tree. Pearson’s correlation analysis was performed to estimate the correlation between module eigengenes and clinical features to identify the key module most relevant to CSVD. The Search Tool for the Retrieval of International Genes (STRING) database was used to analyze protein-protein interactions (PPI), and Cytoscape software (version 3.9.0) was used to process and visualize the resulting PPI network. SPSS 22.0 was used for the statistical analysis of demographic and laboratory examination data. The measurement data that conformed to normal distribution were expressed as mean ± standard deviation (mean ± s) and analyzed by the independent sample t-test; the measurement data that did not conform to the normal distribution were expressed as medians and upper and lower quartiles [M(QL, QU)], and the nonparametric test was used; Categorical variables are expressed as quantities (n, %), for qualitative dichotomous variables, when T ≥ 5, we used Pearson’s chi-square test; when T < 5 but T ≥ 1, we used the continuity-adjusted chi-square test; when T < 1, we used Fisher’s exact test. P < 0.05 was regarded for statistical significance. Machine learning Machine learning was used to filter the important features and build predictive diagnostic models. Standardized data were used to filter the characteristics, where 70% were randomly used as the training set and 30% as the validation set. We used the least absolute shrinkage and selection Operation (LASSO) algorithm to select important features by adjusting the penalty parameter with a 10-fold cross-validation[ 21 ]. Simultaneously, the support vector machine recursive feature elimination (SVM-RFE) algorithm for searching lambda with minimal classification was also used for screening[ 22 ]. The receiver operator characteristic curves were plotted and the area under the curve (AUC) was calculated using the R packages "pROC"[ 23 ]. The AUC and accuracy (ACC) values were used to evaluate the accuracy and precision of the diagnostic model. To keep the model as simple as possible, the number of key genes was limited to 12. Finally, we used the intersection of the two algorithms for the subsequent analysis. Validation by enzyme-linked immunosorbent assay (ELISA) Samples from the two new-independent groups (CSVD and control) were used for ELISA verification to quantify the serum concentrations of the selected protein signatures. The participants were re-screened according to the eligibility criteria. ELISA kits for the following human proteins were used to measure serum protein changes in participants from the CSVD and control groups: coagulation factor Ⅸ (F9, CSB-E08443h), apolipoprotein B (APOB, SEKH-0515), tetranectin (CLEC3B, CSB-EL005531HU), cystatin C (Cys-C, CSB-E08384h), complement 1q (C1q, CSB-E10118h), and pantetheinase (VNN1, CSB-EL025883HU). Results Baseline characteristics of the participants A workflow chart of the study is shown in Fig. 1 -A. A total of 96 participants were enrolled in this study. These were consecutively enrolled participants who underwent complete medical history taking, magnetic resonance imaging, laboratory testing, and cognitive assessment within 2 days after admission. The demographic and clinical characteristics of the participants are summarized in Table 1 . The mean age of the CSVD group was 65.21 ± 9.11 years, and 42 (72.41%) participants were men. Compared with the control group, participants in the CSVD group tended to be older, male, and had a higher prevalence of hypertension, diabetes, and high levels of homocysteine, HbAlc, and FBG. Table 1 Clinical data of study participants. Demographics, clinical characteristics, and laboratory data of the participants of the cerebral small-vessel disease (CSVD) and control groups. The measurement data that conformed to the normal distribution are expressed as mean ± standard deviation (mean ± s); the measurement data that did not conform to the normal distribution are expressed as medians and upper and lower quartiles [M (QL, QU)]. P < 0.05 is considered statistically significant. * indicates a statistically significant difference. Variables Total Count n = 96 Control n = 38 CSVD n = 58 P value Age (year) 60.75 ± 9.83 53.95 ± 6.44 65.21 ± 9.11 <0.001* Sex (male, %) 55(57.29) 13(34.21) 42(72.41) <0.001* Hypertension (n, %) 53(55.21) 2(5.26) 51(87.93) <0.001* Diabetes (n, %) 15(15.63) 0(0) 15(25.86) 0.001* Coronary Artery Disease, CAD (n %) 7(7.29) 2(5.26) 5(8.62) 0.828 Dyslipidemia (n, %) 36(37.50) 17(44.74) 19(32.76) 0.236 Family history of stroke (n, %) 11(11.46) 6(15.79) 5(8.62) 0.281 Smoking (n, %) 21(21.88) 5(13.16) 16(27.59) 0.094 Homocysteine(µmol/L) 12.99(10.86,17.74) 10.90(9.02,13.11) 14.41(12.59,19.33) <0.001* Glycated hemoglobin, HbAlc (%) 5.90(5.58,6.21) 5.70(5.50,5.93) 6.0(5.70,6.43) 0.001* Fast blood glucose, FBG (mmol/L) 4.92(4.60,5.45) 4.83(4.58,5.00) 5.12(4.65,5.82) 0.018* Total Cholesterol, TCHO (mmol/L) 3.89 ± 0.97 4.09 ± 1.08 3.76 ± 0.87 0.103 Triglyceride, TG (mmol/L) 1.20(0.89,1.76) 1.12(0.86,1.83) 1.23(0.90,1.77) 0.913 High-density Lipoprotein cholesterol, HDL-C (mmol/L) 1.21 ± 0.32 1.23 ± 0.30 1.20 ± 0.34 0.609 Low-density Lipoprotein cholesterol, LDL-C (mmol/L) 2.27 ± 0.83 2.44 ± 0.94 2.16 ± 0.74 0.110 Proteome alterations in the sera of patients with CSVD and controls A total of 818 proteins were detected by mass spectrometry in 96 patients, and 766 proteins remained after the removal of contaminating proteins. Partial least squares discrimination analysis (PLS-DA), a supervised analysis method[ 24 ], was used to evaluate the general differences in protein expression between and within groups. The results showed that the patients with CSVD and controls were divided into two groups (Fig. 1 -B). Of the 766 proteins obtained after filtration, 760 were found in the CSVD group and 755 in the control group, with 749 overlapping proteins between the two groups (Fig. 1 -C). We excluded proteins with more than 30% missing values and after conducting K-nearest neighbors filling and normalization, 524 proteins were retained for subsequent analysis (Supplementary Table 1 and Supplementary Fig. 1). Identification and functional enrichment analysis of DEPs Differential abundance analysis revealed that 52 differentially expressed proteins (DEPs) were significantly dysregulated in the CSVD group than in the control group, of which 23 were upregulated and 29 were downregulated. The volcano plot and heatmap in Figs. 1 -D and E, respectively, show the protein expression trends in the CSVD and control groups. To gain insight into the biological significance of these DEPs, GO and Reactome enrichment analyses were performed using R. Figure 2 A and Supplementary Table 3 show the top 10 terms with the smallest P-values for biological processes (BP), cellular components (CC), and molecular functions (MF), respectively. The results showed that the DEPs were mainly enriched in blood coagulation and hemostasis of BPs, collagen-containing extracellular matrix and blood microparticles of CCs, and peptidase regulator activity of MFs. In the Reactome pathway enrichment analysis (Supplementary Table 2), we identified five important biological pathways (Fig. 2 -B), including the regulation of insulin-like growth factor (IGF) transport and uptake by insulin-like growth factor-binding proteins (IGFBPs), platelet degranulation, and response to elevated platelet cytosolic Ca 2+ . To determine the interactive relationships among the DEPs, we used the STRING database to construct a PPI network (Fig. 2 C). The results, visualized using Cytoscape, showed a strong interaction between the DEPs. The CytoHubba plug-in identified the top 10 hub proteins as plasminogen (PLG), antithrombin-III (SERPINC1), prothrombin (F2), APOB, kininogen-1 (KNG1), serum amyloid P component (APCS), complement component 9 (C9), coagulation factor IX (F9), alpha-2 antiplasmin (SERPINF2), and inter-alpha-trypsin inhibitor heavy chain H1 (ITIH1) (Fig. 2 D). We evaluated the correlation coefficients of 25 recorded clinical features and 52 DEP levels in both groups to demonstrate their relationship (Fig. 3 ). The results indicated that most DEPs were significantly associated with the disease group, imaging findings, and cognitive function. Among them, the severity of imaging abnormalities and cognitive impairment was positively correlated with upregulated DEPs but negatively correlated with downregulated DEPs (Fig. 3 ). Weighted co-expression network analysis A weighted co-expression network analysis was performed using the "WGCNA" package to further screen for key proteins related to CSVD clinical features. All proteins were categorized into six co-expression modules, including the gray module (Fig. 4 A and 4 B). The upregulation and downregulation of differentially expressed proteins in each module are shown in Fig. 4 C. The correlation between the modules and clinical features was measured using the correlation between the module eigengene values and clinical features (Fig. 4 D). The blue module contained the largest proportion of DEPs (13/78) and was significantly associated with multiple clinical features (group, age, sex, PWMH, total WMH score, LI, Montreal cognitive assessment score, Alzheimer's disease-8 (AD-8) score, hypertension, and hyperlipidemia); hence, we further identified and analyzed the proteins within. GO and Reactome analyses showed that the terms associated with the blue module proteins were consistent with the enrichment analysis results of DEPs between CSVD and controls (Fig. 4 E and 4 F), including the complement cascade reaction, coagulation regulation, and the regulation of IGF transport and uptake by IGFBPs. This confirmed that these pathways are likely to play important roles in the course of CSVD (Supplementary Table 4). Machine learning We constructed receiver operating characteristic (ROC) curves based on the validation set to determine the optimal protein signatures. The results of the LASSO algorithm showed that the model had the best diagnostic performance (AUC = 0.9893, ACC = 0.9642) when a protein signature with the following top 11 proteins was selected: CLEC3B, lipocalin prostaglandin D synthetase (PTGDS), APOB, F9, cystatin C (CST3), a disintegrin and metalloproteinase with thrombospondin motifs-like 4 (ADAMTSL4), alpha chain of type XVIII collagen (COL18A1), apolipoprotein L1 (APOL1), fetuin-B (FETUB), complement C1q subcomponent subunit B (C1QB), and VNN1 (Fig. 5 A and Supplementary Fig. 2). In contrast, the SVM-RFE model performed the best (AUC = 0.9091, ACC = 0. 0.8929) when a protein signature with the following top 10 proteins was selected: eukaryotic translation initiation factor 4 gamma 3 (EIF4G3), VNN1, F9, CST3, C1QB, APOB, keratin 1 (KRT1), CLEC3B, KRT5, and paraoxonase-1 (PON1) (Fig. 5 B and Supplementary Fig. 2). Among these, F9, APOB, CST3, C1QB, CLEC3B, and VNN1 overlapped in the best diagnostic models constructed using the two algorithms, and F9 and APOB were identified as the top 10 hub proteins in the PPI network (Fig. 5 C and Supplementary Fig. 3). To test the diagnostic values of the six overlapping proteins, we conducted ROC analyses and their combination served as a good model to distinguish patients with CSVD from controls, with an AUC of 0.929 and 95% confidence interval (CI) was 0.876–0.971 (Fig. 5 D Model 1). In addition, considering the age-related characteristics of CSVD, when age was added as a variable in the new model, we found that the AUC of the new model reached 0.947 and the 95%CI was 0.895–0.978 (Fig. 5 D Model 2). Furthermore, we noted that these protein signatures were also localized to the major pathways identified above, such as regulation of IGF transport and uptake by IGFBPs, complement, and the coagulation cascade. Validation of protein signatures performance by ELISA To examine the diagnostic performance of the protein signature with six proteins (F9, APOB, CST3, C1QB, CLEC3B, and VNN1) and the diagnostic models, we further validated using ELISA and sera from the two groups. In the CSVD group, C1QB, CST3, and F9 levels showed an upward trend, whereas CLEC3B level showed a downward trend. This was consistent with the proteomics results, and the differences between the two groups were statistically significant. However, the expression of the other two proteins (APOB and VNN1) was not significantly different between the two groups (Table 2 ). This may be related to the experimental method or limited sample size. Subsequently, we correlated the serum concentration of these six proteins with the clinical characteristics of the samples (Fig. 6 ). The general trend of correlation between six selected proteins and grouping was similar to the proteomic results; however, expression of only C1QB, CLEC3B, CST3, and F9 was statistically significant and correlated with the severity of LI and CMB. Furthermore, the concentrations of C1QB, CST3, and F9 were significantly positively correlated with the severity of total WMH and EPVS. Table 2 Validation results of enzyme-linked immunosorbent assay (ELISA) for six selected proteins. Results for the determination of concentrations of six selected proteins in serum samples obtained from 27 patients with cerebral small-vessel disease (CSVD) and 27 control individuals using ELISA. The measurement data that conformed to the normal distribution were expressed as mean ± standard deviation (mean ± s); the measurement data that did not conform to the normal distribution were expressed as medians and upper and lower quartiles [M (QL, QU)]. P < 0.05 is considered statistically significant. * indicates a statistically significant difference. Variables Control CSVD P value (n = 27) (n = 27) APOB (ng/ml) 168.14(129.91,370.89) 163.15(136.56,219.65) 0.829 C1QB (µg/ml) 28.63(26.13,33.94) 35.47(30.23,42.21) < 0.001* CLEC3B (mg/L) 5.83(5.00,6.89) 4.91(3.99,6.38) 0.032* CST3 (µg/ml) 0.97(0.93,1.03) 1.05(0.97,1.10) 0.013* F9 (ng/ml) 31.18(25.67,37.99) 33.94(28.98,50.12) 0.039* VNN1 (pg/ml) 283.90(102.88,439.06) 294.25(102.88,506.30) 0.672 Discussion CSVD is complex as it involves several underlying cellular and molecular mechanisms and varies in clinical manifestations, leading to difficulties in early diagnosis and specific treatment. In this study, we identified 52 DEPs between the CSVD and control groups that were related to the severity of the CSVD imaging burden and cognitive impairment. Notably, many DEPs have been identified as potential CSVD markers in many studies. For example, COL18A1, up-regulated in the CSVD group, was confirmed to be involved in (micro) vascular wall pathology and remodeling in autopsied brain tissue from patients with CSVD[ 25 , 26 ]. Therefore, to some extent, our differential protein expression profile in CSVD confirmed previous findings and provides a direction and inspiration for future research. Furthermore, functional enrichment, protein interaction, and weighted co-expression network analyses showed that these DEPs may be involved in CSVD by participating in the overactivation of complement and coagulation cascades and dysregulation of insulin-like growth factor-binding proteins on insulin-like growth factor. Furthermore, we identified a protein signature of six proteins in CSVD using the LASSO and SVM-RFE algorithms, and verified them experimentally. First, the complement cascade plays an important role in maintaining healthy brain homeostasis, contributing to the removal of invading pathogens and apoptotic cells, pruning inappropriate synapses, and limiting inflammatory immune responses[ 27 – 29 ], and when unregulated, has adverse effects, often exacerbating disease[ 30 ]; therefore, the altered expression of complement components and complement regulatory proteins in our CSVD group may be of great significance. The upregulation of CIQB, C9, FCN2, and CFHR3 and downregulation of complement factor I in the CSVD group may indicate overactivation or under regulation of the complement cascade[ 31 ], which may participate in the pathological processes of CSVD. Several studies have shown that persistent chronic complement activation can drive a robust neuroinflammatory response associated with synaptic degeneration and progressive cognitive decline[ 32 ]. The specific mechanisms may include: 1) inducing neutrophil infiltration and increasing the secretion of pro-inflammatory cytokines, leading to the destruction of cell homeostasis and tissue damage; 2) activation of microglial phagocytosis of synapses, leading to synaptic loss and neuronal death, thereby destroying the integrity of functional neural circuits and affecting cognitive functions[ 33 ]; and 3) aggravating the original hypoxic-ischemic injury by increasing complement activation products, including opsonins, anaphylatoxins, and the membrane attack complex, ultimately delaying nerve repair and worsening prognosis[ 27 , 34 ]. Thus, we conclude that a better understanding of the process of complement involvement in CSVD may help in the prevention and monitoring of CSVD, and interception of the complement cascade may be a potential therapeutic modality. In this regard, emerging complement-targeted therapeutics for neurological diseases can provide ideas and references[ 35 – 37 ]. Second, compared to the overactivation of the complement system, the regulatory system of IGF by IGFBPs may be downregulated in the CSVD group. IGFs, regulated by six high-affinity IGFBPs, play essential roles in the regulation of growth, development, homeostasis, and neuroplastic changes in the brain, and both the vascular and nervous systems of the brain are their important targets[ 38 ]. Clinical evidence and animal studies have shown that aging can lead to the downregulation of insulin and IGF/IGFBP signaling, which can lead to cerebrovascular disease and age-related cognitive impairment[ 39 , 40 ]. Therefore, we speculate that downregulation of the IGF/IGFBP system is a key trigger for the development and progression of CSVD during aging, after which multiple downstream pathways are involved. Since inflammation may also induce a wide response in the neuroendocrine system[ 41 , 42 ], our alternative hypothesis is that the inhibition of IGF/IGFBPs is related to chronic inflammation in CSVD; however, the causal relationship between the two is unknown. Notably, although increasing evidence suggests that the IGF/IGFBPs system has many downstream pathways and that it plays an important role in lipid metabolism and glucose metabolism[ 43 , 44 ], our data suggest that the major abnormality occurring in CSVD is associated with lipid metabolism (Supplementary Table 3), such as cholesterol metabolic process (APOB, PON1, APOL1, and proprotein convertase subtilisin/kexin type 9 [PCSK9]), lipoprotein metabolic process (APOB, APOL1, and PCSK9), and glycerolipid metabolic process (APOB, PON1, and PCSK9). However, the clinical laboratory results did not show significantly abnormal lipid metabolism in the CSVD group (Table 1 ), which may be related to the widespread use of statins in patients with CSVD. In combination with previous studies, we propose that manipulating the regulatory activity of IGFBP on IGF has the potential to become a therapeutic strategy for CSVD by reducing dyslipidemia and insulin resistance[ 45 , 46 ]. Third, we identified the protein signature of six proteins. Machine learning is a key productivity tool in modern omics research that uses specific modeling and prediction methods to identify patterns in high-throughput sequencing datasets[ 47 ]. In our study, by integrating the 11 proteins obtained after LASSO screening and 10 proteins obtained after SVM-RFE screening, we identified a protein signature of six proteins (F9, APOB, CST3, C1QB, CLEC3B, and VNN1), some of which have already been implicated in CSVD, confirming previously published findings. Specifically, reduced APOB expression has previously been associated with occipital periventricular WMH[ 48 ] and an increased WMH score in the current correlation study (Fig. 3 ). In a community-based study, CST3 was associated with CSVD, and Mendelian randomization analysis showed that a genetically predicted higher CST3 level was associated with an increased risk of lacunar stroke[ 49 ]. C1QB is a key gene in healthy brain aging, whose activation in the brain may contribute to the progression of age-related cognitive dysfunction[ 50 ]. Correspondingly, the expression of CST3 and C1QB was upregulated in the CSVD group. However, F9, CLEC3B, and VNN1 are potential new biomarkers that need to be identified in future studies. The combination of these six proteins is a good model for the diagnosis of CSVD, and the diagnostic performance of this combination can be made more robust by including age as a variable. Our study had a few limitations. First, because this was a cross-sectional study, we cannot conclude causality or explore whether the abnormalities demonstrated in our results are pathogenic or compensatory responses to CSVD. Second, Further investigation is warranted to explore the mechanisms by which the proteins and pathways identified in our study affect CSVD, as well as targeted therapeutic approaches. We believe that our study makes a significant contribution to the literature because this is the most in-depth study on CSVD proteomics to date, and suggests that the overactivation of the complement cascade and the dysregulation of IGFBP on- IGF may be closely correlated with the occurrence and progression of CSVD, offering the potential to develop peripheral blood biomarkers and providing new insights into the biological basis of CSVD. Statements & Declarations Abbreviations BP: biological processes; CC: cellular components; CMB: cerebral microbleeds; CSVD: cerebral small-vessel disease; DEPs: differentially expressed proteins; DWMH: deep white matter hyperintensities; ELISA: enzyme-linked immunosorbent assay; GO: Gene ontology; IGF: insulin-like growth factor; IGFBP: insulin-like growth factor-binding proteins; LASSO: least absolute shrinkage and selection operator; LC-MS/MS: liquid chromatography/tandem mass spectrometry; LI: lacunar infarction; MF: molecular functions; PLS-DA: partial least squares discrimination analysis; PWMH: periventricular white matter hyperintensities; PPI: protein-protein interaction; ROC: receiver operating characteristic; STRING: Search Tool for the Retrieval of International Genes; SVM-RFE: support vector machine recursive feature elimination; WMH: white matter hyperintensities. Declarations Funding This work was supported by the National Natural Science Foundation of China to Dr. Yu-ming Xu [grant numbers 92249305]. Competing Interests The authors have no relevant financial or non-financial interests to disclose. Author Contributions YC. Wang, YM. Xu, JH. Yang conceived this study; YC. Wang and HH. Zhu designed wrote the first draft of the manuscript; YC. Wang and LC. He carried out sample collection and the clinical part of the study; YC. Wang processed the samples; L. Zhang and XL. Xue completed the mass spectrometry detection and retrieval.; HH. Zhu, LC. He and YT. Yao performed the statistical analyses; JY. Li, L.Z, JF. Chen, B.S, CH. Shi, YS. Li, Y. Gao, JH. Yang and YM. Xu reviewed and extensively edited the manuscript; All authors critically reviewed the manuscript and approved the final draft to be published. Data Availability All relevant data are described in the paper. Data can be requested from the corresponding author by all interested researchers. Ethics approval This study was approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University, China (No. 2021KY-0067-001). Consent to participate Informed consent was obtained from all individual participants included in the study. Consent to publish All co-authors approved the final version of the manuscript and agreed to submit it to Molecular Neurobiology. References Pantoni L. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9(7):689–701. Chen X, et al. Cerebral small vessel disease: neuroimaging markers and clinical implication. J Neurol. 2019;266(10):2347–62. Hachinski V, et al. Preventing dementia by preventing stroke: The Berlin Manifesto. Alzheimer's Dement J Alzheimer's Assoc. 2019;15(7):961–84. Hanash S. Disease proteomics. Nature. 2003;422(6928):226–32. Li X, Wang W, Chen J. Recent progress in mass spectrometry proteomics for biomedical research. Sci China Life Sci. 2017;60(10):1093–113. Aebersold R, Mann M. Mass-spectrometric exploration of proteome structure and function. Nature. 2016;537(7620):347–55. Mann M. Origins of mass spectrometry-based proteomics. Nat Rev Mol Cell Biol. 2016;17(11):678. Riley NM, Hebert AS, Coon JJ. Proteomics Moves into the Fast Lane. Cell Syst. 2016;2(3):142–3. de Jesus JR, et al. Depleting high-abundant and enriching low-abundant proteins in human serum: An evaluation of sample preparation methods using magnetic nanoparticle, chemical depletion and immunoaffinity techniques. Talanta. 2017;170:199–209. Millioni R, et al. High abundance proteins depletion vs low abundance proteins enrichment: comparison of methods to reduce the plasma proteome complexity. PLoS ONE. 2011;6(5):e19603. Staals J, et al. Stroke subtype, vascular risk factors, and total MRI brain small-vessel disease burden. Neurology. 2014;83(14):1228–34. Klarenbeek P, et al. Ambulatory blood pressure in patients with lacunar stroke: association with total MRI burden of cerebral small vessel disease. Stroke. 2013;44(11):2995–9. Duering M, et al. Neuroimaging standards for research into small vessel disease-advances since 2013. Lancet Neurol. 2023;22(7):602–18. Charidimou A, et al. White matter perivascular spaces: an MRI marker in pathology-proven cerebral amyloid angiopathy? Neurology. 2014;82(1):57–62. Greenberg SM, et al. Cerebral microbleeds: a guide to detection and interpretation. Lancet Neurol. 2009;8(2):165–74. Biffi A, et al. Aspirin and recurrent intracerebral hemorrhage in cerebral amyloid angiopathy. Neurology. 2010;75(8):693–8. von Elm E, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg (London England). 2014;12(12):1495–9. The Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res, 2019. 47(D1): pp. D330–8. Milacic M, et al. The Reactome Pathway Knowledgebase 2024. Nucleic Acids Res. 2024;52(D1):D672–8. Langfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559. Li Z, Sillanpää MJ. Overview of LASSO-related penalized regression methods for quantitative trait mapping and genomic selection. TAG. Theoretical and Applied Genetics. Theoretische Und Angewandte Genetik. 2012;125(3):419–35. Sanz H, et al. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018;19(1):432. Robin X, et al. pROC: an open-source package for R and S + to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77. Lee LC, Liong C-Y, Jemain AA. Partial least squares-discriminant analysis (PLS-DA) for classification of high-dimensional (HD) data: a review of contemporary practice strategies and knowledge gaps. Analyst. 2018;143(15):3526–39. Khoshneviszadeh M, et al. Microvascular damage, neuroinflammation and extracellular matrix remodeling in Col18a1 knockout mice as a model for early cerebral small vessel disease. Matrix Biology: J Int Soc Matrix Biology. 2024;128:39–64. Freedman BI, et al. APOL1 renal-risk variants associate with reduced cerebral white matter lesion volume and increased gray matter volume. Kidney Int. 2016;90(2):440–9. Alawieh A, et al. Modulation of post-stroke degenerative and regenerative processes and subacute protection by site-targeted inhibition of the alternative pathway of complement. J Neuroinflamm. 2015;12:247. Coulthard LG, Hawksworth OA, Woodruff TM. Complement: The Emerging Architect of the Developing Brain. Trends Neurosci. 2018;41(6):373–84. Stevens B, et al. The classical complement cascade mediates CNS synapse elimination. Cell. 2007;131(6):1164–78. Toutonji A, et al. Chronic complement dysregulation drives neuroinflammation after traumatic brain injury: a transcriptomic study. Acta Neuropathol Commun. 2021;9(1):126. Józsi M, et al. Factor H-related proteins determine complement-activating surfaces. Trends Immunol. 2015;36(6):374–84. Alawieh A, et al. Complement Drives Synaptic Degeneration and Progressive Cognitive Decline in the Chronic Phase after Traumatic Brain Injury. J Neuroscience: Official J Soc Neurosci. 2021;41(8):1830–43. Mastellos DC. Complement emerges as a masterful regulator of CNS homeostasis, neural synaptic plasticity and cognitive function. Exp Neurol. 2014;261:469–74. Hajishengallis G, Lambris JD. Crosstalk pathways between Toll-like receptors and the complement system. Trends Immunol. 2010;31(4):154–63. Carpanini SM, Torvell M, Morgan BP. Therapeutic Inhibition of the Complement System in Diseases of the Central Nervous System. Front Immunol. 2019;10:362. Dalakas MC, Alexopoulos H, Spaeth PJ. Complement in neurological disorders and emerging complement-targeted therapeutics. Nat Reviews Neurol. 2020;16(11):601–17. Alawieh A, et al. Acute Complement Inhibition Potentiates Neurorehabilitation and Enhances tPA-Mediated Neuroprotection. J Neuroscience: Official J Soc Neurosci. 2018;38(29):6527–45. Sonntag WE, et al. Insulin-like growth factor-1 in CNS and cerebrovascular aging. Front Aging Neurosci. 2013;5:27. Khan AS, et al. Growth hormone, insulin-like growth factor-1 and the aging cardiovascular system. Cardiovascular Res. 2002;54(1):25–35. Moloney AM, et al. Defects in IGF-1 receptor, insulin receptor and IRS-1/2 in Alzheimer's disease indicate possible resistance to IGF-1 and insulin signalling. Neurobiol Aging. 2010;31(2):224–43. Straub RH. Interaction of the endocrine system with inflammation: a function of energy and volume regulation. Arthritis Res Therapy. 2014;16(1):203. Straub RH, et al. Energy regulation and neuroendocrine-immune control in chronic inflammatory diseases. J Intern Med. 2010;267(6):543–60. Saukkonen T, et al. Effects of recombinant human IGF-I/IGF-binding protein-3 complex on glucose and glycerol metabolism in type 1 diabetes. Diabetes. 2006;55(8):2365–70. García-Fernández M, et al. Low doses of insulin-like growth factor I improve insulin resistance, lipid metabolism, and oxidative damage in aging rats. Endocrinology. 2008;149(5):2433–42. Bake S, et al. Blood brain barrier and neuroinflammation are critical targets of IGF-1-mediated neuroprotection in stroke for middle-aged female rats. PLoS ONE. 2014;9(3):e91427. Pirger Z et al. Reversal of age-related learning deficiency by the vertebrate PACAP and IGF-1 in a novel invertebrate model of aging: the pond snail (Lymnaea stagnalis). The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 2014. 69(11): pp. 1331–8. Goecks J et al. How Mach Learn Will Transform Biomed Cell, 2020. 181(1). Duan D, et al. Regional Gray Matter Atrophy Coexistent with Occipital Periventricular White Matter Hyper Intensities. Front Aging Neurosci. 2016;8:214. Yao D, et al. Association of Serum Cystatin C With Cerebral Small Vessel Disease in Community-Based Population. Stroke. 2022;53(10):3123–32. Xu J, Zhou H, Xiang G. Identification of Key Biomarkers and Pathways for Maintaining Cognitively Normal Brain Aging Based on Integrated Bioinformatics Analysis. Front Aging Neurosci. 2022;14:833402. Additional Declarations No competing interests reported. Supplementary Files SFigure1.tif Supplementary Figure 1. Data cleaning and standardization process. (A) Bar graph (left) and violin plots (right) representing the numbers of quantified proteins in individual serum samples CSVD and Control groups, respectively. (B) 766 proteins were quantified in this study, covering more than nine orders of magnitude of MS signal. (C) Cue plot of missing values filled proportionally. (D) Boxplot of protein normalization. (E) Density map of protein normalization. SFigure2.tif Supplementary Figure 2. Details of Two Machine Learning Algorithms. (A) Frequency bar plots of the top twelve ranked proteins in the LASSO algorithm. (B) Bar plot of AUC values when 3-12 protein features were selected in SVM-REF. (C-D) AUC, ACC, recall and precision values when 3-12 protein features were selected in LASSO (C) and SVM-REF (D). SFigure3.tif Supplementary Figure 3. Distribution of log2 abundance values of the six selected proteins. (A-F) Above is the boxplot of the log2 abundance values of the corresponding protein. The bold black line in the middle of the box is the median, the top and bottom of the box represent the upper and lower quartile values. Below are the receiver operating characteristic (ROC) curves for the proteins. SupplementaryTable1normalizeddata.xlsx SupplementaryTable2DEPslimma.xlsx SupplementaryTable3DEPsenrichmentanalyse.xlsx SupplementaryTable4WGCNAmoduleinformation.xlsx Cite Share Download PDF Status: Published Journal Publication published 11 Feb, 2025 Read the published version in Translational Stroke Research → Version 1 posted Editorial decision: Revision requested 19 Dec, 2024 Reviewers agreed at journal 18 Dec, 2024 Reviews received at journal 17 Dec, 2024 Reviewers agreed at journal 05 Dec, 2024 Reviews received at journal 02 Dec, 2024 Reviewers agreed at journal 22 Nov, 2024 Reviewers invited by journal 20 Nov, 2024 Editor assigned by journal 18 Nov, 2024 Submission checks completed at journal 14 Nov, 2024 First submitted to journal 12 Nov, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5439901","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":386920524,"identity":"c08aa5f5-26c3-432d-bdb6-83dbbdc76847","order_by":0,"name":"Yun-chao Wang","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yun-chao","middleName":"","lastName":"Wang","suffix":""},{"id":386920525,"identity":"42e01bc2-b828-468b-a51c-900a0acc7681","order_by":1,"name":"Hang-hang Zhu","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Hang-hang","middleName":"","lastName":"Zhu","suffix":""},{"id":386920526,"identity":"ceef70c9-328f-48db-b0ca-95fdae276798","order_by":2,"name":"Liu-Chang He","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Liu-Chang","middleName":"","lastName":"He","suffix":""},{"id":386920527,"identity":"ab3bf23e-8081-4adf-9b0c-b5b66a95b345","order_by":3,"name":"Ya-ting Yao","email":"","orcid":"","institution":"Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Ya-ting","middleName":"","lastName":"Yao","suffix":""},{"id":386920528,"identity":"3f801ef6-46a7-4521-bfca-6761bd91bebb","order_by":4,"name":"Lei Zhang","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Lei","middleName":"","lastName":"Zhang","suffix":""},{"id":386920529,"identity":"16671b7f-d28e-4fa5-942f-4cd2b4e65097","order_by":5,"name":"Xin-li Xue","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Xin-li","middleName":"","lastName":"Xue","suffix":""},{"id":386920530,"identity":"800de460-81a2-4b91-858c-62d9215996d6","order_by":6,"name":"Jing-yi Li","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Jing-yi","middleName":"","lastName":"Li","suffix":""},{"id":386920531,"identity":"77304b2c-0e87-4642-92e6-945462cca323","order_by":7,"name":"Li Zhang","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Li","middleName":"","lastName":"Zhang","suffix":""},{"id":386920532,"identity":"3a8c1218-88ba-4bbb-9e63-fba26a791a45","order_by":8,"name":"Jin-feng Chen","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Jin-feng","middleName":"","lastName":"Chen","suffix":""},{"id":386920533,"identity":"bd6ff3c2-fdef-4dd9-ab9f-5c402da2e6c9","order_by":9,"name":"Bo Song","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Bo","middleName":"","lastName":"Song","suffix":""},{"id":386920534,"identity":"dd03b1f0-f87d-494f-97fb-c3c0f88210e0","order_by":10,"name":"Chang-he Shi","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Chang-he","middleName":"","lastName":"Shi","suffix":""},{"id":386920535,"identity":"52ed70bc-0fb8-4358-a8c9-617bae9de2a9","order_by":11,"name":"Yu-sheng Li","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yu-sheng","middleName":"","lastName":"Li","suffix":""},{"id":386920536,"identity":"0d1ae85c-4347-4280-8475-0891f38ff401","order_by":12,"name":"Yuan Gao","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Yuan","middleName":"","lastName":"Gao","suffix":""},{"id":386920537,"identity":"899f4561-f571-42bc-8a0a-c09d4173db0a","order_by":13,"name":"Jing-hua Yang","email":"","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":false,"prefix":"","firstName":"Jing-hua","middleName":"","lastName":"Yang","suffix":""},{"id":386920538,"identity":"ced546bb-d7da-4b7e-b6b8-c04bb7054900","order_by":14,"name":"Yu-ming Xu","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0klEQVRIiWNgGAWjYNACAwYGfgiLmQQtkg2kaQHpOkCsFoMbOWZSNwru2G0+f/yZBEOFdWID+9kDeLVIzsgxk84xeJa87cCBNAmGM+mJDTx5CXi18EuAtRxONjvYcEyCse1wYoMEjwFeLWwwLcbNjG0SjP+I0AKzxc6AjZlNgrGBCC2SPc+KrYFaEiTOsDFbJBxLN27jycGvxeB48sbbOX8O2/P3H39440ONtWw/+xn8WhgYOMAKEhtAZALIdwTUAwH7AxBpT1jhKBgFo2AUjFgAABaNP7R4h5vMAAAAAElFTkSuQmCC","orcid":"","institution":"The First Affiliated Hospital of Zhengzhou University","correspondingAuthor":true,"prefix":"","firstName":"Yu-ming","middleName":"","lastName":"Xu","suffix":""}],"badges":[],"createdAt":"2024-11-12 13:23:16","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5439901/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5439901/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s12975-025-01332-6","type":"published","date":"2025-02-11T15:57:40+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":71730913,"identity":"9d9134ca-f360-494f-a3dd-8d2ecfbc4d74","added_by":"auto","created_at":"2024-12-18 06:50:51","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":204957,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eStudy overview and preliminary proteome analysis.\u003c/strong\u003e(A) Workflow for collecting and processing samples and analyzing proteomic data. (B) Partial least squares discrimination analysis plot showing protein sorting of CSVD versus control. (C) Venn diagram of the number of proteins examined in CSVD and control groups. (D) Volcano plot of DEPs (\u003cem\u003eP\u003c/em\u003e \u0026lt; 0.05) between CSVD and control. (E) The heatmap shows the serum levels of 52 DEPs.\u003c/p\u003e","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/63df727dfb4fd078b59e586b.png"},{"id":71730103,"identity":"59834727-0f0f-4053-b31d-a5813befcd4c","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":226950,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003ePPI network construction. \u003c/strong\u003e(A) Functional enrichment analysis of DEPs based on Gene ontology. (B) Functional enrichment analysis of DEPs based on Reactome. (C) PPI network constructed based on the interaction analysis of DEPs using STRING database. (D) The top 10 hub proteins identified.\u003c/p\u003e","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/60fad937b2ab90986607ffc0.png"},{"id":71730105,"identity":"fe67ab3c-470b-422e-976d-cbee4409b9e8","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":146879,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation analysis of clinical features and DEPs.\u003c/strong\u003e *, \u003cem\u003eP\u003c/em\u003e-value \u0026lt;0.05; **, \u003cem\u003eP\u003c/em\u003e-value\u0026lt; 0.01; and ***,\u003cem\u003e P\u003c/em\u003e-value \u0026lt; 0.001\u003c/p\u003e","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/1f312aa938d21bf4b7f774a4.png"},{"id":71730112,"identity":"8f6417d1-c3b6-4cc2-be6f-1ec344d6c89e","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":270085,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eWeighted co-expression network analysis. (\u003c/strong\u003eA\u003cstrong\u003e)\u003c/strong\u003e Scale-free fitting index graph and average connectivity graph. (B) Hierarchical clustering dendrogram of module identification. (C) DEPs in each module. (D) Correlation heatmap between modular features and clinical features. The enrichment analysis of proteins in the blue module using Gene ontology (E) and Reactome (F), respectively.\u003c/p\u003e","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/9f317df0215dbe14698a1cf7.png"},{"id":71730107,"identity":"9d5dcb5a-06f6-462f-8634-6984283db701","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":119773,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eMachine learning and diagnostic model. \u003c/strong\u003e(A) LASSO. (B) SVM-RFE. (C) Venn diagram of the best protein signatures identified by the LASSO and SVM-RFE algorithms and the top 10 hub genes identified by CytoHubba. (D) The receiver operating characteristic curves.\u003c/p\u003e","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/38ba20305478f23a8ff851dc.png"},{"id":71730110,"identity":"5f04a4bd-fd51-46a4-849a-d83fd1d644da","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":164090,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCorrelation analysis between serum concentrations of the six selected proteins and clinical features. \u003c/strong\u003e*,\u003cem\u003e P\u003c/em\u003e-value \u0026lt;0.05; **,\u003cem\u003e P\u003c/em\u003e-value\u0026lt; 0.01; and ***, \u003cem\u003eP\u003c/em\u003e-value \u0026lt; 0.001.\u003c/p\u003e","description":"","filename":"OnlineFigure6.png","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/0c8cad3ba5df84479387e026.png"},{"id":76487533,"identity":"8f60384b-5911-4e6a-8e5d-20a4833e525c","added_by":"auto","created_at":"2025-02-17 16:08:45","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":2666826,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/59bbad14-3224-4c51-891c-cd40feb73091.pdf"},{"id":71730914,"identity":"871b8eba-d535-41ab-a740-6b4572daeb7a","added_by":"auto","created_at":"2024-12-18 06:50:51","extension":"tif","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":15598372,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 1. Data cleaning and standardization process. \u003c/strong\u003e(A) Bar graph (left) and violin plots (right) representing the numbers of quantified proteins in individual serum samples CSVD and Control groups, respectively. (B) 766 proteins were quantified in this study, covering more than nine orders of magnitude of MS signal. (C) Cue plot of missing values filled proportionally. (D) Boxplot of protein normalization. (E) Density map of protein normalization.\u003c/p\u003e","description":"","filename":"SFigure1.tif","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/eff16f7122aa462c9999a469.tif"},{"id":71730108,"identity":"d67fd61f-dd01-4185-be32-268688e6f642","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"tif","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":16854472,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 2. Details of Two Machine Learning Algorithms. \u003c/strong\u003e(A) Frequency bar plots of the top twelve ranked proteins in the LASSO algorithm. (B) Bar plot of AUC values when 3-12 protein features were selected in SVM-REF. (C-D) AUC, ACC, recall and precision values when 3-12 protein features were selected in LASSO (C) and SVM-REF (D).\u003c/p\u003e","description":"","filename":"SFigure2.tif","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/8395160d934e82abf8f7eb8a.tif"},{"id":71730116,"identity":"876c9cfd-9a86-4865-9e9e-4b9876cc2a2e","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"tif","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":14796416,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eSupplementary Figure 3. Distribution of log2 abundance values of the six selected proteins. \u003c/strong\u003e(A-F) Above is the boxplot of the log2 abundance values of the corresponding protein. The bold black line in the middle of the box is the median, the top and bottom of the box represent the upper and lower quartile values. Below are the receiver operating characteristic (ROC) curves for the proteins.\u003c/p\u003e","description":"","filename":"SFigure3.tif","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/12f5131c9c95469ac1ecbe6d.tif"},{"id":71730106,"identity":"e7dd2c8f-8639-4acb-ab26-7126fe374fca","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":665478,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable1normalizeddata.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/ed6ef9b8a8db424442166e27.xlsx"},{"id":71730915,"identity":"dd9d52b4-349e-49f7-9992-ad54dfdee62e","added_by":"auto","created_at":"2024-12-18 06:50:51","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":10067,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable2DEPslimma.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/6c8cde70213c416d0d4bb5b9.xlsx"},{"id":71730115,"identity":"8118efa0-1610-4b64-84a7-8d88fce62b4e","added_by":"auto","created_at":"2024-12-18 06:42:51","extension":"xlsx","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":80620,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable3DEPsenrichmentanalyse.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/93a7014f25b0c550b49000a2.xlsx"},{"id":71730916,"identity":"e59a9ea5-a762-4fd2-85d6-75b25060d757","added_by":"auto","created_at":"2024-12-18 06:50:51","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":25352,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTable4WGCNAmoduleinformation.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-5439901/v1/a79ec6f7cfce522333e738d5.xlsx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Proteome Profiling of Serum Reveals Pathological Mechanisms and Biomarker Candidates for Cerebral Small Vessel Disease","fulltext":[{"header":"Introduction","content":"\u003cp\u003eCerebral small vessel disease (CSVD) is a syndrome with clinical, neuroimaging, and neuropathological manifestations in the brain[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. Computed tomography or magnetic resonance imaging (MRI) reveals CSVD as prominent in lacunar infarction (LI), white matter hyperintensities (WMH), cerebral microbleeds (CMB), MRI-visible enlarged perivascular spaces (EPVS), and atrophy[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. As a cerebrovascular disease, CSVD is characterized by ischemic and hemorrhagic strokes, gait and balance dysfunction, and behavioral and neuropsychiatric disorders, which impose a great social burden[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eHowever, the molecular pathophysiological mechanisms of CSVD have not been thoroughly investigated, and there are currently no specific interventions for this condition. It is highly likely to involve multiprotein interactions, resulting in a wide range of clinical phenotypes and neuropathological manifestations. Since proteins are the direct executors of most cellular functions and the proteome profile after extensive transcription and translation more closely resembles the final clinical phenotype[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e], a comprehensive and systematic exploration of the CSVD protein profile will further promote our understanding of the pathological mechanism of CSVD, thereby contributing to the discovery of new biomarkers and therapeutic targets. Liquid chromatography/tandem mass spectrometry (LC-MS/MS)-based proteomics, can specifically identify and quantify proteins in biological or clinical samples, thus being a potentially promising tool for the discovery of key molecular alterations[\u003cspan additionalcitationids=\"CR7\" citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eTo this end, we collected serum samples from 58 consecutively enrolled patients with CSVD and 38 controls and developed a unique serum proteome analysis workflow for detection[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. Through an in-depth and comprehensive analysis of the serum proteomics of patients with CSVD, we aimed to provide new insights into CSVD pathogenesis, contributing to the discovery of new targets and prevention and treatment strategies for the individualized clinical treatment of CSVD.\u003c/p\u003e"},{"header":"Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eParticipant recruitment\u003c/h2\u003e \u003cp\u003eA total of 96 adults were consecutively recruited from The First Affiliated Hospital of Zhengzhou University between 2018 and 2020. The inclusion criteria for the CSVD group (n\u0026thinsp;=\u0026thinsp;58) included age\u0026thinsp;\u0026ge;\u0026thinsp;45 years and total image burden score of CSVD\u0026thinsp;\u0026ge;\u0026thinsp;3 points[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. Conversely, the control group (n\u0026thinsp;=\u0026thinsp;38) comprised participants enrolled during the same period without any imaging abnormalities on brain MRI. The exclusion criteria for all patients were as follows: (1) cerebral infarction lesions (lesion diameter of diffusion-weighted imaging\u0026thinsp;\u0026gt;\u0026thinsp;20 mm), (2) acute cerebral or subarachnoid hemorrhage, (3) dementia due to confirmed neurodegenerative diseases, such as Alzheimer's or Parkinson's diseases, (4) evident white matter lesions of non-vascular origin, such as multiple sclerosis, adult white matter dysplasia, and metabolic encephalopathy, (5) mental diseases diagnosed according to the Diagnostic and Statistical Manual of Mental Disorders, and (6) intracranial infection, traumatic brain injury, or tumors. This study was approved by the Ethics Committee of The First Affiliated Hospital of Zhengzhou University.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eClinical data collection\u003c/h3\u003e\n\u003cp\u003eWe collected basic information from all participants, including demographic data, medical history, smoking history, and family history of stroke. Fasting venous blood samples were collected at baseline to detect homocysteine, total cholesterol, total triglyceride, high-density lipoprotein cholesterol, low-density lipoprotein cholesterol, fasting blood glucose (FBG) levels, and glycosylated hemoglobin (HbAlc) value.\u003c/p\u003e \u003cp\u003eIn addition, two experienced neurologists, who were blinded to the participants\u0026rsquo; clinical data, independently evaluated the MRI images of all participants and reached a consensus according to the Standards for Reporting Vascular Changes on Neuroimaging-2[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Periventricular and deep WMH (PWMH and DWMH, respectively) were assessed according to the 4-point Fazekas rating scale\u003csup\u003e11\u003c/sup\u003e. EPVS in the centrum semiovale and basal ganglia were assessed using a validated 4-point visual rating scale (0\u0026thinsp;=\u0026thinsp;no EPVS; 1\u0026thinsp;=\u0026thinsp;5\u0026ndash;10 EPVS; 2\u0026thinsp;=\u0026thinsp;11\u0026ndash;20 EPVS; 3\u0026thinsp;=\u0026thinsp;21\u0026ndash;40 EPVS; and 4\u0026thinsp;=\u0026thinsp;\u0026gt;\u0026thinsp;40 EPVS)[\u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e]. CMB presence and number were evaluated on susceptibility-weighted imaging according to current consensus criteria and scored using a 4-point scale (0\u0026thinsp;=\u0026thinsp;0; 1\u0026thinsp;=\u0026thinsp;2\u0026ndash;4; 2\u0026thinsp;=\u0026thinsp;5\u0026ndash;9; and 3\u0026thinsp;=\u0026thinsp;\u0026ge;\u0026thinsp;10)[\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e, \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e].\u003c/p\u003e\n\u003ch3\u003eSample processing, mass spectrometry detection, and retrieval\u003c/h3\u003e\n\u003cp\u003ePeripheral venous blood from all participants was collected and centrifuged for 10 min at 3200 rpm at 4\u0026deg;C; of this, 8 \u0026micro;L of serum from each enrolled participant was subjected to removal of HSA and immunoglobulin using the High-Select \u0026trade; HSA/Immunoglobulin Depletion Resin (Thermo Scientific\u0026trade;, A36369). The remaining proteins were then precipitated with acetone and redissolved using ultrasound in 25 mM ammonium bicarbonate. Dithiothreitol was added at a final concentration of 5 mmol. Trypsinization was performed twice for 12 h and 2 h. The peptides were desalted and eluted using a C18 column. Finally, label-free unlabeled quantitative proteomic technology combined with high-performance LC-MS/MS and Thermo QE-HFX mass spectrometry was used for peptide detection and analysis. The Sequest search engine of Proteome Discoverer 2.3 software was used to search for original data generated by the Q-Exactive HF-X mass spectrometer.\u003c/p\u003e\n\u003ch3\u003eData processing and statistical analysis\u003c/h3\u003e\n\u003cp\u003eStatistical and bioinformatic analyses were performed using R Studio and SPSS 22.0. All data analyses were performed under the guidance of professional statisticians and by following relevant guidelines and regulations[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. All abundance values were preprocessed using the R package \"edge R\" to facilitate the subsequent analysis. The R package \"limma\" was used to identify differentially expressed proteins (DEPs) between groups. Gene ontology (GO)[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e] and Reactome[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e] enrichment analyses were conducted using R packages. \u0026ldquo;Weighted co-expression network analysis was implemented by R package \"WGCNA\"[\u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. We correlated the clinical phenotype data with the protein expression data in the reconstructed sample clustering tree. Pearson\u0026rsquo;s correlation analysis was performed to estimate the correlation between module eigengenes and clinical features to identify the key module most relevant to CSVD. The Search Tool for the Retrieval of International Genes (STRING) database was used to analyze protein-protein interactions (PPI), and Cytoscape software (version 3.9.0) was used to process and visualize the resulting PPI network.\u003c/p\u003e \u003cp\u003eSPSS 22.0 was used for the statistical analysis of demographic and laboratory examination data. The measurement data that conformed to normal distribution were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;s) and analyzed by the independent sample t-test; the measurement data that did not conform to the normal distribution were expressed as medians and upper and lower quartiles [M(QL, QU)], and the nonparametric test was used; Categorical variables are expressed as quantities (n, %), for qualitative dichotomous variables, when T\u0026thinsp;\u0026ge;\u0026thinsp;5, we used Pearson\u0026rsquo;s chi-square test; when T\u0026thinsp;\u0026lt;\u0026thinsp;5 but T\u0026thinsp;\u0026ge;\u0026thinsp;1, we used the continuity-adjusted chi-square test; when T\u0026thinsp;\u0026lt;\u0026thinsp;1, we used Fisher\u0026rsquo;s exact test. \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was regarded for statistical significance.\u003c/p\u003e\n\u003ch3\u003eMachine learning\u003c/h3\u003e\n\u003cp\u003eMachine learning was used to filter the important features and build predictive diagnostic models. Standardized data were used to filter the characteristics, where 70% were randomly used as the training set and 30% as the validation set. We used the least absolute shrinkage and selection Operation (LASSO) algorithm to select important features by adjusting the penalty parameter with a 10-fold cross-validation[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. Simultaneously, the support vector machine recursive feature elimination (SVM-RFE) algorithm for searching lambda with minimal classification was also used for screening[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. The receiver operator characteristic curves were plotted and the area under the curve (AUC) was calculated using the R packages \"pROC\"[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. The AUC and accuracy (ACC) values were used to evaluate the accuracy and precision of the diagnostic model. To keep the model as simple as possible, the number of key genes was limited to 12. Finally, we used the intersection of the two algorithms for the subsequent analysis.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eValidation by enzyme-linked immunosorbent assay (ELISA)\u003c/h2\u003e \u003cp\u003eSamples from the two new-independent groups (CSVD and control) were used for ELISA verification to quantify the serum concentrations of the selected protein signatures. The participants were re-screened according to the eligibility criteria. ELISA kits for the following human proteins were used to measure serum protein changes in participants from the CSVD and control groups: coagulation factor Ⅸ (F9, CSB-E08443h), apolipoprotein B (APOB, SEKH-0515), tetranectin (CLEC3B, CSB-EL005531HU), cystatin C (Cys-C, CSB-E08384h), complement 1q (C1q, CSB-E10118h), and pantetheinase (VNN1, CSB-EL025883HU).\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec10\" class=\"Section2\"\u003e \u003ch2\u003eBaseline characteristics of the participants\u003c/h2\u003e \u003cp\u003eA workflow chart of the study is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-A. A total of 96 participants were enrolled in this study. These were consecutively enrolled participants who underwent complete medical history taking, magnetic resonance imaging, laboratory testing, and cognitive assessment within 2 days after admission. The demographic and clinical characteristics of the participants are summarized in Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. The mean age of the CSVD group was 65.21\u0026thinsp;\u0026plusmn;\u0026thinsp;9.11 years, and 42 (72.41%) participants were men. Compared with the control group, participants in the CSVD group tended to be older, male, and had a higher prevalence of hypertension, diabetes, and high levels of homocysteine, HbAlc, and FBG.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eClinical data of study participants.\u003c/b\u003e Demographics, clinical characteristics, and laboratory data of the participants of the cerebral small-vessel disease (CSVD) and control groups. The measurement data that conformed to the normal distribution are expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;s); the measurement data that did not conform to the normal distribution are expressed as medians and upper and lower quartiles [M (QL, QU)]. P\u0026thinsp;\u0026lt;\u0026thinsp;0.05 is considered statistically significant. * indicates a statistically significant difference.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"7\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003eTotal Count n\u0026thinsp;=\u0026thinsp;96\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003cp\u003e n\u0026thinsp;=\u0026thinsp;38\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c5\"\u003e \u003cp\u003eCSVD\u003c/p\u003e \u003cp\u003e n\u0026thinsp;=\u0026thinsp;58\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colspan=\"1\" nameend=\"c7\" namest=\"c7\"\u003e\u0026nbsp;\u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAge (year)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e60.75\u0026thinsp;\u0026plusmn;\u0026thinsp;9.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e53.95\u0026thinsp;\u0026plusmn;\u0026thinsp;6.44\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e65.21\u0026thinsp;\u0026plusmn;\u0026thinsp;9.11\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSex (male, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e55(57.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e13(34.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e42(72.41)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHypertension (n, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e53(55.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2(5.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e51(87.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDiabetes (n, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e15(15.63)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0(0)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e15(25.86)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCoronary Artery Disease, CAD (n %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e7(7.29)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2(5.26)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5(8.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.828\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eDyslipidemia (n, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e36(37.50)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e17(44.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e19(32.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.236\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFamily history of stroke (n, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e11(11.46)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e6(15.79)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5(8.62)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.281\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eSmoking (n, %)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e21(21.88)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5(13.16)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e16(27.59)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.094\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHomocysteine(\u0026micro;mol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e12.99(10.86,17.74)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e10.90(9.02,13.11)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e14.41(12.59,19.33)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e\u0026lt;0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eGlycated hemoglobin, HbAlc (%)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e5.90(5.58,6.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e5.70(5.50,5.93)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e6.0(5.70,6.43)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.001*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eFast blood glucose, FBG (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e4.92(4.60,5.45)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.83(4.58,5.00)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e5.12(4.65,5.82)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.018*\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTotal Cholesterol, TCHO (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e3.89\u0026thinsp;\u0026plusmn;\u0026thinsp;0.97\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e4.09\u0026thinsp;\u0026plusmn;\u0026thinsp;1.08\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e3.76\u0026thinsp;\u0026plusmn;\u0026thinsp;0.87\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.103\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eTriglyceride, TG (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e1.20(0.89,1.76)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.12(0.86,1.83)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.23(0.90,1.77)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.913\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eHigh-density Lipoprotein cholesterol, HDL-C (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e1.21\u0026thinsp;\u0026plusmn;\u0026thinsp;0.32\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e1.23\u0026thinsp;\u0026plusmn;\u0026thinsp;0.30\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e1.20\u0026thinsp;\u0026plusmn;\u0026thinsp;0.34\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.609\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eLow-density Lipoprotein cholesterol, LDL-C (mmol/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colspan=\"2\" nameend=\"c3\" namest=\"c2\"\u003e \u003cp\u003e2.27\u0026thinsp;\u0026plusmn;\u0026thinsp;0.83\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e2.44\u0026thinsp;\u0026plusmn;\u0026thinsp;0.94\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c5\"\u003e \u003cp\u003e2.16\u0026thinsp;\u0026plusmn;\u0026thinsp;0.74\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c6\"\u003e \u003cp\u003e0.110\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c7\"\u003e\u0026nbsp;\u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eProteome alterations in the sera of patients with CSVD and controls\u003c/h2\u003e \u003cp\u003eA total of 818 proteins were detected by mass spectrometry in 96 patients, and 766 proteins remained after the removal of contaminating proteins. Partial least squares discrimination analysis (PLS-DA), a supervised analysis method[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e], was used to evaluate the general differences in protein expression between and within groups. The results showed that the patients with CSVD and controls were divided into two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-B).\u003c/p\u003e \u003cp\u003eOf the 766 proteins obtained after filtration, 760 were found in the CSVD group and 755 in the control group, with 749 overlapping proteins between the two groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-C). We excluded proteins with more than 30% missing values and after conducting K-nearest neighbors filling and normalization, 524 proteins were retained for subsequent analysis (Supplementary Table\u0026nbsp;1 and Supplementary Fig.\u0026nbsp;1).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eIdentification and functional enrichment analysis of DEPs\u003c/h2\u003e \u003cp\u003eDifferential abundance analysis revealed that 52 differentially expressed proteins (DEPs) were significantly dysregulated in the CSVD group than in the control group, of which 23 were upregulated and 29 were downregulated. The volcano plot and heatmap in Figs.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e-D and E, respectively, show the protein expression trends in the CSVD and control groups. To gain insight into the biological significance of these DEPs, GO and Reactome enrichment analyses were performed using R. Figure\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA and Supplementary Table\u0026nbsp;3 show the top 10 terms with the smallest P-values for biological processes (BP), cellular components (CC), and molecular functions (MF), respectively. The results showed that the DEPs were mainly enriched in blood coagulation and hemostasis of BPs, collagen-containing extracellular matrix and blood microparticles of CCs, and peptidase regulator activity of MFs. In the Reactome pathway enrichment analysis (Supplementary Table\u0026nbsp;2), we identified five important biological pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e-B), including the regulation of insulin-like growth factor (IGF) transport and uptake by insulin-like growth factor-binding proteins (IGFBPs), platelet degranulation, and response to elevated platelet cytosolic Ca\u003csup\u003e2+\u003c/sup\u003e.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo determine the interactive relationships among the DEPs, we used the STRING database to construct a PPI network (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC). The results, visualized using Cytoscape, showed a strong interaction between the DEPs. The CytoHubba plug-in identified the top 10 hub proteins as plasminogen (PLG), antithrombin-III (SERPINC1), prothrombin (F2), APOB, kininogen-1 (KNG1), serum amyloid P component (APCS), complement component 9 (C9), coagulation factor IX (F9), alpha-2 antiplasmin (SERPINF2), and inter-alpha-trypsin inhibitor heavy chain H1 (ITIH1) (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD).\u003c/p\u003e \u003cp\u003eWe evaluated the correlation coefficients of 25 recorded clinical features and 52 DEP levels in both groups to demonstrate their relationship (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). The results indicated that most DEPs were significantly associated with the disease group, imaging findings, and cognitive function. Among them, the severity of imaging abnormalities and cognitive impairment was positively correlated with upregulated DEPs but negatively correlated with downregulated DEPs (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eWeighted co-expression network analysis\u003c/h2\u003e \u003cp\u003eA weighted co-expression network analysis was performed using the \"WGCNA\" package to further screen for key proteins related to CSVD clinical features. All proteins were categorized into six co-expression modules, including the gray module (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). The upregulation and downregulation of differentially expressed proteins in each module are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC. The correlation between the modules and clinical features was measured using the correlation between the module eigengene values and clinical features (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD). The blue module contained the largest proportion of DEPs (13/78) and was significantly associated with multiple clinical features (group, age, sex, PWMH, total WMH score, LI, Montreal cognitive assessment score, Alzheimer's disease-8 (AD-8) score, hypertension, and hyperlipidemia); hence, we further identified and analyzed the proteins within. GO and Reactome analyses showed that the terms associated with the blue module proteins were consistent with the enrichment analysis results of DEPs between CSVD and controls (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE and \u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF), including the complement cascade reaction, coagulation regulation, and the regulation of IGF transport and uptake by IGFBPs. This confirmed that these pathways are likely to play important roles in the course of CSVD (Supplementary Table\u0026nbsp;4).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eMachine learning\u003c/h2\u003e \u003cp\u003eWe constructed receiver operating characteristic (ROC) curves based on the validation set to determine the optimal protein signatures. The results of the LASSO algorithm showed that the model had the best diagnostic performance (AUC\u0026thinsp;=\u0026thinsp;0.9893, ACC\u0026thinsp;=\u0026thinsp;0.9642) when a protein signature with the following top 11 proteins was selected: CLEC3B, lipocalin prostaglandin D synthetase (PTGDS), APOB, F9, cystatin C (CST3), a disintegrin and metalloproteinase with thrombospondin motifs-like 4 (ADAMTSL4), alpha chain of type XVIII collagen (COL18A1), apolipoprotein L1 (APOL1), fetuin-B (FETUB), complement C1q subcomponent subunit B (C1QB), and VNN1 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA and Supplementary Fig.\u0026nbsp;2). In contrast, the SVM-RFE model performed the best (AUC\u0026thinsp;=\u0026thinsp;0.9091, ACC\u0026thinsp;=\u0026thinsp;0. 0.8929) when a protein signature with the following top 10 proteins was selected: eukaryotic translation initiation factor 4 gamma 3 (EIF4G3), VNN1, F9, CST3, C1QB, APOB, keratin 1 (KRT1), CLEC3B, KRT5, and paraoxonase-1 (PON1) (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB and Supplementary Fig.\u0026nbsp;2). Among these, F9, APOB, CST3, C1QB, CLEC3B, and VNN1 overlapped in the best diagnostic models constructed using the two algorithms, and F9 and APOB were identified as the top 10 hub proteins in the PPI network (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eC and Supplementary Fig.\u0026nbsp;3).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003eTo test the diagnostic values of the six overlapping proteins, we conducted ROC analyses and their combination served as a good model to distinguish patients with CSVD from controls, with an AUC of 0.929 and 95% confidence interval (CI) was 0.876\u0026ndash;0.971 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD Model 1). In addition, considering the age-related characteristics of CSVD, when age was added as a variable in the new model, we found that the AUC of the new model reached 0.947 and the 95%CI was 0.895\u0026ndash;0.978 (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eD Model 2). Furthermore, we noted that these protein signatures were also localized to the major pathways identified above, such as regulation of IGF transport and uptake by IGFBPs, complement, and the coagulation cascade.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eValidation of protein signatures performance by ELISA\u003c/h2\u003e \u003cp\u003eTo examine the diagnostic performance of the protein signature with six proteins (F9, APOB, CST3, C1QB, CLEC3B, and VNN1) and the diagnostic models, we further validated using ELISA and sera from the two groups. In the CSVD group, C1QB, CST3, and F9 levels showed an upward trend, whereas CLEC3B level showed a downward trend. This was consistent with the proteomics results, and the differences between the two groups were statistically significant. However, the expression of the other two proteins (APOB and VNN1) was not significantly different between the two groups (Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). This may be related to the experimental method or limited sample size. Subsequently, we correlated the serum concentration of these six proteins with the clinical characteristics of the samples (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e). The general trend of correlation between six selected proteins and grouping was similar to the proteomic results; however, expression of only C1QB, CLEC3B, CST3, and F9 was statistically significant and correlated with the severity of LI and CMB. Furthermore, the concentrations of C1QB, CST3, and F9 were significantly positively correlated with the severity of total WMH and EPVS.\u003c/p\u003e \u003cp\u003e \u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e \u003ccaption language=\"En\"\u003e \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e \u003cdiv class=\"CaptionContent\"\u003e \u003cp\u003e\u003cb\u003eValidation results of enzyme-linked immunosorbent assay (ELISA) for six selected proteins.\u003c/b\u003e Results for the determination of concentrations of six selected proteins in serum samples obtained from 27 patients with cerebral small-vessel disease (CSVD) and 27 control individuals using ELISA. The measurement data that conformed to the normal distribution were expressed as mean\u0026thinsp;\u0026plusmn;\u0026thinsp;standard deviation (mean\u0026thinsp;\u0026plusmn;\u0026thinsp;s); the measurement data that did not conform to the normal distribution were expressed as medians and upper and lower quartiles [M (QL, QU)]. \u003cem\u003eP\u003c/em\u003e\u0026thinsp;\u0026lt;\u0026thinsp;0.05 is considered statistically significant. * indicates a statistically significant difference.\u003c/p\u003e \u003c/div\u003e \u003c/caption\u003e \u003ccolgroup cols=\"4\"\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e \u003cdiv align=\"left\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e \u003cthead\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c1\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003eVariables\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003eControl\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003eCSVD\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c4\" morerows=\"1\" rowspan=\"2\"\u003e \u003cp\u003e\u003cem\u003eP\u003c/em\u003e value\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003ctr\u003e \u003cth align=\"left\" colname=\"c2\"\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;27)\u003c/p\u003e \u003c/th\u003e \u003cth align=\"left\" colname=\"c3\"\u003e \u003cp\u003e(n\u0026thinsp;=\u0026thinsp;27)\u003c/p\u003e \u003c/th\u003e \u003c/tr\u003e \u003c/thead\u003e \u003ctbody\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eAPOB (ng/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e168.14(129.91,370.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e163.15(136.56,219.65)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.829\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eC1QB (\u0026micro;g/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e28.63(26.13,33.94)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e35.47(30.23,42.21)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e\u0026lt;\u0026thinsp;0.001*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCLEC3B (mg/L)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e5.83(5.00,6.89)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e4.91(3.99,6.38)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.032*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eCST3 (\u0026micro;g/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e0.97(0.93,1.03)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e1.05(0.97,1.10)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.013*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eF9 (ng/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e31.18(25.67,37.99)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e33.94(28.98,50.12)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.039*\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003ctr\u003e \u003ctd align=\"left\" colname=\"c1\"\u003e \u003cp\u003eVNN1 (pg/ml)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c2\"\u003e \u003cp\u003e283.90(102.88,439.06)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c3\"\u003e \u003cp\u003e294.25(102.88,506.30)\u003c/p\u003e \u003c/td\u003e \u003ctd align=\"left\" colname=\"c4\"\u003e \u003cp\u003e0.672\u003c/p\u003e \u003c/td\u003e \u003c/tr\u003e \u003c/tbody\u003e \u003c/colgroup\u003e \u003c/table\u003e\u003c/div\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eCSVD is complex as it involves several underlying cellular and molecular mechanisms and varies in clinical manifestations, leading to difficulties in early diagnosis and specific treatment. In this study, we identified 52 DEPs between the CSVD and control groups that were related to the severity of the CSVD imaging burden and cognitive impairment. Notably, many DEPs have been identified as potential CSVD markers in many studies. For example, COL18A1, up-regulated in the CSVD group, was confirmed to be involved in (micro) vascular wall pathology and remodeling in autopsied brain tissue from patients with CSVD[\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e]. Therefore, to some extent, our differential protein expression profile in CSVD confirmed previous findings and provides a direction and inspiration for future research. Furthermore, functional enrichment, protein interaction, and weighted co-expression network analyses showed that these DEPs may be involved in CSVD by participating in the overactivation of complement and coagulation cascades and dysregulation of insulin-like growth factor-binding proteins on insulin-like growth factor. Furthermore, we identified a protein signature of six proteins in CSVD using the LASSO and SVM-RFE algorithms, and verified them experimentally.\u003c/p\u003e \u003cp\u003eFirst, the complement cascade plays an important role in maintaining healthy brain homeostasis, contributing to the removal of invading pathogens and apoptotic cells, pruning inappropriate synapses, and limiting inflammatory immune responses[\u003cspan additionalcitationids=\"CR28\" citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e], and when unregulated, has adverse effects, often exacerbating disease[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]; therefore, the altered expression of complement components and complement regulatory proteins in our CSVD group may be of great significance. The upregulation of CIQB, C9, FCN2, and CFHR3 and downregulation of complement factor I in the CSVD group may indicate overactivation or under regulation of the complement cascade[\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e], which may participate in the pathological processes of CSVD. Several studies have shown that persistent chronic complement activation can drive a robust neuroinflammatory response associated with synaptic degeneration and progressive cognitive decline[\u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. The specific mechanisms may include: 1) inducing neutrophil infiltration and increasing the secretion of pro-inflammatory cytokines, leading to the destruction of cell homeostasis and tissue damage; 2) activation of microglial phagocytosis of synapses, leading to synaptic loss and neuronal death, thereby destroying the integrity of functional neural circuits and affecting cognitive functions[\u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e]; and 3) aggravating the original hypoxic-ischemic injury by increasing complement activation products, including opsonins, anaphylatoxins, and the membrane attack complex, ultimately delaying nerve repair and worsening prognosis[\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e, \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e34\u003c/span\u003e]. Thus, we conclude that a better understanding of the process of complement involvement in CSVD may help in the prevention and monitoring of CSVD, and interception of the complement cascade may be a potential therapeutic modality. In this regard, emerging complement-targeted therapeutics for neurological diseases can provide ideas and references[\u003cspan additionalcitationids=\"CR36\" citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eSecond, compared to the overactivation of the complement system, the regulatory system of IGF by IGFBPs may be downregulated in the CSVD group. IGFs, regulated by six high-affinity IGFBPs, play essential roles in the regulation of growth, development, homeostasis, and neuroplastic changes in the brain, and both the vascular and nervous systems of the brain are their important targets[\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. Clinical evidence and animal studies have shown that aging can lead to the downregulation of insulin and IGF/IGFBP signaling, which can lead to cerebrovascular disease and age-related cognitive impairment[\u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. Therefore, we speculate that downregulation of the IGF/IGFBP system is a key trigger for the development and progression of CSVD during aging, after which multiple downstream pathways are involved. Since inflammation may also induce a wide response in the neuroendocrine system[\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e], our alternative hypothesis is that the inhibition of IGF/IGFBPs is related to chronic inflammation in CSVD; however, the causal relationship between the two is unknown. Notably, although increasing evidence suggests that the IGF/IGFBPs system has many downstream pathways and that it plays an important role in lipid metabolism and glucose metabolism[\u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e, \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e44\u003c/span\u003e], our data suggest that the major abnormality occurring in CSVD is associated with lipid metabolism (Supplementary Table\u0026nbsp;3), such as cholesterol metabolic process (APOB, PON1, APOL1, and proprotein convertase subtilisin/kexin type 9 [PCSK9]), lipoprotein metabolic process (APOB, APOL1, and PCSK9), and glycerolipid metabolic process (APOB, PON1, and PCSK9). However, the clinical laboratory results did not show significantly abnormal lipid metabolism in the CSVD group (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e), which may be related to the widespread use of statins in patients with CSVD. In combination with previous studies, we propose that manipulating the regulatory activity of IGFBP on IGF has the potential to become a therapeutic strategy for CSVD by reducing dyslipidemia and insulin resistance[\u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e45\u003c/span\u003e, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e46\u003c/span\u003e].\u003c/p\u003e \u003cp\u003eThird, we identified the protein signature of six proteins. Machine learning is a key productivity tool in modern omics research that uses specific modeling and prediction methods to identify patterns in high-throughput sequencing datasets[\u003cspan citationid=\"CR47\" class=\"CitationRef\"\u003e47\u003c/span\u003e]. In our study, by integrating the 11 proteins obtained after LASSO screening and 10 proteins obtained after SVM-RFE screening, we identified a protein signature of six proteins (F9, APOB, CST3, C1QB, CLEC3B, and VNN1), some of which have already been implicated in CSVD, confirming previously published findings. Specifically, reduced APOB expression has previously been associated with occipital periventricular WMH[\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e48\u003c/span\u003e] and an increased WMH score in the current correlation study (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). In a community-based study, CST3 was associated with CSVD, and Mendelian randomization analysis showed that a genetically predicted higher CST3 level was associated with an increased risk of lacunar stroke[\u003cspan citationid=\"CR49\" class=\"CitationRef\"\u003e49\u003c/span\u003e]. C1QB is a key gene in healthy brain aging, whose activation in the brain may contribute to the progression of age-related cognitive dysfunction[\u003cspan citationid=\"CR50\" class=\"CitationRef\"\u003e50\u003c/span\u003e]. Correspondingly, the expression of CST3 and C1QB was upregulated in the CSVD group. However, F9, CLEC3B, and VNN1 are potential new biomarkers that need to be identified in future studies. The combination of these six proteins is a good model for the diagnosis of CSVD, and the diagnostic performance of this combination can be made more robust by including age as a variable.\u003c/p\u003e \u003cp\u003eOur study had a few limitations. First, because this was a cross-sectional study, we cannot conclude causality or explore whether the abnormalities demonstrated in our results are pathogenic or compensatory responses to CSVD. Second, Further investigation is warranted to explore the mechanisms by which the proteins and pathways identified in our study affect CSVD, as well as targeted therapeutic approaches.\u003c/p\u003e \u003cp\u003eWe believe that our study makes a significant contribution to the literature because this is the most in-depth study on CSVD proteomics to date, and suggests that the overactivation of the complement cascade and the dysregulation of IGFBP on- IGF may be closely correlated with the occurrence and progression of CSVD, offering the potential to develop peripheral blood biomarkers and providing new insights into the biological basis of CSVD.\u003c/p\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eStatements \u0026amp; Declarations\u003c/h2\u003e \u003c/div\u003e"},{"header":"Abbreviations","content":"\u003cp\u003e\u003cstrong\u003eBP:\u0026nbsp;\u003c/strong\u003ebiological processes; \u003cstrong\u003eCC:\u0026nbsp;\u003c/strong\u003ecellular components; \u003cstrong\u003eCMB:\u0026nbsp;\u003c/strong\u003ecerebral microbleeds; \u003cstrong\u003eCSVD:\u0026nbsp;\u003c/strong\u003ecerebral small-vessel disease; \u003cstrong\u003eDEPs:\u0026nbsp;\u003c/strong\u003edifferentially expressed proteins; \u003cstrong\u003eDWMH:\u0026nbsp;\u003c/strong\u003edeep white matter hyperintensities; \u003cstrong\u003eELISA:\u0026nbsp;\u003c/strong\u003eenzyme-linked immunosorbent assay; \u003cstrong\u003eGO:\u0026nbsp;\u003c/strong\u003eGene ontology; \u003cstrong\u003eIGF:\u0026nbsp;\u003c/strong\u003einsulin-like growth factor; \u003cstrong\u003eIGFBP:\u0026nbsp;\u003c/strong\u003einsulin-like growth factor-binding proteins; \u003cstrong\u003eLASSO:\u0026nbsp;\u003c/strong\u003eleast absolute shrinkage and selection operator; \u003cstrong\u003eLC-MS/MS:\u0026nbsp;\u003c/strong\u003eliquid chromatography/tandem mass spectrometry; \u003cstrong\u003eLI:\u0026nbsp;\u003c/strong\u003elacunar infarction; \u003cstrong\u003eMF:\u0026nbsp;\u003c/strong\u003emolecular functions; \u003cstrong\u003ePLS-DA:\u0026nbsp;\u003c/strong\u003epartial least squares discrimination analysis; \u003cstrong\u003ePWMH:\u0026nbsp;\u003c/strong\u003eperiventricular white matter hyperintensities; \u003cstrong\u003ePPI:\u0026nbsp;\u003c/strong\u003eprotein-protein interaction;\u003cstrong\u003e\u0026nbsp;ROC:\u0026nbsp;\u003c/strong\u003ereceiver operating characteristic; \u003cstrong\u003eSTRING:\u0026nbsp;\u003c/strong\u003eSearch Tool for the Retrieval of International Genes; \u003cstrong\u003eSVM-RFE:\u0026nbsp;\u003c/strong\u003esupport vector machine recursive feature elimination; \u003cstrong\u003eWMH:\u0026nbsp;\u003c/strong\u003ewhite matter hyperintensities.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Natural Science Foundation of China to Dr. Yu-ming Xu [grant numbers 92249305].\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eYC. Wang, YM. Xu, JH. Yang conceived this study; YC. Wang and HH. Zhu designed wrote the first draft of the manuscript; YC. Wang and LC. He carried out sample collection and the clinical part of the study; YC. Wang processed the samples; L. Zhang and XL. Xue completed the mass spectrometry detection and retrieval.; HH. Zhu, LC. He and YT. Yao performed the statistical analyses; JY. Li, L.Z, JF. Chen, B.S, CH. Shi, YS. Li, Y. Gao, JH. Yang and YM. Xu reviewed and extensively edited the manuscript; All authors critically reviewed the manuscript and approved the final draft to be published.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll relevant data are described in the paper. Data can be requested from the corresponding author by all interested researchers.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis study was approved by the Ethics Committee of the First Affiliated Hospital of Zhengzhou University, China (No. 2021KY-0067-001).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInformed consent was obtained from all individual participants included in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to publish\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll co-authors approved the final version of the manuscript and agreed to submit it to Molecular Neurobiology.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003ePantoni L. Cerebral small vessel disease: from pathogenesis and clinical characteristics to therapeutic challenges. Lancet Neurol. 2010;9(7):689\u0026ndash;701.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eChen X, et al. Cerebral small vessel disease: neuroimaging markers and clinical implication. J Neurol. 2019;266(10):2347\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHachinski V, et al. Preventing dementia by preventing stroke: The Berlin Manifesto. Alzheimer's Dement J Alzheimer's Assoc. 2019;15(7):961\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHanash S. Disease proteomics. Nature. 2003;422(6928):226\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi X, Wang W, Chen J. Recent progress in mass spectrometry proteomics for biomedical research. Sci China Life Sci. 2017;60(10):1093\u0026ndash;113.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAebersold R, Mann M. Mass-spectrometric exploration of proteome structure and function. Nature. 2016;537(7620):347\u0026ndash;55.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMann M. Origins of mass spectrometry-based proteomics. Nat Rev Mol Cell Biol. 2016;17(11):678.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRiley NM, Hebert AS, Coon JJ. Proteomics Moves into the Fast Lane. Cell Syst. 2016;2(3):142\u0026ndash;3.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ede Jesus JR, et al. Depleting high-abundant and enriching low-abundant proteins in human serum: An evaluation of sample preparation methods using magnetic nanoparticle, chemical depletion and immunoaffinity techniques. Talanta. 2017;170:199\u0026ndash;209.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMillioni R, et al. High abundance proteins depletion vs low abundance proteins enrichment: comparison of methods to reduce the plasma proteome complexity. PLoS ONE. 2011;6(5):e19603.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStaals J, et al. Stroke subtype, vascular risk factors, and total MRI brain small-vessel disease burden. Neurology. 2014;83(14):1228\u0026ndash;34.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKlarenbeek P, et al. Ambulatory blood pressure in patients with lacunar stroke: association with total MRI burden of cerebral small vessel disease. Stroke. 2013;44(11):2995\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuering M, et al. Neuroimaging standards for research into small vessel disease-advances since 2013. Lancet Neurol. 2023;22(7):602\u0026ndash;18.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCharidimou A, et al. White matter perivascular spaces: an MRI marker in pathology-proven cerebral amyloid angiopathy? Neurology. 2014;82(1):57\u0026ndash;62.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGreenberg SM, et al. Cerebral microbleeds: a guide to detection and interpretation. Lancet Neurol. 2009;8(2):165\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBiffi A, et al. Aspirin and recurrent intracerebral hemorrhage in cerebral amyloid angiopathy. Neurology. 2010;75(8):693\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003evon Elm E, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg (London England). 2014;12(12):1495\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eThe Gene Ontology Resource: 20 years and still GOing strong. Nucleic Acids Res, 2019. 47(D1): pp. D330\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMilacic M, et al. The Reactome Pathway Knowledgebase 2024. Nucleic Acids Res. 2024;52(D1):D672\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLangfelder P, Horvath S. WGCNA: an R package for weighted correlation network analysis. BMC Bioinformatics. 2008;9:559.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLi Z, Sillanp\u0026auml;\u0026auml; MJ. \u003cem\u003eOverview of LASSO-related penalized regression methods for quantitative trait mapping and genomic selection.\u003c/em\u003e TAG. Theoretical and Applied Genetics. Theoretische Und Angewandte Genetik. 2012;125(3):419\u0026ndash;35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSanz H, et al. SVM-RFE: selection and visualization of the most relevant features through non-linear kernels. BMC Bioinformatics. 2018;19(1):432.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eRobin X, et al. pROC: an open-source package for R and S\u0026thinsp;+\u0026thinsp;to analyze and compare ROC curves. BMC Bioinformatics. 2011;12:77.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eLee LC, Liong C-Y, Jemain AA. Partial least squares-discriminant analysis (PLS-DA) for classification of high-dimensional (HD) data: a review of contemporary practice strategies and knowledge gaps. Analyst. 2018;143(15):3526\u0026ndash;39.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhoshneviszadeh M, et al. Microvascular damage, neuroinflammation and extracellular matrix remodeling in Col18a1 knockout mice as a model for early cerebral small vessel disease. Matrix Biology: J Int Soc Matrix Biology. 2024;128:39\u0026ndash;64.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eFreedman BI, et al. APOL1 renal-risk variants associate with reduced cerebral white matter lesion volume and increased gray matter volume. Kidney Int. 2016;90(2):440\u0026ndash;9.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlawieh A, et al. Modulation of post-stroke degenerative and regenerative processes and subacute protection by site-targeted inhibition of the alternative pathway of complement. J Neuroinflamm. 2015;12:247.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCoulthard LG, Hawksworth OA, Woodruff TM. Complement: The Emerging Architect of the Developing Brain. Trends Neurosci. 2018;41(6):373\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStevens B, et al. The classical complement cascade mediates CNS synapse elimination. Cell. 2007;131(6):1164\u0026ndash;78.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eToutonji A, et al. Chronic complement dysregulation drives neuroinflammation after traumatic brain injury: a transcriptomic study. Acta Neuropathol Commun. 2021;9(1):126.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eJ\u0026oacute;zsi M, et al. Factor H-related proteins determine complement-activating surfaces. Trends Immunol. 2015;36(6):374\u0026ndash;84.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlawieh A, et al. Complement Drives Synaptic Degeneration and Progressive Cognitive Decline in the Chronic Phase after Traumatic Brain Injury. J Neuroscience: Official J Soc Neurosci. 2021;41(8):1830\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMastellos DC. Complement emerges as a masterful regulator of CNS homeostasis, neural synaptic plasticity and cognitive function. Exp Neurol. 2014;261:469\u0026ndash;74.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eHajishengallis G, Lambris JD. Crosstalk pathways between Toll-like receptors and the complement system. Trends Immunol. 2010;31(4):154\u0026ndash;63.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eCarpanini SM, Torvell M, Morgan BP. Therapeutic Inhibition of the Complement System in Diseases of the Central Nervous System. Front Immunol. 2019;10:362.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDalakas MC, Alexopoulos H, Spaeth PJ. Complement in neurological disorders and emerging complement-targeted therapeutics. Nat Reviews Neurol. 2020;16(11):601\u0026ndash;17.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eAlawieh A, et al. Acute Complement Inhibition Potentiates Neurorehabilitation and Enhances tPA-Mediated Neuroprotection. J Neuroscience: Official J Soc Neurosci. 2018;38(29):6527\u0026ndash;45.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSonntag WE, et al. Insulin-like growth factor-1 in CNS and cerebrovascular aging. Front Aging Neurosci. 2013;5:27.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eKhan AS, et al. Growth hormone, insulin-like growth factor-1 and the aging cardiovascular system. Cardiovascular Res. 2002;54(1):25\u0026ndash;35.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eMoloney AM, et al. Defects in IGF-1 receptor, insulin receptor and IRS-1/2 in Alzheimer's disease indicate possible resistance to IGF-1 and insulin signalling. Neurobiol Aging. 2010;31(2):224\u0026ndash;43.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStraub RH. Interaction of the endocrine system with inflammation: a function of energy and volume regulation. Arthritis Res Therapy. 2014;16(1):203.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eStraub RH, et al. Energy regulation and neuroendocrine-immune control in chronic inflammatory diseases. J Intern Med. 2010;267(6):543\u0026ndash;60.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eSaukkonen T, et al. Effects of recombinant human IGF-I/IGF-binding protein-3 complex on glucose and glycerol metabolism in type 1 diabetes. Diabetes. 2006;55(8):2365\u0026ndash;70.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGarc\u0026iacute;a-Fern\u0026aacute;ndez M, et al. Low doses of insulin-like growth factor I improve insulin resistance, lipid metabolism, and oxidative damage in aging rats. Endocrinology. 2008;149(5):2433\u0026ndash;42.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eBake S, et al. Blood brain barrier and neuroinflammation are critical targets of IGF-1-mediated neuroprotection in stroke for middle-aged female rats. PLoS ONE. 2014;9(3):e91427.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003ePirger Z et al. \u003cem\u003eReversal of age-related learning deficiency by the vertebrate PACAP and IGF-1 in a novel invertebrate model of aging: the pond snail (Lymnaea stagnalis).\u003c/em\u003e The Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 2014. 69(11): pp. 1331\u0026ndash;8.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eGoecks J et al. How Mach Learn Will Transform Biomed Cell, 2020. 181(1).\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eDuan D, et al. Regional Gray Matter Atrophy Coexistent with Occipital Periventricular White Matter Hyper Intensities. Front Aging Neurosci. 2016;8:214.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eYao D, et al. Association of Serum Cystatin C With Cerebral Small Vessel Disease in Community-Based Population. Stroke. 2022;53(10):3123\u0026ndash;32.\u003c/span\u003e\u003c/li\u003e \u003cli\u003e\u003cspan\u003eXu J, Zhou H, Xiang G. Identification of Key Biomarkers and Pathways for Maintaining Cognitively Normal Brain Aging Based on Integrated Bioinformatics Analysis. Front Aging Neurosci. 2022;14:833402.\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"translational-stroke-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"trsr","sideBox":"Learn more about [Translational Stroke Research](http://jcmr-online.biomedcentral.com)","snPcode":"12975","submissionUrl":"https://submission.nature.com/new-submission/12975/3","title":"Translational Stroke Research","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"cerebral small vascular disease, complement cascade, insulin-like growth factor, proteomics","lastPublishedDoi":"10.21203/rs.3.rs-5439901/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5439901/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground\u003c/h2\u003e \u003cp\u003eCerebral small vessel disease (CSVD) is a global brain disorder that is characterized by a series of clinical, neuroimaging, and neuropathological manifestations. However, the molecular pathophysiological mechanisms of CSVD have not been thoroughly investigated. Liquid chromatography-tandem mass spectrometry-based proteomics has broad application prospects in biomedicine. It is used to elucidate disease-related molecular processes and pathophysiological pathways, thus providing an important opportunity to explore the pathophysiological mechanisms of CSVD.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003e Serum samples were obtained from 96 participants (58 with CSVD and 38 controls) consecutively recruited from The First Affiliated Hospital of Zhengzhou University. After removing high-abundance proteins, the serum samples were analyzed using high-resolution mass spectrometry. Bioinformatics methods were used for in-depth analysis of the obtained proteomic data, and the results were verified experimentally.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eCompared with the control group, 52 proteins were differentially expressed in the sera of the CSVD group. Furthermore, analyses indicated the involvement of these differentially expressed proteins in CSVD through participation in the overactivation of complement and coagulation cascades and dysregulation of insulin-like growth factor-binding proteins. The proteomic biomarker panel identified by the machine learning model combined with clinical features is expected to facilitate the diagnosis of CSVD (AUC\u0026thinsp;=\u0026thinsp;0.947,95% CI\u0026thinsp;=\u0026thinsp;0.895\u0026ndash;0.978).\u003c/p\u003e\u003ch2\u003eConclusions\u003c/h2\u003e \u003cp\u003eThe study is the most in-depth study on CSVD proteomics to date, and suggests that the overactivation of the complement cascade and the dysregulation of IGFBP on- IGF may be closely correlated with the occurrence and progression of CSVD, offering the potential to develop peripheral blood biomarkers and providing new insights into the biological basis of CSVD.\u003c/p\u003e","manuscriptTitle":"Proteome Profiling of Serum Reveals Pathological Mechanisms and Biomarker Candidates for Cerebral Small Vessel Disease","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-12-18 06:42:46","doi":"10.21203/rs.3.rs-5439901/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-12-19T06:58:43+00:00","index":"","fulltext":""},{"type":"reviewerAgreed","content":"152762654193200416273091786109753040467","date":"2024-12-18T10:08:41+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-17T12:59:22+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"220993869537707938589233938364943506993","date":"2024-12-05T13:51:46+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-02T06:45:28+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"39437795345570357663342725690800193431","date":"2024-11-22T11:11:25+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-20T11:04:46+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-18T15:41:09+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-14T06:55:13+00:00","index":"","fulltext":""},{"type":"submitted","content":"Translational Stroke Research","date":"2024-11-12T13:11:54+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"translational-stroke-research","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"trsr","sideBox":"Learn more about [Translational Stroke Research](http://jcmr-online.biomedcentral.com)","snPcode":"12975","submissionUrl":"https://submission.nature.com/new-submission/12975/3","title":"Translational Stroke Research","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"325cb25f-0170-40f4-9009-9017eefc8235","owner":[],"postedDate":"December 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-02-17T16:02:34+00:00","versionOfRecord":{"articleIdentity":"rs-5439901","link":"https://doi.org/10.1007/s12975-025-01332-6","journal":{"identity":"translational-stroke-research","isVorOnly":false,"title":"Translational Stroke Research"},"publishedOn":"2025-02-11 15:57:40","publishedOnDateReadable":"February 11th, 2025"},"versionCreatedAt":"2024-12-18 06:42:46","video":"","vorDoi":"10.1007/s12975-025-01332-6","vorDoiUrl":"https://doi.org/10.1007/s12975-025-01332-6","workflowStages":[]},"version":"v1","identity":"rs-5439901","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5439901","identity":"rs-5439901","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}