Extracellular Vesicles Play a Central Role in Cerebral Venous Disease‐Associated Brain Atrophy

Abstract Cerebral venous abnormalities, distinct from traditional arterial diseases, have been linked to brain atrophy in a previous community‐based cohort study, specifically in relation to the reduction of deep medullary veins (r‐DMVs). To better understand the properties and biological functions of serum extracellular vesicles (EVs) in cerebral venous disease‐associated brain atrophy, EVs are extracted from the serum of both participants with r‐DMV and normal controls and analyzed their proteomic profiles using Tandem Mass Tag label quantitation analysis. Phenotypic experiments showed that EVs from individuals with r‐DMVs are able to disrupt the normal functions of neurons, endothelial cells, and smooth muscle cells, and induce A1 reactive astrocytes. Additionally, this study provided a comprehensive characterization of the proteomic profile of DMV EVs and found that the collagen hydroxyproline is upregulated, while complement C3 is downregulated in the r‐DMV group, suggesting that r‐DMV may not be a simple pathological phenomenon and highlighting the potential involvement of EVs in the progression of brain atrophy in r‐DMVs which has implications for the development of future therapeutic strategies.


Introduction
Venous collagenosis (VC), caused by abnormal collagenous thickening of the venous walls, was first described in 1995 in a pathological study of postmortem brains and found to be most prominent in regions adjacent to the lateral ventricles. [1]VC is classified as an etiopathogenic type of cerebral small vessel disease (CSVD).Recently, venules have attracted considerable research interest because of their involvement in perivenous lymphatic drainage in neurodegenerative diseases. [2]VC has been found in some neurodegenerative diseases, such as Alzheimer's disease, [3] and has been described as the pathogenesis of white matter hyperintensity seen in Magnetic resonance imaging (MRI).
VC may lead to venous stenosis and obstruction, accounting for the decreased number of venules visualized in susceptibility-weighted imaging (SWI) in MRI.Consequently, an increasing number of in vivo studies have investigated cerebral venous diseases by counting visible venules, especially deep medullary veins (DMVs).DMVs are the most feasible to be counted because of their regular arrangement perpendicular to the ventricles.Previous studies have revealed that a decreased number of DMVs is independently related to white matter hyperintensity as well as brain atrophy. [4]However, common vascular risk factors are mainly derived from arterial studies, leaving risk factors relative to VC largely unknown.We previously investigated DMVs using brain MRI in a general population and found a stable correlation between reduced DMV numbers and increased age, but no significant association between DMV numbers and any other common vascular risk factors. [5,6]These results highlight the importance of cerebral venous abnormalities and their differences from traditional arterial diseases and warrant further investigation of associated venous risk factors and biomarkers.
Extracellular vesicles (EVs) are small cargo-bearing vesicles that are released by cells into the extracellular space.Intercellular communication through EVs appears to be involved in the pathogenesis of various disorders, including neurodegeneration and inflammatory diseases. [7]EVs can serve as potential biomarkers because their content is a molecular signature of the cell of origin, and exosome-based diagnostic tests are being used for the early detection of some diseases. [8]Investigating serum EVs from patients with DMV-reduction and brain atrophy may provide clues for potential risk factors related to venous pathology.
To explore this issue, we first conducted a proteomic analysis of serum EVs to identify potential biomarkers in a populationbased cohort of 1586 participants comprising participants with reduced DMV and an age-and sex-matched control group.We then studied the biological properties of EVs and their physiological functions in DMV reduction and brain atrophy.

Cerebral Deep Medullary Vein Reduction (r-DMV) is Associated with Brain Atrophy
To understand the relationship between DMV number and brain atrophy, we selected the 25% of the Shunyi cohort with the largest number of veins (DMV ≥20) as the control DMV group (c-DMV), and the 25% population with the fewest veins (DMV ≤17.5) as the r-DMV group.It was previously reported that the number of DMV decreases with age [6] ; therefore, we matched the sex and age of the two groups to control for confounding factors.Finally, 119 participants were included in the c-DMV group (37.0%males), and 118 subjects were included in the r-DMV group (36.4% males).Because of the effect of deoxygenated hemoglobin on DMV susceptibility, we measured hemoglobin (HGB) levels to exclude the potential influence of deoxygenated hemoglobin on DMV imaging.As shown in Figure 1A, there were no significant differences between the c-DMV and r-DMV groups after excluding the confounding factors.
We also identified indicators of traditional cerebrovascular disease risk, such as BMI, hypertension, diabetes mellitus, and hyperlipidemia.There were no significant differences in these indicators between the c-DMV and r-DMV groups, suggesting that DMV reduction and traditional cerebrovascular disease have no shared risk factors.We found that the BPF in the r-DMV group was significantly lower than that in the c-DMV group.In our previous cohort study, we detected a positive correlation between BPF and DMV numbers. [5]Therefore, we performed minimental state examination (MMSE) and montreal cognitive assessment (MOCA) cognitive analyses, which showed no significant differences in terms of cognitive function between the two groups (Figure 1B).We reanalyzed the included population by sex and found a greater decline in the Brain parenchymal fraction (BPF) in females (Figure 1C).The clinical characteristics of participants with c-DMV (n = 119) and r-DMV (n = 118) information are shown in Table S1, Supporting Information.

Extraction Efficiency and Verification of EVs
Proteins in EVs have been associated with brain atrophy in patients with clinically manifested vascular disease, [9] suggesting that EVs may be involved in the process of DMV reductionassociated brain atrophy.We first isolated the EVs from pooled serum samples of the participants using an IZON qEV original column.Ultrastructural analysis by transmission electron microscopy showed the serum-derived EVs were in the size range of 100-200 nm in diameter and displayed a lipid bilayer membrane with a cup-shaped morphology (Figure 2A).[12] Contrastingly, the serum contained a high level of albumin (ALB), which was hardly detectable in the EV fractions and therefore, served as a negative EV marker (Figure 2B).Coomassie Blue staining of separated proteins confirmed that equal amounts of proteins were loaded from each sample (Figure 3C).Nanoparticle tracking analysis (NTA) of the serum-derived EVs showed median sizes in the control DMV (c-DMV), reduced DMV (r-DMV), females from control DMV group (c-DMV-f), females from reduced DMV group (r-DMV-f), males from control DMV group (c-DMV-m), and males from reduced DMV group (r-DMV-m) of 141.23 nm, 139.91 nm, 128.01 nm, 157.71 nm, 149.94 nm, and 149.15 nm, respectively (Figure 2D).The particle diameter ranged from 100 to 200 nm, which was consistent with the size of EVs.Overall, these data revealed that the purity of our extracted EVs conformed with the Minimal Information for Studies of Extracellular Vesicles 2018 (MISEV2018). [13]We performed western blotting (WB) and flow cytometry assays of different cell-specific markers, endothelial cells (CD144) [14] , pericytes (PDGFR) [15,16] , and vascular smooth muscle cells (Myh11) [17] to investigate the changes of EV types between c-DMV and r-DMV group.The results revealed no significant differences between c-DMV and r-DMV groups (c-DMV vs. r-DMV): CD144 + (0.483% ± 0.018% vs. 0.417% ± 0.048%), PDGFR + (0.930% ± 0.017% vs. 0.907% ± 0.012%), and Myh11 + (0.533% ± 0.019% vs. 0.583% ± 0.007%) (Figure 2E-G).In order to investigate whether EVs from r-DMVs have an impact on brain cells, we first verified whether serum-derived EVs from c-DMVs and r-DMVs can penetrate the brain, PKH26-labelled EVs were intravenously injected into C57/BL6 mice, we found that serum-derived EVs not only penetrated the brain but also entered organs such as the liver, spleen, and lungs (Figure 2H).The schematic diagram of the experimental design was shown in Figure 2I.

r-DMV-EVs-Treated Neurons Lose Many Normal Functions
EVs absorbed by neurons have been reported to modulate neuronal functions. [18]Since brain atrophy is caused by the loss of neurons and connections between neurons, we suspected that EVs from the r-DMV population may affect the survival of neurons.To test this hypothesis, we isolated pooled serum-derived EVs from r-DMV participants and investigated their effects on neurons using c-DMV EVs as controls.Neuronal morphology plays a significant role in determining neuronal function and communication. [19]Thus, we first evaluated the morphological changes in rat cortical neurons treated with serum-derived EVs using the neuron-specific microtubule-associated protein 2 (MAP2), which is an abundant protein of cytoskeletal component, to image the neurons [20] (Figure 3A-a,c).Imaris software was used to visualize the architecture of the neuron network and neurite outgrowth (Figure 3A-b,d).We found that EVs of r-DMV impaired the dendritic arborization of the neurons, which was reflected in a shorter total dendrite length, fewer dendrite branches, smaller dendrite areas, and reduced intersection numbers in the Sholl analysis (Figure 3B).We also detected the effect of EVs on synapse and dendritic spine formation by treating primary cortical neurons for 24 h with 10 μg r-DMV or c-DMV EVs and examining the co-localization of presynaptic synapsin I and postsynaptic PSD 95.We observed a decrease in the density of colocalized synapsin-PSD 95 puncta in neurons treated with r-DMV EVs compared with those treated with the c-DMV EVs (Figure 3C,D).We then performed phalloidin staining to explore the function of r-DMV EVs in neuronal dendritic spines.Treatment of cortical neurons with r-DMV EVs induced a decrease in spines compared with the effects of the c-DMV EVs (Figure 3E,F).Deposits of fibrillar -amyloid peptide (A) and intracellular bundles of selfassembled hyperphosphorylated tau proteins (p-tau) are two hallmark lesions of neurodegenerative diseases, such as Alzheimer's disease. [21]We also investigated the association between r-DMV EVs and neurodegeneration by analyzing the expression of A42 and A40 in treated SY5Y cells.Treatment with r-DMV EVs significantly increased the expression level of A42 in SY5Y cells compared with the effects of c-DMV EVs treatment, while there was no significant difference in the expression level of A40 between the two groups (Figure 3G).Furthermore, there were no significant differences in the protein levels of total tau and p-tau (S396) between the two groups (Figure 3H,I).We further substantiate the impact of r-DMV EVs on neurons in vivo by intravenous injection of EVs into mice.Our findings revealed an augmented expression of cleaved caspase 3, suggesting that r-DMV EVs exhibited a greater propensity to provoke neuronal apoptosis in comparison to c-DMV EVs (Figure 3J,K).Taken together, these   a,c).Imaris-rendered neuron morphology is immunostained with MAP2 (yellow) in primary culture neurons (b,d).B) Dendrite length, branch level, and area, and Sholl analysis quantification in neurons; n = 14 neurons per group.C) Representative images of rat primary neurons immunostained with the pre-and post-synaptic markers synapsin I (red) and PSD95 (green), respectively.Co-localization of staining (yellow puncta) was counted as a structural synapse; example synapses are indicated with arrows.D) Quantification of synapse density from images; n = 9-10 dendrites.E) Representative confocal images of dendritic spines of rat primary neurons immunostained with phalloidin (green) (a,d).Images of the selected region of interest (ROI) were reconstructed using Imaris (b,e).Skeleton images of the selected ROI using Image J software (c,f).F) Quantification findings suggest that r-DMV EVs play a critical role in brain atrophy by disrupting normal neuronal function.

r-DMV EVs Induce A1 Reactive Astrocytes
Astrocytes play supportive roles in central nervous system (CNS) functions, such as neurotransmitter homeostasis and synapse development. [22]Some CNS diseases can induce two different reactive astrocyte states, A1 and A2. [23]A1 reactive astrocytes amplify inflammatory responses and produce neurotoxicity effects, which is involved in brain atrophy. [24]The purity of the primary cultured astrocytes isolated from the neonatal rat cortex was confirmed to be satisfactory for use in our study using GFAP staining (Figure S1A, Supporting Information).Next, we determined the effects of r-DMV EVs on the reactive state of astrocytes by assessing mRNA levels of pan-reactive, A1-specific, and A2-specific markers.The transcript level of pan-reactive (Gfap, Lcn2, Timp1, S1pr3, and Steap4) and A1-reactive (C3, Fbln5, H2-T23, Amigo2, Ggta1, Serping, and Psmb8) were significantly increased, whereas those of A2-reactive markers (Tm4sf1, Tgm1, S100a10, Clcf1, and Ptx3) were significantly decreased in r-DMV EVs treated astrocytes (Figure 4A and S1B, Supporting Information).Furthermore, both r-DMV-f and r-DMV-m EVs significantly increased the mRNA expression levels of pan-reactive and A1-specific markers (Figure 4B and S1C, Supporting Information).Astrocytes are not only important for the formation of synapses, but are also essential for the phagocytic elimination of synapses. [25]It has been previously reported that A1 reactive astrocytes have decreased phagocytic capacity. [23]Therefore, to compare the phagocytic abilities of astrocytes treated with r-DMV-EVs and c-DMV-EVs, we measured synaptosome engulfment.The r-DMV-EV-treated astrocytes engulfed fewer synaptosomes than the c-DMV-EV-treated astrocytes.Both r-DMV-f and r-DMV-m EVs treatments also significantly decreased the phagocytic ability of astrocytes compared with the effects of c-DMVf and c-DMV-m EVs (Figure 4C,D).This phagocytic deficit corresponded to decreased mRNA expression level of phagocytic receptors including Mertk, Megf10, Gas6, and Axl (Figure 4E).These data reveal that r-DMV EVs are capable of inducing A1 reactive astrocytes.Taken together, these data indicate that r-DMV EVs may be involved in brain atrophy by inducing A1 reactive astrocytes.

r-DMV-derived EVs Induce Endothelial Cell and Smooth Muscle Cell Dysfunction
Studies have shown that the blood-brain barrier (BBB) leakage increases brain atrophy. [26]The central elements of the BBB struc-ture are tight junctions of endothelial cells.Therefore, we speculated that r-DMV EVs may be involved in brain atrophy by affecting endothelial cell function.With tight junctions (TJ) playing a central role in BBB integrity, we first compared the levels of mRNA encoding BBB-specific TJ proteins in bEnd.3 cells were treated with 100 μg serum-derived EVs from the r-DMV or c-DMV groups.Compared with cells treated with c-DMV EVs group, ZO-1, Ocln, and Cldn5 were downregulated in the cells treated with r-DMV EVs (Figure 5A).Furthermore, CCK-8 assays revealed prominent suppression of bEnd.3 cell growth following treatment with r-DMV EVs at 100 and 250 μg mL −1 , whereas there were no significant differences in the growth patterns of cells treated with r-DMV EVs at 25 and 50 μg mL −1 (Figure 5B).The in vitro tube formation assay showed the ability of endothelial cells to form tube-like structures reflecting the potential for cell migration and the establishment of cell-cell and cell-matrix contractions.To assess the effect of r-DMV EVs on endothelial tube formation, bEnd.3 cells were treated with 100 μg mL −1 serumderived EVs from the r-DMV or c-DMV groups.bEnd.3 cells cultured for 6 h on gelled Matrigel containing s-DMV EVs formed more tubule-like networks than cells cultured in the presence of r-DMV EVs (Figure 5C).These data showed that the tube number and branching of endothelial cells were significantly lower following treatment with r-DMV EVs than after treatment with c-DMV EVs, although there was no significant difference in tube length between the two groups (Figure 5D).Briefly, our data show that the r-DMV EVs inhibited endothelial cell migrating capabilities.
VC is a pathological phenomenon caused by abnormal deposition of collagen in the vessel wall.Since collagen is known to increase in synthetic smooth muscle cells, [27] we investigated whether the reduction in DMVs was due to increased collagen secretion by smooth muscle cells, resulting in abnormal deposition on the vessel wall.After treating rat thoracic aortic smooth muscle cells (A-10) with different concentrations of EVs (50 μg mL −1 , 100 μg mL −1 , and 200 μg mL −1 ) from the c-DMV and r-DMV groups for 48 h, we observed lower levels of contractile markers (Acta-2, Myh11, Tag1n, and Cnn1) and higher levels of synthetic markers (Spp1 and Ereg) in the r-DMV group than those in the c-DMV group (Figure 5E and S2, Supporting Information).
Hydroxyproline, which is required for proline hydroxylation as a post-translational modification, [28] is found in few proteins other than collagen and provides stability to the triple-helical structure by forming hydrogen bonds.Analysis of the levels of hydroxyproline in the pooled serum-derived EVs from the c-DMV (n = 119) and r-DMV (n = 118) groups revealed that the levels of hydroxyproline in the r-DMV EVs were higher than those in the c-DMV EVs (Figure 5F).According to our proteomic profiles, there was no difference in the collagen protein abundance between the c-DMV EVs and r-DMV EVs (Figure 5G) whereas we of spine density from images; n = 9-11 dendrites.Images in (A, C, and E) were adjusted for brightness and contrast.A-F) Rat primary neurons treated with serum extracellular vesicles derived from c-DMV (pooled samples of 119 participants) or r-DMV (pooled samples of 118 participants).G) SH-SY5Y treated with EVs analyzed by ELISA using antibodies for the specific detection of A40 and A42.H) Western blot analysis confirmed no significant difference in p-tau (S396) between serum EVs from c-DMV and r-DMV, with total tau as the internal control of protein expression.I) Densitometric analyses of the immunoreactivities to the antibodies shown in (H).G-I) SH-SY5Y treated with serum extracellular vesicles derived from c-DMV (pooled samples of 119 participants) or r-DMV (pooled samples of 118 participants).J) Representative confocal images of mouse neurons immunostained with neuron marker NeuN (red), cleaved caspase-3 (green), and DAPI (blue).Co-localization of staining (yellow) was counted as the apoptotic neurons; for example, apoptotic neurons are indicated with arrows.K) Quantification of apoptotic neurons from images; n = 5 per group.c-DMV, control DMV; r-DMV, reduced DMV.**P < 0.01, ***P < 0.001; ns, not significant.Student's t-test was used as appropriate.Data represent the mean ± SEM. detected more collagen hydroxyproline in r-DMV EVs than in c-DMV EVs (Figure 5H).Thus, our results indicate that r-DMV EVs switch smooth muscle cells to synthetic phenotypes and yield more collagen than c-DMVs.Moreover, serum EVs are derived from r-DMVs, the collagen of which may be more stable.

Quantitative Characteristics of Protein Profiles
To understand the functional role of EVs in cerebral venous disease-associated brain atrophy, we investigated differences in the composition of serum-derived EV proteins between the  S2, Supporting Information).Among the identified EV proteins, 519 proteins (77.6%) overlapped with ExoCarta protein list (http://www.exocarta.org,release date: July 29, 2015) (Figure 6A).Gene Ontology (GO) classification indicated that most of the identified EV proteins were derived from exosomes, with molecular functions in complement activity as well as the biological processes of immune response, protein metabolism, and cell growth and/or maintenance (Figure 6B).These categories were consistent with the reported functions of EVs (a complete list of all enriched GO terms is shown in Table S3, Supporting Information).Principal component analysis (PCA) of the proteomic profiles showed a clear separation between the r-DMV and c-DMV groups, confirming the distinct differences between the two types of serum-derived EVs (Figure 6C).With a 1.2-fold change cut-off, 78 (25 upregulated and 53 downregulated) differentially expressed proteins (DEPs) were identified in the r-DMV group compared with the c-DMV group (Figure 6D), 221 DEPs (104 upregulated and 117 downregulated) were identified in the r-DMV-f group compared with the c-DMV-f group (Figure 6E), and 211 DEPs (71 upregulated and 140 downregulated) were identified in the r-DMV-m group compared with the c-DMV-m group (Figure 6F).The details of the DEPs are listed in Table S4, Supporting Information.

C3 is Related to the Reduction of DMV, Especially in Females
To clarify the mechanism by which r-DMV EVs induce A1 reactive astrocytes and disrupted neuronal and endothelial cell function, we performed an in-depth bioinformatics analysis of the proteomic profiles of serum-derived EVs from the r-DMV and c-DMV groups, as well as their sex-grouped EVs.When r-DMV EVs were compared with c-DMV EVs, the 78 DEPs between the r-DMV EVs and c-DMV EVs were significantly enriched in synapse organization, complement, and immune response (Figure 7A).
To gain insights into the potential functions of these DEPs, we employed the WEB-based GEneSeTAnaLysis Toolkit (http: //www.webgestalt.org/)to map the gene symbols of the DEPs to the Reactome, KEGG, and Wiki pathway databases.The 78 DEPs were predominantly enriched in the regulation of the complement cascade, complement and coagulation cascades, and human complement activation pathways (Figure 7B).To further place these DEPs in the context of known protein-protein interactions and gain insights into the coordinated roles of these proteins, we investigated their potential interactions using the online resource STRING (https://string-db.org/).Protein-protein interaction networks were visualized using Cytoscape according to their STRING scores.Among these, complement C3 (C3) was a hub protein, and complement factor H (CFH), properdin (CFP), and complement factor B (CFB) were complement-associated proteins (Figure 7C).In our previous analysis, we found that DMV-reduction in females was more strongly associated with brain atrophy than in males (Figure 1C).We identified a total of 328 DEPs between DMV-f (226 DEPs) and DMV-m (195 DEPs), of which 93 were shared (Figure 7D), suggesting the existence of sex-related differences between the serum-derived EVs in males and females with DMV-reduction.To further investigate the sexrelated differences among the EV proteins, we performed a cluster analysis (Figure 7E).Eight distinct protein clusters were identified based on fold changes in protein expression.In groups 1 and 8 (G1 and G8), these proteins showed similar expression trends in both female and male EVs groups.Groups 2 and 3 (G2 and G3) contained proteins that were DEPs in r-DMV-f compared with c-DMV-f, but not in r-DMV-m compared with c-DMV-m.In group 4 (G4), the proteins were upregulated in r-DMV-f and downregulated in r-DMV-m, whereas, in group 5 (G5), proteins were downregulated in r-DMV-f and upregulated in r-DMV-m.The proteins in groups 6 and 7 (G6 and G7) only showed changes in expression in the DMV-f group, but not in the DMV-m group.The DEPs in G1 and G8 were considered sex-consistent, whereas those in G2-G7 were considered sex-specific.Expression levels of sex-specific DEPs in the female and male groups of r-DMV and c-DMV EVs are shown in Figure 7F.Among these, G2 and G3 were female-specific DEPs.We further explored the potential function of female-specific DEPs in DMV-reduction.WikiPathway enrichment analysis indicated that in addition to the complementassociated pathway, cells, and molecules involved in local acute inflammatory response also played a pivotal role (Figure 7G).Next, we clustered all the G2 and G3 proteins into GO categories using bioinformatics tools.In terms of biological process, the female-specific DEPs were also enriched in inflammatory response (Figure 7H).Among the affected proteins involved in the inflammatory response term, C3 was downregulated and SER-PINA3 was upregulated as hub proteins in the serum-derived EVs of r-DMV-f compared with c-DMV-f (Figure 7I).We selected C3 and SERPINA3 for representative MS/MS identification, which yielded results that were consistent with the proteomic data (Figure S6, Supporting Information).C3 protein is the central component of the complement system, making it one of the most abundant proteins in circulation. [29]To investigate the contribution of C3 to the clinical characteristics or progression of DMV-reduction, we determined its levels in individual serum and EVs samples by ELISA.No significant differences were observed in serum C3 levels between the r-DMV and c-DMV groups.Additionally, there were no significant differences in the serum C3 levels of females and males of both the r-DMV and c-DMV groups (Figure 7J).We further detected the expression level of C3 in pooled serum EV samples from 119 c-DMV (75 females by bEnd.3 compared with the c-DMV group.D) Quantification of tube number, branching point, and tube length to assess the tube formation ability.n = 10 per group.E) Quantitative PCR analysis of contractile and synthetic phenotype markers of smooth muscle cells treated with 50 μg mL −1 EVs for 48 h.n = 3 per group.F) Proline hydroxylation assay of serum EVs derived from c-DMV (pooled samples of 119 participants) and r-DMV (pooled samples of 118 participants).n = 3 per group.G) The abundance of collagen expressed in c-DMV and r-DMV derived EVs.H) Representative MS/MS spectral data of collagen proline hydroxylation peptide fragments.c-DMV, control DMV; r-DMV, reduced DMV.*P < 0.05, **P < 0.01, ***P < 0.001; ns, not significant.Student's t-test was used as appropriate.Data represent the mean ± SEM. and 44 males) and 118 r-DMV (75 females and 43 males) participants.C3 levels were decreased in r-DMV EVs and r-DMV-f EVs, whereas there was no significant difference between c-DMV EVs and r-DMV-m EVs (Figure 7K).Further, ROC curve analysis of C3 expression levels in EVs from 119 c-DMV samples and 118 r-DMV samples showed that the performance of EV C3 expression levels for the diagnosis of DMV reduction was good, with AUC values of 0.78 (total), 0.81 (female), and 0.72 (male) (Figure 7L).These results suggested that C3 in EVs may serve as a biomarker for diagnosing DMV-reduction and assessing brain atrophy.

Discussion
Although an association between DMV reduction and brain atrophy was described in our previous cohort study, [5] the underlying mechanism is unclear.EVs cargo proteins have been identified as risk markers for the progression of vascular diseases and brain atrophy. [9]In this study, we discovered that serumderived EVs from participants with DMV reduction promoted the loss of many normal neuronal functions, induced A1 reactive astrocytes, and caused endothelial and smooth muscle cell dysfunction more significantly than those derived from normal control participants.Proteomic profiles of these EVs were ob-tained by TMT-based MS to clarify the mechanism of these cellular phenotypic changes and to further elucidate the role of EVs cargo proteins in brain atrophy.The DEPs identified in serum EV from r-DMV participants were predicted to be involved in multiple processes and functions related to the complement system and inflammation, particularly in females.We hypothesized that serum-derived EVs from r-DMV participants promote brain atrophy by regulating inflammation-related proteins, particularly by downregulating C3 and upregulating SERPINA3.
Our results demonstrated that serum EVs from participants with a reduced number of DMVs significantly impaired dendrite morphology and reduced the synaptic number and dendritic spine formation.Neuronal death and axonal destruction can result in brain atrophy, [30] and we speculated that serum EVs may be involved in brain atrophy by affecting neuronal function.Thus, we used primary rat cortical neurons for in vitro functional studies and intravenous injection of EVs to mice for in vivo functional studies.DMV abnormalities have been reported in patients with mild cognitive impairment and early Alzheimer's disease. [31]In Alzheimer's disease, the accumulation of A peptides gradually leads to the hyperphosphorylation of tau proteins, which then form neurofibrillary tangles, resulting in brain atrophy.Our data showed that serum EVs significantly promoted the production of A42 in the SH-SY5Y cell line (Figure 3G) but had no obvious effect on the phosphorylation of Tau-S396 (Figure 3H).Further studies of the functional effects of serum EVs on human-derived neurons are required to validate our findings.
Astrocytes, which are abundant in the CNS, [32] shape the microarchitecture of the brain, form neuronal-glial-vascular units, and regulate the BBB. [33]Given the vital roles of astrocytes, it is not surprising that impairment of these functions has been implicated in the pathogenesis of many diseases.Particularly, A1 astrocytes, which are highly neurotoxic, have been linked to neurodegenerative diseases, including Alzheimer's disease. [34]herefore, we investigated the ability of serum EVs from participants with reduced DMV levels to change the reactive state of astrocytes.Our data revealed that the serum EVs from DMVreduction participants induced the upregulation of A1 reactive astrocytes.Based on these findings, we speculate that serum EVs from participants with DMV-reduction participate in brain atrophy by inducing A1 reactive astrocyte-mediated neurotoxicity.
BBB permeability, which is associated with brain atrophy, is regulated by tight junction proteins in brain endothelial cells. [26,35]Therefore, the expression and function of brain endothelial tight junction proteins, including occludin, claudins, and members of the membrane-associated guanylate kinase protein family such as ZO-1, are often used as metrics to assess BBB integrity.Claudin-5 and occludin are transmembrane proteins, while ZO-1 is a peripheral membrane protein.The loss of claudins and ZO-1 is associated with BBB breakdown in human neurodegenerative disorders. [36,37]We speculated that EVs may be involved in the brain atrophy of participants with reduced DMV by affecting the function of endothelial cells.In accordance with this hypothesis, we found that the mRNA expression levels of claudin-5, occludin, and ZO-1 were decreased in the r-DMV EV-treated endothelial cells, suggesting that r-DMV derived EVs perturb the tight junctions of endothelial cells.Vascular defects are associated with decreased endothelial cell proliferation and tube formation ability. [38]Based on our findings, it is conceivable that r-DMV serum-derived EVs participate in the process of brain atrophy by inhibiting the proliferation, tube formation, and tight junction function of endothelial cells.
Under physiological conditions, vascular smooth muscle cells are contractile and maintain the elasticity of blood vessel walls and regular blood flow.However, under pathological conditions, these cells can dedifferentiate into a synthetic phenotype associated with the secretion of EV, proliferation, and migration to repair injury. [39]This process, known as phenotypic switching, is often the first step in vascular pathology. [40]In this study, we demonstrated that the mRNA levels of contractile markers were downregulated in smooth muscle cells after treatment with r-DMV EVs, while the expression of synthetic markers was upregulated, indicating a switch to the synthetic phenotype.Therefore, we speculate the decreased number of DMVs in r-DMV participants might be caused by increased synthesis of collagen, causing venous stenosis and obstruction.Regulation of HYP sites within the collagen structure alters fiber conformation and integrity and protects the collagen structure from degeneration. [28]According to our proteomics profiles, proline hydroxylation of collagen was increased in r-DMV EVs, indicating that the collagen in the r-DMV group was more likely to be deposited in the venous walls and appeared as a reduction of DMVs in SWI.
To explore the mechanism underlying the function of r-DMV EVs, we compared the proteomic profiles of serum-derived EVs from participants in the r-DMV and c-DMV groups, as well as the corresponding groups divided according to sex.We found that multiple complement-associated proteins were significantly enriched in the EVs of the r-DMV group.Activation of the complement pathway leads to neuronal damage in various neurological diseases [41] and can contribute to neurodegenerative phenotypes and brain atrophy. [42]Among these complement-associated proteins, we focused on C3 and SERPINA3 protein as the two proteins that were identified as female-specific DEPs.As a central component of the complement system, C3 protein has been identified as a plasma biomarker of brain atrophy in Alzheimer's disease. [43]In this study, we found that C3 protein was downregulated in EVs from r-DMV participants compared with the c-DMV group.It must be noted that very few studies have directly addressed the associations between serum-derived EVs of complement-associated proteins and their levels in the CNS.Interestingly, EVs from the r-DMV group promoted C3 mRNA expression in astrocytes to a greater extent than EVs from the c-DMV group.We speculate that there may be proteins other than C3 in the EVs from r-DMV group that induce A1 reactive astrocytes.SERPINA3 is a glycoprotein belonging to the serine protease inhibitor family of acute phase proteins [44] and was found to be markedly upregulated in the serum-derived EVs from r-DMV participants.SERPINA3 has been reported to be involved in the complement cascade and inflammation. [44]Although the basal level of SERPINA3 expression in the brain is low, it is upregulated in activated astrocytes.Several studies have shown a correlation between increased level of circulating SERPINA3 and brain microbleeds in elderly patients. [45]These data indicate that peripherally derived SERPINA3 directly influences the brain.Notably, the EV C3 levels were lower in the r-DMV group than in the c-DMV group, particularly in females.AUC values of 0.78 in r-DMV and 0.81 in r-DMV-f, indicate that serum EV C3 expression levels can be used to discriminate individuals with DMV-reduction from those with normal numbers of DMVs without the need for imaging to reduce the cost and not limited by instruments.
Changes in EVs may be a common upstream mechanism leading to both reduced deep medullary veins and brain atrophy.The findings of this study revealed that EVs from individuals with r-DMVs play a role in the impairment of brain vascular cells and neurons.Additionally, in vivo experiments have demonstrated that r-DMV EV induces abnormal neuronal activity and apoptosis and activation of astrocytes in mice.This phenomenon can be attributed to the cargo carried by EVs, which can transport diverse bioactive molecules, thereby influencing the functionality of the target cells. [7]Additionally, we observed significant differences in the protein composition and EV content between individuals with r-DMVs and those with c-DMVs.These findings suggest that r-DMV EVs may provoke pathological alterations by inflicting damage on brain vascular cells, releasing EVs that can enter the local brain tissue or bloodstream, and perpetuating a detrimental cycle that exacerbates brain injury and contributes to the onset and progression of brain atrophy.Thus, the relationship between r-DMV and atrophy is complex and the EV may play a role in this network and whether it could be developed as a therapeutical target or bio-marker awaits further investigation.
The limitations of our study should also be considered.This study was based on an ongoing prospective community-based cohort; therefore, our findings must be reproduced using other samples.Further, the absence of a suitable model for brain atrophy hinders further in-depth research into this condition.While animal models can provide some understanding regarding brain atrophy to some extent, they are unable to fully replicate its complexity.Nevertheless, we elucidated the proteomic characteristics of serum-derived EVs in c-DMV and r-DMV participants, and our findings expand the current knowledge of the roles of EVs in the cerebral DMV reduction-associated brain atrophy.

Conclusion
In summary (Figure 8), serum-derived r-DMV EVs promote the dysfunction of neurons, endothelial cells, and smooth muscle cells and induce A1 reactive astrocytes.These phenotypes are associated with brain atrophy.The expression profile of the serum EV proteome in r-DMVs differed from that of c-DMVs.This difference may be attributed to changes in complement-related proteins, particularly C3.Furthermore, we identified C3 as a potential biomarker of DMV reduction.

Experimental Section
Reagents: All antibodies, commercial assays, and media in this study are listed in Table S5, Supporting Information.L-glutamine (SH15140-122, Gibco), and 0.5% penicillin-streptomycin (15140-122, Gibco).Rat primary astrocytes were cultured in DMEM supplemented with 20% FBS and 1% penicillin-streptomycin.All cells were cultured at 37 °C in a humidified atmosphere containing 5% CO 2 .At 80% confluence, the cells were trypsinized and suspended in a fresh medium.
EV Staining with PKH26 Dye: Extracellular vesicles were derived from 1 mL serum per experimental group using size exclusion chromatography (qEV original/35 nm, ICO-35, IZON Science) as previously described.The resulting vesicles were concentrated using an Amicon Ultra-4 30K centrifugal filter unit (UFC8030, Merck).A total of 1 mL of concentrated EVs, containing ≈800 μg of protein (determined by BCA test), were subjected to staining with PKH26 using PKH26 Red Fluorescent Cell Linker Kits for General Cell Membrane Labelling (MINI26, Sigma-Aldrich).The manufacturer's instructions were followed, whereby EVs were incubated with PKH26 at 37 °C for 10 min.The reaction was then terminated using 1% BSA.Subsequently, the labeled EVs were centrifuged at 200 000 g for 90 min at 4 °C to remove excess dye and resuspended in PBS buffer for in vivo imaging.
Biodistribution of EVs in Mice: EVs labeled with PKH26 dye were injected intravenously into 4-week-old female C57/BL6 mice at a dose equivalent to 23 mg kg −1 body weight.After a period of 72 h postinjection, the mice were euthanized.The organs of interest, including the brain, heart, liver, lung, kidney, and spleen were fixed in 4% paraformaldehyde, followed by gradient dehydration in 20%−30% sucrose solution, embedding in O.C.T compound and stored at −80 °C for subsequent immunofluorescence staining.
Immunofluorescence Staining: For in vitro assays, rat primary neurons and astrocytes were cultured on coverslip pre-coated with poly-lysine.For in vivo assays, the frozen organs were cut at a thickness of 12 μm sections.The cells on a coverslip and the tissue sections were fixed in 4% PFA for 30 min, and washed three times with PBS.After permeabilizing with 3% Triton X-100 for 10 min, the cells were blocked with 5% BSA for 30 min and then incubated overnight at 4 °C with primary antibodies against MAP2 were added for 1 h at room temperature.Cells were stained with DAPI to visualize the nuclei.The total dendrite length was measured for primary dendrites and all dendritic branches.For neuronal morphology, the number of intersections was assessed by Sholl analysis. [48]All measurements and 3D reconstruction of neurons were performed using the Imaris software.
Quantitative Real-Time PCR (RT-qPCR): Total RNA was extracted from cultured cells using TRIzol reagent (15 596 026, Thermo Fisher Scientific) according to the manufacturer's instructions.Extracted RNA (500 ng) was then used as a template for cDNA synthesis in 10 μL volume with the PrimeScript RT reagent kit (RR036A, TaKaRa).Quantitative real-time PCR was performed using cDNA in 20 μL reaction with 10 μL TB Green Premix Ex Taq II (Tli RNase H Plus). Thermocycling conditions were 95 ο C for 30 s, followed by 40 cycles at 95 ο C for 5 s, 60 ο C for 30 s, and 95 ο C for 10 s. -actin and GAPDH served as the reference gene and relative gene expression was calculated with the 2 −ΔΔCT method. [49]The primer sequences are shown in Table S6, Supporting Information.
Synaptosome Engulfment Assay: Synaptosomes were isolated from rat brain cortex using the Synaptic Extraction Reagent (Syn-PER Synaptic Protein Extraction Reagent, 87 793, Thermo Fisher Scientific) according to the manufacturer's instructions and conjugated with pHrodo Red, succinimidyl ester (P36600, Thermo Fisher Scientific) at room temperature with gentle agitation.After incubation for 2 h, unbound pHrodo was removed by centrifugation.Astrocytes were then incubated with pHrodoconjugated synaptosomes in glial medium with c-DMV or r-DMV EVs for 48 h and imaged using a Leica time-lapse microscope.For image analysis, we used a 20× objective lens to select 10 images per group from random areas of a 24-well plate and plotted the total sum of the fluorescence intensity of the objects in the image.Data were normalized to the degree of engulfment per image.
ELISA: Cell pellets and serum-derived EVs were lysed in RIPA buffer and total proteins were extracted by sonication.C3 levels were detected using an ELISA kit (CSB-E08665h, Cusabio).After treatment with EVs from r-DMV and c-DMV, the level of soluble A1-40 and A1-42 in SH-SY5Y cell pellets were detected using A1-40 (CSB-E08299 h, Cusabio) and A1-42 (CSB-E10684 h, Cusabio) ELISA kits.All kits were used according to the manufacturer's instructions.
Cell Proliferation Assay: bEnd.3 cells were seeded in 96-well plates (1800 cells per well) and incubated in serum-free medium for 2 h at 37 °C.EVs were then added to stimulate cells for 1 h before the medium was replaced with a medium containing 10% FBS.After incubation for 72 h, 10 μL of CCK8 (HY-K0301, MCE) was added into each well and the cells were incubated for an additional 2 h before cell growth was determined by measuring the absorbance at 450 nm using Multiskan FC Microplate Photometer (Thermo Fisher Scientific).
Tube Formation Assay: The formation of capillary-like structures was determined using the basement membrane extract (Corning Matrigel Growth Factor Reduced Basement Membrane Matrix, CLS354230).After thawing at 4 °C, the Matrigel was immediately layered into 96-well cell culture plates and incubated at 37 °C for 30 min to allow the basement membrane to gel.Subsequently, bEnd.3 (4.5 × 10 3 cells) containing 100 μg mL −1 serum-derived EVs from the c-DMV or r-DMV groups were seeded onto the Matrigel.After incubation at 37 °C for 6 h, images were captured using bright-field microscopy.The tube number, branching point, and tube length were quantified using the Wimasis Analysis Platform (https://mywim.wimasis.com/).
Hydroxyproline Assay: The hydroxyproline content of serum-derived EVs was assessed using a hydroxyproline assay kit (MAK008, Sigma-Aldrich) according to the manufacturer's protocol.Briefly, EVs isolated from 200 μL serum from the c-DMV (n = 119) and r-DMV (n = 118) groups were pooled and transferred to a pressure-tight polypropylene vial containing 100 μL of concentrated hydrochloric acid and the PTFE-lined cap was closed tightly.Aliquots (100 μL) of the solution were then transferred to a 96-well plate and placed at 60 °C to dry samples.The amount of hydroxyproline in the samples was detected by a chromogenic reaction using chloramine-T and 4-dimethylaminobenzaldehyde as the substrates, and the absorbance was measured at 570 nm.
Proteomic Analysis-Tandem Mass Tag (TMT) Labeling: For TMT labeling, the lysates of EVs from the six sample groups (c-DMV, control DMV; r-DMV, reduced DMV; c-DMV-f, females from control DMV group; r-DMVf, females from reduced DMV group; c-DMV-m, males from control DMV group; and r-DMV-m, males from reduced DMV group) were diluted to 1 mg mL −1 with 8 M urea.The TMT kit (90 068, Thermo Fisher Scientific) was used for labeling according to the manufacturer's protocol.For each group, an aliquot of protein (100 μg) was reduced by incubation with 10 mmol L −1 dithiothreitol at 37 °C for 30 min, and then alkylated with 25 mmol L −1 iodoacetamide at room temperature for 30 min in the dark.Protein digestion was completed by incubation with Trypsin/Lys-C (enzyme: protein mass ratio 1:25) for 14 h at 37 °C.After terminating the digestion by heating at 60 °C for 30 min, protein digests were desalted using reverse-phase column cheomatography (Oasis HLB, WAT09425, Waters), dried with a SpeedVac vacuum concentrator, and finally dissolved in 200 mM triethylammonium bicarbonate buffer (pH 8.5) for labeling with 0.8 mg TMT reagents at room temperature.The groups were labelled as follows: c-DMV, TMT-126; r-DMV, TMT-127; c-DMV-f, TMT-128; r-DMV-f, TMT-129; c-DMV-m, TMT-130; r-DMV-m, TMT-131.After labeling, samples were mixed, desalted, dried as described previously, and finally dissolved in 0.1% formic acid for high-performance liquid chromatography (HPLC) fractionation.
Proteomic Analysis-High-Performance Liquid Chromatography (HPLC) Fractionation: The TMT-labelled protein digests (100 μL in 0.1% TFA) were fractionated by HPLC (UltiMate 3000 UHPLC, Thermo Scientific) equipped with an Xbridge BEH300 C18 column (4.6 mm × 250 mm, 5 μm) maintained at 40 °C with a flow rate of 1.0 mL min −1 .Peptides were eluted using a gradient acetonitrile elution buffer consisting of H 2 O with ammonium hydroxide (mobile phase A, pH 10), and 98% acetonitrile with ammonium hydroxide (mobile phase B, pH 10).A total of 47 fractions were collected at 1.5 min intervals, dried in a vacuum concentrator, and combined into 12 samples according to the peptide abundance.The samples were dissolved in 20 μL of 0.1% TFA for subsequent liquid chromatography (LC)-mass spectrometry with tandem mass spectrometry (LC-MS/MS) analysis.
Proteomic Analysis-LC-MS/MS Analysis: The Ultiate U3000 system, directly connected to a Thermo Orbitrap Fusion Lumos mass spectrometer, was used for LC-MS/MS analysis.The analytical column was a homemade fused silica capillary column (75 μm ID, 150 mm length; Upchurch, Oak Harbor, WA, USA) filled with C-18 resin (300 Å, 2 μm; Varian, Lexington, MA, USA).The peptides were separated using a 135-minute gradient elution with mobile phase A consisting of H 2 O with ammonium hydroxide (pH 10), and mobile phase B consisting of 98% acetonitrile with ammonium hydroxide (pH 10) at a flow rate of 0.3 μL min −1 .Xcalibur 4.1 software was used to operate the Orbitrap Fusion mass spectrometer in the data-dependent acquisition mode with a single full-scan mass spectrum in the Orbitrap (350-1500 m/z, 120000 resolution) followed by 3-s datadependent MS/MS scans in an ion routing multipole at 35% normalized higher-energy collisional dissociation (HCD).
Proteomic Analysis-Data Processing: MS/MS spectra were analyzed using Proteome Discoverer 2.4 software (Thermo Scientific).The spectra were searched against the UniProt/SwissProt human proteome database (released on August 11, 2022).The following parameters were set as the search criteria: full tryptic specificity with a maximum of two missed cleavages were allowed; precursor mass tolerance was set at 20 ppm; and fragment mass tolerance was set at 0.02 Da.Static modifications include carbamidomethylation on C and TMT 6-plex on peptide N-terminus.Dynamic modifications include TMT 6-plex on K and oxidation on M. In the current study, identified proteins were defined as proteins with at least two unique peptides.The reported monoisotopic m/z was tuned according to the raw spectral data.The hydroxylated proline modification site localization probabilities were obtained by repeated searches of the MS/MS data using the UniProt/SwissProt human collagen database (released on August 23, 2022) and allowing for dynamic proline oxidation modification.
Proteomic Analysis-Bioinformatics Analysis: For proteomic analysis of human serum-derived EVs, the relative protein abundance was presented as the ratio of TMT-127/126 (r-DMV/c-DMV group), TMT-129/128 (r-DMV-f/c-DMV-f group), and TMT-131/130 (r-DMV-m/c-DMV-m group).The differential expression threshold was defined as a 1.2-fold change in expression.Details of the MS proteomic data were submitted to the ProteomeXchange Consortium with the identifier PXD 03 9873.GO functional enrichment analysis was conducted using the Database for Annotation, Visualization, and Integrated Discovery (DAVID) (https://david.ncifcrf.gov/)and FunRich software (3.1.3).GO analysis was performed using the Cytoscape plug-in ClueGO (version 3.9.1).Pathway mapping was performed using the WEB-based GEneSeT Toolkit (http://www.webgestalt.org/).Pathways with at least three target genes and a P-value of <0.05 were considered significant.The STRING (Search Tool for the Retrieval of Interacting Genes/Proteins) database (https://string-db.org/)was used to predict protein networks.Networks were downloaded and visualized using Cytoscape plug-in MCODE (version 3.9.1).
Statistical Analysis: For the proteome data, the proteins with missing values in all samples were removed.Stringent criteria were applied for high protein confidence (unique peptide ≥ 2) and contaminants from keratin proteins were eliminated.All the statistical analyses and data presentations were performed using R (version 4.1.0)and GraphPad Prism 8 software (GraphPad Software, Inc., La Jolla, CA, USA).All assays were repeated at least three times and the data were presented as mean ± standard error of mean (SEM).Unpaired two-tailed Student's t-test was used to compare the two experimental groups.P-value < 0.05 was set as the threshold for statistical significance.

Figure 1 .
Figure 1.The reduction of deep medullary vein (r-DMV) is associated with brain atrophy.A) Samples to exclude confounding factors.B) Characteristics of the 237 participants.C) Brain atrophy-related indicators for each sex.HGB, hemoglobin; BMI, body mass index; DMV, deep medullary vein; WMH, white matter hyperintensity; BPF, brain parenchyma fraction; GMF, gray matter fraction; WMF, white matter fraction; MMSE, Mini-Mental State Examination; MOCA, Montreal Cognitive Assessment.Binomial variables are expressed as n (%); continuous variables with normal distribution are expressed as mean ± standard deviation (SD); continuous variables without normal distribution are expressed as median (interquartile range, IQR); c-DMV, control DMV; r-DMV, reduced DMV.*P < 0.05, ***P < 0.001.Student's t-test was used as appropriate.

Figure 2 .
Figure 2. Extraction efficiency and verification of EVs.A) Electron microscope image showing the intact morphology of EVs.B) Western blot showing enrichment of the EV markers LAMP2, ALIX, CD63, and ALB (albumin).C) Coomassie Blue staining was used as a control to access standardized loading.D) Nanoparticle tracking analyses (NTA) of serum EVs.E) Western blot and F,G) flow cytometry analysis of endothelial-specific marker CD144, pericyte-specific marker PDGFR, and vascular smooth muscle cell marker Myh11 in serum EVs derived from c-DMV and r-DMV.F) Representative flow cytometry plots displaying CD144 + , PDGFR + , and Myh11 + EVs.G) Quantification of cell-specific EVs from images; n = 3 per group.H) EV-positive organs and EV-negative organs intravenous injection of fluorescence labeled EVs.Representative images of the brain, liver, spleen, lung, heart, and kidney after intravenous EV injection are shown.The nuclei of the cells are marked in blue (DAPI).EVs appear in red (PKH26).I) Schematic diagram of the experimental design.c-DMV, control DMV; r-DMV, reduced DMV; c-DMV-f, females from control DMV; r-DMV-f, females from reduced DMV group; c-DMV-m, males from control DMV group; and r-DMV-m, males from reduced DMV group.ns, not significant.Student's t-test was used as appropriate.Data represent the mean ± SEM.

Figure 3 .
Figure3.EVs of r-DMV-treated neurons lose many normal functions.A) Representative confocal images of rat primary neurons immunostained with neuron marker MAP2 (red) and DAPI (blue) (a,c).Imaris-rendered neuron morphology is immunostained with MAP2 (yellow) in primary culture neurons (b,d).B) Dendrite length, branch level, and area, and Sholl analysis quantification in neurons; n = 14 neurons per group.C) Representative images of rat primary neurons immunostained with the pre-and post-synaptic markers synapsin I (red) and PSD95 (green), respectively.Co-localization of staining (yellow puncta) was counted as a structural synapse; example synapses are indicated with arrows.D) Quantification of synapse density from images; n = 9-10 dendrites.E) Representative confocal images of dendritic spines of rat primary neurons immunostained with phalloidin (green) (a,d).Images of the selected region of interest (ROI) were reconstructed using Imaris (b,e).Skeleton images of the selected ROI using Image J software (c,f).F) Quantification

Figure 5 .
Figure 5.The r-DMV-derived EVs promote endothelial cell and smooth muscle cells dysfunction.A) Quantitative PCR analysis of tight junction proteins showing lower transcript levels in the r-DMV group than in the c-DMV group.n = 3 per group.B) EVs (100 μg mL −1 and 250 μg mL −1 ) from the r-DMV group inhibited endothelial cell proliferation compared with the c-DMV group.n = 4 per group.C) EVs from the r-DMV group suppressed tube formation

Figure 6 .
Figure 6.Quantitative characteristics of protein profiles.A) Venn diagram showing the overlap of identified proteins with ExoCarta proteins (519; Released date: 29 July 2015).B) The five most enriched categories and the enriched significance (-log (P-value), P < 0.05) of identified proteins in the cellular components, biological process, and molecular functions categories.C) Principal component analysis (PCA) showing differences in the proteome profiles between c-DMV, r-DMV, c-DMV-f, r-DMV-f, c-DMV-m, and r-DMV-m.D-F) Scatter plot showing the distribution of upregulated (red dots) and downregulated (blue dots) differentially expressed proteins (DEPs).DEPs in r-DMV and c-DMV (D); DEPs in r-DMV-f and c-DMV-f (E); DEPs in r-DMV-m and c-DMV-m (F).c-DMV, control DMV; r-DMV, reduced DMV; c-DMV-f, females from control DMV group; r-DMV-f, females from reduced DMV group; c-DMV-m, males from control DMV group; and r-DMV-m, males from reduced DMV group.