Plasma exosome proteomics reveals the pathogenesis mechanism of post-stroke cognitive impairment

Exploration and utilization of exosome biomarkers and their related functions provide the possibility for the diagnosis and treatment of post-stroke cognitive impairment (PSCI). To identify the new diagnostic and prognostic biomarkers of plasma exosome were uzed label-free quantitative proteomics and biological information analysis in PSCI patients. Behavioral assessments were performed, including the Mini-Mental Status Examination (MMSE), the Montreal Cognitive Assessment (MoCA), the Barthel index, the Morse Fall Seale (MFS) between control group (n = 10) and PSCI group (n = 10). The blood samples were collected to analyse the biomarker and differentially expressed proteins of plasma exosome using label-free quantitative proteomics and biological information. The exosomes marker proteins were determined by Western blot. The exosome morphology was observed by transmission electron microscopy. The scores of MMSE and MoCA were significantly decreased in the PSCI group. The PT% and high-density lipoprotein decreased and the INR ratio increased in PSCI group. The mean size of exosome was approximately 71.6 nm and the concentration was approximately 6.8E+7 particles/mL. Exosome proteomics identified 259 differentially expressed proteins. The mechanisms of cognitive impairment are related to regulate the degradation of ubiquitinated proteins, calcium dependent protein binding, cell adhesive protein binding, formation of fibrin clot, lipid metabolism and ATP-dependent degradation of ubiquitinated proteins in plasma exosome of PSCI patients. Plasma levels of YWHAZ and BAIAP2 were significantly increased while that of IGHD, ABCB6 and HSPD1 were significantly decreased in PSCI patients. These proteins might be target-related proteins and provide global insights into pathogenesis mechanisms of PSCI at plasma exosome proteins level.


INTRODUCTION
Ischemic stroke is induced by cerebral artery occlusion, which causes damage to endothelial cells, vascular smooth muscle, glial cells, neurons and related neuro-vascular units, and ultimately leads to brain tissue death and focal neurological damage [1,2]. Post-stroke cognitive impairment (PSCI) is a type of vascular cognitive impairment that manifests throughout 6 months following a stroke. Patients suffering from AGING stroke lesions taking place in various regions of the brain that are not traditionally cognition-included may also result in development of PSCI. The plasma exosome biomarkers of PSCI have been emphasized. Plasma exosome biomarkers exploration and utilization and their associated functions allowed PSCI diagnosis and treatment possible.
Exosomes are small membrane vesicles existed in extracellular fluid and contain important genetic materials such as DNA, RNA, protein, lipid and miRNA [3]. Exosomes exist in extracellular fluid and contain important genetic materials such as DNA, RNA, protein, lipid and miRNA [3]. Exosomes can directly or indirectly act on target cells through the release of membrane contents or signal molecules for intercellular information transmission. In different pathological stages of ischemic stroke, exosomes released by different types of nerve cells contain specific signal molecules. Exosome miRNA-122-5p and miR-300-3p were biomarkers for the diagnosis of hyperacute (less than 6 h), subacute (8-14 d) and convalescent (greater than 14 d) ischemic cerebral infarction [4]. Exosomes secreted by circulating EPCs can transfer their inclusion to recipient endothelial cells, which contain miRNA related to PI3K/Akt signaling pathway and miRNA related to angiogenesis, such as miR-126 and miR-296. In the brain, exosomes secreted by cultured glioma cells provide angiogenic proteins, mRNAs and miRNAs to cerebrovascular endothelial cells and induce angiogenesis [5]. Neurons and glial cells interact with exosomes released by them to transfer biomolecules, regulate axonal growth and myelin sheath formation, and participate in brain nerve remodeling. Exosomes derived from multipluripotent mesenchymal stromal cells can effectively promote vascular and nerve regeneration, reduce inflammatory response and improve traumatic brain injury [6,7]. Proteomics is an indispensable omics science to elucidate the proteome diversity. Faced with the limitations of diagnosis and treatment of PSCI, the task of finding effective methods to diagnose, predict and prevent PSCI has become more important.
In order to find a new prognostic and diagnostic plasma exosome biomarkers utilizing label-free quantitative proteomics and analysis of biological information in individuals with PSCI. In this study, a variety of psychological evaluations, which include the Mini-Mental Status Examination (MMSE), the Montreal Cognitive Assessment (MoCA), the Barthel index and blood biomarker detection were performed. The plasma exosome from control participants and PSCI patients were collected and analyzed by label-free quantitative proteomics. The differentially expressed proteins and its biological information analysis were conducted to establish global prospective on the PSCI pathogenesis mechanisms at proteins level.

Study design
The Chinese Clinical Trial Registry (ChiCTR) has registered and recorded this clinical trial (registration number: ChiCTR1900023739, registration date: June 10, 2019), and the research protocol was approved by the Ethics Committee of Dongzhimen Hospital, a department of Beijing University of Chinese Medicine (approval number: DZMEC-KY-2019-04). Patients suffering from acute ischemic stroke were registered from Dongzhimen Hospital (eastern area) associated with Beijing University of Chinese Medicine between June 9, 2019 to December, 2019. The present investigation was conducted based on the ethical principles outlined in the Helsinki Declaration of 1975 (and as modified in 2013). This study contains the control group and PSCI group.

Participants
This investigation included 20 participants who signed an informed consent form. There are 10 participants in the PSCI group and control group, respectively. PSCI patients' inclusion criteria involved the following: (1) Age ≥35 years and ≤70 years; (2) Every patient has either an MRI or CT scan to validate the acute ischemic stroke; (3) Cognitive evaluation was done by MMSE within the first 7 days after the development of acute ischemic stroke. Patients having MMSE score ≤26 were determined as the PSCI group. (4) Cognitive impairment occurred after stroke [8,9]. The exclusion criteria comprise the subsequent aspects : (1) Symptoms exhibited a range of complex to severe primary disorders of the heart, kidney, liver and hematopoietic system; (2) Consciousness acute disturbance; (3) Dementia and brain, different causes are because of mental or physiological diseases; (4) With severe vision, hearing or even speech impairment conflicts with rehabilitation; and (5) Furthermore the onset of cognitive impairment, there were no additional focal indications of cerebrovascular disease.

Behavioral assessment
The informed consent forms were signed by all participants and a variety of behavioral evaluations were conducted, such as the MMSE, the MoCA, the Barthel index, the Morse Fall Seale (MFS), and The Braden Scale for the prediction of pressure sore risk, as well as the Padua Prediction Score. The patients received AGING antiplatelet therapy (aspirin, 100 mg, QD) before assessments.

Blood biomarker detection
After a fasting period of 12-h, blood samples were obtained in the morning. Two milliliters of whole blood were drawn from peripheral vein of each participant and kept in a polypropylene tube including EDTA. The Dongzhimen Hospital affiliated with Beijing University of Chinese Medicine was requested to furnish the blood specimens. Following that, the fibrinogen, red blood cell specific volume (HCT), total cholesterol (CHO), triglyceride (TG), high-density lipoprotein (HDL), lowdensity lipoprotein (LDL), and other different markers were measured using whole blood samples.

The isolation and determination of plasma exosomes characteristics
The isolation procedure of exosomes The isolation of plasma exosomes using TiO2 with a slight modification [10]. Remove the plasma sample from storage and place it on ice. Centrifuge the plasma sample at 2000 × g at room temperature for 20 min to extract cells and debris. Transfer 100 μL plasma to a new tube and processed using 0.2 mm pore size syringe filters (PALL Life Sciences, USA) for extracting apoptotic bodies and the large microvesicles. Following that, 5 mg of TiO2 microspheres were combined with the plasma sample (GL Science Inc, Japan) and mixed the sample thoroughly by vortexing at 4°C for a period of 5 min. The mixture underwent centrifugation at a force of 20,000 × g for 3 min at a temperature of 4°C, subsequently, the supernatant was extracted. The exosomes on the TiO2 microspheres were hydrated with PBS three times to eliminate unspecific contaminants. After washing with PBS, exosomes were lysed and digested directly with trypsin (Promega, Madison, WI) from the surface of microspheres at 37°C in 50 mM ammonium bicarbonate for 16 h. A new tube was used to transfer the supernatant and the TiO2 microspheres were hydrated twice with 100 μL of 0.1% formic acid after centrifugation at 12,000 g at 25°C for 5 min. The washing fraction was extracted and pooled with the supernatant. NanoDrop (Thermo Fisher, USA) was utilized to measure the concentration of peptide at an absorbance of 280 nm.

Western blot analysis of exosomes marker proteins
Detect the quality of plasma protein in 0.5 mL fraction sample. The fraction of vesicles with the least plasma protein content was selected for subsequent analysis. The concentration of protein samples was measured utilizing the BCA method. 12% SDS-PAGE was used for separating 10 μg of the protein, then the protein transferred to a 0.45 μm PVDF membrane, and inhibited utilizing a blocking solution including 5% bovine serum albumin for a period of 1 h at room temperature. Exosomal marker protein antibodies were add and incubate at a temperature of 4°C overnight, containing anti-CD63 rabbit polyclonal antibody (1:600) (

Transmission electron microscopy (TEM)
The exosome morphology was observed utilizing TEM. Exosome sample drops have the ability to adsorb for 5 min on formvar-coated EM grids, and were stained negatively utilizing 2% (w/v) phosphotungstic acid for 1 min. At 80 kV of acceleration voltage, transmission electron microscope (H-7650; Hitachi, Ltd., Tokyo, Japan) was utilized to perform TEM analysis.

Nanoparticle tracking analyzer (NTA) and Dynamic light scattering (DLS) analysis
The particle size and concentration analysis of model exosomes were performed on ZetaView Nanoparticle Tracker (Particle Metrix, Meerbusch, Germany). Calibrate the instrument with polystyrene particles with approximately 100 nm of a particle size, Dilute the model exosomes to approximately 1 × 10 8 particles/mL, and put them into the instrument for analysis. Each group of samples is automatically scanned 11 times to remove abnormal data. The data was recorded and analyzed by ZetaView 8.03.04.01 software.

Label-free quantitative proteomics
MALDI-TOF-MS/MS and database searching were utilized for the identification of proteins. An online liquid chromatography-tandem mass spectrometry (LC-MS/MS) setup including an EasynanoLC system and a Q-Exactive mass spectrometer (Thermo Scientific, Germany) installed with a nanoelectrospray ion source was utilized for all LC-MS/MS investigations.
(1) Mobile phase A included 0.1% FA, 2% acetonitrile dissolved within water, and mobile phase B included 0.1% FA, 98% acetonitrile in water. The flow rate has been measured to be 300 nL/min.
(2) At 2 kV, the source was operated. In order to perform a full MS survey scan, AGC target was 3e6, scan range was from m/z 300 to 1400 and the result of resolution was 70,000. The 50 highest intense peaks with charge state 2 and above were chosen in order to sequence and fragmented in the ion trap by HCD with normalized collision energy of 27%. Exclude isotope item was enabled and dynamic exclusion time was adjusted as 18 s.
(3) MaxQuant software was utilized to search the Raw MS files against UniProt database. (Version 1.5.2.8). The fixed modification was C (carbamidomethyl) and the variable modification was M (oxidation) and protein N-term (acetyl). The tolerance for the first search peptide was 20 ppm and the tolerance for the main search peptide was 6 ppm. The MS/MS tolerance result was 0.02Da. The PSM and protein false discovery level result was 1%. The employed among the runs and minimum score required for modified peptides was 40.

Protein-protein interaction (PPI) networks analysis
To gain a deeper comprehension from the perspective of the biological context of differentially expressed proteins, the protein interaction analysis was conducted by utilizing the free web-based search tool STRING10.5. The STRING database is a fundamental data resource within the ELIXIR's core, containing both identified and anticipated protein interactions. These data will be collected and integrated by the STRING database, with consolidating identified and anticipated protein-protein correlation data for the organism's large number [11]. STRING was needed to incorporate additional predicted functional partners for the purpose of improving the PPI networks formation. The protein IDs list that was differentially expressed was entered into the STRING database (https://string-db.org) to determine identified and anticipated protein functional correlation networks.

GO analysis
For exhibiting the differentially expressed proteins presence, to categorize the proteins in accordance with their biological process, protein classification, cellular composition, and pathway, GO enrichment analysis was conducted. In order to find out how these experimentally discovered proteins are distributed, GO enrichment annotation tools were used to examine each protein with higher speed and to comprehend the relationship between protein and biological function in its entirety. GO function enrichment analysis of protein describes this distribution comparison with the overall protein distribution, confirming which biological processes or molecular functions were significantly enriched with experimentally identified proteins, many regulatory, metabolic, and signal transduction pathways are resistant within the organism, and these pathways frequently form various pathways. Pathway analysis permits the identification of the highest significant biochemical-metabolic and signal transduction pathways in which the proteins were contained.

Statistical analysis
Each value was reported as the mean ± standard deviation (SD). Statistical analysis was conducted by SPSS20.0. The Demographic data and MMSE, MoCA, blood marker were analyzed with independent-samples t-test between control group and PSCI group. The Chisquare statistical test was utilized to investigate potential differences in gender variation between two distinct groups. P-values < 0.05 were reported as statistically significant.

Demographics, behavioral assessment and detection of blood biomarker results
There were no variations in age (P = 0.182), gender (P = 0.653) or education (P = 0.072) among the PSCI and the control groups. The MMSE and MoCA scores were significantly reduced throughout the PSCI groups (P < 0.01) ( Table 1). According to the control group, the PT% and high-density lipoprotein reduced (P < 0.05) and the INR ratio reduced (P < 0.05) in the PSCI group (Table 2).

Isolation and characterization results of plasma exosomes
The content detection results of plasma protein in 0.5 mL fraction sample were shown in Figure 1A. The content of plasma protein in fractions 6-10 is relatively lower, and the purity of exosomes is relatively high. Therefore, the 5 vesicle fractions 6-10 were selected for subsequent analysis. Western blot analysis was utilized to detect the exosomal protein markers in all the plasma exosome samples. The exosomal markers of protein expression levels (TSG101, CD9, CD63, and CD81) were shown in Figure 1B. The exosome morphology and size were visualized by TEM ( Figure 1C). TEM observation revealed highly homogenous exosome combination with a regular round morphology having a diameter range of 30-200 nm. A representative laser scattering microscopy image of isolated exosomes is shown in Figure 1D. NTA was exploited to measure the size distribution of particles and view the isolated vesicle samples ( Figure 1E, 1F). The samples fluorescent detection examined in the scatter mode. The size is calculated by the diffusion behavior. The study determined that the mean size of the particles was 71.6 nm, while the concentration was estimated to be around 6.8E+7 particles/mL.

Differentially expressed exosome proteins determination among the control and PSCI groups
In total, 259 differentially expressed exosome proteins have been measured and determined between the control group and PSCI group by Label-free quantitative proteomics, of which 131 proteins demonstrated up-regulated expression and 128 proteins showed down-regulated expression. The differentially expressed proteins with PSCI/control ratios that are high or less than 1.2-fold change having a set P-value < 0.05 were shown to be significantly changed. The differentially expressed proteins volcano plot was presented in Figure 2. The heat map of the whole differential expressed proteins were presented in Figure 3. The identification results of main 30 upregulated proteins were presented in Table 3 Table 4 and the other 98 down-regulated proteins were shown in Supplementary Table 2. The results of these proteins are summarized in detail. According to Table 5, the biological process category of 30 up-regulated proteins and molecular function were conducted according to biological function. The biological process category of 27 down-regulated proteins and molecular function were shown in Table 6.

GO analysis outcomes of the differentially expressed proteins
A GO analysis was employed to categorize the protein classification, molecular function, cellular composition, biological process, and mechanism.

PPI networks analysis results of the up-regulated proteins
Utilizing STRING analysis, a controlled PPI network with high-quality was constructed. 127 up-regulated proteins were suitable for PPI network analysis (focus molecule) and PPI networks with high-quality were built according to the STRING database. A complete PPI regulation network with 127 up-regulated proteins were presented in Figure 8. The network clustering results showed that the PPI network consists of six specified function clusters that comprise proteins with similar functions and are expressed by various colors (Figure 9). These six visualized interaction function clusters (sub-networks) were related to degradation of ubiquitinated proteins and folding of proteins (lime green), calcium-dependent protein binding and ESCRT III complex disassembly (yellow), cytoskeleton reorganization and platelet aggregation (green), phospholipid scrambling of phosphatidylserine in platelets and ATP mediates synaptic transmission (purple), lipid binding, signal transduction (red), and  Table 7.

PPI networks analysis results of the down-regulated proteins
118 down-regulated proteins were suitable for PPI network analysis (focus molecule) and a controlled PPI networks with high-quality were constructed according to the STRING database. A complete regulation of PPI network by down-regulated proteins were presented in Figure 10. The network clustering result showed that the PPI network consists of six specified function clusters that comprise proteins with similar functions and are expressed by various colors (Figure 11). These six visualized interaction function clusters  A SLe(x)-type proteoglycan, which through high affinity, calcium-dependent interactions, mediates rapid rolling of leukocytes over vascular surfaces in inflammation.
428 Figure 10. The PPI regulation network of down-regulated proteins in the PSCI group.
(sub-networks) were related to protein localization to juxtaparanode region of axon (yellow), cell adhesive protein binding, Fibrin Clot formation (green), mRNA splicing and RNA recognition (red), complement activation, lipid metabolism (purple), protein trafficking and cytoskeleton remodeling (lime green), and ATPdependent degradation of ubiquitinated proteins (blue), respectively. The anticipated functional intermediate partners in the PPI network are presented in Table 8.

ELISA quantitative determination results of plasma proteins
Compared with control group, human 14-3-3 protein zeta/delta (YWHAZ) and human brain-specific angiogenesis suppressor 1 correlated protein 2 (BAIAP2) levels of plasma were significantly increased (P < 0.01) while that of human IgD (IGHD), human ATP binding cassette subfamily B member 6, mitochondrial (ABCB6) and human heat shock protein 60 (HSPD1) were significantly decreased in patients with and without PSCI (P < 0.01) ( Figure 12).

General comments
To further explore the molecular mechanism of cognitive impairment, label-free quantitative proteomics were employed to analyse the differential expressed Figure 11. The PPI network means clustering of down-regulated proteins in the PSCI group. The PPI network is clustered to a specified number of clusters.
www.aging-us.com Table 8. The symbols and full names of the predicted functional intermediate partners in the PPI network of down-regulated expressed proteins shown in Figure 10.  proteins of plasma exosome in PSCI patients. Proteomics identified 259 differentially expressed proteins, containing 131 upregulated proteins and 128 downregulated proteins. These upregulated proteins are connected to ubiquitinated proteins degradation, calcium dependent protein binding, reorganization of cytoskeleton and platelet aggregation and blood coagulation. These downregulated proteins are related to protein localization to juxtaparanode region of axon, cell adhesive protein binding, fibrin clot formation, complement activation, lipid metabolism and ATPdependent degradation of ubiquitinated proteins. The mechanisms of cognitive impairment of PSCI are related to blood flow regulation, energy metabolism, protein folding and degradation, cell apoptosis, synaptic plasticity. These were discussed in detail below.

Blood flow regulation associated proteins
Plasma kallikrein, a multifunctional serine protease associated with activation of contact coagulation [12]. Plasma kallikrein mechanisms of action can be utilized to support pro-thrombotic or anti-thrombotic characteristics. The kallikrein-kinin system suppresses thrombin-induced platelet activation, indicating an antithrombotic function [13]. Plasma kallikrein decreased collagen-induced platelet activation via binding collagen [14]. Whereas, the effect of pro-thrombotic is suggested by the plasma kallikrein critical role in contact activation by conversion of FXII to FXIIa. Additionally, plasma kallikrein converts prorenin to renin, which then converts angiotensinogen to angiotensin I [15]. Plasma kallikrein had been implicated in contributing to both hematoma expansion and thrombosis in stroke [16]. The outcomes of this investigation revealed that the plasma kallikrein expression was downregulated proteins. Plasma kallikrein may influence the occurrence and development of acute stroke through the activation and transformation pathway of FXII.
Platelet glycoprotein V (gpV), is a membrane constituent which containing an 82 kDa relative molecule mass, correlates with the leucine-rich proteins family. It is only expressed in platelets and megakaryocytes, and is non-covalently correlated with the gpIb-IX complex to develop a receptor for von Willebrand factor and thrombin [17,18]. Hence, the GPIb-V-IX complex serve as the vWF receptor and modulates adhesion of vWF-dependent platelet to blood vessels. Platelet adhesion to damaged vascular surfaces in the arterial circulation is a crucial initiating event in hemostasis [19]. Platelet glycoprotein V may utilize as a platelet activation in vivo marker in thrombotic conditions. The expression of plateletglycoprotein V in patients suffering from acute stroke is elevated, which is consistent with the study of Amin HM et al. [20]. Therefore, platelet glycoprotein V may play a protective role in the brain through blood coagulation. Fibrinogen like protein 1 (FGL1) is a released protein having mitogenic effect on primary hepatocytes. FGL1 includes an N-terminal signal recognition peptide, a potential N-terminal coil-coil domain, a C-terminal fibrinogen related domain (FReD) and multiple cysteines presumably utilized for inter and intra molecular disulfide bonds [21]. FGL1 may perform a potential function in these processes such as proliferation, angiogenesis, apoptosis and extracellular matrix modulation like structurally comparable proteins (angiopoietins, tenascins, fibrinogen) [22,23]. Furthermore, the presence of FGL1 in the serum of rats after cytokine stimulation indicates that it could function as a significant biomarker for systemic inflammation [24]. Therefore, FGL1 may mediate the inflammatory response directly or indirectly in acute stroke. Our result of the increased expression of this protein just confirms this hypothesis.

Energy metabolism
ATP-binding cassette sub-family B member 6 (ABCB6), a member of adenosine triphosphate-binding cassette (ABC) transporter the family. It binds with heme and porphyrins and have a function in their ATPdependent uptake into the mitochondria and plays a crucial role in the synthesis of heme [25]. Some researches have found that ABCB6 expression is protective against various results elevating oxidative stress, including exposure of arsenite [26,27]. The outcomes of this investigation revealed that the expression level of ABCB6 reduced in patients suffering from acute stroke. The ABCB6 expression and function of closely related to the oxidation mechanism of mitochondria [25]. Therefore, ABCB6 may serve a protective function in the brain via antioxidant mechanisms.

Synaptic plasticity
Brain-specific angiogenesis inhibitor 1-associated protein 2 (BAIAP2), adapter protein that connects membrane-bound small G-proteins to cytoplasmic effector proteins. Subsequent researches have conclusively proven BAIAP2 serves as an essential regulator of membrane and actin dynamics at subcellular structures rich in actin, such as filopodia and lamellipodia [28][29][30]. Actin skeleton and its dynamics play an important role in the excitatory synaptic transmission and plasticity regulation [31,32]. The results of this investigation revealed that the expression level of BAIAP2 increased in individuals with acute stroke. In the brain, BAIAP2 may perform a neuroprotective function by improving synaptic function.

CONCLUSION
In conclusion, the present study found 259 differentially expressed proteins such as 131 upregulated proteins and 128 downregulated proteins using label-free quantitative proteomics approach in plasma exosome of PSCI patients. The findings suggested that the mechanism of cognitive impairment may be related to blood flow regulation, energy metabolism, protein folding and degradation, cell apoptosis, synaptic plasticity, stress response and protein phosphorylation in PSCI patients. Therefore, these proteins may be target-related proteins and shed light on pathogenesis mechanisms on a global scale of cognitive impairment at plasma exosome proteins level in PSCI patients. The disorders of plasma exosome proteomics may be explained the cognitive impairment in PSCI patients. Further association studies need to be clarified.

CONFLICTS OF INTEREST
The authors declare no conflicts of interest related to this study.