Changes in the Proteome of the Circle of Willis during Aging Reveal Signatures of Vascular Disease

Approximately 70% of all strokes occur in patients over 65 years old, and stroke increases the risk of developing dementia. The circle of Willis (CoW), the ring of arteries at the base of the brain, links the intracerebral arteries to one another to maintain adequate cerebral perfusion. The CoW proteome is affected in cerebrovascular and neurodegenerative diseases, but changes related to aging have not been described. Here, we report on a quantitative proteomics analysis comparing the CoW from five young (2–3-month-old) and five aged male (18–20-month-old) mice using gene ontology (GO) enrichment, ingenuity pathway analysis (IPA), and iPathwayGuide tools. This revealed 242 proteins that were significantly dysregulated with aging, among which 189 were upregulated and 53 downregulated. GO enrichment-based analysis identified blood coagulation as the top biological function that changed with age and integrin binding and extracellular matrix constituents as the top molecular functions. Consistent with these findings, iPathwayGuide-based impact analysis revealed associations between aging and the complement and coagulation, platelet activation, ECM–receptor interaction, and metabolic process pathways. Furthermore, IPA analysis revealed the enrichment of 97 canonical pathways that contribute to inflammatory responses, as well as 59 inflammation-associated upstream regulators including 39 transcription factors and 20 cytokines. Thus, aging-associated changes in the CoW proteome in male mice demonstrate increases in metabolic, thrombotic, and inflammatory processes.


Introduction
Aging is a complex biological process characterized by numerous physiological and structural changes.Age-related alterations in the cerebral arteries of humans predispose to neurodegenerative diseases and cognitive decline and promote cerebrovascular diseases such as intracerebral hemorrhage, microbleeds, and stroke [1,2,3].While resistance arteries, penetrating and parenchymal arterioles, and cerebral capillaries are the principle sites for age-dependent loss of functionality and disease development, the larger arteries of the mouse circle of Willis (CoW) display reduced cerebral blood flow and elastin deposition with aging, as well as increased arterial stiffness and collagen deposition [4].In the aged human middle cerebral artery, functional changes in both vasodilation and constriction have been reported [5].Elucidating changes in the proteome provides insight into the mechanisms that trigger arterial aging and is expected to facilitate the development of treatments for, or the prevention of, cerebrovascular diseases that are linked with old age.
The CoW is a ringlike arterial structure at the base of the brain and ensures collateral blood flow between the posterior and anterior cerebral arterial systems [6].It consists of the two branches of the internal carotid artery, the vertebrobasilar artery, and the anterior and posterior communicating arteries [7].Healthy circulation is vital for cerebral perfusion and function and to prevent ischemia-associated damage [8,9].Several molecular and physiological studies revealed changes in the structure of the CoW and its surface branches and in protein expression within the CoW in healthy aging [10,11] and in cerebrovascular and neurodegenerative diseases [10,12].Also, studies of brains of aged mice and humans have reported hypoplasia of the vascular wall [13], decreases in cerebral blood flow [14], ischemic stroke [15], changes in the thickness of the arterial wall, and loss of elasticity of the CoW arteries [16].In animal models, age-induced changes in brain microvessels included disruption of the blood-brain barrier [17] with changes of the basement membrane [18], decreased blood velocity [19], impairment of bioenergetic pathways [20], and increases in oxidative stress [18].However, information on age-induced impairment specifically in the CoW is scarce.Although recent quantitative analyses of the proteome have contributed to our understanding vascular and neurodegenerative disorders [21,22], few studies have focused on agerelated changes in the cerebral arteries of the CoW [16,23].
In this study, we performed an unbiased, quantitative analysis of the proteomes in the cerebral arteries of the CoW of young (2-3-month-old) and aged (18−20-month-old) mice.The primary goal was to characterize the impact of aging on expression and relative abundance of proteins in cerebral arteries of young and aged healthy mice; the secondary goal was to identify protein networks and pathways that may cause age-related changes in vascular structure and function and increase the risk of cerebrovascular dysfunction in the elderly population; and the tertiary goal was to identify age-related changes in the proteomes of specifically endothelial and smooth muscle cells of the CoW arteries.

Materials and Methods
2.1.Animals and Housing.All experimental procedures were approved by the Institutional Animal Care and Use Committees of the University of Iowa and the Iowa City VA Health Care System, and they complied with the standards of the Institute for Laboratory Animal Research.To avoid genderrelated confounding differences, only male C57BL/6J mice were used.Five mice each were allocated to the young and aged groups, for analysis of the CoW proteome at 2-3 months of age and 18-20 months of age, respectively.Mice were housed in temperature-controlled rooms and maintained on a dark/light cycle of 12 hr, standard rodent chow, and water ad libitum.

Isolation of CoW and Preparation of Protein Samples.
Mice were euthanized by inhalation of 100% CO 2 followed by harvest of the brain according to an institutionally approved euthanasia protocol.After the brain was isolated, the cerebral arteries of the CoW (anterior, posterior and middle) and their main branches were surgically removed from the base of the brain [24].Thereafter, each isolated CoW was rinsed with cold Dulbecco's phosphate-buffered saline solution (DPBS) (Gibco # 2430024) and stored at −80°C until further use.For protein extraction, the CoW was mixed with 80 μL of RIPA buffer (Fisher Scientific # R0278) containing protease (PierceTM # A32963) and phosphatase inhibitor (PierceTM # A32957) cocktails and transferred to fresh Eppendorf tubes (Corning # 3207).The mixture was vortexed three times for 30 s (Fisher Scientific # 0215370) and gently agitated for 30 min at 15-20°C.Samples were then vortexed and incubated (VWR # 12621-112) at 95°C for 10 min.Samples were spun down and subjected to tissue lysis using a Covaris E220 focused ultrasonicator (Covaris # 500239).For this purpose, samples were transferred to Covaris microfiber Screwcap tubes (Covaris # 520216) and kept on ice during the procedure.The instrument parameters for shearing were as follows: water level set point 10, water temperature 6°C, peak incident power 175 W, duty factor 10%, cycles per burst 200, and duration 300 s.The homogenized lysates were transferred to fresh Eppendorf tubes and centrifuged at 16,100x g (Eppendorf # 5415R) for 30 min.The supernatant was collected and stored at −80°C until further use.The total protein concentration was measured using a Bicinchoninic acid (BCA) protein assay kit (PierceTM # 23225).A 30-µg sample of protein from each biological replicate was used for quantitative proteomics analysis.

Protein Digestion.
Protein digestion was performed according to a previous publication [25] using an single pot, solidphase enhanced sample preparation (SP3) method [26].For each sample, 30 μg of protein in 150 μL lysis buffer was incubated with 10 μL of SP3 beads (1 : 1 mix of Sera-Mag Speed Beads A and B (Thermo Scientific)).Pure acetonitrile (ACN) (VWR Chemicals) was added (directly to the samples) to a final concentration of 70% (v/v).Samples were incubated in a thermal shaker for 18 min at 800 rpm and then transferred to the magnet rack for 3 min to immobilize the SP3 beads.The supernatant was discarded, and the SP3 beads were rinsed three times with 1 mL of 80% (v/v) ethanol/water and once with 800 μL of ACN.The bound proteins were reduced by adding 100 μL of 10 mM 1,4-dithiothreitol (DTT) (Sigma) in 50 mM ammonium bicarbonate (Sigma), pH 8.0, incubated at 37°C with shaking at 800 rpm for 1 hr.Proteins were alkylated with 55 mM 2-chloroacetamide (CAA) (Sigma) for 1 hr at 37°C, in the dark.Finally, 1 μg of trypsin (Thermo Scientific) was added, and samples were incubated overnight at 37°C with shaking at 800 rpm.After protein digestion, samples were acidified with formic acid (FA, Carlo Erba) at a final concentration of 1% (v/v), dried in vacuo, and stored at −80°C until further use.

Automated
Off-Line Fractionation.For whole proteome analysis, peptides were resuspended in 110 μL of buffer A (25 mM ammonium formate (Sigma), pH 10), and subjected to high pH reverse phase fractionation using the AssayMAP Bravo platform and 5 mL RP-S cartridges (Agilent).The cartridges were primed sequentially with 150 μL isopropanol (Chemsolute), ACN, and buffer B (80% ACN in 10 mM ammonium formate, pH 10), at a flow rate of 50 μL/min.Subsequentially, cartridges were equilibrated with 100 μL of buffer A, and peptides were then loaded at 5 μL/min.The flow-through (FT) was collected.Peptides were then eluted with 25 mM ammonium formate, pH 10, using increasing concentrations of ACN (5%, 10%, 15%, 20%, 25%, 30%, and 80%).The seven flow-through fractions were pooled into The match-in-between runs and the second peptide options were enabled.The MaxLFQ algorithm [28] was used for labelfree quantification (LFQ).The mass spectrometry proteomics data have been deposited in the ProteomeXchange Consortium via the PRIDE partner repository [29] with the dataset identifier PXD043001.
2.7.Proteomics Data Analysis.Data analysis was performed using the Perseus software (version 2.0.7.0.).Protein identifications were filtered to remove contaminants and decoy hits before performing data normalization of the log2-transformed LFQ intensity values by median centering, as implemented in Perseus.For statistical analysis, only proteins that had been quantified in at least four biological replicates were retained, and missing values for the fifth replicate were imputed from the normal distribution in Perseus, using default parameters.For all samples, the amount of imputed data was below 10%.Supplementary 1 lists all measurements (before imputation) and Supplementary 2 data sets after imputation.Proteins found to be significantly regulated were identified using the Student's t-test with an S0 parameter of 0.1 and corrected for multiple hypotheses using a permutation-based FDR of 5%.Significantly dysregulated proteins were defined as those for which at least two unique peptides were present at a ratio of <0.77-fold or >1.30-fold following correction of p values according to previous publications [30,31,32].
2.8.Pathway Enrichment Analysis.The hierarchical cluster analysis of all dysregulated proteins was performed using GraphBio [33] and SR plot, according to a previous publication [34].A volcano plot of differentially expressed proteins was also constructed using VolcaNoseR [35].Enrichment analysis of differentially expressed proteins was performed to identify the biological processes, cellular locations, molecular functions, and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that they influence.These analyses were done using the HemI 2.0 (https://hemi.biocuckoo.org/)[36], ExpressAnalyst (www.expressanalyst.ca),and Enrichr [37] webtools.The canonical pathway and upstream regulator analyses were performed using the INGENUITY Pathway Analysis (IPA) software (http://www.INGENUITY.com)[38].For the analysis, we applied a threshold of −log (p value) >2.Pathways with a z-score of >2.0 were considered activated and those with a score of <-2.0 inhibited.The impact pathway analysis on differentially expressed proteins was executed using iPathwayGuide [39,40].iPathwayGuide pathway annotations were obtained from the KEGG database, release 100.0+/11−12(Nov 21), and gene ontology annotations from the Gene Ontology Consortium database, release 2021 November 4 [41].The pathways identified were considered to be affected if the FDR-corrected p value threshold was <0.05.Analysis of protein functions and protein-protein interactions, as well as functional characterization of dysregulated proteins, was performed by coupling the STRING and Reactome databases (http://string-db.org)[42,43].To identify the age-associated changes in the mitochondrial proteome of the CoW, data were mapped to MitoCarta 3.0 database, an inventory of mammalian mitochondrial proteins and pathways [44].In addition, we performed a comparison of proteomics data from our study and from a previous study [16].Novel proteins that were identified in our dataset only and all identified proteins from current study and previous study were analyzed using ExpressAnalyst webtool (www.expressanalyst.ca)for pathway enrichment analysis (KEGG pathway) and gene ontology analysis for cellular components.Top enriched pathways, total number of proteins associated to pathway, and their cellular locations in aging CoW were studied.Top 10 enriched KEGG pathways and cellular components from two datasets were selected and compared to each other.Age-associated changes in the endothelial and vascular smooth muscle proteins of the CoW were identified by mapping the significantly dysregulated proteins to an in-house human microvascular endothelial cell Oxidative Medicine and Cellular Longevity proteome dataset (unpublished study) and a published vascular smooth muscle proteome dataset [45].
2.9.Statistical Analysis.All experiments were performed in five biological replicates.The data were expressed as mean with standard error, and the statistical analysis was performed using GraphPad Prism 9.0 software.Normal distribution was assessed by D'Agostino-Pearson omnibus normality test.An unpaired Student's t-test or the nonparametric Mann-Whitney test was used to determine statistical significance for comparison of the two groups.Differences were considered as significant if the p values were <0.05.

Proteomic Profiling of the Mouse CoW.
To understand age-induced changes in protein expression in CoW arteries, we performed quantitative proteomics using LC-MS/MS on samples from 2-3 months old (young) and 18-20 months old (aged) mice (Figure 1(a)).We identified 6,874 proteins in the two groups, of those, 3,926 proteins were quantified with at least two unique peptides (Supplementary 1 and 2).Labelfree quantitation (LFQ) of the measured proteins among the replicates was highly reproducible, as assessed by the Pearson correlation (Figure 1

(b)). Principal component analysis (PCA)
showed that age is a key variable (Figure 1(c)).The distribution of unique peptides among the identified, quantified, and significantly dysregulated proteins is shown in Figure 1(d).A protein was classified as identified if a valid MS/MS spectrum was available for at least one of its peptides: as quantified if the protein was identified in all five biological replicates and as significantly dysregulated if its relative abundance in the two groups differed with statistical significance.Among 3,926 quantified proteins for which at least two unique peptides were available (Supplementary 3), 242 proteins had significantly different expression profiles in the aged versus young groups (AE1.3-fold;p <0:05) (Table 1).We generated a hierarchical cluster heatmap using the LFQ intensity of all significantly dysregulated proteins.This revealed a strong difference in the protein expression in the cerebral artery of the CoW between young and aged mice (Figure 2(a)).Among the 242 proteins, 189 were significantly upregulated and 53 downregulated (Supplementary 4).Volcano plots of all significantly down-or upregulated proteins are shown in Figures 2b) and 2(c).To understand the biological significance of the observed dysregulation, we analyzed the list of up-and downregulated proteins using Enrichr-KG bioinformatics webtool, with an FDR of p <0:05 for biological process.The processes driven by the 189 upregulated proteins included platelet degranulation (GO:0002576), regulated exocytosis (GO:0045055), negative regulation of blood coagulation (GO:0030195), and extracellular matrix organization (GO:0030198).Those driven by the 53 downregulated proteins contribute to intermediate filament bundle assembly (GO:0045110) and supramolecular fiber organization (GO:0097435).
Given that aging alters the mitochondrial proteome [46,47], we mapped our data to the mouse MitoCarta mitochondria proteome database.Of 761 mitochondrial proteins identified, 548 were quantified.Among these, 21 proteins were significantly dysregulated in the aged CoW (Table 2), among 20 proteins were upregulated.All dysregulated mitochondrial proteins were involved in fatty acid oxidation (GO:0019395), fatty acid catabolism (GO:0009062), of fatty acid β-oxidation by acyl-CoA dehydrogenase (GO:0033539).
A previous study had analyzed the proteomes of arteries of the CoW in 6-month-old mice [16].It identified a total of 6,627 proteins and 2,188 with at least two unique high-scoring peptides.That study deployed two proteomics approaches, gel-free nano-LC-mass spectrometry (MS)/MS and gel-based GelLC-MS/MS with spectrometry.In our analysis, we used the Adaptive Focused Acoustics (AFA) technology to prepare the samples.Our analysis revealed an additional 3208 proteins not identified by previous study (Supplementary 5).By combining the previously identified proteins and those uniquely identified in our study, we created a CoW proteome database (Supplementary 5) for a list of 9835 CoW proteins.Gene enrichment analysis of the 3,208 novel proteins using the KEGG pathway revealed that the newly identified proteins are involved in metabolic pathways (n = 350), endocytosis (n = 75), MAPK signaling (n = 62), spliceosome (n = 56), and RNA transport (n = 54) (Supplementary 6).Furthermore, the current study detected additional proteins by KEGG pathways and GO analysis for cellular components in pathways previously implicated.(Supplementary 6 and 7).Further analyses were performed to identify proteins specific to vascular endothelial and smooth muscle cells.All 242 significantly dysregulated proteins were compared to our in-house generated human microvascular endothelial cell proteome (unpublished data) and published vascular smooth muscle proteome data [45].We mapped 148 significantly dysregulated proteins to the human microvascular endothelial cell proteome, with 111 of these upregulated and the other 37 downregulated.Similarly, 22 significantly dysregulated proteins mapped to the vascular smooth muscle proteome, with 16 upregulated and 6 downregulated.This result suggests that aging-induced changes are more pronounced in endothelial than vascular smooth muscle cells (Supplementary 8 and 9).Gene ontology (GO) analysis of the dysregulated proteins revealed association (based on corrected p-value) for 33 GO terms for biological process, most prominently blood coagulation and cell-matrix adhesion and 22 GO terms for molecular function, such as integrin binding and extracellular matrix structural constituents (Figures 3(a) and 3(b)).Dysregulated proteins also enriched for 27 GO terms for cellular components, including collagen-containing extracellular matrix, extracellular space (Figure 3(c)).

Canonical Pathway and Upstream Regulator Analysis.
Canonical pathways that are altered in the aged CoW were identified by core analysis using the IPA software.Ninetyseven enriched canonical pathways were affected by applying the −log (p value) >2 threshold.The top 15 of the canonical pathways are shown in Figure 4(a) and Supplementary 10.Most of the activated canonical pathways contribute to inflammation, immune response, cytokine, and integrin-mediated signaling, and the top three were phagosome formation, neutrophil     Next, we performed an analysis of upstream regulators including transcription regulators, ligand-dependent nuclear receptors, cytokines, and growth factors.Application of a p value of overlap <0.05 as a threshold revealed enrichment of 273 transcription regulators and 63 cytokines, including 29 that were activated and 10 that were inactivated.The top five activated transcription regulators were IRF1, IRF3, IRF7, STAT1, and BHLHE40.The top inhibited transcription regulators were IRF2BP2, TRIM24, ETV6, SMARCA5, and ETV5 (Figure 4(b)).Further reactome pathway analysis showed that most activated transcription regulators are involved in the JAK/STAT, PDGF, IFNγ, and interleukin signaling pathways (Supplementary 11).For the predicted cytokines, 19 were classified as activated and one as inhibited.The top three activated cytokines are IFNγ, TNF, and IFNA2, and the networks of differentially expressed genes regulated by these cytokines are shown in Figures 4(c), 4(d), and 4(e).The reactome pathway analysis revealed that the activated cytokines are involved in processing and signaling by IL1, IL4, IL6, IL10, and IL13 (Supplementary 12).Oxidative Medicine and Cellular Longevity 13

Discussion
In this study, we performed an unbiased, quantitative analysis of the proteomes in cerebral arteries of the CoW of young (2-3-month-old) and aged (18−20-month-old) mice to characterize the impact of aging on expression and relative abundance of proteins in cerebral arteries.The age ranges of the mice we analyzed (2-3 months and 18-20 months) correspond to approximately 16-20 and 56-60 years of age in humans, representing young adulthood and late middle age.
Of note, the incidence of stroke and dementia is increasing in this age group at a faster rate in on older patients [50,51].Of 6,874 proteins that we identified in the CoW arteries, 242 proteins changed in abundance between these time points.Most of these proteins are known to contribute to blood coagulation, platelet dysfunction, altered extracellular matrix composition, and metabolic processes.Activated canonical pathways relate to inflammation and immune responses, including cytokine-and integrin-mediated signaling and activation of transcription factors (e.g., IRF1, IRF3, STAT1) and cytokines (e.g., IFNα, IFNγ, and TNF).Aging also affected the metabolism of CoW arteries, including increasing the abundance of proteins that catalyze various steps of fatty acid βoxidation in the mitochondria.Our study adds to previous reports on the cerebral artery proteome, specifically in the CoW, by providing data on changes induced by aging.A previous study by Badhwar and colleagues analyzed the CoW proteome in 6-month-old male mice of the same genotype [16].Three thousand sixty-two of the proteins we detected overlapped with those reported in the previous study, and the five most abundant and significant in both datasets were Col6a1 (collagen alpha-1(VI) chain), Lama5 (laminin subunit alpha-5), Lamb2 (laminin subunit alpha-2), Myh9 (myosin-9), and Pdlim7 (PDZ and LIM domain protein 7).We detected 3,208 additional proteins that were not identified in the other study, many of them involved in metabolic pathways, endocytosis, and spliceosome.In our data set, 148 endothelial cell and 22 smooth muscle cell proteins were Oxidative Medicine and Cellular Longevity dysregulated during aging.Several highly dysregulated endothelial and smooth muscle cell proteins had previously been related to aging by methods other than proteomics.Among those that are endothelial cell lactadherin [52] and smooth muscle protein thrombospondin-1 [53].Among those reduced were collagen α-1 and α-2 [54].
Changes in the lipid metabolism play an important role in the aging process [55,56].Our data demonstrate that various acyl-CoA dehydrogenases are upregulated and fatty acid synthase is downregulated with aging.A prior study reported that the relative abundance of glycolytic proteins was reduced in cerebral microvessels of 20-month-old versus 14-month-old mice [18].Our study design does not allow us to attribute metabolic changes to specific cell types of the vascular wall.Of note, differential metabolic activity has been reported in cells of the vascular wall: differentiated smooth muscle cells in the intact vascular wall use oxidative phosphorylation to fuel vascular reactivity [57], whereas endothelial cells rely on glycolysis for their baseline metabolic needs and on mitochondrial respiration for specialized processes such as transendothelial Oxidative Medicine and Cellular Longevity migration [58].How a shift toward fatty acid catalysis with aging as implicated by our data affects vascular wall physiology remains to be elucidated.Whereas the proteomes of the cerebral macro-and microvasculature have been established, age-related changes are incompletely studied [16,59,60,61].One study combined samples from middle cerebral arteries and small mesenteric resistance arteries of 3-and 14-month-old C57Bl/6J mice, and the identified 31 proteins were significantly affected by age [23].The most up-and downregulated proteins included myosin, Ig kappa chain V, laminin, and fibulin, of which other paralogs were also identified in our data set.Additional biological processes identified in our dataset included thrombosis, inflammation, and ECM remodeling all of which have been implicated in aging of the vasculature [1,62,63,64].
Our study has several limitations.Firstly, we did not perform extensive validation of our data with an alternative method due to the limited quantity of protein in CoW samples.Secondly, our data were generated in only male mice, and therefore, the findings cannot be extrapolated to females without experimental confirmation.Lastly, we studied mice at only two ages, corresponding in humans to young adults and adults at about 60 years.An understanding of proteome changes associated with aging will require investigation of protein abundance at more advanced ages, corresponding to human age of 65-76 years and above.The current evidence is not sufficient to answer whether preventing stroke and cerebral hemorrhage in older versus younger patients will require additional or different vascular care.Further research on the progression of proteomic changes in the CoW of elderly patients is needed to produce the evidence base necessary to address this question.Oxidative Medicine and Cellular Longevity

FIGURE 1 :
FIGURE 1: Quantitative proteomics analysis of the CoW.(a) Workflow for high-throughput identification and quantification of proteins from the circle of Willis (CoW) of young and aged mice: CoW isolation, tissue lysis and protein extraction, sample preparation, mass spectrometry, and data analysis.(b) Matrix representation of Pearson correlation values based on label-free quantification (LFQ) intensities, quantifying reproducibility in samples from young and aged mice.(c) Plot of principal component analysis (PCA) outcomes for proteins identified in biological replicates of young and aged mice.(d) Distribution of the number of unique peptides in all identified, quantified, and significantly dysregulated proteins.

FIGURE 3 :
FIGURE 3: Gene ontology (GO) enrichment analysis of proteins that are dysregulated in the aged CoW.(a) GO biological process, (b) GO molecular function, and (c) GO cellular components.The top 10 significantly enriched GO terms are shown with their respective corrected p values and E ratios.Colored dots (dark to light) indicate enriched terms with their corrected p values, and size represents the E ratio for each enriched term.The analysis was performed using HemI 2.0 webtool.

FIGURE 4 :
FIGURE 4: Ingenuity pathway analysis (IPA) of proteins that are dysregulated in the aged CoW.(a) Top 15 predicted canonical pathways based on z-score and B-H p value <0.05 (B-H p value was determined using Fischer's exact test and adjusted for multiple comparisons according to the method from Benjamini-Hochberg).A positive z-score implies activation (orange), and a negative z-score inhibition (blue), of the pathway; longer bars indicate stronger significance than the shorter bars (http://www.INGENUITY.com).(b) Heatmap of the top 10 transcription regulators predicted to be activated or inhibited based on z-score (>2 for activated and <-2 for inhibited).(c-e)Predicted activation or inhibition of inflammatory molecules in the aged CoW, and their gene networks, for (c) IFNα2, () IFNγ, and (e) TNF with their gene networks.Genes predicted, based on parameters described in (b), to be upregulated are marked in red, and those predicted to be downregulated are marked in green, as indicated in the prediction legend.

FIGURE 5 :
FIGURE 5: Impact pathway (iPathway) analysis of proteins that are dysregulated in the aged CoW.(a-d) Bar graphs of dysregulated genes that map to the top four pathways identified by iPathway analysis based on FDR-corrected p value significance: (a) complement and coagulation cascades, (b) platelet activation, (c) ECM-receptor interaction, and (d) neutrophil extracellular trap formation.All differentially expressed genes in each pathway were ranked based on their absolute value of log-fold change.Upregulated genes are shown in red, and downregulated genes in blue.The box-and-whisker plots on the left summarize the distributions of all the differentially expressed genes in their pathways.The box represents the 1 st quartile, median, and 3 rd quartile; circles represent the outliers (© Advaita Corporation 2023).

TABLE 1 :
List of all significantly dysregulated proteins in the aged cerebral arteries of the circle of Willis.

TABLE 2 :
List of all significantly dysregulated mitochondrial proteins in the aged cerebral arteries of the circle of Willis.