Impact of Age on the Cerebrovascular Proteomes of Wild-Type and Tg-SwDI Mice

The structural integrity of cerebral vessels is compromised during ageing. Abnormal amyloid (Aβ) deposition in the vasculature can accelerate age-related pathologies. The cerebrovascular response associated with ageing and microvascular Aβ deposition was defined using quantitative label-free shotgun proteomic analysis. Over 650 proteins were quantified in vessel-enriched fractions from the brains of 3 and 9 month-old wild-type (WT) and Tg-SwDI mice. Sixty-five proteins were significantly increased in older WT animals and included several basement membrane proteins (nidogen-1, basement membrane-specific heparan sulfate proteoglycan core protein, laminin subunit gamma-1 precursor and collagen alpha-2(IV) chain preproprotein). Twenty-four proteins were increased and twenty-one decreased in older Tg-SwDI mice. Of these, increases in Apolipoprotein E (APOE) and high temperature requirement serine protease-1 (HTRA1) and decreases in spliceosome and RNA-binding proteins were the most prominent. Only six shared proteins were altered in both 9-month old WT and Tg-SwDI animals. The age-related proteomic response in the cerebrovasculature was distinctly different in the presence of microvascular Aβ deposition. Proteins found differentially expressed within the WT and Tg-SwDI animals give greater insight to the mechanisms behind age-related cerebrovascular dysfunction and pathologies and may provide novel therapeutic targets.


Introduction
The structural integrity of the cerebrovasculature is altered during ageing. These changes can be accelerated by vascular pathologies, most notably cerebral amyloid angiopathy (CAA) [1], [2], [3], [4], [5], [6], [7]. Ageing alone can lead to vessels supplying deep white matter to become tortuous. Thickening of the veins and venules is also commonly observed along with increased thickening of the basement membrane [1]; [8]. A decline in vascular density, specifically capillary loss, has been observed in both the elderly and aged animal models [9], [10] along with increases in arteriole and capillary diameter [11]. Many of these same age-related pathologies are recapitulated in brains with CAA, which is relatively common in the elderly population, presenting in 10-40% of the non-demented individuals over the age of 60 [12]. CAA can be categorized into two types identified pathologically by the accumulation of Ab in leptomeningeal and cortical vessels including capillaries (Type 1) or with the exception of capillaries (Type 2) [13], [14]. CAA can lead to the loss of smooth muscle cells and thickening of the vessel walls, which in turn causes the blood vessel to become compromised, leading to intracerebral hemorrhage [15].
The presence of Ab leads to very different protein changes that are independent of age. This includes the modulation of proteins that are involved in movement, aggregation and degradation of Ab. APOE is a protein that is intimately involved in all of these processes [16]: the expression of the APOE4 allele leads to a propensity for increased Ab aggregation in sporadic AD [17] and increased risk of Ab deposition in capillaries [13]. APOE plays a direct role in Ab efflux across the blood brain barrier to circulating plasma [18], [19], and its overexpression promotes Ab proteolysis via enzymes such as neprilysin and insulin degrading enzyme [20]. Overexpression of neprilysin and insulin degrading enzyme decrease Ab levels in transgenic amyloid precursor protein (APP) mouse models and improve cognition [21].
The Tg-SwDI transgenic mouse model of amyloidosis expresses human APP with three different mutations: the Swedish, Dutch and Iowa mutations. The double Swedish mutation leads to increased production and secretion of Ab [22]. The Dutch and Iowa mutations, found in independent familial forms of CAA, increase the propensity for Ab to accumulate in the vessels and are associated with cortical hemorrhages [23]. The Tg-SwDI model expresses the human APP gene at levels similar to endogenous mouse APP levels [23], and yet develops a temporal accumulation of Ab. This latter accumulation of Ab is potentially attributed to properties of the mutant Ab preventing normal clearance [24].
As the Ab accumulates over time, it associates closely with microvessels, (seen in Type 1 CAA) and is primarily fibrillar in nature [23], [25]. This microvascular deposition is closely associated with an increased inflammatory response [23]. These alterations observed in the Tg-SwDI, resulting from advancing age and accumulating Ab, should be underpinned by changes in various proteins, including those associated with the neurovascular unit and the blood brain barrier [7], [26], [27], [28], [29] yet the scope of these changes are not fully understood.
To better understand how cerebral vessels globally respond to both age and accumulating Ab, we have used a quantitative labelfree proteomics approach to investigate the proteomic signatures of vessel-enriched fractions from the brains of WT and Tg-SwDI mice at 3 and 9 months of age. Our vascular-focused approach has revealed distinctly different protein expression patterns as a result of age in the presence of microvascular Ab accumulation.

Ethics Statement
All procedures were authorized under the Home Office approved project license number 60/3722. This license was approved by the University of Edinburgh's Ethical Review Committee and the Home Office and adhered to regulations specified in the Animals (Scientific Procedures) Act (1986).

Animals
Tg-SwDI mice were obtained from Jackson Laboratories (Bar Harbour, MA). These mice were generated and previously described [23] on a C57Bl/6J background. Homozygous Tg-SwDI mice were bred and maintained. C57Bl/6J (WT) mice were obtained from Jackson laboratories through their UK distributor Charles River.

Vessel Enrichment
Vessels were enriched using a protocol based on 2 sources [30], [31]. Hemibrains were homogenized in 1 mL of ice-cold PBS using a loose fit dounce homogenizer using 15 strokes. Homogenates were transferred to a 15 mL falcon tube and centrifuged for 10 minutes at 2506g. The pellet was re-suspended in 17.5% Ficoll solution (Sigma) and centrifuged for 25 minutes at 3,2006g to collect the pellet (vessel containing S1 fraction), which was retained on ice. The supernatant was centrifuged again for 25 minutes at 32006g to collect a second pellet and the pellets from both spins were combined in 1 mL of 1% BSA-PBS and re-suspended. The suspension was centrifuged for 10 minutes at 20006g. The pellet was subsequently washed in 1 mL of ice-cold PBS before being flash-frozen on dry ice and later stored at 270uC until further processing.
Immunoblots S1 fractions obtained from hemibrains of 3 month-old WT mice (n = 3) for vessel enrichment confirmation were homogenized in a lysis buffer of 9 M Urea/4% CHAPS with Complete Protease Inhibitor Cocktail (Roche, Manheim, Germany). Protein levels were determined using a Pierce BCA protein assay kit (Thermo Scientific, Cramlington, Northumberland, UK). Proteins were separated by SDS-PAGE and transferred onto nitrocellulose membranes. S1 fractions obtained from hemibrains of 3, 12 and 17 monthold Tg-SwDI mice (n = 4/group) for APOE and HTRA1 confirmation were homogenized in a sucrose buffer with Complete Protease Inhibitor Cocktail (Roche, Manheim, Germany). Protein levels were determined as above and proteins were separated by SDS-PAGE and transferred onto PVDF membranes (Bio-Rad, Germany).

Cerebral Vessel Ab Levels
Total Ab was extracted from independent S1 fractions taken from independent cohorts of 3, 12 and 17 month-old Tg-SwDI mice (n = 4/group) using a homogenisation buffer of the following constituents: 5 M guanidine and 50 mM Tris/HCL and mixed for 3-4 hours at RT and later stored at 220uC. The solution was diluted 1:10 in ice-cold reaction buffer of the following constituents: 0.2 g/L KCL, 0.2 g/L KH2PO4, 8 g/L NaCl, 1.15 g/L Na2HPO4, 5% BSA, 0.03% Tween 20 and 1X protease inhibitor cocktail (Calbiochem). The solution was centrifuged at 16,0006g for 20 minutes at 4uC. Protein levels were quantified with Pierce BCA Protein Assay Kit (Thermo Scientific). Ab1-40 levels were measured using the Invitrogen Human Ultrasensitive ELISA kit (Life Technologies, Paisley, UK) following the manufacturer's instructions.

LC-MS Analysis of Vessel-enriched Fractions
All chemicals were purchased from Sigma-Aldrich (UK) unless otherwise stated. Acetonitrile and water for LC-MS/MS and sample preparation were HPLC quality (Fisher, UK). Formic acid was Suprapure 98-100% (Merck, Darmstadt, Germany) and trifluoroacetic acid was 99% purity sequencing grade. Trypsin and Lys-C (modified, sequencing grade) were purchased from Roche Diagnostics (West Sussex, UK). All HPLC-MS connector fittings were from Upchurch Scientific or Valco (Hichrom and RESTEK, UK). 100 mg of protein extract from the vessel-enriched fractions from 3 and 9 month-old, Tg-SwDI and WT mice (n = 7-8/group) was re-suspended in 45 mL of 8 M urea, 2.5 mL of 200 mM DTT and 5 mL of 1 M ammonium bicarbonate. The samples were reduced at room temperature for 30 minutes, and then 2.5 mL of 500 mM iodoacetamide was added. 0.5 mg of Lys-C was added and digested overnight in 6 M urea followed by dilution to 2 M urea and overnight digestion with 2 mg of trypsin.
Capillary-HPLC -MSMS data were collected on an online system consisting of a micropump (1200 binary HPLC system, Agilent, Edinburgh, UK) coupled to a hybrid LTQ-Orbitrap XL instrument (Thermo-Fisher, Loughborough, UK). HPLC-MS methods have been described previously [32]. Peptides were reconstituted in 10 mL of loading buffer before injection and 8 mL was loaded. The peptide mixture was separated using a 140 min gradient form 0% to 100% Buffer B (Buffer A: 2.5% acetonitrile 0.1% formic acid and buffer B 90% acetonitrile, 0.025% trifluoracetic acid, and 0.1% formic acid).
The mass spectrometer was operated in 'data-dependent mode', with a single MS scan (400 to 2,000 m/z) in FT mode 60 K resolution followed by 5 MS/MS scans in the linear ion trap on the most abundant ions and dynamic excluded for 120 seconds.

Protein Identification and Quantification
Conversions from RAW to MGF files were performed as described previously [32], [33]. MS/MS data were searched using MASCOT Versions 2.3 (Matrix Science Ltd., London, UK) against a mouse plus contaminant database downloaded from NCBI with 30,061 sequences (From 12 January 2011) and amended with human APP with the Swedish, Dutch and Iowa mutations due to the transgenic nature of the mouse model. Variable methionine oxidation, STY phosphorylation, protein Nterminal acetylation, and fixed cysteine carbamidomethylation were used in all searches. Precursor mass tolerance was set to 7 ppm and MS/MS tolerance to 0.4 amu. LC-MS label-free protein were all significantly enriched in the S1 fraction compared to whole brain samples (n = 3/gp; p,0.001; t-test). WB = whole brain fraction; S1 = vessel-enriched fraction. doi:10.1371/journal.pone.0089970.g001 abundance data derived from MS-speak data were calculated using Progenesis software.
Regarding the label free quantification, the total number of Features (i.e. intensity signal at a given retention time and m/z) was reduced to MS/MS peaks with charge of 2, 3, or 4+ and we only kept the five most intense MS/MS spectra per ''Feature''. The subset of multi-charged ions (2+, 3+ and 4+) was extracted from each LC-MS run and the ions intensities summed for normalization. Protein quantification was performed as follows; for a specific protein, the associated unique peptide ions were summed to generate an abundance value. The final list of measured protein abundances were transformed using an ArcSinH function (as the method of detection can generate a significant amount of near zero measurement for which a log transform is not ideal). The within group means were calculated to determine the fold change and the transformed data was then used to calculate the p-values. Differentially expressed proteins were considered most meaningful under the following conditions: Only proteins detected by two or more peptides and p,0.01 associated with the protein change.

Gene Ontology Analysis
Proteins found to be significantly different at p,0.01 within the WT and Tg-SwDI cohorts were uploaded to WebGestalt Gene Set Analysis Toolkit version 2 (http://bioinfo.vanderbilt.edu/ webgestalt, Vanderbilt University) for enrichment analysis based on gene ontology [34], [35]. Cut-off criteria included a p,0.01 adjusted with hypergeometric statistical analysis using Benjamini & Hochberg multiple test adjustment with a minimum molecule cut-off of 2.

KEGG Pathway Analysis
For pathway analysis, the p-value threshold was set at p,0.05 to strengthen network analysis. Proteins found to be significantly different at p,0.05 within the WT and Tg-SwDI cohorts were uploaded to WebGestalt Gene Set Analysis Toolkit version 2 for enrichment analysis based on the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway maps using the same criteria above.

Immunohistochemistry
At 3 and 9 months of age, Tg-SwDI and WT mice were deeply anaesthetised with 5% isoflurane and transcardially perfused with 20 ml 0.9% heparinized phosphate buffered saline followed by 20 ml of 4% paraformaldehyde in 0.1% phosphate buffer. Brains were removed and post-fixed in 4% PFA over-night and paraffin embedded.
Coronal tissue sections from each respective cohort were deparaffinized using standard procedures. Briefly, sections were incubated at 60uC for 30 minutes, followed by further dewaxing in xylene. Slides were then placed sequentially in 100%, 90% and finally 70% alcohol before being rinsed in running water. Sections were processed for antigen retrieval by being incubated in 10 mM citric acid for 10 minutes at 100uC in a pressurized retrieval machine. For APOE immunostaining sections were blocked with normal donkey serum and incubated sequentially with goat polyclonal APOE (Calbiochem) along with mouse monoclonal b-dystroglycan (Leica Microsystems) followed by Alexa Flour488Hlabelled mouse monoclonal 6E10 (Covance) overnight at 4uC. APOE and b-dystroglycan were visualized with Cy3-labelled and Cy5-labelled secondary antibodies respectively. For HTRA1 immunostaining sections were blocked with normal goat serum and incubated sequentially with rabbit polyclonal HTRA1 (Abcam) along with mouse monoclonal b-dystroglycan (Leica Microsystems), followed by Alexa Flour488H-labelled mouse monoclonal 6E10 (Covance) overnight at 4uC. HTRA1 and bdystroglycan were visualized with Alexa Flour 546 and 646 respectively. Sections were analyzed using a laser scanning confocal microscope (Leica SP5). All images were acquired with a 406oil-immersion objective within the thalamus. Image analysis was carried out using NIH image J 1.47d software.

Statistical Analysis
Data were analysed using student's t-test or one-way ANOVA where experimental design exceeded 2 comparisons. For the LC-MS data significance was set at p,0.01. For all other analyses significance was set at p,0.05. Vessel enriched fractions were collected from 3 month-old and 9 month-old WT and 3 month-old and 9 month-old Tg-SwDI mice for proteomic analysis. (B) Over 1,000 proteins were identified, with over 70% being identified by more than two peptides. 654 proteins were quantified by at least two peptides and were statistically analysed using Progenesis software. 65 proteins were found to be up-regulated in the WT animals. 45 proteins were found to be up-and down-regulated (21 and 24 respectively) in the Tg-SwDI animals. Those proteins that were found to be significantly different within the respective cohorts were analysed further with WebGestalt web-based software. doi:10.1371/journal.pone.0089970.g002 Table 1. Proteins significantly increased in older wild-type mice (9 months) relative to young wild-type mice (3 months).

Results
Vessel Enrichment: S1 Fractions are Enriched in Vesselassociated Proteins Vascular enrichment was conducted using modifications of previously published protocols [30], [31]. To verify that there was vascular enrichment, S1 fractions from 3 month-old WT mice (n = 3) were compared to whole brain (WB) fractions and the levels of four cerebrovascular-associated proteins were measured. These proteins were selected due to their critical importance in multiple cellular compartments of cerebral vessels; a marker of vascular smooth muscle (SMA); a marker of endothelial tight junctions (occludin), an endothelial marker (PECAM) and a marker of astrocytes (GFAP). The data demonstrated that S1 fractions were enriched by 3-12 fold [Occludin (7 fold), PECAM (11 fold), SMA (12 fold), GFAP (3 fold)] (Figure 1).

Temporal Increase in Vascular Ab1-40
The levels of Ab1-40 were measured since this is the most prominent Ab species in the cerebrovasculature of Tg-SwDI animals [23], whose negative charge in its mutant form contributes to its increased pathogenicity [36], [37] and its reduced clearance due to its lower affinity to LRP [38]. As predicted, the mean levels of Ab1-40 (pg/mg of total tissue 6 S.E.M) were significantly higher in vessel fractions from the 12 and 17 month-old Tg-SwDI mice (220.866.29 and 226619.22 respectively) compared to the 3 month-old mice (10.72610.72; p,0.0001; one-way ANOVA).

Proteomic Analysis of Vessel Fractions
1024 proteins were identified within the S1 fractions from young and older WT and Tg-SwDI mice. 654 were quantified by at least 2 peptides across all groups (a list of all proteins quantified by 2 peptides can be found in Table S1). The proteomic workflow along with a summary of results is presented (Figure 2). 65 proteins were significantly different (p,0.01) between the 3 month-old WT and 9 month-old WT mice. All of these proteins were up-regulated (Table 1). 45 proteins were found to be significantly different (p,0.01) between the 3 month-old Tg-SwDI and 9 month-old Tg-SwDI mice. 24 proteins were up-regulated and 21 were down-regulated ( Table 2).
Of those proteins that were significantly different within the cohorts, only 6 proteins overlapped between the WT and Tg-SwDI cohorts. This indicates differential responses to age within the WT and Tg-SwDI cohorts (Tables 1&2).

KEGG Pathway Analysis
Proteins found to be significantly different (p,0.05) between the 3 month-old and 9 month-old WT animals (178 proteins) along with those significantly different (p,0.05) between the 3 monthold Tg-SwDI and 9 month-old Tg-SwDI animals (154 proteins) were uploaded to the WebGestalt Gene Set Analysis Toolkit version 2 for cohort-specific KEGG pathway analysis.
KEGG pathway analysis revealed similar pathways being represented within the WT and Tg-SwDI cohorts. The top biological processes for both cohorts included: metabolic pathways and several neurodegenerative related pathways: Alzheimer's disease, Parkinson's disease and Huntington disease (See figure 3). Both 9 month-old WT and 9 month-old Tg-SwDI animals showed upregulation of proteins represented by the oxidative phosphorylation category as well as the gap junction category. Closer inspection of the proteins contained within certain categories reveal a divergence in the proteins being modulated within the respective cohorts, with only about a third of the proteins being shared within the metabolic pathways and oxidative phosphorylation categories. Other categories (e.g. gap junction) show much more convergence.

Age Effects on the WT Vascular Proteome
In WT animals, age led to a marked up-regulation of proteins associated with the basement membrane based on gene ontology (cellular compartment; Figure S1), particularly nidogen-1, basement membrane-specific heparan sulfate proteoglycan core protein, laminin subunit gamma-1 precursor and collagen alpha-2(IV) chain preproprotein (Figure 4).  Table 2. Proteins significantly increased (positive fold change) and decreased (negative fold change) in older Tg-SwDI mice (9 months) relative to young Tg-SwDI mice (3 months). Age also led to the up-regulation of several white matterassociated proteins in WT animals (and to a lesser extent in Tg-SwDI animals, including myelin-associated oligodendrocyte basic protein, myelin-oligodendrocyte glycoprotein, myelin proteolipid protein, myelin associated glycoprotein, contactin-associated protein-1 precursor and myelin basic protein isoform 1 ( Figure S2).

Age Effects on the Tg-SwDI Vascular Proteome
HTRA1 and APOE were markedly increased in the 9 monthold as compared to 3 month-old Tg-SwDI animals (*p,0.0001) ( Figure 5A, B). There were no differences detected between 3 month and 9 month-old WT mice. Serine protease HTRA1 was the most up-regulated protein within the Tg cohort (fold change = 6.26) followed by APOE (fold change = 3. 86). To confirm that these alterations in APOE and HTRA1 were a robust finding and increasing with age, the levels of these proteins were measured by Western blot in separate cohorts to those in which LC-MS was conducted at ages 3, 12 and 17 months (Tg-SwDI n = 4/grp). Immunoblot analysis confirmed a age-dependent increase in HTRA1 (*p,0.05; **p,0.01) ( Figure 5C) and in APOE levels (p = 0.096) in vascular fractions ( Figure 5D). However, there was no correlation between Ab1-40 levels and HTRA1 or APOE levels in the same vessel fractions ( Figure 5E, F).
Age led to the extensive down-regulation of several proteins associated with the spliceosome (note figure 3) and RNA processing in the Tg-SwDI animals based on gene ontology (biological process; Figure S3). These include several members of the heterogeneous nuclear ribonucleoprotein family of proteins ( Figure 6).

APOE Localizes to Cerebral Vessels and Accumulated Ab
Tissue from independent cohorts of 3 month-old WT, 9 monthold WT, 3 month-old Tg-SwDI and 9 month-old Tg-SwDI mice were immunolabelled with antibodies to b-dystroglycan, Ab and APOE. APOE immunolabelling was mostly observed in 9 monthold Tg-SwDI animals ( Figure 7A-D). APOE co-localized to both accumulated Ab and cerebral vessels, particularly those vessels proximate to Ab deposits within the thalamus.

HTRA1 is Located Proximately to Ab-laden Vessels in the Thalamus
Tissue from independent cohorts of 3 month-old WT, 9 monthold WT, 3 month-old Tg-SwDI and 9 month-old Tg-SwDI mice (as above) were immunolabelled with antibodies to b-dystroglycan, Ab and HTRA1. HTRA1 immunolabelling was mostly observed in 9 month-old WT and Tg-SwDI sections and appeared to be expressed cytoplasmically (Figure8 A-D). Only in 9 month-old Tg-SwDI mice was HTRA1 found to be located proximately to Abladen vessels ( Figure 8E-J) within the thalamus and appeared extracellular.

Discussion
Brain ageing leads to alterations in the protein components of the cerebrovascular milieu, which can lead to changes in vascular integrity and decreased functioning of the NVU and the BBB [6], [7], [8]. Pathological insults such as CAA can accelerate these changes, [39] [40], [41]. Quantitative proteomics allowed an in-depth and global investigation of the protein profiles of the cerebrovascular milieu of WT and Tg-SwDI animals at distinctly different ages. We were able to identify proteins that might be contributing to these age-dependent changes by looking specifically at vascular enriched fractions. These vascular fractions were enriched in proteins associated with cerebral vessels (PECAM, occludin, SMA, GFAP, ABCB1A, nidogen, claudin 11 and AQP4, Table S1) using techniques similar to those used under physiological conditions [31] and in genetically altered models with cerebral haemorrhage propensity [42].
The number of proteins that were changed between the 3 and 9 month-old WT animals (n = 65) and the 3 and 9 month-old Tg-SwDI animals (n = 45) were substantial in number, yet the protein lists generated from the two cohorts had very little overlap, suggesting that age affected the vascular proteomic profile of the WT and Tg-SwDI animals in very different ways. We hypothesized that age would have a differential effect on the cerebrovascular proteome in the presence of microvascular Ab.
KEGG pathway analysis suggested that many of the pathways enriched by significantly modulated proteins were in fact very similar between the WT and Tg-SwDI cohorts. Further scrutiny of these revealed minimal overlap of the proteins the categorical terms represented, again confirming differential regulation of protein expression within the two cohorts. Examples of this include two of the most significantly enriched pathways for both WT and Tg-SwDI cohorts: metabolic pathways and oxidative phosphorylation indicating changes in multiple mitochondrial proteins. Only 30% of those mitochondrial proteins that were found to be differentially modulated with the cohorts overlapped.
Despite the lack of overlap, modulations of mitochondrial proteins are of great importance to age-related vascular changes. The cerebrovasculature endothelium is highly enriched in mitochondria due to the high energy demands required for neurovascular coupling [43] as well as for the multitude of ATPdependent transporters at the BBB required for bidirectional transport of molecules to the brain [6]. Mitochondrial dysfunction is associated with both brain ageing and accumulating Ab and is closely linked to increases in oxidative stress [44], [45], [46], [47]. Such modulations like those observed in the older WT and Tg-SwDI animals suggest that the age does lead to changes in the expression of certain mitochondrial proteins, especially those that make up the electron transport chain.
Basement membrane thickening due to increases in proteins such as collagen has been observed with age and AD [8], [48], [49]. Of the 65 proteins up-regulated in 9 month-old WT animals, 10% are intimately linked to the basement membrane (nidogen-1, nidogen-2, laminin subunit gamma-1 precursor, collagen alpha-2(1V) chain preproprotein, laminin subunit beta-2 precursor, basement membrane specific heparin sulphate proteoglycan core protein, agrin). At 9 months of age the mice are not aged by convention, yet there is still evidence of a temporal increase in key basement membrane proteins over a 6 month period. This increase is not seen in the Tg-SwDI mice. Instead, comparison of these protein levels across all 4 groups indicates the abundance of these same proteins is typically greater in 3 month-old Tg-SwDI animals as compared to 3 month-old WT animals ( Figure 4) and remained stable as the animals aged and Ab accumulated.
In the present study, alterations in protein signatures were studied in vascular-enriched fractions which allows for the identification of proteins in cellular compartments in close association with the vasculature. Notably white matter-associated proteins were identified and found to be altered in older WT and   . HTRA is located proximately to Ab-laden vessels in the thalamus of 9 mos Tg-SwDI mice. Tissue sections representative of independent 3 and 9 month-old WT (A&B) and 3 and 9 month-old Tg-SwDI (C&D) animals were triple-labelled with HTRA1 (red), 6E10 (green) and bdystroglycan (magenta) and analysed with a laser scanning confocal microscope (4006). Immunohistochemical analysis of the four different cohorts show HTRA1 closely associated with Ab-laden vessels [white arrow, (D)] within the thalamus of 9 month-old Tg-SwDI mice (E-J) and appeared extracellular. doi:10.1371/journal.pone.0089970.g008 TgSWDI mice as compared to the respective younger cohorts. This included the significant up-regulation of myelin-oligodendrocyte glycoprotein, myelin proteolipid protein, myelin associated glycoprotein, contactin-associated protein-1 precursor and myelin basic protein isoform 1 in older WT mice. Only myelin-associated oligodendrocyte basic protein, the third most prevalent protein in myelin [50] was found to be significantly up-regulated in both older WT and Tg-SwDI mice. Furthermore, in older Tg-SwDI mice, a down-regulation in myelin expression factor 2 isoform 1, a suppressor of MBP expression [51] was observed, which was not seen in the older WT mice. This is an indication that these white matter-associated proteins are differentially altered in older WT mice as compared to those found to be altered in the older TgSWDI mice.
The integrity of white matter has been reported to be compromised in ageing and in models of amyloidosis [52], [53], [54]. Whilst studies have previously shown alterations in myelin proteins with ageing the results are conflicting with some studies showing a reduction in specific proteins like myelin basic protein [55] and others finding the up-regulation or complete lack of change in proteins such as 29,39 cyclic nucleotide 39-phosphodiesterase or proteolipid protein respectively [56]. Furthermore the spatial organisation of proteins on myelinated axons, such as those found within paranodal regions, may be altered and can respond dynamically to various challenges including cerebral hypoperfusion and amyloid plaques [57], [58], [59]. In the present study we found a marked up-regulation of a greater number of myelinassociated proteins in older WT mice compared to TgSWDI. These differences are likely due to whn and how these changes are measured and suggest complex regulatory control of myelin protein expression. One of the most up-regulated proteins within the 9 month-old Tg-SwDI animals was APOE. APOE is known to bind to Ab and is involved in its deposition and transcystosis across the blood brain barrier via its interaction with LDL-related protein (LRP) [60], [61]. APOE expression has been manipulated in various ways within the Tg-SwDI mouse model; though very little data exists showing how the endogenous mouse APOE acts naturally within the model. Expression of human APOE in the Tg-SwDI mouse leads to a very different pattern of Ab deposition than does endogenous APOE, such that human APOE reduces microvascular accumulation of Ab [62]. Furthermore, when APOE is blocked from interacting with Ab in Tg-SwDI mice, a reduction in vascular fibrillar amyloid is observed [63]. We show that the APOE co-localizes to Ab deposits closely associated with cerebral vessels. Perivascular drainage of Ab is thought to play a critical role in the clearance of Ab [64] and APOE has been shown to co-localize to perivascular spaces as well as with astrocytic endfeet in transgenic mouse models of amyloidosis [65]. Recent work undertaken using specific antibodies to APOE suggests that APOE might compete with clearance mechanisms for Ab at the BBB, and suggests potential novel targets for therapeutic intervention to reduce amyloidosis [66], [67].
We have also found a significant increase in the HTRA1 within the vascular fractions of the Tg-SwDI mice at 9 months of age. HTRA1 is a member of the trypsin family of serine proteases that has properties similar to many heat shock proteins and typically acts to prevent the aggregation of misfolded proteins [68]. It is known to inhibit Tgfb signalling [69] and missense mutations are associated with cerebral autosomal recessive arteriopathy with subcortical infarcts and leukoencephalopathy [70]. HTRA1 cleaves multiple extracellular matrix (ECM) proteins including fibronectin, collagen and aggrecan, which in turn could lead to ECM remodelling [71]. HTRA1 has also been shown to degrade proteins aggregates such as tau [72] and has been shown previously to associate with Ab, and specifically vascular Ab [73]. Further, when HTRA1 is inhibited in astrocyte cultures that secrete Ab 1-40, there is a dose-dependent increase in the level of Ab1-40, the major Ab species observed in CAA [73]. The presence of HTRA1 in the vessels may indicate a potential alternative clearance method of Ab or its involvement in ECM remodelling around the blood vessels in reaction to accumulating Ab.
Tg-SwDI animals also showed a very significant downregulation of proteins associated with RNA binding, processing and transport [e.g. spliceosome proteins, heterogeneous nuclear ribonucleoproteins (HNRNPs), TAR DNA-binding protein-43 (TDP-43)] at 9 months of age compared to 3 months of age. The spliceosome is required for proper protein synthesis, as it is involved in the formation of functional mRNA through excision of introns and splicing of exons [74], [75]. Heterogeneous nuclear ribonucleoproteins are involved in RNA binding and processing [76], [77]. A large number of RNA-binding proteins have prionlike domains (e.g. HNRNPA1, HNRNPAB, HNRNPA2B1, TDP-43 [78]), some of these, such as TDP-43, have been implicated as major disease proteins in neurodegenerative diseases such as amyotrophic lateral sclerosis and frontotemporal lobar degeneration and as secondary features of diseases like Alzheimer's disease and Parkinson's disease [79].
Age has distinctly different effects on the proteomic profiles of cerebral vessels from WT and Tg-SwDI animals. This is likely due to the transgenic nature of the Tg-SwDI animals, which leads to the vascular accumulation of Ab1-40. These differential effects include changes in proteins from various cellular compartments and with diverse cellular functions. Further investigation of these proteins is warranted and may give greater insight to the mechanisms behind age-related cerebrovascular dysfunction and pathologies as well as provide novel therapeutic targets. Figure S1 Gene ontology analysis of proteins found to be significantly different at p,0.01 within the WT cohort were uploaded to WebGestalt Gene Set Analysis Toolkit version 2. (TIF) Figure S2 A number of white matter-associated proteins are significantly up-regulated in 9 month-old WT and Tg-SwDI animals (**p,0.01; t-test). See Table 2 for full list.