Vulnerability of white matter structure and function to chronic cerebral hypoperfusion and the effects of pharmacological modulation

Chronic cerebral hypoperfusion, a sustained modest reduction in cerebral blood flow, is associated with damage tomyelinated axons and cognitive decline with ageing. Oligodendrocytes (the myelin producing cells) and their precursor cells(OPCs) may be vulnerable to the effects of hypoperfusion and in some forms of injury OPCs have the potential to respondand repair damage by increased proliferation and differentiation. Using a mouse model of cerebral hypoperfusion we havecharacterised the acute and long term responses of oligodendrocytes and OPCs to hypoperfusion in the corpus callosum.Following 3 days of hypoperfusion, numbers of OPCs and mature oligodendrocytes were significantly decreased comparedto controls. However following 1 month of hypoperfusion, the OPC pool was restored and increased numbers ofoligodendrocytes were observed. Assessment of proliferation using PCNA showed no significant differences betweengroups at either time point but showed reduced numbers of proliferating oligodendroglia at 3 days consistent with the lossof OPCs. Cumulative BrdU labelling experiments revealed higher numbers of proliferating cells in hypoperfused animalscompared to controls and showed a proportion of these newly generated cells had differentiated into oligodendrocytes in asubset of animals. Expression of GPR17, a receptor important for the regulation of OPC differentiation following injury, wasdecreased following short term hypoperfusion. Despite changes to oligodendrocyte numbers there were no changes to themyelin sheath as revealed by ultrastructural assessment and fluoromyelin however axon-glial integrity was disrupted afterboth 3 days and 1 month hypoperfusion. Taken together, our results demonstrate the initial vulnerability ofoligodendroglial pools to modest reductions in blood flow and highlight the regenerative capacity of these cells. Citation: McQueen J, Reimer MM, Holland PR, Manso Y, McLaughlin M, et al. (2014) Restoration of Oligodendrocyte Pools in a Mouse Model of Chronic CerebralHypoperfusion. PLoS ONE 9(2): e87227. doi:10.1371/journal.pone.0087227Editor: Ken Arai, Massachusetts General Hospital/Harvard Medical School, United States of AmericaReceived July 16, 2013; Accepted December 25, 2013; Published February 3, 2014Copyright: 2014 McQueen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.Funding: The funding of the Lifelong Health and Wellbeing Initiative (supported by the BBSRC, EPSRC, ESRC and MRC), The Disconnected Mind (supported byAge UK) and Alzheimer’s Research UK (ARUK) is gratefully acknowledged. JHF is supported by a research fellowship from the Alzheimer’s Society. Confocalimaging was performed at the IMPACT Imaging facility and at the Euan MacDonald Centre at the University of Edinburgh. The funders had no role in study design,data collection and analysis, decision to publish, or preparation of the manuscript.Competing Interests: The authors have declared that no competing interests exist.* E-mail: [email protected] IntroductionOligodendrocytes are the myelin producing cells of the CNSand are critical for maintaining and regulating the myelination ofaxons. Oligodendrocyte survival and the integrity of myelinatedaxons is essential for maintaining saltatory conduction, neuronalcommunication and normal cognitive function (for review see [1]).A single oligodendrocyte can myelinate up to 50 axonal segments[2] and thus damage to individual oligodendrocytes could have amajor effect on myelination of axons and efficiency of the relay ofinformation.Oligodendroglia appear to be particularly vulnerable to bloodflow reductions and in animal models of cerebral ischemia andsevere hypoperfusion a marked loss of oligodendrocytes occursrapidly in response to severe reductions in blood flow [3–6].Additionally, in vitro models of hypoxia and oxygen-glucosedeprivation, common pathways in cerebral ischaemia, havedemonstrated the susceptibility of oligodendrocytes to theseconditions [7,8] and it is now thought that damage to oligoden-drocytes is mediated by oxidative stress, inflammation andexcitotoxicity (for review see [9]). Indeed, damage to myelinatedaxons and oligodendrocytes is prominent in various conditions inwhich cerebral blood flow is compromised such as the ageing brain[10,11], Alzheimer’s disease [12,13] and stroke [4,14] and maycontribute to a functional impairment.Despite the initial degeneration of oligodendrocytes followinginjury there is now evidence to indicate that oligodendrocyteprecursor cells (OPCs) can proliferate and differentiate and as aresult may serve to replenish the loss of damaged oligodendrocytesand potentially repair functional deficits. In neonatal models ofhypoxic-ischaemic injury, cell proliferation is increased and newoligodendrocytes are generated up to several weeks after the initialinjury [15]. In the adult brain there also appears to be anendogenous capacity to generate new oligodendrocytes in responseto cerebral ischaemia. In models of focal cerebral ischaemia wheneither blood flow is restored with reperfusion or in the peri-infarctregion where there is sufficient collateral flow, increased numbersof OPCs are detectable [6,16,17]. Interestingly, in aged humanbrain, increased numbers of oligodendrocytes and OPCs occur inareas adjacent to white matter disruption where blood flow iscompromised [18] and increased numbers of oligodendrocyteshave been demonstrated in cases of vascular cognitive impairment PLOS ONE | www.plosone.org1February 2014 | Volume 9 | Issue 2 | e87227 [19]. Together these studies suggest that OPCs may respond toreduced cerebral blood flow in an attempt to ameliorate whitematter damage.There are a number of mechanisms that may regulate OPCproliferation and differentiation. Relevant to blood flow alterationsboth glutamate and ATP, the extracellular levels of which may beincreased with reduced blood flow, have been shown to beimportant in regulating OPC proliferation and differentiation[20,21]. In addition, recent studies have identified a role for a G-coupled protein receptor, GPR17, as an important mediator ofOPC differentiation and white matter repair [22,23]. GPR17 hasbeen shown to be expressed by a subset of OPCs [23,24] and it isthought that these cells may operate as an early sensor of braindamage whereby they are activated by uracil nucleotides andcysteinyl leukotrienes which are increased in response to cerebralischemia. In support of this, GPR17 positive cells are upregulatedin response to cerebral ischemia and associated with oligodendro-cyte differentiation [23].The present study sought to determine whether the pools ofoligodendrocytes and OPCs would be influenced by modestreductions in blood flow more akin to those occurring in theageing brain. We utilised a mouse model of cerebral hypoperfu-sion induced by permanent bilateral carotid stenosis which wehave previously shown to result in diffuse white matter pathology[12,25,26]. Importantly, these mice also exhibit impaired spatialworking memory [12,27], providing a link between white matterdisruption and cognitive decline. More recently, microarrayanalysis in mice subject to cerebral hypoperfusion has revealedincreased expression of several genes involved in cell proliferation[25] which may underlie a potential white matter repairmechanism. We therefore also investigated the extent of OPCproliferation and differentiation and whether this was mediated byGPR17. In addition, ultrastructural analysis and myelin labellingwere carried out to determine whether alterations to oligodendro-cyte pools influenced myelin sheath thickness. Materials and MethodsEthics statementAll procedures were authorised under the Home Officeapproved Project Licenses, ‘Pathophysiology of Alzheimer’sdisease: link to cerebrovascular disease’ (licence number 60/3722) and ‘Pathophysiology of vascular cognitive impairment andAlzheimer’s disease’ (licence number 60/4350) held by Prof. K.Horsburgh. The licences were approved by the University ofEdinburgh’s Ethical Review Committee and the Home Office,and adhered to regulations specified in the Animals (ScientificProcedures) Act (1986). Animals and surgeryAdult male C57Bl/6J mice (aged 3–5 months old, 25–30 g)were obtained from Charles River Laboratories Inc, UK. Animalswere subject to chronic cerebral hypoperfusion as previouslydescribed [25–28]. In brief, wire microcoils (0.18 mm internaldiameter, Sawane Spring Co., Japan) were applied to bothcommon carotid arteries under isoflurane anaesthesia (induced at5%, and maintained at 1.2–1.6%). A 30 minute interval was leftbetween left and right coil application. Sham-operated animalsunderwent identical procedures with the exception that coils werenot placed around the arteries. Housing of animals and allprocedures were carried out in pathogen-free animal units.BrdU labellingAnimals from the 1 month cohort were given intraperitonealinjections (35 mg/kg body weight) of 59-bromo-29-deoxyuridine(BrdU; Fluka, UK) twice daily for the first 3 days following surgeryto label proliferating cells during this period by an individualblinded to surgical group. Laser speckle contrast imagingAn additional cohort of animals underwent measurement ofcerebral blood flow using laser speckle flowmetry. Animals wereanaesthetised with 5% isoflurane in oxygen for 1.5 minutes in ananaesthetic chamber. Animals were then transferred to astereotaxic frame and their heads were fixed into position.Anaesthesia was maintained at 2–2.5% isoflurane via a nose coneand body temperature was monitored and regulated. An incisionwas made to expose the skull and the skin overlying the skull wasreflected. The skull was moistened using saline and a small amountof water-based gel (37uC) was spread evenly onto the skull. AmoorFLPI2 Speckle Contrast Imager (Moor Instruments, UK)was positioned 20 cm above the head. Image sequences wereacquired at a resolution of 7526580 pixels and a frequency of 1frames/second (20 ms/frame). Following stabilisation of perfusionreadings, a 2 minute perfusion recording was carried out.Raw speckle contrast sequences were analysed usingmoorFLPI2 Review software (v4.0). Regions of interest wereconsistent between each mouse and made 1 mm to 22 mm fromBregma with care taken to avoid any artefacts on the skull surface.Data were measured in blood perfusion units (PU) and calculatedfor each mouse as the percentage change relative to baseline(Figure S1). Tissue preparation and immunohistochemistryAt 3 days or 28 days post-surgery, mice were deeplyanaesthetised with 5% isoflurane and transcardially perfused with20 ml 0.9% heparinised saline followed by 20 ml 4% paraformal-dehyde (PFA) in 0.1% phosphate buffer (PB, pH 7.4). Followingperfusion, the brains were removed and post-fixed in 4% PFAovernight. Brains were then transferred to PB and stored overnightat 4uC. The brains were cut along the midline and free-floating50mm sagittal sections were cut using a vibrating blade microtome(Hydrax V50, Zeiss, Germany). Sections were stored in cryopro-tective medium (30% glycerol/30% ethylene glycol in PB) at220uC until required. Different cohorts of animals were used dueto the sensitivity of some antibodies to tissue fixation (n = 13 sham,12 hypoperfused and n = 9 sham, 10 hypoperfused for 3 daystudies; n = 10 sham, 11 hypoperfused and n = 9 sham, 9hypoperfused for 1 month studies). Occasionally animals wereexcluded from analysis if there was an absence of cellular stainingor the quality was deemed too poor to perform accurate analysis.The following primary antibodies were used in this study: anti-BrdU (1:200, AB6326, Abcam, UK), anti-CC1 (APC 1:20, OP80,Calbiochem, USA), anti-GFAP (1:1000, Z0334, Dako UK), anti-GPR17 (1:200, 10136, Cayman Chemical, USA) anti-Iba1 (1:100,ab5076, Abcam UK), anti-MAG (1:100, sc-9543, Santa Cruz,USA) anti-NG2 (1:100, AB5320, Millipore, UK), anti-Olig2(1:500, Ab9610, and 1:100, MABN50, both Millipore, UK),anti-PCNA (1:500 ab29, Abcam UK), anti-PDGFRa (1:100,558774, BD Pharmingen UK), and anti-PDGFRb (1:100,AF1042, R and D Systems, UK). Cy2, Cy3, DyLight 488 andAlexa Fluor 488 and 647 (all 1:200) conjugated secondaryantibodies were purchased from Jackson ImmunoResearchLaboratories Inc (USA). Alexa Fluor 488 and 546 conjugatedsecondary antibodies (1:500) were purchased from Life Technol-ogies Ltd (UK). Double labelling experiments were carried out toOligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org2February 2014 | Volume 9 | Issue 2 | e87227 confirm the cellular specificity of antibodies used for OPC andoligodendrocyte labelling. Non-specific labelling was blocked using3% normal serum and sections were incubated in primary andsecondary antibodies overnight at 4uC. Sections were mountedonto SuperFrost slides and mounted using Vectashield hard setmounting medium containing the nuclear stain 49,6-diamidino-2-phenylindole (DAPI) (H-1500, Vector Laboratories, USA).For BrdU labelling experiments, an additional antigen retrievalstep was required. Following 3 washes in PBS, sections wereincubated for 30 minutes in 2M HCl at 37uC to allowdenaturation of DNA. Following this, sections were given three5 minute washes in 0.1 M sodium borate buffer (Na2B4O7,pH 8.5). For PCNA labelling, sections were retrieved in 10 mMcitrate buffer for 30 minutes at 85uC and blocked using 10%normal serum and 0.5% bovine serum albumin.Fluoromyelin staining was used to assess myelin integrityfollowing hypoperfusion. Free-floating sections were washed inPBS and mounted onto slides. Following rehydration in PBS,sections were incubated in Fluoromyelin Green (1:200, Invitrogen)for 1 hour at room temperature. At the outset the conditions wereoptimised as recommended by the manufacturer. This protocolwas determined to be optimal for studying myelin alterations inthick vibratome sections and is a slight modification of thatsuggested by the manufacturer for thin paraffin sections. Confocal laser scanning microscopy and image analysisImmunolabelled 50 mm sections were imaged using confocallaser scanning microscopy (Zeiss Axioskop LSM 510 or ZeissLSM710, Zeiss, Germany). All images were acquired using a 206objective (numerical aperture 0.75) representing an area of4606460mm. Images were obtained at a resolution of102461024 pixels. Z-stacks of a minimum of 7mm were acquiredwith a step size of 1mm. The region of the corpus callosum, insagittal sections, was imaged above the lateral ventricle at thestereotactic co-ordinates, lateral 2.4060.1 mm, Bregma21.560.1 mm.Stereological cell counting was performed using ImageJsoftware (version 1.42q) (National Institutes of Health, USA).Cells were identified based on expression of the immunolabel(s) ofinterest co-localised with the nuclear stain (DAPI). Cells weremanually identified and counted using the ImageJ Cell Counterplugin. To prevent over-counting, cells crossing the left and topsides of the region of interest were included, but any cells crossingthe right or bottom boundaries were not counted. Images from thetop of the Z-stack were excluded from analysis whilst counts fromthe bottom image from the stack were included. Cell counts areexpressed as the percentage of sham controls or as the number ofcells/0.01mm.To assess the intensity of GPR17 and fluoromyelin staining,sections were imaged using identical gain and offset settings on theconfocal microscope which ensured a common threshold was setin the acquisition of all images. For GPR17 staining, individualGPR17 cells within the corpus callosum were manually outlinedand the mean gray value measured and expressed as the averageper animal. To assess the intensity of fluoromyelin staining, thecorpus callosum was manually outlined and mean gray valuemeasured. All intensity measurements were carried out intriplicate and values averaged. For analysis of MAG immuno-staining, images were acquired using identical confocal settingsand background subtraction was applied before calculating thepercentage area of positive staining. All experiments andsubsequent analysis were carried out blind to surgical condition.Western blot analysisIn a separate cohort of mice after one month of hypoperfusion(n = 8) or a sham (n = 5) procedure, after decapitation, the brainwas rapidly removed, the cerebellum discarded and the remainingbrain frozen in liquid nitrogen. Myelin-enriched fractions wereprepared by sucrose density centrifugation [29] and the totalprotein concentration was determined using Pierce BCA ProteinAssay Kit (Thermo Fisher Scientific, UK). Proteins were separatedby Bis-Tris 4–12% SDS-PAGE (NuPageH NovexH, Life Technol-ogies) and transferred onto PVDF membrane (Immobilon-FL,Millipore). Immunobloting was performed using the OdysseyInfrared Imaging System (LiCor Biosciences, Lincoln, NE, USA).Membranes were blocked 1 hour at room temperature in Odysseyblocking buffer (diluted 1:1 with phosphate-buffered saline),washed in phosphate-buffered saline–Tween (phosphate-bufferedsaline with 0.1% Tween) and incubated over-night at 4uC withMBP (1:10.000, Millipore). After gentle washing, membranes werethen incubated 1 hour at room temperature with GAPDH(1:14.000, Sigma) which was used as a loading control.Membranes were then incubated for 45 minutes with theappropriate fluorescent secondaries (1:3000, LiCor Biosciences).The western blots were analysed using the LiCOR BioscienceOdyssey system and software. The four MBP isoforms wereanalysed together and normalized to GAPDH. Electron microscopyTo determine whether 3 days or 1 month of chronic cerebralhypoperfusion led to alterations in myelin sheath thickness in thecorpus callosum, transmission electron microscopy was carriedout.Animals were transcardially perfused with 5% glutaraldehyde/4% paraformaldehyde, 3 days (n = 5 sham, 6 hypoperfused) or 28days (n = 7 sham, 7 hypoperfused) after the onset of cerebralhypoperfusion. The brains were then cut into 1 mm thick sectionsand the corpus callosum manually dissected out. Corpus callosumsamples were fixed in 3% glutaraldehyde in 0.1M sodiumcacodylate buffer (pH 7.3) for 2 hours and then washed 3 timesin the same buffer. Samples were then post-fixed in 1% osmiumtetroxide in 0.1M sodium cacodylate. Following dehydration,sections were embedded in araldite resin. Samples were then cutfrom the midline into the corpus callosum and 60 nm ultrathintransverse sections were cut using a Reichert OMU4 ultramicro-tome (Leica Microsystems (UK) Ltd, Milton Keynes) and stainedin uranyl acetate and lead citrate. Sections were viewed using aPhilips CM120 transmission electron microscope (FEI UK Ltd.,Cambridge, England) and images taken at a magnification of35006using a Gatan CCD camera (Gatan UK, Oxon, England).To quantify changes in myelin sheath thickness, a lined grid of0.9560.95 mm was overlaid onto each image using Image Jsoftware (v1.42q) and fibres were selected for analysis if theirmyelin sheath was intersected by a grid line. Using the freehandtool, the perimeters of each fibre and axon were manually tracedand whole fibre area and axonal area were measured. Thesevalues were used to calculate corresponding fibre and axonaldiameters. To calculate g-ratio, axonal diameter was divided bywhole fibre diameter. For each animal a minimum of 137 fibreswere analysed with the observer blind to surgical condition. Statistical analysisData were analysed using Student’s unpaired t-test or theMann-Whitney U-test depending on parametric or non-paramet-ric distribution. Statistical analysis was performed using GraphPadPrism 5 software (GraphPad Software, San Diego, USA). AOligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org3February 2014 | Volume 9 | Issue 2 | e87227 probability (p) value #0.05 was considered to be statisticallysignificant. ResultsDecreased numbers of OPCs and matureoligodendrocytes during the early response to cerebralhypoperfusionTo assess the acute response of OPCs to chronic cerebralhypoperfusion, NG2 labelling was carried out and numbers ofNG2 OPCs were counted. We determined that NG2 specificallylabelled OPCs in the corpus callosum through colocalisation withthe OPC marker PDGFRa (Figure S2A). In addition, NG2 cellslacked PDGFRb expression, a marker of pericytes (Figure S2B).Two distinct populations of NG2 cells were identified: onepopulation showed circular immunoreactivity around the nucleusand few processes, corresponding to ‘early’ stage OPCs, whilst theother were more intensely stained and displayed more extensivecellular processes, corresponding to ‘late’ stage OPCs (Figure 1A).Cell counting of both NG2 populations revealed that during theacute response to hypoperfusion, numbers of early NG2 cellswere significantly decreased compared to sham controls(p = 0.015) (Figures 1B, S3A). Numbers of NG2 cells displayinglate OPC morphology were unchanged after 3 days (p = 0.532)(Figure 1C).The early effects of cerebral hypoperfusion on matureoligodendrocyte populations were then examined using CC1immunolabelling (Figure 1D, Figure S3B). This revealed asignificant decrease in oligodendrocyte number in the hypoper-fused group compared to sham controls (p = 0.027) (Figure 1E). Ithas been reported that astrocytes may express the CC1 antigen[30], but in this study CC1/GFAP double labelling determinedthat only 0.8% of CC1 cells were GFAP in the corpus callosum(Figure S4A). Furthermore the numbers of astrocytes wereunchanged after 1 month of cerebral hypoperfusion (FiguresS4B-C). Together these results show that rapid alterations in OPCand oligodendrocyte populations occur in response to modestreductions in cerebral blood flow. Restoration of the precursor pool and increased numbersof oligodendrocytes after long term cerebralhypoperfusionThe longer term responses of the oligodendrocyte pools tocerebral hypoperfusion were next investigated. NG2 labelled cellswere counted and showed no differences in numbers of early(p = 0.245) or late (p = 0.860) NG2 cells between sham andhypoperfused groups (Figure 2A and 2B respectively). In contrastto the loss of mature oligodendrocytes observed after 3 days ofhypoperfusion, CC1 labelling (Figures 2C, S3B) revealed asignificant increase (19%) in the number of mature oligodendro-cytes in the hypoperfused animals compared to sham operatedanimals (p = 0.007) (Figure 2D). Together these results indicatethat in response to longer term cerebral hypoperfusion areplacement mechanism is acting to restore OPCs and in thecase of mature oligodendrocytes, increase numbers of cells. Proliferation and differentiation of OPCs in response tocerebral hypoperfusionTo characterise levels of proliferation in response to hypoper-fusion, PCNA labelling was carried out to determine numbers ofproliferating cells at 3 days and 1 month of hypoperfusion(Figure 3A). This showed that overall numbers of proliferating cellswere not different between groups at either 3 days (p = 0.20)(Figure 3B) or 1 month (p = 0.564) (Figure 3C). We next sought todetermine the extent of OPC proliferation and whether thiscontributed to the restoration of oligodendroglial pools followinghypoperfusion. For technical reasons PCNA/NG2 double label-ling could not be carried out therefore PCNA/Olig2 labellingexperiments were carried out in 3 day and 1 month cohorts(Figure 3D). Whilst Olig2 is expressed throughout the oligoden-drocyte lineage, only OPCs can proliferate and thus expressPCNA. The number ofPCNA/Olig2 cells were significantlydecreased in 3 day hypoperfused animals compared to shamcontrols (p = 0.02) (Figure 3E) while no differences in proliferatingOPCs were observed after 1 month of hypoperfusion (p = 0.323)(Figure 3F). Together these results show that whilst overall levels ofproliferation were unchanged after 3 days of hypoperfusion,numbers of proliferating OPCs were decreased in hypoperfusedanimals compared to controls.To further investigate the early proliferative responses tocerebral hypoperfusion, animals from the 1 month cohort receivedinjections of BrdU for the first three days following surgery to labelall proliferating cells within this period (Figure 3G). BrdU cellswere present in 2 of the 9 (22%) of the sham control group and in5 out of 10 (50%) of the hypoperfused group (Figure 3H) althoughthe difference between groups was not statistically significant(p = 0.157). BrdU and NG2 double labelling failed to demonstrateany proliferating OPCs (data not shown).We next sought to determine whether BrdU cells generatedearly in response to hypoperfusion had differentiated into matureoligodendrocytes and carried out BrdU/CC1 labelling(Figure 3I). This showed thatBrdU/CC1 cells were present in3 of the 10 (30%) 1 month hypoperfused mice but were completelyabsent in sham-operated controls (Figure 3J). Although thedifference was not statistically significant (p = 0.095), the presenceof these double labelled cells in a proportion of hypoperfusedanimals indicates that OPC differentiation has occurred in a subsetof animals in response to reduced cerebral blood flow. CBFresponses in individual mice may vary which may account fordifferences in the proliferative/differentiation responses. Anindication of the individual CBF responses was investigated indifferent cohorts of mice to those used to assess pathology at 3 daysand 1 month hypoperfusion (see Figure S1). Decreased expression of GPR17 in response to cerebralhypoperfusionTo elucidate a possible mechanism involved in OPC differen-tiation in response to cerebral hypoperfusion, we analysed theexpression of the G protein-coupled receptor GPR17 (Figures 4A,S3D). Previous studies have shown that in the intact brain GPR17is expressed by premyelinating oligodendrocytes and a subset ofOPCs, and receptor activation has been demonstrated to play apermissive role in OPC differentiation [22,23]. GPR17/NG2double labelling could not be performed due to the antibodiesbeing raised in the same species therefore we confirmed GPR17expression by oligodendrocyte lineage cells using Olig2. Thisshowed approximately 64% of GPR17 cells co-expressed Olig2(Figure S5), which is in agreement with a previous report [23].Numbers of GPR17 cells were not significantly different betweengroups at either time point examined (Figures 4B and 4C). Wenext assessed expression of the receptor, as indicated by theintensity of staining. This revealed a decrease in the GPR17labelling intensity after 3 days in hypoperfused as compared toshams (p = 0.007) (Figure 4D). However, after 1 month ofhypoperfusion there was no difference between groups(p = 0.363) (Figure 4E).Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org4February 2014 | Volume 9 | Issue 2 | e87227 Impairment of axon-glial integrity, increased microgliaand absence of gross myelin alterations in response tocerebral hypoperfusionTo determine whether alterations in oligodendrocyte numbersmay be paralleled by alterations in myelin density followingcerebral hypoperfusion, myelin status was assessed using thefluorescent lipophilic dye fluoromyelin (Figure 5A). However,there were no significant differences in staining intensity followingeither 3 days of cerebral hypoperfusion (p = 0.598) or after 1month of cerebral hypoperfusion (p = 0.063) (Figure 5B & C).Investigation of myelin-enriched extracts using Western blotanalysis additionally indicated that there was no significantdifference in MBP levels between sham and one monthhypoperfused mice (Figure S6).To further investigate the myelin integrity, electron microscopywas carried out and measurements of myelin sheath thicknessrelative to fibre diameter, i.e. g-ratio, were conducted in separatecohorts of 3 day and 1 month hypoperfused animals (Figure 5D).This revealed no significant difference in g-ratio values followingboth short and long term hypoperfusion compared to respectivesham-operated controls (Figure 5E). As a note since the 3 days and1 month cohorts underwent surgery and tissue processing atdifferent times no statistical comparisons can be made betweensham cohorts or the hypoperfused cohorts.Despite an absence of gross myelin alterations, in our group wehave consistently determined that cerebral hypoperfusion results ina disruption of axon-glial integrity [12,25,26] and thus sought toverify whether similar alterations occur in the white matter in thisstudy. Axon-glial integrity was assessed by myelin associatedglycoprotein (MAG), a key myelin protein involved in themaintenance of axon-glial integrity (Figure 6A). There was areduction in the density of MAG staining in the corpus callosumafter 3 days (p = 0.008) and 1 month (p = 0.027) of chroniccerebral hypoperfusion (Figures 6B and 6C).Figure 1. Reduced numbers of OPCs and mature oligodendrocytes after 3 days of chronic cerebral hypoperfusion. (A) Representativeconfocal images showing morphology of early and late NG2 OPCs in the corpus callosum. (B) A significant decrease in numbers of early NG2 OPCswas found in the corpus callosum after 3 days of chronic cerebral hypoperfusion. (C) No significant differences in numbers of late NG2 cells wereobserved after 3 days. (D) Representative confocal images of CC1 labelling of oligodendrocyte cell bodies in the corpus callosum. (E) A significantdecrease in CC1 oligodendrocytes was found after 3 days of hypoperfusion. n = 13 sham, 12 hypoperfused for NG2 labelling experiments; n = 9sham, 10 hypoperfused for CC1 labelling. Scale bars = 10 mm * p,0.05.doi:10.1371/journal.pone.0087227.g001Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org5February 2014 | Volume 9 | Issue 2 | e87227 A pronounced microglial response as determined by Iba1immunoreactivity has also been a robust finding in our model ofhypoperfusion (Figure 6D). In support of this after 1 month ofhypoperfusion, numbers of microglia were significantly increasedcompared to sham controls (p = 0.0011) (Figure 6E), whilstnumbers of microglia were unchanged after 3 days of cerebralhypoperfusion (p = 0.425) (Figure 6F), consistent with earlierstudies using this mouse model [12,28].Taken together these data indicate that whilst gross myelinmorphology remains intact and increased numbers of oligoden-drocytes are observed following hypoperfusion, axon-glial integrityis impaired supporting our previous observations [25]. DiscussionPrevious studies have demonstrated the susceptibility ofoligodendrocytes to severe reductions in cerebral blood flow(.70% to that of baseline levels) with profound oligodendrocyteloss occurring early in response to the insult [6,17]. The presentstudy additionally highlights the vulnerability of these oligoden-droglial cells to more modest reductions in blood flow comparableto those observed in the ageing brain.In the present study we investigated the pools of OPCs andmature oligodendrocytes. We observed two different populationsof NG2 cells identified by differences in morphology. Onepopulation of NG2 cells showed circular reactivity around thenucleus and few processes, corresponding to ‘early’ stage OPCs,whilst the other population of cells was more intensely stained andmore processed, corresponding to ‘late’ stage OPCs. Interestinglyonly the ‘early’ stage OPCs were affected by hypoperfusion. It ispossible that these early OPCs have responded rapidly tohypoperfusion by extending more processes and have thus beenclassified as late OPCs. Mature oligodendrocytes were alsoreduced in response to hypoperfusion, highlighting the vulnera-bility of these glial cells to even modest reductions in cerebralblood flow. Although the mechanisms underlying this earlyoligodendroglial loss with hypoperfusion remain to be determined,these may involve damage to the oligodendrocytes and OPCs as aresult of oxidative stress [31,32] or inflammation [33,34], both ofwhich are known inducers of oligodendroglial damage and/ordeath (for review see [9,35]). In the present study, consistent withour previous work, we demonstrated a marked microglial responsewith hypoperfusion. Previously we have also shown alterations inindices of hypoxia [25] in white matter after hypoperfusion.Localised alterations in glutamate levels as a result of compromisedblood flow could also contribute to the loss of oligodendrocytesand OPCs via NMDA receptor activation leading to intracellularCa-dependent injury to oligodendroglia [36,37]. However itshould be noted that whilst we propose that the loss of CC1+ andNG2+ labelling of cells is an indicator of cell loss, a loss of cellularantigenicity could also account for the reduction in cellularstaining.With sustained hypoperfusion, marked alterations in oligoden-droglial pools were observed. The OPC pool was restored andFigure 2. Restoration of the NG2 precursor pool and increased numbers of mature oligodendrocytes after 1 month of chroniccerebral hypoperfusion. (A) No significant differences in numbers of early or (B) late NG2 cells were observed after 1 month of cerebralhypoperfusion. (C) Representative confocal images of CC1 labelling of oligodendrocyte cell bodies in the corpus callosum. Scale bar = 10 mm. (D) Asignificant increase in CC1 oligodendrocytes in the corpus callosum was observed following 1 month of chronic cerebral hypoperfusion. n = 10sham, 11 hypoperfused for NG2 and CC1 labelling. ** p,0.01.doi:10.1371/journal.pone.0087227.g002Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org6February 2014 | Volume 9 | Issue 2 | e87227 numbers of mature oligodendrocytes were increased whenexamined after 1 month hypoperfusion. This suggests that thereis sufficient capacity with the adult brain to overcome the initialloss of oligodendrocyte pools. Similarly, in models of focal cerebralischaemia when either blood flow is restored with reperfusion or inthe peri-infarct region where there is sufficient collateral flow,increased numbers of OPCs are detectable [16]. In contrast, otherstudies of mouse cerebral hypoperfusion have shown thatoligodendrocyte numbers remain reduced at one month ofhypoperfusion [38,39]. There are notable differences betweenthe model of cerebral hypoperfusion in our group compared toothers [28,38,39]. Importantly we do not detect demyelination butinstead a robust disruption of axon-glial integrity and apronounced microglial response in white matter [12,25,26]. Theremay also be differences in the time course of progression ofoligodendrocyte changes between our studies and others. The levelof reduction in cerebral blood flow may be a critical factor whichinfluences the extent of pathology and proliferation/differentiationresponse. We used laser speckle imaging and demonstrated that inour hands the reduction in CBF was approximately 36% that ofbaseline at 3 days and then restored to 22% that of shams at 1month. These are slightly greater than the levels reportedpreviously by Shibata et al. 2004 [28] where the maximal levelsof CBF reduction as assessed by laser Doppler flowmetry were30% although are consistent with more recent studies such as thatby Duan et al. [40] who have reported reductions of 37% fromthat of baseline using Laser Speckle imaging. However, there are anumber of other factors that may influence the outcome anddifferences in pathology in models between different laboratoriesincluding the anaesthetic used; the background strain of mice(influences differences in cerebrovasculature) and environment(pathogen status, temperature). Notably however, in these studiesthat report sustained reductions in oligodendrocyte numbers[38,39,41,42] the pools can be restored by either pharmacologicalintervention [39,41,42] or bone marrow cell treatment [38]indicating that there is restorative capacity of oligodendrocytes inthe model. In a previous study we detected a marked increase ingenes associated with cell proliferation in white matter in responseto hypoperfusion [25] and as a consequence expected to observemarked increases in cell proliferation. However, assessment usingthe acute proliferation marker PCNA showed that overall levels ofproliferation were unchanged with hypoperfusion but revealedthat proliferation of Olig2 cells was decreased in 3 dayhypoperfused animals compared to controls. It is important tonote that whilst proliferation of OPCs was reduced after 3 days,there may be a proliferative response of other cells within the whitematter although characterisation of this was beyond the scope ofthe current study.Figure 3. Low numbers of proliferating cells are observed in response to cerebral hypoperfusion however a proportion of newlygenerated cells differentiate into mature oligodendrocytes within 1 month of cerebral hypoperfusion. (A) Representative confocalimages showing PCNA labelling of proliferating cells in the corpus callosum. (B) No significant differences in numbers of PCNA cells were observedafter 3 days or (C) 1 month of cerebral hypoperfusion. (D) Confocal images showing Olig2/PCNA labelling of proliferating oligodendroglia in thecorpus callosum. (E) Decreased numbers of Olig2/PCNA cells were observed after 3 days of cerebral hypoperfusion. (F) No significant differences innumbers of Olig2/PCNA cells were found after 1 month of hypoperfusion. (G) Representative confocal images showing BrdU labelled cells in thecorpus callosum. (H) BrdU cells were present in 2 out of 9 (22%) sham control animals but were observed in 5 out of 10 (50%) hypoperfused animals.(I) Representative confocal images showing BrdU/CC1 cells in the corpus callosum. (J) CC1/BrdU double labelled cells were present in 3 out of 10(30%) of the hypoperfused cohort but were completely absent from the sham control group. n = 13 sham, 11 hypoperfused for PCNA and Olig2labelling at 3 days. One animal was excluded from analysis on the basis of poor cellular staining; n = 9 sham, 9 hypoperfused for PCNA/Olig2 and CC1/BrdU labelling at 1 month. Scale bars = 10 mm. * p,0.05.doi:10.1371/journal.pone.0087227.g003Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org7February 2014 | Volume 9 | Issue 2 | e87227 To further characterise the early proliferative responses tocerebral hypoperfusion, BrdU incorporation was used to assess thetotal number of proliferating cells during the first 3 days after theonset of hypoperfusion. This revealed that proliferating cells weredetectable in 50% of the hypoperfused cohort compared to 22% ofsham operated animals suggesting a modest proliferative responseearly in response to hypoperfusion. However, this low level ofproliferation would be insufficient to account for the restoration ofthe oligodendroglial pool. It is possible that the BrdU labellingprotocol used in this study has not adequately labelled allproliferating cells and thus has underestimated the extent ofproliferation. Although it has been reported that BrdU at 200 mg/kg body weight is a saturating dose and does not result incytotoxicity [43], the dosage used in this study (70 mg/kg bodyweight/day) is within the standard range of 50–100 mg/kgreported in many studies. However, the use of only two injectionsper day coupled with the short bioavailability of BrdU may havebeen insufficient to label all proliferating cells within the injectionperiod. In addition, a comprehensive characterisation of cellproliferation would require continuous administration of BrdU (forexample in drinking water) and assessment of BrdU cells atvarious time points.The lack of convincing evidence of proliferation in response tocerebral hypoperfusion raises the question as to whether differen-Figure 4. Chronic cerebral hypoperfusion does not alter numbers of GPR17-expressing cells. (A) Representative confocal imagesshowing GPR17 labelling in the corpus callosum following 3 days and 1 month of hypoperfusion. Scale bar = 10 mm. (B) Numbers of GPR17 cells inthe corpus callosum were unchanged after 3 days and (C) 1 month of cerebral hypoperfusion. (D) Intensity of GPR17 labelling was significantlydecreased after 3 days of cerebral hypoperfusion. (E) No difference in GPR17 labelling intensity was observed after 1 month of cerebralhypoperfusion. n = 13 sham and 11 hypoperfused for 3 day analysis; n = 9 sham and 9 hypoperfused for 1 month analysis. ** p,0.01.doi:10.1371/journal.pone.0087227.g004Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org8February 2014 | Volume 9 | Issue 2 | e87227 tiation of pre-existing OPCs may contribute to the restoration ofoligodendrocyte pools. Consistent with the hypothesis thatglutamate and ATP may be involved in white matter disruptionfollowing cerebral hypoperfusion, there is growing evidenceimplicating these two neurotransmitters in the regulation ofOPC proliferation and differentiation. In vitro studies havedemonstrated that glutamate inhibits OPC proliferation [21] butpromotes OPC differentiation via NMDA receptor activation [44].Similarly it has been demonstrated that ATP and relatedderivatives also inhibit proliferation whilst promoting OPCdifferentiation [20]. Therefore it is possible that the low numbersof proliferating cells in this study may be due to extracellular ATPand glutamate acting to limit OPC proliferation whilst promotingdifferentiation. In the current study we have demonstrated that asmall proportion of newly generated cells had differentiated intomature oligodendrocytes however this is unlikely to exclusivelyaccount for the significant increase in oligodendrocyte numbersobserved after 1 month. It is possible that differentiation of pre-existing OPCs has also occurred to boost oligodendrocytenumbers in response to cerebral hypoperfusion. Interestingly, thisis consistent with a study in a model of cerebral ischaemia whichsimilarly showed repopulation of the oligodendrocyte pool in theabsence of significant OPC proliferation [6].To investigate a potential mechanism involved in differentiationof OPCs in response to hypoperfusion we examined the expressionof GPR17, a novel dual uracil nucleotide and cysteinyl-leukotrieneG protein-coupled receptor which has been implicated inmediating OPC responses to injury such as ischaemia [22]. Ithas previously been demonstrated that expression of GPR17 isupregulated in response to ischaemia and it has been suggestedthat receptor activation may act as a sensor of local damage [22–24]. In the current study, we have shown that numbers of GPR17-expressing cells were unchanged but expression of the receptor wasdecreased during the acute response to cerebral hypoperfusion.There is however conflicting evidence regarding the effects ofreceptor activation on OPC differentiation. Studies using trans-genic mice have shown that receptor over-expression results indecreased myelination whereas knockout leads to increasedmyelination suggesting that GPR17 is a negative regulator ofOPC differentiation [45]. In contrast, it has been shown in vitrothat receptor agonism results in increased numbers of matureoligodendrocytes, implying that receptor activation promotesdifferentiation, whilst antagonism increases the proportion ofOPCs [24]. Additionally it has also been demonstrated that whencultured OPCs are transferred to medium which promotesdifferentiation, GPR17 expression is increased [46]. Despite thisapparently permissive role of GPR17 in OPC differentiation, it hasalso been reported that receptor expression may also increase thesusceptibility of a cell to ATP-induced death [46]. Thus althoughGPR17 activation does not appear to mediate OPC responses tocerebral hypoperfusion at the time points examined, its downreg-ulation after 3 days may represent an attempt to limit cell damageand/or death induced by hypoperfusion.Figure 5. Alterations to numbers of oligodendrocytes as a result of chronic cerebral hypoperfusion does not impact on myelinationof axons within the corpus callosum. (A) Representative confocal images showing fluoromyelin labelling in the corpus callosum after 3 days and1 month of cerebral hypoperfusion. Scale bar = 20 mm (B) No significant difference in fluoromyelin intensity was observed after 3 days of chroniccerebral hypoperfusion. (C) Fluoromyelin intensity was not significantly different between groups after 1 month of cerebral hypoperfusion. n = 13sham, 12 hypoperfused for 3 day analysis and n = 9 sham, 10 hypoperfused for 1 month analysis. (D) Representative electron micrographs showingmyelinated fibres in the corpus callosum after 3 days and 1 month of cerebral hypoperfusion. Scale bars = 1 mm. (E) G-ratio values were unchangedfollowing 3 days and 1 month of hypoperfusion compared to respective sham controls. n = 5 sham and 6 hypoperfused for 3 day analysis; n = 7 shamand 7 hypoperfused for 1 month analysis.doi:10.1371/journal.pone.0087227.g005Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org9February 2014 | Volume 9 | Issue 2 | e87227 Myelinated axons in the corpus callosum were examined at theultrastructural level to determine whether alterations to oligoden-drocyte numbers had an impact on myelin sheaths. We observedno significant differences in g-ratio, an index of myelin sheaththickness, between sham and 3 day hypoperfused mice despitehypoperfusion inducing around a 15% reduction in matureoligodendrocytes at this time point. Furthermore there were nodifferences in myelin levels as assessed by fluoromyelin-labelling ofmyelin in the corpus callosum or MBP levels measured in myelin-enriched extracts. Limitations of these approaches may haveprecluded detection of myelin alterations after hypoperfusion. Inthe present study, fluoromyelin staining was conducted in thicksections using a slightly modified protocol to that of previouslypublished work which did detect alterations in myelin in thehypoperfusion model [38,41]. Thus, differences in fluoromyelinstaining methods might explain the different outcomes. Addition-ally, the present study assessed MBP alterations in myelin-enrichedfractions [29] prepared from both white matter and grey matter.Arguably there may have been a dilutional effect of the greymatter on the measurement of MBP levels since this model hasbeen shown to exhibit predominantly a white matter pathologywithout robust grey matter pathology within 1 month after thesurgery [12,27,28]. However, this finding (loss of oligodendrocytesand absence of myelin changes), is consistent with a recent studywhich used diphtheria toxin to specifically ablate oligodendrocytenumbers by approximately 26% and reported that myelin wasFigure 6. Axon-glial integrity is disrupted and increased numbers of microglia after cerebral hypoperfusion. (A) Representativeconfocal images showing MAG labelling in the corpus callosum following 3 days and 1 month of chronic cerebral hypoperfusion. Scale bar = 20 mm.(B) A significant decrease in the density of MAG labelling was observed following 3 days of cerebral hypoperfusion compared to sham controls. (C)Similarly, decreased density of MAG was also observed following 1 month of hypoperfusion. n = 12 sham, 12 hypoperfused for 3 day analysis. Oneanimal was excluded due to a lack of positive MAG labelling; n = 9 sham, 8 hypoperfused for 1 month analysis. (D) Representative confocal imagesshowing 1ba1 labelling 3 days and 1 month after the onset of cerebral hypoperfusion. Scale bar = 10 mm. (E) Numbers of microglia were unchangedfollowing 3 days of cerebral hypoperfusion. (F) Following 1 month of cerebral hypoperfusion, a significant increase in microglial number wasobserved compared to sham controls. n = 13 sham, 12 hypoperfused for 3 day analysis; n = 9 sham, 10 hypoperfused for 1 month analysis.doi:10.1371/journal.pone.0087227.g006Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org10February 2014 | Volume 9 | Issue 2 | e87227 preserved despite this extensive oligodendrocyte loss [47]. Thislack of effect on myelination despite profound alterations inoligodendrocyte pools is also consistent with previous studies byour group which have determined that there are no grossalterations to the protein levels of myelin basic protein as assessedby immunohistochemistry [25]. Instead, in this model, hypoper-fusion results in disruption of MAG and breakdown of axon-glialintegrity associated with disruption of the paranodal septatejunctions [25]. The current study used an antibody against theCC1 protein which labels oligodendrocyte cell bodies but notprocesses, and as a result cannot reliably distinguish betweenmyelinating and non-myelinating oligodendrocytes. One possibleexplanation for our findings is that in response to cerebralhypoperfusion, a non-myelinating population of CC1 oligoden-drocytes are lost and thus no disruption to myelin has beenobserved. Another possibility is that neighbouring oligodendro-cytes may compensate for oligodendrocyte loss by extendingprocesses and ‘filling in’ non-myelinated areas [48]. We alsoobserved that g-ratio was similarly unchanged following 1 monthof hypoperfusion and so it remains to be determined whetherincreased numbers of oligodendrocytes observed at this time pointare surplus to requirements. Overall these findings suggest thateven in the presence of cerebral hypoperfusion, the CNS cantolerate significant alterations to mature oligodendrocyte poolswithout any apparent detriment to myelin thickness or density.Thus in conclusion this study provides further support thatoligodendrocytes are vulnerable to modest blood flow reductionsand evidence supporting their regenerative capacity. This turnoverof the pool of oligodendrocytes appears to occur in the absence ofchanges in myelination but it would be interesting to investigate infuture studies whether there are any functional consequences ofthese changes within the white matter. Supporting InformationFigure S1 Reduced cerebral blood flow in hypoperfusedanimals. Cerebral blood flow was measured using laser speckleflowmetry prior to surgery (baseline) and at 3 days and 1 monthfollowing surgery to assess the extent of hypoperfusion at the timeswhite matter alterations were investigated. (A) Representativeimages showing speckle images at baseline, 3 days and 1 month fora sham and hypoperfused mouse. Images show the average of 100frames. Baseline CBF was not significantly different (p,0.05)between groups (1130636 perfusion units in shams vs 1072624perfusion units in hypoperfused group). (B&C;) After 3 days and 1month, CBF was significantly decreased in hypoperfused com-pared to sham animals to (p,0.001 and p,0.005 respectively).Data are calculated for each mouse as the percentage changerelative to baseline. (B) Cerebral blood flow was decreased byapproximately 36% to that of shams following 3 days of cerebralhypoperfusion. (C) Following 1 month of hypoperfusion, CBFvalues had recovered to approximately 22% of sham levels. Data isshown for each mouse, n = 8 sham, 6 hypoperfused.(TIF)Figure S2 NG2 is a specific marker of OPCs. NG2 andPDGFRa labelling confirmed the specificity of NG2 as an OPCmarker in the corpus callosum. (A) NG2 and PDGFRa co-labelledOPCs in the corpus callosum. Scale bar = 10mm. (B) Confocalimages showing representative NG2 and PDGFRb labelling. NoPDGFRb labelling of pericytes was observed in the corpuscallosum and only occasional PDGFRb labelling was observed inthe cortex. NoNG2/PDGFRb cells were observed in eitherregion. Scale bar = 20mm.(TIF)Figure S3 Low magnification images of NG2, CC1,Olig2 and GPR17 labelling. (A) Low magnification confocalimages showing NG2 labelling of OPCs in the corpus callosum(CC). (B) Confocal images showing CC1 labelling of matureoligodendrocytes in the corpus callosum. (C) Confocal imagesshowing Olig2 labelling of oligodendroglia in the corpuscallosum. (D) Confocal images showing GPR17 labelling in thecorpus callosum. Scale bars = 50mm. CTx = Cortex, CC = corpuscallosum, CPu = caudate putamen.(TIF)Figure S4 Astrocytes are not labelled with CC1 andnumbers are unchanged following 1 month of cerebralhypoperfusion. It has been reported that a subpopulation ofastrocytes can express the CC1 (APC) antigen, therefore CC1 andGFAP double labelling was carried out to determine numbers ofCC1 astrocytes. (A) Representative confocal images showingCC1/GFAP double labelling in the corpus callosum (CC). Scalebar = 50mm, inset scale bar = 10 mm. Cell counts of numbers ofdouble labelled cells revealed approximately 0.8% of CC1 cellsexpressed GFAP thus confirming the high specificity of CC1 as amarker of mature oligodendrocytes. (B) Representative confocalimages showing GFAP labelling of astrocytes in the corpuscallosum (CC). Scale bar = 50mm. (C) Numbers of GFAPastrocytes are unchanged following 1 month of chronic cerebralhypoperfusion. CTx = Cortex, CC = corpus callosum, CPu = cau-date putamen.(TIF)Figure S5 GPR17 is expressed by oligodendroglia and asmall number of microglia. (A) Representative confocalimages showingGPR17/Olig2 double labelling in the corpuscallosum. Arrows indicate double labelled cells. (B) Representativeconfocal images showingGPR17/Iba1 labelling in the corpuscallosum. (C) Cell counting revealed approximately 64.463.33%of GPR17 cells express Olig2 in the corpus callosum. Only8.2960.75% of GPR17 cells co-expressed Iba1. Scalebars = 50mm, inset scale bars = 10 mm. CTx = Cortex,CC = corpus callosum, CPu = caudate putamen.(TIF)Figure S6 MBP levels are unchanged following 1 monthof cerebral hypoperfusion. (A) Representative Western blotfrom myelin enriched extracts showing the different MBP isoforms(25–16 kDa) and GAPDH, the later used as a loading control. (B)Analysis of fluorescent intensity relative to GAPDH showed nosignificant changes in MBP levels after 1 month of cerebralhypoperfusion. n = 5 sham, n = 8 hypoperfused.(TIF) AcknowledgmentsWe would like to thank the Euan MacDonald Centre for the use ofconfocal facilities, and Julia Edgar and Maj-lis McCulloch for their inputand advice regarding electron microscopy studies. Author ContributionsConceived and designed the experiments: KH JM. Performed theexperiments: JM MR JF PH YM MM. Analyzed the data: JM JF KHYM. Wrote the paper: JM KH.Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org11February 2014 | Volume 9 | Issue 2 | e87227 References1. Nave KA (2010) Myelination and support of axonal integrity by glia. Nature468: 244–252.2. Rivers LE, Young KM, Rizzi M, Jamen F, Psachoulia K, et al. (2008)PDGFRA/NG2 glia generate myelinating oligodendrocytes and piriformprojection neurons in adult mice. Nature Neuroscience 11: 1392–1401.3. Tomimoto H, Ihara M, Wakita H, Ohtani R, Lin JX, et al. (2003) Chroniccerebral hypoperfusion induces white matter lesions and loss of oligodendrogliawith DNA fragmentation in the rat. Acta Neuropathology 106: 527–534.4. Valeriani V, Dewar D, McCulloch J (2000) Quantitative assessment of ischemicpathology in axons, oligodendrocytes, and neurons: attenuation of damage aftertransient ischemia. Journal of Cerebral Blood Flow and Metabolism 20: 765–771.5. Pantoni L, Garcia JH, Gutierrez JA (1996) Cerebral white matter is highlyvulnerable to ischemia. Stroke 27: 1641–1646; discussion 1647.6. McIver SR, Muccigrosso M, Gonzales ER, Lee JM, Roberts MS, et al. (2010)Oligodendrocyte degeneration and recovery after focal cerebral ischemia.Neuroscience 169: 1364–1375.7. Lyons SA, Kettenmann H (1998) Oligodendrocytes and microglia are selectivelyvulnerable to combined hypoxia and hypoglycemia injury in vitro. Journal ofCerebral Blood Flow and Metabolism 18: 521–530.8. McDonald JW, Levine JM, Qu Y (1998) Multiple classes of the oligodendrocytelineage are highly vulnerable to excitotoxicity. Neuroreport 9: 2757–2762.9. Merrill JE, Scolding NJ (1999) Mechanisms of damage to myelin andoligodendrocytes and their relevance to disease. Neuropathology and AppliedNeurobiology 25: 435–458.10. Head D, Buckner RL, Shimony JS, Williams LE, Akbudak E, et al. (2004)Differential vulnerability of anterior white matter in nondemented aging withminimal acceleration in dementia of the Alzheimer type: evidence from diffusiontensor imaging. Cerebral Cortex 14: 410–423.11. Kohama SG, Rosene DL, Sherman LS (2011) Age-related changes in humanand non-human primate white matter: from myelination disturbances tocognitive decline. Age DOI 10.1007/s11357-011-9357-7.12. Coltman R, Spain A, Tsenkina Y, Fowler JH, Smith J, et al. (2011) Selectivewhite matter pathology induces a specific impairment in spatial workingmemory. Neurobiology of Aging 32: 2324 e2327–2312.13. Desai MK, Mastrangelo MA, Ryan DA, Sudol KL, Narrow WC, et al. (2010)Early oligodendrocyte/myelin pathology in Alzheimer’s disease mice constitutesa novel therapeutic target. American Journal of Pathology 177: 1422–1435.14. Aboul-Enein F, Rauschka H, Kornek B, Stadelmann C, Stefferl A, et al. (2003)Preferential loss of myelin-associated glycoprotein reflects hypoxia-like whitematter damage in stroke and inflammatory brain diseases. Journal ofNeuropathology and Experimental Neurology 62: 25–33.15. Zaidi AU, Bessert DA, Ong JE, Xu H, Barks JD, et al. (2004) Newoligodendrocytes are generated after neonatal hypoxic-ischemic brain injury inrodents. Glia 46: 380–390.16. Tanaka K, Nogawa S, Ito D, Suzuki S, Dembo T, et al. (2001) Activation ofNG2-positive oligodendrocyte progenitor cells during post-ischemic reperfusionin the rat brain. Neuroreport 12: 2169–2174.17. Tanaka K, Nogawa S, Suzuki S, Dembo T, Kosakai A (2003) Upregulation ofoligodendrocyte progenitor cells associated with restoration of matureoligodendrocytes and myelination in peri-infarct area in the rat brain. BrainResearch 989: 172–179.18. Simpson JE, Fernando MS, Clark L, Ince PG, Matthews F, et al. (2007) Whitematter lesions in an unselected cohort of the elderly: astrocytic, microglial andoligodendrocyte precursor cell responses. Neuropathology and Applied Neuro-biology 33: 410–419.19. Back SA, Kroenke CD, Sherman LS, Lawrence G, Gong X, et al. (2011) Whitematter lesions defined by diffusion tensor imaging in older adults. Annals ofNeurology 70: 465–476.20. Agresti C, Meomartini ME, Amadio S, Ambrosini E, Volonte C, et al. (2005)ATP regulates oligodendrocyte progenitor migration, proliferation, anddifferentiation: involvement of metabotropic P2 receptors. Brain ResearchReviews 48: 157–165.21. Yuan X, Eisen AM, McBain CJ, Gallo V (1998) A role for glutamate and itsreceptors in the regulation of oligodendrocyte development in cerebellar tissueslices. Development 125: 2901–2914.22. Ciana P, Fumagalli M, Trincavelli ML, Verderio C, Rosa P, et al. (2006) Theorphan receptor GPR17 identified as a new dual uracil nucleotides/cysteinyl-leukotrienes receptor. The EMBO Journal 25: 4615–4627.23. Lecca D, Trincavelli ML, Gelosa P, Sironi L, Ciana P, et al. (2008) The recentlyidentified P2Y-like receptor GPR17 is a sensor of brain damage and a new targetfor brain repair. PLoS One 3: e3579.24. Boda E, Vigano F, Rosa P, Fumagalli M, Labat-Gest V, et al. (2011) TheGPR17 receptor in NG2 expressing cells: Focus on in vivocell maturation andparticipation in acute trauma and chronic damage. Glia 59: 1958–1973.25. Reimer MM, McQueen J, Searcy L, Scullion G, Zonta B, et al. (2011) Rapiddisruption of axon-glial integrity in response to mild cerebral hypoperfusion.Journal of Neuroscience 31: 18185–18194.26. Holland PR, Bastin ME, Jansen MA, Merrifield GD, Coltman RB, et al. (2011)MRI is a sensitive marker of subtle white matter pathology in hypoperfusedmice. Neurobiology of Aging 32: 2325.e2321–e2326.27. Shibata M, Yamasaki N, Miyakawa T, Kalaria RN, Fujita Y, et al. (2007)Selective impairment of working memory in a mouse model of chronic cerebralhypoperfusion. Stroke 38: 2826–2832.28. Shibata M, Ohtani R, Ihara M, Tomimoto H (2004) White matter lesions andglial activation in a novel mouse model of chronic cerebral hypoperfusion.Stroke 35: 2598–2603.29. Yool DA, Klugmann M, McLaughlin M, Vouyiouklis DA, Dimou L, et al.(2001) Myelin proteolipid proteins promote the interaction of oligodendrocytesand axons. J Neurosci Res 63: 151–164.30. Bhat RV, Axt KJ, Fosnaugh JS, Smith KJ, Johnson KA, et al. (1996) Expressionof the APC tumor suppressor protein in oligodendroglia. Glia 17: 169–174.31. Back SA, Gan X, Li Y, Rosenberg PA, Volpe JJ (1998) Maturation-dependentvulnerability of oligodendrocytes to oxidative stress-induced death caused byglutathione depletion. Journal of Neuroscience 18: 6241–6253.32. Husain J, Juurlink BH (1995) Oligodendroglial precursor cell susceptibility tohypoxia is related to poor ability to cope with reactive oxygen species. BrainResearch 698: 86–94.33. Mabuchi T, Kitagawa K, Ohtsuki T, Kuwabara K, Yagita Y, et al. (2000)Contribution of microglia/macrophages to expansion of infarction and responseof oligodendrocytes after focal cerebral ischemia in rats. Stroke 31: 1735–1743.34. Deng Y, Lu J, Sivakumar V, Ling EA, Kaur C (2008) Amoeboid microglia in theperiventricular white matter induce oligodendrocyte damage through expressionof proinflammatory cytokines via MAP kinase signaling pathway in hypoxicneonatal rats. Brain Pathology 18: 387–400.35. McTigue DM, Tripathi RB (2008) The life, death, and replacement ofoligodendrocytes in the adult CNS. Journal of Neurochemistry 107: 1–19.36. Alix JJ, Fern R (2009) Glutamate receptor-mediated ischemic injury ofpremyelinated central axons. Annals of Neurology 66: 682–693.37. Stys PK, Lipton SA (2007) White matter NMDA receptors: an unexpected newtherapeutic target? Trends in Pharmacological Sciences 28: 561–566.38. Fujita Y, Ihara M, Ushiki T, Hirai H, Kizaka-Kondoh S, et al. (2010) Earlyprotective effect of bone marrow mononuclear cells against ischemic whitematter damage through augmentation of cerebral blood flow. Stroke; a journalof cerebral circulation 41: 2938–2943.39. Maki T, Ihara M, Fujita Y, Nambu T, Miyashita K, et al. (2011) Angiogenic andvasoprotective effects of adrenomedullin on prevention of cognitive decline afterchronic cerebral hypoperfusion in mice. Stroke; a journal of cerebral circulation42: 1122–1128.40. Duan W, Ran H, Zhou Z, He Q, Zheng J (2012) Adenosine A2A receptordeficiency up-regulates cystatin F expression in white matter lesions induced bychronic cerebral hypoperfusion. PLoS One 7: e52566.41. Miyamoto N, Maki T, Pham LD, Hayakawa K, Seo JH, et al. (2013) OxidativeStress Interferes With White Matter Renewal After Prolonged CerebralHypoperfusion in Mice. Stroke.42. Miyamoto N, Pham LD, Hayakawa K, Matsuzaki T, Seo JH, et al. (2013) Age-Related Decline in Oligodendrogenesis Retards White Matter Repair in Mice.Stroke.43. Taupin P (2007) BrdU immunohistochemistry for studying adult neurogenesis:paradigms, pitfalls, limitations, and validation. Brain Res Rev 53: 198–214.44. Cavaliere F, Urra O, Alberdi E, Matute C (2012) Oligodendrocytedifferentiation from adult multipotent stem cells is modulated by glutamate.Cell Death and Disease 3: e268.45. Chen Y, Wu H, Wang S, Koito H, Li J, et al. (2009) The oligodendrocyte-specific G protein-coupled receptor GPR17 is a cell-intrinsic timer ofmyelination. Nature Neuroscience 12: 1398–1406.46. Ceruti S, Vigano F, Boda E, Ferrario S, Magni G, et al. (2011) Expression of thenew P2Y-like receptor GPR17 during oligodendrocyte precursor cell maturationregulates sensitivity to ATP-induced death. Glia 59: 363–378.47. Oluich LJ, Stratton JA, Lulu Xing Y, Ng SW, Cate HS, et al. (2012) Targetedablation of oligodendrocytes induces axonal pathology independent of overtdemyelination. Journal of Neuroscience 32: 8317–8330.48. Richardson WD, Young KM, Tripathi RB, McKenzie I (2011) NG2-glia asmultipotent neural stem cells: fact or fantasy? Neuron 70: 661–673.Oligodendrocyte Pools after Cerebral Hypoperfusion PLOS ONE | www.plosone.org12February 2014 | Volume 9 | Issue 2 | e87227 Neurobiology of DiseaseRapid Disruption of Axon–Glial Integrity in Response toMild Cerebral Hypoperfusion Michell M. Reimer,1,3 Jamie McQueen,1,3 Luke Searcy,1,2 Gillian Scullion,1 Barbara Zonta,1 Anne Desmazieres,1Philip R. Holland,1,3 Jessica Smith,1 Catherine Gliddon,1 Emma R. Wood,2 Pawel Herzyk,4 Peter J. Brophy,1James McCulloch,2,3 and Karen Horsburgh1,31University of Edinburgh, Centre for Neuroregeneration, Edinburgh EH16 4SB, United Kingdom, 2University of Edinburgh, Centre for Cognitive and NeuralSystems, Edinburgh EH8 9JZ, United Kingdom,3University of Edinburgh, Centre for Cognitive Ageing and Cognitive Epidemiology, Edinburgh EH8 9JZ,United Kingdom, and4University of Glasgow, Institute of Molecular, Cell and Systems Biology College of Medical, Veterinary and Life Sciences, GlasgowG12 8QQ, United Kingdom Myelinated axons have a distinct protein architecture essential for action potential propagation, neuronal communication, and main-taining cognitive function. Damage to myelinated axons, associated with cerebral hypoperfusion, contributes to age-related cognitivedecline. We sought to determine early alterations in the protein architecture of myelinated axons and potential mechanisms afterhypoperfusion. Using a mouse model of hypoperfusion, we assessed changes in proteins critical to the maintenance of paranodes, nodesof Ranvier, axon– glial integrity, axons, and myelin by confocal laser scanning microscopy. As early as 3 d after hypoperfusion, theparanodal septate-like junctions were damaged. This was marked by a progressive reduction of paranodal Neurofascin signal and a lossof septate-like junctions. Concurrent with paranodal disruption, there was a significant increase in nodal length, identified by Nav1.6staining, with hypoperfusion. Disruption of axon– glial integrity was also determined after hypoperfusion by changes in the spatialdistribution of myelin-associated glycoprotein staining. These nodal/paranodal changes were more pronounced after 1 month of hypo-perfusion. In contrast, the nodal anchoring proteins AnkyrinG and Neurofascin 186 were unchanged and there were no overt changes inaxonal and myelin integrity with hypoperfusion. A microarray analysis of white matter samples indicated that there were significantalterations in 129 genes. Subsequent analysis indicated alterations in biological pathways, including inflammatory responses, cytokine-cytokine receptor interactions, blood vessel development, and cell proliferation processes. Our results demonstrate that hypoperfusionleads to a rapid disruption of key proteins critical to the stability of the axon– glial connection that is mediated by a diversity of molecularevents. IntroductionThe integrity of the brain’s white matter, comprised mainly ofmyelinated axons, is critical in regulating efficient neuronal com-munication and maintaining cognitive function (Nave, 2010).Myelination of axons, by oligodendrocytes, determines the local-ization of key proteins along the axon and segregates the axonalmembrane into defined regions: the node of Ranvier, paranode,juxtaparanode, and internode (Rios et al., 2003; Susuki and Ras-band, 2008). In the adult brain, the nodes of Ranvier located atdefined points along the axolemma are critical for action poten-tial propagation. They are comprised of voltage-gated sodiumchannels that are clustered at a high density within a diffusionbarrier of septate-like junctions. These paranodal septate-likejunctions, consisting of glial Neurofascin (Nfasc)155, axonalcontactin, and contactin-associated protein (Caspr), are criticalto the maintenance of axon– glial junctions at the paranodes. Adynamic relationship exists between axons and glia, in additionto highly coordinated signaling with vascular components (en-dothelial cells, pericytes) that comprise the oligovascular niche(Arai and Lo, 2009), which, if disrupted, could impede neuronalcommunication and cognitive abilities.With advancing age, there is a general decline in white matterintegrity associated with cognitive decline (Bartzokis et al., 2004;Bastin et al., 2009). Although there is evidence that the proteinarchitecture of myelinated axons may be altered at nodal regionsin the aging brain (Lasiene et al., 2009), as yet the underlyingmolecular changes that occur early are not known. Cerebral hy-poperfusion is suggested to contribute to the development ofwhite matter changes in the aging brain (Fernando et al., 2006).In response to severe hypoperfusion as a result of stroke, axon–glial integrity is impaired, as demonstrated by a loss of myelin-associated glycoprotein (MAG) (Waxman, 2006), a proteinReceived Sept. 28, 2011; revised Oct. 28, 2011; accepted Nov. 2, 2011.Author contributions: M.M.R., L.S., and K.H. designed research; M.M.R., J.McQu, L.S., G.S., P.R.H., J.S., and C.G.performed research; B.Z., A.D., and P.B. contributed unpublished reagents/analytic tools; M.M.R., J.McQ., L.S.,E.R.W., P.H., J.McC. and K.H. analyzed data; M.M.R. and K.H. wrote the paper.This work was supported by the Disconnected Mind program (supported by Age, UK and Alzheimer’s ResearchUK). We gratefully acknowledge the University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiol-ogy (part of the cross council Lifelong Health and Wellbeing Initiative).Correspondence should be addressed to Dr. Karen Horsburgh, Centre for Neuroregeneration, The Medical School,University of Edinburgh, Chancellor’s Building, 49 Little France Crescent, Edinburgh EH16 4SB, United Kingdom.E-mail: [email protected]. Smith’s present address: Alzheimer’s Society, Devon House, London E1W 1LB, United Kingdom.DOI:10.1523/JNEUROSCI.4936-11.2011Copyright © 2011 the authors 0270-6474/11/3118185-10$15.00/0The Journal of Neuroscience, December 7, 2011 • 31(49):18185–18194 • 18185