alterations in microglial phenotype and hippocampal neuronal...

Brain, Behavior, and Immunity 45 (2015) 80–97

Contents lists available at ScienceDirect

Brain, Behavior, and Immunity

journal homepage: www.elsevier .com/locate /ybrbi

Alterations in microglial phenotype and hippocampal neuronal functionin transgenic mice with astrocyte-targeted production of interleukin-10

http://dx.doi.org/10.1016/j.bbi.2014.10.0150889-1591/� 2014 Elsevier Inc. All rights reserved.

Abbreviations: BBB, blood brain barrier; BSA, bovine serum albumin; CNS,central nervous system; DAPI, 40 ,6-diamidino-2-phenylindol; EAE, experimentalautoimmune encephalomyelitis; fEPSPs, field excitatory post-synaptic potentials;GFAP, glial fibrillary acidic protein; HFS, high frequency stimulation; hGH, humangrowth hormone gene; IL-10, interleukin-10; IL-10R, interleukin-10 receptor; LTP,long-term potentiation; MCAO, middle cerebral artery occlusion; MHC, majorhistocompatibility complex; PBS, phosphate buffer solution; PCR, polymerase chainreaction; pH3, phospho-histone 3; qRT-PCR, quantitative real time-polymerasechain reaction; RPA, ribonuclease protection assay; RT, room temperature; SOCS3,suppressor of cytokine signalling 3; TBS, Tris-buffered saline.⇑ Corresponding author at: Unitat d’Histologia, Torre M5, Facultat de Medicina,

Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain. Tel.: +34 935811826;fax: +34 935812392.

E-mail address: [email protected] (B. Almolda).

Beatriz Almolda a,⇑, Carmen de Labra a, Iliana Barrera a, Agnès Gruart b, Jose M. Delgado-Garcia b,Nàdia Villacampa a, Antonietta Vilella c, Markus J. Hofer d, Juan Hidalgo a, Iain L. Campbell d,Berta González a, Bernardo Castellano a

a Department of Cell Biology, Physiology and Immunology, Institute of Neuroscience, Universitat Autònoma de Barcelona, Bellaterra 08193, Spainb Division of Neurosciences, Pablo de Olavide University, Seville 41013, Spainc Department of Biomedical, Metabolic and Neural Sciences, Università degli Studi di Modena e Reggio Emilia, 41125, Italyd School of Molecular Bioscience, The University of Sydney, Sydney, NSW 2006, Australia

a r t i c l e i n f o a b s t r a c t

Article history:Received 11 July 2014Received in revised form 24 September2014Accepted 25 October 2014Available online 31 October 2014

Keywords:Transgenic miceCytokineNeuroinflammationGliosisMicroglial plasticityLTPIba1CD16/32CD11bF4/80CD150GFAPIL-10R

Interleukin-10 (IL-10) is a cytokine classically linked with anti-inflammatory and protective functions inthe central nervous system (CNS) in different neurodegenerative and neuroinflammatory conditions. Inorder to study the specific role of local CNS produced IL-10, we have created a new transgenic mouse linewith astrocyte-targeted production of IL-10 (GFAP-IL10Tg). In the present study, the effects of local CNSIL-10 production on microglia, astrocytes and neuronal connectivity under basal conditions were inves-tigated using immunohistochemistry, molecular biology techniques, electrophysiology and behaviouralstudies. Our results showed that, in GFAP-IL10Tg animals, microglia displayed an increase in densityand a specific activated phenotype characterised by morphological changes in specific areas of the brainincluding the hippocampus, cortex and cerebellum that correlated with the level of transgene expressedIL-10 mRNA. Distinctively, in the hippocampus, microglial cells adopted an elongated morphology fol-lowing the same direction as the dendrites of pyramidal neurons. Moreover, this IL-10-induced microglialphenotype showed increased expression of certain molecules including Iba1, CD11b, CD16/32 and F4/80markers, ‘‘de novo’’ expression of CD150 and no detectable levels of either CD206 or MHCII. To evaluatewhether this specific activated microglial phenotype was associated with changes in neuronal activity,the electrophysiological properties of pyramidal neurons of the hippocampus (CA3-CA1) were analysedin vivo. We found a lower excitability of the CA3-CA1 synapses and absence of long-term potentiation(LTP) in GFAP-IL10Tg mice. This study is the first description of a transgenic mouse with astrocyte-tar-geted production of the cytokine IL-10. The findings indicate that IL-10 induces a specific activatedmicroglial phenotype concomitant with changes in hippocampal LTP responses. This transgenic animalwill be a very useful tool to study IL-10 functions in the CNS, not only under basal conditions, but alsoafter different experimental lesions or induced diseases.

� 2014 Elsevier Inc. All rights reserved.

1. Introduction

Interleukin 10 (IL-10) is one of the most crucial immunoregula-tory cytokines in the periphery where in general it has anti-inflam-matory functions (Couper et al., 2008; Moore et al., 2001). In thecentral nervous system (CNS), IL-10 expression has been reportedto be upregulated under a variety of neuroinflammatory and path-ological situations including traumatic brain injury (Kamm et al.,2006), excitotoxicity (Gonzalez et al., 2009), middle cerebral arteryocclusion (MCAO) (Zhai et al., 1997), Alzheimer’s disease (Apeltand Schliebs, 2001), multiple sclerosis (Hulshof et al., 2002) andexperimental autoimmune encephalomyelitis (EAE) (Ledeboer

http://crossmark.crossref.org/dialog/?doi=10.1016/j.bbi.2014.10.015&domain=pdf

http://dx.doi.org/10.1016/j.bbi.2014.10.015

mailto:[email protected]

http://dx.doi.org/10.1016/j.bbi.2014.10.015

http://www.sciencedirect.com/science/journal/08891591

http://www.elsevier.com/locate/ybrbi

B. Almolda et al. / Brain, Behavior, and Immunity 45 (2015) 80–97 81

et al., 2003). Noteworthy, and consistent with the anti-inflamma-tory role attributed to IL-10 in the periphery (Couper et al., 2008;Moore et al., 2001), upregulation of IL-10 expression after EAEoccurs mostly during the recovery phase (Issazadeh et al., 1995;Ledeboer et al., 2003) and in the MCAO model coincides with adecrease of pro-inflammatory mediators (Zhai et al., 1997).

Expression of IL-10 receptor (IL10R) has been described in neu-rons (Lim et al., 2013), microglia (Ledeboer et al., 2002; Nordenet al., 2014), astrocytes (Gonzalez et al., 2009; Ledeboer et al.,2002; Norden et al., 2014) and oligodendrocytes (Cannella andRaine, 2004), in both basal conditions and after injury, indicatinga potentially broad range of actions of this cytokine in the CNS.Indeed, administration of IL-10 has been found to reduce astroglialactivation (Balasingam and Yong, 1996; Pang et al., 2005), microg-lial production of LPS-induced inflammatory mediators(Balasingam and Yong, 1996; Kremlev and Palmer, 2005; Lodgeand Sriram, 1996; Molina-Holgado et al., 2001b; Norden et al.,2014; Pang et al., 2005; Sawada et al., 1999) and leukocyte infiltra-tion (Ooboshi et al., 2005). Moreover, IL-10 exerts a neuroprotectiverole in several in vitro and in vivo models of CNS injury, such as trau-matic brain injury, excitotoxicity and MCAO among others(Arimoto et al., 2007; Bachis et al., 2001; Brewer et al., 1999;Knoblach and Faden, 1998; Ledeboer et al., 2000; Molina-Holgadoet al., 2001a; Ooboshi et al., 2005; Park et al., 2007; Spera et al.,1998; Xin et al., 2011). Remarkably, the route of IL-10 administra-tion has been reported to be essential for IL-10 protection. Thus,EAE was completely prevented by IL-10 administration only whenIL-10 was administered intraparenchymatically, but when adminis-tered systemically there was no effect (Cua et al., 2001) or in somecases, worsened the disease (Cannella et al., 1996). Similarly,peripherally administered IL-10 fails to rescue facial motor neuronsfrom the axotomy-induced cell death observed in IL-10 KO mice(Xin et al., 2011). Furthermore, after acute injuries such as spinalcord excitotoxicity or traumatic brain injury, intraspinal or intra-cerebroventricular IL-10 administration worsens disease (Breweret al., 1999) or has no effect (Knoblach and Faden, 1998), whereassystemic administration improves the functional outcome oflesions (Brewer et al., 1999; Knoblach and Faden, 1998).

In this context and in order to obtain a better understanding ofthe role played by intrathecally produced IL-10 specifically in theCNS, we have generated a new transgenic mouse model in whichthe cDNA encoding for murine IL-10 was placed under the regula-tory control of a glial fibrillary acidic protein (GFAP) transgene inastrocytes. The main goal of the present study was to determinethe effects that local IL-10 production exerts on the populationsof neurons, microglia and astrocytes under basal conditions.

2. Material and methods

2.1. Construction of the GFAP-IL10 fusion gene and production oftransgenic mice

The full-length cDNA encoding murine IL-10 was cloned into aconstruct containing the mouse glial fibrillary acidic protein(GFAP) promoter and the polyadenylation signal sequence fromthe human growth hormone gene (hGH), as previously described(Stalder et al., 1998) (Fig. 1A).

The GFAP-IL10 construct was microinjected into fertilised eggsfrom SJL/L mice. Transgenic offspring were identified by PCR ongenomic DNA extracted from tail biopsies using primers againstthe hGH sequence (Fig. 1B). The F1 offspring were backcrossedwith the C57BL/6 strain for at least 10 generations to obtain trans-genic mice on the C57BL/6 background.

A total of 48 GFAP-IL10 transgenic (GFAP-IL10Tg) animals andtheir corresponding wild-type (WT) littermates (n = 49) of bothsexes were used in this study. Animals were maintained with food

and water ad libitum in a 12 h light/dark cycle during theexperiment.

All experimental animal work was conducted according toSpanish regulations (Ley 32/2007, Real Decreto 1201/2005, Ley9/2003 and Real Decreto 178/2004) in agreement with EuropeanUnion directives (86/609/CEE, 91/628/CEE and 92/65/CEE) andwas approved by the Ethical Committee of the AutonomousUniversity of Barcelona.

2.2. Tissue processing for PCR analysis

DNA was extracted from tail biopsies using the DNA extractionkit (740.952.250, Macherey–Nagel) following the manufacturer’sinstructions. Briefly, tail samples were incubated for 2 h at 56 �Cin 180 ll T1 buffer and 25 ll proteinase-K. The supernatantobtained after centrifugation for 5 min at 12,000 rpm was trans-ferred to a new tube and 200 ll of lysis buffer 3 and 200 ll of100% ethanol added and mixed gently. DNA was separated usingspecific columns provided in the kit that were centrifuged at12,000 rpm for 3 min. After 2 washes and centrifugation rounds(12,000 rpm for 3 min) with washing buffer and 1 with buffer 5,the DNA was eluted from the column and recovered in a new tubeby adding buffer BE and centrifuging at 12,000 rpm for 2 min.

2.3. Tissue processing for ribonuclease protection assay (RPA) analysis

For RPA, 4 month-old WT (n = 3) and GFAP-IL10Tg animals(n = 3) were anaesthetized with ketamine (80 mg/kg) and xylazine(20 mg/kg; 0.015 ml/g) solution and intracardially perfused withphosphate buffer solution (PBS). The brain was quickly removedand the hippocampus, cerebral cortex cerebellum, thalamus, brain-stem and spinal cord areas were dissected out and processed sep-arately. RNA was prepared using Tri-reagent (T9424, SigmaAldrich) performed according to the manufacturer’s instructions.RPA was performed as described previously (Campbell et al.,1994). The RNA samples (3 lg of total RNA) were hybridized with[32P]UTP-labelled probe sets, containing cRNA probes for IL-10,CD11b, SOCS3, GFAP and L32. For quantification, autoradiographswere scanned and analysed by densitometry using NIH Image Jsoftware (Wayne Rasband, National Institutes of Health, USA).The densitometry value for each transcript was expressed as a ratioto the L32 RNA, which served as a control for RNA loading.

2.4. Tissue processing for Bioplex protein microarray

Adult (6 months) GFAP-IL10Tg animals (n = 8) and their corre-sponding WT littermates (n = 8) were euthanised under anaesthe-sia (as described above) and perfused intracardially for 30 s withphosphate buffer solution (PBS), the entire hippocampus was dis-sected out quickly, snap frozen in liquid nitrogen and stored at�80 �C. Total protein was extracted by solubilisation of samplesin Lysis buffer containing 250 mM HEPES (pH 7.4), 0.2% Igepal,5 mM MgCl2 , 1.3 mM EDTA, 1 mM EGTA, 1 mM PMSF and proteaseand phosphatase inhibitor cocktails (1:100, Sigma Aldrich). Follow-ing solubilisation, samples were clarified by centrifugation at13,000 rpm for 5 min and the supernatant retained. Total proteinconcentration was determined with a commercial Pierce BCA Pro-tein Assay kit (23225, Thermo Scientific) according to the manufac-turer’s protocol. Protein lysates were stored aliquoted at �80 �Cuntil used for Bio-plex protein microarray.

2.5. Analysis of IL-10 with Bio-plex protein microarray

The cytokine IL-10 was analysed using a Bio-plex Pro TM Mousecytokine GrpI panel kit (M60-009RDPD, Bio-rad) according to themanufacturer’s instructions. Briefly, 50 ll of hippocampus extracts

Fig. 1. Generation of the GFAP-IL10Tg mice. (A) Representation of the GFAP-IL10 fusion vector used for the generation of GFAP-IL10Tg mice. (B) Representative PCR resultobtained from tail biopsies. Note that transgenic mice are characterised by the detection of the PCR product hGH Poly A found in the construct. Detection of a house-keepinggene (fatty acid-binding protein 2 intestinal) is used as a control of the technique and therefore is found in both WT and GFAP-IL10Tg animals.

82 B. Almolda et al. / Brain, Behavior, and Immunity 45 (2015) 80–97

were diluted with sample diluent (1:3), added to the plate, alongwith the standards in separate wells, containing 50 ll of customfluorescent beads and incubated for 90 min at room temperature(RT) in a plate-shaker (750 rpm). After three washes with washbuffer (1�), plates were incubated with 25 ll of biotinylated anti-body reagent for 30 min at RT in a plate-shaker (750 rpm). Subse-quently, the plate was washed three times with wash buffer andincubated for 10 min at RT in plate-shaker (750 rpm) with 50 llof streptavidin-PE conjugated reagent. Finally, a Luminex� Mag-pix� device with the xPONENT� 4.2 software was used to readthe plate. Data was analysed using the Milliplex� Analyst 5.1 soft-ware. Data was corrected as picograms/milligram of protein.

2.6. Sample processing for qRT-PCR

Adult (6 months) GFAP-IL10Tg animals (n = 3) and their corre-sponding WT littermates (n = 3) were deeply anaesthetized witha solution of xylazine (20 mg/kg) and ketamine (80 mg/kg) injectedintraperitoneally (0.015 ml/g) and intracardially perfused with asolution containing phosphate buffer solution (PBS) and 0.1% ofdiethyl pyrocarbonate water (DEPC, Sigma, 40718). Brains werequickly dissected out, frozen using dry ice and stored at �80 �Cuntil use. The hippocampus was lysed by mechanical disruptionin Trizol reagent (Qiagen) and homogenised following the proce-dure provided by the manufacturer. Isolated mRNA was reversetranscribed to cDNA using a first-strand synthesis kit and M-MLVReverse Transcriptase (Promega Corporation, MA, USA). Sampleswere heated at 70 �C for 5 min to eliminate any secondary struc-tures, then incubated at 23 �C for 10 min, 1 h at 37 �C and 5 minat 95 �C before being chilled at 4 �C using a thermocycler T Gradi-ent (Whatman, Biometra). The amount of cDNA was quantifiedwith iTaq Universal SYBR� Green Supermix (Bio Rad) using a BioRad qRT-PCR iCycler. Each PCR reaction was performed in triplicatewith the following cycling parameters: 10 min at 95 �C and 40cycles of 1 min at 95 �C, 1 min at 60 �C and 1 min at 72 �C, followedby a melting curve analysis. qRT-PCR primers were designed in twodifferent exons; primer length was comprised between 18 and30 bp, GC content was between 40% and 60% and nonspecific pri-mer annealing and mismatches were minimised. The presence of

non-specific products of amplification and primer–dimer presencewere evaluated by melting curve analysis during qRT-PCR primervalidation. The following primers were used to amplify the tran-scripts of interest:

GADPH 50: CATCAAGAAGGTGGTGAAGCGADPH 30: ACCACCCTGTTGCTGTAGIL10R 50: TCACGACGGAGCAGTATTTCAIL10R 30: GAAGACCAGGACTGTAGGCA

Fold differences of expression were calculated using the com-parative method, also referred as DDCt Method (Livak andSchmittgen, 2001). The media of the WT samples was used asthe reference sample or calibrator.

2.7. Tissue processing for histological analysis

Adult (6–10 months-old) WT (n = 11) and GFAP-IL10Tg animals(n = 10) used for immunohistochemical analysis were euthanisedunder anaesthesia (as described above) and perfused intracardiallyfor 10 min with 4% paraformaldehyde in 0.1 M PBS (pH 7.4). Foreach animal, the telencephalon, brainstem and cerebellum weredissected immediately and postfixed for 4 h at 4 �C in the same fix-ative. Afterwards, samples were cryoprotected in 30% sucrose solu-tion in 0.1 M phosphate buffer for 48 h at 4 �C and, finally, frozen inice-cold methylbutane (320404, Sigma–Aldrich). Frozen parallelcoronal sections (30 lm thick) were obtained using a CM 3050SLeica cryostat and stored free-floating at �20 �C in Olmos anti-freeze buffer.

2.8. Toluidine blue staining

One series of parallel coronal sections were mounted onto gel-atinized slides, air dried at RT for one hour and were then incu-bated for 1 min in a solution containing 0.1% toluidine bluediluted in Walpole’s buffer (0.05 M, pH 4.5). After washes in dis-tilled water, sections were dehydrated in graded alcohols, N-butylalcohol and after xylene treatment, coverslipped with DPX mount-ing media.


2.9. Single and double immunohistochemistry

Parallel free-floating cryostat sections were processed for thevisualisation of microglial cells and the characterisation of theiractivation phenotype using different antibodies: Iba1, CD11b,CD16/32, F4/80, CD150, CD206, MHCII, Pu.1 and phospho-histone3 (pH3). For the study of astrocytes sections were immunostainedwith GFAP and with synaptophysin for the study of synapsis den-sity (Table 1). After 10 min of endogenous peroxidase blockingwith 2% H2O2 in 70% methanol, sections were incubated for 1 hin a blocking solution of 0.05 M Tris-buffered saline (TBS), pH7.4, containing 10% foetal calf serum, 3% bovine serum albumin(BSA) and 1% Triton X-100. In the case of Pu.1, previous to theblocking solution incubation, an antigen retrieval treatment wasused incubating the sections with sodium citrate buffer (pH 8.5)at 80 �C for 40 min. In the case of synaptophysin, previous to theblocking step, sections were incubated overnight at 4 �C and 1 hat RT in F’ab blocking solution (1:100; 315-007-003; JacksonImmunoresearch) diluted in TBS. Then, sections were incubatedovernight at 4 �C and 1 h at RT with the corresponding primaryantibody (Table 1) diluted in the same blocking solution. In thecase of IL10Ra, sections were incubated for 48 h at 4 �C and 1 hat RT. Sections incubated in media lacking the primary antibodywere used as negative control, and spleen sections as positive con-trol. After washes with TBS+ 1% Triton (3 � 5 min), sections wereincubated at RT for 1 h with either biotinylated anti-rabbit second-ary antibody or biotinylated anti-rat secondary antibody (Table 1).For the study of the blood brain barrier integrity, after endogenousperoxidase blocking and blocking solution incubation, sectionswere incubated for 1 h in biotinylated anti-mouse IgG (1:500)diluted in the same blocking solution. After 1 h in streptavidin-per-oxidase, the reaction was visualised by incubating the sections in aDAB kit (SK-4100; Vector Laboratories, Inc; Burlingame, CA) fol-lowing the manufacturer’s instructions. Finally, sections weremounted on slides, dehydrated in alcohol and after xylene treat-ment, coverslipped with DPX.

Double-immunolabelling was carried out by firstly processingthe sections for Iba1 or IL-10Ra immunolabelling as describedabove, but using, respectively, AlexaFluor� 488-conjugated anti-rabbit as secondary antibody or AlexaFluor� 555-conjugated strep-tavidin. After several washes and incubation in blocking solution

Table 1List of antibodies and reagents used for immunohistochemistry.

Target antigen Host

Primary antibodies Iba1 RabbitCD16/32 RatCD11b RatF4/80 RatCD150 RatCD206 RatMHCII (IA) Rat hybridomaPu.1 RabbitGFAP RabbitGFAP MousepH3 RabbitSynaptophysin MouseIL10Ra Rabbit

Secondary antibodies Biotinylated RabbitBiotinylated RatBiotinylated MouseAlexa 488 RabbitAlexa 555 RatAlexa 488 Mouse

Streptavidin-Alexa Fluor 555

Streptavidin-HRP

DAPI

for an hour, sections were incubated overnight at 4 �C and 1 h atRT with one of the primary antibodies specified in Table 1. Thiswas followed by 1 h incubation in AlexaFluor� 555-conjugatedanti-rat or AlexaFluor� 488-conjugated anti-mouse secondaryantibodies. Finally, sections were stained with DAPI (40,6-diamidi-no-2-phenylindol) (Table 1), mounted on slides, dehydrated ingraded alcohol and coverslipped in DPX. Sections were analysedusing a fluorescence Nikon Eclipse E600 microscope and a ZeissLSM 700 confocal microscope.

2.10. Densitometric analysis

Densitometric analysis was performed on sections immunola-belled with Iba1, CD11b, CD16/32, F4/80 and GFAP. A minimumof three WT and three GFAP-IL10Tg animals were used. For eachimmunohistochemistry technique and animal, a total of 12 photo-graphs from the CA1 area from 2 different hippocampal sections,12 photographs from the hindlimb cortex area of 2 different cere-bral cortex sections and 12 photographs from 2 different cerebellarcortex sections were captured at 20� magnification with a DXM1200F Nikon digital camera mounted on a brightfield Nikon Eclipse80i microscope using the software ACT-1 2.20 (Nikon corporation).By means of analySIS� software, both the percentage of area occu-pied by the immunolabelling as well as the intensity of the immu-noreaction (Mean Grey Value Mean) was recorded for eachphotograph. For each marker, the AI index (Almolda et al., 2014)was calculated by multiplying the percentage of the immunola-belled area by the Mean Grey Value Mean.

Quantification of microglial cell density was performed on sec-tions immunostained for Pu.1. At least three WT and three GFAP-IL10Tg animals were analysed. For each animal, a minimum of15 photographs from the CA1 area of 2 different hippocampus sec-tions, 8 photographs from the hindlimb area of 2 different cerebralcortex sections and 10 photographs of 2 different cerebellar cortexsections were captured at 20�magnification with the same deviceand software referred to above. The number of Pu.1+ cells per pho-tograph (0.0058 mm2 frame) was obtained using the ‘‘Automaticnuclei counter (ITCN)’’ plug-in from NIH Image J� software (WayneRasband, National Institutes of Health, USA), averaged and con-verted to cells/mm2.

Dilution Cat Number Manufacturer

1:3000 019-19741 Wako1:1000 553142 BD Pharmingen1:1000 MCA74GA AbD Serotec1:250 ab6640 AbCam1:125 MCA2274 AbD Serotec1:500 MCA2235GA AbD Serotec1:25 TIB-120 ATCC1:400 2258S Cell signalling1:1800 Z0334 Dakopatts1:5000 G3893 Sigma–Aldrich1:3000 06-570 Millipore1:200 S5768 Sigma–Aldrich1:50 sc-985 Santa Cruz Biotechnology

1:500 BA-1000 Vector Laboratories1:500 BA-4001 Vector Laboratories1:500 BA-2001 Vector Laboratories1:1000 A21206 Molecular probes1:1000 A21434 Molecular probes1:1000 A11029 Molecular probes

1:1000 S21381 Molecular probes

1:500 SA-5004 Vector Laboratories

1:10000 D9542 Sigma Aldrich


Quantification of synaptic density was performed on sectionsstained for synaptophysin. At least three WT and three GFAP-IL10Tg animals were analysed. For each animal, a minimum of 20photographs from the CA1 area of 2 different hippocampus sec-tions was captured at 20�magnification with a Zeiss LSM 700 con-focal microscope. By means of NIH Image J� software (WayneRasband, National Institutes of Health, USA), both the percentageof area occupied by the immunolabelling as well as the intensityof the immunofluorescence (Mean Grey Value Mean) was recordedfor each photograph.

2.11. Morphometric analysis

Morphometric analysis of microglial cells was done on sectionsimmunolabelled for Iba1. At least three WT and three GFAP-IL10Tganimals were analysed. For each animal, a total of 60 representa-tive microglial cells from each brain region were randomly chosenfrom 12 different photographs from the hippocampus, the cerebralcortex or the cerebellar cortex and photographed at 20� magnifi-cation. Using the analySIS� software, individual cells were isolatedand different parameters including the area occupied by each cell,shape factor and elongation were recorded for each cell.

2.12. Electrophysiological analysis

6 month-old WT and GFAP-IL10Tg mice (n = 10 per group) wereanesthetised with 4% chloral hydrate. Animals were implantedwith bipolar stimulating electrodes in the right Schaffer collateralpathway of the dorsal hippocampus (2 mm lateral and 1.5 mmposterior to Bregma; depth from brain surface, 1.0–1.5 mm) andwith two recording electrodes in the ipsilateral stratum radiatumunderneath the CA1 area (1.2 mm lateral and 2.2 mm posteriorto Bregma; depth from brain surface, 1.0–1.5 mm) (Fig. 9A). Elec-trodes were made of 50 lm, Teflon-coated tungsten wire (AdventResearch Materials Ltd., Eynsham, England). The final position ofhippocampal electrodes was determined using as a guide the fieldpotential depth profile evoked by paired (40 ms interval) pulsespresented at the Schaffer’s collateral pathway. A bare silver wire(0.1 mm) was affixed to the skull as ground. Wires were solderedto a 6-pin socket and the socket was fixed to the skull with the helpof two small screws and dental cement (Gruart et al., 2006).

2.12.1. Recording and stimulation proceduresField excitatory post-synaptic potentials (fEPSPs) were recorded

using Grass P511 differential amplifiers through a high-impedanceprobe (2 � 1012 X, 10 pF). Electrical stimuli presented to Schaffercollaterals consisted of 100 ls, square, biphasic pulses presentedalone, paired, or in trains. Stimulus intensities ranged from 0.02to 0.4 mA for the construction of the input/output curves. For thepaired pulse test, the stimulus intensity was set well below thethreshold for evoking a population spike, usually 35% of the inten-sity was necessary for evoking a maximum fEPSP response(Gureviciene et al., 2004). Paired pulses were presented at six(10, 20, 40, 100, 200, and 500) increasing pulse intervals. Forlong-term potentiation (LTP) induction in alert behaving mice,the stimulus intensity was also set at 35% of its asymptotic value.Baseline records were collected for 15 min at a rate of 1 stimulus/20 s. For LTP induction, each animal was presented with a high fre-quency stimulation (HFS) protocol consisting of five trains (200 Hz,100 ms) of pulses at a rate of 1/s. This protocol was presented 6times in total, at intervals of 1 min. The 100 ls, square, biphasicpulses used to evoke LTP were applied at the same intensity usedfor baseline records. LTP evolution was followed for up to 72 h afterHFS (Gruart et al., 2006). A criterion for selecting stimulus intensityfor LTP induction was that a second stimulus, presented 40 ms

after a conditioning pulse, evoked a larger (>20%) synaptic fieldpotential than the first one (Bliss and Gardner-Medwin, 1973).

2.13. Behavioural studies

Both 6 month-old WT (n = 14) and GFAP-IL10Tg (n = 14) ani-mals were studied in the open field, passive avoidance test, andthe acquisition of two selective associative learning tasks, duringthe light period (8:00–20:00 h).

Spontaneous motor activities were determined in a 30 � 30 cmarena provided with a 16 � 16 infrared detecting system (Actifot 8,Cibertec, Madrid, Spain). Both the percentage of time spent in thecentre of the arena and the accumulated motor activities weredetermined automatically, stored on-line in a computer and tota-lised every 3 min, for a total period of 15 min.

For the passive avoidance test and in accordance to a previousstudy from our group (Eleore et al., 2007), each animal was placedin darkness 5 min before training. Then, mice were placed individ-ually in an illuminated box (10 � 13 � 15 cm) connected to a darkbox of the same size equipped with an electric grid floor, and sep-arated by an automatic door. This door was opened 60 s later.Entrance of mice into the dark box was punished by an adequateelectric foot-shock (0.5 mA, 1 s). Mice that did not enter the darkcompartment (cut-off time = 300 s) were excluded from subse-quent experimentation. After 1 h, pre-trained mice were placedinto the illuminated box again and observed for up to 300 s. Micethat avoided the dark compartment during the total time of theexperiment were considered to remember the task. The time(step-through latency) that the mice took to enter the dark boxwas noted, and the mean time was calculated for each experimen-tal group. Mice were retested 48 h later. The step-through latency(in seconds) was determined automatically by the passive avoid-ance device (Ugo Basile, Comerio, Italy).

For the associative learning task training and testing were per-formed in six Skinner box operant chambers (Jurado-Parras et al.,2012; Madronal et al., 2010). Before training, mice were handleddaily for 7 days and food-deprived to 80% of their free-feedingweight. Training took place for 20 min to a maximum of 10 days,in which mice learnt to press a lever to receive pellets from a foodtray using a fixed-ratio (1:1) schedule. A more complex condition-ing task was carried out for 10 days using a light/dark protocol, inwhich only during the lighted period there was a pellet reward. Inaddition, lever presses carried out during the dark period restartedthe dark protocol for an additional time (10–20 s).

2.14. Statistical analysis

All experimental results were expressed as mean values ± stan-dard error. Statistics were performed using the Graph Pad Prism�

software. Either unpaired Student’s-t test, to compare betweenWT and GFAP-IL10Tg animals, or one-way ANOVA with Tukey’spost hoc test, to compare among different brain areas, was usedto determine statistically significant differences.

3. Results

3.1. General characteristics of GFAP-IL10Tg mice

Our analysis of GFAP-IL10Tg animals indicated that these ani-mals were born at the expected Mendelian ratio, with no changesin fertility or mating. Transgenic animals also showed a normaldevelopment, without any noticeable difference in body size,weight and life expectancy when compared with their WTlittermates.


3.2. Increase in IL-10, CD11b, SOCS3 and GFAP mRNA expression levelsin GFAP-IL10Tg mice

Different areas of the brain including the cerebral cortex, cere-bellum, hippocampus, thalamus, brainstem and spinal cord of bothWT and GFAP-IL10Tg animals were processed for RPA analysis todetermine the levels of the mRNA for IL-10, CD11b, SOCS-3 andGFAP. Our results (Fig. 2) showed that, while IL-10 mRNA wasundetectable in all brain regions examined from WT mice, IL-10mRNA was detected in GFAP-IL10Tg mice in these brain areas.Among these regions, the hippocampus was the area where IL-10mRNA was the highest.

Fig. 2. RPA analysis and IL-10 protein levels. (A) Representative RPA gel showing the leveboth WT and GFAP-IL10Tg animals. L32 was used as control and applied for the normalisfor IL-10 (B), CD11b (C), SOCS-3 (D) and GFAP (E) in the cerebral cortex, cerebellum, hippIL10Tg animals (black bars). Note that IL-10 mRNA is only detected in GFAP-IL10Tg micedifferences between WT and GFAP-IL10Tg mice in the levels of CD11b and SOCS-3 expredifferences between WT and GFAP-IL10Tg animals in GFAP mRNA levels are only observehistogram corresponding to the quantification of IL-10 protein levels in the hippocampulevels in transgenic animals.

In comparison with samples from WT mice, a significantincrease in CD11b and SOCS-3 mRNA levels was detected inGFAP-IL10Tg animals in the cerebral cortex, cerebellum andhippocampus. The GFAP mRNA levels were increased only in thehippocampus but not the other regions of transgenic brain whencompared with the WT.

3.3. Increase in IL-10 protein levels in GFAP-IL10Tg animals

The entire hippocampus of both WT and GFAP-IL10Tg animalswas processed for Bio-plex protein microarray to determine thesynthesis of IL-10 protein. Our results clearly showed that IL-10

ls of IL-10, CD11b, SOCS-3, GFAP mRNAs detected along different areas of the CNS ofation of mRNA values. Histograms in (B–E) show the quantification of mRNA levelsocampus, thalamus, brainstem and spinal cord of both WT (white bars) and GFAP-

but not in WT, and shows the highest level of expression in the hippocampus. Majorssion are detected in the cerebral cortex, cerebellum and hippocampus. In contrast,d in the hippocampus (⁄p 6 0.05; ⁄⁄p 6 0.005; ⁄⁄⁄p 6 0.0001). Inset in (B) shows the

s of WT (white bar) and GFAP-IL10Tg animals (black bar). Note the increase in IL-10


protein levels were significantly higher in the hippocampus ofGFAP-IL10Tg animals than in WT mice (Fig. 2 inset in B).

3.4. Increase in IL-10Ra levels in GFAP-IL10Tg animals

Using qRT-PCR quantification, we observed that although IL-10Ra mRNA is expressed constitutively in the hippocampus ofWT animals, the expression was substantially higher in GFAP-IL10Tg mice (Fig. 3A). By immunohistochemistry, we determinedthat, whereas in WT animals the major part of IL-10Ra+ cells cor-responded to neurons, in GFAP-IL10Tg animals, in addition to neu-ronal expression, a high number of IL-10Ra+/GFAP+ astrocyteswere detected, specially in the molecular layer of the hippocampalCA1 region (Fig. 3B–E).

Fig. 3. IL-10Ra expression. (A) Histogram showing the IL-10Ra mRNA levels. Levels are ein GFAP-IL10Tg animals (black bars) was almost twice than in WT (white bars) (⁄p 6hippocampus of WT (B) and GFAP-IL10Tg animals (D). (C and E) Representative imagesGFAP (shown in green) in the CA1 region of the hippocampus of both WT (C) and GFAP-(For interpretation of the references to colour in this figure legend, the reader is referre

Fig. 4. Toluidine blue staining. Representative images showing the toluidine blue stainingboth WT (A–C) and GFAP-IL10Tg mice (D–F). No changes between WT and GFAP-IL10Tg abar = 50 lm. (For interpretation of the references to colour in this figure legend, the rea

3.5. Transgene-encoded IL-10 does not induce changes in the braincyto-architecture or the blood brain barrier permeability

To investigate for possible changes in the brain cyto-architec-ture, synaptic density or the integrity of the blood brain barrier(BBB) induced by transgene-encoded IL-10 production in the CNSa microscopic study was performed on toluidine blue, synapto-physin and IgG stained sections. In general, our analysis showedno detectable changes between WT and GFAP-IL10Tg animals ineither the disposition and size of neuronal layers or the total num-ber of cells (Fig. 4) as well as in the synaptic density (data notshown). Moreover, no discernible changes in BBB, monitored byIgG immunostaining, were observed in any of the cerebral or cere-bellar areas studied (data not shown).

xpressed as fold changes compared to the WT value. Note that IL-10Ra mRNA levels0.05). (B and D) Representative images showing single IL-10Ra staining in the

of double-immunofluorescence staining of IL-10Ra (shown in red) combined withIL10Tg animals (E). Yellow-orange staining shows colocalization. Scale bar = 30 lm.d to the web version of this article.)

of the cerebral cortex (A and D), cerebellum (B and E) and hippocampus (C and F) ofnimals were observed in the cytoarchitecture of the three brain areas studied. Scaleder is referred to the web version of this article.)

Fig. 5. Iba1 immunohistochemistry. Representative images showing the expression of Iba1 in the cerebral cortex (A, D, G and H), cerebellum (B, E, I and J) and hippocampus(C, F, K and L) of both WT (A–C, G, I and K) and GFAP-IL10Tg animals (D–F, H, J and L). Note the major differences in microglial morphology and Iba1 expression between WTand GFAP-IL10Tg in the three areas (arrows in G–L). Remarkably, in the CA1 area of the hippocampus of GFAP-IL10Tg animals major part of microglial cells displayed acharacteristic elongated morphology (arrow in L). (M and N) Histograms showing the quantification of the AI values of Iba1 immunolabelling (M) and the area occupied bylabelled cells (N). (O and P) Histograms showing the shape factor (high values indicate round shape and low values ramified morphology) (O) and the elongation values (valueequal to 1 indicates round morphology and high values increased elongation) (P) calculated for individual cells. In all histograms WT (white bars) and GFAP-IL10Tg mice(black bars) in the cerebral cortex, cerebellum and hippocampus (#p 6 0.15; ⁄p 6 0.05; ⁄⁄p 6 0.005). Scale bar (A–F) = 50 lm; (G–L) = 10 lm. (CA: cornus ammonis, HF:hippocampal fissure, DG: dentate gyrus).



3.6. Transgene-encoded IL-10 induces a specific microglial cellphenotype

Characterisation of microglial cells in WT and GFAP-IL10Tg ani-mals was assessed by immunohistochemistry, evaluating their dis-tribution and morphology with Iba1, their cell density with Pu.1 andtheir phenotype with different markers of microglial activationsuch as CD11b, CD16/32, F4/80, CD150, CD206 and MHCII. In accor-dance with the RPA results, the cerebral cortex, cerebellum and hip-pocampus were the three brain areas where the major differencesbetween WT and GFAP-IL10Tg animals were detected; hence theremaining studies were focused on these specific brain regions.

3.6.1. Morphology and density of microglial cellsThe study of sections immunolabelled for Iba1 revealed signifi-

cant changes between GFAP-IL10Tg and WT animals in the expres-sion levels of this molecule as well as in the morphology ofmicroglial cells (Fig. 5).

A significant increase in Iba1 labelling was found in GFAP-IL10Tganimals when compared with WT (Fig. 5M). While WT microglialcells had a characteristic ramified morphology and showed ahomogenous distribution in both the grey and the white matterareas (Fig. 5A–C, G, I and K), microglial cells in the cerebral cortex,the cerebellum and the hippocampus of GFAP-IL10Tg mice dis-played an enlargement of the cell body and had thicker processes(Fig. 5D–F, H, J and L). Noticeably, in the CA1 region of the hippo-campus of transgenic animals, microglial cells displayed a distinc-tive elongated morphology following the same direction as themain dendritic trees of pyramidal neurons of this region (Fig. 5L).These qualitative morphological changes were quantified by a mor-phometric analysis of different parameters such as the area, theshape factor and the elongation of individual Iba1+ microglial cells.Confirming the qualitative findings, GFAP-IL10Tg animals showed a

Fig. 6. Pu.1 immunohistochemistry. Representative images showing the number of Pu.1both WT (A–C) and GFAP-IL10Tg animals (D–F). (G) Histogram showing the quantificationIL10Tg animals (black bars) was almost twice than in WT (white bars) (⁄p 6 0.05; ⁄⁄p 6

significant increase in the area occupied by individual microglialcells in the cerebral cortex and hippocampus (Fig. 5N). Moreover,microglial cells in transgenic animals showed a less ramified mor-phology, as indicated by higher values of shape factor (Fig. 5O),especially in the cerebellum and hippocampus and an increase inthe elongation of individual cells only in the hippocampus (Fig. 5P).

Concomitant with the morphological differences of microglialcells, a significant increase in the density of Pu.1+ microglial/mac-rophages cells was detected in the cerebral cortex, cerebellum andhippocampus of GFAP-IL10Tg mice (Fig. 6). In the three brainregions, the number of Pu.1+ cells in GFAP-IL10Tg animals wasnearly twice that in WT mice (Fig. 6G). Analysis of pH3, a prolifer-ation marker, denoted no differences in the number of proliferatingmicroglia between WT and GFAP-IL10Tg in any of the areas studied(data not shown).

3.6.2. Phenotypic analysis of microglial cellsTo better characterise the microglial phenotype induced by the

transgenic overproduction of IL-10, different markers commonlyrelated to the activation state of microglia such as CD11b, CD16/32, F4/80, CD150, CD206 and MHCII were analysed by single anddouble immunohistochemistry.

Microglial cells of WT animals exhibited low CD11b (Fig. 7A–D)and very low CD16/32 (Fig. 7H–K) and F4/80 (Fig. 7O–R) expres-sion in both the grey and white matters of the cerebral cortex, cer-ebellar cortex and hippocampus, while CD150 (Fig. 8A–C), CD206(Fig. 8G–I) and MHCII (Fig. 8M–O) were undetectable in WTmicroglial cells in the three brain areas studied.

Notably, when compared with WT mice, in GFAP-IL10Tg ani-mals there was a significant increase in CD11b (Fig. 7A, E–G),CD16/32 (Fig. 7H, L–N) and F4/80 expression (Fig. 7O, S–U) andinduction of CD150 (Fig. 8D–F) in microglial cells of the cerebralcortex, cerebellum and hippocampus. However, similar to the

in the cerebral cortex (A and D), cerebellum (B and E) and hippocampus (C and F) ofof Pu.1+ cell density in the three brain areas. Note that the number of cells in GFAP-

0.005). Scale bar = 30 lm.


observations in WT animals, microglial cells from transgenic micedid not have detectable CD206 (Fig. 8J–L) or MHCII (Fig. 8P–R).

In conclusion, this immunohistochemical study revealed dra-matic changes in microglial morphology, increase in cell densityand the acquisition of a specific phenotype in the cerebral cortex,cerebellar cortex and hippocampus of GFAP-IL10Tg animals.

3.7. Transgene-encoded IL-10 is associated with changes in astrocytesin the hippocampus

To extend our analysis of the effects of local IL-10 production inCNS to other glial cell populations, we studied the potential effectsof transgenic production of this cytokine on the astrocyte popula-tion. Confirming the RPA results above, the analysis of GFAP immu-nostained sections only revealed a significant increase in the GFAPexpression in the hippocampus of GFAP-IL10Tg animals when com-pared with WT mice (Fig. 9). This increase in GFAP was attributed tothe higher number of ramifications observed in transgenic animals.

3.8. GFAP-IL10Tg animals have lower excitability of CA1-CA3 synapseand absence of LTP

Due to the remarkable changes observed in the morphology andphenotype of microglial cells of GFAP-IL10Tg animals, especially inthe hippocampus, the next question we addressed was to deter-mine whether or not these microglial modifications exerted anyeffect on hippocampal synaptic transmission.

The spectral analysis of local field potentials recorded in thehippocampal CA1 area when the animals were located in the Skin-ner box (with lever and feeder removed) did not show any signif-icant differences between WT and GFAP-IL10Tg animals for thedifferent (delta, theta, alpha, beta, gamma) frequency bands(Fig. 10C and D).

In a first step, we studied the input/output curves evoked at theCA3-CA1 synapse in the two groups of animals. Both WT and GFAP-IL10Tg mice had similar increases in the slope of fEPSPs evoked atthe CA1 area following the presentation of single pulses of increas-ing intensity to the ipsilateral Schaffer’s collaterals (Fig. 11A).Although in both the cases, these relationships were fitted by sig-moid curves for both genotypes, WT mice had steeper values thanGFAP-IL10Tg animals, mostly at the middle range of stimulusintensities. These results were indicative of a decreased excitabilityof hippocampal synapses in GFAP-IL10Tg animals with respect totheir littermate controls.

Paired-pulse stimulation is a form of short-term synaptic mod-ulation used as an indirect measurement of changes in the proba-bility of neurotransmitter release at the presynaptic terminal(Madronal et al., 2009). As illustrated in Fig. 11B, both WT andGFAP-IL10Tg animals showed a significant increase of the responseto the 2nd pulse at short (10–100 ms) time intervals. Although val-ues collected from GFAP-IL10Tg mice were slightly lower, no sig-nificant differences between the two experimental groups wereobserved at any of the selected intervals. In accordance, the prob-ability of neurotransmitter release in glutamatergic hippocampalsynapses was not modified in GFAP-IL10Tg mice.

As shown recently, LTP is indicative of the activity-dependentplasticity of hippocampal circuits (Gruart et al., 2006). As illus-trated in Fig. 11C, WT animals presented a significant LTP for thefirst recording session [F(20, 47) = 2.694; p 6 0.001]. In contrast,GFAP-IL10Tg mice had smaller LTP values that were significantlydifferent [F(5, 20) = 10.175; p 6 0.03] than those presented by WT.These results are indicative of diminished plastic properties atthe CA3-CA1 synapse of GFAP-IL10Tg mice (Fig. 11C).

In conclusion, the electrophysiological studies indicated thatGFAP-IL10Tg mice have a decreased excitability of hippocampal

circuits as well as a limited capability to experimentally evokedchanges in synaptic strength.

3.9. GFAP-IL10Tg animals show a lower rate of exploratory behaviourbut no changes in learning

Finally, to extend our study to the potential effects of transgenicIL-10 production on the function of the hippocampus, we also per-formed different behavioural tests to evaluate different tasksinvolving the hippocampus, such as spatial memory and associa-tive learning.

In a first behavioural test, we analysed the performance of bothgroups of mice in an open field. As shown in Fig. 12A, GFAP-IL10Tganimals remained for longer periods of time in the centre of thearena, although they showed a significantly lower mobility in theopen field test than WT animals (Fig. 12B).

We also checked putative differences between the two groupsof mice with the passive avoidance test (Fig. 12C). During the firsttraining session (training trial), there was no difference betweengroups in the time spent before entering the dark compartment.For the retention sessions both groups of animals showed signifi-cant differences when compared with their corresponding trainingtrial. Although GFAP-IL10Tg mice had a slightly larger step-through latency than their WT littermates, the differences didnot reached statistical significance (Fig. 12C).

Animals were also trained in a Skinner box to obtain a food pel-let every time they press a lever located nearby the feeder, using afixed ratio (1:1) schedule (Fig. 12D upper diagram). As shown inFig. 12E and F, WT mice acquired the fixed ratio (1:1) schedule atthe same rate [F(4, 43) = 21.756; p 6 0.001] than GFAP-IL10Tg ani-mals [F(4, 54) = 35.240; p 6 0.001]. Animals from the two groupsthat reached criterion for the fixed ratio task in less than 7 dayswhere further trained in a more complex situation, where animalswere rewarded only during the period in which a small light bulb,located over the lever, was switched on (Fig. 12D lower diagram).As illustrated in Fig. 12G, WT mice acquired this complex task atthe same rate as the GFAP-IL10Tg group.

In conclusion, GFAP-IL10Tg animals had a larger prevention tomove into open spaces, a lower spontaneous mobility than theirWT littermates and a slightly (although non-significant) largerretention memory during the passive avoidance test. Interestinglyenough, no significant differences were observed in the selectedassociative learning tasks.

4. Discussion

Our observations showed that astrocyte-targeted production ofIL-10 has a significant impact, under basal conditions, on bothmicroglial cells and to, a lesser extent, astrocytes in regions oftransgene expression, i.e., hippocampus, cerebral cortex and cere-bellum. In these regions, GFAP-IL10Tg microglial cells showedchanges in morphology and displayed a particular activated phe-notype characterised by upregulation of Iba1, CD11b, CD16/32and F4/80, de novo expression of CD150 although no expressionof either CD206 or MHCII. Astrocytes in the hippocampus of trans-genic mice showed an increase in GFAP. These alterations inmicroglial cells and astrocytes were associated with electrophysio-logical modifications in the hippocampal synaptic transmission oftransgenic animals, characterised by a lower excitability of the hip-pocampal CA3-CA1 synapse and absence of LTP evoked by HFS ofthe Schaffer’s collateral projections onto the CA1 area. Moreover,transgenic mice showed a lower exploratory behaviour in the openfield than their WT littermates, but no differences in either the pas-sive avoidance or the operant conditioning tasks.

Fig. 7. CD11b, CD16/32 and F4/80 immunohistochemistry. Quantification of the AI values for CD11b (A), CD16/32 (H) and F4/80 (O) markers shows an important andstatistically significant increase of the three markers in microglial cells of GFAP-IL10Tg animals (black bars) (⁄p 6 0.05; ⁄⁄p 6 0.005; ⁄⁄⁄p 6 0.0005). Representative images ofdouble-immunofluorescence staining of Iba1 (shown in green) combined with CD11b (B–G), CD16/32 (I–N) and F4/80 (P–U) (shown in red) in the CA1 region of thehippocampus of both WT (B–D, I–K and P–R) and GFAP-IL10Tg animals (E–G, L–N and S–U). Yellow–orange staining shows colocalization. Although expression of the threemarkers was observed in Iba1+ microglial cells of WT and GFAP-IL10 Tg animals, the expression levels were always higher in GFAP-IL10Tg mice. Note that some CD11b+/Iba1- cells (arrowheads in E–G) were found in transgenic animals. Blue is DAPI. Scale bar = 10 lm. (For interpretation of the references to colour in this figure legend, thereader is referred to the web version of this article.)


Fig. 8. CD150, CD206 and MHCII immunohistochemistry. Representative images showing the double immunofluorescence for Iba1 (shown in green) and CD150 (A–F), CD206(G–L) and MHCII (M–R) in the CA1 region of the hippocampus of both WT (A–C, G–I and M–O) and GFAP-IL10Tg mice (D–F, J–L and P–R). Yellow–orange staining showscolocalization. Expression of CD150, CD206 and MHCII was not observed in microglial cells of WT animals. Microglial cells of GFAP-IL10Tg animals only showed expression ofCD150 (arrows in D–F), whereas remained negative for CD206 and MHCII. Blue is DAPI. Scale bar = 10 lm. (For interpretation of the references to colour in this figure legend,the reader is referred to the web version of this article.)


Fig. 9. GFAP immunohistochemistry. (A–D) Representative images showing the GFAP staining in the hippocampus of WT (A–B) and GFAP-IL10Tg animals (C–D). Highmagnifications pictures are shown in B and D. Note the increase in GFAP expression observed in GFAP-IL10Tg animals as shown in the histogram (E) where AI is represented(⁄p 6 0.05). Scale bar (A and C) = 50 lm; (B and D) = 30 lm.

Fig. 10. Experimental design for the electrophysiological experiments. (A) Draft showing the specific area where animals were chronically implanted with stimulating (St.)electrodes on the Schaffer collaterals/commissural pathway and with recording (Rec.) electrodes in the hippocampal CA1 area. Abbreviations: DG, dentate gyrus; Schaffer coll.,Schaffer collaterals, PP, perforant pathway. (B) Representative photomicrographs illustrating the location (arrows) of recording (left) and stimulating (right) electrodes. Scalebar: 100 lm. (C and D) Spectral analysis of local field potentials recorded in the hippocampal CA1 area with the animals (WT, light line in C and grey bars in D; GFAP-IL10Tg,heavy line in C, and black bars in D) located in the Skinner box in the absence of the lever and the feeder. Note that no significant differences were observed between WT andtransgenic animals.


4.1. IL-10-induced microglial phenotype

Microglial precursors coming from the yolk salk invade the CNSduring the embryonic stages and subsequently proliferate, migrateand differentiate into ramified adult microglial cells (Dalmau et al.,

1998a,b, 2003). Microglia are classically considered as the residentmacrophages of the CNS (Gordon and Taylor, 2005; Ransohoff andCardona, 2010). Macrophage responses have been commonly clas-sified as M1 (classical activation) and M2 (alternative activation)responses, where the M1 response is driven by an exposure to

Fig. 11. Changes in long term potentiation in GFAP-IL10Tg mice compared to WT littermates. (A) Input/output curves (best sigmoidal adjustments) for the CA3-CA1 synapse.Note the larger fEPSP values reached by WT mice (black line) during the central part of the selected stimulus intensities in comparison to GFAP-IL10Tg mice (dashed line). (B)Paired-pulse facilitation. The data shown are mean ± SEM slopes of the second fEPSP expressed as a percentage of the first for six (10, 20, 40, 100, 200, 500 ms) inter-stimulusintervals. The two groups of mice (WT, black circles; GFAP-IL10Tg, white circles) presented similar paired-pulse facilitation at intervals of 20–100 ms. Representative fEPSPscollected from the two types of mouse are illustrated at the top. (C) The upper right recordings illustrate examples of fEPSPs collected from representative WT (black circles)and GFAP-IL10Tg (white circles) animals before (baseline, B) and after (days 1–4) HFS of Schaffer collaterals. The bottom graph illustrates the time course of LTP evoked in theCA1 area (fEPSP mean ± SEM) following HFS for WT and GFAP-IL10Tg mice. The HFS was presented after 15 min of baseline recordings, at the time marked by the dashed line.The fEPSP are given as a percentage of the baseline (100%) slope. Only WT mice presented a significant increase in fEPSP slopes following HFS when compared with baselinerecords. Values collected from the WT group were significantly (⁄p 6 0.05) larger than those collected from GFAP-IL10Tg mice.


microorganisms and characterised by increase in pro-inflamma-tory mediators, such as IL-1b, TNF-a and iNOS (Appel et al.,2010). Alternatively, the M2 response comprises at least three dif-ferent phenotypes (M2a, M2b and M2c) based upon their geneexpression profile and activation cues (Appel et al., 2010; Bocheet al., 2013; Gordon and Martinez, 2010). The M2a phenotype, isdriven by an exposure to IL-4 and IL-13 and involves the upregula-tion of factors such as mannose-receptor, arginase-1, YM1 andFizz1; the M2b phenotype, directed by exposure to immunocom-plexes and Toll-like receptors, induces an increase in MHCII,CD80 and CD86 expression; and the M2-deactivated or wound-healing phenotype (M2c), induced by IL-10 or TGF-b, that inducedthe increase in molecules such as mannose-receptor, scavengerreceptors A and B, CCR2, TGF-b, CD14 and CD150 (Boche et al.,2013; Gordon and Martinez, 2010; Mantovani et al., 2004).

Microglial activation is a highly plastic process strongly influ-enced by the microenvironment in which they are located(Carson et al., 2007; Kettenmann et al., 2011; Ransohoff andPerry, 2009). Therefore, it is largely assumed that activated microg-lial cells may display different morphological and functional phe-notypes and express M1 and M2 markers (Kettenmann et al.,2011; Prinz et al., 2014; Ransohoff and Cardona, 2010). Remark-ably, in our study we found a specific and distinctive Iba1high,CD11bhigh, CD16/32high, F4/80high, CD150+, CD206- and MHCII-phenotype in the microglia of GFAP-IL10Tg animals. This pheno-type, on one hand, exhibited markers commonly associated with

the M2c microglia/macrophage phenotype such as CD150, a glyco-protein receptor mediating the signalling between T-cells and anti-gen-presenting cells (Howie et al., 2002), but on another handshowed an increase in markers usually related with the M1-pheno-type, such as CD16/32 and F4/80. Furthermore, in our study we didnot observe microglial expression of other expected M2c-associ-ated molecules, such as the mannose-receptor (CD206), a C-typelectin involved in the endocytosis of certain glycoproteins, patho-gen recognition and antigen presentation (Taylor, 2001). Ourresults are in agreement with an increasing number of studiesshowing that microglia are able to express M1 and M2 markersat the same time (Crain et al., 2013; Olah et al., 2012) and rein-forced the idea, advocated by some authors (Biber et al., 2014;Shechter and Schwartz, 2013), that microglia are very versatileand plastic cells with a broad spectrum of functions and pheno-types that cannot be just classified as a M1 or a M2 phenotype.

Iba1, CD11b, CD16/32 and F4/80 are molecules constitutivelyexpressed by microglial cells and whose upregulation has beencommonly associated with the activation of these cells under neur-oinflammatory conditions (Kettenmann et al., 2011). Therefore,this specific IL10-induced microglial phenotype found in GFAP-IL10Tg animals, with high expression of Iba1, CD11b, CD16/32and F4/80, can be considered as a particular activated state of theseglial cells comparable to other chronic states of activated microgliadescribed in some particular conditions. In the aged brain, forinstance, microglia display chronic morphological changes and

Fig. 12. Performance of WT and GFAP-IL10Tg mice in the open field, the passive avoidance test and during two operant conditioning tasks. (A) Time spent by the two groupsof animals (WT, grey bars; GFAP-IL10Tg, black bars) in the centre of the open field. Note that GFAP-IL10Tg mice spent more time in the centre of the arena than their WTlittermates (⁄p 6 0.05). (B) Total motor activity (in arbitrary units) of both groups of mice (WT, light grey line; GFAP-IL10Tg, black line) in the open field (⁄p 6 0.05). (C)Latency to enter the dark compartment (step-through latency) for WT and GFAP-IL10Tg groups was measured to a maximum of 300 s. Note that, following the training trial(white bars), the step-through (i.e., retention) latency was similar for the two groups (⁄p 6 0.05). (D) Schematic drafts showing the two different operant conditioning tasks inwhich the animals were evaluated: a fixed-ratio (1:1) schedule (top diagram) and a light/dark test; in this case, rewarding only when the light bulb was switched on (bottomdiagram). (E) Data collected from the first 5 days of training with a fixed ration (1:1) schedule. (F) Time (in days) spent by the two groups of mice to reach the selectedcriterion (obtaining 20 pellets/20 min session for two successive days). (G) Performance during the light/dark test. The light/dark coefficient was calculated as follows:(number of lever presses during the light period � number of lever presses during the dark period)/total number of lever presses). Note that there was no significantdifference in performance between WT and GFAP-IL10Tg animals in neither the fixed 1:1 ratio task nor the light/dark test.


increased levels of certain molecules such as Iba1, Toll-like recep-tors, CD11b and MHCII, as well as release of inflammatory media-tors like IL-1 b, TNF-a and IL-6 (Biber et al., 2014; Norden andGodbout, 2013). Usually, this chronically activated phenotype ofmicroglia in the healthy aged brain has been associated with a dis-tinctive capacity of microglial cells to respond to CNS lesions(Norden and Godbout, 2013). In this way, we think that it will bevery interesting to analyse, in future studies, how the IL-10-induced microglial phenotype responds to different types of CNSlesions and, if they respond differently, whether this microglialresponse influences the outcome of these lesions.

The altered microglial phenotype found in transgenic animalsmay be due to a direct effect of astrocyte-targeted IL-10 productionon microglia, since expression of IL-10R in these cells has beendescribed under both steady-state (Gonzalez et al., 2009) andinflammatory conditions (Hulshof et al., 2002; Strle et al., 2002).Consistent with this hypothesis, downregulation of pro-inflamma-tory cytokine secretion has been demonstrated in LPS-activatedmicroglial cultures treated with IL-10 (Kremlev and Palmer,2005; Lodge and Sriram, 1996; Norden et al., 2014; Park et al.,2007). However, in our study, we did not detected IL-10R expres-sion in microglial cells in GFAP-IL10Tg animals, but rather, and inagreement with previous studies (Ledeboer et al., 2002; Nordenet al., 2014), we found expression in neurons and astrocytes. Theseresults suggest that changes found in microglial phenotype couldbe due to an indirect effect, at least in this paradigm, induced byIL-10 on neurons and/or astrocytes. Actually, in our study we foundan increase in GFAP expression in astrocytes of GFAP-IL10Tg

indicating an effect of IL-10 on the astroglial population. Differentstudies have reported the importance of molecules such as TGF- b,M-CSF and GM-CSF produced by astrocytes in the maintenance ofmicroglia in their ramified state (Abutbul et al., 2012; Schillinget al., 2001; Walker et al., 2014). Although the exact role exertedby IL-10 on the astrocyte function has not been extensively charac-terised, a recent in vitro study from Norden et al. has reported thatastrocytes treated with IL-10 reduced the secretion of the pro-inflammatory cytokine IL-1 b by LPS-activated microglia througha mechanism involving TGF- b secretion (Norden et al., 2014), sug-gesting that IL-10 is able to modify the expression of some mole-cules related to the astroglia-microglia communication.

4.2. IL-10 induces an increase in microglial cell number

Another interesting result of our study was that, in addition tomorphological changes, the density of microglial cells monitoredby Pu.1 expression increased in the three different CNS areas ana-lysed. Pu.1 is a member of the Ets transcription factor family foundpredominantly in mast cells, B cells, neutrophils, hematopoieticstem cells and macrophages (Moreau-Gachelin, 1994), includingmicroglia (Ellis et al., 2010; Walton et al., 2000). In microglial cells,Pu.1 regulates, important molecules related with the macrophage/microglial function such as CD11b, CD18 and FcRs (Pahl et al.,1993; Perez et al., 1994; Rosmarin et al., 1995). In agreement withthis, in our study we observed an increase in CD11b and CD16/32expression. Moreover, Pu.1 is also important for microglial devel-opment, as deficiency of Pu.1 leads to complete microglial absence


and early death of deficient animals (Beers et al., 2006). Althoughin normal situations the turn-over of microglial cells in the adultCNS is very low (Lawson et al., 1992), during development, oncemicroglial cell precursors enter the CNS through the meningesand blood vessels, they proliferate, transform into ramified cellsand a percentage of them die by an apoptotic mechanism(Dalmau et al., 2003; Harry, 2013; Navascues et al., 2000). As wedid not observe microglial proliferation differences between theadult WT and GFAP-IL10Tg animals, it is likely that the increasednumber of microglial cells in GFAP-IL10Tg animals may be due toincreased proliferation during development. However, based onsome in vitro studies demonstrating that IL-10 induces an increasein microglial survival (Strle et al., 2002) but has no effects onmicroglial proliferation (Sawada et al., 1999; Strle et al., 2002),an alternative explanation would be that IL-10 produced in trans-genic animals might increase microglial survival during postnataldevelopment resulting in a higher number of Pu.1+ cells in theadult GFAP-IL10Tg animals. Another possible explanation wouldbe that IL-10 might induce an increase in the number of microglialprecursors entering the CNS during development, leading to amajor number of ramified microglia in the adult. Further studiesto address all these questions are warranted.

4.3. Changes in synaptic transmission and exploratory behaviour

A notable finding of this study was that the specific microglialphenotype observed in transgenic animals and the changes inGFAP content in astrocytes was associated with functional modifi-cations in neuronal activity in the hippocampus. Our results indi-cated that GFAP-IL10Tg mice exhibited a lower excitability at thehippocampal CA3-CA1 synapse as well as a noticeable absence ofLTP evoked by HFS of Schaffer collateral projections onto the CA1area. Based on the fact that a low LTP is usually associated with

Fig. 13. Phenotype of microglial cells in WT and GFAP-IL10Tg animals. Principalcharacteristics regarding the specific phenotype of microglial cells in GFAP-IL10Tganimals in comparison to WT were outlined in this drawing. In addition tomorphological changes, microglial cells in transgenic animals showed an increasedexpression of Iba1, CD11b, CD16/32 and F4/80, de novo expression of CD150 but nodetectable (n.d) expression of either CD206 or MHCII. Moreover, an increase in thenumber of cells, stated by Pu.1 labelling, was found in this study. In the figure, theintensity of colour denotes intensity of AI whereas the rectangle length indicatesnumber of cells expressing each molecule. (For interpretation of the references tocolour in this figure legend, the reader is referred to the web version of this article.)

a decrease in the number of synapses or their functional capabili-ties, we expected a reduction in either the synaptic density or thefunction of these synapses in the hippocampus of transgenic ani-mals. However, no detectable changes were found in synapto-physin expression in GFAP-IL10Tg animals. If we discard changesin synaptic density, then putative modifications in synaptic trans-mission may be the cause of the altered LTP. In agreement, emerg-ing evidence pointed towards a role of microglia in the control ofneuronal function. It has been recently published that microglialprocesses continuously make transient physiological contacts withhighly active neurons, enwrap their soma and induces a reductionin the activity of the targeted neuron (Li et al., 2013, 2012). Simi-larly, in some pathological conditions such as facial nerve axotomy,microglial cells are able to enwrap and disconnect injured neuronsin the so-called phenomenon of synaptic stripping (Blinzinger andKreutzberg, 1968; Kreutzberg, 1996), reinforcing the idea of theessential role of microglia in the control of neuronal connectivity.

In conclusion, this study demonstrates that intrathecal expres-sion of IL-10 exerts an important influence over microglial pheno-type, inducing an increase in the number of microglial cellsshowing a particular activated phenotype (Fig. 13) in the cerebralcortex, cerebellar cortex and hippocampus. This specific IL-10-induced microglial phenotype associated with modifications inneuronal activity in the hippocampus leading to a decrease in theLTP responses, suggesting a role of microglial cells in the regulationof neuronal connectivity. In this context, it will be interesting infurther studies to analyse how this special IL-10-induced microgli-al phenotype impacts the microglial response to alterations in thehomeostasis of the CNS after experimental injury or induced dis-eases, and how this might alter the evolution of these neuropatho-logic states.

Acknowledgments

The authors would like to thank Miguel A. Martil, Isabella Appi-ah, B. Ferrer, J. Carrasco, Sue Ling Lim and María Sánchez-Enciso fortheir outstanding technical help. This work was supported by theSpanish Ministry of Science and Innovation (BFU2011-27400),the Spanish MINECO (BFU2011-29089 and BFU2011-29286) andJunta de Andalucía (BIO122, CVI 2487, and P07-CVI-02686),SAF2008-00435, SAF2011-23272 and an NHMRC project grant632754. The authors declare no competing financial interests.

References

Abutbul, S., Shapiro, J., Szaingurten-Solodkin, I., Levy, N., Carmy, Y., Baron, R., Jung,S., Monsonego, A., 2012. TGF-beta signaling through SMAD2/3 induces thequiescent microglial phenotype within the CNS environment. Glia 60, 1160–1171.

Almolda, B., Villacampa, N., Manders, P., Hidalgo, J., Campbell, I.L., Gonzalez, B.,Castellano, B., 2014. Effects of astrocyte-targeted production of interleukin-6 inthe mouse on the host response to nerve injury. Glia 62, 1142–1161.

Apelt, J., Schliebs, R., 2001. Beta-amyloid-induced glial expression of both pro- andanti-inflammatory cytokines in cerebral cortex of aged transgenic Tg2576 micewith Alzheimer plaque pathology. Brain Res. 894, 21–30.

Appel, S.H., Beers, D.R., Henkel, J.S., 2010. T cell-microglial dialogue in Parkinson’sdisease and amyotrophic lateral sclerosis: are we listening? Trends Immunol.31, 7–17.

Arimoto, T., Choi, D.Y., Lu, X., Liu, M., Nguyen, X.V., Zheng, N., Stewart, C.A., Kim,H.C., Bing, G., 2007. Interleukin-10 protects against inflammation-mediateddegeneration of dopaminergic neurons in substantia nigra. Neurobiol. Aging 28,894–906.

Bachis, A., Colangelo, A.M., Vicini, S., Doe, P.P., De Bernardi, M.A., Brooker, G.,Mocchetti, I., 2001. Interleukin-10 prevents glutamate-mediated cerebellargranule cell death by blocking caspase-3-like activity. J. Neurosci. 21, 3104–3112.

Balasingam, V., Yong, V.W., 1996. Attenuation of astroglial reactivity by interleukin-10. J. Neurosci. 16, 2945–2955.

Beers, D.R., Henkel, J.S., Xiao, Q., Zhao, W., Wang, J., Yen, A.A., Siklos, L., McKercher,S.R., Appel, S.H., 2006. Wild-type microglia extend survival in PU.1 knockoutmice with familial amyotrophic lateral sclerosis. Proc. Natl. Acad. Sci. U.S.A. 103,16021–16026.

http://refhub.elsevier.com/S0889-1591(14)00512-1/h0005




























Biber, K., Owens, T., Boddeke, E., 2014. What is microglia neurotoxicity (not)? Glia62, 841–854.

Blinzinger, K., Kreutzberg, G., 1968. Displacement of synaptic terminals fromregenerating motoneurons by microglial cells. Z. Zellforsch. Mikrosk. Anat. 85,145–157.

Bliss, T.V., Gardner-Medwin, A.R., 1973. Long-lasting potentiation of synaptictransmission in the dentate area of the unanaestetized rabbit followingstimulation of the perforant path. J. Physiol. 232, 357–374.

Boche, D., Perry, V.H., Nicoll, J.A.R., 2013. Review: activation patterns of microgliaand their identification in the human brain. Neuropathol. Appl. Neurobiol. 39,3–18.

Brewer, K.L., Bethea, J.R., Yezierski, R.P., 1999. Neuroprotective effects ofinterleukin-10 following excitotoxic spinal cord injury. Exp. Neurol. 159, 484–493.

Campbell, I.L., Hobbs, M.V., Kemper, P., Oldstone, M.B., 1994. Cerebral expression ofmultiple cytokine genes in mice with lymphocytic choriomeningitis. J.Immunol. 152, 716–723.

Cannella, B., Gao, Y.L., Brosnan, C., Raine, C.S., 1996. IL-10 fails to abrogateexperimental autoimmune encephalomyelitis. J. Neurosci. Res. 45, 735–746.

Cannella, B., Raine, C.S., 2004. Multiple sclerosis: cytokine receptors onoligodendrocytes predict innate regulation. Ann. Neurol. 55, 46–57.

Carson, M.J., Bilousova, T.V., Puntambekar, S.S., Melchior, B., Doose, J.M., Ethell, I.M.,2007. A rose by any other name? The potential consequences of microglialheterogeneity during CNS health and disease. Neurotherapeutics 4, 571–579.

Couper, K.N., Blount, D.G., Riley, E.M., 2008. IL-10: the master regulator of immunityto infection. J. Immunol. 180, 5771–5777.

Crain, J.M., Nikodemova, M., Watters, J.J., 2013. Microglia express distinct M1 andM2 phenotypic markers in the postnatal and adult central nervous system inmale and female mice. J. Neurosci. Res. 91, 1143–1151.

Cua, D.J., Hutchins, B., LaFace, D.M., Stohlman, S.A., Coffman, R.L., 2001. Centralnervous system expression of IL-10 inhibits autoimmune encephalomyelitis. J.Immunol. 166, 602–608.

Dalmau, I., Finsen, B., Zimmer, J., Gonzalez, B., Castellano, B., 1998a. Development ofmicroglia in the postnatal rat hippocampus. Hippocampus 8, 458–474.

Dalmau, I., Vela, J.M., Gonzalez, B., Castellano, B., 1998b. Expression of purinemetabolism-related enzymes by microglial cells in the developing rat brain. J.Comp. Neurol. 398, 333–346.

Dalmau, I., Vela, J.M., Gonzalez, B., Finsen, B., Castellano, B., 2003. Dynamics ofmicroglia in the developing rat brain. J. Comp. Neurol. 458, 144–157.

Eleore, L., Lopez-Ramos, J.C., Yi, P.J., Delgado-Garcia, J.M., 2007. The cognitiveenhancer T-588 partially compensates the motor associative learningimpairments induced by scopolamine injection in mice. Behav. Neurosci. 121,1203–1214.

Ellis, S.L., Gysbers, V., Manders, P.M., Li, W., Hofer, M.J., Muller, M., Campbell, I.L.,2010. The cell-specific induction of CXC chemokine ligand 9 mediated by IFN-gamma in microglia of the central nervous system is determined by the myeloidtranscription factor PU.1. J. Immunol. 185, 1864–1877.

Gonzalez, P., Burgaya, F., Acarin, L., Peluffo, H., Castellano, B., Gonzalez, B., 2009.Interleukin-10 and interleukin-10 receptor-I are upregulated in glial cells afteran excitotoxic injury to the postnatal rat brain. J. Neuropathol. Exp. Neurol. 68,391–403.

Gordon, S., Martinez, F.O., 2010. Alternative activation of macrophages: mechanismand functions. Immunity 32, 593–604.

Gordon, S., Taylor, P.R., 2005. Monocyte and macrophage heterogeneity. Nat. Rev.Immunol. 5, 953–964.

Gruart, A., Munoz, M.D., Delgado-Garcia, J.M., 2006. Involvement of the CA3-CA1synapse in the acquisition of associative learning in behaving mice. J. Neurosci.26, 1077–1087.

Gureviciene, I., Ikonen, S., Gurevicius, K., Sarkaki, A., van Groen, T., Pussinen, R.,Ylinen, A., Tanila, H., 2004. Normal induction but accelerated decay of LTP inAPP + PS1 transgenic mice. Neurobiol. Dis. 15, 188–195.

Harry, G.J., 2013. Microglia during development and aging. Pharmacol. Ther. 139,313–326.

Howie, D., Okamoto, S., Rietdijk, S., Clarke, K., Wang, N., Gullo, C., Bruggeman, J.P.,Manning, S., Coyle, A.J., Greenfield, E., Kuchroo, V., Terhorst, C., 2002. The role ofSAP in murine CD150 (SLAM)-mediated T-cell proliferation and interferongamma production. Blood 100, 2899–2907.

Hulshof, S., Montagne, L., De Groot, C.J., Van Der Valk, P., 2002. Cellular localizationand expression patterns of interleukin-10, interleukin-4, and their receptors inmultiple sclerosis lesions. Glia 38, 24–35.

Issazadeh, S., Ljungdahl, A., Hojeberg, B., Mustafa, M., Olsson, T., 1995. Cytokineproduction in the central nervous system of Lewis rats with experimentalautoimmune encephalomyelitis: dynamics of mRNA expression for interleukin-10, interleukin-12, cytolysin, tumor necrosis factor alpha and tumor necrosisfactor beta. J. Neuroimmunol. 61, 205–212.

Jurado-Parras, M.T., Gruart, A., Delgado-Garcia, J.M., 2012. Observational learning inmice can be prevented by medial prefrontal cortex stimulation and enhancedby nucleus accumbens stimulation. Learn. Memory 19, 99–106.

Kamm, K., Vanderkolk, W., Lawrence, C., Jonker, M., Davis, A.T., 2006. The effect oftraumatic brain injury upon the concentration and expression of interleukin-1beta and interleukin-10 in the rat. J. Trauma 60, 152–157.

Kettenmann, H., Hanisch, U.K., Noda, M., Verkhratsky, A., 2011. Physiology ofmicroglia. Physiol. Rev. 91, 461–553.

Knoblach, S.M., Faden, A.I., 1998. Interleukin-10 improves outcome and altersproinflammatory cytokine expression after experimental traumatic braininjury. Exp. Neurol. 153, 143–151.

Kremlev, S.G., Palmer, C., 2005. Interleukin-10 inhibits endotoxin-induced pro-inflammatory cytokines in microglial cell cultures. J. Neuroimmunol. 162, 71–80.

Kreutzberg, G.W., 1996. Microglia: a sensor for pathological events in the CNS.Trends Neurosci. 19, 312–318.

Lawson, L.J., Perry, V.H., Gordon, S., 1992. Turnover of resident microglia in thenormal adult mouse brain. Neuroscience 48, 405–415.

Ledeboer, A., Breve, J.J., Poole, S., Tilders, F.J., Van Dam, A.M., 2000. Interleukin-10,interleukin-4, and transforming growth factor-beta differentially regulatelipopolysaccharide-induced production of pro-inflammatory cytokines andnitric oxide in co-cultures of rat astroglial and microglial cells. Glia 30, 134–142.

Ledeboer, A., Breve, J.J., Wierinckx, A., van der Jagt, S., Bristow, A.F., Leysen, J.E.,Tilders, F.J., Van Dam, A.M., 2002. Expression and regulation of interleukin-10and interleukin-10 receptor in rat astroglial and microglial cells. Eur. J.Neurosci. 16, 1175–1185.

Ledeboer, A., Wierinckx, A., Bol, J.G., Floris, S., Renardel de Lavalette, C., De Vries,H.E., van den Berg, T.K., Dijkstra, C.D., Tilders, F.J., van dam, A.M., 2003. Regionaland temporal expression patterns of interleukin-10, interleukin-10 receptorand adhesion molecules in the rat spinal cord during chronic relapsing EAE. J.Neuroimmunol. 136, 94–103.

Li, Y., Du, X.F., Du, J.L., 2013. Resting microglia respond to and regulate neuronalactivity in vivo. Commun. Integr. Biol. 6, e24493.

Li, Y., Du, X.F., Liu, C.S., Wen, Z.L., Du, J.L., 2012. Reciprocal regulation betweenresting microglial dynamics and neuronal activity in vivo. Dev. Cell 23, 1189–1202.

Lim, S.H., Park, E., You, B., Jung, Y., Park, A.R., Park, S.G., Lee, J.R., 2013. Neuronalsynapse formation induced by microglia and interleukin 10. PLoS ONE 8,e81218.

Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data usingreal-time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods 25,402–408.

Lodge, P.A., Sriram, S., 1996. Regulation of microglial activation by TGF-beta, IL-10,and CSF-1. J. Leukoc. Biol. 60, 502–508.

Madronal, N., Gruart, A., Delgado-Garcia, J.M., 2009. Differing presynapticcontributions to LTP and associative learning in behaving mice. Front. Behav.Neurosci. 3, 7.

Madronal, N., Lopez-Aracil, C., Rangel, A., del Rio, J.A., Delgado-Garcia, J.M., Gruart,A., 2010. Effects of enriched physical and social environments on motorperformance, associative learning, and hippocampal neurogenesis in mice. PLoSONE 5, e11130.

Mantovani, A., Sica, A., Sozzani, S., Allavena, P., Vecchi, A., Locati, M., 2004. Thechemokine system in diverse forms of macrophage activation and polarization.Trends Immunol. 25, 677–686.

Molina-Holgado, E., Vela, J.M., Arevalo-Martin, A., Guaza, C., 2001a. LPS/IFN-gammacytotoxicity in oligodendroglial cells: role of nitric oxide and protection by theanti-inflammatory cytokine IL-10. Eur. J. Neurosci. 13, 493–502.

Molina-Holgado, F., Grencis, R., Rothwell, N.J., 2001b. Actions of exogenous andendogenous IL-10 on glial responses to bacterial LPS/cytokines. Glia 33, 97–106.

Moore, K.W., de Waal Malefyt, R., Coffman, R.L., O’Garra, A., 2001. Interleukin-10and the interleukin-10 receptor. Annu. Rev. Immunol. 19, 683–765.

Moreau-Gachelin, F., 1994. Spi-1/PU.1: an oncogene of the Ets family. Biochim.Biophys. Acta 1198, 149–163.

Navascues, J., Calvente, R., Marin-Teva, J.L., Cuadros, M.A., 2000. Entry, dispersionand differentiation of microglia in the developing central nervous system. An.Acad. Bras. Cienc. 72, 91–102.

Norden, D.M., Fenn, A.M., Dugan, A., Godbout, J.P., 2014. TGFbeta produced by IL-10redirected astrocytes attenuates microglial activation. Glia 62, 881–895.

Norden, D.M., Godbout, J.P., 2013. Review: microglia of the aged brain: primed to beactivated and resistant to regulation. Neuropathol. Appl. Neurobiol. 39, 19–34.

Olah, M., Amor, S., Brouwer, N., Vinet, J., Eggen, B., Biber, K., Boddeke, H.W.G.M.,2012. Identification of a microglia phenotype supportive of remyelination. Glia60, 306–321.

Ooboshi, H., Ibayashi, S., Shichita, T., Kumai, Y., Takada, J., Ago, T., Arakawa, S.,Sugimori, H., Kamouchi, M., Kitazono, T., Iida, M., 2005. Postischemic genetransfer of interleukin-10 protects against both focal and global brain ischemia.Circulation 111, 913–919.

Pahl, H.L., Scheibe, R.J., Zhang, D.E., Chen, H.M., Galson, D.L., Maki, R.A., Tenen, D.G.,1993. The proto-oncogene PU.1 regulates expression of the myeloid-specificCD11b promoter. J. Biol. Chem. 268, 5014–5020.

Pang, Y., Rodts-Palenik, S., Cai, Z., Bennett, W.A., Rhodes, P.G., 2005. Suppression ofglial activation is involved in the protection of IL-10 on maternal E. coli inducedneonatal white matter injury. Brain Res. Dev. Brain Res. 157, 141–149.

Park, K.W., Lee, H.G., Jin, B.K., Lee, Y.B., 2007. Interleukin-10 endogenouslyexpressed in microglia prevents lipopolysaccharide-inducedneurodegeneration in the rat cerebral cortex in vivo. Exp. Mol. Med. 39, 812–819.

Perez, C., Coeffier, E., Moreau-Gachelin, F., Wietzerbin, J., Benech, P.D., 1994.Involvement of the transcription factor PU.1/Spi-1 in myeloid cell-restrictedexpression of an interferon-inducible gene encoding the human high-affinity Fcgamma receptor. Mol. Cell. Biol. 14, 5023–5031.

Prinz, M., Tay, T.L., Wolf, Y., Jung, S., 2014. Microglia: unique and common featureswith other tissue macrophages. Acta Neuropathol. 128, 319–331.

Ransohoff, R.M., Cardona, A.E., 2010. The myeloid cells of the central nervoussystem parenchyma. Nature 468, 253–262.

Ransohoff, R.M., Perry, V.H., 2009. Microglial physiology: unique stimuli, specializedresponses. Annu. Rev. Immunol. 27, 119–145.














































































































































































Rosmarin, A.G., Caprio, D., Levy, R., Simkevich, C., 1995. CD18 (beta 2 leukocyteintegrin) promoter requires PU.1 transcription factor for myeloid activity. Proc.Natl. Acad. Sci. U.S.A. 92, 801–805.

Sawada, M., Suzumura, A., Hosoya, H., Marunouchi, T., Nagatsu, T., 1999.Interleukin-10 inhibits both production of cytokines and expression ofcytokine receptors in microglia. J. Neurochem. 72, 1466–1471.

Schilling, T., Nitsch, R., Heinemann, U., Haas, D., Eder, C., 2001. Astrocyte-releasedcytokines induce ramification and outward K+ channel expression in microgliavia distinct signalling pathways. Eur. J. Neurosci. 14, 463–473.

Shechter, R., Schwartz, M., 2013. Harnessing monocyte-derived macrophages tocontrol central nervous system pathologies: no longer ‘if’ but ‘how’. J. Pathol.229, 332–346.

Spera, P.A., Ellison, J.A., Feuerstein, G.Z., Barone, F.C., 1998. IL-10 reduces rat braininjury following focal stroke. Neurosci. Lett. 251, 189–192.

Stalder, A.K., Carson, M.J., Pagenstecher, A., Asensio, V.C., Kincaid, C., Benedict, M.,Powell, H.C., Masliah, E., Campbell, I.L., 1998. Late-onset chronic inflammatoryencephalopathy in immune-competent and severe combined immune-deficient(SCID) mice with astrocyte-targeted expression of tumor necrosis factor. Am. J.Pathol. 153, 767–783.

Strle, K., Zhou, J.H., Broussard, S.R., Venters, H.D., Johnson, R.W., Freund, G.G.,Dantzer, R., Kelley, K.W., 2002. IL-10 promotes survival of microglia withoutactivating Akt. J. Neuroimmunol. 122, 9–19.

Taylor, M.E., 2001. Structure and function of the macrophage mannose receptor.Results Probl. Cell Differ. 33, 105–121.

Walker, F.R., Beynon, S.B., Jones, K.A., Zhao, Z., Kongsui, R., Cairns, M., Nilsson, M.,2014. Dynamic structural remodelling of microglia in health and disease: areview of the models, the signals and the mechanisms. Brain Behav. Immun. 37,1–14.

Walton, M.R., Gibbons, H., MacGibbon, G.A., Sirimanne, E., Saura, J., Gluckman, P.D.,Dragunow, M., 2000. PU.1 expression in microglia. J. Neuroimmunol. 104, 109–115.

Xin, J.P., Wainwright, D.A., Mesnard, N.A., Serpe, C.J., Sanders, V.M., Jones, K.J., 2011.IL-10 within the CNS is necessary for CD4(+) T cells to mediate neuroprotection.Brain Behav. Immun. 25, 820–829.

Zhai, Q.H., Futrell, N., Chen, F.J., 1997. Gene expression of IL-10 in relationship toTNF-alpha, IL-1beta and IL-2 in the rat brain following middle cerebral arteryocclusion. J. Neurol. Sci. 152, 119–124.






































alterations in microglial phenotype and hippocampal neuronal...

Documents