Available online at www.sciencedirect.com

www.elsevier.com/locate/brainres

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5

0006-8993/$ - see frohttp://dx.doi.org/10.

Abbreviations: AD

FJB, fluoro-Jade B;

chromatography; IL

SLPI, secretory leunCorrespondence

Fax: +1 206 884 1210nnCorresponding auE-mail addresses:

Research Report

Progranulin promotes activationof microglia/macrophage after pilocarpine-inducedstatus epilepticus

Shanshan Zhua,b,n, Chao Taia,c, Terri L. Petkaud, Si Zhanga, Chengyong Liaoa,Zhifang Donga,e, Wendy Wena, Qing Changa, Yu Tian Wanga,Brian A. MacVicara, Blair R. Leavittd, William Jiaa, Max S. Cynadera,nn

aBrain Research Centre, University of British Columbia, 2211 Wesbrook Mall, Vancouver, British Columbia, CanadabCenter for Integrative Brain Research, Seattle Children's Research Institute, Seattle, Washington, United StatescDepartment of Pharmacology, University of Washington, Seattle, Washington, United StatesdCentre for Molecular Medicine & Therapeutics, University of British Columbia, British Columbia, Vancouver, CanadaeChongqing Key Laboratory of Translational Medical Research in Cognitive Development and Learning and MemoryDisorders, Children's Hospital of Chongqing Medical University, Chongqing 400014, China

a r t i c l e i n f o

Article history:

Accepted 15 July 2013

Progranulin (PGRN) haploinsufficiency accounts for up to 10% of frontotemporal lobe dementia.

PGRN has also been implicated in neuroinflammation in acute and chronic neurological

Available online 22 July 2013

Keywords:

Progranulin

Microglial/macrophage activation

Pilocarpine

Status epilepticus

Frontotemporal lobe dementia

nt matter Crown Copyri1016/j.brainres.2013.07.02

, Alzheimer's disease;

FTLD, frontotemporal l

-6, interleukin 6; KO, k

kocyte protease inhibitorto: Center for Integrative.thor. Fax: +1 604 822 0361shanshanzhupku@hotm

a b s t r a c t

disorders. Here we report that both protein and mRNA levels of cortical and hippocampal

PGRN are significantly enhanced following pilocarpine-induced status epilepticus. We also

identify intense PGRN immunoreactivity that colocalizes with CD11b in seizure-induced

animals, suggesting that PGRN elevation occurs primarily in activated microglia and macro-

phages. To test the role of PGRN in activation of microglia/macrophages, we apply recombinant

PGRN protein directly into the hippocampal formation, and observe no change in the number of

CD11b+ microglia/macrophages in the dentate gyrus. However, with pilocarpine-induced status

epilepticus, PGRN application significantly increases the number of CD11b+ microglia/macro-

phages in the dentate gyrus, without affecting the extent of hilar cell death. In addition, the

number of CD11b+ microglia/macrophages induced by status epilepticus is not significantly

different between PGRN knockout mice and wildtype. Our findings suggest that status

epilepticus induces PGRN expression, and that PGRN potentiates but is not required for

seizure-induced microglia/macrophage activation.

Crown Copyright & 2013 Published by Elsevier B.V. All rights reserved.

ght & 2013 Published by Elsevier B.V. All rights reserved.3

ALS, amyotrophic lateral sclerosis; BSA, bovine serum albumin; DIV, days in vitro;

obe dementia; GFAP, glial fibrillary acidic protein; HPLC, high performance liquid

nockout; PGRN, progranulin; LPS, lipopolysaccharide; SE, status epilepticus;

; WT, wildtypeBrain Research, Seattle Children's Research Institute, 1900 9th Ave, Seattle, WA 98101, USA.

.ail.com (S. Zhu), cynader@brain.ubc.ca (M.S. Cynader).

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 55

1. Introduction

More than 60 different null mutations of the progranulin gene(PGRN) have been identified in patients with frontotemporal lobedementia (FTLD) (Sun and Eriksen, 2011) and PGRN haploinsuffi-ciency appears to be a major cause of FTLD. The PGRN geneencodes a 68 KDa secretable protein that is widely expressed byneurons and microglia. Recent findings suggest critical roles forPGRN in the central nervous system. First, recombinant PGRNprotein promotes neuronal survival and neurite growth incortical and motor neuron cultures (Van Damme et al., 2008).Loss-of-function studies also support the neurotrophic role ofPGRN. Knocking down PGRN with siRNA increases the vulner-ability of cultured neuron to NMDA excitotoxicity (Guo et al.,2010). Cultured slices from PGRN knockout mice are moresensitive to lipopolysaccharide (LPS)-induced cell loss (Yinet al., 2010). In fact, PGRN knockout mice have a significantlyshorter lifespan (Ghoshal et al., 2012; Wils et al., 2012). Secondly,upregulation of PGRN has been detected in chronic diseases suchas Alzheimer's disease (AD) (Baker et al., 2006) and amyotrophiclateral sclerosis (ALS) (Philips et al., 2010), and has also beenreported following acute lesions such as sciatic nerve axotomy(Moisse et al., 2009), spinal cord contusion (Naphade et al., 2010),and the quinolinic acid-induced striatal lesion (Petkau et al.,2010). These findings suggest that PGRN may regulate patholo-gical responses in a variety of chronic and acute neurologicaldisorders.

The first aim of this study is to broaden our knowledge ofPGRN pathology in the acute response to status epilepticus (SE).SE causes acute injury of the hippocampus, accompanied bymicroglial activation and infiltration of macrophages (Fabeneet al., 2010). Using a well-characterized pilocarpine SE model,we have established the kinetics of PGRN induction. Secondly,in view of evidence showing FTLD-associated symptoms inPGRN-deficient mice (Ghoshal et al., 2012; Petkau et al., 2012;Wils et al., 2012; Yin et al., 2010), supplementing PGRN in vivomay become a valuable therapeutic strategy. Few studies haveaddressed the side effect of such a strategy. Thus we aim toinvestigate how recombinant PGRN might affect neuronal survi-val and activation of microglia/macrophages both in the normalbrain and in the pilocarpine SE model.

Our findings demonstrate the induction of PGRN expres-sion in microglia following pilocarpine-induced SE. Usingcannula-guided administration of PGRN protein, we foundthat PGRN promoted SE-induced microglial activation, but didnot affect SE-induced cell death in the dentate gyrus. Ourfindings indicate the upregulation of PGRN protein by epilep-tic insults, and reveal potential side effects for any PGRN-supplementing strategy.

2. Results

2.1. Induction of PGRN protein and mRNA afterpilocarpine-induced SE

Seizures have been shown to increase expression of numerousgrowth factors such as nerve growth factor (Holtzman andLowenstein, 1995), brain-derived neurotrophic factor (Rudge

et al., 1998), vascular endothelial growth factor (Nicoletti et al.,2008), and basic fibroblast growth factor (Riva et al., 1994; VanDer Wal et al., 1994). As a growth factor that plays a key role inthe central nervous system, PGRN may be regulated by seizureactivity. We tested this hypothesis in the rat pilocarpine modelof SE. In the cortex, the PGRN protein levels first remainedunchanged at 3 h or 12 h, elevated at 24 h, reached a maximallevel at 48 h, and persisted until 96 h after SE (2.070.1 fold at24 h, 4.670.7 fold at 48 h, and 3.270.2 fold at 96 h post-SEcompared with non-SE controls; Fig. 1B). Using real time PCR,we observed a similar induction pattern of PGRN mRNA levels(unaffected at 3 h or 12 h, 1.370.1 fold at 24 h, 2.470.2 fold at48 h and 2.370.2 fold at 96 h post-SE compared with non-SEcontrols; Fig. 1C). In the hippocampus, the mRNA and proteinlevels of PGRN were also enhanced by SE in the followingmanner: first insignificant at 3 h or 12 h, evident at 24 h(1.970.2 fold for protein and 1.570.1 fold for mRNA, comparedwith respective non-SE controls; Fig. 1B and C), further ele-vated at 48 h (3.170.2 fold for protein and 2.370.1 fold formRNA, compared with non-SE controls; Fig. 1B and C), andreaching the maximal levels at 96 h (4.570.3 fold for proteinand 4.670.2 fold for mRNA compared with non-SE controls;Fig. 1B and C). Our data indicated that PGRN expression wasmarkedly increased at 24–96 h following SE at both thetranscriptional and translational level.

2.2. Localization of PGRN in microglia/macrophagesafter SE

When PGRN was highly expressed in the brain at 48 h post-SE, reactive gliosis was also observed. We found increasedlevels of glial fibrillary acidic protein (GFAP), a marker forhypertrophic astrocytes, in the cortical and hippocampalbrain lysates at 48 h and 96 h after SE in rats (SupplementaryFig. 2A–2D). We also found greater numbers of CD11b+

microglia/macrophages at 48 h post-SE than in non-SE con-trols (Supplementary Fig. 2E). The Ox42 antibody recognizesCD11b expressed on the surface of microglia and numerousleukocytes including monocytes and macrophages(Kettenmann et al., 2011). Previous study has shown thatleukocytes infiltrate the brain in the pilocarpine model of SE(Fabene et al., 2010). Thus our study does not differentiatebetween activated microglia and infiltrating macrophages.

We next asked whether PGRN upregulation could beassociated with reactive gliosis using double immunohisto-chemistry. We found PGRN-positive cells that co-stained withCD11b antibody at 48 h post-SE in rats (Fig. 2A). PGRN+/CD11b+ cells significantly outnumbered those in non-SEcontrols by 4.9 fold in the hippocampus, 30.5 fold in thethalamus, and 31.5 fold in the cortex (Fig. 2D). However, thePGRN+ cells did not colocalize with GFAP+ cells at 48 h post-SE (Fig. 2C). Our results thus indicate that activated microgliabut not astrocytes strongly express PGRN after pilocarpine-induced SE.

2.3. PGRN protein increased CD11b+ microglia/macrophages after seizure induction

Another approach to study the association between PGRNand microglial activation is to apply PGRN protein in the

Fig. 1 – Induction of PGRN following SE. Total protein andmRNA were extracted from the cortex and hippocampus fromrats at 3, 12, 24, 48, 96 h after the onset of pilocarpine-inducedSE. A, Representative Western blots of PGRN and the actincontrol are shown. B, Densitometric measurements of PGRNimmunoblots in the panel A after normalization to actin. PGRNprotein levels in non-SE 48 h control groups (con) were set to1.0, and PGRN protein levels in SE groups were expressed asrelative fold change compared to controls. PGRN protein levelsfrom the cortex (white bars) and the hippocampus (black bars)were upregulated after the onset of SE at 24 h (n¼6), 48 h (n¼6),and 96 h (n¼5), but not at 3 h (n¼8) or 12 h (n¼5), comparedwith controls (n¼6). C, Quantitative real-time PCR assessmentof PGRN mRNA levels in the cortex (white bars) and thehippocampus (black bars). The mRNA levels of control groupswere set to 1.0, and mRNA levels of SE groups were calculatedas fold changes compared to controls. PGRN mRNA levels wereupregulated after the onset of SE at 24 h (n¼4), 48 h (n¼4), and96 h (n¼4), but not at 3 h (n¼6) or 12 h (n¼4), compared withnon-SE controls. npo0.05, nnpo0.01 and nnnpo0.001 comparedwith cortical controls using one-way ANOVA. ##po0.01 and###po0.001 compared with hippocampal controls using one-way ANOVA.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 556

pilocarpine model of SE. Human PGRN shows 75% similarityto rat PGRN protein, and exerts functional responses in rat asshown by many groups (Guo et al., 2010; Pickford et al., 2011;Tapia et al., 2011; Van Damme et al., 2008). We administeredhuman PGRN (100 ng) or bovine serum albumin (BSA, 100 ng)bilaterally into the hippocampal region in rats before seizureinduction. To avoid the effect of endogenously upregulatedPGRN on that of recombinant PGRN, we narrowed ourinspection window to no later than 24 h following SE, whenendogenous PGRN had not yet been induced (Fig. 1). PGRNincreased the number of CD11b+ microglia/macrophages,depending on the conditions of seizures (F(PGRN/BSA treatment)

¼57.80, po0.001; F(seizure)¼7.47, po0.01; F(treatment� seizure)¼6.06, po0.05; Fig. 3B). Application of PGRN had no effect onmicroglial activation in control groups without pilocarpinetreatment (p¼0.56 compared with BSA-control; Fig. 3). In SEgroups, CD11b+ cells on the BSA side were not evident at 6 hpost-SE (p¼0.60 between BSA-6 h SE and BSA-control; Fig. 3B),but were significantly increased at 12 h post-SE (1.7 fold overBSA-control; po0.05; Fig. 3B) and at 24 h post-SE (2.8 fold overBSA-control; po0.001; Fig. 3B). Interestingly, administrationof PGRN significantly activated microglia/macrophages at 6 h(2.7 fold over BSA-6 h SE; po0.01; Fig. 3B), and furtherenhanced the number of activated microglia/macrophagesat 12 h post-SE (2.0 fold over BSA-12 h SE; po0.01; Fig. 3B) andat 24 h post-SE (1.6 fold over BSA-24 h SE; po0.001; Fig. 3B).Our results thus suggest that administration of PGRN proteinpotentiates and accelerates pilocarpine-induced activation ofmicroglia/macrophages in vivo.

2.4. PGRN protein had no effect on dentate hilar cell deathafter pilocarpine administration

PGRN can act as a neurotrophic factor (Van Damme et al.,2008) and depletion of PGRN has been shown to rendercultured neurons more vulnerable to NMDA or H2O2 toxicity(Guo et al., 2010). Therefore we proceeded to test whetherrecombinant PGRN could protect against neuronal degenera-tion in the pilocarpine model of SE in rats. Since we injectedPGRN near the dentate gyrus (Fig. 7), where we observed cleareffects of PGRN on microglial activation, we measured celldeath in the polymorphic cell layer of dentate gyrus. Fluoro-Jade B (FJB)-positive degenerating cells were clearly increasedat 6 h, 12 h and 24 h after SE compared with controls(F(seizure)¼27.88, po0.001; Fig. 4). Treatment with PGRN led tothe same levels of cell death as with BSA at all time pointstested (F(PGRN/BSA treatment)¼0.20, p¼0.66; F(treatment� seizure)¼0.17,p¼0.92; Fig. 4B). Therefore, our results show that administrationof PGRN protein has no effect on neuronal death whilepotentiating activation of microglia/macrophages in the pilo-carpine SE model.

2.5. The density of CD11b+ microglia/macrophagesfollowing pilocarpine-induced SE was not different in PGRNdeficient mice

To determine whether PGRN was required for SE-inducedmicroglial activation, we induced seizures in PGRN�/� andPGRN+/+ mice with 320 mg/kg pilocarpine. At 48 h after SE,immunolabeling of PGRN was obvious in wildtype mice but

Fig. 2 – PGRN expression is upregulated in CD11b+ microglia after SE. A, Representative sections from hippocampus, cortexand thalamus of non-SE controls (ctrl) and of rats at 48 h after the onset of SE (SE) were double stained with anti-PGRN (red)and anti-CD11b (green). DAPI (blue) labeled the nucleus. Scale bar: 50 lm. B and C, High magnification images indicates clearupregulation of PGRN in CD11b-positive microglia (B) but not in GFAP-positive astrocytes (C). Scale bar: 20 lm. D, The numberof PGRN+ microglia per 0.4 square millimeter was quantified in indicated areas from the panel A (2 sections per rat, and n¼3rats per group). npo0.05 and nnpo0.01 compared with non-SE controls using Student's t test.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 57

was absent in PGRN knockout mice (Fig. 5A). However, therewas no significant difference in the density of CD11b+cellsbetween genotypes (Fig. 5B and C).

2.6. LPS and ATP induced PGRN expression in mixedneuronal-glial cultures

Next we attempted to address the factors contributing toPGRN expression. SE-induced brain damage has been asso-ciated with increased energy demand, imbalanced excitatoryand inhibitory signals, and inflammatory responses. Forexample, glutamate concentrations rise during seizures(Meldrum et al., 1999). The GABA receptor antagonist bicucul-line and the glutamate receptor agonist kainic acid have beenshown to induce generalized seizures (Ben-Ari et al., 1981).Extracellular ATP levels also increase with seizures(Wieraszko and Seyfried, 1989) and contribute to cytokineand chemokine release from glial cells (Bianco et al., 2005;Honda et al., 2001; Kucher and Neary, 2005; Panenka et al.,2001). LPS, a component of the Gram-negative bacterial wall,

induces the release of inflammatory factors that are alsoemitted after seizures (Vezzani, 2005). Therefore, we thentested whether glutamate, bicuculline, ATP, or LPS couldaffect PGRN expression in mixed rat neuronal-glial cultures.

We found that LPS significantly elevated PGRN proteinlevels (F(LPS dose)¼8.38, po0.01; Fig. 6A and B), while sucheffect remained the same for one-day or two-day treatment(F(time)¼1.57, p¼0.23; F(time�dose)¼0.24, p¼0.79; Fig. 6A and B).We also found a less significant effect of ATP on increasingPGRN levels, independent of the treatment time tested(F(ATP)¼6.87, po0.05; F(time)¼0.05, p¼0.83; F(time�ATP)¼0.38,p¼0.55; Fig. 6C and D). Treatment with glutamate (1 mM) orbicuculline (50 μM) for one or two days caused no significantdifference in PGRN protein expression (p40.05; Fig. 6C and D).

3. Discussion

Growth factors modulate seizure-induced reactive gliosis(Fabene et al., 2010; Holtzman and Lowenstein, 1995;

Fig. 3 – Administration of PGRN increased the number of microglia in animals with SE but not in controls. A, Sterilized PGRNprotein (100 ng, lower panels) or BSA (100 ng, upper panels) was injected into the contralateral hippocampus in all four groups(control, 6 h, 12 h, and 24 h). Thirty minutes later, SE was induced in rats with pilocarpine, while controls received vehicleinstead of pilocarpine and were processed 24 h later. Seizure sections were obtained at 6 h, 12 h, and 24 h after SE induction.Sections were immunostained with CD11b (green) for microglia/macrophages and DAPI (blue) for the nucleus. Scale bar:50 lm. B, The number of microglia/macrophages per 0.4 square millimeter was quantified in the dentate gyrus (n¼5 rats forcontrol, n¼3 rats for 6 h, n¼4 rats for 12 h, and n¼3 rats for 24 h). For each animal, cell counts were averaged from foursections spaced throughout the injection site (one image field per section). nnpo0.01 and nnpo0.001 after two-way repeatedmeasures ANOVA.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 558

Nicoletti et al., 2008; Riva et al., 1994; Rudge et al., 1998; VanDer Wal et al., 1994; Vezzani, 2005). Here we report increasedexpression of PGRN in activated microglia/macrophages inthe pilocarpine model of SE. We also show that application ofPGRN protein increases the density of CD11b+microglia/macrophages following pilocarpine-induced SE, withoutaffecting seizure-induced neuronal injury. PGRN deficiencyin mice does not affect SE-induced increase in the density ofCD11b+ microglia/macrophages. In addition, LPS and ATP, butnot glutamate or bicuculline, induces PGRN expression inmixed neuronal-glial cultures. Our data indicate that PGRNlevels are regulated by epileptic and inflammatory stimula-tion, and suggest a chemotaxic role for external but notendogenous PGRN protein during seizures.

Our data strongly support the notion that inflammationplays a key role in regulating PGRN protein expression.Firstly, activated microglia/macrophages are considered asneuroinflammation markers (Kettenmann et al., 2011) andthey express PGRN after SE (Fig. 2). Consistent with ourfindings, enhanced PGRN expression in activated microglia/macrophages are found in other neurodegenerative models

such as the Huntington's disease (Petkau et al., 2010), AD(Baker et al., 2006), and ALS models (Philips et al., 2010).Traumatic brain injury also induces PGRN expression inmicroglia (Tanaka et al., 2013). Secondly, LPS is a well-known inducer of inflammation (Vezzani, 2005) and ATPhas also been shown to activation microglia and astrocyte(Bianco et al., 2005; Honda et al., 2001; Kucher and Neary,2005; Panenka et al., 2001). Both of them increased PGRNexpression in mixed neuronal-glial cultures (Fig. 5), support-ing that inflammation could contribute to PGRN upregulation.

PGRN mRNA levels are also elevated following status epi-lepticus (Fig. 1), suggesting the transcriptional regulation ofPGRN gene expression. The PGRN promoter contains severalinflammation-related regulatory elements, such as the IL-6response elements (Bhandari et al., 1996). Increased levels ofIL-6 are detected after pilocarpine-induced SE (Holtman et al.,2013) and in the cerebrospinal fluid of seizure patients (Peltolaet al., 1998). IL6-dependent PGRN expression has been reportedin cholangiocarcinoma cell lines (Frampton et al., 2012). Thusinterleukin 6 (IL-6) might be a potential candidate that regulatesSE-induced PGRN gene transcription.

Fig. 4 – Injection of PGRN had no effect on cell death in thehilus of dentate gyrus. Sterilized PGRN (100 ng, 1 ll) or BSA(100 ng, 1 ll) was injected bilaterally into the hippocampusinto all control and SE groups in rats. Control animals (n¼6rats) received vehicle instead of pilocarpine. At 6 h (n¼5rats), 12 h (n¼5 rats), and 24 h (n¼3 rats) after SE initiation,the dentate gyrus sections near injection sites were stainedwith the Fluoro-Jade B (FJB). FJB+ cells in the delineatedpolymorphic cell layer of dentate gyrus were averaged fromthree sections per animal (one image field per section).A, Representative sections are shown. Scale bar: 100 lm.B, Degenerating cells are expressed as the number of FJB+

cells. nnpo0.01 and nnnpo0.001 compared with controls aftertwo-way repeated measures ANOVA.

Fig. 5 – Unaffected microglial activation by PGRN deficiencyafter pilocarpine administration. SE was induced in thewildtype mice (WT) and PGRN knockout mice (KO) withpilocarpine. A, Immunohistochemical staining of PGRN inthe dentate gyrus at 48 h after SE induction, showingsignificant deletion of PGRN in KO mice. Scale bar: 100 lm.B, At 48 h post-SE, sections from KO mice (n¼3) and WTmice (n¼3) were immunostained with anti-CD11b antibodyto label microglia. Cell counts were averaged from threeevenly spaced sections from approximately Bregma �2.8 to�3.8 mm in each animal (one image field per section). DAPIwas used to label nucleus. Representative sections from thedentate gyrus are shown. Scale bar: 50 lm. C, The numbersof CD11b+ microglial cells in the dentate gyrus from thepanel B are shown. p40.05 with the Student's t test.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 59

Post-translational modulation could also contribute to SE-stimulated PGRN expression. Neutrophil elastase and macro-phages elastase have been shown to cleave PGRN intogranulins, while the secretory leukocyte protease inhibitor(SLPI) blocks the cleavage (Suh et al., 2012; Zhu et al., 2002).A recent study reported that inflammatory factors stimulatedcultured astrocytes to release SLPI (Suh et al., 2012). Oncespecific antibodies recognizing granulins become available,it will be interesting to examine whether the processing ofprogranulin to granulins is regulated following pilocarpine-induced SE. In addition, Sortilin has been shown to bind toPGRN protein and mediate endocytosis of PGRN into neurons(Hu et al., 2010). Without Sortilin, PGRN accumulates in brain(Hu et al., 2010). Neurotensin competes with PGRN for thePGRN-binding domain on Sortilin (Zheng et al., 2011).

Neurotensin has been shown to increase in hippocampus ina seizure model (Shulkes et al., 1988). Thus neurotensinmight compete with PGRN for Sortilin binding, which mightattenuate PGRN endocytosis and increase PGRN levelsafter SE.

Fig. 6 – Differential roles of LPS, ATP, bicuculline, and glutamate in regulating PGRN expression. On DIV 3, we treated ratmixed neuronal-glial cultures with the indicated drugs for one or two days. Untreated cultures were used as controls. Valuesare shown as fold changes compared to those of one-day controls that were set to 1.0. A and B, PGRN and actin levels incultures treated with 100 ng/mL or 500 ng/mL LPS were measured by Western blots. Representative immunoblots are shownin the panel A and the pooled data are shown in the panel B. nnpo0.01 using two-way ANOVA. C and D, Representativeimmunoblots (C) and quantified results (D) of PGRN protein levels in cultures treated with ATP (50 lM), bicuculline (50 lM),and glutamate (1 mM). npo0.05 using two-way ANOVA. All values are averages of results from at least three separateexperiments.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 560

We found that PGRN protein administration potentiatedthe number of activated microglia/macrophages at 6, 12, and24 h after pilocarpine-induced SE, when endogenous PGRNhad not yet been induced. Surprisingly, PGRN protein alonehad no effect on microglial activation in vivo. This is contra-dictory to Pickford's report showing a chemoattractive role ofPGRN both in vitro and in vivo (Pickford et al., 2011).Comparing our studies reveals several interesting differences.Firstly, we injected the PGRN protein into the hippocampalformation while they targeted the cortex. Secondly, we usedpurified PGRN protein in our in vivo studies, while Pickford'sgroup used PGRN-expressing lentiviral vectors. Previousreports have shown that lentiviral proteins such as Tat orgp41 cause migration of microglia (Lokensgard et al., 2001;Speth et al., 2000). Indeed their control GFP-expressing lenti-viral vectors also induced PGRN expression and activatedmicroglia. Thirdly, we successfully repeated their in vitrofindings showing that PGRN protein promoted the migrationof purified cultured microglia (unpublished data). Note-worthy, cultured microglial cells are generally considered tobe in a more activated state than resident microglia in vivo.Therefore, the difference between our results can beexplained by: (1) region-specific impacts of PGRN on micro-glial migration; (2) implications that PGRN does not activateresident microglia but rather modulates microglial activationwhen they are in an activated state such as under stressinduced by seizures, lentiviral infection, or culture conditions.

Noteworthy, when interpreting our findings with intra-hippocampal PGRN administration, we should take a fewconcerns into account. Firstly, the injected PGRN might becleaved into granulins and what we observed following PGRN

administration could results from the mixed effects of bothPGRN and granulins. This is a hypothesis that we cannot testnow due to the lack of granulin antibodies. Secondly, we usedvehicle controls in Figs. 3 and 4, thus we cannot occlude thepossibility that the nontonic-clonic effects of pilocarpinemight also intervene with the effect of PGRN proteins onmicroglial activation.

Numerous studies of PGRN knockout mice and of PGRNmutations in FTLD-U patients have shown that reduced PGRNprotein levels progressively cause neural injury and subse-quent cognitive impairment (Matsuwaki et al., 2011; Petkauet al., 2012; van Swieten and Heutink, 2008). Few studies haveaddressed the biological responses to supplementing PGRNprotein in vivo, a strategy with potential clinical benefits.Our study clearly demonstrates that acute supplementationof PGRN protein in healthy brain, at least in the hippocam-pus, does not cause microglial migration or neural injury. Ourstudy shows that in the pilocarpine SE model PGRN poten-tiated microglial activation, while the extent of neural injuryremained unaffected. Our findings may be relevant to futurestudies using PGRN supplementation to treat dementiapatients with PGRN haploinsufficiency.

We also found that exogenous PGRN exerted no effect ondentate cell death while activating microglia/macrophages afterpilocarpine-induced SE. Activation of microglia/macrophages isconsidered to result from neuronal injury (Kettenmann et al.,2011) and we detected neuronal injury earlier than microglial/macrophage activation (Figs. 3 and 4). However, whetheractivation of microglia/macrophages causes neuronal injuryremains in doubt. It has been reported that minocycline-induced inhibition of microglial activation did not affect

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 61

SE-induced neuron degeneration (Yang et al., 2010). In an ALSmice model, ablation of proliferating microglia also did not alterthe time course of motor neuron death (Gowing et al., 2008). Inaddition, our group has recently reported that in PGRN�/� micethe extent of acute cell death in the kainic acid or pilocarpinemodel of SE were not affected (Petkau et al., 2013). Therefore,while PGRN protein activates microglia/macrophages, it couldeither be too late to activate a neuronal survival pathway orcould induce pathways that are irrelevant to neuronalprotection.

We have also shown that seizure-induced chemotaxis wasunaffected by PGRN deficiency using PGRN�/� mice. Similarnegative results have been reported in the PBS lesion modelusing PGRN�/� mice (Pickford et al., 2011). Together, weconclude that PGRN is a premise but is not required formicroglial migration in the pilocarpine SE model. Interest-ingly, both Pickford's and our studies were conducted in adult(2–3 month old) but not aged animals. Perhaps the negativeresults can be explained by an age-dependent function ofPGRN in the CNS, a hypothesis suggested by the lack ofdementia phenotypes in young adult PGRN knockout miceand in young FTLD patients with PGRN null mutations(Matsuwaki et al., 2011; Petkau et al., 2012; van Swieten andHeutink, 2008).

In conclusion, we demonstrate the regulation of PGRNlevels by epileptic and inflammatory insults, and suggest achemotaxic effect of supplementing PGRN protein duringseizures.

4. Experimental procedures

4.1. Ethical approval

We performed our animal experiments strictly following theguidelines of the Canadian Council on Animal Care. Allanimal studies were approved by the University of BritishColumbia Animal Care Committee. Appropriate measureswere taken during experimental procedures to minimize painand discomfort.

4.2. Cannula implantation

Adult male Sprague Dawley rats (200–250 g, Center for Dis-ease Modeling, UBC, Canada) were housed at room tempera-ture with a 12 h light/dark cycle. Surgeries were performedunder 2% isoflurane anesthesia. Rats were implanted withtwo 26-gauge stainless steel cannulae (PlasticsOne, Roanoke,VA, USA) bilaterally into the hippocampal region (3.4 mmposterior to Bregma, 1.5 mm lateral to the midline, and2.8 mm below the skull surface) as previously described(Chen et al., 2007). The cannulae were secured to the skullwith three stainless steel screws (Stoelting, Wood Dale, IL,USA). Rats were allowed to recover for at least 7 days beforeseizure induction.

4.3. Generation of PGRN-deficient mice

The generation of PGRN-deficient mice carrying a gene-trapallele (denoted ‘pFleo’ allele) has been previously described

(Petkau et al., 2010). Constitutive PGRN-knockout mice forthis study were generated by crossing pFleo mice to Flp-expressing mice (Jackson Laboratories, strain name: B6.Cg-Tg(ACTFLPe)9205Dym/J; Jax #005703) and subsequently to Cre-expressing mice (generous gift from E.Simpson, originallyfrom Jackson Laboratories, strain name: FVB/N-Tg(ACTB-cre)2Mrt/J, Jax #003376; then backcrossed for 5 generations toC57Bl/6 before being crossed to mice carrying the Flp-recombined allele). Successful Flp and Cre recombinationwas detected by PCR on tail tip DNA. After backcrossingheterozygous Flp- and Cre-recombined mice (denotedPGRN+/�) to C57BL/6 mice for 5 generations, we producedPGRN�/� and PGRN+/+ littermates for our experiments bycrossing two heterozygous parents.

4.4. The pilocarpine model of status epilepticus

Seizures were induced in Sprague Dawley rats, and in PGRN+/

+ and PGRN�/� mice by intraperitoneal (i.p.) injection of320 mg/kg pilocarpine as previously described (Goffin et al.,2007). Atropine methylbromide (5 mg/kg, i.p.; Sigma) wasadministered 30 min before pilocarpine to minimize periph-eral convulsions. Seizures were scored according to Racine'sstandards (Luttjohann et al., 2009; Racine, 1972): 0—noresponse; 1—eye blinking and mouth clonus; 2—head nod-ding with facial jerking; 3—brief and involuntary jerks of neckand forelimbs; 4—bilateral clonic convulsions with sittingand rearing; 5—clonic convulsions with rearing, falling andloss of body control. We determined the onset of SE by thetime point at which generalized seizures (class 4 and 5) werecontinuously observed in animals. Rats scored at 0–3 inresponse to pilocarpine were used as non-SE controls. Toterminate the SE, we injected diazepam (4 mg/kg, i.p.) 3 hafter the onset of continual seizures. Non-SE controls showedsimilar base levels of PGRN protein expression as did vehiclecontrols (Supplementary Fig. 3), suggesting that the nontonic-clonic effects of pilocarpine do not lead to PGRN upregulation.

To study the effect of PGRN in vivo (Figs. 3 and 4), weapplied PGRN (100 ng, 1 μl) into one side of the hippocampusand bovine serum albumin (BSA; 100 ng, 1 μl) into the oppo-site side at a constant speed of 0.5 μl/30 min before atropineinjection (Fig.7). Pilocarpine was applied 30 min after atropineas described above. Because our unpublished data estimateda half life of 12 h for PGRN protein, we reapplied the samedose of PGRN or BSA at 12 h after SE for the 24 h post-seizuregroup. The control groups in these studies (Figs. 3 and 4)received an equal volume of vehicle instead of pilocarpine.

4.5. Expression, purification and verification of PGRNprotein

We expressed the PGRN protein as previously described withminor modifications (He and Bateman, 1999). Human PGRNcDNA was cloned into pcDNA3.1 vector with a HIS tag at the5’-end (a generous gift from Dr. Andrew Bateman, McGillUniversity). HEK 293 T cells were transfected with pcDNA3.1-PGRN plasmid purified with endotoxin-free maxi kit (Qiagen).Culture medium collected from transfected cells was passedthrough a Ni-NTA Agarose column (Invitrogen), washed, andeluted in 250 mM imidazole in 50 mM NaH2PO4 and 2.5 M

Fig. 7 – PGRN and BSA injection sites in the hippocampus.Each black round point represents the estimated center of aninjection site in rats (BSA and PGRN side). Brain diagramswere adapted from the Paxinos and Watson atlas (Paxinosand Watson, 2005). From the top: Bregma �3.0 mm,�3.24 mm, and �3.6 mm.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 562

NaCl, pH 8.0. The PGRN medium was concentrated with anAmicon Ultra-15 Centrifugal Filter Unit (Ultracel-30 filtermembrane, Millipore). Crude acidified PGRN medium wasfractionated using a reverse phase high performance liquidchromatography (HPLC) column (C4, 10 mm, 250�4.6 mm,Grace Davison Discovery Sciences, Deerfield, IL, USA). Afterwashing, the PGRN eluate was collected at 215 nm UVabsorbance, and was further concentrated in PBS (pH 7.4)using an Amicon filter. The purity and identity of the PGRNprotein was confirmed by SDS–PAGE and Coomassie bluestaining and immunoblotting with anti-PGRN antibody (Sup-plementary Fig. 1), and was also confirmed with liquidchromatography-mass spectrometry (not shown).

4.6. Preparation of mixed neuronal-glial cultures

Mixed neuronal-glial cultures were prepared from cerebralcortex of embryonic day 18 Sprague Dawley rats (Center forDisease Modeling, UBC, Canada) as previously described(Kartvelishvily et al., 2006; Zhu et al., 2009). Briefly, aftercortices were isolated, trypsinized, and triturated, cells wereplated at a density of 1�106 per well in 12-well plate.Incubated at 37 1C in a humidified 5% CO2 chamber, mixedcultures were maintained for 3–5 days in vitro (DIV) in DMEMsupplemented with 10% FBS and 1% (vol/vol) penicillin/streptomycin. The mixed culture contained 52% MAP2+ neu-rons and 3% CD11b+ microglia.

4.7. Western blots

Rat cerebral cortex and hippocampus were isolated in ice-cold buffer containing (in mM): 75 sucrose, 87 NaCl, 25

NaHCO3, 25 D-glucose, 2.5 KCl, 1.25 NaH2PO4, 0.5 CaCl2, 7.0MgCl2, pH 7.3. Tissues were then homogenized in extractionbuffer containing 50 mM Tris base pH 7.2, 150 mM NaCl, 1 mMEDTA, 1 mM EGTA, 10% NP-40, 0.1% sodium dodecyl sulfate,0.5% sodium deoxycholate, and the protease inhibitor tablet(1 tablet/10 mL buffer, Roche, Laval, Quebec, Canada). Cul-tured cells were homogenized in 1� sample buffer contain-ing 124 mM Tris–HCl pH 6.8, 50% glycerol, 4% SDS, and 0.08%Bromophenol blue. The protein was resolved by SDS poly-acrylamide gel electrophoresis, and was transferred to nitro-cellulose membranes (Bio-Rad, Hercules, CA). After blocking,protein bands were detected using primary antibodies includ-ing sheep polycolonal anti-PGRN antibody (1:1000; R&D Sys-tems, Minneapolis, MN, USA), anti-GFAP antibody (1:5000;Santa Cruz Biotechnology, CA, USA), and anti-actin antibody(1:2000; Cell signal technology, Danvers, MA, USA). Afterwashing, we applied horseradish peroxidase-conjugated sec-ondary antibodies (PerkinElmer, Wellesley, MA) and visua-lized protein bands with enhanced chemiluminescencereagents (PerkinElmer). We used NIH ImageJ software(http://rsb.info.nih.gov/ij/) to measure the intensity of proteinbands. Actin bands were used as loading controls.

4.8. Quantitative real time PCR

RNA extraction and RT-PCR were performed as previouslydescribed (Zhu et al., 2009). We used Trizols Reagent (Invi-trogen) to extract total RNA from isolated brain tissues. 1 mgof RNA was reverse-transcribed to cDNA using SuperScript™II Reverse Transcriptase (Invitrogen). Duplicates of three4-fold serial dilutions of cDNA samples were used. Each25 mL RT-PCR reaction buffer contained 12.5 mL SYBRs GreenPCR master mix (Applied Biosystems, Foster City, CA, USA),300 nM primers, and 2 mL diluted cDNA. Fold inductions werecalculated using the formula 2� (ΔΔCt), where ΔΔCt is theΔCt(SE) -ΔCt(control), ΔCt is Ct(PGRN) - Ct(actin), and Ct is the cycleat which the threshold is crossed. The gene specific primerpairs were as follows: rat PGRN gene forward 5′–CACACGC-GATGCATTTCAC-3′ and reverse 5′–CTGCCCTGTTGGTCCTT-TGT-3′; actin gene forward 5′–ACGAGGCCCAGAGCAAGAG-3′and reverse 5′–TCTCCATGTCGTCCCAGTTG-3′. The quality ofthe PCR product was monitored by post-PCR melt curveanalysis.

4.9. Immunohistochemistry

Animals were transcardially perfused with saline followed by4% paraformaldehyde under urethane anesthesia. Brainswere postfixed, cryopreserved and sectioned coronally at20 μm. Free-floating sections were immunostained as pre-viously described (Ryu et al., 2009). After blocking, we incu-bated sections overnight at 4 1C with one of the primaryantibodies: mouse anti-rat CD11b (OX42 antibody, 1:150; AbDSerotec, Raleigh, NC, USA), rat anti-mouse CD11b (1:50; AbDSerotec), and sheep anti-PGRN (1:50; R&D Systems). We thenwashed and incubated sections with respective secondaryantibodies conjugated with Alexa Fluor 488 or 546 (1:1000,Invitrogen) at room temperature for 1 h. Sections were driedand mounted in DAPI-Fluoromount-G (Southern Biotech).

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 63

To identify the cell types expressing PGRN, we firstincubated free-floating sections with sheep anti-PGRN(1:50; R&D Systems) followed by donkey anti-sheep second-ary antibody conjugated with Alexa Fluor 546 (1:1000; Invi-trogen). After washing, sections were incubated with eithermouse anti-rat CD11b (OX42 antibody, 1:150; AbD Serotec)or rabbit anti-GFAP (1:500; Santa Cruz Biotechnology).Sections were then washed and incubated with Alexa Fluor488-conjugated goat anti-mouse or goat anti-rabbit (1:1000;Invitrogen), respectively. Samples were dried and sealed inDAPI-Fluoromount-G.

Images were captured using an Olympus Fluoview FV1000confocal microscope at 20� or 40� magnification with afixed exposure time in a given run. Focus was adjusted in theDAPI channel to avoid photobleaching. Images were analyzedin Image J. The number of CD11b+ and CD11B+/PGRN+ cells incortex, thalamus or dentate gyrus was determined for twosections per animal at approximately �2.8 to �3.8 mm fromBregma. The number of CD11b+ cells following PGRN injec-tion was counted in the dentate gyrus area near the injectionsite for four sections evenly spaced through the injection site(3.4 mm posterior to Bregma). To determine the number of SE-activated CD11b+ microglia/macrophages in the dentate gyrusof PGRN knockout mice, three sections at approximately �2.8 to�3.8 mm from Bregma were analyzed for each animal. Thenumber of cells was divided by the sampled area to measuredensity. Quantification was conducted in a blinded manner.

4.10. Fluoro-jade B staining

SE-induced neuronal death was detected by Fluoro-Jade Bstaining as previously described (Chen et al., 2007). Briefly,brain sections mounted on slides were subsequentlyimmersed for 1 min in 100% ethanol, 1 min in 70% ethanol,and 1 min in distilled water, and 15 min in 0.06% potassiumpermanganate. After washing, slides were incubated for30 min in 0.0004% Fluoro-Jade B (Histo-Chem, Jefferson, AR,USA) and 0.1% acetic acid. Slides were dried, mounted, andimaged with an Olympus Fluoview FV1000 confocal micro-scope at 10� magnification. The number of Fluoro-JadeB-positive cells was counted in the polymorphic cell layer ofdentate gyrus and divided by the sampled area for threeevenly spaced sections (at approximately �2.8 to �3.8 mmfrom Bregma) in a blinded manner.

4.11. Statistical analysis

Results were expressed as means7SEM. One-way or two-wayANOVA was used to compare data among groups if the groupnumber is greater than two. When ANOVAs indicated sig-nificant differences among groups, least significant differencepost-hoc test was used to generate p values for specificcomparison. Two-tailed Student's t tests were conductedwhen data from only two groups (treatment and control)are compared. Differences were considered significant atpo0.05.

Acknowledgment

This work was supported by grants from the Savoy Founda-tion and Pacific Alzheimer Research Foundation (PARF). Wethank Dr. Andrew Bateman from McGill University for pro-viding the plasmid encoding the human progranulin cDNA.

Appendix A. Supporting information

Supplementary data associated with this article can be foundin the online version at http://dx.doi.org/10.1016/j.brainres.2013.07.023.

r e f e r e n c e s

Baker, M., Mackenzie, I.R., Pickering-Brown, S.M., Gass, J.,Rademakers, R., Lindholm, C., Snowden, J., Adamson, J.,Sadovnick, A.D., Rollinson, S., Cannon, A., Dwosh, E., Neary, D.,Melquist, S., Richardson, A., Dickson, D., Berger, Z., Eriksen, J.,Robinson, T., Zehr, C., Dickey, C.A., Crook, R., McGowan, E.,Mann, D., Boeve, B., Feldman, H., Hutton, M., 2006. Mutations inprogranulin cause tau-negative frontotemporal dementialinked to chromosome 17. Nature 442, 916–919.

Ben-Ari, Y., Tremblay, E., Riche, D., Ghilini, G., Naquet, R., 1981.Electrographic, clinical and pathological alterations followingsystemic administration of kainic acid, bicuculline orpentetrazole: metabolic mapping using the deoxyglucosemethod with special reference to the pathology of epilepsy.Neuroscience 6, 1361–1391.

Bhandari, V., Daniel, R., Lim, P.S., Bateman, A., 1996. Structuraland functional analysis of a promoter of the human granulin/epithelin gene. Biochem. J. 319 (Pt 2), 441–447.

Bianco, F., Pravettoni, E., Colombo, A., Schenk, U., Moller, T.,Matteoli, M., Verderio, C., 2005. Astrocyte-derived ATP inducesvesicle shedding and IL-1 beta release from microglia. J.Immunol. 174, 7268–7277.

Chen, Q., He, S., Hu, X.L., Yu, J., Zhou, Y., Zheng, J., Zhang, S.,Zhang, C., Duan, W.H., Xiong, Z.Q., 2007. Differential roles ofNR2A- and NR2B-containing NMDA receptors in activity-dependent brain-derived neurotrophic factor gene regulationand limbic epileptogenesis. J. Neurosci. 27, 542–552.

Fabene, P.F., Bramanti, P., Constantin, G., 2010. The emerging rolefor chemokines in epilepsy. J. Neuroimmunol. 224, 22–27.

Frampton, G., Invernizzi, P., Bernuzzi, F., Pae, H.Y., Quinn, M.,Horvat, D., Galindo, C., Huang, L., McMillin, M., Cooper, B.,Rimassa, L., DeMorrow, S., 2012. Interleukin-6-drivenprogranulin expression increases cholangiocarcinoma growthby an Akt-dependent mechanism. Gut 61, 268–277.

Ghoshal, N., Dearborn, J.T., Wozniak, D.F., Cairns, N.J., 2012. Corefeatures of frontotemporal dementia recapitulated inprogranulin knockout mice. Neurobiol. Dis. 45, 395–408.

Goffin, K., Nissinen, J., Van Laere, K., Pitkanen, A., 2007. Cyclicityof spontaneous recurrent seizures in pilocarpine model oftemporal lobe epilepsy in rat. Exp. Neurol. 205, 501–505.

Gowing, G., Philips, T., Van Wijmeersch, B., Audet, J.N., Dewil, M.,Van Den Bosch, L., Billiau, A.D., Robberecht, W., Julien, J.P.,2008. Ablation of proliferating microglia does not affect motorneuron degeneration in amyotrophic lateral sclerosis causedby mutant superoxide dismutase. J. Neurosci. 28, 10234–10244.

Guo, A., Tapia, L., Bamji, S.X., Cynader, M.S., Jia, W., 2010.Progranulin deficiency leads to enhanced cell vulnerability

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 564

and TDP-43 translocation in primary neuronal cultures. BrainRes. 1366, 1–8.

He, Z., Bateman, A., 1999. Progranulin gene expression regulatesepithelial cell growth and promotes tumor growth in vivo.Cancer Res. 59, 3222–3229.

Holtman, L., van Vliet, E.A., Aronica, E., Wouters, D., Wadman, W.J.,Gorter, J.A., 2013. Blood plasma inflammation markers duringepileptogenesis in post-status epilepticus rat model for temporallobe epilepsy. Epilepsia 54, 589–595.

Holtzman, D.M., Lowenstein, D.H., 1995. Selective inhibition ofaxon outgrowth by antibodies to NGF in a model of temporallobe epilepsy. J. Neurosci. 15, 7062–7070.

Honda, S., Sasaki, Y., Ohsawa, K., Imai, Y., Nakamura, Y., Inoue, K.,Kohsaka, S., 2001. Extracellular ATP or ADP induce chemotaxisof cultured microglia through Gi/o-coupled P2Y receptors. J.Neurosci. 21, 1975–1982.

Hu, F., Padukkavidana, T., Vaegter, C.B., Brady, O.A., Zheng, Y.,Mackenzie, I.R., Feldman, H.H., Nykjaer, A., Strittmatter, S.M.,2010. Sortilin-mediated endocytosis determines levels of thefrontotemporal dementia protein, progranulin. Neuron 68,654–667.

Kartvelishvily, E., Shleper, M., Balan, L., Dumin, E., Wolosker, H.,2006. Neuron-derived D-serine release provides a novel meansto activate N-methyl-D-aspartate receptors. J. Biol. Chem. 281,14151–14162.

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

Kucher, B.M., Neary, J.T., 2005. Bi-functional effects of ATP/P2receptor activation on tumor necrosis factor-alpha release inlipopolysaccharide-stimulated astrocytes. J. Neurochem. 92,525–535.

Lokensgard, J.R., Hu, S., Hegg, C.C., Thayer, S.A., Gekker, G.,Peterson, P.K., 2001. Diazepam inhibits HIV-1 Tat-inducedmigration of human microglia. J. Neurovirol. 7, 481–486.

Luttjohann, A., Fabene, P.F., van Luijtelaar, G., 2009. A revisedRacine's scale for PTZ-induced seizures in rats. Physiol. Behav.98, 579–586.

Matsuwaki, T., Asakura, R., Suzuki, M., Yamanouchi, K.,Nishihara, M., 2011. Age-dependent changes in progranulinexpression in the mouse brain. J. Reprod. Dev. 57, 113–119.

Meldrum, B.S., Akbar, M.T., Chapman, A.G., 1999. Glutamatereceptors and transporters in genetic and acquired models ofepilepsy. Epilepsy Res. 36, 189–204.

Moisse, K., Volkening, K., Leystra-Lantz, C., Welch, I., Hill, T.,Strong, M.J., 2009. Divergent patterns of cytosolic TDP-43 andneuronal progranulin expression following axotomy:implications for TDP-43 in the physiological response toneuronal injury. Brain Res. 1249, 202–211.

Naphade, S.B., Kigerl, K.A., Jakeman, L.B., Kostyk, S.K., Popovich, P.G.,Kuret, J., 2010. Progranulin expression is upregulated after spinalcontusion in mice. Acta Neuropathol. 119, 123–133.

Nicoletti, J.N., Shah, S.K., McCloskey, D.P., Goodman, J.H., Elkady, A.,Atassi, H., Hylton, D., Rudge, J.S., Scharfman, H.E., Croll, S.D.,2008. Vascular endothelial growth factor is up-regulated afterstatus epilepticus and protects against seizure-induced neuronalloss in hippocampus. Neuroscience 151, 232–241.

Panenka, W., Jijon, H., Herx, L.M., Armstrong, J.N., Feighan, D.,Wei, T., Yong, V.W., Ransohoff, R.M., MacVicar, B.A., 2001.P2� 7-like receptor activation in astrocytes increaseschemokine monocyte chemoattractant protein-1 expressionvia mitogen-activated protein kinase. J. Neurosci. 21,7135–7142.

Paxinos, G., Watson, C., 2005. The Rat Brain in StereotaxicCoordinates. Elsevier.

Peltola, J., Hurme, M., Miettinen, A., Keranen, T., 1998. Elevatedlevels of interleukin-6 may occur in cerebrospinal fluid frompatients with recent epileptic seizures. Epilepsy Res. 31,129–133.

Petkau, T.L., Neal, S.J., Orban, P.C., MacDonald, J.L., Hill, A.M.,Lu, G., Feldman, H.H., Mackenzie, I.R., Leavitt, B.R., 2010.Progranulin expression in the developing and adult murinebrain. J. Comp. Neurol. 518, 3931–3947.

Petkau, T.L., Neal, S.J., Milnerwood, A., Mew, A., Hill, A.M., Orban, P.,Gregg, J., Lu, G., Feldman, H.H., Mackenzie, I.R., Raymond, L.A.,Leavitt, B.R., 2012. Synaptic dysfunction in progranulin-deficientmice. Neurobiol. Dis. 45, 711–722.

Petkau, T.L., Zhu, S., Lu, G., Fernando, S., Cynader, M., Leavitt, B.R.,2013. Sensitivity to neurotoxic stress is not increased inprogranulin-deficient mice. Neurobiol. Aging.

Philips, T., De Muynck, L., Thu, H.N., Weynants, B., Vanacker, P.,Dhondt, J., Sleegers, K., Schelhaas, H.J., Verbeek, M.,Vandenberghe, R., Sciot, R., Van Broeckhoven, C.,Lambrechts, D., Van Leuven, F., Van Den Bosch, L.,Robberecht, W., Van Damme, P., 2010. Microglial upregulation ofprogranulin as a marker of motor neuron degeneration. J.Neuropathol. Exp. Neurol. 69, 1191–1200.

Pickford, F., Marcus, J., Camargo, L.M., Xiao, Q., Graham, D., Mo, J.R.,Burkhardt, M., Kulkarni, V., Crispino, J., Hering, H., Hutton, M.,2011. Progranulin is a chemoattractant for microglia andstimulates their endocytic activity. Am. J. Pathol. 178, 284–295.

Racine, R.J., 1972. Modification of seizure activity by electricalstimulation. II. Motor seizure. Electroencephalogr. Clin.Neurophysiol. 32, 281–294.

Riva, M.A., Donati, E., Tascedda, F., Zolli, M., Racagni, G., 1994.Short- and long-term induction of basic fibroblast growthfactor gene expression in rat central nervous system followingkainate injection. Neuroscience 59, 55–65.

Rudge, J.S., Mather, P.E., Pasnikowski, E.M., Cai, N., Corcoran, T.,Acheson, A., Anderson, K., Lindsay, R.M., Wiegand, S.J., 1998.Endogenous BDNF protein is increased in adult rathippocampus after a kainic acid induced excitotoxic insult butexogenous BDNF is not neuroprotective. Exp. Neurol. 149,398–410.

Ryu, J.K., Cho, T., Choi, H.B., Wang, Y.T., McLarnon, J.G., 2009.Microglial VEGF receptor response is an integral chemotacticcomponent in Alzheimer's disease pathology. J Neurosci. 29,3–13.

Shulkes, A., Harris, Q.L., Lewis, S.J., Vajda, J.E., Jarrott, B., 1988.Regional brain concentrations of neurotensin followingamygdaloid kindled and cortical suprathreshold stimulation-induced seizures in the rat. Neuropeptides 11, 77–81.

Speth, C., Joebstl, B., Barcova, M., Dierich, M.P., 2000. HIV-1envelope protein gp41 modulates expression of interleukin-10and chemokine receptors on monocytes, astrocytes andneurones. AIDS 14, 629–636.

Suh, H.S., Choi, N., Tarassishin, L., Lee, S.C., 2012. Regulation ofprogranulin expression in human microglia and proteolysis ofprogranulin by matrix metalloproteinase-12 (MMP-12). PLoSOne 7, e35115.

Sun, L., Eriksen, J.L., 2011. Recent insights into the involvement ofprogranulin in frontotemporal dementia. Curr.Neuropharmacol. 9, 632–642.

Tanaka, Y., Matsuwaki, T., Yamanouchi, K., Nishihara, M., 2013.Increased lysosomal biogenesis in activated microglia andexacerbated neuronal damage after traumatic brain injury inprogranulin-deficient mice. Neuroscience.

Tapia, L., Milnerwood, A., Guo, A., Mills, F., Yoshida, E., Vasuta, C.,Mackenzie, I.R., Raymond, L., Cynader, M., Jia, W., Bamji, S.X.,2011. Progranulin deficiency decreases gross neuralconnectivity but enhances transmission at individualsynapses. J. Neurosci. 31, 11126–11132.

Van Damme, P., Van Hoecke, A., Lambrechts, D., Vanacker, P.,Bogaert, E., van Swieten, J., Carmeliet, P., Van Den Bosch, L.,Robberecht, W., 2008. Progranulin functions as a neurotrophicfactor to regulate neurite outgrowth and enhance neuronalsurvival. J. Cell Biol. 181, 37–41.

b r a i n r e s e a r c h 1 5 3 0 ( 2 0 1 3 ) 5 4 – 6 5 65

Van Der Wal, E.A., Gomez-Pinilla, F., Cotman, C.W., 1994. Seizure-associated induction of basic fibroblast growth factor and itsreceptor in the rat brain. Neuroscience 60, 311–323.

van Swieten, J.C., Heutink, P., 2008. Mutations in progranulin (GRN)within the spectrum of clinical and pathological phenotypes offrontotemporal dementia. Lancet Neurol. 7, 965–974.

Vezzani, A., 2005. Inflammation and epilepsy. Epilepsy Curr. 5, 1–6.Wieraszko, A., Seyfried, T.N., 1989. Increased amount of

extracellular ATP in stimulated hippocampal slices of seizureprone mice. Neurosci Lett. 106, 287–293.

Wils, H., Kleinberger, G., Pereson, S., Janssens, J., Capell, A., VanDam, D., Cuijt, I., Joris, G., De Deyn, P.P., Haass, C., VanBroeckhoven, C., Kumar-Singh, S., 2012. Cellular ageing,increased mortality and FTLD-TDP-associated neuropathologyin progranulin knockout mice. J. Pathol. 228, 67–76.

Yang, F., Liu, Z.R., Chen, J., Zhang, S.J., Quan, Q.Y., Huang, Y.G.,Jiang, W., 2010. Roles of astrocytes and microglia in seizure-induced aberrant neurogenesis in the hippocampus of adultrats. J. Neurosci. Res. 88, 519–529.

Yin, F., Banerjee, R., Thomas, B., Zhou, P., Qian, L., Jia, T., Ma, X.,Ma, Y., Iadecola, C., Beal, M.F., Nathan, C., Ding, A., 2010.Exaggerated inflammation, impaired host defense, andneuropathology in progranulin-deficient mice. J. Exp. Med.207, 117–128.

Zheng, Y., Brady, O.A., Meng, P.S., Mao, Y., Hu, F., 2011. C-terminus of progranulin interacts with the beta-propellerregion of sortilin to regulate progranulin trafficking. PLoS One6, e21023.

Zhu, J., Nathan, C., Jin, W., Sim, D., Ashcroft, G.S., Wahl, S.M.,Lacomis, L., Erdjument-Bromage, H., Tempst, P., Wright, C.D.,Ding, A., 2002. Conversion of proepithelin to epithelins: rolesof SLPI and elastase in host defense and wound repair. Cell111, 867–878.

Zhu, S., Tai, C., MacVicar, B.A., Jia, W., Cynader, M.S., 2009.Glutamatergic stimulation triggers rapid Krupple-like factor 4expression in neurons and the overexpression of KLF4sensitizes neurons to NMDA-induced caspase-3 activity. BrainRes. 1250, 49–62.