Inflammation without neuronal death triggers striatal neurogenesis

Neurobiology of Disease 83 (2015) 1–15

Contents lists available at ScienceDirect

Neurobiology of Disease journal homepage: www.elsevier.com/locate/ynbdi

Inflammation without neuronal death triggers striatal neurogenesis comparable to stroke Katie Z. Chapman a,1, Ruimin Ge a,1, Emanuela Monni a, Jemal Tatarishvili a, Henrik Ahlenius b, Andreas Arvidsson a,c, Christine T. Ekdahl d, Olle Lindvall a,2, Zaal Kokaia a,⁎,2 a

Laboratory of Stem Cells and Restorative Neurology, Lund Stem Cell Center, University Hospital, SE-221 84 Lund, Sweden Stem cells, Aging and Neurodegeneration Group, Lund Stem Cell Center, University Hospital, SE-221 84 Lund, Sweden Division of Neurology, University Hospital, SE-221 84 Lund, Sweden d Inflammation and Stem Cell Therapy Group, Wallenberg Neuroscience Center, University Hospital, SE-221 84 Lund, Sweden b c

a r t i c l e

i n f o

Article history: Received 6 July 2015 Revised 11 August 2015 Accepted 17 August 2015 Available online 20 August 2015 Keywords: CXCL13 CXCR5 Inflammation Lipopolysaccharide Microarray Microglia Neurogenesis Striatum Stroke

a b s t r a c t Ischemic stroke triggers neurogenesis from neural stem/progenitor cells (NSPCs) in the subventricular zone (SVZ) and migration of newly formed neuroblasts toward the damaged striatum where they differentiate to mature neurons. Whether it is the injury per se or the associated inflammation that gives rise to this endogenous neurogenic response is unknown. Here we showed that inflammation without corresponding neuronal loss caused by intrastriatal lipopolysaccharide (LPS) injection leads to striatal neurogenesis in rats comparable to that after a 30 min middle cerebral artery occlusion, as characterized by striatal DCX+ neuroblast recruitment and mature NeuN+/BrdU+ neuron formation. Using global gene expression analysis, changes in several factors that could potentially regulate striatal neurogenesis were identified in microglia sorted from SVZ and striatum of LPS-injected and stroke-subjected rats. Among the upregulated factors, one chemokine, CXCL13, was found to promote neuroblast migration from neonatal mouse SVZ explants in vitro. However, neuroblast migration to the striatum was not affected in constitutive CXCL13 receptor CXCR5−/− mice subjected to stroke. Infarct volume and pro-inflammatory M1 microglia/macrophage density were increased in CXCR5−/− mice, suggesting that microglia-derived CXCL13, acting through CXCR5, might be involved in neuroprotection following stroke. Our findings raise the possibility that the inflammation accompanying an ischemic insult is the major inducer of striatal neurogenesis after stroke. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Ischemic stroke following cerebral artery occlusion is a leading cause of chronic disability in humans, and effective treatment to promote recovery is lacking. It is well established that neural stem/progenitor cells (NSPCs) in the subventricular zone (SVZ) of adult rodents continuously produce new neuroblasts that migrate into the injured striatum for several months after stroke (Arvidsson et al., 2002; Thored et al., 2006, 2007; Parent et al., 2002). These neuroblasts differentiate to mature neurons, become integrated (Yamashita et al., 2006), project to substantia nigra (Sun et al., 2012), and seem to be functional (Hou et al., 2008). There is also evidence for enhanced SVZ cell proliferation and neuroblast formation after stroke in humans (Macas et al., 2006; Marti-Fabregas et al., 2010; Minger et al., 2007). The interest in this ⁎ Corresponding author at: Laboratory of Stem Cells and Restorative Neurology, Lund Stem Cell Center, University Hospital, SE-221 84 Lund, Sweden. E-mail address: [email protected] (Z. Kokaia). 1 K.Z.C. and R.G. contributed equally to this work. 2 O.L. and Z.K. have shared senior authorship. Available online on ScienceDirect (www.sciencedirect.com).

http://dx.doi.org/10.1016/j.nbd.2015.08.013 0969-9961/© 2015 Elsevier Inc. All rights reserved.

potential self-repair mechanism was further increased by a recent report showing that neuroblasts from SVZ enter striatum and become interneurons in adult humans under normal conditions (Ernst et al., 2014). However, in rats, only a fraction of stroke-induced neurons survive long-term (Arvidsson et al., 2002), and it is unclear whether they contribute to the spontaneous functional recovery after the insult (Lagace, 2012). Stroke is associated with inflammation, which exerts a complex influence on several steps of striatal neurogenesis (Tobin et al., 2014). Several months following stroke in rats, activated microglia/ macrophages continue to be localized in the ipsilateral SVZ concomitant with the continuous production of new neuroblasts migrating into the striatum (Thored et al., 2009). Factors released from activated microglia/macrophages can either stimulate NSPC proliferation in the SVZ, as with IGF-1 (Thored et al., 2009; Yan et al., 2006) and IL-15 (Gomez-Nicola et al., 2011), or in the case of TNF-α signaling through TNF-R1, suppress it (Iosif et al., 2008). Activated microglia/macrophages are also involved in directing neuroblasts to the damaged area by secreting CXCL12 (Robin et al., 2006), MCP-1 (Yan et al., 2007), and osteopontin (Yan et al., 2009). Finally, one study indicates that

2

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

inflammation associated with stroke contributes to the poor survival of the new striatal neurons (Hoehn et al., 2005), similar to what has been described for new hippocampal neurons in other inflammatory environments (Ekdahl et al., 2003; Monje et al., 2003). The number of recruited neuroblasts correlates with the volume of striatal injury after stroke of different severities (Thored et al., 2006), but whether it is the injury per se or the associated inflammation that induces striatal neurogenesis is unknown. We show here that inflammation without neuronal death, evoked by intrastriatal LPS-injection, is sufficient to trigger striatal neurogenesis similar to that after stroke in rats. Using global gene expression analysis on sorted rat microglia, we identified several potential regulators of this response with changes observed both after LPS and middle cerebral artery occlusion (MCAO). One of the upregulated factors, CXCL13, improved migration of mouse SVZ neuroblasts in vitro. However, knockout (KO) mice for CXCR5, the receptor for CXCL13, showed no impairment of neuroblast migration after stroke but more extensive injury and pro-inflammatory phenotype of microglia/macrophages.

The internal carotid artery (ICA) was temporarily occluded with a microvascular clip. A small incision was made in the common carotid and a heat-blunted nylon microfilament was advanced into the ICA until resistance was felt (approx. 19 mm). Animals recovered from anesthesia during the occlusion. 30 min after occlusion, animals were re-anesthetized and the filament was withdrawn. Temperature was maintained at 37 ± 0.5 °C while animals were under anesthesia. Sham surgeries were carried out in the same way but the filament was only advanced 2 mm inside the ICA. MCAO animals that did not fulfill predefined inclusion criteria for successful 30 min occlusion (N40% striatal damage; No cortical damage; No subarachnoid hemorrhage) were excluded following NeuN staining. In mice, the procedure was modified as follows: Right carotid arteries were isolated, and the common carotid artery and the external carotid artery were ligated. The ICA was temporarily occluded with a microvascular clip, and a silicon-coated microfilament was placed into the external carotid artery via a small incision and advanced into the ICA until resistance was felt (approx. 9 mm). Occlusion was maintained for 35 min before the filament was withdrawn.

2. Materials and methods 2.3. Immunohistochemistry 2.1. Animals and experimental design All procedures were carried out in accordance with the guidelines set by the Malmö-Lund Ethical Committee for the use of laboratory animals, and were conducted in accordance with the European Union directive on the subject of animal rights. Procedures were carried out on male Wistar rats (250–300 g, Charles River, Germany), male C57BL/6 J mice (25–30 g, Charles River, Germany), and male CXCR5−/− mice (25–30 g, The Jackson Laboratory, http://jaxmice. jax.org/strain/006659.html), housed under 12 h light/12 h dark cycle with ad libitum access to food and water. In order to characterize immunohistochemically the effect of LPS and MCAO, rats were randomly assigned to one of 8 groups (n = 4–8 per group): LPS, vehicle (veh), 30 min MCAO or sham with either a 2 or 6 weeks recovery. Each animal was given intraperitoneal (i.p) injections of BrdU (50 mg/kg, Sigma) in phosphate-buffered saline (PBS) twice daily for 2 weeks beginning on the evening following surgery. For analysis of the gene profiles in striatal or SVZ microglia, animals were randomly assigned to LPS, MCAO or naïve groups. Due to low cell numbers, notably in SVZ, each ‘n’ (4) comprised 3 pooled animals. For examining the role of CXCL13 and its receptor CXCR5 in vivo, wild-type and CXCR5−/− mice (n = 6 and 4, respectively) were subjected to 35 min MCAO. Animals were sacrificed at 2 weeks after surgery. 2.2. Surgical procedures Animals were anesthetized with isoflurane (3.5% induction; 1.5% maintenance) in 70% N2O/30% O2. All animals received locally injected lidocaine for pain relief. While under anesthesia and in the early recovery period (2 h), animals were placed on a heat pad maintained at 37 °C. LPS from Salmonella enterica, serotype abortus equi (Sigma-Aldrich; 15 μg in 1.5 μl of artificial CSF (aCSF)) or vehicle (aCSF) was stereotaxically injected using a self-made glass microneedle fixed to a gas-tight syringe (Hamilton company) into the right striatum (coordinates: 1.2 mm rostral, 2.5 mm lateral to bregma, 4.5 mm ventral from brain surface, toothbar at −3.3 mm) (Paxinos and Watson, 1998). In a pilot experiment in mice, a dose–response curve was established with 0.01 to 100 μg LPS administered in ten-fold increasing concentrations as above (coordinates: 0.9 mm rostral, 1.6 mm lateral to bregma and 3.5 mm ventral from brain surface, toothbar at 0 mm). The intraluminal filament technique was used to induce transient MCAO (Koizumi et al., 1986). In rats, the right carotid arteries were isolated and the common and external carotids were proximally ligated.

Rats and mice were deeply anesthetized with an overdose of pentobarbital and transcardially perfused with saline followed by 4% paraformaldehyde (PFA). Brains were post-fixed overnight in 4% PFA, placed in 20% sucrose for 24 h, and then cut into 8 series of 30 μm thick coronal sections on dry ice. For each staining, one full series was used. Fluorescence double-staining was used for visualization of BrdU +/DCX +, BrdU +/NeuN +, Iba1 +/NeuN +, Iba1 +/Ki67 +, Iba1 +/ED1 +, CD16/32 +/Iba1 +, DCX +/HuD +, DCX +/PDGFRα +, DCX +/S100B and GFAP +/Nestin + cells. All sections for BrdU staining were pre-treated with 1 M HCl for 10 min at 65 °C and 20 min at room temperature. All double stains were carried out according to the following protocol: Free-floating sections were pre-incubated with the appropriate serum and then incubated with primary antibodies overnight at 4 °C. Sections were incubated for 2 h in the dark with secondary antibodies conjugated with Cy3 or Alexa Fluor 488 (1:200, Molecular Probes, Life Technologies), Cy3-conjugated donkey anti-rat/goat anti-rabbit/donkey anti-mouse (all 1:200, Jackson ImmunoResearch), or biotinylated horse anti-mouse/anti-goat (both 1:200, Vector Laboratories). The primary antibodies used are listed in Table 1. Single labeling for NeuN was performed with biotinylated horse anti-mouse antibody and visualized with avidin–biotin–peroxidase complex (Elite ABC kit, Vector Laboratories), followed by peroxidasecatalyzed diaminobenzidine reaction.

Table 1 Primary antibodies used for immunohistochemical staining. Antibody

Host species

Concentration

Company

Anti-BrdU Anti-NeuN Anti-PDGFRα Anti-DCX Anti-Iba1 Anti-ED1 Anti-ki67 Anti-HuD Anti-S100B Anti-GFAP Anti-nestin Anti-CD16/32 Anti-iNOS Anti-RECA

Rat Mouse Mouse Goat Rabbit Rat Mouse Rabbit Rabbit Rabbit Mouse Rat Rabbit Mouse

1:200 1:100 1:300 1:400 1:1000 1:200 1:500 1:200 1:200 1:400 1:200 1:200 1:200 1:400

Abcam Merck Millipore Santa Cruz Biotechnology Santa Cruz Biotechnology Wako AbD Serotec Novocastra, Leica Biosystems Sigma-Aldrich Sigma-Aldrich Zymed, Life Technologies Merck Millipore BD Biosciences BD Biosciences AbD Serotec

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

2.4. Microscopical analysis All assessments were performed blinded to treatment conditions. Neuronal death was assessed by an estimation of the total number of remaining NeuN + cells in the striatum using the optical fractionator technique (West et al., 1991). This was carried out using the Computer Assisted Stereological Toolbox (C.A.S.T-GRID) software (Olympus, Denmark) as described previously (Arvidsson et al., 2003), but with sampling from three coronal sections at approx + 0.9 mm, + 1.2 mm (LPS injection site) and +1.5 mm from Bregma (Paxinos and Watson, 1998). In brief, images from the microscope were acquired with a digital camera and displayed live on a monitor screen. Using a 1.25× objective, the striatum was delineated on the screen according to predefined criteria: dorsal and lateral boundaries along the corpus callosum and the medial sides of claustrum and dorsal endopiriform nucleus; ventral boundary along a line drawn from the border of the dorsal endopiriform nucleus at the level of the flexure of the piriform cortex to the anterior commissure, or at more caudal levels along a line following the posterior part of the anterior commissure; and the medial boundary along a line drawn from the anterior commissure to the ventral tip of the lateral ventricle and the lateral side of the ventricle, or at more caudal levels along the lateral side of globus pallidus and the lateral side of the lateral ventricle. The thickness of each section was measured at high magnification at multiple locations within the delineated striatum using a microcator attached to the stage of the microscope. The striatum was then systematically sampled at high magnification and cells at each sampling point were counted using a three-dimensional probe (counting frame combined with optical dissector) following accepted stereological cell counting methods (Gundersen, 1978; West, 1999). Counting frame area and stepping distances were chosen to sample 100–200 cells per striatum, keeping the number of cells counted at each sampling point as close to 1 as possible. Number of cells per striatum was calculated by dividing the number of cells counted with the sampling fraction. Images of SVZ in sections with NeuN and cresyl violet staining from the same aforementioned rostrocaudal levels were first taken under 40 × magnification. The SVZ was then defined by cells stained only with cresyl violet and area was measured using Visiopharm software (Visiopharm, Denmark). For volume measurement, striatum was first delineated using the pre-defined criteria described above. Area was then measured using Visiopharm software (Visiopharm, Denmark) in coronal sections from + 2.2 mm to −0.4 mm from bregma. Striatal volume was estimated by multiplying the areas with the distance between sections (240 μm). Numbers of Iba1+, BrdU+/DCX+, BrdU−/DCX+, BrdU+/NeuN+ single or double-labeled cells in the rat striatum were counted using a 0.0625 mm2 quadratic grid on an epifluorescence microscope with a 40 × objective on the three coronal sections described above. Cell counts are presented as the total number in these 3 sections. Because striatal volume was decreased in MCAO animals (by 20% and 38% at 2 and 6 weeks after the insult, respectively, compared to sham groups), the NeuN +, Iba1 +, BrdU +/DCX +, BrdU −/DCX + and BrdU +/ NeuN + cell counts in 3 coronal sections were, therefore, multiplied with the shrinkage ratio to compensate for the effect of shrinkage on cell density. The shrinkage ratio was calculated by dividing Iba1+ cell density in all striatal sections (+ 2.2 mm to − 0.4 mm from bregma) with Iba1 + cell density in the 3 coronal sections (+ 0.9 mm, +1.2 mm and +1.5 mm from bregma). Distribution of cells was calculated as described previously (Thored et al., 2006). Additionally, 50 DCX + cells per animal were analyzed using epifluorescence microscopy to assess co-expression with HuD, PDGFRα or S100β. Since it was not feasible to count astrocyte numbers due to the extensive astrogliosis seen in both MCAO and LPS-injected animals, each animal was given a score of 0–3 by a blinded observer based on a semi-quantitative scale of astrocyte activation as obverved by GFAP and nestin staining: Score 0: Astrocytes appear branched and thin with no aggregation; no double positive nestin/GFAP cells; no obvious

3

increase in astrocyte numbers compared to contralateral side. Score 1: Some astrocytes exhibit a more swollen, ‘active’ phenotype but with minimal aggregation; active astrocytes are limited to less than 1/3 of striatum with only small increase in numbers; less than 1/3 of ‘active’ GFAP positive cells are nestin positive. Score 2: Astrocytes exhibit ‘active’ phenotype and aggregation, with clear increase in numbers; more than 1/3 but less than 2/3 of ‘active’ GFAP positive cells are nestin postive; activation localized or more diffuse. Score 3: Astrocytes exhibit ‘active’ phenotype and aggregation, with clear increase in numbers; more than 2/3 of ‘active’ GFAP postive cells are nestin postive and activation is widespread. In mice, the quantification of DCX + cell number/distribution, microglia/macrophage density and infarct volume was performed in three coronal sections at + 0.02, + 0.5, + 0.98 mm from bregma. DCX + cell number/distribution was quantified using a 0.0625 mm2 quadratic grid on an epifluorescence microscope with a 40 × objective. Iba1+, Iba1+/ED1+, CD16/32+ (M1 marker) microglia/macrophage densities were assessed by first defining the striatal region, and counting positive cells using the Visiopharm software (Visiopharm, Denmark), and then calculating cell density by dividing cell number by striatal area. For infarct volume estimation, images of NeuN-DAB stained sections were first taken under 4 × magnification. Intact areas identified by NeuN + cells in the ipsilateral and contralateral hemispheres were delineated and then measured using Visiopharm software. The area of unlesioned tissue in the ipsilateral hemisphere was subtracted from that of the contralateral hemisphere to get infarct area, and this area was subsequently multiplied by the distance between the sections (240 μm) to get infarct volume. Mouse and rat cells double-labeled with different markers in epifluorescence microscopy were randomly selected and co-expression validated by confocal microscopy (Carl Zeiss Jena GmbH, Germany) using orthogonal views of single optical sections from confocal Z-series. 2.5. Microglia isolation and flow cytometry Microglia were isolated according to a modified version of a previously described method (Cardona et al., 2006). Rats or mice were decapitated, brains were rapidly removed and placed in Leibovitz-15 (L-15) media. Brains were then placed in a brain matrix and cut into 1 mm coronal sections and the striatum and SVZ were then microdissected in L15 media. All solutions and instruments were kept ice-cold until this point. In a laminar hood, tissue was diced and re-suspended in a + 37 °C papain, neutral protease (dispase II), DNAse (PPD) solution and incubated for 30 min at + 37 °C. The PPD solution was prepared as follows: 2.5 U/ml papain (Worthington Biochemical Corporation), 250 U/ml DNAse I (Worthington Biochemical Corporation), and 1 U/ml dispase II (Roche) were dissolved in DMEM containing 4.5 g/l glucose at + 37 °C, filter-sterilized and stored at −20 °C prior to use. Tissue was then triturated, and excess DMEM/F12 with glutamine (500 μl/50 ml) and 10% fetal bovine serum (FBS) medium was added. Cells were washed by centrifugation, re-suspended in medium and strained through a 40 μm strainer. Cells were then re-centrifuged and re-suspended in 4 ml 37% percoll. 4 ml 70% percoll was slowly underlaid and 30% percoll added on top followed by an additional 2 ml of media. A gradient was then run for 40 min, 200 ×g, +18 °C. Minimal acceleration and brake settings were used. The thick viscous layer of debris formed was removed and the halo-like ring of brain-microglia formed between the 70% and 37% gradients was collected and washed by centrifugation in media. Cells were then re-suspended in FACS block buffer (0.1% FBS in PBS) with antibodies for 30 min at + 4 °C. For rat microglia sorting, Allophycocyanin (APC)-conjugated mouse anti-rat CD11b (1:100; Life Technologies) and R-Phycoerythrin (RPE)-conjugated mouse anti-rat CD45 (1:10; AbD Serotec) antibodies were used. For mouse microglia sorting, Brilliant Violet 421-conjugated rat anti-mouse/human

4

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

CD11b (1:200, BioLegend) and Brilliant Violet 510-conjugated rat anti-mouse CD45 (1:20, BioLegend) antibodies were used. Cells were then washed by centrifugation at + 4 °C and re-suspended in 400 μl FACS buffer (1% BSA in PBS) to be ready for FACS sorting (BD FACSAria™ III, Becton Dickinson, Franklin lakes, NJ). 2 min prior to sorting, 2 μl propidium iodide (PI) was added to the sample for the identification of dead cells. A minimum of 50 000 and 10 000 cells were collected for striatum and SVZ samples, respectively. Cells were directly sorted into RLT buffer (Qiagen) containing 1% beta-Mercaptoethanol and were immediately frozen on powdered dry ice. RNA was extracted from these isolated microglia with the micro RNeasy kit (Qiagen) according to manufacturer's instructions.

and the SVZs were dissected from the lateral wall of the anterior horn of the lateral ventricle and cut into small explants (approx. 100–200 μm in diameter). These were then mixed with Matrigel (Corning) and cultured in four-well dishes. After polymerization (25 min), 500 μL of Neurobasal medium supplemented with B-27, N2-supplement, glutamine, and penicillin/streptomycin (all from Gibco-Life Technologies) were added. Cultures were maintained in a humidified, 5% CO2, 37 °C incubator with mouse CXCL13 (2, 5, 10, or 20 μg/mL; R&D Systems) or CXCL12 (50 ng/mL; R&D Systems). The length of migratory chains was measured from the edge of the explants at three angles using Image J software (NIH, Bethesda, MD, USA) after 24 h. 2.10. Statistical analysis

2.6. RNA extraction and quantitative PCR Total RNA was extracted from cells or tissue using a RNeasy Plus micro kit (Qiagen) or a RNeasy Lipid Tissue Mini Kit (Qiagen), respectively, and then reversed to cDNA using a iScript Advanced cDNA Synthesis Kit (Bio-rad). For quantitative PCR, TaqMan Gene expression master mix (Life Technologies) and TaqMan probe for β-actin and CXCL13 were used. The DNA band was examined after running in a 2% agarose gel at 90 mV for 1 h.

Comparisons were performed using Prism 6 software (GraphPad Software, Inc.) by one-way or two-way ANOVA followed by Bonferroni's post hoc test, or Student's unpaired t test. To achieve normal distribution of data, counts of DCX+ cells in stroke and LPS-treated groups values were subjected to log 10 transformation and parametric statistical analysis was then performed. Data are presented as means ± SEM, and differences are considered significant at p b 0.05. 3. Results

2.7. GeneChip microarray assay Sample preparation for microarray hybridization was carried out as described in the NuGEN Ovation Pico WTA System V2 and Encore Biotin Module manuals (NuGEN Technologies, Inc, San Carlos, CA). In brief, 2 to 10 ng of total RNA was reverse transcribed into double-stranded cDNA in a two-step process, introducing a SPIA tag sequence. Good quality of RNA and cDNA was confirmed by the company performing microarray analysis (KFB—Center of Excellence for Fluorescent Bioanalytics, Regensburg, Germany; www.kfb-regensburg.de). Bead-purified cDNA was amplified by a SPIA amplification reaction followed by an additional bead purification. 3.0 μg of SPIA cDNA were fragmented, terminally biotin-labeled and hybridized to an Affymetrix Rat Gene 1.1 ST Array Plate. For hybridization, washing, staining and scanning, an Affymetrix GeneTitan® system was used (Affymetrix, Inc., Santa Clara, CA). Sample processing was performed at an Affymetrix Service Provider and Core Facility, “KFB—Center of Excellence for Fluorescent Bioanalytics” (Regensburg, Germany; www.kfb-regensburg.de). 2.8. Global gene expression microarray analysis Summarized probe set signals were calculated by the RMA (Arvidsson et al., 2002) algorithm with the Affymetrix GeneChip Expression Console Software. Average signal values, comparison fold changes and significance P values were calculated using affy and limma package in R software by comparing SVZ/Striatum samples of LPS or MCAO groups with that of naïve groups. Only genes with changes greater than 1.5 fold and with an adjusted p value b 0.05 were identified. Genes coding secreted proteins were identified by GO:0005576. Common upregulated or downregulated genes in SVZ and striatum of LPS and MCAO groups were identified using VENNY software (http://bioinfogp.cnb.csic.es/tools/venny/index.html). Among genes that were significantly changed in striatal microglia sorted from MCAO or LPS condition compared with Naïve one, we identified factors that differed significantly between MCAO and LPS conditions (p b 0.05) and for which the fold difference was N1.5. 2.9. Migration assay and immunohistochemistry in SVZ explants This procedure has been described elsewhere (Wang et al., 2012). In brief, mouse pups were decapitated at postnatal days 4–5 (P4–P5), brains were removed, placed in ice-cold L-15 medium and cut into 250 μm sections on a vibratome. Sections containing SVZ were collected

3.1. Intrastriatal LPS injection causes massive inflammation without neuronal loss in rat striatum We first determined whether intracerebral administration of LPS could induce microglia/macrophage activation without causing neuronal cell death. In a pilot experiment, we injected different doses of LPS ranging from 0.01 to 100 μg into mouse striatum. Substantial microglia/macrophage activation without any neuronal death was found in the striatum 2 weeks after injection of 10 μg LPS in mice (data not shown). Because the neurogenic response after stroke is more robust in rats, we wanted to explore whether inflammation without injury could be triggered by LPS injection also in rats. We confirmed that we could induce microglia/macrophage activation without significant neuronal death in the rat striatum using 15 μg LPS. At 2 weeks after LPS injection, there was a 2-fold increase in the total number of Iba1 + microglia/macrophages compared to vehicle in rat striatum (Fig. 1A, B). The difference between LPS and vehicle did not remain at 6 weeks. The degree of brain damage was assessed using stereological methods to estimate the total number of NeuN + mature neurons in the striatum. No significant differences between LPS- and vehicleinjected animals were detected at either 2 or 6 weeks (Fig. 2A, B). We did notice, however, in two out of eight LPS-injected animals a thin line of neuronal death around the injection tract, which was not observed in any of the vehicle-injected animals (data not shown). We also measured stereologically the volume of the striatum in the animals injected with LPS and vehicle. We observed no differences in striatal volume either between ipsi- and contralateral sides in LPS-injected animals (15.9 ± 0.4 and 15.4 ± 0.4 mm3, respectively, at 2 weeks, and 14.4 ± 0.5 and 15.4 ± 0.7 mm3, respectively, at 6 weeks), or between LPS- and vehicle-injected animals ipsilateral to the injection (15.9 ± 0.4 and 17.3 ± 0.4 mm3, respectively, at 2 weeks, and 14.4 ± 0.5 and 15.7 ± 0.4 mm3, respectively, at 6 weeks). Striatal volume also did not differ between time points. Taken together, these findings do not provide any evidence for a significant loss of striatal volume, possibly caused by degradation of neuronal connections, after LPS injection. 3.2. Striatal neurogenesis is similar after stroke and LPS injection Stroke was induced by 30 min MCAO and lead to injury confined to the striatum, mostly in its dorsolateral part. Using stereological

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

5

Fig. 1. LPS-injection and MCAO cause massive inflammation characterized by microglia/macrophage and astrocyte activation in the rat striatum. A, Representative images of Iba1 + microglia/macrophages at 2 weeks after vehicle (Veh), LPS, Sham and MCAO treatment. B–D, Numbers of Iba1 + cells in the striatum at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. E, Astrocyte activation score after MCAO and LPS treatment at 2 and 6 weeks. Means ± SEM. 2 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 8; 6 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 4. * p b 0.05, One-way ANOVA with Bonferroni's post hoc test. Scale bar, 200 μm.

methods, we detected a loss of mature NeuN+ neurons in the striatum by 81.7% and 66.4% after MCAO compared to sham animals at 2 and 6 weeks, respectively (Fig. 2A, C). Iba1+ microglia/macrophage numbers were markedly increased at 2 weeks after stroke, but at 6 weeks had returned to sham levels (Fig. 1A, C). There was no difference in the number of Iba1 + cells between LPS and MCAO groups either at 2 or 6 weeks (Fig. 1D).

We assessed the activation of astrocytes after stroke and LPS injection using a semi-quantitative scale. At 2 weeks both MCAO and LPS had induced a marked increase in astrocyte activation compared with controls, which all scored zero (Fig. 1E). By 6 weeks, the level of activation had decreased in both LPS and MCAO animals but had not reached baseline levels. No differences in astrocyte activation scores were observed between MCAO and LPS animals at either time-point.

Fig. 2. LPS-injection causes no injury but MCAO leads to extensive neuronal death in the rat striatum. A, Representative images of NeuN-DAB staining at 2 weeks after Veh, LPS, Sham and MCAO treatment. B–D, Numbers of NeuN+ cells in the striatum at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. Means ± SEM. 2 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 8; 6 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 4. * p b 0.05, One-way ANOVA with Bonferroni's post hoc test. Scale bar, 0.5 mm. STR, striatum. Dashed line, infarct region.

6

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

For identifying newly generated cells, we used BrdU, which was given to the animals during the first two weeks after MCAO, LPS injection or control procedures. Both MCAO and LPS gave rise to the recruitment of large numbers of new DCX + neuroblasts into the striatum (Fig. 3A–C), whereas only very few DCX + cells were detected in animals subjected to sham surgery or vehicle injection. At 2 weeks, the majority of the DCX + cells co-expressed BrdU, indicating that they had been formed after the insults. The high number of DCX + cells was maintained at 6 weeks in both MCAO and LPS animals but at this time-point, 4 weeks after the cessation of BrdU injections, only a minority of DCX + cells were also BrdU +. Because DCX is expressed for only 2–3 weeks after neurons have been formed (Brown et al., 2003), our findings indicate that following LPS injection there is, as shown previously after 2 h MCAO (Thored et al., 2006), a long-lasting production of neuroblasts entering the striatum. No differences in the number of DCX + cells between MCAO and LPS animals were detected in the striatum either at 2 or 6 weeks (Fig. 3D). At both 2 and 6 weeks, the number of Ki67+ cells in SVZ was similar to that in controls following both stroke and LPS injections (Fig. 3E). Thus, we obtained no evidence of increased SVZ NSPC proliferation at

either of the two timepoints. However, whereas the area of SVZ was increased in MCAO compared to sham-operated animals, consistent with our previous report (Thored et al., 2006), we detected no difference in SVZ area between LPS- and vehicle-injected groups (Fig. 3F). We did not observe any DCX + cells in the striatum which co-expressed the oligodendrocyte progenitor marker PDGFRα or the astrocyte marker S100β following LPS injection, confirming that the DCX + cells were of neuronal lineage. Moreover, using confocal microscopy, we found that 94% of randomly selected DCX + cells also expressed the neuronal marker HuD (data not shown). We explored whether the insults had given rise to the formation of mature neurons by counting the number of NeuN +/BrdU + cells in the striatum. At 2 weeks such cells were scarce but at 6 weeks a substantial number of NeuN+/BrdU+ cells were found in both stroke and LPS groups (MCAO, 2 weeks: 6.01 ± 1.42, 6 weeks: 94.74 ± 49.43; LPS, 2 weeks: 0, 6 weeks: 34 ± 10.3, Fig. 4A–D). Thus, the neurogenic response after MCAO and LPS injection leads to the formation of mature neurons in the striatum. We estimated the percentage of neuroblasts giving rise to mature neurons after stroke and LPS injection by calculating the ratio between the number of NeuN +/BrdU + cells at 6 weeks and the number of DCX +/BrdU + cells at 2 weeks. Interestingly, we found a

Fig. 3. LPS-injection and MCAO trigger similar neuroblast recruitment to the striatum but no change in SVZ cell proliferation. A, Representative images of DCX+ neuroblast migration in the striatum at 2 weeks after Veh, LPS, Sham and MCAO treatment. B–D, Numbers of DCX+/BrdU+ and DCX+/BrdU− cells in the striatum at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. E, Numbers of Ki67+ cells in the SVZ at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. F, SVZ area at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. Means ± SEM. 2 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 8; 6 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 4. In B–D, values were subjected to log 10 transformation before statistical analysis. One-way ANOVA with Bonferroni's post hoc test. * p b 0.05, Scale bar, 100 μm. STR, striatum; LV, lateral ventricle.

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

7

Fig. 4. LPS-injection and MCAO lead to the formation of mature neurons in the rat striatum. A, Representative images of NeuN+/BrdU+ cells in the rat striatum at 6 weeks after LPS and MCAO treatment. B–D, Numbers of NeuN+/BrdU+ cells at 2 and 6 weeks after Veh, LPS, Sham and MCAO treatment. Means ± SEM. 2 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 8; 6 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 4. *p b 0.05, One-way ANOVA with Bonferroni's post hoc test. Scale bar, 10 μm.

Fig. 5. New neuroblasts and mature neurons are located closer to the SVZ in MCAO compared to LPS rats. Distribution of new neuroblasts in the striatum at 2 and 6 weeks (A, B) and mature neurons at 6 weeks (C) after LPS-injection and MCAO. Means ± SEM. 2 weeks: LPS n = 4, MCAO n = 8; 6 weeks: LPS n = 4, MCAO n = 4. *p b 0.05, Two-way ANOVA with Bonferroni's post hoc test.

8

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

substantially higher ratio following stroke (20%) than after LPS injection (5%), suggesting that the differentiation and/or survival of the newly formed neuroblasts differ between the two inflammatory environments. Finally, we compared the migration within the striatum of new neuroblasts and mature neurons generated after MCAO and LPS injection. Both at 2 and 6 weeks after the insult, a higher percentage of DCX+ cells were localized close to the SVZ in MCAO group compared with LPS group (Fig. 5A, B). Similarly, more NeuN+/BrdU+ cells were found close to the SVZ at 6 weeks in the stroke-affected than in the LPS-injected rats (Fig. 5C). This reduction in neuroblast migration is unlikely to be caused by the occurrence of a lesion in MCAO-subjected rats, as there was no correlation between lesion size and the percentage of DCX+ cells close to SVZ (2 week: r = −0.5362, p = 0.1707; 6 week: r = 0.5161, p = 0.4839). 3.3. Inflammation after LPS injection and stroke is associated with changes in microglia phenotype in both subventricular zone and striatum We previously observed a long-lasting increase in microglia/ macrophage numbers in the SVZ following 2 h MCAO (Thored et al., 2009). Here we found that following 30 min MCAO and LPS injection, the number of Iba1 + microglia/macrophages in the SVZ did not differ from sham and vehicle groups, respectively, at either 2 or 6 weeks (data not shown). We compared the morphological

characteristics of microglia between MCAO, LPS and respective controls in both ipsilateral striatum and SVZ. Microglia were classified into one of four increasingly activated phenotypic states: from ramified through intermediate, and amoeboid to round (Lehrmann et al., 1997). In both control groups, close to 100% of microglia were classified as ramified at both time-points in both brain regions. In contrast, both MCAO and LPS administration lead to a substantial switch toward the more activated amoeboid/round phenotypes in both striatum and SVZ at 2 weeks (Fig. 6A–C). In both treatment groups, microglia/macrophages shifted to a less activated profile over time in the striatum, whereas the elevated levels of activation were virtually unchanged at 6 weeks in the SVZ. Because of the difference between MCAO and LPS groups in the survival/differentiation of the newly formed neurons at 6 weeks, we also assessed the densities of pro-inflammatory microglia/macrophages in the striatum. However, at both 2 and 6 weeks after stroke and LPSinjection, we found no difference between the two groups in the densities of iNOS+ pro-inflammatory M1 microglia/macrophages (MCAO, 2 weeks: 30.79 ± 19.49, 6 weeks: 2.51 ± 0.94; LPS, 2 weeks: 28.4 ± 6.49, 6 weeks: 4.16 ± 1.42 cells/mm2). 3.4. Microglia sorting and subsequent global gene expression analysis identify potential regulators of stroke- and LPS-induced striatal neurogenesis To identify microglia-derived factors that could potentially regulate striatal neurogenesis after both MCAO and LPS-injection, we first

Fig. 6. LPS-injection and MCAO give rise to increased percentage of activated microglia/macrophages in both the striatum and SVZ. A, Representative images of Iba1+ ramified, intermediate, amoeboid and round microglia/macrophages. B, C, Percentage of ramified, intermediate, amoeboid and round microglia/macrophages in the SVZ and striatum at 2 and 6 weeks after Veh, LPS, sham and MCAO treatment. Means ± SEM. 2 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 8; 6 weeks: Veh n = 4, LPS n = 4, Sham n = 4, MCAO n = 4. *p b 0.05, Two-way ANOVA with Bonferroni's post hoc test. Scale bar, 10 μm.

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

purified, using percoll gradients and FACS, microglia from the SVZ and striatum of naïve, MCAO and LPS-injected rat brains using CD45low CD11b + as microglial markers 2 weeks after stroke and LPS injection. For the microglia/macrophages sorted from the striatum, 98.8%, 95.7% and 90% of the total number of CD11b + microglia/macrophages were CD45lowCD11b + microglia in naïve, MCAO and LPS groups, respectively (Fig. 7A). The RNA of sorted cells was then extracted and microarray analysis was performed using Affymetrix Rat Gene 1.1 ST Array. We hypothesized that genes coding for secreted factors that are changed in microglia sorted from striatum and/or SVZ of both LPSinjected and MCAO-subjected rats could potentially regulate different steps of neurogenesis. We identified several secreted protein-coding genes that were either upregulated or downregulated in microglia sorted from both SVZ and striatum (Table 2), only in microglia sorted from SVZ (Table 3) or only in microglia sorted from striatum (Table 4), in both MCAO-subjected and LPS-injected rats compared to the naïve condition. We also identified genes in sorted striatal microglia for which expression levels differed significantly between MCAO and LPS conditions (Table 5). Among the changed genes, we focused on one factor, CXCL13, which was upregulated in microglia sorted from both SVZ and striatum of both MCAO and LPS groups. CXCL13 has been found to promote human neural progenitor cell migration in vitro (Weiss et al., 2010). Moreover, since CXCL13 belongs to the same superfamily as CXCL12, a well-established stimulator of neuroblast migration (Thored et al., 2006), we hypothesized that microglia-derived CXCL13 might induce striatal neuroblast migration after MCAO. We confirmed that Cxcl13 mRNA was also expressed on CD45lowCD11b + microglia sorted from mouse striatum after stroke (Fig. 7B), and then dissected the role of CXCL13 on neuroblast migration using in vitro and in vivo mouse models (see below). We were also interested in the expression of Cysteinyl leukotriene receptor 1 (CysLT1) signaling molecules, since CysLT1 signaling has been reported to be involved in regulating progenitor proliferation as well as neurogenesis after traumatic brain injury in zebrafish (Kyritsis et al., 2012). To test whether signaling by Leukotriene C4 through

9

CysLT1 is involved in the neuroblast recruitment observed after MCAO and LPS injection, we analyzed expession of Ltc4s, the gene coding the enzyme LTC4S for LTC4 synthesis, in our microarray data. However, in sorted SVZ rat microglia we found that the mRNA level of Ltc4s was not changed in MCAO and downregulated in LPS compared with naïve conditions. In striatum samples, Ltc4s mRNA level was downregulated in both MCAO and LPS groups (Fig. 7C). 3.5. CXCL13 has a mild stimulatory effect on migration of subventricular zone-derived neuroblasts in vitro but neuroblast recruitment and migration in striatum are unimpaired in vivo in CXCR5−/− mice following stroke In order to examine the role of CXCL13 on neuroblast migration, we cultured SVZ explants from neonatal mice in the presence of CXCL13 or CXCL12 (Fig. 8A). After a 24 h incubation, we measured the length of the neuroblast migratory chains extending from the explants. CXCL13 increased chain length with doses higher than 5 μg/ml. We found that CXCL12 was considerably more potent in this assay, inducing a similar level of neuroblast migration at a dose of 50 ng/ml (Fig. 8B). The in vitro migration assay indicated that CXCL13 might be involved in neuroblast migration after stroke. While CXCL13 has been shown to bind also to CXCR3, it is the only known ligand for CXCR5 (Jenh et al., 2001). We therefore took advantage of CXCR5 knockout mice for exploring the role of CXCL13–CXCR5 signaling in striatal neuroblast migration in vivo. CXCR5−/− and wild-type mice were subjected to 35 min MCAO, which gives rise to ischemic injury in the striatum and overlying cerebral cortex, and were sacrificed after 2 weeks. Because CXCR5 has been reported to alter ventricular cell proliferation in zebrafish (Kizil et al., 2012), we analyzed cell proliferation in the SVZ of the two animal groups following stroke. However, no differences in number of Ki67+ cells were detected between ipsi- and contralateral SVZ or between CXCR5−/− and wild-type mice (CXCR5−/−, ipsilateral: 805 ± 83.29, contralateral: 831 ± 26.64; wild-type, ipsilateral: 794 ± 37.15, contralateral: 721 ± 43.03). We also found that the deletion of the Cxcr5 gene did not lead to any difference in the magnitude of

Fig. 7. Microglia sorting and subsequent global gene expression analysis identify potential regulators of MCAO- and LPS-induced striatal neurogenesis. A, Representative FACS plots of rat striatal CD45lowCD11b+ microglia sorted using PE-conjugated anti-CD45 and APC− conjugated anti-CD11b antibodies in naïve, MCAO and LPS rats 2 weeks after treatment. B, Q-PCR gel band of Cxcl13 mRNA in CD45lowCD11b+ microglia sorted from striatum of MCAO-subjected mice with spleen tissue as positive control. C, Relative Ltc4s mRNA level in microglia sorted from SVZ and striatum of naïve, MCAO-subjected, and LPS-injected rats. Means ± SEM. Naïve n = 4, LPS n = 4, MCAO n = 4, *p b 0.05, One-way ANOVA with Bonferroni's post hoc test.

10

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

Table 2 Genes coding for secreted proteins that are upregulated or downregulated (p b 0.05, fold change N1.5) in microglia sorted from both the SVZ and striatum in both MCAO and LPS conditions compared with the naïve condition. Genes listed alphabetically. Gene name

Gene full name

Gene accession

Fold change in MCAO (SVZ)

Fold change in MCAO (Striatum)

Fold change in LPS (SVZ)

Fold change in LPS (Striatum)

Upregulated genes Adm Adrenomedullin Anxa2 Annexin A2 Apoc1 Apolipoprotein C-I C3 Complement component 3 Csf1 Colony stimulating factor 1 (macrophage) Cxcl13 Chemokine (C–X–C motif) ligand 13 F10 Coagulation factor X Fgf2 Fibroblast growth factor 2 Glipr1 GLI pathogenesis-related 1 Il1rn Interleukin 1 receptor antagonist Lgals3bp Lectin, galactoside-binding, soluble, 3 binding protein Npb Neuropeptide B Plau Plasminogen activator, urokinase Tnfsf12 Tumor necrosis factor ligand superfamily member 12

NM_012715 NM_019905 NM_012824 NM_016994 NM_023981 NM_001017496 NM_017143 NM_019305 NM_001011987 NM_022194 NM_139096 NM_153293 NM_013085 NM_001001513

4.87 12.79 6.74 1.98 10.30 3.64 3.11 6.39 3.80 7.19 5.61 2.08 7.74 2.08

4.44 10.50 10.41 2.18 15.22 20.55 3.12 5.25 5.78 6.46 4.64 2.61 11.76 1.98

3.51 2.88 12.94 3.49 2.76 106.77 3.10 1.54 2.03 12.33 4.87 3.67 3.02 2.37

2.45 3.45 18.90 3.31 3.19 326.08 3.26 1.89 3.31 9.13 3.62 9.05 2.97 1.85

Downregulated genes Crispld1 Cysteine-rich secretory protein LCCL domain containing 1 Enpp2 Ectonucleotide pyrophosphatase/phosphodiesterase 2 Fam5b Family with sequence similarity 5, member B Glipr2 GLI pathogenesis-related 2 Grem1 Gremlin 1, cysteine knot superfamily, homologue (Xenopus laevis) Hbegf Heparin-binding EGF-like growth factor Il6 Interleukin 6 Slit2 Slit homologue 2 (Drosophila)

NM_001134963 NM_057104 NM_173115 ENSRNOT00000019916 NM_019282 NM_012945 NM_012589 NM_022632

2.99 1.94 2.20 1.87 1.80 2.73 4.71 2.61

3.94 2.20 3.70 2.85 3.42 3.40 5.80 2.90

6.59 2.41 4.85 2.56 1.57 3.42 4.51 5.35

3.48 1.53 4.11 2.59 3.14 3.44 4.38 3.46

neuroblast recruitment to the striatum after stroke (CXCR5−/−: 201 ± 89.36, wild-type: 212.4 ± 40.79). We then counted the DCX + cell distribution pattern in three different 250 μm wide columns extending laterally from the SVZ. In contrast to in vitro data, we could not detect any effect of deleting the Cxcr5 gene on neuroblast migration in vivo (data not shown). 3.6. Ischemic injury and M1 density of microglia/macrophages in striatum are increased in CXCR5−/− mice following stroke Finally, we determined whether deletion of CXCR5 influenced the size of the ischemic lesion and characteristics of the inflammatory response following stroke. Interestingly, CXCR5−/− mice showed a 45.5% larger injury volume compared to wild-type animals (CXCR5 −/−: 11.64 ± 0.56 mm 3 , wild-type: 6.341 ± 1.67 mm3 ; Fig. 9A). In both wild-type and CXCR5−/− mice, the ischemic lesion extended to but never involved the SVZ (4 out of 6 wild-type mice, 2 out of 4 CXCR5−/− mice). In agreement, we detected no difference in SVZ size between the two animal groups (wild-type: 0.074 ± 0.007, CXCR5 −/−: 0.071 ± 0.003, mm2). There was no difference Table 3 Genes coding for secreted proteins that are upregulated or downregulated (p b 0.05, fold change N1.5) only in SVZ microglia (not in striatal microglia), in both MCAO and LPS conditions compared with the naïve condition. Genes listed alphabetically. Gene name

Gene full name

Upregulated genes Aebp1 AE binding protein 1 Apoa1bp Apolipoprotein A-I binding protein Dpp7 Dipeptidylpeptidase 7 Plaur Plasminogen activator, urokinase receptor Tuba4a Tubulin, alpha 4A Downregulated genes Il1a Interleukin 1 alpha Pdyn Prodynorphin Vegp2 Von Ebners gland protein 2

Gene accession

Fold change Fold change in MCAO (SVZ) in LPS (SVZ)

NM_001100970 2.05 NM_001106440 2.28

2.96 1.88

NM_031973 NM_134352

3.27 1.73

1.55 2.75

NM_001007004 3.73

2.20

NM_017019 NM_019374 NM_053574

3.96 2.09 1.61

2.08 1.66 1.63

between the two groups in the density of either Iba1+ or Iba1+/ED1+ microglia/macrophages in the ischemic striatum (CXCR5−/−, Iba1+: 1309 ± 137.6, Iba1+/ED1+: 357.2 ± 10.54; wild-type, Iba1+: 1452 ± 218.4, Iba1 +/ED1 +: 395.1 ± 81.36 cells/per mm2 ). We also found no difference in astrocyte activation between wild-type and CXCR5−/− mice using the semi-quantitative scale (wild-type: 1.889 ± 0.111, CXCR5−/−: 2.0 ± 0.136). However, in CXCR5−/− mice we found a markedly higher density of CD16/32 + M1 microglia/ macrophages (CXCR5−/− : 868.8 ± 124.1, wild-type: 437.8 ± 99.07; Fig. 9B), almost all of which were Iba1 + (96%, Fig. 9C). Thus, deletion of CXCR5 gene caused a more severe injury and a more pro-inflammatory environment in the ischemic striatum.

4. Discussion We show here that brain inflammation is sufficient for inducing a regenerative response from NSPCs located in the SVZ of the adult rodent brain. LPS-induced striatal inflammation without neuronal loss triggered the migration of neuroblasts into the striatum and their differentiation into mature neurons. The magnitude of neuroblast production in response to inflammation was similar to that observed after a 30 min MCAO, which caused major neuronal loss in the striatum. Stroke and LPS-induced inflammation gave rise to a long-lasting (at least 6 weeks) recruitment of neuroblasts into the striatum, indicating that the molecular mechanisms regulating the neurogenic response after both insults are still operating at 6 weeks. It should be emphasized that LPS and MCAO are known to trigger inflammation via different mechanisms but through some overlapping signaling pathways. LPS interacts with LPS-binding protein (LBP) to form LPS–LBP complex, which is recognized by toll-like receptor 4 (TLR4) with two co-receptors (CD14 and MD2) leading to nuclear factor-κB (NF-κB), mitogenactivated protein kinases (MAPK) and other signaling molecules activation then induction of pro-inflammatory genes (Morris et al., 2014). After stroke, on the other hand, various damage-associated molecular patterns (DAMPs) released from dead cells are recognized by pattern recognition receptors (PRRs) such as TLRs and scavenger receptors (Iadecola and Anrather, 2011). Activation of these receptors triggers upregulation of pro-inflammatory cytokines and chemokines

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

11

Table 4 Genes coding for secreted proteins that are upregulated or downregulated (p b 0.05, fold change N1.5) only in striatal microglia (not in SVZ microglia), in both MCAO and LPS conditions compared with the naïve condition. Genes listed alphabetically. Gene name

Gene accession

Fold change in MCAO (striatum)

Fold change in LPS (striatum)

Upregulated genes Acpp Acid phosphatase, prostate Angptl4 Angiopoietin-like 4 Cd14 CD14 molecule Cxcl10 Chemokine (C–X–C motif) ligand 10 Cxcl9 Chemokine (C–X–C motif) ligand 9 Il1b Interleukin 1 beta Olr1 Oxidized low density lipoprotein (lectin-like) receptor 1 Rnpep Arginyl aminopeptidase (aminopeptidase B) Serping1 Serine (or cysteine) peptidase inhibitor, clade G, member 1 Slc1a5 Solute carrier family 1 (neutral amino acid transporter), member 5 Timp1 TIMP metallopeptidase inhibitor 1

Gene full name

NM_020072 NM_199115 NM_021744 NM_139089 NM_145672 NM_031512 NM_133306 NM_031097 NM_199093 NM_175758 NM_053819

4.75 2.09 3.02 2.56 1.87 2.47 5.96 3.06 3.00 3.42 3.81

1.76 2.04 8.88 3.55 2.51 5.08 1.93 1.75 2.76 2.53 2.47

Downgulated genes Epdr1 Ependymin related protein 1 (zebrafish) Il4ra Interleukin 4 receptor, alpha Il6r Interleukin 6 receptor, alpha Resp18 Regulated endocrine-specific protein 18

NM_001007625 NM_133380 NM_017020 NM_019278

2.33 1.65 1.77 2.14

2.00 2.54 2.44 1.86

via NF-κB, MAPK, type I interferon and other signaling pathways (Chen and Nunez, 2010). We found that the level of microglia/macrophage activation, as assessed by counts of Iba1 + cells, was similar after LPS and stroke. After LPS injection and 30 min MCAO, we did not observe any increased number of microglia/macrophages in the SVZ as previously shown after 2 h MCAO (Thored et al., 2009). In the striatum, microglia/macrophages numbers were transiently increased at 2 weeks but had returned to control levels at 6 weeks. The increased microglia/macrophage activation in the striatum after both insults at 2 weeks had shifted to a less activated profile at 6 weeks, whereas activation was maintained in the SVZ at this time-point. Moreover, our microarray data showed that around 40% of the genes with significant changes in microglia from striatum and SVZ after LPS treatment are also altered in the same direction after stroke, indicating partial similarity of microglia activation pattern after LPS and MCAO. Based on these findings, we hypothesize that also the inflammation associated with stroke could be essential for triggering striatal neurogenesis. However, whether this is the case has to be determined in future studies. Cell proliferation in the SVZ was similar to control at both timepoints after stroke and LPS, probably because the increased proliferation triggered by the insults had occurred before 2 weeks (Thored et al., 2006; Zhang et al., 2004). The new neuroblasts migrated less in the striatum in animals subjected to MCAO than in those given LPS injection, with more cells remaining close to the SVZ. We obtained no evidence that this was due to stroke-induced striatal injury hindering the migration of neuroblasts. Interestingly, the microarray analysis indicated that both Ccl2, a factor previously reported to recruit neuroblasts into ischemic striatum (Yan et al., 2007), and Cxcl13 (see below) had higher expression level in microglia sorted from striatum after LPS compared with MCAO condition (Table 5). Besides, the spatial arrangement of specific reactive astrocytes and their barrier formation and secreted molecules might also contribute to the extent of migration of SVZ neuroblasts (Wanner et al., 2013). Hypothetically, differences between MCAO and stroke in levels of chemokines recruiting neuroblasts and the barrier characteristics of the astroglial response could underlie the differences in migration of the newly formed neurons. When we estimated the percentage of neuroblasts giving rise to mature neurons after stroke and LPS delivery, we found substantially higher value following stroke (20%) than after LPS (5%). Although the inflammatory environments induced by LPS-injection and MCAO share many common features, we also observed discrepancies that most likely underlie the differences in differentiation/survival of the newly formed neuroblasts and neurons. For example, the microarray analysis provided evidence that the expression of the gene coding

IL-1β, a factor detrimental for neuronal survival (Rothwell, 2003), was higher in microglia sorted from striatum of LPS-injected as compared to those from MCAO-subjected rats. In contrast, gene expression of IGF-1, a factor beneficial for neuronal survival (Zheng et al., 2000), was higher in microglia from striatum of stroke-affected compared with LPS-injected rats (Table 5). In agreement with our results, Kyritsis and collaborators (Kyritsis et al., 2012) recently demonstrated that not only traumatic brain injury, but also the induction of an inflammatory response without cell death, gave rise to enhanced neurogenesis in the adult zebrafish brain. LTC4 signaling was found to be the major mediator, and the injection of LTC4 increased progenitor proliferation and neurogenesis. Importantly, inflammation was shown to initiate the cellular regenerative response, which was distinct from constitutive neurogenesis. Our study provides the first experimental evidence for a similar role of inflammation in triggering neurogenesis in the adult mammalian brain. It is unlikely however, that this effect is mediated by LTC4 released from microglia or other immune cells. First, our global gene expression data indicated that the expression of the LTC4 synthase gene in sorted microglia was either unchanged or decreased in the SVZ and striatum. Second, whereas radial glial cells in the zebrafish predominantly express the CysLT1 receptor (Kyritsis et al., 2012), adult rat NSPCs express the leukotriene receptor GPR17 but CysLT1 is undetectable (Huber et al., 2011). Administering the CysLT1 antagonist Pranlukast in zebrafish reduced proliferation and neurogenesis (Kyritsis et al., 2012). In contrast, the inhibitor of GPR17 and CysLT1, Montelukast, strongly elevated rat NSPCs proliferation in vitro (Huber et al., 2011). We have focused on the role of microglia/macrophages, as a continuation of our previous work (Thored et al., 2009), in which we demonstrated long-term accumulation of microglia/macrophages in the SVZ concomitant with persistent striatal neurogenesis after stroke. However, the astrocyte activation as part of the neuroinflammation after both LPS and stroke could also be involved in the regulation of neurogenesis. In line with this idea, Zamanian et al. found a core of gene expression changes (166 genes with more than 2-fold upregulation) in reactive mouse brain astrocytes, which were shared between ischemic stroke and systemic LPS injections (Zamanian et al., 2012). Possibly, such common changes could contribute to triggering the neurogenic responses observed after both stroke and LPS. The microarray data on sorted microglia from the SVZ and striatum pointed to a number of secreted factors which were altered in both MCAO and LPS conditions and could potentially act as regulators of striatal neurogenesis. Several of the upregulated factors in both SVZ and striatum have previously been reported to influence SVZ NSPC proliferation/differentiation and adult neurogenesis (Table 2). For

12

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

Table 5 Genes coding for secreted proteins that are changed in MCAO or LPS condition compared with naïve one (p b 0.05, fold change N1.5) and exhibit different expression levels between MCAO and LPS conditions (p b 0.05, fold difference N1.5) in microglia sorted from striatum. Genes listed alphabetically. Gene name

Gene accession

Fold difference

Higher expression in LPS compared to MCAO C3 Complement component 3 Capg Capping protein (actin filament), gelsolin-like Ccl12 Chemokine (C–C motif) ligand 12 Ccl17 Chemokine (C–C motif) ligand 17 Ccl2 Chemokine (C–C motif) ligand 2 Ccl22 Chemokine (C–C motif) ligand 22 Ccl7 Chemokine (C–C motif) ligand 7 Cd14 CD14 molecule Cfp Complement factor properdin Cp Ceruloplasmin Cx3cl1 Chemokine (C–X3–C motif) ligand 1 Cxcl13 Chemokine (C–X–C motif) ligand 13 Dpp4 Dipeptidylpeptidase 4 Fas Fas (TNF receptor superfamily, member 6) Fuca2 Fucosidase, alpha-L-2, plasma Hpse Heparanase Htra1 HtrA serine peptidase 1 Il15 Interleukin 15 Il1b Interleukin 1 beta Lcn2 Lipocalin 2 Lif Leukemia inhibitory factor Lyz2 Lysozyme 2 Meteorin, glial cell differentiation regulator-like Metrnl Npb Neuropeptide B Pcsk1n Proprotein convertase subtilisin/kexin type 1 inhibitor Pf4 Platelet factor 4 Plbd1 Phospholipase B domain containing 1 Prf1 Perforin 1 (pore forming protein) Ptn Pleiotrophin S100a9 S100 calcium binding protein A9 Serpini1 Serine (or cysteine) peptidase inhibitor, clade I, member 1 Smpdl3a Sphingomyelin phosphodiesterase, acid-like 3A Wfdc21 WAP four-disulfide core domain 21

Gene full name

NM_016994 NM_001013086 NM_001105822 NM_057151 NM_031530 NM_057203 NM_001007612 NM_021744 NM_001106757 NM_012532 NM_134455 NM_001017496 NM_012789 NM_139194 NM_001004218 NM_022605 NM_031721 NM_013129 NM_031512 NM_130741 NM_022196 NM_012771 NM_001014104 NM_153293 NM_019279 NM_001007729 NM_001013927 NM_017330 NM_017066 NM_053587 NM_053779 NM_001005539 NM_001003706

LPS vs MCAO 1.52 2.13 6.12 1.73 3.15 4.85 5.85 2.95 2.99 4.27 14.15 15.87 1.76 1.93 1.59 9.15 1.62 2.17 2.05 10.77 1.65 2.13 1.58 3.46 2.15 4.36 12.23 1.85 1.57 5.04 2.82 3.97 2.84

Higher expression in MCAO compared to LPS Adm Adrenomedullin Ang1 Angiogenin, ribonuclease A family, member 1 Anxa2 Annexin A2 C1r Complement component 1, r subcomponent C1rl Complement component 1, r subcomponent-like C4b Complement component 4, gene 2 Cd34 CD34 molecule Cfd Complement factor D (adipsin) Chad Chondroadherin Clstn1 Calsyntenin 1 Crim1 Cysteine rich transmembrane BMP regulator 1 (chordin like) Csf1 Colony stimulating factor 1 (macrophage) Cxcl14 Chemokine (C–X–C motif) ligand 14 Dpp7 Dipeptidylpeptidase 7 Edn3 Endothelin 3 Fcrla Fc receptor-like A Fgf2 Fibroblast growth factor 2 Gas6 Growth arrest specific 6 Glipr1 GLI pathogenesis-related 1 Htra3 HtrA serine peptidase 3 Igf1 Insulin-like growth factor 1 Il1rl1 Interleukin 1 receptor-like 1 Interleukin 33 Il33 Lamb2 Laminin, beta 2 Ltbp3 Latent transforming growth factor beta binding protein 3 Mfge8 Milk fat globule-EGF factor 8 protein Olr1 Oxidized low density lipoprotein (lectin-like) receptor 1 Osm Oncostatin M Pdgfa Platelet-derived growth factor alpha polypeptide Plau Plasminogen activator, urokinase Pnp Purine nucleoside phosphorylase Pros1 Protein S (alpha) Pycard PYD and CARD domain containing Qsox1 Quiescin Q6 sulfhydryl oxidase 1 Rnase10 Ribonuclease, RNase A family, 10 (non-active) Rnase4 Ribonuclease, RNase A family 4 St14 Suppression of tumorigenicity 14 (colon carcinoma) St3gal2 ST3 beta-galactoside alpha-2,3-sialyltransferase 2 Tac4 Tachykinin 4

NM_012715 NM_001006992 NM_019905 NM_001134555 NM_001002804 NM_001002805 NM_001107202 NM_001077642 NM_019164 ENSRNOT00000022100 NM_001169103 NM_023981 NM_001013137 NM_031973 NM_001077650 NM_001100682 NM_019305 NM_057100 NM_001011987 ENSRNOT00000010852 NM_001082478 NM_001127689 NM_001014166 NM_012974 NM_001191561 NM_001040186 NM_133306 NM_001006961 NM_012801 NM_013085 NM_001106031 NM_031086 NM_172322 NM_001109898 NM_001012467 NM_020082 NM_053635 NM_031695 NM_172328

MCAO vs LPS 1.82 3.15 3.04 4.64 1.93 5.53 18.37 2.66 7.17 1.73 2.85 4.76 3.24 2.11 2.11 4.83 2.78 2.64 1.75 24.47 5.29 2.27 2.14 2.05 2.10 2.43 3.09 2.19 1.57 3.96 1.84 2.02 1.94 2.00 2.70 4.94 5.66 2.59 1.78

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

13

Table 5 (continued) Gene name

Gene full name

Higher expression in MCAO compared to LPS Tgfb1 Transforming growth factor, beta 1 Timp1 TIMP metallopeptidase inhibitor 1 Vcl Vinculin Vegfb Vascular endothelial growth factor B

example, FGF2 has been shown to maintain a slow-dividing neural stem cell pool in adult mouse SVZ (Zheng et al., 2004). Another factor, TNFSF12, inhibits adult mouse SVZ NSPC proliferation and promotes neuronal differentiation (Scholzke et al., 2011). Moreover, Complement 3 (C3) knockout mice show decreased ischemia-induced neurogenesis in SVZ (Rahpeymai et al., 2006). Among the downregulated genes in both structures (Table 2), IL-6 has been demonstrated to inhibit hippocampal neurogenesis in vitro (Monje et al., 2003), and SLIT2 was identified as negative regulator of neurite growth (Byun et al., 2012). If the gene changes are translated to the corresponding alterations in protein levels, these factors could contribute to the occurrence of striatal neurogenesis in both conditions. Interestingly, our microarray analysis found that Timp1, a factor identified as inhibiting the Notch shedding enzyme ADAM10 (Rapti et al., 2008; Muraguchi et al., 2007; Amour et al., 2000), was upregulated in microglia sorted from the striatum of LPS-injected and MCAO-subjected rats (Table 4). As Notch signaling was recently shown to inhibit striatal neurogenesis from local astrocytes in mice (Magnusson et al., 2014), our findings suggest that microglia-derived factors might be involved in regulating this neurogenic program in striatal astrocytes following stroke. An upregulation of CXCL13 was detected in microglia isolated from SVZ and striatum after both stroke and LPS. Previous studies have shown that CXCL13 via its receptor CXCR5 is a potent chemokine for recruitment of human neural precursor cells across brain endothelial

Gene accession

Fold difference

NM_021578 NM_053819 NM_001107248 NM_053549

2.06 1.54 1.97 1.64

cells (Weiss et al., 2010). Moreover, neuroblastoma cells expressing CXCR5 have been found to migrate toward CXCL13 in vitro (Airoldi et al., 2008; Del Grosso et al., 2011). In accordance, we found that CXCL13 promoted neuroblast migration from neonatal mouse SVZ explants, although the effect was weaker than that of CXCL12, which has been shown to promote neuroblast migration after stroke (Thored et al., 2006). Inconsistent with the in vitro data, the number of neuroblasts recruited from SVZ to the striatum and their migration toward the injury following stroke did not differ between the CXCR5−/− and wild-type mice. This is in contrast to a recent report showing that intact CXCR5−/− mice exhibited increased numbers of DCX+ cells but decreased proliferation in the dentate gyrus (Stuart et al., 2014). Meanwhile, using immunohistochemistry we found that CXCR5 protein co-localized with DCX in neuroblasts from P5 mouse SVZ explants (data not shown), but we could not detect either CXCR5 mRNA using Q-PCR in DCX+ neuroblasts sorted from the SVZ of intact DCX-GFP mice or CXCR5 protein on DCX + neuroblasts in SVZ of adult mice subjected to stroke with immunohistochemistry (data not shown). Taken together, these findings suggest that CXCR5 expression is reduced to undetectable levels in SVZ neuroblasts in adult mice, probably causing the discrepancies between our in vitro and in vivo results. The CXCR5−/− mice exhibited a much larger ischemic injury in both the striatum and cerebral cortex and had microglia/macrophages with a more proinflammatory phenotype, illustrated by a higher M1 density,

Fig. 8. CXCL13 promotes neuroblast migration from neonatal mouse SVZ explants. A, Representative images of neuroblasts migrating from SVZ explants at 24 h after Control (PBS), CXCL13 and CXCL12 treatment. B, Length of neuroblast migratory chain after Control, CXCL13 and CXCL12 treatment. Means ± SEM. Control n = 27; CXCL13 2 μg/ml n = 24, 5 μg/ml n = 14, 10 μg/ml n = 12, 20 μg/ml n = 8; CXCL12 n = 18. *p b 0.05, One-way ANOVA with Bonferroni's post hoc test. Scale bar, 100 μm.

14

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15

Fig. 9. CXCR5−/− mice exhibit larger lesion and more pro-inflammatory microglia/macrophage profile after MCAO. Quantification and representative images of NeuN-DAB staining (A) and CD16/32+ M1 microglia/macrophages (B) at 2 weeks after 35 min MCAO in the brains of wild-type and CXCR5−/− mice. Representative images of CD16/32+/Iba1+ microglia/macrophages (C). Scale bars, 0.5 mm (A), 10 μm (B, C). STR, striatum. Dashed line, infarct region.

compared to wild-type animals after stroke. CXCR5 is expressed on virtually all B lymphocytes (Kim et al., 2001), and CXCL13 signaling through CXCR5 is chemotactic for B lymphocytes and controls their circulation and redistribution in lymphoid organs (Legler et al., 1998; Gunn et al., 1998). B lymphocytes are recruited to the ischemic hemisphere after stroke (Ren et al., 2011; Lehmann et al., 2014), and regulatory B cells secreting IL-10 have been reported to limit the infarct volume and inhibit microglia/macrophage activation after MCAO in mice (Ren et al., 2011). Based on the similar phenotype in CXCR5−/− and B cell-deficient μMT−/− mice after stroke (Ren et al., 2011), it is tempting to speculate that the neuroprotective action of CXCL13– CXCR5 signaling, for the first time indicated by our findings, is mediated by B cells. Although previous studies have established that inflammation can influence different steps of neurogenesis, i.e., cell proliferation, migration, differentiation, survival and functional synaptic integration, our study provides the first experimental evidence that inflammation per se triggers a neurogenic response in the mammalian brain. These findings support the notion that the cross-talk between immune cells and NSPCs and their progeny is of crucial importance for efficient neuroregenerative responses following disease and damage in the adult brain (Kokaia et al., 2012). Better understanding of the molecular mechanisms involved in this dialog could provide novel therapeutic targets to promote more efficient repair of the damaged and degenerated brain.

Conflict of interest disclosure The authors declare no conflicts of interests.

Acknowledgments We thank Stefan Lang for assisting with microarray analysis, Zhi Ma and Teona Roschupkina for help with cell sorting, Zhaolu Wang for help with cell counting, and Linda Jansson for technical assistance. This work was supported by the Swedish Research Council, the European Union project TargetBraIn (279017), AFA Foundation (100231), Torsten Söderberg Foundation and Swedish Government Initiative for Strategic Research Areas (StemTherapy).

References Airoldi, I., Cocco, C., Morandi, F., Prigione, I., Pistoia, V., 2008. CXCR5 may be involved in the attraction of human metastatic neuroblastoma cells to the bone marrow. Cancer Immunol. Immunother. 57 (4), 541–548. Amour, A., Knight, C.G., Webster, A., Slocombe, P.M., Stephens, P.E., Knauper, V., et al., 2000. The in vitro activity of ADAM-10 is inhibited by TIMP-1 and TIMP-3. FEBS Lett. 473 (3), 275–279. Arvidsson, A., Collin, T., Kirik, D., Kokaia, Z., Lindvall, O., 2002. Neuronal replacement from endogenous precursors in the adult brain after stroke. Nat. Med. 8 (9), 963–970. Arvidsson, A., Kirik, D., Lundberg, C., Mandel, R.J., Andsberg, G., Kokaia, Z., et al., 2003. Elevated GDNF levels following viral vector-mediated gene transfer can increase neuronal death after stroke in rats. Neurobiol. Dis. 14 (3), 542–556. Brown, J.P., Couillard-Despres, S., Cooper-Kuhn, C.M., Winkler, J., Aigner, L., Kuhn, H.G., 2003. Transient expression of doublecortin during adult neurogenesis. J. Comp. Neurol. 467 (1), 1–10. Byun, J., Kim, B.T., Kim, Y.T., Jiao, Z., Hur, E.M., Zhou, F.Q., 2012. Slit2 inactivates GSK3beta to signal neurite outgrowth inhibition. PLoS ONE 7 (12), e51895. Cardona, A.E., Huang, D., Sasse, M.E., Ransohoff, R.M., 2006. Isolation of murine microglial cells for RNA analysis or flow cytometry. Nat. Protoc. 1 (4), 1947–1951. Chen, G.Y., Nunez, G., 2010. Sterile inflammation: sensing and reacting to damage. Nat. Rev. Immunol. 10 (12), 826–837. Del Grosso, F., Coco, S., Scaruffi, P., Stigliani, S., Valdora, F., Benelli, R., et al., 2011. Role of CXCL13–CXCR5 crosstalk between malignant neuroblastoma cells and Schwannian stromal cells in neuroblastic tumors. Mol. Cancer Res. 9 (7), 815–823.

K.Z. Chapman et al. / Neurobiology of Disease 83 (2015) 1–15 Ekdahl, C.T., Claasen, J.H., Bonde, S., Kokaia, Z., Lindvall, O., 2003. Inflammation is detrimental for neurogenesis in adult brain. Proc. Natl. Acad. Sci. U. S. A. 100 (23), 13632–13637. Ernst, A., Alkass, K., Bernard, S., Salehpour, M., Perl, S., Tisdale, J., et al., 2014. Neurogenesis in the striatum of the adult human brain. Cell 156 (5), 1072–1083. Gomez-Nicola, D., Valle-Argos, B., Pallas-Bazarra, N., Nieto-Sampedro, M., 2011. Interleukin-15 regulates proliferation and self-renewal of adult neural stem cells. Mol. Biol. Cell 22 (12), 1960–1970. Gundersen, H.J., 1978. Estimators of the number of objects per area unbiased by edge effects. Microschim. Acta 81 (2), 107–117. Gunn, M.D., Ngo, V.N., Ansel, K.M., Ekland, E.H., Cyster, J.G., Williams, L.T., 1998. A B-cell-homing chemokine made in lymphoid follicles activates Burkitt's lymphoma receptor-1. Nature 391 (6669), 799–803. Hoehn, B.D., Palmer, T.D., Steinberg, G.K., 2005. Neurogenesis in rats after focal cerebral ischemia is enhanced by indomethacin. Stroke 36 (12), 2718–2724. Hou, S.W., Wang, Y.Q., Xu, M., Shen, D.H., Wang, J.J., Huang, F., et al., 2008. Functional integration of newly generated neurons into striatum after cerebral ischemia in the adult rat brain. Stroke 39 (10), 2837–2844. Huber, C., Marschallinger, J., Tempfer, H., Furtner, T., Couillard-Despres, S., Bauer, H.C., et al., 2011. Inhibition of leukotriene receptors boosts neural progenitor proliferation. Cell. Physiol. Biochem. 28 (5), 793–804. Iadecola, C., Anrather, J., 2011. The immunology of stroke: from mechanisms to translation. Nat. Med. 17 (7), 796–808. Iosif, R.E., Ahlenius, H., Ekdahl, C.T., Darsalia, V., Thored, P., Jovinge, S., et al., 2008. Suppression of stroke-induced progenitor proliferation in adult subventricular zone by tumor necrosis factor receptor 1. J. Cereb. Blood Flow Metab. 28 (9), 1574–1587. Jenh, C.H., Cox, M.A., Hipkin, W., Lu, T., Pugliese-Sivo, C., Gonsiorek, W., et al., 2001. Human B cell-attracting chemokine 1 (BCA-1; CXCL13) is an agonist for the human CXCR3 receptor. Cytokine 15 (3), 113–121. Kim, C.H., Rott, L.S., Clark-Lewis, I., Campbell, D.J., Wu, L., Butcher, E.C., 2001. Subspecialization of CXCR5+ T cells: B helper activity is focused in a germinal center-localized subset of CXCR5+ T cells. J. Exp. Med. 193 (12), 1373–1381. Kizil, C., Dudczig, S., Kyritsis, N., Machate, A., Blaesche, J., Kroehne, V., et al., 2012. The chemokine receptor cxcr5 regulates the regenerative neurogenesis response in the adult zebrafish brain. Neural Dev. 7, 27. Koizumi, J.Y.Y., Nakazawa, T., Ooneda, G., 1986. Experimental studies of ischemic brain edema. 1. A new experimental model of cerebral embolism in rats in which recirculation can be introduced in the ischemic area. Jpn. J. Stroke 8, 1–8. Kokaia, Z., Martino, G., Schwartz, M., Lindvall, O., 2012. Cross-talk between neural stem cells and immune cells: the key to better brain repair? Nat. Neurosci. 15 (8), 1078–1087. Kyritsis, N., Kizil, C., Zocher, S., Kroehne, V., Kaslin, J., Freudenreich, D., et al., 2012. Acute inflammation initiates the regenerative response in the adult zebrafish brain. Science 338 (6112), 1353–1356. Lagace, D.C., 2012. Does the endogenous neurogenic response alter behavioral recovery following stroke? Behav. Brain Res. 227 (2), 426–432. Legler, D.F., Loetscher, M., Roos, R.S., Clark-Lewis, I., Baggiolini, M., Moser, B., 1998. B cell-attracting chemokine 1, a human CXC chemokine expressed in lymphoid tissues, selectively attracts B lymphocytes via BLR1/CXCR5. J. Exp. Med. 187 (4), 655–660. Lehmann, J., Hartig, W., Seidel, A., Fuldner, C., Hobohm, C., Grosche, J., et al., 2014. Inflammatory cell recruitment after experimental thromboembolic stroke in rats. Neuroscience 279, 139–154. Lehrmann, E., Christensen, T., Zimmer, J., Diemer, N.H., Finsen, B., 1997. Microglial and macrophage reactions mark progressive changes and define the penumbra in the rat neocortex and striatum after transient middle cerebral artery occlusion. J. Comp. Neurol. 386 (3), 461–476. Macas, J., Nern, C., Plate, K.H., Momma, S., 2006. Increased generation of neuronal progenitors after ischemic injury in the aged adult human forebrain. J. Neurosci. 26 (50), 13114–13119. Magnusson, J.P., Goritz, C., Tatarishvili, J., Dias, D.O., Smith, E.M., Lindvall, O., et al., 2014. A latent neurogenic program in astrocytes regulated by Notch signaling in the mouse. Science 346 (6206), 237–241. Marti-Fabregas, J., Romaguera-Ros, M., Gomez-Pinedo, U., Martinez-Ramirez, S., Jimenez-Xarrie, E., Marin, R., et al., 2010. Proliferation in the human ipsilateral subventricular zone after ischemic stroke. Neurology 74 (5), 357–365. Minger, S.L., Ekonomou, A., Carta, E.M., Chinoy, A., Perry, R.H., Ballard, C.G., 2007. Endogenous neurogenesis in the human brain following cerebral infarction. Regen. Med. 2 (1), 69–74. Monje, M.L., Toda, H., Palmer, T.D., 2003. Inflammatory blockade restores adult hippocampal neurogenesis. Science 302 (5651), 1760–1765. Morris, M.C., Gilliam, E.A., Li, L., 2014. Innate immune programing by endotoxin and its pathological consequences. Front. Immunol. 5, 680. Muraguchi, T., Takegami, Y., Ohtsuka, T., Kitajima, S., Chandana, E.P., Omura, A., et al., 2007. RECK modulates Notch signaling during cortical neurogenesis by regulating ADAM10 activity. Nat. Neurosci. 10 (7), 838–845. Parent, J.M., Vexler, Z.S., Gong, C., Derugin, N., Ferriero, D.M., 2002. Rat forebrain neurogenesis and striatal neuron replacement after focal stroke. Ann. Neurol. 52 (6), 802–813.

15

Paxinos, G., Watson, C., 1998. The Rat Brain in Stereotaxic Coordinates. Academic Press, San Diego. Rahpeymai, Y., Hietala, M.A., Wilhelmsson, U., Fotheringham, A., Davies, I., Nilsson, A.K., et al., 2006. Complement: a novel factor in basal and ischemia-induced neurogenesis. EMBO J. 25 (6), 1364–1374. Rapti, M., Atkinson, S.J., Lee, M.H., Trim, A., Moss, M., Murphy, G., 2008. The isolated N-terminal domains of TIMP-1 and TIMP-3 are insufficient for ADAM10 inhibition. Biochem. J. 411 (2), 433–439. Ren, X., Akiyoshi, K., Dziennis, S., Vandenbark, A.A., Herson, P.S., Hurn, P.D., et al., 2011. Regulatory B cells limit CNS inflammation and neurologic deficits in murine experimental stroke. J. Neurosci. 31 (23), 8556–8563. Robin, A.M., Zhang, Z.G., Wang, L., Zhang, R.L., Katakowski, M., Zhang, L., et al., 2006. Stromal cell-derived factor 1alpha mediates neural progenitor cell motility after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 26 (1), 125–134. Rothwell, N., 2003. Interleukin-1 and neuronal injury: mechanisms, modification, and therapeutic potential. Brain Behav. Immun. 17 (3), 152–157. Scholzke, M.N., Rottinger, A., Murikinati, S., Gehrig, N., Leib, C., Schwaninger, M., 2011. TWEAK regulates proliferation and differentiation of adult neural progenitor cells. Mol. Cell. Neurosci. 46 (1), 325–332. Stuart, M.J., Corrigan, F., Baune, B.T., 2014. Knockout of CXCR5 increases the population of immature neural cells and decreases proliferation in the hippocampal dentate gyrus. J. Neuroinflammation 11, 31. Sun, X., Zhang, Q.W., Xu, M., Guo, J.J., Shen, S.W., Wang, Y.Q., et al., 2012. New striatal neurons form projections to substantia nigra in adult rat brain after stroke. Neurobiol. Dis. 45 (1), 601–609. Thored, P., Arvidsson, A., Cacci, E., Ahlenius, H., Kallur, T., Darsalia, V., et al., 2006. Persistent production of neurons from adult brain stem cells during recovery after stroke. Stem Cells 24 (3), 739–747. Thored, P., Wood, J., Arvidsson, A., Cammenga, J., Kokaia, Z., Lindvall, O., 2007. Long-term neuroblast migration along blood vessels in an area with transient angiogenesis and increased vascularization after stroke. Stroke 38 (11), 3032–3039. Thored, P., Heldmann, U., Gomes-Leal, W., Gisler, R., Darsalia, V., Taneera, J., et al., 2009. Long-term accumulation of microglia with proneurogenic phenotype concomitant with persistent neurogenesis in adult subventricular zone after stroke. Glia 57 (8), 835–849. Tobin, M.K., Bonds, J.A., Minshall, R.D., Pelligrino, D.A., Testai, F.D., Lazarov, O., 2014. Neurogenesis and inflammation after ischemic stroke: what is known and where we go from here. J. Cereb. Blood Flow Metab. 34 (10), 1573–1584. Wang, Z., Andrade, N., Torp, M., Wattananit, S., Arvidsson, A., Kokaia, Z., et al., 2012. Meteorin is a chemokinetic factor in neuroblast migration and promotes strokeinduced striatal neurogenesis. J. Cereb. Blood Flow Metab. 32 (2), 387–398. Wanner, I.B., Anderson, M.A., Song, B., Levine, J., Fernandez, A., Gray-Thompson, Z., et al., 2013. Glial scar borders are formed by newly proliferated, elongated astrocytes that interact to corral inflammatory and fibrotic cells via STAT3-dependent mechanisms after spinal cord injury. J. Neurosci. 33 (31), 12870–12886. Weiss, N., Deboux, C., Chaverot, N., Miller, F., Baron-Van Evercooren, A., Couraud, P.O., et al., 2010. IL8 and CXCL13 are potent chemokines for the recruitment of human neural precursor cells across brain endothelial cells. J. Neuroimmunol. 223 (1–2), 131–134. West, M.J., 1999. Stereological methods for estimating the total number of neurons and synapses: issues of precision and bias. Trends Neurosci. 22 (2), 51–61. West, M.J., Slomianka, L., Gundersen, H.J., 1991. Unbiased stereological estimation of the total number of neurons in thesubdivisions of the rat hippocampus using the optical fractionator. Anat. Rec. 231 (4), 482–497. Yamashita, T., Ninomiya, M., Hernandez Acosta, P., Garcia-Verdugo, J.M., Sunabori, T., Sakaguchi, M., et al., 2006. Subventricular zone-derived neuroblasts migrate and differentiate into mature neurons in the post-stroke adult striatum. J. Neurosci. 26 (24), 6627–6636. Yan, Y.P., Sailor, K.A., Vemuganti, R., Dempsey, R.J., 2006. Insulin-like growth factor-1 is an endogenous mediator of focal ischemia-induced neural progenitor proliferation. Eur. J. Neurosci. 24 (1), 45–54. Yan, Y.P., Sailor, K.A., Lang, B.T., Park, S.W., Vemuganti, R., Dempsey, R.J., 2007. Monocyte chemoattractant protein-1 plays a critical role in neuroblast migration after focal cerebral ischemia. J. Cereb. Blood Flow Metab. 27 (6), 1213–1224. Yan, Y.P., Lang, B.T., Vemuganti, R., Dempsey, R.J., 2009. Osteopontin is a mediator of the lateral migration of neuroblasts from the subventricular zone after focal cerebral ischemia. Neurochem. Int. 55 (8), 826–832. Zamanian, J.L., Xu, L., Foo, L.C., Nouri, N., Zhou, L., Giffard, R.G., et al., 2012. Genomic analysis of reactive astrogliosis. J. Neurosci. 32 (18), 6391–6410. Zhang, R., Zhang, Z., Zhang, C., Zhang, L., Robin, A., Wang, Y., et al., 2004. Stroke transiently increases subventricular zone cell division from asymmetric to symmetric and increases neuronal differentiation in the adult rat. J. Neurosci. 24 (25), 5810–5815. Zheng, W.H., Kar, S., Dore, S., Quirion, R., 2000. Insulin-like growth factor-1 (IGF-1): a neuroprotective trophic factor acting via the Akt kinase pathway. J. Neural Transm. Suppl. 60, 261–272. Zheng, W., Nowakowski, R.S., Vaccarino, F.M., 2004. Fibroblast growth factor 2 is required for maintaining the neural stem cell pool in the mouse brain subventricular zone. Dev. Neurosci. 26 (2–4), 181–196.