Brain, Behavior, and Immunity 41 (2014) 239–250

Contents lists available at ScienceDirect

Brain, Behavior, and Immunity journal homepage: www.elsevier.com/locate/ybrbi

Microglia and their CX3CR1 signaling are involved in hippocampal- but not olfactory bulb-related memory and neurogenesis Ronen Reshef, Tirzah Kreisel, Dorsa Beroukhim Kay, Raz Yirmiya ⇑ Department of Psychology, The Hebrew University, Jerusalem, Israel

a r t i c l e

i n f o

Article history: Received 4 March 2014 Received in revised form 15 April 2014 Accepted 26 April 2014 Available online 13 June 2014 Keywords: Microglia CX3CR1 Learning and memory Hippocampus Olfactory bulb Neurogenesis Environmental enrichment

a b s t r a c t Recent studies demonstrate that microglia play an important role in cognitive and neuroplasticity processes, at least partly via microglial CX3C receptor 1 (CX3CR1) signaling. Furthermore, microglia are responsive to environmental enrichment (EE), which modulates learning, memory and neurogenesis. In the present study we examined the role of microglial CX3CR1 signaling in hippocampal- and olfactorybulb (OB)-related memory and neurogenesis in homozygous mice with microglia-specific transgenic expression of GFP under the CX3CR1 promoter (CX3CR1 / mice), in which the CX3CR1 gene is functionally deleted, as well as heterozygous CX3CR1+/ and WT controls. We report that the CX3CR1-deficient mice displayed better hippocampal-dependent memory functioning and olfactory recognition, along with increased number and soma size of hippocampal microglia, suggestive of mild activation status, but no changes in OB microglia. A similar increase in hippocampal-dependent memory functioning and microglia number was also induced by pharmacological inhibition of CX3CR1 signaling, using chronic (2 weeks) i.c.v. administration of CX3CR1 blocking antibody. In control mice, EE improved hippocampal-dependent memory and neurogenesis, and increased hippocampal microglia number and soma size, whereas odor enrichment (OE) improved olfactory recognition and OB neurogenesis without changing OB microglia status. In CX3CR1-deficient mice, EE and OE did not produce any further improvement in memory functioning or neurogenesis and had no effect on microglial status. These results support the notion that in the hippocampus microglia and their interactions with neurons via the CX3CR1 play an important role in memory functioning and neurogenesis, whereas in the OB microglia do not seem to be involved in these processes. Ó 2014 Elsevier Inc. All rights reserved.

1. Introduction Microglia serve as the representatives of the immune system in the brain, and therefore most research on these cells focused on their roles in infection, injury, and neurodegenerative diseases. Under these conditions, ‘‘activated’’ microglia change their morphology from a ramified (‘‘resting’’ or ‘‘quiescent’’) to an amoeboid shape, assume phagocytic capabilities, produce and secrete pro-inflammatory cytokines (e.g., TNFa, IL-1b and IL-6), express inflammatory-related cell surface markers (e.g., MHC class II), display increased migratory capabilities and also begin to proliferate (Kettenmann et al., 2011). Because the brain’s delicate chemical and structural balance requires strict, context-specific regulation over microglial status, there are numerous processes that prevent ⇑ Corresponding author. Address: Department of Psychology, The Hebrew University of Jerusalem, Jerusalem 91905, Israel. Tel.: +972 2 5883695; fax: +972 2 5882947. E-mail address: [email protected] (R. Yirmiya). http://dx.doi.org/10.1016/j.bbi.2014.04.009 0889-1591/Ó 2014 Elsevier Inc. All rights reserved.

microglial over-activation, including signaling by the CX3C receptor-1 (CX3CR1). In the brain, CX3CR1 is exclusively expressed by microglia, binds the neuronally-derived CX3CL1 (fractalkine) chemokine, and usually exerts tonic inhibition over microglial activation (Harrison et al., 1998; Jung et al., 2000; Ransohoff et al., 2007; Wolf et al., 2013). Recent studies demonstrated that in their ‘‘resting’’ state microglia are highly motile and actively participate in important neurophysiological processes, including synapse development and function (e.g., activity-dependent synaptic pruning) and micro-damage repair (Schafer et al., 2012; Tremblay et al., 2011), as well as ‘‘shaping’’ of hippocampal neurogenesis (e.g., by eliminating non-functional apoptotic newborn cells) (Ekdahl et al., 2009; Sierra et al., 2010). Furthermore, hippocampal microglia were found to be involved in the beneficial effects of rearing in enriched environment (Ziv et al., 2006; Ziv and Schwartz, 2008), known to enhance memory functioning and neurogenesis in a modality specific manner (Rochefort et al., 2002; van Praag et al., 2000). Specifically, exposure to environmental enrichment (EE)

240

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

induced concomitant increases in neurogenesis and in the number of hippocampal microglia, which assumed a neuroprotective phenotype via regulation by the T cells-derived cytokines IFNc and IL-4. Moreover, the EE-induced increase in neurogenesis was attenuated by chronic treatment with the microglial inhibitor minocycline (Butovsky et al., 2006; Ziv et al., 2006). In the subventricular zone (SVZ)-olfactory bulb neurogenic niche, evidence for a role of microglia in neurogenesis is derived from studies reporting that under resting conditions SVZ microglia display an intermediate state of constitutive activation and proliferation compared with non-neurogenic cortical areas (Goings et al., 2006), and that microglia or their conditioned medium supports the proliferation and differentiation of cultured SVZ cells to newborn neurons (Aarum et al., 2003; Walton et al., 2006). Another indirect evidence for a role for microglia in cognition and neurogenesis was recently provided by several studies examining the effects of CX3CR1 deficiency on hippocampal-related memory and neurogenesis, as well as its modulation by EE (Bachstetter et al., 2009; Maggi et al., 2011; Rogers et al., 2011). These studies utilized transgenic mice expressing enhanced green fluorescent protein (EGFP) in conjunction with the CX3CR1. Homozygous mice with the CX3CR1-GFP mutation have a functional deletion of the microglial CX3CR1 (Jung et al., 2000). Although all of these studies demonstrated that CX3CR1-deficient mice exhibit cognitive and neurogenic alterations, the results of these studies were somewhat contradictory: In some studies CX3CR1-deficient mice exhibited impaired spatial learning, hippocampal neurogenesis and LTP and consistently young adult rats injected i.c.v. with CX3CR1 blocking antibody displayed reduced neurogenesis (Bachstetter et al., 2009; Rogers et al., 2011). However, in another study the CX3CR1-deficient mice displayed improved hippocampal LTP and spatial learning capabilities (Maggi et al., 2011). Moreover, in the latter study, EE, which significantly improved spatial memory functioning in WT mice, had no such effect in the CX3CR1-deficient mice. To study the generality of the cognitive and neurogenic effects of microglia, we examined the involvement of these cells, as well as CX3CR1 signaling, in memory, neurogenesis and their modulation by enriched environment, not only in hippocampal-related functions, but also in the olfactory bulb (OB) system. In parallel, we examined for the first time the effects of enrichment on hippocampal and OB microglia in CX3CR1-deficient and control mice. 2. Methods 2.1. Subjects Subjects were 2–3 months old male CX3CR1-GFP transgenic mice (on C57BL/6 genetic background) and their wild type (WT) controls. The GFP gene was knocked-in under the CX3CR1 promoter (Jung et al., 2000). We used homozygous CX3CR1-GFP mice, with CX3CR1 deficiency (CX3CR1 / ), as well as heterozygous CX3CR1-GFP mice, which express both the functional CX3CR1 gene and GFP (CX3CR1+/ mice) and wild-type (WT) mice as controls. Animals were housed in air conditioned rooms (23° C), with food and water ad libitum, and were kept in a reversed light/dark cycle, with lights off from 8 a.m. to 8 p.m. All the experiments were approved by the Hebrew University Ethics Committee on Animal Care and Use. 2.2. Environmental enrichment procedure Groups of CX3CR1 / , CX3CR1+/ and WT mice were randomly assigned into two subgroups (n = 8/subgroup), reared in either enriched environmental conditions for 6 weeks or standard

housing conditions (i.e., 3  2 experimental design). In the enriched condition, 8 mice were housed in large (60  60  40 cm) transparent cages with running wheels, plastic-tube mazes, and ladders. In the standard condition, 2–3 mice were housed in a regular laboratory cage (25  20  15 cm). 2.3. Odor enrichment procedure Groups of CX3CR1 / , CX3CR1+/ and WT mice were randomly assigned into two subgroups (n = 8/subgroup), reared in either an odor-enriched environment for 40 days or in standard housing conditions (i.e., 3  2 experimental design). Odor-enriched mice were exposed daily for 24 h to different odoriferous substances (Table 1) that were placed in a tea-ball hanging from the filtering cover of standard animal cages. The standard housing subgroups were reared under the same conditions, except that the tea-balls were left empty. Three days before the behavioral experiments, mice were housed singly and familiarized with the test procedure. All of the behavioral experiments were conducted in the home cage of the tested animals during the dark phase of the circadian cycle under dimmed illumination. 2.4. Behavioral measurements 2.4.1. Water maze learning and memory Animals were trained in a circular pool (160 cm in diameter) filled with 23 ± 2 °C water mixed with non-toxic gouache paint to make it opaque. In the spatial memory experiment, mice were trained to find the location of a hidden platform, submerged 1 cm below the water surface, using extra maze visual cues. Training consisted of three trials per day, with a one-hour break between trials, for 3 days. Both experiments were conducted using a random protocol, in which the entrance point to the maze was varied randomly between trials and the platform remained in a permanent position. In the fourth day of the experiment, the platform was removed and mice were placed in the water maze for 60 s (probe test). Memory in the probe test was measured by calculating the Gallagher’s proximity index, i.e., the average distance of the animal from the previous location of the platform in each second, averaged across the whole 60 s trial (Gallagher et al., 1993). A video camera above the pool was connected to a computerized video tracking system (Noldus Ethovision SE, The Netherlands), which automatically monitored the latency to reach the platform in each training trial and the location of the mice in the probe trial (see (Avital et al., 2003; Goshen et al., 2009) for details). 2.4.2. Contextual and auditory-cued fear conditioning The fear conditioning system consisted of two transparent square conditioning cages (25  21  18 cm), with a grid floor wired to a shock generator and scrambler (Kinder scientific, U.S.). Mice were placed in the cage for 120 s, and then a pure tone

Table 1 Chart of all the common odors used to enrich the olfactory environment and to test olfactory memory. Olfactory enrichment Lavender Cinnamon Marjoram Paprika Cumin Mace Clove Tarragon Cardamom Basil leaves

Olfactory memory test Anise Chocolate Orange Lemon Ginger Garlic Thyme Rosemary Banana Celery

Jasmine Roses Mint Vanilla Coffee Caraway Chrysanthemum Lemon verbena Fenugreek

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

(2.9 kHz) was sound for 20 s, followed by a 2 s, 0.5 mA foot-shock. This procedure was then repeated and 30 s after the delivery of the second foot-shock mice were returned to their home cage. 48 h later, fear conditioning was assessed by measuring the time spent in freezing (complete immobility) during exposure to the conditioned stimuli (i.e., either the context or the tone), using an automated system (Kinder scientific, U.S.). The mice tested for contextual fear conditioning were placed in the original conditioning cage, and freezing was measured for 5 min. The mice tested for the auditory-cued fear conditioning were placed in a different context – a triangular shaped cage with no grid floor. As a control for the influence of the novel environment, freezing was measured for 2.5 min in this new cage, and then the tone was sounded for 2.5 min, during which conditioned freezing was measured. Freezing was also measured during the first 120 s of the conditioning trial, before the tone and shock administration, in order to assess possible strain differences in baseline freezing. 2.4.3. Olfactory memory Memory for odors was measured by the Olfactory Recognition (OR) test. Odors were presented by placing 6 ll of the odor stimulus onto a circle of blotting paper, which was inserted into a 1 ml BD Plastipak syringe. This syringe was then inserted into a 10 ml BD Plastipak syringe that was permanently present inside the cage via attachment to the cage top, such that the blotting paper was 8 cm from the floor. This procedure allowed changing the odor stimulus without disturbing the animal and without creating contact between the odor stimulus and the permanent (10 ml) outer syringe. All odors were dissolved in mineral oil and freshly prepared before each experiment. Moreover, each odor preparation was inserted into a new 1 ml syringe to prevent odor contamination. In the three days preceding the OR test animals were familiarized with the testing procedure by exposing them to odor stimuli different from those subsequently used in the testing sessions, using the odorous blotting paper within the syringes, as detailed above. In the following days, each animal underwent 4 testing sessions, two for specific assessment of odor memory and two for assessing non-specific odor-related behavior in the testing context (as control). In the specific memory testing sessions mice were presented with an odor for 5 min and then presented with the same odor at a delay interval of either 30 or 360 min (these intervals were based on previous literature (Rochefort et al., 2002) and preliminary examination of various intervals in our laboratory). The order of the sessions with the two intervals was counterbalanced and they were separated by at least 24 h. Different odors were used in each test and counter-balanced across the different delay conditions. The time that the animals spent rearing and sniffing at the odorous blotting paper was recorded and the OR score was computed as the percent time of investigation following the delay divided by the investigation time during the first encounter with the odor. A significant decrease in the OR score during the second presentation indicates that mice were able to recognize an odor that had been presented previously. To assess the specificity of odor recognition, the same animals were also tested in the non-specific testing sessions, whose procedure was identical to the specific memory testing except that 30 and 360 min following the presentation of the first odor the mice were presented with a different odor. 2.4.4. BrdU administration To determine the rate of olfactory neurogenesis, 5-bromo-2’deoxyuridine (BrdU; 10 mg/ml; Sigma, St. Louis, MO), a marker of cell proliferation, was administered i.p. (0.1 mg/g of body weight). Mice received 4 daily injections of BrdU on days 20–24 of the 40 days olfactory enrichment procedure. The number of

241

BrdU-labeled neurons in the OB and SVZ was determined by immunohistochemical methods, as described below. 2.4.5. Administration of CX3CR1 blocking antibodies Adult WT mice were anesthetized with Isoflurane (Terrell, USA) and placed in a stereotaxic apparatus. A burr hole was drilled posterior to bregma, using the following formula: [ 0.4– 0.66(bl 3.8)] mm (bl = distance between bregma and lambda), 1.5 mm lateral to the midline, and a 30 gauge brain-infusion cannula (Alzet, CA, USA) was lowered 2.2 mm below the skull surface into the right lateral ventricle. Brain cannulation is a non-trivial procedure, which is known to activate microglia in the cannula area. The unilateral cannulation approach was used in order to differentiate between the net effect of the antibody (which readily reaches the non-operated, quiescent hemisphere) and the effect of the antibody on top of the cannulation-induced microglial activation in the injected hemisphere. The guide cannula was secured to the skull with three stainless-steel screws and dental cement. An osmotic minipump (Alzet., CA, USA) was implanted subcutaneously and attached with catheter tubing to the cannula. The minipumps delivered artificial CSF for 7 days, while allowing the animals to recover from the procedure. On day 7 post implantation, the minipumps were exchanged with new ones filled with CX3CR1 blocking antibody (Torrey pines bio labs, NJ, USA; 6 lg per day) or artificial CSF (vehicle) for 14 additional days. Behavioral experiments were performed at 8–14 days following the initiation of antibody delivery. 2.4.6. Immunohistochemistry After termination of the enrichment protocols and behavioral testing, animals were perfused trans-cardially with cold phosphate buffer solution (PBS), followed by 4% paraformaldehyde in 0.1 M PBS, and the brains were quickly removed and placed in 4% paraformaldehyde. After 24 h, the brains were placed in 30% sucrose solution in PBS for 48 h and then frozen in OCT. Coronal sections (8 lm) were serially cut along the rostro-caudal axis of the olfactory bulbs and the dorsal hippocampus using a cryostat (Leica) and mounted on slides. To determine the rate of neurogenesis in the OB and SVZ, sections were double-labeled for BrdU and the neuronal marker NeuN. Slides were first pretreated by DNA denaturation (2 N HCl for 30 min at 37 °C), and then incubated for 48 h at 4 °C with the primary antibodies: rat monoclonal anti-BrdU antibody (1:200; Accurate Scientific, Harlan Sera-Lab, UK) and mouse monoclonal anti-NeuN antibody (1:200; Chemicon, Temecula, CA). Sections were then incubated with a secondary antibody (goat anti rat IgG conjugated to Alexa 555, and goat anti-mouse IgG conjugated to Alexa 488 (both at 1:500; obtained from Invitrogen (Brazil). The rate of neurogenesis in the hippocampus was assessed by staining for doublecortin (DCX), which is expressed by immature neurons for only 2–3 weeks after their division. Sections were incubated in 3% H2O2 in methanol for 20 min in 20°. Following three washes with PBS, sections were incubated in 0.5% TritonX100 in PBS for 1 h. After one more wash with PBS, sections were incubated with the primary antibody (goat anti-DCX1:1000, Millipore, Chemicon, Temecula, CA, USA) for 24 h at 4°. Sections were then incubated with the secondary antibody (biotin-SP-conjugated donkey anti-Guinea pig, 1:200; Jackson Laboratories, West grove, PA, USA) for 1 h at RT and visualized using a conjugated streptavidin (Jackson Laboratories, West grove, PA, USA). As a first attempt to determine the phenotype of the hippocampal microglia, slides were stained for the activation markers Iba-1, MHC-II and TNFa, as well as for Brain Derived Neurotrophic Factor (BDNF). To determine microglial Iba-1 expression, sections were fixed with 2% PFA, followed by blocking with 3% NGS in 5% BSA. The primary antibody, rabbit anti iba-1 (1:200, Wako) was added

242

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

for 24 h in 4 °C. The secondary antibody, goat anti rabbit IgG (Molecular Probes), was added for 1 h at RT. For assessment of MHC-II expression, sections were incubated in a solution of 0.1% Triton-X100 in PBS for 1 h, followed by incubation with the primary antibody (mouse anti MHC-II 1:50, Millipore, Chemicon), diluted in 0.1% Triton X-100 for 24 h in 4 °C. Sections were then incubated with the secondary antibody (goat anti mouse IgG conjugated to Alexa 555) for 1 h in RT. To determine microglial TNFa expression, sections were incubated in a solution of 3% H2O2 in methanol for 20 min in 20 °C. Sections were then blocked in 2% BSA in PBS for 1 h at RT, followed by incubation with the primary antibody (goat anti TNFa ata dilution of 1:100, R&D Systems, Minneapolis, USA) for 24 h at 4 °C. Sections were then incubated with the secondary antibody (donkey anti goat IgG conjugated to Alexa 555) for 1 h at RT. To determine microglial BDNF expression, sections were incubated with Background Buster blocking (Innovex biosciences) for 30 min, followed by incubation with the primary antibody (rabbit anti-BDNF 1:250, Millipore, Chemicon), diluted in 0.3% Triton X-100 for 24 h at RT. Sections were then incubated with the secondary antibody (goat anti rabbit IgG conjugated to Alexa 555) for 1 h at RT. All slides were counterstained with DAPI. 2.4.7. Estimation of microglial number and soma area Images were captured using Nikon Eclipse microscope and camera, at 10 and 40 magnifications. In each slide the number of GFP-labeled microglia (that were also counter-stained with DAPI to ascertain nuclear staining) as well as the microglial soma area were automatically measured in a defined area exclusively containing the entire hippocampal dentate gyrus, the entire olfactory bulb (including the glomerular, plexiform, mitral and granule layers) and the sub-ventricular zone (SVZ) (situated throughout the lateral walls of the lateral ventricles), using Nikon Imaging Elements Software (NIS-Elements). These measurements were conducted by an experimenter who was blind with respect to the group assignment of the animals. For each brain, microglia number and morphology were assessed in 16 sections along the entire rostro-caudal axis of the olfactory bulbs, 16 sections along the rostro-caudal axis of the two SVZ’s and additional 16 sections along the entire rostro-caudal axis of the dorsal hippocampus (in both hemispheres), and the number of microglia and their soma areas were averaged. 2.5. Statistical analysis All data are presented as mean ± SEM. Statistical comparisons were computed using SPSS 19.0 software and consisted of t-tests, one-way and two-way analyses of variance (ANOVAs), followed by the Fisher’s least significant difference (LSD) post hoc analyses, when appropriate. The results of the water maze experiment were analyzed by a three way ANOVA, with strain and enrichment status as between subjects factors and the trials as a within subjects, repeated measures factor. The results of the olfactory recognition test were also analyzed by a three way ANOVA, with odor specificity and enrichment status as between subjects factors and the interval (30 vs. 360 min) as a within subjects, followed by planed contrasts between the specific and non-specific conditions. 3. Results 3.1. Effects of environmental enrichment on hippocampal-dependent learning, memory and neurogenesis When tested in the contextual fear conditioning (FC) task, enriched WT and CX3CR1+/ displayed significantly improved contextual memory, i.e., longer freezing time in the context in which

they were shocked 48 h earlier, in comparison with their nonenriched WT and CX3CR1+/ controls. Non-enriched CX3CR1 / showed longer freezing time then non-enriched WT and CX3CR1+/ mice (the latter strains did not differ from each other), but enrichment did not have any further memory facilitating effect in the CX3CR1 / strain (p < 0.05) (Fig. 1a). This effect was not the result of inherent strain differences in the tendency to freeze or differential anxiety levels, because at baseline (i.e., before the shock administration) there were no differences in freezing time between any of the groups (with freezing times ranging from 0% to 5% in the various groups, p > 0.1). In contrast with the contextual FC results, in the auditory-cued FC paradigm enrichment caused memory improvement in all strain groups (p < 0.05) and there were no strain or strain by enrichment interaction effects (p > 0.1) (Fig. 1b). The lack of group differences during the 2.5 min preceding the tone presentation also corroborates the absence of differential tendency to freeze or of anxiety levels (p > 0.1). In the Morris water maze task, non-enriched CX3CR1 / mice showed facilitated learning compared with non-enriched WT and CX3CR1+/ controls (Fig. 1c,e). Enriched WT and CX3CR1+/ mice displayed significantly improved spatial learning, i.e., faster decline in the latencies to reach the platform, compared with their nonenriched WT and CX3CR1+/ controls (p < 0.05), but enriched CX3CR1 / mice did not differ significantly from their nonenriched controls (Fig. 1d,e). The learning effects did not seem to result from differential search strategies or sensorimotor abilities between the groups, as no group differences were obtained in the levels of thigmotaxis (the tendency to cling or follow the wall around the outer perimeter of the tank) or swim speed (p > 0.1). Despite the learning differences, the groups did not differ in their memory performance during the probe trial, displaying similar Gallagher’s proximity index (average distance to the previous location of the platform) (p > 0.1) (Fig. 1f). This result may be explained by the attainment of asymptotic spatial performance in all groups during the last day of training. In non-enriched mice, the number of adult-born (doublecortin (DCX)-labeled) neurons in the hippocampal dentate gyrus was higher in WT than in the CX3CR1+/ or CX3CR1 / groups (p < 0.05). Following EE, neurogenesis was increased in enriched WT and CX3CR1+/ mice, as compared with their respective nonenriched controls, whereas in CX3CR1 / mice EE had no effect on neurogenesis (Fig. 2a–c). To clarify the specific stage at which CX3CR1 signaling affects the neurogenesis process we divided the DCX-labeled cells to neuronal progenitors (i.e., 1–10 days old cells that are still in polarization, migration or the early stages of neurite growth) vs. young neurons (i.e., 10–14 days old cells, which already resemble mature granule cells morphology yet are not fully connected to the network) (insets in Fig. 2c). Separate analyses on the two populations revealed similar strain and EE effects to those obtained in the overall analysis. 3.2. Effects of environmental enrichment on hippocampal microglia Because in the behavioral tests WT and CX3CR1+/ mice displayed similar memory functioning and enrichment effects, and because analysis of microglia number and morphology was based on the CX3CR1-GFP labeling, studies on microglia alterations following EE focused on comparisons between the CX3CR1+/ and CX3CR1 / groups. In EE-exposed CX3CR1+/ mice the number of DG microglia (Fig. 2d–f) and their soma area (Fig. 2g–i) were significantly increased compared with their non-enriched controls (p > 0.05). Non-enriched CX3CR1 / mice displayed elevated number of DG microglia and their soma area, compared with non-enriched CX3CR1+/ mice, however, EE did not produce any further microglial changes in the CX3CR1 / mice. These effects did not seem to depend on GFP gene dosage effect because

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

243

Fig. 1. Effects of environmental enrichment and CX3CR1 deficiency on learning and memory functioning. (a) Contextual fear conditioning (reflected by % freezing during the memory retention test) was significantly increased following enrichment in WT and CX3CR1+/ mice, as compared with their non-enriched controls. Non-enriched CX3CR1 / mice displayed improved memory in this task compared to non-enriched WT and CX3CR1+/ mice, but they showed no further improvement following enrichment. These findings were reflected by a significant strain by enrichment interaction (F(2,40) = 4.148, p < 0.05) (n = 8/group). (b) Auditory-cued fear conditioning was significantly increased following enrichment in all strains, as compared with their non-enriched controls (F(1,40) = 15.185, p < 0.05). In this task, there were no differences in memory functioning among the three non-enriched control groups. (c) In the water maze paradigm, non-enriched CX3CR1 / mice exhibited improved spatial learning, i.e., their latencies to reach the submerged platform were shorter compared to their non-enriched controls. (d) Enriched WT and CX3CR1+/ mice exhibited improved spatial learning, i.e., their latencies to reach the submerged platform were shorter compared to their non-enriched controls. However, CX3CR1 / mice showed no further improvement following enrichment. The findings presented in (c) and (d) were analyzed together, using a three-way ANOVA (with strain and enrichment status as between subjects factors and the trials as a within subjects, repeated measures factor), demonstrating a significant strain by enrichment interaction (F(2,39) = 4.45, p < 0.05). Furthermore, a separate ANOVA on the control condition revealed a significant difference between the CX3CR1+/ mice and the other two groups (F(2,19) = 3.79, p < 0.05). (e) For clarity of presentation, the latencies to reach the platform, shown in (c) and (d), were averaged over the three training days, demonstrating the enrichment-induced learning improvement in WT and CX3CR1+/ mice, as well as the superior learning of non-enriched CX3CR1 / mice, which did not display further enrichment-induced learning facilitation. These findings were reflected by a significant strain by enrichment status interaction (F(2,40) = 4.069, p < 0.005) (n = 7–8/group). (f) The mean distance from the platform during the probe trial (also termed Gallagher’s proximity index) did not differ among the groups, probably due to the attainment of asymptotic spatial performance in all groups during the last day of training. ⁄ p < 0.05 compared with the respective non-enriched control group. ⁄⁄p < 0.05 compared with the non-enriched WT and CX3CR1+/ control groups.

immunohistochemical labeling with Iba-1 showed that the morphology of microglia cells was very similar using either GFP or Iba-1 staining (insets within Fig. 2h–i), as we described previously (Kreisel et al., 2014). To verify the difference in microglia number between CX3CR1 / and WT controls and to further exclude the possibility of a gene dosage effect, sections from WT, CX3CR1+/ and CX3CR1 / were also stained with the microglial marker Iba-1. Consistent with the findings obtained with CX3CR1-GFP labeling, the number of DG microglia in CX3CR1 /

mice was significantly greater than the number in either WT or CX3CR1+/ mice (Fig. 2j–l). In contrast with the effects of EE, odor enrichment had no effect on hippocampal microglia in any strain (data not shown). Elevated microglial numbers and enlargement of their soma area are usually indicative of an activated phenotype. However, staining of hippocampal sections with an antibody to the microglial activation markers MHC-II did not detect any stained cells (although in positive control animals injected with LPS many

244

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

Fig. 2. Effects of environmental enrichment and CX3CR1 deficiency on hippocampal neurogenesis and microglial status. (a) Hippocampal neurogenesis, determined by counting the number of DCX-labeled cells in the dentate gyrus (DG), was significantly higher in non-enriched WT mice, as compared with non-enriched CX3CR1+/ and CX3CR1 / mice. Following enrichment, neurogenesis was significantly increased in WT and CX3CR1+/ mice, whereas in CX3CR1 / mice enrichment did not influence neurogenesis. These findings were reflected by significant strain, enrichment, and strain by enrichment interaction effects (F(2,23) = 19.5, p < 0.05; F(1,23) = 10.299, p < 0.05; F(2,23) = 3.795, p < 0.05, respectively) (n = 5/group). (b) Representative pictures of the DG from a non-enriched CX3CR1+/ control mouse and (c) an enriched CX3CR1+/ mouse demonstrate the increased number of DCX-labeled new neurons, particularly in the sub-granular zone of the DG, following enrichment. This increase was observed with respect to both neuronal progenitors and to young neurons, exemplified in the left and right insets, respectively (Blue = DAPI-labeled cell bodies, red = DCX-labeled new neurons). (d) In enriched CX3CR1+/ mice, the density of hippocampal microglia (number/mm2 DG) was significantly increased, as compared with their non-enriched controls. Non-enriched CX3CR1 / mice displayed greater density of hippocampal microglia compared with non-enriched CX3CR1+/ controls, but enrichment did not produce a further increase in microglial density in this strain (n = 5/group). (e) Representative pictures of the DG in a non-enriched CX3CR1+/ control mouse and (f) an enriched CX3CR1+/ mouse, demonstrating the enrichment-induced increase in microglial number (Blue = DAPI-labeled cell bodies, Green = CX3CR1-GFP-labeled microglia). (g) In enriched CX3CR1+/ mice, microglial soma area was significantly increased, as compared with their non-enriched controls. Non-enriched CX3CR1 / mice displayed greater microglial soma area compared with non-enriched CX3CR1+/ controls, but enrichment did not produce further increase in microglial soma size in this strain. These findings were reflected by significant strain by enrichment interactions (F(1,16) = 4.830, p < 0.05; F(1,20) = 5.83, p < 0.05; for the microglia number and soma area, respectively) (n = 4–5/group). (h) A representative picture of a microglia cell, double-labeled with CX3CR1-GFP (green) and Iba-1 (red), from a non-enriched CX3CR1+/ mouse, as compared with (i) a microglia cell from an enriched CX3CR1+/ mouse, showing larger soma area. (j) In non-enriched CX3CR1 / mice, the density of hippocampal Iba-1 positive microglia (number/mm2 DG) was significantly increased, as compared with non-enriched WT and CX3CR1+/ animals (F(2,9) = 20.08, p < 0.05) (h–j) (n = 4/group). (k) Representative pictures of the DG in a non-enriched WT control mouse and (l) a non-enriched CX3CR1 / mouse, demonstrating the increase in microglial number in CX3CR1-deficient mice (Blue = DAPI-labeled cell bodies, Green = Iba-1 labeled microglia). ⁄p < 0.05 compared with the corresponding non-enriched control group; ⁄⁄p < 0.05 compared with the WT non-enriched control group. ⁄⁄⁄p < 0.05 compared with all other groups. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

microglia were found to express this molecule). Staining for the pro-inflammatory cytokine TNFa as well as to the neurotrophic and neuroprotective marker BDNF revealed scattered stained cells, but no GFP-labeled microglia were double-labeled with either of these markers in any of the groups (data not shown). 3.3. Effects of CX3CR1 blocking antibody on learning and memory functioning, neurogenesis and microglial status To corroborate and extend the findings on the positive effects of CX3CR1 signaling deficiency on hippocampal-dependent learning and memory and to exclude the possibility that the behavioral, neurogenic and microglial alterations exhibited by knock-in animals involve developmental problems we examined the effects of chronic administration of CX3CR1 blocking antibody on memory, neurogenesis and microglial status in WT mice. In the contextual fear conditioning (FC) task, mice that were infused i.c.v. with CX3CR1 blocking antibody for 14 days displayed significantly enhanced freezing during the test (i.e., 48 h following acquisition), as compared with vehicle-treated mice (p < 0.05) (Fig. 3a), whereas in the auditory-cued FC paradigm there was no difference between the groups (p > 0.1; Fig. 3b). This effect was not the result of CX3CR1 blocking antibody differences in the tendency to freeze or differential anxiety levels, because at baseline (i.e., before of the shock administration) there were no differences in freezing time between the groups (with freezing times ranging from 3% to 6% in the two groups) (p > 0.1). In the water maze paradigm, mice that were treated with the CX3CR1 blocking antibody displayed significantly improved spatial memory, i.e., faster decline in latencies to reach platform, as compared with vehicle-treated mice (p < 0.05) (Fig. 3c). The learning effects did not seem to result from differential search strategies between the groups, as no group differences were obtained in the levels of thigmotaxis were found (p > 0.1). There were no differences between the groups in the probe memory test. Specifically, the mean ± SEM Gallagher’s proximity index (representing the mean distance from the previous location of the platform throughout the probe trial) of the vehicle and CX3CR1-Ab were 57 (±18.2) and 51(±4.7), respectively, p > 0.1). The number of adult-born (doublecortin-labeled) neurons did not differ between the groups (p > 0.1; Fig. 3d–f). Analysis of microglia number revealed a marked difference between the numbers of microglia in the left vs. the right hippocampus of vehicle-treated mice, reflecting an increased microglial number in the cannula-implanted hemisphere. In the non-implanted hippocampus, the number of microglia in CX3CR1 blocking antibody-treated mice was greater than in the corresponding vehicle treated mice (p < 0.05), but in the implanted hemisphere microglia number was not further increased by the blocking antibody administration (Fig. 3g–k). 3.4. Effects of odor enrichment on odor recognition memory Odor recognition was measured by assessing the reductions in the investigation time for the same odor presented either 30 or 360 min earlier (specific recognition), as compared with changes in investigation of a different odor than the one presented 30 or 360 min earlier (due to non-specific factors, e.g., habituation to the manipulation). In both WT and CX3CR1+/ mice, the control and enriched groups displayed specific odor recognition at the relatively short (30 min) interval, however in the longer interval (360 min) only the enriched mice displayed specific recognition whereas the control mice could no longer recognize the same odor (Fig. 4a,b). In CX3CR1 / mice, both control and enriched animals displayed specific odor recognition at both intervals, i.e., in this

245

strain the control non-enriched mice displayed odor recognition also at the longer interval (Fig. 4c). 3.5. Effects of odor enrichment on OB and SVZ neurogenesis Following odor enrichment (OE), the number of adult-born neurons (double labeled with BrdU and NeuN) was increased in the OB of enriched WT and CX3CR1+/ mice, as compared with their non-enriched controls (p < 0.05). In contrast, OE had no effect on OB neurogenesis in CX3CR1 / mice (Fig. 5a–c). Non-enriched CX3CR1 / mice displayed similar neurogenesis levels to those exhibited by non-enriched WT and CX3CR1+/ mice. To examine whether the effect of enrichment in WT and CX3CR1+/ mice was related to the survival of the adult-born cells after their migration to the OB or to neural progenitor cells proliferation in the SVZ, BrdU was injected 1 day before the termination of the 40 days odor enrichment period and cell proliferation in the rostral SVZ was measured. No differences were found in the numbers of progenitor cells between the enriched and non-enriched WT and CX3CR1+/ groups (Fig. 5d–f). Similarly to the findings in the OB there were no differences between the numbers of SVZ progenitor cells in non-enriched CX3CR1 / and control mice, and no difference in progenitor cell number between enriched and nonenriched CX3CR1 / mice (Fig. 5d). 3.6. Effects of odor and environmental enrichment on OB microglia In the olfactory bulb, microglia number and soma area did not differ between the WT, CX3CR1+/ and CX3CR1 / groups and were not influenced by either odor or environmental enrichment (p > 0.5) (Fig. 5g–h). Similarly, there were no strain or enrichment effects on SVZ microglia (p > 0.1) (Fig. 5i). 4. Discussion The results of this study demonstrate that increases in hippocampal microglia number and soma area, induced by environmental enrichment, as well as by microglial CX3CR1 signaling deficiency, are associated with improved hippocampal-dependent learning and memory processes. Chronic administration of CX3CR1 blocking antibody also facilitated hippocampal-dependent memory functioning and positively regulated microglial status. In contrast, olfactory bulb microglia were not altered in mice exposed to odor enrichment or in CX3CR1-deficient mice, although these mice displayed markedly improved odor memory. In control mice neurogenesis in the hippocampus and the OB (but not SVZ) was facilitated following EE and OE, respectively, however, in CX3CR1-deficient mice these enrichment procedures did not produce any effect on neurogenesis. These findings suggest that microglia and CX3CR1 signaling are involved in modulation of learning, memory and neurogenesis in a region-specific manner. Previous studies demonstrated that the number of microglia is enhanced in the hippocampal DG of environmentally-enriched rats (Ziv et al., 2006), and that the proliferation of cortical microglia is increased following 4-weeks exposure to running wheels in mice (Ehninger and Kempermann, 2003). In contrast, voluntary exercise in running wheels over 10 days did not induce proliferation and activation of DG microglia (Olah et al., 2009), and similar exposure to running wheels for 14 days produced an increase in hippocampal microglia number that did not reach statistical significance (Vukovic et al., 2012). Together with the findings of the present study, it may be suggested that long-term (several weeks) exposure to environmental enrichment (which includes availability to running wheels) or to voluntary exercise by itself induces hippocampal microglia proliferation and activation, whereas shorter

246

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

Fig. 3. Effects of CX3CR1 blocking antibody on memory functioning, neurogenesis and microglia status. (a) Contextual fear conditioning (reflected by % freezing during the memory retention test) was significantly increased following administration of CX3CR1 blocking antibody (CX3CR1-Ab) in WT mice, as compared with their vehicle-treated controls (t(16) = 2.43, p < 0.05) (n = 8–9/group). (b) Auditory-cued fear conditioning memory was not affected by the CX3CR1-Ab administration. (c) In the water maze paradigm, CX3CR1-Ab treated mice exhibited improved spatial learning, i.e., displayed reduced latencies to reach the platform as compared with vehicle-treated mice (F(1,15) = 3.49, p < 0.05) (n = 8–9/group). (d) Hippocampal neurogenesis, determined by counting the number of doublecortin (DCX)-labeled cells in the dentate gyrus (DG), was not affected by the CX3CR1-Ab administration. (e) Representative pictures of the DG from a vehicle-treated control mouse and (f) a CX3CR1-Ab treated mouse, demonstrating no differences in the number of DCX-labeled new neurons in the DG (Blue = DAPI-labeled cell bodies, Red = DCX-labeled new neurons). (g) The density of hippocampal microglia (number/mm2 DG) in vehicle-treated mice was greater in the implanted vs. the non-implanted (intact) hemisphere. In CX3CR1-Ab treated mice hippocampal microglial density in the intact hemisphere was greater than in the corresponding hemisphere of vehicle treated mice, but it was not further increased in the implanted hemisphere. These findings were reflected by treatment (vehicle vs. CX3CR1 blocking antibody) by implantation side interaction (F(1,12) = 13.1, p < 0.05)(n = 5/group). (h) Representative pictures of microglia density in the DG of the intact hemisphere and (i) the injection cannula-implanted hemisphere of vehicle-treated mice, showing the marked implantation-induced microglial activation. (j) Representative picture of microglia density in the DG of the intact hemisphere and (i) the injection cannula-implanted hemisphere of CX3CR1-Ab treated mice, showing the small Ab-induced increase in microglia density in the intact hemisphere, (k) with no further increase in density in the implanted hemisphere. ⁄p < 0.05 compared with vehicle-treated mice. ⁄⁄p < 0.05 compared with the DG on the intact hemisphere in vehicle-treated mice. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

Fig. 4. Effects of olfactory enrichment and CX3CR1 deficiency on olfactory memory. (a) Following the short-delay interval (30 min), enriched and non-enriched WT mice, as well as (b) enriched and non-enriched CX3CR1+/ mice displayed significant odor recognition, reflected by reduced investigation time of the same odor (specific condition), but not a different odor (non-specific condition), during its second presentation over the investigation time on the first presentation. Following the long delay interval (360 min) enriched WT and CX3CR1+/ mice still exhibited significant odor recognition, which was no longer evident in non-enriched WT and CX3CR1+/ non-enriched mice. These findings were reflected by significant 3-way interactions between odor specificity by enrichment by interval (F(1,23) = 4.3, p < 0.05 and F(1,20) = 7.1, p < 0.05 for the WT and CX3CR1+/ groups, respectively) (n = 7–8/group). (c) Non-enriched and enriched CX3CR1 / mice displayed significant odor recognition both following the short-delay (30 min) and the long-delay (360 min) intervals. These findings were reflected by a significant specificity effect (F(1,24) = 15.39, p < 0.05), with no further interactions with the enrichment or the interval variables (n = 8/group). ⁄p < 0.05 compared with the corresponding odor recognition in the non-specific condition.

periods of exercise are not sufficient to produce microglial changes. Ample evidence demonstrates that rearing in EE and habitual physical activity result in improved hippocampal-dependent learning, memory and neurogenesis (van Praag et al., 2000), but only three previous studies assessed the temporal association between the effects of these manipulations on microglia and on neurogenesis (cognitive functioning was not assessed in these studies), with somewhat conflicting results. Specifically, in one study the EEinduced increases in microglia number was temporally associated with enhanced neurogenesis (Ziv et al., 2006), in a second study exercise induced neurogenesis facilitation without alterations in microglia (Olah et al., 2009), and in a third study EE-induced

247

increases in microglia were not accompanied by elevated neurogenesis (Williamson et al., 2012). The results of the present study support the first report, demonstrating that at least in the EE model cognition and neurogenesis were improved concomitantly with the alterations in the microglial status. In contrast with the effects of enrichment on hippocampal microglia, long-term (6 weeks) odor enrichment produced no effects on microglia in the OB (as well as no effect on microglia in the hippocampus and the SVZ). This lack of effect contrasts with the positive effect of OE on olfactory recognition memory and enhanced OB neurogenesis. The latter effect is related to the enhanced survival of OB adult-born neurons, because OE had no effect on the proliferation of progenitor cells in the SVZ. The regional specificity in the effects of enrichment on microglia in the hippocampus and OB and their relations to memory and neurogenesis can be related to the anatomical and physiological differences between the two systems. Such differences include the neurochemical milieu, for example IGF-1, which is known to regulate plasticity in the hippocampus, is not active in the olfactory bulb (Lledo et al., 2006). Regional differences in microglial morphology (Lawson et al., 1990) and phenotype (evident by the secretion and expression of different molecules) (Ren et al., 1999) have been also reported. It should be noted that OE-induced microglial alterations, parallel to those occurring in the hippocampus following EE, may occur in other regions of the olfactory system, rather than in the OB. Indeed, plasticity of the major components of olfactory perceptual organization and learning is mediated by cortical areas such as the piriform cortex (Lebel et al., 2001; Roman et al., 1993; Wilson et al., 2003). Thus, the possibility that OE induces microglial changes in cortical regions of the olfactory system should be tested in future studies. Signaling via the microglial CX3CR1 constitutes an important regulatory mechanism, by which neuronally-derived CX3CL1 (fractalkine) restrains microglial over-activation during various inflammatory conditions (Biber et al., 2007; Cardona et al., 2006; Harrison et al., 1998; Ransohoff et al., 2007; Wolf et al., 2013). Indeed, CX3CR1 / mice were shown to exhibit increased susceptibility to microglial over-activation (including enhanced IL-1b production) and neurotoxicity following exposure to various inflammatory challenges (Cardona et al., 2006). Our findings demonstrate, for the first time, that under normal physiological conditions genetic CX3CR1 deficiency is also associated with microglial alterations, including increased number and enlargement of the soma. These morphological parameters are suggestive of a mild activation state (Walker et al., 2014; Kreutzberg, 1996), albeit different from the usual inflammatory activation of these cells, because it was not accompanied by expression of classical activation markers, such as MHCII or TNFa. A similar increase in microglial number was also found in the intact hemisphere of CX3CR1 blocking antibody-treated mice. Interestingly, in the cannula implanted hemisphere this antibody prevented any further increase in microglial number. The latter finding may be explained by a suppressive effect of the CX3CR1 blocking antibody on microglial chemoattraction (influencing adjacent areas or infiltrating macrophages, which can enter the brain under certain conditions (Wohleb et al., 2013)). Such an effect has been previously demonstrated in vitro (Harrison et al., 1998), but is shown here for the first time in vivo, in relation to the cannula-induced injury. This effect is probably specific to the CX3CR1 blockade because administration of an irrelevant (control) IgG often produces no microglial alterations and when such alterations are found locally they include microglial activation (e.g., (Wilcock et al., 2003; Wilcock et al., 2004)), rather than suppression. We and others have previously provided experimental evidence and theoretical conceptualization for an inverted U-shape relationship between the hippocampal inflammatory state

248

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

Fig. 5. Effects of olfactory and CX3CR1 deficiency on neurogenesis in the olfactory bulb and SVZ and on microglia density. (a) In WT and CX3CR1+/ mice, OB neurogenesis, determined by counting the number of BrdU/NeuN double labeled cells, was significantly increased following odor enrichment, whereas in CX3CR1 / mice enrichment had no effect on neurogenesis. These findings were reflected by a significant strain by enrichment interaction (F(2,12) = 6.12, p < 0.05) (n = 5/group). (b) A representative picture of new neurons in the olfactory bulb of a non-enriched control CX3CR1+/ mouse (Green = NeuN (a neuronal marker), Red = BrdU, Yellow = double staining). Inset in the right corner shows a part of this OB at a higher magnification. (c) Similar depiction of the OB in an enriched CX3CR1+/ mouse, showing increased number of new OB neurons. (d) The number of progenitor cells (BrdU-labeled) in the subventricular zone (SVZ) was similar in non-enriched and enriched WT, CX3CR1+/ and CX3CR1 / mice (p > 0.1) (n = 4/group). (e) A representative picture of BrdU-labeled (red) cells in the SVZ of aCX3CR1+/ mouse. (f) A picture of the same area, showing CX3CR1-GFP labeled microglia (green) in addition to the BrdU-labeled cells. (g) Microglia density (number/mm2 OB) did not differ between the WT, CX3CR1+/ and CX3CR1 / groups and was not influenced by odor enrichment (p > 0.1) (n = 4/group). (h) A representative picture of OB microglia (Green = CX3CR1-GFP, Blue = DAPI). (i) Microglia density in the SVZ of WT, CX3CR1+/ and CX3CR1 / mice was not altered by odor enrichment (p > 0.1) (n = 4/group). ⁄p < 0.05 compared with the corresponding non-enriched control group. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

(particularly IL-1 production and microglial activation) and cognitive functioning, whereby a small increase in the inflammatory state is beneficial whereas a larger increase is detrimental to behavioral and neural plasticity (Goshen et al., 2007; Goshen and Yirmiya, 2009; Yirmiya and Goshen, 2011). This hypothesis was recently corroborated by an elegant study demonstrating that microglial depletion produces learning impairments in multiple learning and memory tasks (Parkhurst et al., 2013). According to this conceptualization, the increased number and mild activation of microglia observed in CX3CR1 / mice, as well as following exposure to EE and in the intact hemisphere of CX3CR1 blocking antibody-treated control mice, underlies the enhanced learning and memory functioning observed in these mice. (Fig. 6a,b). This model may also explain the lack of EE-induced improvement in hippocampal-dependent learning and memory in the CX3CR1 / mice, as in these mice microglia-associated cognitive facilitation already reaches maximal levels, and therefore any further microglial activation does not elicit further memory improvement (Fig. 6c). Interestingly, the involvement of CX3CR1 signaling in memory seems to be specific to the hippocampus, because in the auditory-cued FC paradigm (which does

not depend on hippocampal mechanisms) CX3CR1 / mice, as well as CX3CR1-blocking antibody-administered WT mice did not show memory superiority. Furthermore, in that task CX3CR1 / mice did undergo EE-induced memory facilitation, similarly to WT and CX3CR1+/ control mice. The present findings and their interpretation are in agreement with the results of a previous study reporting improved spatial memory in CX3CR1 / mice (Maggi et al., 2011), but are inconsistent with the findings of another study (Rogers et al., 2011), demonstrating that CX3CR1 / mice exhibited impairments in contextual and spatial memory, which could be rescued by blockade of IL-1 signaling. The basis for these contradictory results is not clear, but it may be suggested that some environmental or procedural parameters in the various laboratories differentially influenced microglia activation, such that in some instances (Rogers et al., 2011) the over-sensitive microglia in the CX3CR1 / mice were driven to the right side of the inflammation-memory functioning inverted U-curve (i.e., the detrimental zone of activation) (Fig. 6d). CX3CR1 signaling is also important for hippocampal neurogenesis, as evidenced by reduced neurogenesis in non-enriched

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

249

Fig. 6. A model of the relations between hippocampal microglia activation and neurobehavioral functioning. (a) Under normal physiological conditions, hippocampal microglia participate in memory functioning and neuroplasticity processes, exemplified by the finding that microglial depletion (i.e., left side of the curve) produces learning impairments in multiple learning and memory tasks (Parkhurst et al., 2013). (b) A small increase in microglia number and activation, which was evident in CX3CR1 / mice, as well as in control mice exposed to EE or in the intact hemisphere of CX3CR1 blocking antibody-treated mice, induces cognitive facilitation. It should be noted that the EE condition resembles more closely the natural normal habitat of rodents. (c) Exposure of CX3CR1 / mice to EE does not cause a further increase in microglia and cognitive functioning (present results and Maggi et al., 2011), because the microglia are already functioning at optimal levels. (d) In some circumstances, CX3CR1 / mice exhibit overactivation even under regular rearing conditions, resulting in reduced cognitive functioning (e.g., Rogers et al., 2011).

CX3CR1+/ and CX3CR1 / mice compared with WT mice, consistently with previous reports (Bachstetter et al., 2009; Maggi et al., 2011; Rogers et al., 2011; Vukovic et al., 2012). The role of CX3CR1 in hippocampal neurogenesis may be age dependent because in a previous study chronic administration of CX3CR1 blocking antibody was found to reduce hippocampal neurogenesis in young, but not middle aged or old rats (Bachstetter et al., 2009). The findings that in the present study chronic administration of CX3CR1 blocking antibody had no effect in young mice may be explained by different developmental dynamics of neurogenesis modulation in different species (rats in the previous study vs. mice in the present study) and by the fact that in contrast with the previous study the animals in the present study underwent fairly extensive learning and memory testing, which may have interacted with the modulatory effects of CX3CR1 signaling on neurogenesis. Interestingly, the importance of CX3CR1 signaling per se for neurogenesis is restricted to the hippocampus, since in the OB there were no differences among the non-enriched three genotype groups. In addition to its involvement in the neurogenesis process under normal conditions, CX3CR1 signaling may be also important for mediation of the effects of both EE and OE on neurogenesis, since both in the hippocampus and the OB the respective enrichment procedure was associated with increased neurogenesis in WT and CX3CR1+/ but not in CX3CR1 / mice. The latter finding also suggests an age-dependent effect of CX3CR1 signaling deficiency and its interaction with EE because in a previous study adolescent female CX3CR1 / mice (as opposed to the adult CX3CR1 / male mice used in the present study) did show EE-induced neurogenesis facilitation (Maggi et al., 2011). The findings do not support any clear relationship between microglia activation, memory functioning and neurogenesis, because in CX3CR1 / mice the increased microglia activation and improved memory functioning were not correlated with enhanced hippocampal neurogenesis. Furthermore, preliminary results demonstrated that the alterations in hippocampal microglia in the CX3CR1 / and EE-reared CX3CR1+/ mice were not restricted to the DG (the sole location of neurogenesis in the hippocampus), but could be also observed in other hippocampal regions, including CA3 and CA2. Together these findings suggest that microglia proliferation and activation influence memory functioning via mechanisms that do not involve neurogenesis, e.g., by

synapse stabilization, pruning and elimination (Wake et al., 2009; Paolicelli et al., 2011; Tremblay et al., 2011). This interpretation is consistent with recent evidence that adult neurogenesis is necessary for pattern separation and odor discrimination but not for other forms of spatial tasks (Sahay and Hen, 2007; Deng et al., 2010). In the OB, basal levels of neurogenesis were also similar between WT, CX3CR1 / and CX3CR1+/ , despite the better olfactory memory exhibited by the CX3CR1 / mice. These findings suggest that in CX3CR1 / mice there is a complete dissociation between OB-related neurogenesis, memory functioning and microglia status. As mentioned earlier, it is still possible that relations between these parameters do exit in higher centers of the olfactory system that may be more relevant to the neural plasticity underlying the olfactory recognition memory task. In conclusion, our results demonstrate a brain region-specific relations between microglial status and memory functioning. In the hippocampal system, increases in microglial number and soma area induced by either EE or by genetic blockade of CX3CR1 signaling are associated with improved memory functioning. Furthermore, induction of CX3CR1 signaling deficiency by long-term administration of CX3CR1 blocking antibodies was also found to positively regulate microglial status and to have cognitive enhancing properties. In contrast, in the OB system memory functioning is improved both by exposure to OE and by CX3CR1 deficiency, but microglia in this area are not altered by these manipulations. Moreover, although CX3CR1 signaling by itself (in non-enriched mice) does not seem to be related to basal levels of neurogenesis, the increases in DG and the OB neurogenesis following EE or OE, respectively, may be related to CX3CR1 deficiency as they do not occur in CX3CR1-deficient mice. Conflict of interest statement None declared. Acknowledgments We thank Dr. Steffen Jung, the Weizmann Institute, Israel, for the CX3CR1-GFP mice, Ms. Noa Rahamim for help with running the experiments, and Ms. Zehava Cohen for help in preparation

250

R. Reshef et al. / Brain, Behavior, and Immunity 41 (2014) 239–250

of the figures. This research was supported by the Israel Science Foundation – First Program Grant No. 1357/13 and by the Israel Science Foundation Grant No. 206/12 (to R.Y.).

References Aarum, J., Sandberg, K., Haeberlein, S.L., Persson, M.A., 2003. Migration and differentiation of neural precursor cells can be directed by microglia. Proc. Natl. Acad. Sci. USA 100, 15983–15988. Avital, A., Goshen, I., Kamsler, A., Segal, M., Iverfeldt, K., Richter-Levin, G., Yirmiya, R., 2003. Impaired interleukin-1 signaling is associated with deficits in hippocampal memory processes and neural plasticity. Hippocampus 13, 826– 834. Bachstetter, A.D., Morganti, J.M., Jernberg, J., Schlunk, A., Mitchell, S.H., Brewster, K.W., Hudson, C.E., Cole, M.J., Harrison, J.K., Bickford, P.C., Gemma, C., 2009. Fractalkine and CX(3)CR1 regulate hippocampal neurogenesis in adult and aged rats. Neurobiol. Aging. Biber, K., Neumann, H., Inoue, K., Boddeke, H.W., 2007. Neuronal ‘On’ and ‘Off’ signals control microglia. Trends Neurosci. 30, 596–602. Butovsky, O., Ziv, Y., Schwartz, A., Landa, G., Talpalar, A.E., Pluchino, S., Martino, G., Schwartz, M., 2006. Microglia activated by IL-4 or IFN-gamma differentially induce neurogenesis and oligodendrogenesis from adult stem/progenitor cells. Mol. Cell. Neurosci. 31, 149–160. Cardona, A.E., Pioro, E.P., Sasse, M.E., Kostenko, V., Cardona, S.M., Dijkstra, I.M., Huang, D., Kidd, G., Dombrowski, S., Dutta, R., Lee, J.C., Cook, D.N., Jung, S., Lira, S.A., Littman, D.R., Ransohoff, R.M., 2006. Control of microglial neurotoxicity by the fractalkine receptor. Nat. Neurosci. 9, 917–924. Deng, W., Aimone, J.B., Gage, F.H., 2010. New neurons and new memories: how does adult hippocampal neurogenesis affect learning and memory? Nat. Rev. Neurosci. 11, 339–350. Ehninger, D., Kempermann, G., 2003. Regional effects of wheel running and environmental enrichment on cell genesis and microglia proliferation in the adult murine neocortex. Cereb. Cortex 13, 845–851. Ekdahl, C.T., Kokaia, Z., Lindvall, O., 2009. Brain inflammation and adult neurogenesis: the dual role of microglia. Neuroscience 158, 1021–1029. Gallagher, M., Burwell, R., Burchinal, M., 1993. Severity of spatial learning impairment in aging: development of a learning index for performance in the Morris water maze. Behav. Neurosci. 107, 618–626. Goings, G.E., Kozlowski, D.A., Szele, F.G., 2006. Differential activation of microglia in neurogenic versus non-neurogenic regions of the forebrain. Glia 54, 329–342. Goshen, I., Avital, A., Kreisel, T., Licht, T., Segal, M., Yirmiya, R., 2009. Environmental enrichment restores memory functioning in mice with impaired IL-1 signaling via reinstatement of long-term potentiation and spine size enlargement. J. Neurosci. 29, 3395–3403. Goshen, I., Kreisel, T., Ounallah-Saad, H., Renbaum, P., Zalzstein, Y., Ben-Hur, T., Levy-Lahad, E., Yirmiya, R., 2007. A dual role for interleukin-1 in hippocampaldependent memory processes. Psychoneuroendocrinology 32, 1106–1115. Goshen, I., Yirmiya, R., 2009. Interleukin-1 (IL-1): a central regulator of stress responses. Front. Neuroendocrinol. 30, 30–45. Harrison, J.K., Jiang, Y., Chen, S., Xia, Y., Maciejewski, D., McNamara, R.K., Streit, W.J., Salafranca, M.N., Adhikari, S., Thompson, D.A., Botti, P., Bacon, K.B., Feng, L., 1998. Role for neuronally derived fractalkine in mediating interactions between neurons and CX3CR1-expressing microglia. Proc. Natl. Acad. Sci. USA 95, 10896–10901. Jung, S., Aliberti, J., Graemmel, P., Sunshine, M.J., Kreutzberg, G.W., Sher, A., Littman, D.R., 2000. Analysis of fractalkine receptor CX(3)CR1 function by targeted deletion and green fluorescent protein reporter gene insertion. Mol. Cell. Biol. 20, 4106–4114. Kettenmann, H., Hanisch, U.K., Noda, M., Verkhratsky, A., 2011. Physiology of microglia. Physiol. Rev. 91, 461–553. Kreisel, T., Frank, M.G., Licht, T., Reshef, R., Ben-Menachem-Zidon, O., Baratta, M.V., Maier, S.F., Yirmiya, R., 2014. Dynamic microglial alterations underlie stressinduced depressive-like behavior and suppressed neurogenesis. Mol. Psychiatry 19, 699–709. Kreutzberg, G.W., 1996. Microglia: a sensor for pathological events in the CNS. Trends Neurosci. 19, 312–318. Lawson, L.J., Perry, V.H., Dri, P., Gordon, S., 1990. Heterogeneity in the distribution and morphology of microglia in the normal adult mouse brain. Neuroscience 39, 151–170. Lebel, D., Grossman, Y., Barkai, E., 2001. Olfactory learning modifies predisposition for long-term potentiation and long-term depression induction in the rat piriform (olfactory) cortex. Cereb. Cortex 11, 485–489. Lledo, P.M., Alonso, M., Grubb, M.S., 2006. Adult neurogenesis and functional plasticity in neuronal circuits. Nat. Rev. Neurosci. 7, 179–193. Maggi, L., Scianni, M., Branchi, I., D’Andrea, I., Lauro, C., Limatola, C., 2011. CX(3)CR1 deficiency alters hippocampal-dependent plasticity phenomena blunting the effects of enriched environment. Front. Cell. Neurosci. 5, 22.

Olah, M., Ping, G., De Haas, A.H., Brouwer, N., Meerlo, P., Van Der Zee, E.A., Biber, K., Boddeke, H.W., 2009. Enhanced hippocampal neurogenesis in the absence of microglia T cell interaction and microglia activation in the murine running wheel model. Glia 57, 1046–1061. Paolicelli, R.C., Bolasco, G., Pagani, F., Maggi, L., Scianni, M., Panzanelli, P., Giustetto, M., Ferreira, T.A., Guiducci, E., Dumas, L., Ragozzino, D., Gross, C.T., 2011. Synaptic pruning by microglia is necessary for normal brain development. Science 333, 1456–1458. Parkhurst, C.N., Yang, G., Ninan, I., Savas, J.N., Yates 3rd, J.R., Lafaille, J.J., Hempstead, B.L., Littman, D.R., Gan, W.B., 2013. Microglia promote learning-dependent synapse formation through brain-derived neurotrophic factor. Cell 155, 1596– 1609. Ransohoff, R.M., Liu, L., Cardona, A.E., 2007. Chemokines and chemokine receptors: multipurpose players in neuroinflammation. Int. Rev. Neurobiol. 82, 187–204. Ren, L., Lubrich, B., Biber, K., Gebicke-Haerter, P.J., 1999. Differential expression of inflammatory mediators in rat microglia cultured from different brain regions. Brain Res. Mol. Brain Res. 65, 198–205. Rochefort, C., Gheusi, G., Vincent, J.D., Lledo, P.M., 2002. Enriched odor exposure increases the number of newborn neurons in the adult olfactory bulb and improves odor memory. J. Neurosci. 22, 2679–2689. Rogers, J.T., Morganti, J.M., Bachstetter, A.D., Hudson, C.E., Peters, M.M., Grimmig, B.A., Weeber, E.J., Bickford, P.C., Gemma, C., 2011. CX3CR1 deficiency leads to impairment of hippocampal cognitive function and synaptic plasticity. J. Neurosci. 31, 16241–16250. Roman, F.S., Chaillan, F.A., Soumireu-Mourat, B., 1993. Long-term potentiation in rat piriform cortex following discrimination learning. Brain Res. 601, 265–272. Sahay, A., Hen, R., 2007. Adult hippocampal neurogenesis in depression. Nat. Neurosci. 10, 1110–1115. Schafer, D.P., Lehrman, E.K., Kautzman, A.G., Koyama, R., Mardinly, A.R., Yamasaki, R., Ransohoff, R.M., Greenberg, M.E., Barres, B.A., Stevens, B., 2012. Microglia sculpt postnatal neural circuits in an activity and complement-dependent manner. Neuron 74, 691–705. Sierra, A., Encinas, J.M., Deudero, J.J., Chancey, J.H., Enikolopov, G., OverstreetWadiche, L.S., Tsirka, S.E., Maletic-Savatic, M., 2010. Microglia shape adult hippocampal neurogenesis through apoptosis-coupled phagocytosis. Cell Stem Cell 7, 483–495. Tremblay, M.E., Stevens, B., Sierra, A., Wake, H., Bessis, A., Nimmerjahn, A., 2011. The role of microglia in the healthy brain. J. Neurosci. 31, 16064–16069. van Praag, H., Kempermann, G., Gage, F.H., 2000. Neural consequences of environmental enrichment. Nat. Rev. Neurosci. 1, 191–198. Vukovic, J., Colditz, M.J., Blackmore, D.G., Ruitenberg, M.J., Bartlett, P.F., 2012. Microglia modulate hippocampal neural precursor activity in response to exercise and aging. J. Neurosci. 32, 6435–6443. Wake, H., Moorhouse, A.J., Jinno, S., Kohsaka, S., Nabekura, J., 2009. Resting microglia directly monitor the functional state of synapses in vivo and determine the fate of ischemic terminals. J. Neurosci. 29, 3974–3980. Walker, F.R., Beynon, S.B., Jones, K.A., Zhao, Z., Kongsui, R., Cairns, M., Nilsson, M., 2014. Dynamic structural remodelling of microglia in health and disease: a review of the models, the signals and the mechanisms. Brain Behav. Immun. 37C, 1–14. Walton, N.M., Sutter, B.M., Laywell, E.D., Levkoff, L.H., Kearns, S.M., Marshall 2nd, G.P., Scheffler, B., Steindler, D.A., 2006. Microglia instruct subventricular zone neurogenesis. Glia 54, 815–825. Wilcock, D.M., DiCarlo, G., Henderson, D., Jackson, J., Clarke, K., Ugen, K.E., Gordon, M.N., Morgan, D., 2003. Intracranially administered anti-Abeta antibodies reduce beta-amyloid deposition by mechanisms both independent of and associated with microglial activation. J. Neurosci. 23, 3745–3751. Wilcock, D.M., Munireddy, S.K., Rosenthal, A., Ugen, K.E., Gordon, M.N., Morgan, D., 2004. Microglial activation facilitates Abeta plaque removal following intracranial anti-Abeta antibody administration. Neurobiol. Dis. 15, 11–20. Williamson, L.L., Chao, A., Bilbo, S.D., 2012. Environmental enrichment alters glial antigen expression and neuroimmune function in the adult rat hippocampus. Brain Behav. Immun. 26, 500–510. Wilson, C.J., Cohen, H.J., Pieper, C.F., 2003. Cross-linked fibrin degradation products (D-dimer), plasma cytokines, and cognitive decline in community-dwelling elderly persons. J. Am. Geriatr. Soc. 51, 1374–1381. Wohleb, E.S., Powell, N.D., Godbout, J.P., Sheridan, J.F., 2013. Stress-induced recruitment of bone marrow-derived monocytes to the brain promotes anxiety-like behavior. J. Neurosci. 33, 13820–13833. Wolf, Y., Yona, S., Kim, K.W., Jung, S., 2013. Microglia, seen from the CX3CR1 angle. Front. Cell. Neurosci. 7, 26. Yirmiya, R., Goshen, I., 2011. Immune modulation of learning, memory, neural plasticity and neurogenesis. Brain Behav. Immun. 25, 181–213. Ziv, Y., Ron, N., Butovsky, O., Landa, G., Sudai, E., Greenberg, N., Cohen, H., Kipnis, J., Schwartz, M., 2006. Immune cells contribute to the maintenance of neurogenesis and spatial learning abilities in adulthood. Nat. Neurosci. 9, 268–275. Ziv, Y., Schwartz, M., 2008. Immune-based regulation of adult neurogenesis: implications for learning and memory. Brain Behav. Immun. 22, 167–176.