Three-dimensional morphometric analysis of microglial changes in a mouse model of virus encephalitis: age and environmental influences.

European Journal of Neuroscience, Vol. 42, pp. 2036–2050, 2015

doi:10.1111/ejn.12951

CLINICAL AND TRANSLATIONAL NEUROSCIENCE

Three-dimensional morphometric analysis of microglial changes in a mouse model of virus encephalitis: age and environmental influences Aline A. de Sousa,1 Renata R. dos Reis,1 Camila M. de Lima,1 Marcus A. de Oliveira,1 Taiany N. Fernandes,2 ~es,1 Cristovam G. Diniz,3 Marcia C. K. Sosthenes,1 Joa ~o Giovanni F. Gomes,1 Daniel G. Diniz,1 Nara M. Magalha 1 4 Antonio P. Diniz Jr, Pedro F. da C. Vasconcelos4 and Cristovam Wanderley P. Diniz1,5 Bento-Torres, Jose

~o e ~es em Neurodegenerac ^ncias Biolo gicas, Universidade Federal do Para , Laborato rio de Investigac Instituto de Cie ßa ßo ~o no Hospital Universita rio Joa ~o de Barros Barreto, Bele m, Para , Brazil Infecc ßa 2 ^nia, Bele m, Para , Brazil Avenida Alcindo Cancela, Universidade da Amazo 3 ~o, Cie ^ncia e Tecnologia do Para , Braganc , Brazil Instituto Federal de Educac ß a, Para ßa 4 gicas, Ananindeua, Para , Brazil Instituto Evandro Chagas (IEC), Departamento de Arbovirologia e Febres Hemorra 5 Department of Pharmacology, University of Oxford, Mansfield Road, Oxford, OX1 3QT, UK 1

Keywords: aging, behavioral changes, enriched environment, microglial morphometry, sublethal arbovirus encephalitis

Abstract Many RNA virus CNS infections cause neurological disease. Because Piry virus has a limited human pathogenicity and exercise reduces activation of microglia in aged mice, possible influences of environment and aging on microglial morphology and behavior in mice sublethal encephalitis were investigated. Female albino Swiss mice were raised either in standard (S) or in enriched (EE) cages from age 2 to 6 months (young – Y), or from 2 to 16 months (aged – A). After behavioral tests, mice nostrils were instilled with Piryvirus-infected or with normal brain homogenates. Brain sections were immunolabeled for virus antigens or microglia at 8 days postinfection (dpi), when behavioral changes became apparent, and at 20 and 40 dpi, after additional behavioral testing. Young infected mice from standard (SYPy) and enriched (EYPy) groups showed similar transient impairment in burrowing activity and olfactory discrimination, whereas aged infected mice from both environments (EAPy, SAPy) showed permanent reduction in both tasks. The beneficial effects of an enriched environment were smaller in aged than in young mice. Six-hundred and forty microglial cells, 80 from each group were reconstructed. An unbiased, stereological sampling approach and multivariate statistical analysis were used to search for microglial morphological families. This procedure allowed distinguishing between microglial morphology of infected and control subjects. More severe virus-associated microglial changes were observed in young than in aged mice, and EYPy seem to recover microglial homeostatic morphology earlier than SYPy . Because Piry-virus encephalitis outcomes were more severe in aged mice, it is suggested that the reduced inflammatory response in those individuals may aggravate encephalitis outcomes.

Introduction Sublethal encephalitis caused by viral infection is known to affect behavior and the immune response, and viral diseases of the CNS represent a significant proportion of neurological disabilities, particularly in poor countries (Johnston & Hauser, 2008). Emerging virus infections of the CNS are mainly associated with RNA viruses, many of which cause neurological disease (Olival & Daszak, 2005). The rhabdoviruses are part of the broad group of negative-strand RNA viruses, which includes a number of medically relevant viruses such as avian influenza, measles, Ebola and vesicular stomatitis virus (VSV; Kuzmin et al., 2009). Because VSV has limited human pathogenicity, it has been used as a model of rhabdoviruses in both in vitro and in vivo studies investigating viral adaptive and host immune responses (Reiss et al., 1998; van den Pol et al., 2009).

Correspondence: Dr C. W. P. Diniz, 5Department of Pharmacology, as above. E-mail: [email protected] Received 7 December 2014, revised 16 April 2015, accepted 13 May 2015

Upon infection of the CNS with cytopathic neurotropic viruses, such as vesiculoviruses, virus clearance and survival requires the parenchymal penetration of dendritic cells, T-lymphocytes and microglial activation (McCandless et al., 2008; Liu et al., 2009; Steel et al., 2009). T-cell migration to the brain parenchyma enhances viral clearance in VSV encephalitis (Ciavarra et al., 2006). Piry virus is a member of the Vesiculovirus genus from the Rhabdoviridae family, one of the five families in the Order Mononegavirales that is associated with severe human disease (Ebola virus, avian influenza, rabies, measles and VSV; Kuzmin et al., 2009). Because Piry virus has a limited human pathogenicity and showed similar neuroinvasion and pathogenicity to the VSV virus, it was used as a model in a series of previous experiments to investigate the influence of enriched environment on sublethal encephalitis outcomes (de Sousa et al., 2011). It has been demonstrated that changes conferred by environmental enrichment lead to less CNS neuroinvasion and/or more rapid T-cellassociated viral clearance without neuronal damage. It has been found that, upon viral infection, young adult mice housed in an enriched

© 2015 Federation of European Neuroscience Societies and John Wiley & Sons Ltd

Microglial morphology in virus infection 2037 environment exhibit greater T-cell infiltration, less CNS cell infection by the virus, faster viral clearance, less microgliosis and less damage to the extracellular matrix compared with animals housed in standard cages (de Sousa et al., 2011). There is now unambiguous evidence that senescence is associated with functionally relevant changes in the immune system. Among these are changes in T-cell subsets, secreted cytokine profile, cellular replicative capacity and production of antibodies, which together create a proinflammatory environment that is associated with decreased responsiveness to new antigens (de Ara ujo et al., 2013). Based on the previous observation that physically active older adults show less functional decline in T- and B-cellmediated adaptive immunity than age-matched adults with a sedentary lifestyle (Simpson & Guy, 2010), it was hypothesized that aged mice raised in an standard environment (S) would show more intense behavioral and microglial morphological changes than mice raised in an enriched environment (EE) when subjected to viral infection. Therefore, the previous investigation was extended to investigate microglia morphological response in the senescent immune system using a newly established murine model of sublethal arbovirus encephalitis. Typically, microglia are classified into two main morphological types: the first is a ramified morphology, found in microglia in the homeostatic healthy brain; the second is an amoeboid morphology, representative of activated microglia found at sites of brain injury. However, these morphologies are only extreme examples of a dynamic process; microglial morphology may change in association with neuroprotective, proinflammatory, cytotoxic, immunoregulatory and repair functions (Hanisch & Kettenmann, 2007; Benarroch, 2013; Miyamoto et al., 2013; Gomez-Nicola & Perry, 2015). In fact, a variety of different models were proposed to characterize microglial morphological changes after brain damage (Walker et al., 2014). With small variations, it has been learned from these proposals that microglia may displace after injury, at least four different morphological phenotypes: a ramified microglia (Stage 1) changes first into hyper-ramified reactive phenotype (Stage 2), then to a ‘reactive’ state (Stage 3), and finally microglia change to the ‘phagocytic’ morphology (Stage 4; Davis et al., 1994; Streit et al., 1999; Stence et al., 2001). More recently, Yamada & Jinno (2013) proposed a novel classification model of changes in cell morphology following axotomy as a function of time after lesion. However, it was previously demonstrated that, even in the absence of neurological injury, highly reactive microglia phenotypes are expressed during aging as part of an elevated, persistent, proinflammatory profile (Ogura et al., 1994; Godbout & Johnson, 2009). Agerelated physiological changes in microglia include altered cytokine production (Ye & Johnson, 2001; Sierra et al., 2007), altered activation marker expression (Perry et al., 1993; Kullberg et al., 2001) and dystrophic morphologies (Streit et al., 2004). In addition, neuron–microglia crosstalk during aging appears to be deregulated, with a concomitant loss of neuronal-derived factors that control microglial activation (Jurgens & Johnson, 2012). Aged-related induction of proinflammatory microglial profiles in the hippocampus and dentate gyrus are not necessarily associated with behavioral changes (VanGuilder et al., 2011). Moreover, these profiles at least in the dentate gyrus are found more frequently in sedentary subjects than in active subjects (Kohman et al., 2012). Indeed, it has demonstrated that exercise attenuates microglia proliferation, and enhances the expression of a proneurogenic phenotype in the hippocampus and dentate gyrus (Kohman et al., 2012). Furthermore, it has previously been demonstrated that standard laboratory cages provide an impoverished environment that impairs the spatial memory of both adult and aged mice raised in sed-

entary conditions (Diniz et al., 2010). Coherently aged rats raised in similar conditions showed losses in both spatial memory (Long et al., 2009; Bergado et al., 2011) and object recognition memory (Platano et al., 2008). Consistent with this view, an enriched environment and voluntary exercise reduce microglial numbers, suggesting that microglial homeostasis may be also modulated by activity-dependent, functional alterations (Ehninger et al., 2011; Kohman et al., 2012). Taken together these findings suggest that it is important and may be essential to quantify the subtleties of microglial morphology to gain insight into functional variability (Karperien et al., 2013b). Thus, the microglial morphological changes after induction of Piryviral encephalitis in aged albino Swiss mice that have been housed either in an enriched or standard environment to mimic active and sedentary lifestyles, respectively, were examine, and outcomes were compared with those of young mice raised in similar conditions to distinguish the effects of aging from the effects of environment.

Materials and methods All procedures were approved by the institutional animal care committee of the Federal University of Para. One-hundred and seventyfour 2-week-old albino Swiss mice were obtained from the Animal Care Facility of Instituto Evandro Chagas. All animals were handled in accordance with the ‘Guide for the Care and Use of Laboratory Animals’ http://grants.nih.gov/grants/olaw/Guide-for-the-care-anduse-of-laboratory-animals.pdf. Behavioral analysis The young adult mice (n = 59) were kept either in standard environment (S, n = 31) or in an enriched environment (EE, n = 28) for 3 months. Aged mice (n = 115) were maintained in standard (SA, n = 51) or enriched (EA, n = 64) environment for 16–20 months. Mice were then subjected to the following tests: open-field, burrowing, and olfactory memory. The enriched environmental conditions corresponded to plastic cages (32 9 639 9 616.5 cm) with chopped rice straw on the floor, and rod bridges, tunnels, running wheels, and toys made of plastic, wood or metal with a variety of forms and colors that were changed every week. The standard environment cages corresponded to plastic cages with the same dimensions and chopped rice straw on the floor, but without equipment or toys. Each cage housed 12–15 mice. All mice had free access to water and food, and 12 h dark and light cycles were maintained. Tests were performed during the light cycle. Burrowing test For a 2-h daily testing period (from 09:00 to 11:00 h), for 3 consecutive days before inoculation and from post-inoculation days 2 to 35, each of the animals was placed in individual plastic cages (32 9 39 9 16.5 cm) containing a cylindrical PVC tube (20 cm long, 7.2 cm in diameter). These tubes were filled with 250 g of normal diet food pellets. The open end of the tube was supported 3 cm above the floor. After the testing period, the food remaining in the cylinders was weighed and the mice were returned to their collective cage (Deacon et al., 2001). Open-field test The testing apparatus consisted of a gray PVC box (30 9 30 9 30 cm) with the floor divided into nine squares of 10 cm2. For 3 consecutive days before inoculation, each animal was placed

© 2015 Federation of European Neuroscience Societies and John Wiley & Sons Ltd European Journal of Neuroscience, 42, 2036–2050

2038 A. A. de Sousa et al. in one corner and allowed to roam within the apparatus for 5 min. One meter above the open-field, a video camera connected to a computer recorded each training session for later analysis by ANYMAZE software (St€ oelting). The following parameters were analysed: distance traveled (m); mean speed (m/s); and immobility time (s). After each session, the open-field of the apparatus was cleaned with 70% ethanol. Olfactory memory test The test was performed in a box divided into three compartments (one central and two lateral), following a previously published protocol (Soffie & Lamberty, 1988). One of the side compartments contained spoiled straw originating from a cage of mice undergoing testing. The other side compartment contained clean straw. The animals were placed one at a time in the central compartment facing the wall, and a camera recorded their behavior for 5 min. The amounts of time spent in the compartments containing the dirty and clean straw were recorded. The compartment that each animal chose to visit first was also recorded. Experimental groups and inoculation Suckling mice were intracerebrally infected with 10 lL of infected brain suspension, and killed after exhibiting symptoms. The presence of the virus in the brains of animals used to prepare the infected brain homogenate was confirmed by electron microscopy. Homogenized brain samples from neonate diseased animals were prepared for analysis with a Zeiss EM 900 transmission electron microscope, as described elsewhere (Diniz et al., 2006; de Sousa et al., 2011). Viruscontaining brain homogenates were obtained as follows. First, 0.02 mL of viral suspension was intracerebrally inoculated into each newborn mouse, and the animals were observed daily. Upon presenting with clinical signs, the animals were killed and immediately stored at 80 °C. Later, the brain tissue (0.2 g/animal) was macerated and mixed with 1.8 mL phosphate-buffered saline (PBS) containing 100 U/mL penicillin and 100 mg/mL streptomycin. The suspension was cleared by centrifugation at 10 000 g for 15 min at 4 °C. Virus titration was carried out by intracerebral inoculation of newborn mice with 0.02 mL of serial 10-fold dilutions of the viral suspension in PBS, and LD50 values were calculated by the method of Reed and Muench (Thakur & Fezio, 1981). The Piry-virus titers in the sample (LD50/0.02 mL) were 8.0 Log10 equivalent to 108.2 PFU/mL. The viral concentration used (103.2 PFU/mL) was carefully chosen to be a non-lethal dose that induces a sublethal encephalitis. Female albino Swiss mice were raised in either a standard or enriched environment from age 2 to 6 months (young, Y) or from age 2 to 16 months (aged, A). After behavioral tests, all animals were anesthetized with intraperitoneal 2,2,2-tribromoethanol 1% (0.01 mL/g of body weight) and intra-nasally challenged with 5 lL of infected brain homogenate viral suspension (1025 v/v in 100 U/ mL penicillin, 100 mg/mL streptomycin) or with normal, uninfected brain homogenate as a control (10% v/v in 100 U/mL penicillin, 100 mg/mL streptomycin). After recovering, animals were housed in enriched environment or standard cages, and maintained in the Instituto Evandro Chagas (Belem – PA) animal house, where they remained until the end of the experiments. All animals were kept in the care facility at a temperature of 23 2 °C, with access to food and water ad libitum, and with a 12 h light/dark cycle. At 20 or 40 days post-infection (dpi), mice were again subjected to testing (open-field test, burrowing test, and olfactory test).

Statistical analysis of behavioral changes To evaluate the statistical significance of differences observed between groups with respect to the behavioral tasks, three-way ANOVA, two-way ANOVA and two-tail t-tests were used to investigate the influences of age, infection and environment on behavioral outcomes, with differences between groups defined to be significant with a 95% confidence level cutoff (P < 0.05) using IBM SPSS statistics or BIOESTAT 5.0 (Ayres et al., 2007) software. To apply three-way factorial ANOVA or two-way ANOVA to the results of the olfactory discrimination test, the percentage values of the time spent in the spoiled straw were used, taking the sum of the time spent in the clean and in the spoiled straw as 100%. Differences between groups were defined to be significant at 95% confidence level (P < 0.05). Similarly, the three-way, two-way and two-tail t-tests statistical analyses were used to measure the influences of environment, age and infection on burrowing activity and significant differences between the experimental groups. Histology and immunohistochemistry When each animal reached the designated survival time (for young or aged animals), final behavioral tests were performed, and then the mice were weighed and anesthetized with intraperitoneal 2,2,2-tribromoethanol (0.04 mL/g of body weight) and transcardially perfused with heparinized saline, followed by 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.2–7.4). Serial sections (70 mm thickness) were obtained using a Vibratome (Leica VT1000, MK, UK), and were immunolabeled using polyclonal or monoclonal antibodies. To assess the distribution of Piry-viral antigens, CD3-immunolabeled T-cells and microglia in the mouse brain at the different time points, immunohistochemistry was performed on all infected animals and on five uninfected controls. Specific antibodies against Piry-virus species were produced by the Department of Arbovirus and Hemorrhagic Fevers at the Instituto Evandro Chagas, as described elsewhere (Travassos da Rosa et al., 2001; Diniz et al., 2006). In brief, freefloating sections were rinsed in 0.1 M phosphate buffer and placed in a solution of 0.2 M boric acid (pH 9.0) at 70 °C for 1 h for antigen retrieval, then rinsed in 0.1 M PBS with 5% Triton X-100, and incubated in a solution of methanol and 0.3% hydrogen peroxide for 10 min. After washing in PBS, the Mouse-on-Mouse (MOM) Blocking Kit (Vector Laboratories, Burlingame, CA, USA) was used as follows: MOM IgG blocking for 1 h followed by primary antibody for 72 h. The following primary antibodies and dilutions were used: antiCD3 T-lymphocytes 1 : 1000 (AbD Serotec, Oxford, England, UK); anti-IBA-1 (2 lg/mL); and anti-Piry 1 : 20 (Instituto Evandro Chagas, PA, Brazil) in 0.1 M PBS, pH 7.2–7.4 for 3 days at 4 °C, with gentle, continuous agitation. Sections were then incubated in MOM biotinylated anti-mouse IgG reagent for 12 h, washed in PBS and transferred to avidin-biotin-peroxidase complex (ABC) solution (Vector Laboratories, Burlingame, CA, USA) for 1 h. They were washed again before incubation in 0.2 M acetate buffer (pH 6.0) for 5 min, and revealed in GND solution (diaminobenzidine 0.6 mg/mL, ammonium nickel chloride 2.5 mg/mL and glucose oxidase). All steps were carried out under gentle and constant agitation. As a negative control, normal horse serum was added to some slides in place of each of primary antibody used as a cell marker, and slides were processed for immunohistochemistry as previously described. Microglial cell 3D reconstruction and quantitative morphology A NIKON Eclipse 80i microscope (Nikon, Japan), equipped with a motorized stage (MAC6000, Ludl Electronic Products, Hawthorne,


Microglial morphology in virus infection 2039 NY, USA), was used to analyse brain sections. Microglia from the dorsal CA3 subregion of the hippocampus were analysed under oil immersion with a high-resolution, 100 9 oil immersion, plan fluorite objective (Nikon, NA 1.3, DF = 0.19 lm). Images were acquired with NEUROLUCIDA software (MBF Bioscience, Frederick, MD, USA). Although shrinkage in the z-axis is not a linear event, the software used in the present study corrected for this effect, based on previous evidence of 75% shrinkage in the z-axis (Carlo & Stevens, 2011). Without correction, this shrinkage would significantly distort the length measurements along this axis. Only cells with dendritic trees that were unequivocally complete were included for 3D analysis; cells were discarded when dendrite branches appeared artificially cut or not fully immunolabeled. Terminal branches were typically thin. Although many morphological features were analysed, here only those

that showed significant differences between subjects were describe. Twenty-two microglial morphological parameters were estimated and compared, 10 related to the soma and 11 to the microglial branches, as follows: (1) branch length (lm); (2) total branch length (lm); (3) surface area (lm2); (4) branch volume (lm3); (5) total branch volume; (6) segments/mm; (7) tortuosity; (8) fractal dimensions (k-dim); (9) base diameter of the primary branch (lm); (10) total number of segments; (11) planar angle; (12) number of trees; (13) soma area (lm2); (14) soma perimeter; (15) Feret minimum diameter; (16) Feret maximum diameter; (17) compactness; (18) form factor; (19) solidity; (20) roundness; (21) aspect ratio; and (22) convexity. Thus, all of the microglia detected in each area or in lamina of interest were measured multiple times, and dedicated software (Neuroexplorer, MicroBright Field, Williston, VT, USA) was used to process data obtained with

Table 1. Definitions of the morphological parameters used to quantify and compare dorsal CA3 microglial features in albino Swiss mice Branched structure analysis Segment Segments/mm No. of trees Total no. of segments Branch length Total branch length Tortuosity Surface area Branch volume Total branch volume Base diameter of primary branch Planar angle Fractal dimension Cell body Area Perimeter Feret max/min Aspect ratio

Compactness

Convexity

Form factor

Roundness Solidity

Any portion of microglial branched structure with endings that are either nodes or terminations with no intermediate nodes. Number of segments/total length of the segments expressed in mm. Number of trees in the microglia. Refer to the total number of segments in the tree. Total length of the line segments used to trace the branch of interest. Total length for all branches in the tree Mean = [Length]/[Number of branches]. = [Actual length of the segment]/[Distance between the endpoints of the segment].The smallest value is 1; this represents a straight segment.Tortuosity allows segments of different lengths to be compared in terms of the complexity of the paths they take. Computed by modeling each branch as a frustum (truncated right circular cone). Computed by modeling each piece of each branch as a frustum. Total volume for all branches in the tree. Diameter at the start of the 1st segment. Computed based on the endpoints of the segments.It refers to the change in direction of a segment relative to the previous segment. The ‘k-dim’ of the fractal analysis, describes how the structure of interest fills space.Significant statistical differences in k-dim suggest morphological dissimilarities. Refers to the 2D cross-sectional area contained within the boundary of the cell body. Length of the contour representing the cell body. Largest and smallest dimensions of the cell body as if a caliper was used to measure across the contour. The two measurements are independent of one another and not necessarily at right angles to each other. Aspect ratio = [Min diameter]/[Max diameter]. Indicates the degree of flatness of the cell body: • Range of values is 0–1. • A circle has an aspect ratio of 1. pffiffiffiffiffiffi ðp4ÞArea Compactness = MaxDiam • The range of values is 0–1. • A circle is the most compact shape (compactness = 1). Convexity = [Convex perimeter]/[Perimeter]. • A completely convex object does not have indentations, and has a convexity value of 1 (e.g. circles, ellipses and squares). • Concave objects have convexity values less than 1. • Contours with low convexity have a large boundary between inside and outside areas. Form factor ¼ 4p

Area perimeter2

• As the contour shape approaches that of a perfect circle, this value approaches a maximum of 1.0. • As the contour shape flattens out, this value approaches 0. Roundness = [Compactness]2 Use to differentiate objects that have small compactness values. Solidity = [Area]/[Convex area] The area enclosed by a ‘rubber band’ stretched around a contour is called the convex area. • Circles, squares and ellipses have a solidity of 1. • Indentations in the contour take area away from the convex area, decreasing the actual area within the contour.


2040 A. A. de Sousa et al. Neurolucida. These morphological parameters (defined in Table 1) were used to investigate possible features shared by CA3 microglia. For detailed information and definitions, please see: http://mbfbioscience.com/technical-support-center. CA3 limits and sampling rules To define the limits of the CA3 subregion of the hippocampus, architectonic differences in the neuropil region viewed after Wisteria floribunda histochemistry were used, in which CA2 appears much darker than CA3, with a marked difference in the thickness of the CA3 and CA2 pyramidal layer that is readily visible in Nissl-counterstained sections (Fig. 1). Systematic and random samples were taken from a series of sections containing CA3 dorsal material to guarantee that all regions of CA3 dorsal had the same probability of being included in the samples analysed (Fig. 1). Each box inside the outline of CA3 indicates a site from which a single microglial cell was selected for 3D reconstruction. Using this sampling methodology, 640 microglial cells have been reconstructed from eight experimental groups comprising young and aged individuals raised in standard or enriched environments, with and without Piryvirus infection (SA, SY, SAPy, SYPy, EA, EY, EAPy, EYPy), 80 from each experimental group. Microglia morphometry and statistical analyses Because CA3 is located far from the anatomical olfactory regions heavily targeted by Piry virus, microglial branches and soma could be readily distinguished and reconstructed. First, the presence of morphological features shared by the 3D microglial cell reconstructions within each experimental group was investigated. All of the morphometric microglial quantitative variables were subjected to an initial cluster analysis, which included all of the experimental subjects from all of the experimental groups. Cluster analysis involves grouping a set of objects of interest such that objects in the same

Fig. 1. Low-power photomicrography of hippocampus and dentate gyrus from a section immunolabeled with anti-IBA1 antibody to reveal microglia laminar distribution, counterstained with Cresyl violet to define layers and limits of CA3. Note that Nissl counterstaining of the pyramidal layer readily shows the boundary with CA2, showing a conspicuous reduction of thickness of the layer in CA2. A straight line connecting the tips of the granular layer of dentate gyrus was used to define the limit between CA3 and the polymorphic layer of dentate gyrus. The grid defined by x- and y-axes defines the intervals between the square boxes, and illustrates the random and systematic sampling approach. The number of boxes placed on each section was proportional to the area covered by CA3, and a single microglial cell located inside every box was selected for 3D reconstruction. A total of 640 microglia (80 from each group, 20 from each animal) was selected for 3D reconstruction using this sampling strategy.

group (called a cluster) are more similar to each other than to those in other groups (clusters). Multivariate statistical procedures were applied to the sample of microglia to search for potential microglial classes. Distinct microglial classes that were defined based on morphological similarities, as suggested by the cluster analysis, were further assessed with a forward stepwise discriminant function analysis, performed with STATISTICA 7.0 (Statsoft, Tulsa, OK, USA). Discriminant function analysis is used to identify which variables discriminate between two or more naturally occurring clusters by determining whether clusters differ with regard to the mean of a variable, and then using that variable to predict cluster membership. The arithmetic mean and standard deviation were also calculated for the variables found to be the best predictors for the microglial clustering groups. In addition, three-way ANOVA was used to investigate the influences of age, infection and environment on microglial morphological features, with differences between groups defined to be significant with a 95% confidence level cutoff (P < 0.05; ezANOVA, http://www.cabiatl.com/mricro/ezanova/#defs). Photomicrography For photomicrographs, a digital camera was used (AxioCam – ERc, Zeiss, Gottingen, Germany), coupled to a NIKON Eclipse 80i microscope. Digital photomicrographs were processed with ADOBE PHOTOSHOP software; scaling and adjustments to the brightness and contrast were applied to the whole image. To illustrate morphological differences between microglia of each experimental group, 3D microglial reconstructions with morphometric values close to the mean values for individual cells within each group were selected. All 3D microglial reconstructions were performed with images taken at 8 dpi.

Results Piry-virus neuroinvasion Intranasal instillation of Piry virus into mice nostrils leads to initial infection of the olfactory bulb, meninges and blood vessels at 2 dpi. Virus immunolabeling peaks between days 3 and 5 dpi, and after 8 dpi immunolabeling in the parenchyma was rarely detected. At 5 dpi, immunostaining of sectioned mouse brain with anti-Piry-virus antibodies revealed neuroinvasion of the olfactory bulb, olfactory nuclei, olfactory tubercles, piriform cortex, septum, amygdala, ventral hippocampus, hippocampal fimbria, polymorphic layer of the ventral dentate gyrus and the CA3 ventral hippocampal subfield (see de Sousa et al., 2011 for anatomical details). Figure 2A is an anatomical series of sections from the stereotaxic atlas where virus antigenic immunolabeling has been drawn to illustrate the anatomical sequence of neuroinvasion from 2 to 8 dpi. Figure 2B and C shows photomicrographs at low and high power of Piry-virus-immunolabeled sections to illustrate Piry-virus neuroinvasion in the frontal cortex, bulb and olfactory tract (low power), and some morphological details of virus-immunolabeled cells (high power) at 5 dpi. Virus-immunolabeled cell morphologies suggests that both glial and neurons are targeted by Piry virus. Figure 3 shows CA3 sections immunostained with IBA1 to illustrate microglia morphologies from controls and infected aged and young mice after 8 dpi, raised in either standard or enriched cages. Note that infected young mice show more intense microglial morphological changes than infected aged mice, and those individuals from an enriched environment show less intense morphological changes than the ones from standard cages. The microglia selection to represent each group in Fig. 3 was based on the mean values of multiple measurements of each microglia followed by the definition


Microglial morphology in virus infection 2041 A

B

C

Fig. 2. (A) Schematic representation of Piry-virus antigens distribution on a series of stereotaxic horizontal anatomical sections (http://www.mbl.org/atlas232/ atlas232_frame.html) to illustrate the sequence of virus neuroinvasion. Blue circles refer to meningeal immunolabeling, purple triangles illustrate white matter neuroinvasion, yellow and red dots correspond to gray matter neurons and glial immunolabeled cells. (B) Low- and high-power photomicrographs from immunolabeled sections to illustrate Piry-virus antigens in the frontal cortex, and olfactory bulb and tract. (C) Morphological details of Piry-virus-infected cells in high-power pictures. Scale bars: 250 lm (low power); 25 lm (high power).

of the mean microglia of each group after statistical analysis. Microglial pictures in Fig. 3 were taken from the microglia closer to the mean microglia of each group. Microglial morphological changes were significantly more severe in infected young adult mice than in aged infected mice A total of 640 microglial cells were digitally reconstructed in three dimensions from CA3 dorsal hippocampal images of brain sections from albino Swiss mice, comprising 80 from each experimental group. Figure 4A–G shows graphical representations of distinct microglial morphological features to illustrate age, environment and infection influences on microglial morphology. Three-way ANOVA

between-subject factors, pairwise comparisons between independent groups, and the honestly significant difference (HSD) test were used to investigate the influences of age, infection and environment on microglial morphological features. Significant differences were set for pairwise comparisons as [Q = TukeyHSD: P < 0.05, or P < 0.01], and this analysis revealed that age, environment and infection influenced many of the dorsal CA3 microglial morphological features. The variable infection influenced the following morphometric features: branch length (F1,632 = 111, P < 0.000001); surface area (F1,632 = 232, P < 0.000001); branch volume (F1,632 = 397, P < 0.000001); base diameter of primary branches (F1,632 = 496, P < 0.000001); soma area (F1,632 = 151, P < 0.000001); Feret max (F1,632 = 80.1, P < 0.000001); Feret min (F1,632 = 106, P < 0.000001); planar angle (F1,632 = 26.7, P < 0.000001); fractal


2042 A. A. de Sousa et al.

A

B

C

D

Ai

Bi

Ci

Di

E

F

G

H

Ei

Fi

Gi

Hi

Fig. 3. Photomicrographs from IBA1-immunolabeled sections from CA3 sections of young and aged mice to illustrate environmental and aging influences on microglial morphological response to Piry-virus infection. Comparisons between rows illustrate environmental influences and between columns illustrate age influences. Square boxes in low-power pictures in the 1st and 3rd rows are represented in high power in the 2nd and 4th rows, respectively. To select microglia to represent each group, multiple measures of 3D reconstructed microglia were used, and the morphological features of illustrated cells are close to the mean microglia of each group. (A, Ai, C, Ci, E, Ei, G, Gi) Low- and high-power pictures of microglia from CA3 control groups. (B, Bi, D, Di, F, Fi, H, Hi) Lowand high-power pictures from microglia of correspondent infected groups. Sections were counterstained with Cresyl violet. Scale bars: 250 lm (low power); 10 lm (high power).

dimension (F1,632 = 9.50, P < 0.002148); density of segments (F1,632 = 119, P < 0.000001); number of trees (F1,632 = 12.0, P < 0.000580); and number of segments (F1,632 = 35.6, P < 0.000001). The variable age influenced the following morphometric features: branch length (F1,632 = 54.8, P < 0.000001); surface area (F1,632 = 120, P < 0.000001); branch volume (F1,632 = 164, P < 0.000001); base diameter of primary branches (F1,632 = 42.2, P < 0.000001); Feret max (F1,632 = 4.13, P < 0.042533); Feret min (F1,632 = 5.21, P < 0.022742); k-dim (F1,632 = 17.0, P < 0.0000430); density of segments (F1,632 = 45.8, P < 0.000001); number of trees (F1,632 = 3.83, P < 0.050768); and number of segments (F1,632 = 39.2, P < 0.000001). The variable environment

influenced the following morphometric features: surface area (F1,632 = 13.4, P < 0.000277); branch volume (F1,632 = 16.3, P < 0.000061); base diameter of primary branches (F1,632 = 10.0, P < 0.001635); soma area (F1,632 = 18.5, P < 0.000020); Feret max (F1,632 = 20.3, P < 0.000008); Feret min (F1,632 = 11.0, P < 0.000982); planar angle (F1,632 = 6.27, P < 0.012496); number of segments (F1,632 = 9.27, P < 0.002433); and convexity (F1,632 = 4.50, P < 0.034271). In some cases, all variables (environment, age and infection) interact and influence microglial changes as follows: branch volume (F1,632 = 9.58, P < 0.002); mean branch length (F1,632 = 11.2, P < 0.00088); and segments/mm (F1,632 = 12.2, P < 0.00052).



B

C

D

E

F

G

Fig. 4. Graphical representations of mean SEM values of seven distinct CA3 microglial morphological parameters from 640 microglia (80 of each experimental group). (A) Total volume, (B) density of segments, (C) surface area, (D) number of segments, (E) branch volume, (F) number of trees, and (G) branch length. These morphological features were distinguished by discriminant analysis as the ones that most contributed to the cluster formation. (*), (+) and (#) indicate statistical significance, defined by P-values of three-way ANOVA pairwise comparisons between control vs. infected, standard vs. enriched environments, and adult vs. aged mice, respectively. Individual morphological features are defined in Table 1.

These three morphological features significantly contributed to discriminate between morphological families of microglia. Although age, infection and environment simultaneously interact and affect other microglia morphological features, for example, k-dim (F1,632 = 5.06, P < 0.025), planar angle (F1,632 = 9.35, P < 0.0023), Feret max (F1,632 = 7.23, P < 0.0074), Feret min (F1,632 = 3.97, P < 0.047), convexity (F1,632 = 18.2, P < 0.00002) and base diameter of the primary branch (F1,632 = 18.9, P < 0.000016), these features did not contribute significantly to distinguish between the morphological families of microglia. The average volumes of microglial trees (Fig. 4A: total volume) and trees (Fig. 4E: branch volume) of infected mice were 200–300% higher than the correspondent microglial volumes and trees of uninfected mice, independent of the environment or age. Similarly, infected young mice from both enriched and standard environments significantly increased segment density as compared with controls, with a larger effect in the subjects from the enriched environment (Fig. 4B). With the exception of aged infected mice, who show no differences from the corresponding control groups, the average branch length of microglia from infected animals is smaller than that of uninfected mice, and this effect was greater in young than in aged mice (Fig. 4G). On average, microglia from infected subjects showed thicker primary branches, and this is reflected in their higher values of primary base diameters as compared with that of control animals, independent of age or environment (Fig. S1A). Infection, environment and age interact simultaneously and contribute significantly to this effect. Morphological measurements of the microglial cell bodies show an enlargement of soma diameters of microglia of infected animals, and this effect is promptly recognized in the higher values of Feret diameters, max and min (Fig. S1C and D) of infected animals in comparison with the correspondent diameters of control animals. Although microglia from infected young mice showed higher complexity than those from uninfected mice, this effect was not observed

in aged mice from an enriched environment where infection was associated with lower complex microglia (fractal dimension; Fig. S1E). A similar effect was observed on the convexity values of soma convexity mean values of aged mice from an enriched environment. Figure 5 illustrates 3D reconstructions, with corresponding dendrograms of individual microglia, which were selected to have morphological features close to the mean microglia of each experimental group. As compared with control mice, microglial trees from infected mice showed, on average, higher tree and branch volumes, and surface area, and these effects were more evident in young mice from a standard environment. Aged infected mice, as compared with uninfected age-matched mice, showed microglia with thicker and shorter branches, and this effect was more conspicuous in mice from a standard environment. Dendrograms below 3D reconstructions promptly reveal increased branching of microglial processes particularly in infected young mice from a standard environment. Cluster and discriminant analysis across all experimental groups (Fig. 6A and B) suggested the occurrence of four distinct main clusters of microglia. Indeed, the cluster on the left side of the dendrogram contain microglia from infected subjects, and the other two on the left side microglia from infected subjects (Fig. 6A). However, the cluster in the center of the diagram (second cluster from left to right) contain a mixture of microglia from controls and infected aged subjects. Indeed, three of four aged infected mice from the enriched environment group, and two out of four from the standard environment, are clustered together with control animals, suggesting that microglial morphological changes after infection in those aged subjects were less significant. Figure 6B shows a graphical representation of the discriminant analysis. The majority of control (uninfected) subjects are grouped together on the left side of the y-axis, and the infected subjects are dispersed on the right side of the yaxis. This analysis revealed that the total volume (P < 1 9 106), density of segments (P = 0.026), surface area (P = 0.025), total



Fig. 5. CA3 microglia 3D reconstructions from young and aged mice raised in either standard or enriched environments (top), with corresponding linear dendrograms of each microglia arbor with the length of each branch segment displayed to scale as vertical lines and sister branches horizontally displaced (bottom). The microglia depicted here exhibit morphometric features representative of the mean morphological features of other microglia within the relevant experimental group. The dendrogram was plotted and analysed with Neuroexplorer (MicroBrightField). Branches of the same parental (primary branch) trunk are shown in one color. Note that microglia from infected mice show thicker and shorter primary processes than those of uninfected mice, and this effect was more conspicuous in aged mice from a standard environment. Similarly, microglia cell soma from infected mice were larger than from control animals, and this effect was more intense in aged than in young mice. NBH, normal brain homogenate.

number of segments (P = 0.033), total number of trees (P = 0.0054), branch volume (P = 0.027) and branch length (P = 0.0081) were the morphometric features that most contributed to the formation of the clusters. Behavioral outcomes All animals that had their nostrils instilled with infected brain homogenate showed sickness behavioral signs, including ruffled fur, tremor, hunched posture and less exploratory activity. Based on immunohistochemistry for virus antigens it was confirmed that Piry virus targeted olfactory pathways and the limbic system, including the hippocampus and dentate gyrus. Thus, to quantify behavioral outcomes, olfactory and hippocampal-dependent tasks have been chosen. Permanent losses on burrowing activity in aged Piry-virus-infected mice Figure 7 illustrates burrowing activity in young (right) and aged (left) mice measured before and after Piry-arbovirus inoculation (EYPy, SYPy and EAPy, SAPy) compared with mock-infected ani-

mals (EY, SY and EA, SA). Burrowing behavior is quantified by measuring removal of food pellets from a tube. At 7 dpi, aged mice exhibit a significant and permanent decrease in the amount of burrowed food, independent of the environment in which the subjects were raised. Infected young mice from an enriched environment show no burrowing changes, whereas animals from a standard environment showed a significant decrease at 4, 7 and 9 dpi, suggesting that the environmental enrichment may contribute to avoid reduction of burrowing activity in young but not in aged mice. Three-way ANOVA shows no interactions between aging, infection and environment. Permanent olfactory losses in aged infected mice with Piry virus Figure 8 illustrates the results of olfactory tests conducted before (baseline) and after inoculation of Piry-arbovirus-infected brain homogenate in young (SYPy, EYPy) and aged (EAPy and SAPy) mice. Note that aged infected mice from a standard environment were not able to distinguish between familiar and new odors at any of the post-infection time points (8, 20 and 40 dpi). Aged infected mice from an enriched environment were able to distinguish at



B

Fig. 6. Graphical representations of (A) cluster dendrogram and (B) discriminant analysis of CA3 microglial features. This analysis encompasses all morphological features from microglia of all experimental groups. Note the four main classes of microglia distinguished by Euclidian distances above 5000 (Wards method). Two of these clusters (on the right side) correspond to microglia from infected animals (colored triangles), and two others (on the left side) correspond to microglia from all control animals (colored circles) and some individuals (green and blue triangles) from aged infected mice group. Discriminant analysis graphic representation also indicates control and infected animals by colored circles and triangles, respectively (B). SYPy, young adult infected mice housed in an standard environment; SY, young adult control mice housed in an standard environment; EYPy, young adult infected mice housed in an enriched environment; EY, young adult control mice housed in an enriched environment; SAPy, aged infected mice housed in an standard environment; SA, aged control mice housed in an standard environment; EAPy, aged infected mice housed in an enriched environment; EA, aged control mice housed in an enriched environment.

8 dpi, but lost this ability in the later post-infection time points (20 and 40 dpi). As expected, control groups (EA and SA) did not show any impairment. Infection reduced the ability of SYPy to distinguish new from familiar odors at 9 dpi but did not affect EYPy, suggesting that an enriched environment may protect against olfactory impairment produced by viral infection. Table S1 shows descriptive parametric statistics (mean, standard deviations, standard errors, and t- and P-values to illustrate significant differences in the olfactory tests between experimental groups at each time window). Three-way ANOVA showed no interactions between aging, infection and environment in the olfactory tests. Taken together, the current findings revealed that the severity of microglial morphological changes did not show simple correlations with permanent behavioral deficits in aged mice, suggesting that microglial changes in isolation may not explain CNS damage induced by Piry-virus encephalitis during senescence.

Discussion Previously, a model has been established in which Piry-virusinduced encephalitis in adult albino Swiss mice housed under enriched or standard environmental conditions is used to correlate neuropathological features (quantified using a stereological-based unbiased method) with behavioral changes. Outcomes were compared among animals that had been subjected to intranasal inoculation of Piry-virus-infected brain homogenate and individuals that were mock-infected with healthy brain homogenate, and between animals housed under more or less stimulating environments (de Sousa et al., 2011). In the present report, the investigation was expanded to compare young and aged mice using this model, to test the hypothesis that the intensity of microglial changes may be associated with the severity of observed behavioral impairments. Threedimensional reconstructions of microglial cells that had been observed microscopically were made, and multivariate statistical analysis of cell parameters to analyse the significance of the differences observed was used. It was found that aged infected mice, independently of the environment where they were raised, showed permanently reduced burrowing activity and olfactory discrimination. Microglial morphological changes observed in aged mice were less intense than those seen in young mice, indicating that the reduced inflammatory response in immunosenescent individuals may aggravate encephalitis outcomes. A few reports have used microscopic 3D reconstruction of microglia to describe morphological changes under homeostatic and neuropathological conditions (Papageorgiou et al., 2014; Torres-Platas et al., 2014); however, none of the previous investigations

Fig. 7. Graphical representation of food burrowing behavior as a function of disease progression in young (top) and aged (bottom) mice. Note that all infected aged mice, independently of the environment, showed a permanent and significant decrease in the amount of burrowed food from 7 dpi onwards, whereas young infected mice from an enriched environment (EYPy) recovered burrowed activity up to control levels at 9 dpi. SYPy, young adult infected mice housed in a standard environment; SY, young adult control mice housed in an standard environment; EYPy, young adult infected mice housed in an enriched environment; EY, young adult control mice housed in an enriched environment; SAPy, aged infected mice housed in an standard environment; SA, aged control mice housed in an standard environment; EAPy, aged infected mice housed in an enriched environment; EA, aged control mice housed in an enriched environment. (*) indicates significant differences between each experimental group and its respective baseline mean scores. (#) indicates significant differences between mean scores of experimental infected subjects and respective controls. © 2015 Federation of European Neuroscience Societies and John Wiley & Sons Ltd European Journal of Neuroscience, 42, 2036–2050


Fig. 8. Graphical representation of the results in the task of olfactory discrimination. Note that infected aged mice, independent of the environment in which they are raised, permanently lost their olfactory discrimination from 8 dpi onwards. SYPy, young adult infected mice housed in an standard environment; SY, young adult control mice housed in an standard environment; EYPy, young adult infected mice housed in an enriched environment; EY, young adult control mice housed in an enriched environment; SAPy, aged infected mice housed in an standard environment; SA, aged control mice housed in an standard environment; EAPy, aged infected mice housed in an enriched environment; EA, aged control mice housed in an enriched environment. *P < 0.05; **P < 0.001.

endeavored to quantify subtle microglial changes caused by arbovirus infection of mouse brain tissue, including the effect of host age. Here, an experimental system has been established in which quantitative associations between microglial 3D morphology with stereological sampling approach and behavioral changes can be measured in a murine model of encephalitis. In the adult rodent brain, damage can induce microglial morphological changes that included a transition between a process-bearing or more ramified appearance and a rounded or amoeboid morphology with a continuum with multiple intermediate stages (Karperien et al., 2013b; see Harry & Kraft, 2012 for review). To detect and quantify details along this continuum of morphological possibilities established in the present report, stereological random and systematic sample approaches combined with 3D reconstructions of microglia were used. To the authors’ knowledge, so far there are no previous findings associating an unbiased sample approach with 3D microscopic reconstructions to assess morphological microglial phenotypes. This approach was chosen to guarantee that all regions from the area of interest would have the same probability to be included in the sample (systematic and randomized sample), and that fine anatomical details (from 3D reconstructed microglia) could be quantified in both control uninfected and infected brains with unbiased methods. After that, cluster analysis was used to identify possible morphological groups of microglia followed by discriminant analysis to identify which morphometric features significantly contributed to distinguish between microglial morphological families in both control uninfected and infected mice. From this approach, the authors have learned that a few morphological parameters are enough to distinguish four microglial families in the sample as follow: total volume; number of segments; number of trees; branch volumes; surface area; mean branch length; and density of segments. These variables clearly distinguished infected from uninfected young mice microglia but not, so clearly, aged infected from uninfected mice microglia. These findings suggest that the morphological changes associated with microglial response in aged infected mice were not as conspicuous as those observed in young infected mice. To generate unbiased and statistically valid results, random systematic sampling (to guarantee that all regions of CA3 would contribute to the sample with the same probability) and multivariate statistical analysis were combined. It has been shown here that the severity of the behavioral impairments observed was not directly correlated with the intensity of microglial morphological changes associated with

Piry-virus encephalitis. It has also been demonstrated that behavioral and microglial morphological changes were more intense in animals living under standard conditions than in mice housed under enriched environment conditions. In addition, it has been found that more severe microglial morphological changes take place in young adult than in aged mice upon Piry-virus infection. Microglial morphological changes and behavioral deficits The previous study using Piry-virus infection of adult female Swiss mice as an encephalitis model revealed that microglial host response was more intense and generalized in the brain parenchyma at 8 dpi compared with 20 or 40 dpi (de Sousa et al., 2011). These findings suggested that microglial activation may be one of the key factors in the pathogenesis of Piry-virus encephalitis in adult young mice. In the present report, the subtleties of microglial morphological changes in young adult and aged Piry-virus-infected mice at 8 dpi were measured and compared using 3D cell reconstruction, and the association between microglial morphological changes and correspondent behavioral impairments was evaluated. An alternative morphometric approach applied to 3D microglial reconstructions was used, as previously suggested, to be able to quantitate the subtleties of the morphological changes (Harry & Kraft, 2012; Di Ieva et al., 2013; Karperien et al., 2013b). The current findings were consistent with evidence that the morphology of microglia can be used to characterize changes in their functional status (Karperien et al., 2013b). Indeed, other studies have shown that microglial functional status was associated with a combination of extracellular changes in neuronal activity (Tremblay et al., 2010, 2011; Wake et al., 2011, 2013), chemokine signaling (Liang et al., 2009) and purinergic signaling (Davalos et al., 2005; Dibaj et al., 2010; Fontainhas et al., 2011; Ohsawa & Kohsaka, 2011). An inverse association between age, behavioral impairments and microglial host morphological changes was found. Indeed, microglial changes in young adult mice were much more intense than in aged mice, whereas behavioral changes were more severe in aged mice. It was suggested that the decreased intensity of microglial activation in aged mice leads to poor behavioral outcomes. A few reports described the microglial morphological response to virus infection on the aged mice brain (Davies et al., 2004), and none of them used 3D reconstruction of microglia using a stereological unbiased sampling approach. However, a variety of other studies


Microglial morphology in virus infection 2047 demonstrated that the form-function model is at least a good starting point to investigate possible influences of multivariate factors affecting microglial morphology and function (Nimmerjahn et al., 2005; Wake et al., 2009; Karperien et al., 2013b). In the present report, the influences of aging, environment and infection on CA3 microglial morphology were analysed, and an increase in the base diameters of primary branches, branch volume, surface area, total volume and density of segments in infected mice was found, independent of aging or environment where mice were raised. In contrast, an increase of fractal dimension (k-dim) of the microglial morphology in young but not in aged infected mice was found. This statistical index of microglial morphological complexity was associated with an hyper-ramified microglia in CA3 of young mice, similar to that previously described in the prefrontal cortex of rats submitted to chronic stress in the absence of increased inflammation or neurodegeneration (Hinwood et al., 2012). Thus, two main questions related to those morphological changes remain to be answered: what mechanisms are responsible for those microglial changes in CA3? To what physiological implications are the morphological changes related? It has been suggested that an increase in the expression of b1-integrin (CD29) after lipopolysaccharide (LPS) injection alters microglial branches, and this was associated with an increase of integrin immunoreactivity of ramified microglia in the mouse brain (Kloss et al., 2001). Similarly to that found after viral infection in the present report, those changes in microglial morphology in the rat brain were characterized by the enlargement of the microglial cell body, a thickening of the proximal processes and a reduction in distal ramification (Kloss et al., 2001; Hinwood et al., 2012). Microglial morphology of aged uninfected mice shows similar morphological features to microglia previously described in the rodent aged brain (Perry et al., 1993; Streit & Xue, 2010). However, considering that microglia from aged mice are expected to be closer to the phagocyte-like phenotype, why CA3 microglial morphological changes associated to Piry-virus infection of aged mice was less intense than microglial response of infected young mice? How to explain that a less intense microglial response would be associated with more severe behavioral changes? Although the authors do not have a direct answer to that question, it is important to discuss alternative explanations for future investigations. Primarily, morphologically altered microglia, following Piry-virus neuroinvasion, were observed in the olfactory pathways, septal region, amygdala, ventral CA3 and the polymorphic layer of the ventral dentate gyrus. Ventral and dorsal hippocampal-dependent tasks (Deacon et al., 2002; Bannerman et al., 2003; Deacon & Rawlins, 2005) and an olfactory discriminative test (Carr et al., 1976; Soffie & Lamberty, 1988) were selected to investigate behavioral changes. Although the evidence in the present report is indirect and this study is explicitly correlational, it provides the opportunity to formulate hypotheses about the relationships between microglial morphological changes, cytopathic virus infections and behavioral deficits. In addition, this microglial form-function model provides a good starting point to study the host microglia inflammatory response to virus infection under the influence of distinct environments mimicking sedentary (standard environment) and active (enriched environment) lifestyles. The previous findings in adult young mice demonstrated that changes in burrowing behavior of young adult Piry-virus-infected mice became apparent at 8 dpi, when IYPy animals burrowed significantly less food than IY controls and continued to do so until 13 dpi, when burrowing activity recovered to control levels. No significant differences were detected in burrowing between young adult EYPy and EY control animals (de Sousa et al., 2011). In the present

report, previous observations in young mice were confirmed and the findings for aged mice were expanded. The findings in aged mice demonstrate that, independent of the environment, infected subjects start to burrow less food than respective controls at 7 dpi, and continue with reduced burrowing activity until at least 40 dpi, when the subjects were killed. These results are in line with previous reports demonstrating that systemic inflammation reduces burrowing activity, either when induced by LPS to mimic inflammatory responses to bacterial infection (Teeling et al., 2007) or by the toll-like receptor 3 agonist Poly:IC, to mimic inflammatory responses to systemic viral infection (Konat et al., 2009). Because burrowing activity requires hippocampal integrity, and the ventral hippocampus, dentate gyrus and septal regions are targeted by Piry virus, it is suggested that non-inflammatory permanent damage to these areas may contribute, at least in part, to the sustained reduction in food burrowing behavior observed in aged infected mice. Because olfactory nuclei and projections were intensely targeted by Piry virus and microglial morphological changes were more severe within these regions, it was impossible to reconstruct microglial cells along the olfactory pathways. Because of the severity of these changes, it is suggested that olfactory losses, both those observed transiently in young adult mice and those that are observed to be permanent in aged subjects, may be related to non-inflammatory caspase-dependent cell death along the anatomical pathways related to olfactory function. Similarly, the olfactory pathways were also completely damaged during encephalitis induced by nostril inoculation with VSV, which affects very similar anatomical targets within the anatomical olfactory pathways to those targeted by the Piry virus (Lundh et al., 1987; Huneycutt et al., 1993; Bi et al., 1995; Christian et al., 1996). Technical limitations In the present report, multivariate statistical analysis using microglial morphometric parameters revealed at least three, and possibly five, morphologically distinct groups of microglia in the CA3 hippocampal region of infected and uninfected young and aged mice. Two of these types were found in control animals and the other three in infected subjects. However, an important related question remains to be discussed: is it possible microglial changes were missed due to not studying the critical brain region affecting the behavioral tests? Indeed it has been demonstrated, for example in adult rodent, that ischemia can induce microglia to display either a more ramified and bushy appearance or an amoeboid morphology depending on the level of damage and distance from the infarct site (Harry & Kraft, 2012). Thus, to understand similarities and differences or the functional implications of the subtle morphological changes, it is necessary to characterize both anatomical and functional phenotypes in each specific situation, and this includes viral infections (Karperien et al., 2013a). Because the current study did not reconstruct microglia from olfactory pathways where morphological microglial changes were more intense, it is certain that some of the more severe microglial morphological changes were missed in the current sample. These assumptions, however, cannot explain the results in aged mice. Indeed, because aged mice (as compared with young mice) had more severe outcomes (permanent olfactory losses) but less intense changes in microglial morphology in the olfactory pathways, it was suggested that the damage associated to Piry-virus infection in aged mice may not be associated with an exacerbated microglial morphological response. Coherently with this view, in the hippocampaldependent burrowing activity (Deacon et al., 2002), where


2048 A. A. de Sousa et al. microglial morphological changes were less intense than in the olfactory pathways, there were transitory and permanent changes in burrowing activity in young and aged mice, respectively, with correspondent intense and less intense changes in microglial morphology in young and old mice. Microscopic 3D reconstructions may be affected by mechanical factors associated with the vibratome sectioning and the dehydration procedure, which can induce non-uniform shrinkage in the z-axis of the sections (Hosseini-Sharifabad & Nyengaard, 2007). Thus, estimates of modifications in the x/y-dimensions during tissue processing cannot be linearly extrapolated to the z-dimension. These methodological constraints impose limitations that must be taken into consideration when interpreting the results of the present study. However, a reliable indication of severe shrinkage in the z-axis is the curling of branches, indicating that individual processes did not shrink at the same rate as the slice in which they were located. These effects tend to be of higher amplitude at the surface, decreasing in depth along the z-axis. This pattern, however, was not observed in microglia that were reconstructed in this study. Moreover, the samples were taken from the middle region of the z-axis, where the impact of these changes is expected to be minor. More recently, it has been demonstrated that in the z-axis (perpendicular to the cutting surface), sections shrink by approximately 75% of the cut thickness after dehydration and clearing (Carlo & Stevens, 2011). Based on those findings, all microglial reconstructions of the present report were corrected for z-axis shrinkage, expecting shrinkage to 75% of the original value. No corrections were applied to x/ y-axes, as it was expected that these dimensions did not change after histological dehydration and clearing. Another limitation is that no stereological procedures were performed to estimate the total number of each type of microglia and its degree of activation in the target area of aged mice. However, because the criteria used to select individual microglial cells for 3D reconstruction were systematically blind and randomized, and the number of elements selected for reconstruction was rather large (640 in total, 80 in each group), it is reasonable to suppose that no a priori sample bias was induced by the choice of objects of interest among subjects. The specific mechanisms that induced permanent behavioral changes in aged mice and protection in young mice from an enriched environment remain to be investigated. Detailed cellular and molecular analysis of these processes, including characterization of the inflammatory cells mobilized to the brain parenchyma as well as viral neuroinvasion and clearance mechanisms, may elucidate the pathophysiological basis of these events, improving the understanding of potential targets for treatment of neurological disorders.

Supporting Information Additional supporting information can be found in the online version of this article: Fig. S1. Graphical representations of mean SEM values of seven distinct CA3 microglial morphological parameters from 640 microglia (80 of each experimental group), six from branches (A – branch length, B – planar angle, C – Feret max, D – Feret min, E – fractal dimension, F – convexity). (*), (+) and (#) indicate statistical significance, defined by P-values of three-way ANOVA pairwise comparisons between control vs. infected, standard vs. enriched environments, and adult vs. aged mice, respectively. Individual morphological features are defined in Table 1. Table S1. Descriptive statistics (mean, standard deviation, standard error, t-values from t-tests and correspondent P-values) indicating

significant differences between olfactory test scores for all experimental groups.

Acknowledgements Supporting funds were provided by Fundacß~ao de Amparo a Pesquisa do Para – FADESP; Pr o-Reitoria de Pesquisa e P os-Graduacß~ao da Universidade Federal do Para – PROPESP Edital 02-2014-PIAPA; Coordenacß~ao de Aperfeicßoamento de Pessoal de Nıvel Superior – CAPES; Brazilian Research Council – CNPq grant number: 307749/2004-5 and 471077/2007-0 for CWPD; and Financiadora de Estudos e Projetos – FINEP, Instituto Brasileiro de Neurociências – IBNnet.

Abbreviations dpi, days post-infection; HSD, honestly significant difference; LPS, lipopolysaccharide; MOM, Mouse-on-Mouse; PBS, phosphate-buffered saline; VSV, vesicular stomatitis virus.

References de Ara ujo, A.L., Silva, L.C., Fernandes, J.R. & Benard, G. (2013) Preventing or reversing immunosenescence: can exercise be an immunotherapy? Immunotherapy, 5, 879–893. Ayres, M., Ayres, M. Jr., Ayres, D.L. & Santos, A.S. (2007) BioEstat 5.0: Aplicacß~ oes estatısticas nas areas das ciências biol ogicas e medicas. Sociedade Civil Mamiraua, Brasılia CNPq., Belem. Bannerman, D.M., Grubb, M., Deacon, R.M., Yee, B.K., Feldon, J. & Rawlins, J.N. (2003) Ventral hippocampal lesions affect anxiety but not spatial learning. Behav. Brain Res., 139, 197–213. Benarroch, E.E. (2013) Microglia: multiple roles in surveillance, circuit shaping, and response to injury. Neurology, 81, 1079–1088. Bergado, J.A., Almaguer, W., Rojas, Y., Capdevila, V. & Frey, J.U. (2011) Spatial and emotional memory in aged rats: a behavioral-statistical analysis. Neuroscience, 172, 256–269. Bi, Z., Barna, M., Komatsu, T. & Reiss, C.S. (1995) Vesicular stomatitis virus infection of the central nervous system activates both innate and acquired immunity. J. Virol., 69, 6466–6472. Carlo, C.N. & Stevens, C.F. (2011) Analysis of differential shrinkage in frozen brain sections and its implications for the use of guard zones in stereology. J. Comp. Neurol., 519, 2803–2810. Carr, W.J., Yee, L., Gable, D. & Marasco, E. (1976) Olfactory recognition of conspecifics by domestic Norway rats. J. Comp. Physiol. Psych., 90, 821–828. Christian, A.Y., Barna, M., Bi, Z. & Reiss, C.S. (1996) Host immune response to vesicular stomatitis virus infection of the central nervous system in C57BL/6 mice. Viral Immunol., 9, 195–205. Ciavarra, R.P., Stephens, A., Nagy, S., Sekellick, M. & Steel, C. (2006) Evaluation of immunological paradigms in a virus model: are dendritic cells critical for antiviral immunity and viral clearance? J. Immunol., 177, 492–500. Davalos, D., Grutzendler, J., Yang, G., Kim, J.V., Zuo, Y., Jung, S., Littman, D.R., Dustin, M.L. & Gan, W.B. (2005) ATP mediates rapid microglial response to local brain injury in vivo. Nat. Neurosci., 8, 752–758. Davies, C.A., Gollins, H., Stevens, N., Fotheringham, A.P. & Davies, I. (2004) The glial cell response to a viral vector in the aged brain. Neuropath. Appl. Neuro., 30, 30–38. Davis, E.J., Foster, T.D. & Thomas, W.E. (1994) Cellular forms and functions of brain microglia. Brain Res. Bull., 34, 73–78. Deacon, R.M.J. & Rawlins, J.N.P. (2005) Hippocampal lesions, species-typical behaviours and anxiety in mice. Behav. Brain Res., 156, 241–249. Deacon, R., Raley, J., Perry, V. & Rawlins, J. (2001) Burrowing into prion diasease. NeuroReport, 12, 2053–2057. Deacon, R., Croucher, A. & Rawlins, J. (2002) Hippocampal cytotoxic lesion effects on species-typical behaviours in mice. Behav. Brain Res., 132, 203–213. Di Ieva, A., Grizzi, F., Jelinek, H., Pellionisz, A.J. & Losa, G.A. (2013) Fractals in the neurosciences, Part I: general principles and basic neurosciences. Neuroscientist, 20, 403–417. Dibaj, P., Nadrigny, F., Steffens, H., Scheller, A., Hirrlinger, J., Schomburg, E.D., Neusch, C. & Kirchhoff, F. (2010) NO mediates microglial response to acute spinal cord injury under ATP control in vivo. Glia, 58, 1133– 1144.


Microglial morphology in virus infection 2049 Diniz, J., Nunes, M., Travassos da Rosa, A., Cruz, A., De Souza, W., Medeiros, D., Chiang, J. & Vasconcelos, P. (2006) Characterization of two new rhabdoviruses isolated from midges (Culicoides SPP) in the Brazilian Amazon: proposed members of a new genus Bracorhabdovirus. Arch. Virol., 151, 2519–2527. Diniz, D., Foro, C., Rego, C., Gloria, D., de Oliveira, F., Paes, J., de Sousa, A., Tokuhashi, T., Trindade, L., Turiel, M., Vasconcelos, E., Torres, J., Cunnigham, C., Perry, V., Vasconcelos, P. & Diniz, C. (2010) Environmental impoverishment and aging alter object recognition, spatial learning, and dentate gyrus astrocytes. Eur. J. Neurosci., 32, 509–519. Ehninger, D., Wang, L.P., Klempin, F., R€omer, B., Kettenmann, H. & Kempermann, G. (2011) Enriched environment and physical activity reduce microglia and influence the fate of NG2 cells in the amygdala of adult mice. Cell Tissue Res., 345, 69–86. Fontainhas, A.M., Wang, M., Liang, K.J., Chen, S., Mettu, P., Damani, M., Fariss, R.N., Li, W. & Wong, W.T. (2011) Microglial morphology and dynamic behavior is regulated by ionotropic glutamatergic and GABAergic neurotransmission. PLoS One, 6, e15973. Godbout, J.P. & Johnson, R.W. (2009) Age and neuroinflammation: a lifetime of psychoneuroimmune consequences. Immunol. Allergy Clin., 29, 321–337. Gomez-Nicola, D. & Perry, V.H. (2015) Microglial dynamics and role in the healthy and diseased brain: a paradigm of functional plasticity. Neuroscientist, 21, 169–184. Hanisch, U.K. & Kettenmann, H. (2007) Microglia: active sensor and versatile effector cells in the normal and pathologic brain. Nat. Neurosci., 10, 1387–1394. Harry, G.J. & Kraft, A.D. (2012) Microglia in the developing brain: a potential target with lifetime effects. Neurotoxicology, 33, 191–206. Hinwood, M., Morandini, J., Day, T.A. & Walker, F.R. (2012) Evidence that microglia mediate the neurobiological effects of chronic psychological stress on the medial prefrontal cortex. Cereb. Cortex, 22, 1442–1454. Hosseini-Sharifabad, M. & Nyengaard, J.R. (2007) Design-based estimation of neuronal number and individual neuronal volume in the rat hippocampus. J. Neurosci. Meth., 162, 206–214. Huneycutt, B.S., Bi, Z., Aoki, C.J. & Reiss, C.S. (1993) Central neuropathogenesis of vesicular stomatitis virus infection of immunodeficient mice. J. Virol., 67, 6698–6706. Johnston, S.C. & Hauser, S.L. (2008) Neurological disease on the global agenda. Ann. Neurol., 64, A11–A12. Jurgens, H.A. & Johnson, R.W. (2012) Dysregulated neuronal-microglial cross-talk during aging, stress and inflammation. Exp. Neurol., 233, 40–48. Karperien, A., Ahammer, H. & Jelinek, H.F. (2013a) Quantitating the subtleties of microglial morphology with fractal analysis. Front. Cell Neurosci., 7, 3. Karperien, A., Ahammer, H. & Jelinek, H.F. (2013b) Quantitating the subtleties of microglial morphology with fractal analysis. Front. Cell Neurosci., 7, 3. Kloss, C.U., Bohatschek, M., Kreutzberg, G.W. & Raivich, G. (2001) Effect of lipopolysaccharide on the morphology and integrin immunoreactivity of ramified microglia in the mouse brain and in cell culture. Exp. Neurol., 168, 32–46. Kohman, R.A., DeYoung, E.K., Bhattacharya, T.K., Peterson, L.N. & Rhodes, J.S. (2012) Wheel running attenuates microglia proliferation and increases expression of a proneurogenic phenotype in the hippocampus of aged mice. Brain Behav. Immun., 26, 803–810. Konat, G.W., Borysiewicz, E., Fil, D. & James, I. (2009) Peripheral challenge with double-stranded RNA elicits global up-regulation of cytokine gene expression in the brain. J. Neurosci. Res., 87, 1381–1388. Kullberg, S., Aldskogius, H. & Ulfhake, B. (2001) Microglial activation, emergence of ED1-expressing cells and clusterin upregulation in the aging rat CNS, with special reference to the spinal cord. Brain Res., 899, 169–186. Kuzmin, I.V., Novella, I.S., Dietzgen, R.G., Padhi, A. & Rupprecht, C.E. (2009) The rhabdoviruses: biodiversity, phylogenetics, and evolution. Infect. Genet. Evol., 9, 541–553. Liang, K.J., Lee, J.E., Wang, Y.D., Ma, W., Fontainhas, A.M., Fariss, R.N. & Wong, W.T. (2009) Regulation of dynamic behavior of retinal microglia by CX3CR1 signaling. Invest. Ophth. Vis. Sci., 50, 4444–4451. Liu, J., Gong, N., Huang, X., Reynolds, A.D., Mosley, R.L. & Gendelman, H.E. (2009) Neuromodulatory activities of CD4+CD25+ regulatory T cells in a murine model of HIV-1-associated neurodegeneration. J. Immunol., 182, 3855–3865. Long, L.H., Liu, R.L., Wang, F., Liu, J., Hu, Z.L., Xie, N., Jin, Y., Fu, H. & Chen, J.G. (2009) Age-related synaptic changes in the CA1 stratum radiatum and spatial learning impairment in rats. Clin. Exp. Pharmacol. P., 36, 675–681.

Lundh, B., Kristensson, K. & Norrby, E. (1987) Selective infections of olfactory and respiratory epithelium by vesicular stomatitis and Sendai viruses. Neuropath. Appl. Neuro., 13, 111–122. McCandless, E.E., Zhang, B., Diamond, M.S. & Klein, R.S. (2008) CXCR4 antagonism increases T cell trafficking in the central nervous system and improves survival from West Nile virus encephalitis. Proc. Natl. Acad. Sci. USA, 105, 11270–11275. Miyamoto, A., Wake, H., Moorhouse, A.J. & Nabekura, J. (2013) Microglia and synapse interactions: fine tuning neural circuits and candidate molecules. Front. Cell Neurosci., 7, 70. Nimmerjahn, A., Kirchhoff, F. & Helmchen, F. (2005) Resting microglial cells are highly dynamic surveillants of brain parenchyma in vivo. Science, 308, 1314–1318. Ogura, K., Ogawa, M. & Yoshida, M. (1994) Effects of ageing on microglia in the normal rat brain: immunohistochemical observations. NeuroReport, 5, 1224–1226. Ohsawa, K. & Kohsaka, S. (2011) Dynamic motility of microglia: purinergic modulation of microglial movement in the normal and pathological brain. Glia, 59, 1793–1799. Olival, K.J. & Daszak, P. (2005) The ecology of emerging neurotropic viruses. J. Neurovirol., 11, 441–446. Papageorgiou, I.E., Fetani, A.F., Lewen, A., Heinemann, U. & Kann, O. (2014) Widespread activation of microglial cells in the hippocampus of chronic epileptic rats correlates only partially with neurodegeneration. Brain Struct. Funct., doi: 10.1007/s00429-014-0802-0. [Epub ahead of print]. Perry, V.H., Matyszak, M.K. & Fearn, S. (1993) Altered antigen expression of microglia in the aged rodent CNS. Glia, 7, 60–67. Platano, D., Fattoretti, P., Balietti, M., Bertoni-Freddari, C. & Aicardi, G. (2008) Long-term visual object recognition memory in aged rats. Rejuv. Res., 11, 333–339. van den Pol, A.N., Ozduman, K., Wollmann, G., Ho, W.S., Simon, I., Yao, Y., Rose, J.K. & Ghosh, P. (2009) Viral strategies for studying the brain, including a replication-restricted self-amplifying delta-G vesicular stomatis virus that rapidly expresses transgenes in brain and can generate a multicolor golgi-like expression. J. Comp. Neurol., 516, 456–481. Reiss, C.S., Plakhov, I.V. & Komatsu, T. (1998) Viral replication in olfactory receptor neurons and entry into the olfactory bulb and brain. Ann. NY Acad. Sci., 855, 751–761. Sierra, A., Gottfried-Blackmore, A.C., McEwen, B.S. & Bulloch, K. (2007) Microglia derived from aging mice exhibit an altered inflammatory profile. Glia, 55, 412–424. Simpson, R.J. & Guy, K. (2010) Coupling aging immunity with a sedentary lifestyle: has the damage already been done?–a mini-review. Gerontology, 56, 449–458. Soffie, M. & Lamberty, Y. (1988) Scopolamine effects on juvenile conspecific recognition in rats: possible interaction with olfactory sensitivity. Behav. Process., 17, 181–190. de Sousa, A.A., Reis, R., Bento-Torres, J., Trevia, N., Lins, N.A., Passos, A., Santos, Z., Diniz, J.A., Vasconcelos, P.F., Cunningham, C., Perry, V.H. & Diniz, C.W. (2011) Influence of enriched environment on viral encephalitis outcomes: behavioral and neuropathological changes in albino Swiss mice. PLoS One, 6, e15597. Steel, C.D., Hahto, S.M. & Ciavarra, R.P. (2009) Peripheral dendritic cells are essential for both the innate and adaptive antiviral immune responses in the central nervous system. Virology, 387, 117–126. Stence, N., Waite, M. & Dailey, M.E. (2001) Dynamics of microglial activation: a confocal time-lapse analysis in hippocampal slices. Glia, 33, 256– 266. Streit, W.J. & Xue, Q.S. (2010) The brain’s aging immune system. Aging Dis., 1, 254–261. Streit, W.J., Walter, S.A. & Pennell, N.A. (1999) Reactive microgliosis. Prog. Neurobiol., 57, 563–581. Streit, W.J., Sammons, N.W., Kuhns, A.J. & Sparks, D.L. (2004) Dystrophic microglia in the aging human brain. Glia, 45, 208–212. Teeling, J.L., Felton, L.M., Deacon, R.M.J., Cunningham, C., Rawlins, J.N.P. & Perry, V.H. (2007) Sub-pyrogenic systemic inflammation impacts on brain and behavior, independent of cytokines. Brain Behav. Immun., 21, 836–850. Thakur, A.K. & Fezio, W.L. (1981) A computer program for estimating LD50 and its confidence limits using modified Behrens-Reed-Muench cumulant method. Drug Chem. Toxicol., 4, 297–305. Torres-Platas, S.G., Comeau, S., Rachalski, A., Bo, G.D., Cruceanu, C., Turecki, G., Giros, B. & Mechawar, N. (2014) Morphometric characterization of microglial phenotypes in human cerebral cortex. J. Neuroinflamm., 11, 12.


2050 A. A. de Sousa et al. Travassos da Rosa, A.P., Turell, M.J., Watts, D.M., Powers, A.M., Vasconcelos, P.F., Jones, J.W., Klein, T.A., Dohm, D.J., Shope, R.E., Degallier, N., Popov, V.L., Russell, K.L., Weaver, S.C., Guzman, H., Calampa, C., Brault, A.C., Lemon, A.P. & Tesh, R.B. (2001) Trocara virus: a newly recognized Alphavirus (Togaviridae) isolated from mosquitoes in the Amazon Basin. Am. J. Trop. Med. Hyg., 64, 93–97. Tremblay, M.E., Lowery, R.L. & Majewska, A.K. (2010) Microglial interactions with synapses are modulated by visual experience. PLoS Biol., 8, e1000527. Tremblay, M., Stevens, B., Sierra, A., Wake, H., Bessis, A. & Nimmerjahn, A. (2011) The role of microglia in the healthy brain. J. Neurosci., 31, 16064–16069. VanGuilder, H.D., Bixler, G.V., Brucklacher, R.M., Farley, J.A., Yan, H., Warrington, J.P., Sonntag, W.E. & Freeman, W.M. (2011) Concurrent hippocampal induction of MHC II pathway components and glial activation with advanced aging is not correlated with cognitive impairment. J. Neuroinflamm., 8, 138. 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. Wake, H., Moorhouse, A.J. & Nabekura, J. (2011) Functions of microglia in the central nervous system - beyond the immune response. Neuron. Glia Biol., 7, 47–53. Wake, H., Moorhouse, A.J., Miyamoto, A. & Nabekura, J. (2013) Microglia: actively surveying and shaping neuronal circuit structure and function. Trends Neurosci., 36, 209–217. 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., 37, 1–14. Yamada, J. & Jinno, S. (2013) Novel objective classification of reactive microglia following hypoglossal axotomy using hierarchical cluster analysis. J. Comp. Neurol., 521, 1184–1201. Ye, S.M. & Johnson, R.W. (2001) An age-related decline in interleukin-10 may contribute to the increased expression of interleukin-6 in brain of aged mice. Neuroimmunomodulation, 9, 183–192.


A morphometric study of age-dependent changes in mitochondrial population of mouse liver and heart.

Isolation and analysis of mouse microglial cells.

Age- and location-related changes in microglial function.

Ultrastructural changes of mouse brain neurons infected with Japanese encephalitis virus.

Morphometric characterization of microglial phenotypes in human cerebral cortex.

A Mouse Model of Zika Virus Pathogenesis.

Experimental staphylococcal mastitis in the mouse. A morphometric study of early changes in mammary gland structure.

Age changes in genetic and environmental variation in laying hens.

Principal Component Analysis of the Effects of Environmental Enrichment and (-)-epigallocatechin-3-gallate on Age-Associated Learning Deficits in a Mouse Model of Down Syndrome.

Age and Environment Influences on Mouse Prion Disease Progression: Behavioral Changes and Morphometry and Stereology of Hippocampal Astrocytes.

Quetiapine reduces microglial number in the hippocampus of a transgenic mouse model of Alzheimer's disease.

Suppression of microglial activation is neuroprotective in a mouse model of human retinitis pigmentosa.

Microglial activation during epileptogenesis in a mouse model of tuberous sclerosis complex.

PS1dE9 mouse model of Alzheimer's disease.

Age-related changes in mouse bone permeability.

Morphometric changes in the heights and anteroposterior diameters of the lumbar intervertebral discs with age.

Age-related dermal collagen changes during development, maturation and ageing - a morphometric and comparative study.

The consequences of tubulo-interstitial changes for renal function in glomerulopathies. A morphometric and cytological analysis.

Local and Systemic Immune Responses to Influenza A Virus Infection in Pneumonia and Encephalitis Mouse Models.

Mouse encephalitis induced by avirulent Semliki forest virus.

Age-associated changes in DNA methylation across multiple tissues in an inbred mouse model.

Plasmid CDS5 influences infectivity and virulence in a mouse model of Chlamydia trachomatis urogenital infection.

Regulatory role of TRIM21 in the type-I interferon pathway in Japanese encephalitis virus-infected human microglial cells.

Characterization of Barmah Forest virus pathogenesis in a mouse model.