Cerebellum DOI 10.1007/s12311-015-0669-5

ORIGINAL PAPER

Maternal Immune Activation Produces Cerebellar Hyperplasia and Alterations in Motor and Social Behaviors in Male and Female Mice Tooka Aavani 1 & Shadna A. Rana 1 & Richard Hawkes 2 & Quentin J. Pittman 1

# Springer Science+Business Media New York 2015

Abstract There have been suggestions that maternal immune activation is associated with alterations in motor behavior in offspring. To explore this further, we treated pregnant mice with polyinosinic:polycytidylic acid (poly(I:C)), a viral mimetic that activates the innate immune system, or saline on embryonic days 13–15. At postnatal day (P) 18, offspring cerebella were collected from perfused brains and immunostained as whole-mounts for zebrin II. Measurements of zebrin II+/− stripes in both sexes indicated that prenatal poly(I:C)-exposed offspring had significantly wider stripes; this difference was also seen in similarly treated offspring in adulthood (~P120). When sagittal sections of the cerebellum were immunostained for calbindin and Purkinje cell numbers were counted, we observed greater numbers of Purkinje cells in poly(I:C) offspring at both P18 and~P120. In adolescence (~P40), both male and female prenatal poly(I:C)-exposed offspring exhibited poorer performance on the rotarod and ladder rung tests; deficits in performance on the latter test persisted into adulthood. Offspring of both sexes from poly(I:C) dams Tooka Aavani and Shadna A. Rana are co-first authors. Electronic supplementary material The online version of this article (doi:10.1007/s12311-015-0669-5) contains supplementary material, which is available to authorized users. * Quentin J. Pittman [email protected] 1

Department of Physiology and Pharmacology, Hotchkiss Brain Institute, Cumming School of Medicine, Health Sciences Centre, University of Calgary, 3330 Hospital Drive NW, Calgary T2N 4N1, Alberta, Canada

2

Department of Cell Biology & Anatomy, Genes & Development Research Group, Hotchkiss Brain Institute, Cumming School of Medicine, Health Sciences Centre, University of Calgary, 3330 Hospital Drive NW, Calgary, Alberta T2N 4N1, Canada

also exhibited impaired social interaction in adolescence, but this difference was no longer apparent in adulthood. Our results suggest that maternal immune exposure at a critical time of cerebellum development can enhance neuronal survival or impair normal programmed cell death of Purkinje cells, with lasting consequences on cerebellar morphology and a variety of motor and non-motor behaviors. Keywords Zebrin . Purkinje cell . Poly(I:C) . Development . Cerebellum . Motor behavior

Introduction The fetal brain is highly vulnerable to environmental insults, among which are disturbances in neural developmental processes resulting from maternal immune activation (MIA) that have been implicated in a number of neurodevelopmental disorders such as schizophrenia (e.g., [1–3]) and autism (e.g., [4, 5]). The association between MIA and neurodevelopmental disturbances is not limited to a single infectious agent, suggesting factors common to the immune response such as neuroinflammation during early development may be the critical link (see review [6]). In an established rodent model of MIA, the viral mimic polyinosinic:polycytidylic acid, poly(I:C), is given to pregnant rodents. Like many other immunogens (e.g., lipopolysaccharide [7, 8]), poly(I:C) causes the production of inflammatory cytokines such as IL-1β, TNF-α, and IL-6 both in the plasma and in the placenta (e.g., [9–11]), all of which have been implicated in fetal brain damage [12, 13]. A number of studies have also shown that the fetal brain continues to express IL-1β, IL-6, IL-13, and IL-17, hours after maternal poly(I:C) administration [14–16].

Cerebellum

MIA results in both structural and functional abnormalities in the brains of adult offspring (e.g., [17–19]). Examples include enlarged lateral ventricles [20], increased cortical thickness [21], and cognitive and social dysfunctions (e.g., [22, 23]). A brain region that may be particularly vulnerable to developmental insults is the cerebellum because it develops over a relatively long time window and is not fully developed until after birth in both humans and rodents [24–26]. The morphology of the cerebellar cortex is traditionally described in terms of lobes and lobules separated by fissures. However, lobulation does not reflect the functional organization of the cerebellum, its connectivity, or its embryology. Rather, the fundamental architecture is revealed by intrinsic differences between subsets of Purkinje cells (PCs) through the expression of molecular markers such as ZII/aldolase C [27]. Differential expression of ZII reveals an elaborate and reproducible pattern of zones and stripes that is conserved in birds and a diverse group of mammals [28, 29], which form the scaffold around which the rest of the cerebellar cortex is organized (reviewed in [30]). The basic zone and stripe architecture of the cerebellum can be manipulated genetically during development, but it is highly resistant to postnatal experimental influences (e.g., [31–33]). The formation and maintenance of the adult ZII+/− stripe array are not dependent on afferent input (e.g., [33–36]), in contrast to isocortical patterns such as ocular dominance columns (e.g., [37]) and whisker barrel fields (e.g., [38]). Also, no effects on cerebellar compartmentation have ever been found as a result of rearing conditions. Therefore, the ZII+/− stripe array serves as a reliable and sensitive indicator of early developmental disturbances. Several distinct stages in cerebellar development might be targets of an in utero insult. PCs are born in the fourth ventricle of the anterior metencephalon (E10–E13 in mice [39–41]). Post-mitotic PCs migrate into the cerebellar anlage where they form a stereotyped array of clusters (E14–E18: reviewed in [31, 42–45]). These clusters disperse between E18 and P20, triggered by Reelin signaling, to form the mature pattern of parasagittal stripes (reviewed in [30]). While many studies have identified cognitive and affective deficits in the offspring of MIA dams (see review by [46]), and strong evidence shows that neuroinflammation during fetal development contributes to structural and morphological changes in the cerebellum, little attention has been paid to alterations in cerebellar cytoarchitecture and its association to motor and non-motor behaviors in the affected offspring across the lifespan. Thus, the present study was designed to examine the effect of MIA on cerebellar PC numbers, stripe morphogenesis, and cerebellum-dependent behaviors of the offspring. We hypothesized that maternal inflammation at E13–E15, a critical time during cerebellar development, would affect PC development and adult behavior in both male and female offspring.

Materials and Methods Generation of Animals for the Study Female C57BL/6 mice housed under specified pathogen-free conditions were mated overnight, and the presence of a vaginal plug marked that day as E0. Twenty-eight pregnant females were individually housed and left undisturbed until E13 at which time gestational drug treatment commenced. They were maintained under a 12-h light-dark cycle (lights on 0700 hours), at a room temperature of 20–22 °C, and received ad libitum access to food and water, as were their offspring. A total of 56 male and female offspring were used in the present study. All procedures used were approved by the University of Calgary Animal Care Committee in accordance with guidelines from the Canadian Council on Animal Care. Prenatal administration of poly(I:C) Between E13 and E15, pregnant C57BL/6 mice were administered once daily intraperitoneal injections of 20 mg/kg poly(I:C) (potassium salt; Sigma, St. Louis, MO) freshly dissolved in 0.9 % saline or equivolume saline (1 mL/kg). The dosage has been previously used in MIA studies [14, 47]. The timing of gestational drug treatment coincided with the end of PC birth (E10–E13) [40] and the initial stages of PC migration into embryonic clusters [30]. Postnatal Handling As several studies have shown that there are sex differences in brain structure and behavior as a consequence of in utero insults [48–50], both male and female offspring were used in the present study. The experimental groups consisted of prenatal poly(I:C)-exposed male (N=14) and female (N=14) offspring, and control groups consisted of prenatal salineexposed male (N=14) and female (N=14) offspring. One male and one female per litter contributed to the experimental and control groups (total of 28 litters). The morphological features of the cerebellum were assessed at P18 and~P120 in two different cohorts of animals. Behavioral tests used to assess motor function (rotarod and ladder rung tests) and sociability were conducted with a cohort of animals, mentioned above, at both ~ P40 and ~ P120 (prior to sacrifice for cerebellar analysis). Cerebellar Anatomy During the Neonatal and Adult Periods Twenty poly(I:C)- or saline-exposed male and female mice, at P18 (n=5/sex) and~P120 (n=5/sex), were deeply anesthetized with sodium pentobarbital (60 mg/kg i.p.) prior to transcardiac perfusion with phosphate-buffered saline (PBS;

Cerebellum

pH 7.4) followed by 4 % paraformaldehyde (PFA) in PBS. The brains were removed and stored in buffered 4 % PFA overnight at 4 °C for immunohistochemistry. ZII Whole-Mount Immunohistochemistry of the Cerebellum Whole-mount immunostaining was performed using a modified protocol originally designed by Sillitoe and Hawkes [51, 52]. The cerebella were post-fixed in Dent’s fixative (methanol:dimethylsulfoxide (DMSO) at 4:1) overnight at 4 °C followed by inactivation of endogenous peroxidase in Dent’s bleach (methanol:DMSO:30 % H2O2 at 4:1:1) overnight. After three 100 % methanol washes, the cerebella were passed through five cycles of chilling to −80 °C and warming to room temperature in 100 % methanol. The cerebella were rehydrated through 50 % methanol, 15 % methanol, and PBS washes for 90 min each, and then enzymatically digested in 10 μg/mL proteinase K (>600 units/mL; Boehringer Mannheim Inc., Quebec, Canada) in PBS at room temperature. After the tissue was rinsed with PBS, it was blocked with PBS containing 2 % non-fat skim milk and 0.1 % Triton X-100 (PBSMT) overnight at 4 °C, and then incubated for 48 h at 4 °C with a well-characterized anti-ZII, a mouse monoclonal antibody produced by immunization with a crude cerebellar homogenate from the electric fish Apteronotus [27], used directly from spent hybridoma culture medium, diluted 1:200 in PBS containing 10 % normal goat serum (NGS), 0.1 % Triton X-100, and 5 % DMSO. The cerebella were washed in PBSMT and then reacted for 24 h at 4 °C with anti-mouse immunoglobulins (1:1000; Jackson ImmunoResearch Laboratories, Ltd.) diluted in PBSMT containing 5 % DMSO. The cerebella were washed with PBSMT for 2 h at 4 °C followed by incubation with PBS containing 0.2 % bovine serum albumin and 0.1 % Triton X-100 (PBT) for 2 h at 4 °C. Cerebella were placed in 0.05 % diaminobenzidine (DAB) solution to visualize horseradish peroxidase activity. The reaction was terminated by PBS washes. Photomicrographs of whole-mount ZII-immunostained cerebella were captured with a SPOT digital camera (Diagnostic Instruments, Sterling Heights, MI, USA) mounted on a Zeiss Stemi SV6 microscope (Zeiss, Göttingen, Germany). Montages were constructed in Adobe Photoshop 7.0. The images were cropped and corrected for brightness and contrast but not otherwise manipulated. The lateral width of the cerebellum (Fig. 1a) and widths of individual ZII+ and ZII− stripes (Fig. 1b) were measured from the digital images by using Adobe ImageReady 7.0. Calbindin (CaBP) Section Immunohistochemistry Following the capture of whole-mount photomicrographs, the cerebella were cryoprotected in sucrose and stored overnight

at 4 °C. Cerebella were cut in half at the midline and cryosectioned sagittally at 35 μm. Sections from the medial part of the cerebellum were kept free floating in wells to be used for calbindin (calcium-binding protein, CaBP) immunostaining, a reliable marker for PCs (e.g., [53, 54]). Immunohistochemistry was carried out as described previously [55]. In brief, tissue sections were rinsed thoroughly, blocked with 10 % normal goat serum (Jackson ImmunoResearch Laboratories, Inc.), and incubated in 0.1 M PBS buffer containing 0.1 % Triton-X and 0.005 % bovine serum albumin (blocking solution) and the mouse monoclonal anti-CaBP (1:1000, Sigma Immunochemicals, St. Louis, MO) for 16–18 h at room temperature. The sections were then rinsed with PBS and incubated for 2 h at room temperature in the secondary goat antimouse immunoglobulins diluted in blocking solution. After PBS washes, sections were placed in a 0.05 % DAB solution to visualize peroxidase activity. Sections were mounted onto slides, cleared, and coverslipped. Photomicrographs of CaBPstained sagittal cerebellar sections were captured using a SPOT digital camera. Images were imported into ImageJ (NIH) for quantification of cerebellar PC numbers. Three sections were counted per cerebellum. Behavioral Tests Motor behaviors and sociability were assessed in a total of 36 poly(I:C)- and saline-exposed male and female mice (n=9/ group) at two time points—adolescence (P40–P55) and adulthood (P120–P135). Tests were conducted in the following sequence: rotarod, ladder rung, and sociability, with a minimum 3-day inter-test interval. The behavioral apparatus were cleaned thoroughly with 70 % ethanol between each mouse to remove residual odors. Rotarod Test The rotarod test is widely used to evaluate motor coordination and balance, and is particularly sensitive to cerebellar dysfunction in rodents [56, 57]. Each mouse was given three daily training sessions consisting of five trials during which it was placed on the drum of an accelerating rotarod initially set at 5 rpm and increasing to 25 rpm. If the animal fell, it was placed back on the drum until a maximum allotted time of 60 s was attained. On test day, the animal was placed on the rotating drum initially set at 5 rpm, with an incremental increase of 5 rpm per trial, up to a final speed of 40 rpm. The latency to fall during each trial was recorded by automatic sensors. Ladder Rung Test Cerebellar lesions result in abnormalities in skilled voluntary movement and have recently been shown to increase gait variability in mice [58]. The ladder rung test is sensitive to subtle sensorimotor disturbances in rodents and is used to assess skilled walking and measures limb placement and coordination [59, 60]. The horizontal ladder rung

Cerebellum

Fig. 1 External anatomy of the cerebellum. a The lateral width of the anti-ZII-immunostained cerebellum was measured in prenatal poly(I:C)and saline-exposed offspring. An arrow is drawn to indicate the measurements of cerebellar width (scale bar=500 μm). b ZII+ stripes in the cerebellar vermis as revealed by anti-ZII whole-mount

immunocytochemistry. The distance between the two ZII+ (P2+) stripes was measured, as shown by a black arrow. The width of the two ZII− (P1−) stripes was measured, as shown by red arrows. The lobules visible in the figure are represented by roman numerals (scale bar=250 μm)

apparatus consisted of horizontal metal rungs and clear Plexiglas side walls elevated 90 cm off the floor. The rungs were situated at varying distances of 1.5 to 2.5 cm across the ladder. At one end of the ladder, there was a neutral start point and at the other end a small box containing home cage bedding to motivate the mice to walk along the rungs. The spacing between the rungs was altered each day to ensure that the animals did not learn the pattern [60]. Each animal was given one training session and three daily tests with five trials per day. Three types of errors were identified and scored from video recordings, which included total misses (when the foot missed the rung completely), full slips (when the foot was placed on the rung and slipped off the rung), and half slips (when the foot was placed on the rung and then slipped off the rung but not a complete fall that would interrupt walking). Errors were identified and scored from the video recordings as previously described [60]. The ladder rung apparatus was cleaned and the Bend box^ containing home bedding was changed between each animal.

Data Analysis

Sociability Test The MIA rodent models have revealed that affected offspring exhibit deficits in social behavior (e.g., [61, 62]). To extend these observations to both males and females across developmental time points, we evaluated sociability in a three-chamber social approach apparatus. On the habituation day, mice were allowed unrestricted access to all chambers for 5 min. A similar pretest was performed the following day to ensure there was no side/chamber preference. On the test day, a novel sex- and age-matched conspecific was placed in a meshed cylinder located in one chamber (social side) while an identical empty cylinder was placed in the other (nonsocial) chamber. The social and non-social chambers were randomly alternated between animals. Each test animal was initially placed in the center chamber, followed by removal of the retractable doors. They were allowed to freely explore the chamber for 5 min, and time spent exploring the social and non-social chambers was scored from video recordings.

Cerebellar Anatomy Initially, the lateral width of the cerebellum and the distance between ZII+/− stripes were analyzed using two-by-two analysis of variances (ANOVAs) with between-groups factors sex (male and female) and gestational treatment (saline and poly(I:C)). As the ANOVAs revealed no effect of sex (p>0.40), male and female data were pooled. Subsequent analyses of the lateral width of the cerebellum, width of ZII+ stripes, width of ZII− stripes, and number of PCs were analyzed by using independent t tests. Percent changes in PC numbers in poly(I:C)-exposed offspring relative to controls were also analyzed. Behavioral Tests To assess latency to fall from an accelerating rotarod, a mixed-design ANOVA with between-groups factors of sex (male and female) and gestational treatment (saline and poly(I:C)), as well as within-subjects factor of speed (rpm), followed by two-by-two ANOVAs for each speed, was conducted on~P40 and~P120 data. Where interactions were found, simple effects analyses were conducted. The ladder rung test consisted of three test trial days. The parameters measured on each of the trials included half slip, full slip, total misses, and total errors. Statistical analysis was conducted on the mean of each of the parameters measured. Therefore, a mixed-design ANOVA with between-groups factors of sex and gestational drug, and within-subjects factor of age (beginning at P40 and P120), followed by two-by-two ANOVAs with between-groups factors of sex and gestational drug for each of the four parameters, were conducted on data collected from~P40 and~ P120 ladder rung tests. Where interactions were found, simple effects analyses were conducted. Lastly, the sociability test data were analyzed in a similar fashion to the ladder rung data except that the dependent measure was time spent in a chamber containing a social conspecific. See effect size table for all behavioral tests results (Electronic Supplementary Table 1).

Cerebellum

As there is now considerable evidence that there are alterations in cortical morphological features in maternal immune exposed offspring, we initially asked if cerebellum size was altered in poly(I:C)-exposed offspring. Independent t tests of the lateral width of the cerebellum (Fig. 2) revealed that the cerebellum was wider [t(18) = 6.74, p < 0.001] at P18 [poly(I:C) 4.06 (s=0.063), saline 3.75 (s=0.130); Fig. 2b], but not at~P120 (Fig. 2c) (data collapsed across sexes). This was not due to alterations in gross body weight, as independent t tests revealed that P18 male and female offspring of saline- and poly(I:C)-treated dams were of equivalent body weight (p>0.20). A 2(sex) by 2(gestational treatment) by 2(age) mixed-design ANOVA was used to compare body weights at adolescence and adulthood; it revealed only a sex by age interaction F(1, 32)=5.63, p=0.024. Simple main effects analyses revealed that P40 male and female offspring

were of similar weight; however, by ~ P120, males weighed more than females regardless of gestational exposure (ps