The effect of hemorrhage on the development of the postnatal mouse cerebellum

Experimental Neurology 252 (2014) 85–94

Contents lists available at ScienceDirect

Experimental Neurology journal homepage: www.elsevier.com/locate/yexnr

The effect of hemorrhage on the development of the postnatal mouse cerebellum Ji Young Janice Yoo 1, Gloria K. Mak 1, Daniel Goldowitz ⁎ University of British Columbia, Department of Medical Genetics, Centre for Molecular Medicine and Therapeutics, 950 W. 28th Avenue, Vancouver, BC V5Z 4H4, Canada

a r t i c l e

i n f o

Article history: Received 21 June 2013 Revised 1 October 2013 Accepted 7 November 2013 Available online 16 November 2013 Keywords: Preterm brain injury Preterm cerebellar hemorrhage Mouse model Behavioral phenotype Cerebellar granule cells Gene expression Cerebellar development

a b s t r a c t Recent studies have shown that hemorrhagic injury in the preterm cerebellum leads to long-term neurological sequelae, such as motor, affective, and cognitive dysfunction. How cerebellar hemorrhage (CBH) affects the development and function of the cerebellum is largely unknown. Our study focuses on developing a mouse model of CBH to determine the anatomical, behavioral, and molecular phenotypes resulting from a hemorrhagic insult to the developing cerebellum. To induce CBH in the postnatal mouse cerebellum, we injected bacterial collagenase, which breaks down surrounding blood vessel walls, into the fourth ventricle at postnatal day two. We found a reduction in cerebellar size during postnatal growth, a decrease in granule cells, and persistent neurobehavioural abnormalities similar to abnormalities reported in preterm infants with CBH. We further investigated the molecular pathways that may be perturbed due to postnatal CBH and found a significant upregulation of genes in the inflammatory and sonic hedgehog pathway. These results point to an activation of endogenous mechanisms of injury and neuroprotection in response to postnatal CBH. Our study provides a preclinical model of CBH that may be used to understand the pathophysiology of preterm CBH and for potential development of preventive therapies and treatments. © 2013 Elsevier Inc. All rights reserved.

Introduction The incidence of preterm birth is approximately 11% worldwide (Blencowe et al., 2012). The survivors of premature birth are at an increased risk for various neurodevelopmental disabilities, such as motor deficits (i.e., cerebral palsy), sensory deficits (i.e., visual and auditory impairments), and cognitive, behavioral and socialization deficits (Wood et al., 2005). With the recent improvement in neuroimaging technology and more frequent use of magnetic resonance imaging (MRI), cerebellar abnormalities have become increasingly recognized as one of the major risk factors for these neurodevelopmental problems (Limperopoulos et al., 2007; Volpe, 2009). In particular, cerebellar hemorrhage (CBH) has been reported to be significantly associated with long-term neurological and behavioral sequelae. Recent studies have reported the incidence of CBH in preterm infant is 7–8% (Dyet et al., 2006; Tam et al., 2011), and among the infants who weigh less than 750 g at birth, the incidence rate was found to be as high as 17% (Limperopoulos et al., 2005). The hemorrhagic injury to the preterm cerebellum most frequently occurs from the rupture of fragile immature blood vessels that surround

⁎ Corresponding author at: 950 W. 28th Avenue, Vancouver, BC V5Z 4H4, Canada. Fax: + 1 604 875 3840. E-mail address: [email protected] (D. Goldowitz). 1 These authors contributed equally to the production of this paper and should be considered co-first authors. 0014-4886/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.expneurol.2013.11.010

the cerebellar germinal matrices, the external granule layer (EGL) and ventricular zone (VZ) (Volpe, 2008). During 24 to 40 weeks gestation, granule cells proliferate in the EGL, whereas the interneurons are generated in the VZ. The disruption of this rapid developmental program due to CBH may result in long-term structural and functional alterations in the cerebellum (Volpe, 2009). Indeed, preterm CBH has been shown to cause significant cerebellar volume reduction (Limperopoulos et al., 2010; Messerschmidt et al., 2005), and long-term neurological and behavioral disturbances, such as motor, cognitive, affective, and social function deficits (Limperopoulos et al., 2005, 2007). Despite growing recognition of preterm CBH and its neurodevelopmental outcomes, it is largely unknown how CBH affects the development and function of the cerebellum. To date, detailed examination of the histological and molecular phenotypes that result from CBH has not been possible due to limitations of clinical studies and lack of animal models. Thus, it is clinically important to develop and characterize an animal model of preterm CBH to understand the effects of CBH on the developing postnatal cerebellum. In the present study, we developed an animal model of preterm CBH in neonatal mouse pups. We used pups at postnatal day two (P2), which is developmentally comparable to 26–30 weeks gestation in human (Biran et al., 2012). Using this animal model, we examined anatomical features of the cerebellum, behavioral phenotypes, and expression levels of genes that are relevant to brain injury and neuroprotection after hemorrhagic insult. We hypothesized that postnatal CBH leads to abnormalities in cerebellar development, deficits in cerebellar-specific

86

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

behaviors, and the activation of molecular pathways for brain injury and neuroprotection. Materials and methods Animals This study was conducted in accordance with the guidelines defined by the Canadian Council of Animal Care and was approved by the University of British Columbia Animal Care Committee. In this study, we used ICR (CD1) mice, which first originated from a group of Swiss mice consisting of two males and seven female albino mice derived from a non-inbred stock in the laboratory of Dr. de Coulon, Centre Anticancereux Romand, Lausanne, Switzerland. These animals were imported to the United States by Dr. Clara Lynch of the Rockefeller Institute in 1926. The Hauschka (Ha/ICR) stock was initiated in 1948 at the Institute for Cancer Research (ICR) in Philadelphia from “Swiss” mice of Rockefeller origin. Adult female mice (P60–P150) were individually mated with a male stud. Pregnant females were group housed until E17, upon which they were singly housed for the duration of pregnancy and lactation. The day of birth was considered P0. Pups were kept with their mother until wean age, P21. Determination of method of cerebellar hemorrhage induction and collagenase dose CBH was induced in the pups at P2. Before inducing the hemorrhage, mice were anesthetized with isoflurane (4% isoflurane in oxygen for

induction, 1–2% isoflurane in oxygen for maintenance, from a precision vaporizer). To determine the best method to induce cerebellar hemorrhage in postnatal pups, l μl of autologous blood, 1 μl of bacterial collagenase (Sigma-Aldrich), or 1 μl of sterile saline was injected into the cerebral aqueduct (using a 10 μl Hamilton syringe) of individual mouse pups. The cerebral aqueduct is continuous with the fourth ventricle (Fig. 1A). Injection of autologous blood did not induce pronounced hemorrhaging in the postnatal brain, nor contribute to any area differences in the postnatal cerebellum. Bacterial collagenase was dissolved in sterile saline to obtain a concentration of 2 U/μl. We then serially dilute the 2 U/μl stock to give concentrations of 1 U/μl, 0.5 U/μl, 0.1 U/μl, and 0.05 U/μl. We found that 1 U/μl, 0.5 U/μl, and 0.1 U/μl of collagenase induced 100% mortality in P2 pups when injected into the cerebral aqueduct, whereas 0.05 U/μl of collagenase did not induce hemorrhaging. We then determined whether 0.075 U/μl, 0.06 U/μl, and 0.057 U/μl of collagenase would be more appropriate concentrations of collagenase to inject into the cerebral aqueduct. We found that a concentration of 0.057 U/μl would provide an adequate number of surviving P2 pups within a litter, post-hemorrhage induction, and that appropriate hemorrhaging in the ventricular system of the postnatal brain was present after collagenase delivery. This dose of collagenase (0.057 U/μl) was used for all of the experiments reported in this study. After delivering this dose of collagenase, the syringe remained in place for 2 min to prevent back-leakage before being withdrawn. Controls were injected with 1 μl of saline, using the same injection methods. Animals were placed on a warmed heating pad until active forelimb and hindlimb movement were regained upon gently touching pups. Pups were then returned to their nest with the mother following recovery from

Fig. 1. The induction of cerebellar hemorrhage with collagenase injection. (A) Bacterial collagenase was injected into the cerebral aqueduct (CA), which subsequently would flow into the fourth ventricle (4 V). (B) Dorsal and (C) ventral brain shows blood accumulation on the surface of the cerebellum (CB) 30 min after the induction of the hemorrhage. (D) Spectrophotometric hemoglobin assay results show significantly higher concentration of cyanomethylhemoglobin in CBH brains. Mean ± SEM and *P b 0.05. n = 6 for control and n = 7 for CBH.

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

anesthesia. Although the target number of mice for each set of phenotypes was set at 10, the actual number of surviving pups posthemorrhage induction dictated the final numbers of animals in each group for the various histological, behavioral, and molecular assays. Surviving pups from different litters were used to rule out litter specific effects. Spectrophotometric hemoglobin assay To quantify the extent of the hemorrhage in CBH mice, a spectrophotometric hemoglobin assay was performed using the methods outlined in Choudhri et al. (1997), which involved the use of Drabkin's Reagent (Sigma-Aldrich) and the procedures outlined in the manufacturer's instructions. Tissue collection and processing At varying survival times (6 h, 1, 3, 5, and 13 days) after the collagenase/saline injections, mice were perfusion fixed. Animals were randomly allocated to each survival time and an n = 4 to 10 per treatment group was collected at each time point. Upon completion of behavioral testing, mice were perfused at P75–78. To harvest the brain tissue, animals were overdosed with avertin [1.25% (w/v), 2,2,2tribomoethanol in tert-amyl alcohol] through intraperitoneal injection. Subsequently, mice were transcardially perfused with cold phosphate buffered saline (0.1 M PBS) and 4% paraformaldehyde (PFA). The brains were carefully dissected from the skull and post-fixed in 4% PFA solution overnight. Brains were then cryo-protected using a sucrose/PBS gradient, embedded in OCT (Sakura), and stored at −80 °C. Sagittal wholebrain sections were cut at 14 μm thickness in a cryostat and directly mounted on glass slides. Slides were stored at −20 °C.

87

Image collection and morphometric analysis The immunostained tissues were examined with a Zeiss fluorescence microscope and photomicrographs were taken with Axio Vision Rel. 4.6 software. Morphometric analysis of cresyl violet-stained medial cerebellar sections involved computer-assisted hand delineation of the cerebellum and cerebellar cell layers (ImageJ 1.46; National Institutes of Health). For all morphometric analyses, we used 4 non-contiguous medial (vermian) sagittal cerebellar sections per animal.

Cell quantification The density of granule cells in the inner granule layer (IGL) was measured from the four most medial sagittal cerebellar sections per animal. Each tissue section was separated by 140 μm. Using ImageJ software, a counting grid was superimposed on a 20× magnification image of CV stained IGL region. Each box in the counting grid was 0.05 mm2. The number of granule cells in 8 randomly chosen grids per section was quantified. The average number of cells per 0.05 mm2 per animal was calculated. The interneuron density in the molecular layer (ML) of adult cerebellum was also measured from four most medial sagittal sections per animal. The number of CV stained interneurons using four 20× magnification images taken from the ML of a cerebellar section was quantified. The area of the ML was measured using ImageJ to calculate cell density. Purkinje cell counting was done on four medial sagittal cerebellar sections per animal. Purkinje cells were labeled with anti-Calbindin prior to counting. Purkinje cell nuclei were identified with the aid of a bright field microscope at 40 × objective. Purkinje cells were counted from the whole section and data obtained from four sections were averaged to estimate the number of Purkinje cells per medial sagittal cerebellar section.

Cresyl Violet (CV) staining Behavioral evaluation Brain sections were rinsed in distilled water and immersed in 0.5% CV for 10–15 min. After rinsing in water to remove excess stain, slides were dehydrated and decolorized in 50 and 70% ethanol solution (3 min each). Slides were immersed in a differentiation solution (2 drops glacial acetic acid in 95% ethanol) for further decolorization when necessary. Dehydration was continued in 95% ethanol, and 3 changes of 100% ethanol (3 min each). Sections were then cleared in three consecutive baths of xylenes (3 min each) and glass cover slips were applied with Permount. Immunohistochemistry In order to label particular cell types in the cerebellum, we performed immunohistochemistry with cell-specific antibodies. The tissue sections were rinsed in 0.1 M PBS, and incubated with 0.3% hydrogen peroxide to quench endogenous tissue peroxidases. Sections were incubated in blocking solution, containing 30% bovine serum albumin (1:100, Sigma-Aldrich), normal goat serum (1:10, Bethyl Laboratories), and triton X-100 (1:100, Fisher Scientific) in PBS. After 30 min of blocking, the sections were incubated in primary antibody at 4 °C overnight. The sections were rinsed and incubated for 1 h with either biotinylated goat anti-rabbit or goat anti-mouse IgG (ABC Elite Kit, Vector Laboratories). Sections were next incubated with avidin and biotinylated horseradish peroxidase (Vector Laboratories) for 30 min at room temperature. Immunostaining was visualized with diaminobenzadine (DAB; Sigma-Aldrich). Slides were dehydrated using alcohol and xylenes and mounted in Permount (Sigma-Aldrich). The primary antibodies used in this study include rabbit anti-Myelin Basic Protein (MBP; 1:500, Abcam) for labeling the white matter and rabbit antiCalbindin (1:1000; Millipore) for labeling Purkinje cells.

At P60, motor function, locomotor activity, and anxiety-like behavior of collagenase-injected CBH and saline-injected control animals were assessed. Behavioral testing consisted of open field, rotarod, and horizontal ladder rung, consecutively. The data collected from males (Control n = 10; CBH n = 13) and females (Control n = 10; CBH n = 12) were analyzed separately. Open field General locomotor activity and anxiety were assessed by open field test (Gould et al., 2009) 60 days after the hemorrhagic insult. Mice were placed in the center of a 50 cm × 50 cm × 20 cm arena. The activity of each mouse was digitally recorded for 10 min and subsequently analyzed by Noldus Ethovision XT software (Noldus Information Technology). Total distance traveled, frequency of entering center, time spent in the center, and mean velocity were measured. The center of the arena was defined as a 20 cm × 20 cm central square.

Rotarod The Rotarod (Ugo-Basile) was used to assess motor coordination, balance, and motor learning (Jones and Roberts, 1968). The rotarod was programmed to accelerating (from 4 to 40 rpm over 5 min). The duration spent on the rotarod was recorded for each mouse across 3 consecutive days of testing. On day 1, the mice were given 4 trials (separated by a 30 min inter-trial interval), and on days 2 and 3, 2 trials were given. If a mouse fell off within 5 s from the beginning of the trial, that trial was not counted, and the mouse was given a new trial.

88

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

Horizontal ladder rung walking test The horizontal ladder rung walking test was conducted to assess skilled forelimb and hindlimb walking (Metz and Whishaw, 2002). The test apparatus consisted of clear Plexiglas side walls (1 m long and 20 cm high) and variably spaced metal rungs (3 mm diameter). The ladder was elevated 30 cm above the ground with the home cage at one end of the ladder. All animals crossed the ladder in the same direction. The level of difficulty was modified by varying the pattern of spacing between the rungs (1 to 3 cm). The pattern of spacing was varied each day of testing for 3 days. All animals were trained to habituate to the ladder and the height prior to testing. On the following day, mice underwent 5 trials, separated by a 5 min inter-trial interval, and videorecorded. The video-recordings were inspected using frame-by-frame analysis to score the number of steps and forelimb and hindlimb misplacements. Then, an error rate was calculated using the following formula: Error rate = number of slips/number of steps. The number of slips included forelimb and hindlimb misplacements. Tissue collection for RNA extraction A total of 56 ICR mice (P2, n = 14 per group; P15, n = 14 per group) were used to determine gene expression changes after CBH. The cerebellum was dissected and snap-frozen in liquid nitrogen and stored at −80 °C before RNA extraction was performed. Total RNA was isolated using RNeasy Mini Kit (Qiagen) according to the manufacturer's instructions. The RNA quantity and quality were measured using spectrophotometry and the samples were stored at −80 °C until use.

terminated at 85 °C for 5 min and RNase H was added and subsequently incubated at 37 °C for 30 min. Quantitative real-time PCR (qRT-PCR) was conducted to compare the expression levels of selected genes. Fast SYBR Green Master Mix (Applied Biosystems) was used to set up qRT-PCR reactions (10 μl total volume) containing cDNA as template. The reactions were carried out with the ABI 7500 Fast Real Time PCR System (Applied Biosystems). Amplicons were designed to amplify between 100 and 200 bp in length (see Table 1 for a list of primer sequences). All samples were tested in triplicate with the reference gene 18S to normalize the data to correct for variations in cDNA quantity. Statistical analysis All statistical analyses were performed using IBM SPSS Statistics 20 software (IBM SPSS Statistics). Data were expressed as mean ± 1 standard error of the mean (SEM). Data involving comparisons between more than two factors were analyzed with ANOVA and Tukey's post hoc test was performed if significant main effects were observed. Data were pooled in the case where no significant sex differences were observed to increase the power of the analyses. The Student's t test was used when comparing data involving only two factors. When two groups have unequal variances, the Welch correction was used. Repeated measures of analysis of variance (RM-ANOVA) was used to analyze data collected from the same cohort of animals over multiple trials or days. Results Postnatal CBH affects cerebellar size and granule cell density

Reverse transcriptase PCR and quantitative real-time PCR 1 μg of total RNA was used as a template to perform reverse transcription–polymerase chain reaction (RT–PCR) in a total reaction volume of 20 μl. Reverse transcription reactions were primed by 1 μl of oligo d(T)20 and incubated at 65 °C for 5 min. The reaction tubes were cooled on ice for 2 min, followed by the addition of SuperScript III reverse transcriptase (SuperScript III First-Strand Synthesis Systems; Invitrogen) and incubated at 50 °C for 50 min. The reaction was

Table 1 List of primer sequences for qRT-PCR. Pathway

Gene

Direction

Primer sequence

Apoptotic pathway

Caspase-3

Inflammatory pathway

TNFα

Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse Forward Reverse

TGACTGGAAAGCCGAAACTC AGCCTCCACCGGTATCTTCT ACGGCATGGATCTCAAAGAC GTGGGTGAGGAGCACGTAGT CCCAAGCAATACCCAAAGAA GCTTGTGCTCTGCTTGTGAG TGGCTGACACTTTTGAGCAC CAAAGGCACTTGACTGCTGA GTCTGACTTGGCCTCAAAGC GTGAGGTGGAGTGGGAGGTA CACTGATCGCTGGAGTGAGA ATTGCCTGATGGAAGTCACC TGGAT GCTCT TCAGT TCGTG CAACA CTCAT CCACA ATGCC TCCAAAGCTCACATCCACTG GGGACGTAAGTCCTTCACCA TGTTGTGGGAGGGAAGAAAC TGGCAGGGCTCTGACTAACT ACCATGCCTACCCAACTCAG CCTCAGCCTCAGTCTTGACC GCGGTAACCACTTTCACGAT AGTTGTGCTGCTGATGGATG GCCGCTAGAGGTGAAATTCTT CATTCTTGGCAAATGCTTTCG

IL-1β Neurotrophic factors

BDNF EPOR CNTF IGF1

Sonic hedgehog pathway

SHH GLI1 GLI2 NMYC

Internal control

18S

The induction of CBH in postnatal pups gave rise to a 35% mortality rate within the first 24 h after the injection, with most of the death occurring within the first six hours post-collagenase injection. Among the pups that survived the hemorrhagic insult, approximately 7.5% showed stunted growth. Gross examination of the brain 30 min after the injection revealed the presence of hematomas in the fourth ventricle and on the cerebellar surface (Figs. 1B–C). Hematoma size increased in the fourth ventricle region 1.5 h after the injection and smaller hematomas were consistently observed up to P7. To quantitatively assess the degree of hemorrhaging induced by collagenase administration, a spectrophotometric hemoglobin assay was performed on P2 brains collected six hours after administering collagenase into the ventricular system. We found a significantly higher cyanomethylhemoglobin concentration in the CBH group compared to the control group, which was administered saline (Fig. 1D). Moreover, in the saline injected control animals, we did not observe any death, stunted growth, or hematoma formation. To examine the effects of CBH on the development of the cerebellum, we measured the area of the medial cerebellum, 6 h (P2), 1 day (P3), 3 days (P5), 5 days (P7), 13 days (P15), and 76 days (P78) after collagenase injection. During postnatal cerebellar developmental time points from P2 to P15, we found a significant main effect of age (F4,70 = 909.9, P b 0.0001) and group (F1,70 = 23.55, P b 0.0001). Post hoc analyses revealed that at P7 and P15, the CBH group exhibited a significant decrease in the area of the cerebellum (P = 0.0001 and 0.00001, respectively) (Fig. 2A). Although a sex effect was observed with respect to the area of the cerebellum in adult mice (F1,17 = 5.364, P = 0.0333), there was no significant difference between the CBH and control groups in adult mice (Fig. 2B). We also examined whether CBH leads to structural alterations in specific cell layers of the developing cerebellum. The area of EGL, the cell layer which is in direct contact with the hematoma when CBH occurs, showed a significant main effect of age (F3,53 = 360.0, P b 0.0001) and an interaction between age and group (F3,53 = 6.601, P = 0.0007). Further post hoc analyses revealed a significant difference in EGL area at

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

89

Fig. 2. Postnatal CBH reduces the size of the cerebellum (CB), external granule layer (EGL), and molecular layer (ML) and decreases the density of granule cells in adult males. After CBH induction, the area of: (A) postnatal medial CB, (B) adult CB, (C) EGL, and (D) internal granular layer (IGL) was quantified. (E) Granule cell density, (F) ML area, (G) interneuron density, (H) Purkinje cell count, and (I) WM area were further quantified. Mean ± SEM and *P b 0.05. For the control groups, n = 4 at P2, n = 9 at P3, n = 8 at P5, n = 10 at P7, n = 9 at P15, n = 5 adult males, and n = 5 adult females. For the CBH groups, n = 5 at P2, n = 10 at P3, n = 5 at P5, n = 10 at P7, n = 10 at P15, n = 6 adult males, and n = 5 adult females. For histological measures without a significant main effect of sex, measurements from males and females were pooled.

postnatal day 7 (P = 0.0001) (Fig. 2C). As the cerebellum matures, the EGL diminishes and the internal granule layer (IGL) becomes more prominent. When we measured the area of the IGL at P15, we found no significant difference between the CBH and control groups (t16 = 0.4349, P = 0.6695) (Fig. 2D). In measuring the area of the IGL in the adult cerebellum, we found no main effect of sex (F1,17 = 2.664, P = 0.121) and no significant difference between CBH and control groups (F1,17 = 1.394, P = 0.254) (Fig. 2D). Given that there is no significant main effect of sex involving the area of the IGL, area measurements for males and females were pooled. When we further assessed the

density of granule cells in the adult IGL, we found a main effect of sex (F1,17 = 28.789, P = 0.000051) and post hoc analyses revealed a difference between CBH and control groups in adult male mice. (P = 0.01) (Fig. 2E). We then measured the area of the ML, where the inhibitory interneurons, parallel fibers and Purkinje cell dendrites reside. We found a significant reduction in the ML area at P15 in the CBH group (t12 = 3.421, P = 0.0051) (Fig. 2F). In the adult CBH mice, we found no significant main effect of sex in the area of the ML (F1,17 = 2.732, P = 0.117) and also no significant difference in ML area between the

90

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

CBH and control groups (F1,17 = 0.308, P = 0.586). Similarly, we found no significant main effect of sex in interneuron density (F1,17 = 1.667, P = 0.214) and no interneuron density difference between CBH and control groups (F1,17 = 2.325, P = 0.146) (Fig. 2G). When we counted

the number of Purkinje cells in the adult cerebellum, we found no significant main effect of sex (F1,17 = 2.145, P = 0.161) and no difference between CBH and control groups (F1,17 = 0.011, P = 0.919) (Fig. 2H). Given that there were no sex differences in ML area, interneuron

Fig. 3. Postnatal CBH leads to decreased anxiety and impaired motor function. The duration in the (A) center and (B) periphery of the open field was measured to assess the level of anxiety. (C) Rotarod performance in males and (D) in females on Day 1 was evaluated by comparing the duration on the rotarod between the control and CBH groups. Rotarod performance across three days for males (E) and females (F) was also evaluated. Horizontal ladder rung (G) was used to evaluate the fine motor skills. Mean ± SEM and *P b 0.05. n = 10 control males, n = 13 CBH males, n = 10 control females, and n = 12 CBH females.

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

91

density, and Purkinje cell numbers for males and females, data from both sexes were pooled into single analyses. Upon measuring the area of white matter (WM) at P15, we found no difference between the CBH and control groups (t15 = 1.147, P = 0.2693). Moreover, in adult mice, we found a main effect of sex in white matter area (F1,17 = 28.789, P = 0.000051), however no significant difference was observed between the CBH and control groups (F1,17 = 0.046, P = 0.833) (Fig. 2I). Postnatal CBH affects open field behavior and gives rise to motor deficits To investigate the effects of CBH on general movement, activity, and anxiety, we used the open field test. Anxiety was assessed by quantifying the amount of time mice spent in the center and periphery of the open field. We found a significant main effect of sex with respect to time spent in the center and periphery of the open field (Centre: F1,41 = 9.696, P = 0.003; Periphery: F1,41 = 3.113, P = 0.003), where post hoc analyses revealed that CBH males spent more time in the center of the open field (P = 0.01) and less time in the periphery (P = 0.05), compared to control males (Figs. 3A–B). Female mice with postnatal CBH did not show any difference in time spent in the center and periphery of the open field. We also observed that male mice, which experienced postnatal CBH, made more frequent attempts to jump on the walls of the open field and displayed stereotypic circling behavior. When we assessed the total distance traveled, frequency of entering the center, duration moving, and mean velocity in the open field, we did not find any significant effects of postnatal CBH (Suppl. Fig. 1). The effect postnatal CBH may have on motor coordination, balance, and motor learning was assessed using the rotarod test. There was a significant improvement in performance, where all mice demonstrated a greater ability to remain on the rotarod over the four trials conducted on the first day of training (F1,39 = 31.899, P = 0.000002) (Figs. 3C–D). A main effect of sex was also observed during the first day of training (F1,39 = 19.491, P = 0.000078), as well as main effect of group (F1,39 = 6.687, P = 0.014). Post hoc analyses revealed that there was a significant difference between the CBH and control groups during trial 2 and trial 1 in male and in female performance, respectively (Figs. 3C–D). When we further assessed rotarod performance across the three days of testing, we found a significant main effect of sex (F1,34 = 12.319, P = 0.001) and a significant main effect of group (F1,34 = 18.125, P = 0.00154) (Figs. 3E–F). Post hoc analyses revealed significant differences to be present on the first day of rotarod testing in both male and female mice in the CBH group compared to control (P = 0.001 for both males and females). (Figs. 3E–F). Given that postnatal CBH leads to a deficit in motor performance on the rotarod, we used the horizontal ladder rung walking test to more specifically assess limb coordination. We found no significant main effect of sex on horizontal ladder rung performance (F1,41 = 0.35, P = 0.853); however, we did find a significant main effect of group (F1,41 = 4.259, P = 0.045) (Fig. 3G). Post hoc analyses further revealed that there was a significant difference between the CBH and control group on day 2 of horizontal ladder rung testing (P = 0.01). (Fig. 3G). Postnatal CBH increases the expression of genes involved in injury and neuroprotection We examined the changes in specific gene transcript levels involved in the apoptotic, inflammatory, and neuroprotective pathways after postnatal CBH. To determine whether there was an inflammatory response arising from the hemorrhagic insult, we measured the relative quantity of IL-1β and TNFα. We found a significant upregulation of IL1β in the cerebellum at P2 (P = 0.015) (Fig. 4A) due to the hemorrhagic insult. At P15, a trend in elevated expression for both IL-1β and TNFα was displayed in the cerebellum of mice experiencing postnatal CBH (Figs. 4A–B). To assess apoptosis, Caspase-3 transcript expression was

Fig. 4. Postnatal CBH increases mRNA expression of the genes involved in the inflammatory and apoptotic pathways. We compared the expression of (A) IL-1β, (B) TNFα, and (C) Caspase-3 between CBH and control groups at P2 and P15. Mean ± SEM and *P b 0.05; n = 14 for all groups.

examined. Although there was a trend of increase in Caspase-3 expression at P2 and P15, the increase was not significant (P = 0.36 for P2; P = 0.25 for P15) (Fig. 4C). We then investigated the effects postnatal CBH has on the activation of the sonic hedgehog (SHH) pathway, which plays a critical role in cerebellar development and has been reported to be involved in neuroprotection (Amankulor et al., 2009). The SHH transcript expression showed approximately 1.8-fold increase in the cerebellum of P2 mice with CBH (P = 0.016) (Fig. 5A), whereas at P15 a trend of upregulated expression was determined (P = 0.062). The expression level of a mediator of the SHH pathway, GLI1, was found to be elevated at P2 and P15 in the cerebellum of mice with CBH (P = 0.031 and 0.0022, respectively) (Fig. 5B). We then examined another SHH pathway mediator, GLI2, and also found a significant increase in expression in the cerebellum at P2 and P15 in mice with CBH (P = 0.008 and 0.03, respectively) (Fig. 5C). NMYC, the downstream effector of the SHH pathway was also examined. We found a trend for increased NMYC transcript expression in the P2 cerebellum of CBH mice (P = 0.077)

92

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

Fig. 5. Postnatal CBH leads to an up-regulation of sonic hedgehog signaling. We compared the levels of transcript expression of genes involved in the sonic hedgehog pathway, including (A) SHH, (B) GLI1, (C) GLI2, and (D) NMYC between the control and CBH groups at P2 and P15. Mean ± SEM and * P b 0.05; n = 14 for all groups.

and a 2.5-fold increase at P15 in the cerebellum of mice with CBH (P = 0.025) (Fig. 5D). To further investigate the activation of potential endogenous protective mechanisms in the cerebellum of mice with postnatal CBH, we assessed the transcript expression of the following neurotrophic factors, IGF1, CNTF, BDNF, and EPOR, which are all known to play a neuroprotective role (Guan et al., 2003; Lin et al., 1998; Siren et al., 2001). We found a trend of increased transcript expression of the noted neurotrophic factors in the cerebellum of P2 mice with postnatal CBH (P = 0.26 for BDNF, P = 0.17 for EPOR, p = 0.12 for CNTF and P = 0.13 for IGF) (Suppl. Fig. 2). At P15, although EPOR and CNTF expression still showed an increased trend, BDNF and IGF transcript expression was no longer elevated (P = 0.14 for EPOR, P = 0.38, P = 0.51 for BDNF, for CNTF and P = 0.14 for IGF) (Suppl. Fig. 2). Discussion Preterm CBH is a serious neurological condition that can result in infant mortality or devastating neurodevelopmental outcomes (Limperopoulos et al., 2007; Volpe, 2009). The underlying molecular and cellular mechanisms of CBH have not been adequately explored due to the limitations of clinical studies (Scafidi et al., 2009) to address this issue and the lack of preclinical animal models. In our current study, we developed a mouse model of preterm CBH to examine the effects that postnatal hemorrhagic insult has on the development and function of the cerebellum, as well as to explore the gene expression changes associated with a postnatal ischemic insult in the cerebellum. In our animal model of preterm CBH, we intracranially delivered bacterial collagenase, which disrupts the basal lamina of blood vessels (Rosenberg et al., 1993). Subsequent to collagenase administration, a gradual hematoma is formed on the cerebellar surface and in the fourth ventricle, thus replicating the clinical presentation of primary CBH and secondary invasion of intraventricular blood, respectively (Donat et al., 1979; Volpe, 2009). While this collagenase injection paradigm successfully replicates bleeding in the cerebellar germinal matrix, it has some drawbacks, such as variability in the size and location of the hematoma,

which may result in phenotypic variability within the treatment group itself. Although it may be possible to employ other methods to induce a more uniform lesion resulting from a cerebellar hemorrhage, such as the injection of autologous blood directly into the cerebellum, this paradigm does not reproduce spontaneous bleeding from ruptured blood vessels, which is more aligned with what occurs during brain hemorrhages (Bullock et al., 1984; Yang et al., 1994). Further, injection of autologous blood seems to be more useful for studying the biochemical mechanisms that result from the presence of blood in the brain tissue. As mentioned in our Methods section, from our initial studies of determining which method to use to induce CBH, we also found that intracranial delivery of autologous blood into the same area where we inject collagenase did not produce hemorrhaging or a lesion. Therefore, we chose to use collagenase to induce CBH in our study. Ongoing studies in our laboratory currently involve administering collagenase directly into postnatal cerebellum to further investigate the effect of preterm hemorrhage on cerebellar development. Nevertheless, the variability within the collagenase-injected mice may, in fact, accurately reflect the differences in clinical outcomes among individual preterm infants with CBH (Volpe, 2008). The induction of CBH with the administration of collagenase in the postnatal mouse pup, led to decreased cerebellar area and decreased density of cerebellar granule cells. These results are in line with what has been previously reported by Limperopoulos et al. (2005), who showed that 37% of preterm infants with CBH exhibit cerebellar atrophy. It has been further demonstrated that blood deposition on the cerebellar surface and in the fourth ventricle (independent of CBH) is also significantly associated with cerebellar volume reduction (Messerschmidt et al., 2008; Steggerda et al., 2009). These findings suggest that blood deposition may be an important cause of damage to adjacent brain regions. Other studies have also shown that the presence of blood breakdown products result in neuronal loss through lipid peroxidation and the formation of free radicals (Aronowski and Zhao, 2011; Siesjo et al., 1989). Therefore, in our current study, it is possible that the cerebellar granule cells are particularly sensitive to the adverse effects of breakdown products from the blood. In particular, blood deposition on the cerebellar

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

surface, which is in direct contact with the germinal matrices of the cerebellum (the EGL and VZ), may lead to the decreased proliferation and/or survival of granule cell precursors. In addition to the potentially adverse affects of blood breakdown products, the anatomical abnormalities observed in our study may have also been induced from an indirect mechanism, such as the disruption of normal cerebellar circuitry due to the degeneration or death of cerebellar neurons (Uchino et al., 1993, 2006). Cerebellar interneurons receive excitatory inputs from granule cells via parallel fibers and although we did not see a significant decrease in cerebellar interneuron density, the CBH group did show a decreased trend. As such, the decrease in granule cell density observed in our study may lead to a decrease in the number of parallel fibers, which, in turn, may give rise to anterograde transneuronal degeneration of interneurons (Anderson and Flumerfelt, 1986). Furthermore, the reduction of parallel fiber numbers may result in abnormalities in Purkinje cell development, such as decreased Purkinje cell number and abnormal dendrite arborization and orientation (Maricich et al., 1997; Rakic and Sidman, 1973). Although we did not detect any significant reductions in Purkinje cell number in mice which experienced postnatal CBH, we again found a decreased trend for Pukinje cell numbers and a significant decrease in molecular layer area, where a diminished molecular layer is considered to be a reasonable indication for a decrease in the number of synapses between Purkinje cells and granule cells (Anderson, 1994; Greenough and Anderson, 1991). Therefore, postnatal CBH may lead to aberrant development of synaptic connections between Purkinje cells and parallel fibers from granule neurons. It has been previously reported that hemorrhagic injury of the developing cerebellum is significantly associated with deficits in gross and fine motor skills and higher rates of cerebral palsy (Bodensteiner and Johnsen, 2005; Johnsen et al., 2005; Limperopoulos et al., 2007). Moreover, a recent report from Zayek et al. (2011) has shown that preterm infants with CBH in the medial (vermian) part of the cerebellum have increased risk for motor disabilities. Thus, the reduction of medial cerebellum size observed in our study may be directly associated with the motor impairments observed in mice with postnatal CBH. The cerebellum is increasingly recognized as an important structure not only for motor function but also for cognitive and affective processing (Schmahmann, 2001). In our study, we found that mice with postnatal CBH spent more time in the center of the open field. This deviation in thigmotactic behavior is considered to be maladaptive and may be associated with decreased anxiety and an increased propensity of risk-taking behaviors. Interestingly, a subset of male mice, which experienced postnatal CBH, displayed abnormal behavioral phenotypes, such as frequent jumping in the open field and stereotypic circling. Although further behavioral characterization is necessary to determine the cognitive and affective behavioral deficits resulting from postnatal CBH in mice, these observed phenotypes in our study are in accordance with clinical research findings which show that children with isolated CBH are more likely to develop neurodevelopmental disorders and display behavioral problems, such as autism spectrum disorder (ASD) and attention deficit hyperactivity disorders (ADHD) (Limperopoulos et al., 2007). When considering the anatomical and behavioral phenotypes in our mouse model of preterm CBH, our study suggests that males are more susceptible to the adverse effects of CBH. Such a sex-bias for adverse outcome due to preterm CBH has previously been reported in both the human population and animal models of brain injury (Kent et al., 2011; Limperopoulos et al., 2005, 2012; Mayoral et al., 2009; Peacock et al., 2012; van Kooij et al., 2011). These sex differences observed in our study may be modulated by the sex-specific hormones testosterone and estrogen. The elevated levels of testosterone during gestation in humans and during the perinatal period in rodents have been suggested to be associated with enhanced neuronal excitotoxicity and increased long-term adverse outcomes in males (Hines, 2008; Yang et al., 2002). In contrast, estrogen has been shown to have protective effects in adult stroke models (McCullough and Hurn, 2003). Although only

93

minimal estrogen is circulating during perinatal period, the elevated production of estrogen later in development may have some beneficial effects on neural repair and plasticity after injury (Hill and Fitch, 2012). Given that the anatomical and behavioral phenotypes characterized in our mouse model of preterm CBH are similar to what is observed in the human population of preterm infants with CBH, we sought to investigate the developmental genetic mechanisms that may be perturbed due to postnatal CBH. We found increased Il-1β transcript expression due to postnatal CBH and evidence for an increased trend of activation for additional genes involved in the apoptotic and inflammatory pathways. The increase in the expression of genes involved in apoptotic pathway and the mediators of inflammatory pathway have been previously reported in both clinical and preclinical studies for other types of brain injury (Brunswick et al., 2012; Castillo et al., 2002; Xi et al., 2006). The activation of the apoptotic and inflammatory pathway can lead to tissue destruction and secondary injury to the brain (Cahill et al., 2006; Nathan, 2002). In addition to the activation of apoptotic and inflammatory responses after hemorrhagic insults, it has been previously demonstrated that such an ischemic event can initiate a cascade of signaling events in neurons and glia to repair the damage and to protect from further damage. In our study, we found that components of the SHH pathways are also elevated in the cerebellum after postnatal CBH. SHH signaling is known to play an essential role in granule cell proliferation in the developing cerebellum (Wechsler-Reya and Scott, 1999) and has also been reported to play a neuroprotective role after CNS injury (Amankulor et al., 2009; Bambakidis et al., 2003). The increased levels of SHH signaling in the cerebellum after a hemorrhagic insult may arise from the loss of granule neurons or reduction of granule cell proliferation. Therefore, increased SHH signaling may promote the proliferation of granule cells to compensate for this loss (Knoepfler et al., 2002). Such a notion is supported by a previous study which demonstrated that NMYC expression in granule cell neuroblasts enhances SHH signaling to promote granule cell proliferation (Kenney et al., 2003). In our study, we report an increase in NMYC expression in the cerebellum after CBH. SHH has also been demonstrated to stimulate oligodendrocyte precursor cell proliferation in the cerebellum (Bouslama-Oueghlani et al., 2012), thus potentially acting as an endogenous protective mechanism for white matter injury after hemorrhage and possibly providing a rationale as to why we see no change in cerebellar white matter area between CBH and control groups. To further investigate the neuroprotective mechanisms involved in postnatal CBH, we also examined the transcriptional expression changes of neurotrophic factors previously shown to be up-regulated in response to brain injury and to play a role in neuroprotection (Guan et al., 2003; Hicks et al., 1999; Lin et al., 1998; Siren et al., 2001). We showed a trend of up-regulation of BDNF, EPOR, CNTF, and IGF1 transcripts in the cerebellum of postnatal mice with CBH. IGF1 has been shown to play an essential role in promoting granule cell proliferation and anti-apoptotic activity (Chrysis et al., 2001; Ye et al., 1996). CNTF may be involved in stimulating axonal growth (Muler et al., 2007) and reducing glutamate excitotoxicity (Beurrier et al., 2009). In conclusion, we have developed a mouse model for preterm CBH that mimics the cerebellar structural and functional abnormalities displayed by human preterm neonates with CBH. With our mouse model, we also demonstrated that postnatal CBH activates various signaling pathways involved in injury and neuroprotective responses. Therefore, our mouse model of preterm CBH may be useful for understanding and developing potential preventive and treatment regimes. Supplementary data to this article can be found online at http://dx. doi.org/10.1016/j.expneurol.2013.11.010. Acknowledgments This project was funded by the Canadian Institutes of Health Research (CIHR) and the Canadian Institute for Advanced Research (CIFAR).

94

J.Y.J. Yoo et al. / Experimental Neurology 252 (2014) 85–94

References Amankulor, N.M., Hambardzumyan, D., Pyonteck, S.M., Becher, O.J., Joyce, J.A., Holland, E.C., 2009. Sonic hedgehog pathway activation is induced by acute brain injury and regulated by injury-related inflammation. J. Neurosci. 29, 10299–10308. Anderson, B., 1994. The volume of the cerebellar molecular layer predicts attention to novelty in rats. Brain Res. 641, 160–162. Anderson, W.A., Flumerfelt, B.A., 1986. Long-term effects of parallel fiber loss in the cerebellar cortex of the adult and weanling rat. Brain Res. 383, 245–261. Aronowski, J., Zhao, X., 2011. Molecular pathophysiology of cerebral hemorrhage: secondary brain injury. Stroke 42, 1781–1786. Bambakidis, N.C., Wang, R.-Z., Franic, L., Miller, R.H., 2003. Sonic hedgehog-induced neural precursor proliferation after adult rodent spinal cord injury. J. Neurosurg. Spine 99, 70–75. Beurrier, C., Faideau, M., Bennouar, K.-E., Escartin, C., Kerkerian-Le Goff, L., Bonvento, G., Gubellini, P., 2009. Ciliary neurotrophic factor protects striatal neurons against excitotoxicity by enhancing glial glutamate uptake. PLoS One 5, e8550. Biran, V., Verney, C., Ferriero, D.M., 2012. Perinatal cerebellar injury in human and animal models. Neurol. Res. Int. 2012, 1–9. Blencowe, H., Cousens, S., Oestergaard, M.Z., Chou, D., Moller, A.-B., Narwal, R., Adler, A., Vera Garcia, C., Rohde, S., Say, L., Lawn, J.E., 2012. National, regional, and worldwide estimates of preterm birth rates in the year 2010 with time trends since 1990 for selected countries: a systematic analysis and implications. Lancet 379, 2162–2172. Bodensteiner, J.B., Johnsen, S.D., 2005. Cerebellar injury in the extremely premature infant: newly recognized but relatively common outcome. J. Child Neurol. 20, 139–141. Bouslama-Oueghlani, L., Wehrle, R., Doulazmi, M., Chen, X.R., Jaudon, F., LemaigreDubreuil, Y., Rivals, I., Sotelo, C., Dusart, I., 2012. Purkinje cell maturation participates in the control of oligodendrocyte differentiation: role of sonic hedgehog and vitronectin. PLoS One 7, e49015. Brunswick, A.S., Hwang, B.Y., Appelboom, G., Hwang, R.Y., Piazza, M.A., Connolly Jr., E.S., 2012. Serum biomarkers of spontaneous intracerebral hemorrhage induced secondary brain injury. J. Neurol. Sci. 321, 1–10. Bullock, R., Mendelow, A.D., Teasdale, G.M., Graham, D.I., 1984. Intracranial haemorrhage induced at arterial pressure in the rat. Part 1: description of technique, icp changes and neuropathological findings. Neurol. Res. 6, 184–188. Cahill, W.J., Calvert, J.H., Zhang, J.H., 2006. Mechanisms of early brain injury after subarachnoid hemorrhage. J. Cereb. Blood Flow Metab. 26, 1341–1353. Castillo, J., Davalos, A., Alvarez-Sabin, J., Pumar, J.M., Leira, R., Silva, Y., Montaner, J., Kase, C.S., 2002. Molecular signatures of brain injury after intracerebral hemorrhage. Neurology 58, 624–629. Choudhri, T.F., Hoh, B.L., Solomon, R.A., Connolly, E.S., Pinsky, D.J., 1997. Use of a spectrophotometric hemoglobin assay to objectively quantify intracerebral hemorrhage in mice. Stroke 28, 2296–2302. Chrysis, D., Calikoglu, A.S., Ye, P., D'Ercole, A.J., 2001. Insulin-like growth factor-i overexpression attenuates cerebellar apoptosis by altering the expression of bcl family proteins in a developmentally specific manner. J. Neurosci. 21, 1481–1489. Donat, J.F., Okazaki, H., Kleinberg, F., 1979. Cerebellar hemorrhages in newborn infants. Am. J. Dis. Child. 133, 441. Dyet, L.E., Kennea, N., Counsell, S.J., Maalouf, E.F., Ajayi-Obe, M., Duggan, P.J., Harrison, M., Allsop, J.M., Hajnal, J., Herlihy, A.H., Edwards, B., Laroche, S., Cowan, F.M., Rutherford, M.A., Edwards, A.D., 2006. Natural history of brain lesions in extremely preterm infants studied with serial magnetic resonance imaging from birth and neurodevelopmental assessment. Pediatrics 118, 536–548. Gould, T.D., Dao, D.T., Kovacsics, C.E., 2009. The Open Field Test, Mood and Anxiety Related Phenotyes in Mice: Characterization Using Behavioral Tests. 1–20. Greenough, W.T., Anderson, B.J., 1991. Cerebellar synaptic plasticity relation to learning versus neural activity. Ann. N. Y. Acad. Sci. 627, 231–247. Guan, J., Bennet, L., Gluckman, P.D., Gunn, A.J., 2003. Insulin-like growth factor-1 and postischemic brain injury. Prog. Neurobiol. 70, 443–462. Hicks, R.R., Martin, V.B., Zhang, L., Seroogy, K.B., 1999. Mild experimental brain injury differentially alters the expression of neurotrophin and neurotrophin receptor MRNAS in the hippocampus. Exp. Neurol. 160, 469–478. Hill, C.A., Fitch, R.H., 2012. Sex differences in mechanisms and outcome of neonatal hypoxia-ischemia in rodent models: implications for sex-specific neuroprotection in clinical neonatal practice. Neurol. Res. Int. 2012, 1–9. Hines, M., 2008. Early androgen influences on human neural and behavioural development. Early Hum. Dev. 84, 805–807. Johnsen, S.D., Bodensteiner, J.B., Lotze, T.E., 2005. Frequency and nature of cerebellar injury in the extremely premature survivor with cerebral palsy. J. Child Neurol. 20, 60–64. Jones, B.J., Roberts, D.J., 1968. The quantitative measurement of motor inco-ordination in naive mice using an accelerating rotarod. J. Pharm. Pharmacol. 20, 302–304. Kenney, A.M., Cole, M.D., Rowitch, D.H., 2003. Nmyc upregulation by sonic hedgehog signaling promotes proliferation in developing cerebellar granule neuron precursors. Development 130, 15–28. Kent, A.L., Wright, I.M.R., Abdel-Latif, M.E., the New South Wales and Australian Capital Territory Neonatal Intensive Care Units Audit, G., 2011. Mortality and adverse neurologic outcomes are greater in preterm male infants. Pediatrics 129, 124–131. Knoepfler, P.S., Cheng, P.F., Eisenman, R.N., 2002. N-myc is essential during neurogenesis for the rapid expansion of progenitor cell populations and the inhibition of neuronal differentiation. Gene Dev. 16, 2699–2712. Limperopoulos, C., Benson, C.B., Bassan, H., Disalvo, D.N., Kinnamon, D.D., Moore, M., Ringer, S.A., Volpe, J.J., du Plessis, A.J., 2005. Cerebellar hemorrhage in the preterm infant: ultrasonographic findings and risk factors. Pediatrics 116, 717–724. Limperopoulos, C., Bassan, H., Gauvreau, K., Robertson, R.L., Sullivan, N.R., Benson, C.B., Avery, L., Stewart, J., Md, J.S.S., Ringer, S.A., Volpe, J.J., duPlessis, A.J., 2007. Does

cerebellar injury in premature infants contribute to the high prevalence of longterm cognitive, learning, and behavioral disability in survivors? Pediatrics 120, 584–593. Limperopoulos, C., Chilingaryan, G., Guizard, N., Robertson, R.L., Du Plessis, A.J., 2010. Cerebellar injury in the premature infant is associated with impaired growth of specific cerebral regions. Pediatr. Res. 68, 145–150. Limperopoulos, C., Chilingaryan, G., Sullivan, N., Guizard, N., Robertson, R.L., du Plessis, A.J., 2012. Injury to the premature cerebellum: outcome is related to remote cortical development. Cereb. Cortex 1–9. Lin, T.N., Wang, P.Y., Chi, S.I., Kuo, J.S., 1998. Differential regulation of ciliary neurotrophic factor (cntf) and cntf receptor alpha (cntfralpha) expression following focal cerebral ischemia. Mol. Brain Res. 55, 71–80. Maricich, S.M., Soha, J., Trenkner, E., Herrup, K., 1997. Failed cell migration and death of purkinje cells and deep nuclear neurons in the weaver cerebellum. J. Neurosci. 17, 3675–3683. Mayoral, S.R., Omar, G., Penn, A.A., 2009. Sex differences in a hypoxia model of preterm brain damage. Pediatr. Res. 66, 248–253. McCullough, L.D., Hurn, P.D., 2003. Estrogen and ischemic neuroprotection: an integrated view. Trends Endocrinol. Metab. 14, 228–235. Messerschmidt, A., Brugger, P.C., Boltshauser, E., Zoder, G., Sterniste, W., Birnbacher, R., Prayer, D., 2005. Disruption of cerebellar development: potential complication of extreme prematurity. Am. J. Neuroradiol. 26, 1659–1667. Messerschmidt, A., Prayer, D., Brugger, P.C., Boltshauser, E., Zoder, G., Sterniste, W., Pollak, A., Weber, M., Birnbacher, R., 2008. Preterm birth and disruptive cerebellar development: assessment of perinatal risk factors. Eur. J. Paediatr. Neurol. 12, 455–460. Metz, G.A., Whishaw, I.Q., 2002. Cortical and subcortical lesions impair skilled walking in the ladder rung walking test: a new task to evaluate fore- and hindlimb stepping, placing, and co-ordination. J. Neurosci. Methods 115, 169–179. Muler, A., Hauk, T.G., Fischer, D., 2007. Astrocyte-derived cntf switches mature rgcs to a regenerative state following inflammatory stimulation. Brain 130, 3308–3320. Nathan, C., 2002. Points of control in inflammation. Nature 420, 846–852. Peacock, J.L., Marston, L., Marlow, N., Calvert, S.A., Greenough, A., 2012. Neonatal and infant outcome in boys and girls born very prematurely. Pediatr. Res. 71, 305–310. Rakic, P., Sidman, R.L., 1973. Organization of cerebellar cortex secondary to deficit of granule cells in weaver mutant mice. J. Comp. Neurol. 152, 133–161. Rosenberg, G.A., Estrada, E., Kelley, R.O., Kornfeld, M., 1993. Bacterial collagenase disrupts extracellular matrix and opens blood–brain barrier in rat. Neurosci. Lett. 160, 117–119. Scafidi, J., Fagel, D.M., Ment, L.R., Vaccarino, F.M., 2009. Modeling premature brain injury and recovery. Int. J. Dev. Neurosci. 27, 863–871. Schmahmann, J.D., 2001. The cerebrocerebellar system: anatomic substrates of the cerebellar contribution to cognition and emotion. Int. Rev. Psychiatry 13, 247–260. Siesjo, B.K., Agardh, C.D., Bengtsson, F., 1989. Free radicals and brain damage. Cerebrovasc. Brain Metab. Rev. 1, 165–211. Siren, A.-L., Knerlich, F., Poser, W., Gleiter, C.H., Bruck, W., Ehrenreich, H., 2001. Erythropoietin and erythropoietin receptor in human ischemic/hypoxic brain. Acta Neuropathol. 101, 271–276. Steggerda, S.J., Leijser, L.M., Wiggers-de Bruine, F.T., van der Grond, J., Walther, F.J., van Wezel-Meijler, G., 2009. Cerebellar injury in preterm infants: incidence and findings on US and MR images1. Radiology 252, 190–199. Tam, E.W.Y., Rosenbluth, G., Rogers, E.E., Ferriero, D.M., Glidden, D., Goldstein, R.B., Glass, H.C., Piecuch, R.E., Barkovich, A.J., 2011. Cerebellar hemorrhage on magnetic resonance imaging in preterm newborns associated with abnormal neurologic outcome. J. Pediatr. 158, 245–250. Uchino, A., Hasuo, K., Uchida, K., Matsumoto, S., Tsukamoto, Y., Ohno, M., Masuda, K., 1993. Olivary degeneration after cerebellar or brain stem haemorrhage: MRI. Neuroradiology 35, 335–338. Uchino, A., Takase, Y., Nomiyama, K., Egashira, R., Kudo, S., 2006. Brainstem and cerebellar changes after cerebrovascular accidents: magnetic resonance imaging. Eur. Radiol. 16, 592–597. van Kooij, B.J.M., van Pul, C., Benders, M.J.N.L., van Haastert, I.C., de Vries, L.S., Groenendaal, F., 2011. Fiber tracking at term displays gender differences regarding cognitive and motor outcome at 2 years of age in preterm infants. Pediatr. Res. 70, 626–632. Volpe, J.J., 2008. Neurology of the Newborn. Saunder Elsevier, Philadelphia PA. Volpe, J.J., 2009. Cerebellum of the premature infant: Rapidly developing, vulnerable, clinically important. J. Child Neurol. 24, 1085–1104. Wechsler-Reya, R.J., Scott, M.P., 1999. Control of neuronal precursor proliferation in the cerebellum by sonic hedgehog. Neuron 22, 103–114. Wood, N.S., Costeloe, K., Gibson, A.T., Hennessy, E.M., Marlow, N., Wilkinson, A.R., 2005. The epicure study: associations and antecedents of neurological and developmental disability at 30 months of age following extremely preterm birth. Arch. Dis. Child. Fetal Neonatal Ed. 90, F134–F140. Xi, G., Keep, R.F., Hoff, J.T., 2006. Mechanisms of brain injury after intracerebral haemorrhage. Lancet Neurol. 5, 53–63. Yang, G.-Y., Betz, A.L., Chenevert, T.L., Brunberg, J.A., Hoff, J.T., 1994. Experimental intracerebral hemorrhage: relationship between brain edema, blood flow, and blood–brain barrier permeability in rats. J. Neurosurg. 81, 93–102. Yang, S.-H., Perez, E., Cutright, J., Liu, R., He, Z., Day, A.L., Simpkins, J.W., 2002. Testosterone increases neurotoxicity of glutamate in vitro and ischemia-reperfusion injury in an animal model. J. Appl. Physiol. 92, 195–201. Ye, P., Xing, Y., Dai, Z., D'Ercole, A.J., 1996. In vivo actions of insulin-like growth factor-i (IGF-I) on cerebellum development in transgenic mice: evidence that IGF-I increases proliferation of granule cell progenitors. Dev. Brain Res. 95, 44–54. Zayek, M.M., Benjamin, J.T., Maertens, P., Trimm, R.F., Lal, C.V., Eyal, F.G., 2011. Cerebellar hemorrhage: a major morbidity in extremely preterm infants. J. Perinatol. 32, 699–704.