Ventricular size mapping in a transgenic model of schizophrenia

www.elsevier.com/locate/devbrainres

Developmental Brain Resea

Research report

Ventricular size mapping in a transgenic model of schizophrenia

German Torresa, Beth A. Meederb, Brian H. Hallasa, Joseph A. Spernyakc, Richard Mazurchukc,

Craig Jonesd, Kenneth W. Grossd, Judith M. Horowitzb,d,*

aDepartment of Neuroscience, New York College of Osteopathic Medicine of New York Institute of Technology, Old Westbury, New York 11568, USAbClinical Neuroscience Laboratory, Medaille College, Buffalo, New York 14214, USA

cPreclinical MR Imaging Facility, Roswell Park Cancer Institute, Buffalo, New York 14263, USAdDepartment of Molecular and Cellular Biology, Roswell Park Cancer Institute, Buffalo, New York 14263, USA

Accepted 3 August 2004

Available online 27 October 2004

Abstract

Genetically engineered mice have been generated to model a variety of neurological disorders. The chakragati (ckr) mouse is beginning

to provide valuable insights into the structural brain changes underlying certain manifestations of schizophrenia. For instance, these mice

show enlargement of the lateral ventricles, an abnormality frequently reported as a structural aberration in the schizophrenic brain. As neither

the anatomical pattern nor the timing of this ventricular enlargement is known, we used magnetic resonance imaging (MRI) techniques to

non-invasively visualize the development of the ventricular system in 5-, 10- and 30-day-old ckr pups. High-resolution MR images obtained

from these mutants showed a progressive enlargement of the lateral ventricles, starting at day 5 of postnatal life. These emerging deficits were

associated with abnormalities in mid-saggital corpus callosum area and thickness, particularly in 30-day-old adolescent animals. At this time

of development, aberrant behaviors that mimic certain symptoms of schizophrenia also appeared in ckr mice suggesting that structural

changes in ventricular size predates the onset of psychotic-like behaviors. These results are viewed as further indication that pre- and peri-

natal disturbances of the ventricular system and adjacent neural regions may be important pathogenic factors in schizophrenia. Application of

MRI to the ckr mouse is relatively new but has great potential for clarifying the relationship between brain structure changes and genetically

induced vulnerabilities to psychoses.

D 2004 Elsevier B.V. All rights reserved.

Theme: Disorders of the nervous system

Topic: Developmental disorders

Keywords: ckr mouse; Magnetic resonance imaging; Corpus callosum morphology; Circling behavior; Trans-genes; Heterozygote kin

1. Introduction

Schizophrenia is a complex brain disease characterized

by early developmental defects followed by progressive

clinical symptoms [19]. In this regard, a number of

examinations with in vivo imaging techniques indicate

that schizophrenia is associated with structural changes in

the brain parenchyma. By far the most prevalent and

0165-3806/$ - see front matter D 2004 Elsevier B.V. All rights reserved.

doi:10.1016/j.devbrainres.2004.08.011

* Corresponding author. Clinical Neuroscience Laboratory, Medaille

College, 18 Agassiz Circle, Buffalo, New York 14214, USA. Tel.: +1 716

884 3411x229; fax: +1 716 884 0291.

E-mail address: [email protected] (J.M. Horowitz).

consistent regional change is an increased ventricular-brain

ratio, namely an expansion of the lateral ventricles [10,12].

Indeed, structural brain imaging studies suggest that there

is a strong genetic linkage between enlargement of the

lateral ventricles and schizophrenic psychoses [28]. It is

therefore conceivable that when ventricles become

enlarged, among other defects, selective neurons may be

misplaced and neural circuits go awry ultimately producing

severe cognitive and motor deficits [21]. Given the high

incidence of lateral ventricular enlargement in first episode

schizophrenic patients, this defect may represent a rela-

tively simple endophenotype in imaging studies for

understanding one of schizophrenia’s most tractable

rch 154 (2005) 35–44

G. Torres et al. / Developmental Brain Research 154 (2005) 35–4436

signatures. However, it is not clear whether the above

endophenotype predates the onset of the disease or

whether it is affected by anti-psychotic drug treatment

[2]. To differentiate between these two possibilities,

genetically engineered mice might provide specific insights

into the developmental patterns and the timing of such

structural changes. In this particular case, enlargement of

the lateral ventricles is an obvious endophenotype that can

be tested in animals that have been generated and model

certain features of schizophrenia. For instance, studies of

homozygous ckr mice provide evidence for increased

ventricular size; in addition these mice also display

hyperactivity (i.e., increased motor activity) and aberrant

circling behavior [9,25]. Importantly, reversal of the

aforementioned psychotic behaviors can be achieved by

clinically effective neuroleptic drugs [32]. Further, hetero-

zygous mice with a single transgene insertion (i.e., Ren-2d

renin gene) also show enlargement of the lateral ventricles

without the abnormal circling behavior observed in their

homozygous kin. This finding is of significant interest as

relatives of schizophrenics who do not express the clinical

symptoms, nevertheless have larger ventricles than indi-

viduals from families without the brain disease [18,27].

Thus, heterozygous and homozygous ckr mice are ideal

animal models for deconstructing specific features of

schizophrenia: they both contain a variable and heritable

trait (i.e., lateral ventricular enlargement) but only one

mouse genotype displays the full-blown aberrant circling

behavior. This polarity could potentially recapitulate more

precisely the structural changes and psychotic behaviors

observed in schizophrenics and their first-degree relatives.

The first hypothesis of this study was to use high-

resolution magnetic resonance imaging (MRI scans) to

determine the onset of lateral ventricular enlargement in

ckr mouse pups as neither the anatomical pattern nor the

timing of this developmental event have been established.

This initial hypothesis has merit because it might provide

insights not only into the spatio-temporal mapping of

lateral ventricular size area in vivo, but also the temporal

profile of when specific structural changes in schizophre-

nia might emerge. Second, we wished to determine

whether lateral ventricular enlargement in both hetero-

zygous and homozygous ckr mice was associated with

corpus callosum abnormalities. In this regard, there is

evidence that ventricular enlargements influence shape and

displacement of the callosum in first episode schizophrenic

patients [13,24]. Third, we wanted to confirm and extend

earlier findings of hyperactivity and aberrant circling

behavior in homozygous ckr mice, an effect with early

adolescent onset [9]. Finally, we wished to determine if

high-resolution MRI volumes in the lateral ventricles could

be detected from mouse brains of different ages. In

general, our findings suggest that the ckr mutant might

be relevant for understanding the timing, rates and

structural changes thought to occur in the schizophrenic

brain.

2. Methods

2.1. Animals

Mice at three developmental ages were used for the

studies described herein: 5-day-old pups (n=2–5/genotype/

sex; 3.5–4.5 g), 10-day-old pups (n=2–5/genotype/sex; 7.0–

9.0g) and 30-day-old pups (n=2–5/genotype/sex; 25.0–38.0

g). Wild-type (C57BL/10Rospd�C3H/HeRos), heterozy-

gous and homozygous (5- and 10-day-old) pups were

maintained with their respective dams and littermates until

behavioral testing and imaging procedures were performed.

Adolescent (30-day-old) mice were kept in (same-sex/same-

genotype) groups of 3–4/cage and maintained on a light:-

dark cycle of 12:12 h (lights on at 07:00) with free access to

food and water. Mice were never handled or isolated prior

to MRI scans or open-field testing.

The ckr mouse was serendipitously generated by micro-

injection of a 24-kb genomic fragment containing the mouse

Ren-2d renin gene into BCF (C57BL/10Rospd�C3H/

HeRos) fertilized oocytes [25]. Genetic and physical

analysis of this insertion revealed that 2.5 copies of the

transgene, comprising 65–70-kb, had integrated, duplicated

and inverted portions of a particular locus within chromo-

some 16 of the mouse genome. All transgenic mice used for

the studies below were male and female F2 animals of the

mixed genetic background of BCF1 (C57BL/10Rospd�C3H/HeRos). Classification of genotype for both hetero-

zygous and ckr mice was conducted by restriction fragment-

length polymorphism analysis of biopsied tail DNA taken

during the first week of postnatal life [25]. All behavioral

and anatomical procedures were carried out in accordance

with the NIH Guide for the Care and Use of Laboratory

Animals, and with approval from the Roswell Park Cancer

Institute IACUC. All efforts were made to minimize animal

stress and to reduce the number of mice used for the

imaging and behavioral studies.

2.2. Southern blotting

Mouse pup genotypes were determined by Southern

analysis as described previously [29]. Briefly, 5–10 Ag of

genomic DNA was restriction digested with Bgl II and

separated onto 0.8% agarose gels in 1X TAE buffer. Gels

were stained with ethidium bromide and UV irradiated to

break large DNA fragments prior to transfer. Gels were

blotted to nylon membranes (Zeta-Probe GT, Bio-Rad) by

capillary transfer according to the manufacturer’s instruc-

tions. Blots were then hybridized with probe AR6 according

to the method of Church and Gilbert [7] and as modified by

the membrane manufacturer.

2.3. Ventricular area measurements

Adolescent male and female (30-day-old) mice were

sacrificed with CO2, decapitated and brains collected in ice-

Fig. 1. The magnetic resonance imaging concept used in this study. (A) A

General Electric CSI 4.7T/33 cm bore magnet was used to create a multi-

slice two-dimensional model of the mouse brain. The magnet used in this

particular MRI was 4.7 Tesla. (B) The anesthetized mouse (foreground) was

placed in a sliding platform that subsequently migrated to the bore of the

scanner for the duration of the imaging process (~60 min). A series of axial

scans were then generated as average T2-weighted structural images of the

nascent ventricular system.

G. Torres et al. / Developmental Brain Research 154 (2005) 35–44 37

cold 4% paraformaldehyde for five days. Brains were then

placed in 20% sucrose in 0.1 M sodium phosphate buffer for

2 days or until the brains sank. Frozen coronal sections were

cut on a sliding microtome at 40 Am and processed for

Neutral Red histochemical staining. Determination of

ventricular size was accomplished as follows: coronal brain

sections from wild-type, heterozygous or homozygous ckr

mice were mounted on gel-coated-slides and the entire

ventricular system was photographed at 4� with a micro-

scope-mounted digital camera and scanned into Adobe

Photoshop. Ventricular size was then drawn in Camera

lucida, and estimates of ventricular space area were

generated using the NIH IMAGE software package. All

quantifications of ventricular size were performed without

knowledge of animal genotype. Statistical significance was

defined as PV0.05 using one-way ANOVAs and Tukey

post-hoc test comparisons for group significance.

2.4. MR imaging procedures

To assess cerebral ventricular size in vivo, wild-type,

heterozygous and homozygous ckr mice were imaged on

postnatal days 5, 10 and 30. Prior to scanning, mice were

anesthetized with 4% isoflurane and general anesthesia was

maintained during the scanning procedure via an inlet tube

placed in front of the nose. A small vacuum applied to a

second tube served to remove carbon dioxide and excess

anesthetic. Under this anesthetic plane, mice were placed

within a 3-mm diameter, butyrate plastic tube, and the head

was immobilized by applying slight pressure with medical

tape over the top of the skull cushioned with a small, foam

pillow. In order to maintain constant core body temperature

during the scans, a small heating pad (37 8C) was present

underneath the mice. To improve the signal to noise ratio

and to obtain better scan readouts in the 5-day-old pups, a

contrast agent (Magnevist, Berlex) was injected intraper-

itoneally at a concentration of 0.3 mmol/kg 15 min prior to

the imaging procedures.

High-resolution MR imaging scans were acquired using

a General Electric (GE) CSI 4.7T/33 cm horizontal bore

magnet (GE NMR Instruments, Fremont, CA) with

upgraded RF and computer systems incorporating AVANCE

digital electronics (Bruker BioSpec platform with Para-

VisionR Version 2.1 Operating System, Bruker Medical,

Billerica, MA). MR data were acquired using a G060

removable gradient coil insert generating maximum field

strength of 950 mT/m and a custom designed 35 mm RF

transceiver coil. Standard spin echo (SE) and rapid

acquisition with relaxation enhancement (RARE) MR

imaging pulse sequences were used to acquire multi-slice

volume images. A series of preliminary pilot scans were

acquired to obtain positional geometry of the mouse brain.

For anatomical detail, high-resolution T2-weighted RARE

coronal and axial MR scans encompassing the entire brain

were acquired (Fig. 1). T2-weighted images were deemed

superior in providing optimal contrast of the ventricles and

surrounding tissue. Due to the intrinsically long acquisition

times required for T2-weighted imaging, RARE encoding

was applied to reduce imaging times.

Acquisition parameters for coronal and transverse axial

acquisitions consisted of TE/TR=80/3200 ms, 20 averages,

with an echo train length (ETL) of 8. Coronal images were

obtained with a 256�192 matrix, and transaxial images with

a 192�192 matrix. Slice thickness and fields of view (FOV)

were modified for the different animal ages to reflect the

need for higher resolution imaging for mice at younger ages.

Both coronal and axial images were obtained for all three

genotypes. For coronal images generated from 5-day-old

pups, FOV was 2.6�1.9 cm, and slice thickness was 0.7

mm; axial FOV was 1.9�1.9 and slice thickness was 0.7

mm. For 10-day-old mice, coronal FOV was 3.0�2.1 cm

and slice thickness was 0.7 mm; axial FOV was 2.2�2.2 cm

and slice thickness was 0.8 mm. For adolescent coronal

images, FOV was 4.0�3.2 cm and slick thickness was 0.9

mm; for axial images, FOV was 3.2�3.2 and slick thickness

was 0.9 mm. Ventricular space volumes and corpus

callosum banding and volumes were estimated using

Analyze 5.0 (Biomedical Imaging Resource: Mayo Clinic,

Rochester, MN. Statistical significance was defined as

PV0.05 using one-way ANOVAs and Tukey post-hoc test

comparisons for group significance.


2.5. Behavioral testing procedures

Five-, 10-, and 30-day-old pups were used for all of the

behavioral studies described herein. At each of these

developmental time points, wild-type, heterozygous and

homozygous mice of both sexes were videotaped for 1

minute prior to MR imaging. In brief, mice were placed

individually in a novel environment (a clean, cylindrical

bucket, 27 cm diameter�30 cm height) during daytime of

the diurnal cycle (1000) and spontaneous behavior was

recorded on a Sony Model TRV 900 videotape recorder

equipped with a standard 35 mm Sony min-DV tape.

Animals were allowed to explore this novel environment

for 1 min. Recorded spontaneous exploratory behaviors

were then digitized for each mouse using a Peak Motus

Program. This was achieved by marking several constant

points on every animal recorded: the nose, occiput, left and

right shoulders, left and right hips and the base of the tail.

The aforementioned coordinates were followed for 3 sec to

generate a spatial diagram of motor activities by plotting

the range of circular motion angles [14]. In addition, full

3608 rotational turns were visually recorded by two

individuals blind to the genotype of each mouse subject.

After the testing procedures, animals were either imaged or

returned to their respective dams and littermates. Statistical

significance was defined as PV0.05 using one-way

ANOVAs and Tukey post-hoc test comparisons for group

significance.

Fig. 2. Camera lucida drawings derived from light microscope images of the mous

is depicted. Note the striking enlargement of the lateral ventricle as a function of

uniformed in all three genotypes. Representative sections of the ventricular syste

group. Magnification 4�. +/+=Wild-type; +/�=Heterozygous; �/�=Homozygou

3. Results

3.1. Ventricular enlargement in the adolescent brain

We have previously shown that cross-sectional studies of

the heterozygous and homozygous ckr mouse brain are

characterized by conspicuous enlargement of the lateral

ventricles relative to that of the wild-type C57BL/

10Rospd�C3H/HeRos background strain [32]. Here, using

histochemical techniques, we have confirmed and extended

our previous findings by measuring the area (in mm2) of the

lateral ventricles in all three genotypes (Fig. 2; top panel).

Adolescent wild-type mice showed a ventricular area of

0.134F0.01 mm2, whereas the ventricular area of age-

matched heterozygous and homozygous ckr animals was

0.85F0.04 and 0.973F0.03 mm2, respectively. One-way

ANOVA with genotype as the independent variable and

ventricular size as the dependent variable indicated a

significant effect of genotype (F2,8=184.8, PV0.05). As

expected, lateral ventricular area between heterozygous and

homozygous kin did not differ at all [32]. These results

confirm that the insertion of the Ren-2d renin gene in ckr

mice accounts for most of the ventricular size variance in this

model.

To test whether the above differences in ventricular size

were confined exclusively to the lateral as opposed to the

third or fourth ventricles, area variability in both the third

and fourth ventricle was established among adolescent wild-

e ventricular system. For the lateral ventricles, only the first (right) ventricle

genotype. In contrast, the third and fourth ventricles appeared to be grossly

m depicted here were obtained at random from two mice per genotype per

s.


type, heterozygous and homozygous ckr mice. In all cases

examined (4–6 coronal sections at 50 Am measured per

slide), mouse brains stained with Neutral Red showed

meansFSEM in ventricular area that were similar between

all three genotypes (Pz0.05). For instance, the area of the

third ventricle ranged from 0.120 to 0.128 mm2

(0.124F0.01 mm2) among wild-type, heterozygous and

homozygous ckr mice (Fig. 2; middle panel). Along the

same lines, the area of the fourth ventricle also did not differ

significantly among the three genotypes (Pz0.05; Fig. 2;

bottom panel). Here, the area of the metencephalic ventricle

ranged from 0.690 to 0.708 mm2 (0.69F0.01 mm2). Thus,

the ventricular enlargement profile ostensibly seen in

heterozygous and homozygous ckr mice may be specific

to the lateral ventricular system after all. In general, the

relative area of the third and fourth ventricle is not strongly

affected by the insertion of the Ren-2d renin gene in the

mouse genome. This is consistent with the specificity of

lateral ventricular enlargement in the schizophrenic brain as

well [10].

3.2. In vivo imaging of the ventricular system:

developmental aspects

To directly establish the onset of lateral ventricular

enlargement in ckr mice, we acquired MR images from

wild-type, heterozygous and homozygous mouse brains at 5,

Fig. 3. Representative MRI scans of the lateral ventricles (yellow banding) ob

conspicuous enlargement of the lateral ventricles is readily reconstructed by RF pu

function of age. This structural anatomical pattern is seen in both male and fem

ventricular size are seen in heterozygous mouse brains, suggesting a lesion that m

10 and 30 days after birth (Fig. 3). In general, these

postnatal days represent infancy and adolescent milestones

in rodent life-span trajectories [3]. At day 5 of postnatal life,

the lateral ventricles in wild-type mice were barely

detectable by MR imaging. In contrast, subtle, but statisti-

cally significant changes in lateral ventricular size were

detected in age-matched heterozygous and homozygous ckr

mice. For instance, MR images of the lateral ventricles for

wild-type mice showed densities of 0.16F0.04 mm3,

whereas ventricular scans from heterozygous and homozy-

gous pups showed densities of 0.47F0.01 and 0.47F0.04

mm3, respectively. One-way ANOVA with genotype as the

independent variable and ventricular size as the dependent

variable indicated a significant effect of genotype

(F2,9=19.1, PV0.001). These results suggest that an early

degree of progressive lateral ventricular enlargement is

evident in both heterozygous and homozygous mice, and

that the anatomical specificity of ventricular size is present

in both genotypes irrespective of aberrant circling behavior.

At day 10 of postnatal life, lateral ventricular enlarge-

ment was now more conspicuous in heterozygous and

homozygous ckr mice relative to their wild-type cohorts

(Fig. 3; middle panel). Here, mice containing either one or

two copies of the transgene insertion showed larger

ventricular size volumes (3.2F0.69 and 3.0F0.17 mm3,

respectively) than the C57BL/10Rospd�C3H/HeRos back-

ground strain (0.26F0.03 mm3). One-way ANOVAwith the

tained from mice at different stages of development. At day 5 of life a

lse forces. Such a ventricular enlargement is progressively accelerated as a

ale adolescent mice. It should be noted that inter-individual differences in

ay only be partially expressed in this particular mutant.


genotype as the independent variable and ventricular size as

the dependent variable indicated a significant effect of

genotype (F2,5=16.0, PV0.02). Along the same lines, a

similar developmental profile in lateral ventricular enlarge-

ment was seen at day 30 of postnatal life (Fig. 3; bottom

panel). One-way ANOVA with the genotype as the

independent variables and ventricular size again as the

dependent variable showed a significant effect of genotype

over ventricular size (F2,8=14.3, PV0.005). These results

indicate a progressive and pervasive lateral ventricular

enlargement as a function of age in heterozygous and

homozygous ckr mice, and indicate further a dynamic

structural basis for early brain defects produced by a

transgene insertion.

It should be noted that we had hypothesized that lateral

ventricular enlargement in ckr mice would be present at day

1 of postnatal life, reflecting an earlier and more severe

brain abnormality than previously thought. However, MRI

scans acquired from wild-type and mutant 1-day-old pups

were unable to generate imaging parameters to identify

clearly the ventricular system. The reasons for this are

unknown, but may have included (i) natural distortions at

this stage of brain development (e.g., diffuse brain matter),

and/or (ii) poor conspicuity as a result of partial volume

averaging effects (e.g., reduced structural size). Regardless,

in our experimental groups, enlargement of the lateral

ventricles was most pervasive in 5- and 10-day-old

heterozygous and homozygous ckr pups and was observed

before behavioral symptom onset (see below).

3.3. Corpus callosal abnormalities in 30-day-old pups

To begin to understand the relationship between callosal

morphology and ventricular enlargement in heterozygous

and homozygous ckr mouse brains, we measured point

locations along the entire callosal surface of these animals in

three-dimensional axes. Group differences were present for

callosal area (in mm2) and callosal thickness (in mm)

between wild-type, heterozygous and homozygous ckr

mice, irrespective of brain-size corrections. One-way

ANOVA with the genotype as the independent variable

and corpus callosum area as the dependent variable

indicated a significant effect of genotype (F2,10=8.4,

Fig. 4. Callosal morphology drawings derived from high-resolution MRI scans. C

heterozygous and homozygous ckr mice relative to wild-type cohorts (two mice

ventricular enlargement and corpus callosum abnormalities may reflect an overall

weighted fast spoiled gradients failed to clearly reconstruct callosal morpholog

Homozygous.

PV0.01). In general, both heterozygous and homozygous

30-day-old pups showed a dramatic diminution of callosal

surface when compared with wild-type pups. We also

computed corpus callosum thickness among the three

genotypes. Here, one-way ANOVA again revealed a

significant effect of genotype over this morphometric

parameter (F2,10=28.2, PV0.001) with heterozygous and

homozygous mice exhibiting less thickness in callosal

banding than age-matched control animals. Schematic

diagrams of corpus callosum area and thickness for all

three genotypes are shown in Fig. 4 and indicate that

callosal size differences are associated with the traditional

sign of neuropathology, i.e., ventricular size. Thus,

abnormalities in both ventricular size and callosal mor-

phology, as visualized by MR imaging, are core features

of the heterozygous and homozygous ckr mouse brains.

3.4. Aberrant circling behavior: developmental aspects

We have previously shown that adult homozygous ckr

mice display full-blown circling behavior, a phenotypic

feature uncharacteristic of heterozygote kin or wild-type

mice [25,32]. To determine whether spontaneous circling

behavior and hyperactivity in homozygous ckr mice are

detected during the first weeks of postnatal life, we

videotaped 1-, 5-, 10- and 30-day-old homozygous mice

and compared their behavioral activities, including com-

plete, full (right or left) 3608 circles, mean path-length

movements between stops and distanced moved from in-

place activity with similar behavioral parameters displayed

by age-matched heterozygous and wild-type cohorts. At day

1 of postnatal life, during a 1-min behavioral test in a novel

environment (i.e., away from their respective mothers and

littermates), no specific or salient behavioral activity was

apparent among the three genotypes (data not shown). At

days 5 and 10 of postnatal life, the number and position

parameters of behavioral activity was now more pronounced

in single wild-type, heterozygous, and homozygous ckr

mice. However, it was not until day 10 of postnatal life that

a stable pattern of hyperactivity consisting of forward

locomotion was readily apparent in homozygous ckr mice

relative to their heterozygote kin and wild-type mice (Figs. 5

and 6; top panel). One-way ANOVA with the genotype as

allosal shape differences (e.g., area and thickness) are noted in adolescent

per genotype per group; right hemisphere). Associations between lateral

deterioration of brain structure in ckr mutants. It should be noted that T2-

y in 5- and 10-day-old pups. +/+=Wild-type; +/�=Heterozygous; �/�=

Fig. 5. Integrated behavioral ethograms obtained from neonatal mice individually exposed to a novel environment. Spontaneous movements and postures for

each genotype were videotaped for 1 min during the light phase of the diurnal cycle. Note that by day 10 of postnatal life a distinct pattern of hyperactivity is

readily apparent in homozygous ckr mice. This hyperactivity is progressively culminated by aberrant circling behavior in adolescent 30-day-old ckr pups.


the dependent variable and behavioral activity as the

independent variable demonstrated a significant effect of

genotype (F2,17=5.2, PV0.04). At day 30 of postnatal life,

the hyperactivity parameter previously observed in homo-

zygous ckr mice was now augmented by a consistent

circling behavior with individual turning rates ranging from

20 to 50 full body turns per minute (Figs. 5 and 6; bottom

panel). Again, one-way ANOVA with the genotype and

behavioral activity as the independent and dependent

Fig. 6. Representative rotational counts (means F SEM) derived from neonat

heterozygous mice show a paucity of rotational activity, whereas the rotational pro

genotype per group. *PV0.05 when compared with wild-type and heterozygous a

variable respectively, showed a significant effect of geno-

type (F2,17=15.8, PV0.001).It should be noted that the high degree of circling

behavior and hyperactivity of 30-day-old homozygous ckr

mice persisted for more than the 1-min testing period and

was still observed even after the adolescent mice had been

united with their littermates (data not shown). In general,

these results suggest that homozygous ckr mice show a

spontaneous aberrant circling behavior early in postnatal life

al and adolescent mice exposed to a novel environment. Wild-type and

file of homozygous ckr mice is characterized by full body turns. n=3–4 per

nimals. NS=not significant.


(pre-pubertal), at a developmental time when an imaged

structural change has already been identified in their brains.

However, it is unclear at the present time whether enlarge-

ment of the lateral ventricles and the aberrant behavioral

phenotype have a similar mechanism or are independent.

Regardless, the fact that homozygous ckr mice show

increasing brain and behavioral abnormalities suggests that

subtle deficiencies as a result of carrying two copies of a

transgene insertion accumulate during adolescence to cross

the threshold into schizophrenia-like pathologies.

4. Discussion

We have performed a comprehensive anatomical and

behavioral analysis of neonatal and adolescent ckr mutant

mice. We found that during mouse development, a

progressive structural change occurred in the lateral

ventricles, namely increased ventricular size. This dynamic

pathology is intriguing, as it begins early in postnatal life

(~day 5) and intensifies during adolescence (~day 30). This

anatomical profile mirrors what is thought to occur in brains

of schizophrenics in late adolescence or early adulthood

[19]. Indeed, enlargement of the lateral ventricles is among

the most frequently reported imaged structural change in

adult schizophrenia [12]. In addition, recent structural MRI

studies indicate a significant loss of cortical gray matter in

early-onset schizophrenia, a neural event that intensifies

over years of disease progression [31]. Thus, there is a

progressive deterioration of structure in the schizophrenic

brain, a view clearly consistent with earlier reports of

ventricular enlargement and corpus callosum displacement

[12,24]. The fact that our MRI studies uncovered a dynamic

change in the time and rate of lateral ventricular pathology

in homozygous ckr mice suggests that these animals, too,

undergo a progressive deterioration of structure. Thus, the

timing and structural changes in ckr mutant mice are similar

to those observed in schizophrenic patients.

The anatomical profile of homozygous ckr mutant mice

is not only similar to that observed in schizophrenia but also

to periventricular leukomalacia (PVL), a neurological

disease characterized by white matter loss and lateral

ventricular expansion [17,35]. Further, ciliary dyskinesia,

Dandy-Walker malformation and several X-linked disorders

are also associated with pathologically enlarged ventricles

[1,4,6,15,16]. At the clinical level, all the aforementioned

disorders are characterized by mental deficits and/or motor

abnormalities, including stereotypy, catatonia and abnormal

posture and limb movements. This suggests that a complex

set of brain illnesses with similar cognitive and behavioral

syndromes across diverse patient populations share a

common structural phenotype: enlargement of the cerebral

ventricles. Thus, ventricular enlargement could be used as

an early diagnostic marker to identify individuals with a

genetic vulnerability for developing profound cognitive and

motor deficits. The questions now are to (i) ascertain the

degree of genetic linkage between cerebral ventricular size

and disease onset, and (ii) determine why ventricular

pathology would lead to selective clinical symptoms.

Clearly, animal models of this disease could potentially

provide answers that are relevant to schizophrenia. In this

regard, ventricular size in mice is a highly heritable trait

modulated by the additive effects of several genes. For

instance, significant quantitative trait locus (QTL) on

chromosome 8 with epistatic interactions between loci on

chromosomes 4 and 7 is found in the mouse genome [37].

Of significance, the QTLs are located in close proximity to

genes underlying pathologically enlarged ventricles and

may therefore represent exonic sequences related to

mammalian ventricular phenotypes. Unfortunately, the

precise chromosomal rearrangement caused by the insertion

of the Ren-2d renin gene in ckr mice is unknown [29,30].

What is clear, however, is that such an insertion disrupted

certain regulatory units on chromosome 16 which may have

had an impact on additional sequences with strong epistatic

interaction for genes implicated in ventricular size. Studies

addressing this issue are currently on-going projects in our

laboratories.

It is thought that certain structural changes in the

schizophrenic brain arise during fetal development

[10,19]. Enlargement of the lateral ventricles may well be

one of the first neural irregularities associated with the

disease. In this regard, prenatal risk factors for schizophre-

nia such as sepsis, hypoxia and retrovirus exposure may

contribute to the onset and initial progression of ventricular

lesions. This is clearly observed in mouse pups reared under

hypoxic conditions and in mice deficient in adenosine

receptors (A1ARs). In this context, adenosine acting on

A1ARs appears to mediate hypoxia-induced ventricular

enlargement in animal models of PVL [34]. It is conceivable

therefore that once ventricular enlargement sets-in, it may

contribute to focal shrinkage in specific neural circuits

implicated in the pathophysiology of schizophrenia. Indeed

subtle, yet significant, shrinkage in the cortex, striatum and

thalamus is revealed in structural brain imaging studies [12].

These regionally specific reductions of brain parenchyma

are also highly correlated with ventricular enlargement;

deficits that may somehow precipitate the emergence of a

synaptic deficiency syndrome in schizophrenia. Thus

ventricular pathology, as indicated by our homozygous ckr

mouse studies, may contribute to the underlying roots of

schizophrenia, ultimately giving rise to fragmentation of

synaptic networks and clinical manifestations of the disease.

Abnormalities of the corpus callosum are also highly

linked with lateral ventricular enlargement in first episode

schizophrenic patients [5,8,23]. Thus, callosal surface

displacement may represent an additional insult to the

schizophrenic brain, at least in some susceptible individ-

uals [24]. Here, we report that heterozygous and homo-

zygous ckr mice also show abnormalities in callosal

surface area and thickness relative to wild-type controls.

Specifically, T2-weighted MR images show a significant


thinning of callosal banding in both mutant genotypes.

Such findings strongly suggest that enlargement of the

lateral ventricles and callosal abnormalities are highly

linked in mutant mice and support further the tenet that the

spectrum of abnormalities in ckr mutant mice is similar to

that observed in schizophrenia patients. Under this

scenario, defects in both ventricular size and callosal

morphology occurring in the nascent brain could reach a

certain threshold that propels adjacent neurons into a state

of disarray resulting in aberrant cognitive and behavioral

deficits. In this context, the corpus callosum is a para-

ventricular region that physically connects the two

homologous hemispheres of the mammalian brain. It is

thought that selective abnormalities of callosal architecture

may limit the flow of neuronal signals between the two

hemispheres, resulting in impairment of some but not all

region-specific functions for each hemisphere [22,36].

Although our non-invasive MR images provide evidence

for corpus callosum pathology in ckr mice, it is not yet

clear how this deficit relates to any of the cognitive

disorders that are seen in schizophrenia. In this regard, it

should be noted that mutant mice only reproduce certain

aspects of the human disease phenotype. Any mouse

model will be limited until the entire biochemical nature of

schizophrenia is established. Nevertheless, the structural

deteriorations we observe in heterozygous and homozy-

gous ckr mouse brains supports the idea that developmen-

tal abnormalities may play a critical role in shaping the

vulnerability to schizophrenia.

The neuroanatomical abnormalities of homozygous (but

not heterozygous) ckr mice are also associated with aberrant

circling behavior and hyperactivity. These behavioral

deficits are similar to those reported in some currently

available mouse models of schizophrenia [11]. For instance,

mice lacking the dopamine transporter as well as animals

deficient in calcineurin and reelin show multiple abnormal

behaviors related to schizophrenia, including hyperactivity

and impairments in pre-pulse inhibition [11,20,33]. Thus,

ckr mice and other genetically engineered mutants are

valuable models to elucidate certain behavioral traits

deemed to be abnormal in schizophrenia. In this regard,

circling behavior and hyperactivity in un-medicated patients

are features of the disease and thus represent quantitative

differences between the normal expression and the over-

expression of such behavioral traits [26]. Consistent with

this notion, 10-day- and 30-day-old homozygous ckr pups

exhibit differences in circling behavior and hyperactivity

relative to heterozygous and wild-type mice. These findings

suggest that ckr mice, like schizophrenic patients, are

characterized by early developmental defects followed by

progressive behavioral symptoms. In our studies, a video-

based system followed by digitized clips of spontaneous

behaviors show adolescent mutants displaying excessive

bouts of hyperactivity and circling behavior earlier than

previously thought [25]. At day 10 of postnatal life, ckr

mice are already engaging in a series of abnormal actions

and postures listed as ethograms. Of interest, these

behavioral abnormalities can effectively be reduced in

adulthood by neuroleptic medications such as clozapine

and olanzapine [32]. Clearly, a pharmacological model of

the disease provides further insights into the chemical

systems that might be involved in the development of

psychotic behaviors. Finally, the fact that only the

homozygous ckr mouse shows both a structural deficit

and aberrant behavior phenotype opens the way to under-

standing how its heterozygote kin, with a single mutant

chromosome, escapes the manifestation of hyperactivity

and circling behavior. It is thought that the keys to

unlocking the etiology of schizophrenia may lie not in the

patients themselves but in their unaffected relatives. The

homozygous ckr mouse with heterozygote kin offers the

opportunity to (i) test such a hypothesis and (ii) dissect the

disease into component trait complexes. This approach

represents a conceptual shift in the use of available animal

models of schizophrenia.

Acknowledgements

The authors are indebted to Courtney Grim, Valerie

Pawlowski and Aaron Miller (New Media Institute,

Medaille College) and Elizabeth A. Doran (New Technol-

ogies Initiative, NYCOM) for their excellent technical

assistance. This study was supported by an NIH grant

(#1R15MH64513-01A1) to J.M.H., and in part by a grant

(#CA-76561) and an Institute Comprehensive Cancer

Center Support Grant (#CA-16056) from the National

Cancer Institute, Bethesda, MD.

References

[1] M. al-Shroof, A.M. Karnik, A.A. Karnik, J. Longshore, N.A. Sliman,

F.A. Khan, Ciliary dyskinesia associated with hydrocephalus and

mental retardation in a Jordanian family, Mayo Clin. Proc. 76 (2001)

1219–1224.

[2] W.F. Baare, C.J. van Oel, H.E. Hulshoff Pol, H.G. Schnack, S.

Durston, M.M. Sitskoorn, Volumes of brain structures in twins

discordant for schizophrenia, Arch. Gen. Psychiatry 58 (2001)

33–40.

[3] S.A. Bayer, J. Altman, R.J. Russo, X. Zhang, Timetables of

neurogenesis in the human brain based on experimentally determined

patterns in the rat, Neurotoxicology 14 (1993) 83–144.

[4] S.S. Brooks, K. Wisniewski, W.T. Brown, New X-linked mental

retardation (XLMR) syndrome with distinct facial appearance and

growth retardation, Am. J. Med. Genet. 51 (1994) 157–167.

[5] M.F. Casanova, M. Zito, T.E. Goldberg, R.L. Suddath, Corpus

callosum curvature in schizophrenic twins, Biol. Psychiatry 28

(1990) 83–85.

[6] D.P. Cavalcanti, M.A. Salomao, Dandy-Walker malformation with

postaxial polydactyly: further evidence for autosomal recessive

inheritance, Am. J. Med. Genet. 85 (1999) 183–184.

[7] G.M. Church, W. Gilbert, Genomic sequencing, Proc. Natl. Acad. Sci.

U. S. A. 81 (1984) 1991–1995.

[8] J.E. Downhill Jr., M.S. Buchsbaum, T. Wei, J. Spiegel-Cohen, E.A.

Hazlett, M.M. Haznedar, J. Silverman, L.J. Siever, Shape and size of


the corpus callosum in schizophrenia and schizotypal personality

disorder, Schizophr. Res. 42 (2000) 193–208.

[9] L.W. Fitzgerald, A.K. Ratty, M. Teitler, K.W. Gross, S.D. Glick,

Specificity of behavioral and neurochemical dysfunction in the

chakragati mouse: a novel genetic model of a movement disorder,

Brain Res. 608 (1993) 247–258.

[10] W.G. Frankle, J. Lerma, M. Laruelle, The synaptic hypothesis of

schizophrenia, Neuron 39 (2003) 205–216.

[11] R.R. Gainetdinov, A.R. Mohn, L.M. Bohn, M.G. Caron, Gluta-

matergic modulation of hyperactivity in mice lacking the

dopamine transporter, Proc. Natl. Acad. Sci. U. S. A. 98 (2001)

11047–11054.

[12] C. Gaser, I. Nenadic, B.R. Buchsbaum, E. Hazlett, M. Buchsbaum,

Ventricular enlargement in schizophrenia related to volume reduction

of the thalamus, striatum, and superior temporal cortex, Am. J.

Psychiatry. 161 (2004) 154–156.

[13] W.S. Gharaibeh, F.J. Rohlf, D.E. Slice, L.E. Delisi, A geometric

morphometric assessment of change in midline brain structural shape

following a first episode of schizophrenia, Biol. Psychiatry 48 (2000)

398–405.

[14] J.M. Horowitz, A. Goyal, N. Ramdeen, B.H. Hallas, A.T. Horowitz,

G. Torres, Characterization of fluoxetine plus olanzapine treatment in

rats: a behavior, endocrine, and immediate-early gene expression

analysis, Synapse 50 (2003) 353–364.

[15] S. Katsuragi, K. Teralka, K. Ikegami, K. Amano, K. Yamashite, K.

Ishizuka, T. Miyakawa, Late onset X-linked hydrocephalus with

normal cerebrospinal fluid pressure, Psychiatry Clin. Neurosci. 54

(2000) 487–492.

[16] S. Kenwrick, M. Jouet, D. Donnai, X-linked hydrocephalus and

MASA syndrome, J. Med. Genet. 33 (1996) 59–65.

[17] K. Kuban, U. Sanocka, A. Leviton, E. Allred, M. Pagano, O.

Dammann, J. Share, D. Rosenfeld, M. Abiri, D. DiSalvo, P. Doubilet,

R. Kairam, E. Kazam, M. Kirpekar, S. Schonfeld, White matter

disorders of prematurity: Association with intraventricular hemor-

rhage and ventriculomegaly, J. Pediatr. 134 (1990) 539–546.

[18] S.M. Lawrie, S.S. Abukmeil, Brain abnormality in schizophrenia. A

systematic and quantitative review of volumetric magnetic resonance

imaging studies, Br. J. Psychiatry 172 (1999) 110–120.

[19] D.A. Lewis, P. Levitt, Schizophrenia as a disorder of neurodevelop-

ment, Annu. Rev. Neurosci. 25 (2002) 409–432.

[20] T. Miyakawa, L.M. Leiter, D.J. Gerber, R.R. Gainetdinov, T.D.

Sotnikova, H. Zeng, M.G. Caron, S. Tonegawa, Conditional

calcineurin knockout mice exhibit multiple abnormal behaviors

related to schizophrenia, Proc. Natl. Acad. Sci. U. S. A. 100 (2003)

8987–8992.

[21] B. Moghaddam, Bringing order to the glutamate chaos in schizo-

phrenia, Neuron 40 (2003) 881–884.

[22] B. Mohr, F. Pulvermueller, R. Cohen, B. Rockstroh, Interhemispheric

cooperation during word processing: evidence for callosal transfer

dysfunction in schizophrenic patients, Schizophr. Res. 46 (2000)

231–239.

[23] K.L Narr, P.M. Thompson, T. Sharma, J. Moussai, A.F. Cannestra,

A.W. Toga, Mapping morphology of the corpus callosum in

schizophrenia, Cereb. Cortex 10 (2000) 40–49.

[24] K.L. Narr, T.D. Cannon, R.P. Woods, P.M. Thompson, S. Kim, D.

Asunction, T.G.M. van Erp, V. Poutanen, M. Huttunen, J. Lonnqvist,

C. Standerksjold-Nordenstam, J. Kaprio, J.C. Mazziotta, A.W. Toga,

Genetic contributions to altered callosal morphology in schizophrenia,

J. Neurosci. 22 (2002) 3720–3729.

[25] A.K. Ratty, L.W. Fitzgerald, M. Titeler, S.D. Glick, J.J. Mullins, K.W.

Gross, Circling behavior exhibited by a transgenic insertional mutant,

Mol. Brain Res. 8 (1990) 355–358.

[26] A.K. Ratty, K.W. Gross, L.W. Fitzgerald, S.D. Glick, The chakragati

mouse: a model for brain dopaminergic dysfunction, in: H.B. Niznik

(Ed.), Dopamine Receptors and Transporters: Pharmacology, Struc-

ture and Function, Marcel Dekker, New York, 1994.

[27] T. Sharma, E. Lancaster, D. Lee, S. Lewis, T. Sigmundsson, N. Takei,

H. Gurling, P. Barta, G. Pearlson, R. Murray, Brain changes in

schizophrenia. Volumetric MRI study of families multiply affected

with schizophrenia-the Maudsley Family Study 5, Br. J. Psychiatry

173 (1998) 132–138.

[28] L. Shihabuddin, J.M. Silverman, M.S. Buchsbaum, L.J. Seiver, C.

Luu, M.K. Germans, Ventricular enlargement associated with linkage

marker for schizophrenia-related disorders in one pedigree, Mol.

Psychiatry 3 (1996) 215–222.

[29] D.J. Smiraglia, A.K. Ratty, K.W. Gross, Physical characterization of

the chromosomal rearrangements that accompany the transgene

mutation in the chakragati mouse mutant, Genomics 45 (1997)

562–571.

[30] D.J. Smiraglia, C. Wu, M.K. Ellsworth, A.K. Ratty, V.M. Chapman,

K.W. Gross, Genetic characterization of the chromosomal rearrange-

ments that accompany the transgene insertion in the chakragati mouse

mutant, Genomics 45 (1997) 572–579.

[31] P.M. Thompson, K.L. Narr, R.E. Blanton, A.W. Toga, Mapping

structural alterations of the corpus callosum during brain development

and disease. In: The Corpus Callosum. M. Iacoboni, E. Zaidel (eds).

Netherlands: Kluwer Academic Publishers.

[32] G. Torres, B.H. Hallas, V.A. Vernace, C. Jones, K.W. Gross, J.M.

Horowitz, A neurobehavioral screening of the ckr mouse mutant:

implications for an animal model of schizophrenia, Brain Res. Bull.

62 (2004) 315–326.

[33] L. Tremolizzo, G. Carboni, B. Ruzicka, C.P. Mitchell, I. Sugaya, P.

Tueting, R. Sharma, D.R. Grayson, E. Costa, An epigenetic mouse

model for molecular and behavioral neuropathologies related to

schizophrenia vulnerability, Proc. Natl. Acad. Sci. U. S. A. 99

(2002) 17095–17100.

[34] C.P. Turner, M. Seli, L. Ment, W. Stewart, H. Yan, B. Johansson,

B.B. Fredholm, M. Blackburn, S.A. Rivkees, A1 adenosine receptors

mediate hypoxia-induced ventriculomegaly, Proc. Natl. Acad. Sci.

U. S. A. 100 (2003) 11718–11722.

[35] J.J. Volpe, Neurobiology of periventricular leukomalacia in the

premature infant, Pediatr. Res. 50 (2001) 552–562.

[36] P.W.R. Woodruff, M.L. Phillips, T. Rushe, I.C. Wright, Corpus

callosum size and inter-hemispheric function in schizophrenia,

Schizophr. Res. 23 (1997) 189–196.

[37] C.C. Zygourakis, G.D. Rosen, Quantitative trait loci modulate

ventricular size in the mouse brain, J. Comp. Neurol. 461 (2003)

362–369.

Ventricular size mapping in a transgenic model of schizophrenia

Documents

Transcript of Ventricular size mapping in a transgenic model of schizophrenia