Schizophrenia-relevant behaviors in a genetic mouse model of constitutive Nurr1 deficiency

Authors


U. Meyer, Laboratory of Behavioural Neurobiology, Swiss Federal Institute of Technology (ETH) Zurich, Schorenstrasse 16, 8603 Schwerzenbach, Switzerland. E-mail: urmeyer@ethz.ch

Abstract

Nurr1 (NR4A2) is an orphan nuclear receptor highly essential for the dopaminergic development and survival. Altered expression of Nurr1 has been suggested as a potential genetic risk factor for dopamine-related brain disorders, including schizophrenia. In support of this, recent experimental work in genetically modified mice shows that mice with a heterozygous constitutive deletion of Nurr1 show a facilitation of the development of schizophrenia-related behavioral abnormalities. However, the behavioral characterization of this Nurr1-deficient mouse model remains incomplete. This study therefore used a comprehensive behavioral test battery to evaluate schizophrenia-relevant phenotypes in Nurr1-deficient mice. We found that these mice displayed increased spontaneous locomotor activity and potentiated locomotor reaction to systemic treatment with the non-competitive N-methyl-d-aspartate (NMDA) receptor antagonist, dizocilpine (MK-801). In addition, male but not female Nurr1-deficient mice showed significant deficits in the prepulse inhibition and prepulse-elicited reactivity. However, Nurr1 deletion did not induce overt abnormalities in other cardinal behavioral and cognitive functions known to be impaired in schizophrenia, including social interaction and recognition, spatial recognition memory or discrimination reversal learning. Our findings thus suggest that heterozygous constitutive deletion of Nurr1 results in a restricted phenotype characteristic of schizophrenia symptomatology, which primarily relates to motor activity, sensorimotor gating and responsiveness to the psychomimetic drug MK-801. This study further emphasizes a critical role of altered dopaminergic development in the precipitation of specific brain dysfunctions relevant to human psychotic disorder.

Nurr1 (NR4A2) is an orphan member of the steroid hormone receptor family with essential functions in both the final differentiation and the survival of ventral mesencephalic dopaminergic precursor neurons (Kadkhodaei et al. 2009; Saucedo-Cardenas et al. 1998; Zetterström et al. 1996). Nurr1 appears to be an obligatory factor in the mesencephalic dopaminergic neuron development and maturation because constitutive and complete ablation of Nurr1 leads to full agenesis of dopamine but not other monoamine neurons (Zetterström et al. 1996). It also seems to play a critical role in the migration and striatal target area innervation of differentiating mesencephalic dopaminergic cells (Wallén et al. 1999). Furthermore, together with other seminal transcription factors, such as Pitx3 and Lmx1b, Nurr1 exerts a number of functions in postmitotic and mature mesencephalic dopamine neurons, including the regulation of tyrosine hydroxylase, dopamine transporter and vesicular monoamine transporter 2 expression (Kadkhodaei et al. 2009; Saucedo-Cardenas et al. 1998; Smidt & Burbach 2007).

Nurr1 has been suggested as a potential susceptibility gene in schizophrenia. This is based on the observation that two different missense mutations in the third exon of the Nurr1 gene, which were coupled with a 30–40% reduction in in vitro transcriptional activity of Nurr1 dimers, were present in some schizophrenic patients (Buervenich et al. 2000). In addition, Nurr1 messenger RNA was found reduced in the prefrontal cortex in schizophrenia (Xing et al. 2006). In view of the critical importance of Nurr1 in dopamine development and function, altered Nurr1 expression may thus contribute to the dopaminergic pathology present in patients with schizophrenia (Guillin et al. 2007; Howes & Kapur 2009). However, other studies have failed to establish an association between Nurr1 and schizophrenia (Carmine et al. 2003; Ruano et al. 2004), implying that the impact of Nurr1 on schizophrenic risk is currently unclear and requires further exploration.

There is also initial experimental evidence for a role of altered Nurr1 expression in the development of schizophrenia-related behavioral abnormalities in animal models. First, the developmental alterations in Nurr1 expression have been implicated in environmental neurodevelopmental animal models of schizophrenia (Cui et al. 2010; Meyer et al. 2008a; Vuillermot et al. 2010). Furthermore, Nurr1+/− mice have been reported to display spontaneous and psychostimulant-induced hyperactivity (Bäckman et al. 2003; Rojas et al. 2007), increased susceptibility to stress (Eells et al. 2006; Rojas et al. 2007) and altered emotional memory (Rojas et al. 2007). On the basis of this, it has been suggested that Nurr1+/− mice may provide a genetic animal model with substantial face validity for schizophrenia-like behavioral and pharmacological dysfunctions. However, the behavioral characterization of Nurr1-deficient mice with respect to schizophrenia-relevant phenotypes is far from complete. Extending the phenotypic characterization of Nurr1-deficient mice to other tests relevant to human psychotic disease is thus clearly warranted to further ascertain the face validity of the Nurr1-deficient mouse model for schizophrenia and related disorders.

Therefore, this study was undertaken to further investigate whether Nurr1-deficient mice may display multiple schizophrenia-related abnormalities. Nurr1-deficient mice were tested relative to wild-type (WT) mice in a comprehensive battery of behavioral, cognitive and pharmacological tests, which have been designed to assess schizophrenia-relevant functions in translational animal models (Meyer & Feldon 2010). This approach includes tests measuring basal and drug-induced locomotor activity, sensorimotor gating, selective attention, social interaction, spatial memory and cognitive flexibility, all of which have been implicated in human psychotic disorders (Meyer & Feldon 2010).

Materials and methods

Animals

The subjects included male and female mice with a heterozygous deletion of the Nurr1 gene (Nurr1+/−; referred to as heterozygous knockout [KO] animals) and their WT littermates (Nurr1+/+; referred to as WT animals). This study focused on heterozygous Nurr1-deficient mice because mice lacking both alleles of Nurr1 (i.e. Nurr1−/−) die soon after birth and are devoid of dopaminergic neurons in the midbrain (Le et al. 1999a; Zetterström et al. 1997). In contrast, heterozygous Nurr1-deficient mice survive postnatally without obvious locomotor deficits (Le et al. 1999a; Zetterström et al. 1997) and can thus be used for experimental investigations in adulthood. The generation of Nurr1-mutated mice with a 129/Sv genetic background has been fully described before (Zetterström et al. 1997). Nurr1+/− mice were backcrossed for more than 15 generations to the C57BL/6 strain. The genotype of each animal was determined by polymerase chain reaction of tail DNA using specific probes for the Nurr1 gene as previously described (Zetterström et al. 1997).

All animals were kept in a temperature- and humidity-controlled (21 ± 1°C, 55 ± 5%) holding facility under a reversed light/dark cycle (lights off: 0800 to 2000 h). All animals had ad libitum access to food (Kliba 3430, Klibamühlen, Kaiseraugst, Switzerland) and water, and they were group-caged (three to five animals) in Macrolon type III cages unless specified otherwise. All procedures described in this study had been previously approved by the Cantonal Veterinarian's Office of Zurich and are in agreement with the principles of laboratory animal care in the Guide for the Care and Use of Laboratory Animals (National Institutes of Health Publication No. 86-23, revised 1985). All efforts were made to minimize the number of animals used and their suffering.

Behavioral testing: test order

Behavioral testing commenced when the animals reached the adult stage of development, i.e. between postnatal days (PNDs) 70 and 135. All testing was carried out in the dark phase of the light/dark cycle. Three different cohorts of animals were used to accomplish all behavioral, cognitive and pharmacological tests of interest. The use of three different cohorts served to minimize potential confounding factors associated with prolonged behavioral testing and aging. The first cohort was used for assessing (1) spontaneous locomotor activity in the open field, (2) spatial novelty preference in the Y-maze, (3) prepulse inhibition (PPI) of the acoustic startle reflex, (4) social interaction, (5) latent inhibition (LI) in a two-way conditioned active avoidance (CAA) paradigm and (6) discrimination reversal learning in the water T-maze. Hence, animals from this cohort were repeatedly tested according to the specified test order. This test order was designed in such a way that the arguably least stressful test came first (open field) and the most stressful tests last (reversal learning in the water T-maze), with the purpose of minimizing potential stress-related carryover effects from one test to another. The second cohort was used for assessing LI in the conditioned taste aversion (CTA) task and for evaluating the locomotor response to acute systemic treatment with dizocilpine (MK-801) using a within-subject treatment design. All animals from the second cohort were first tested in the CTA LI paradigm; subsequently, a subgroup of the animals was then tested in the MK-801 sensitivity test. Finally, a third cohort of animals served to assess the locomotor response to acute systemic MK-801 treatment using a between-subject treatment design. The precise number of animals used in each experimental group and test is summarized in Table S1.

Behavioral testing: cohort I

Spontaneous locomotor activity in the open field

Spontaneous locomotor activity was evaluated using a standard open-field exploration task. The main rationale of using this test was to confirm the presence of enhanced spontaneous locomotor behavior in our cohort of Nurr1 heterozygous KO animals relative to WT animals (Bäckman et al. 2003; Rojas et al. 2007). Another rationale for investigating the spontaneous locomotor activity in Nurr1-deficient animals relative to WT animals was to explore the behavioral correlates of subcortical dopaminergic imbalances relevant especially to the positive symptoms of schizophrenia (van den Buuse et al. 2005). The apparatus consisted of four identical open-field arenas (40 × 40 × 35-cm high) made of wood and painted white as described before (Meyer et al. 2005). They were located in a testing room under dim diffused lighting (approximately 35 lux as measured in the center of the arenas). A digital camera was mounted directly above the four arenas. Images were captured at a rate of 5 Hz and transmitted to a PC running the EthoVision tracking system (Noldus Information Technology, Wageningen, The Netherlands). Spontaneous locomotor activity was assessed for 1 h.

Spatial recognition memory in the Y-maze

Spatial recognition memory was evaluated by a spatial novelty preference task in the Y-maze. The spatial novelty preference test in the Y-maze assesses short-term spatial memory and uses the natural tendency of rodents to explore novel over familiar spatial environments (Dellu et al. 1992). The primary rationale of using this test was to explore the impact of Nurr1 deficiency on short-term spatial memory in relation to the known impairments of memory for spatial context in schizophrenic patients (Brébion et al. 2004).

The apparatus was made of transparent Plexiglass and consisted of three identical arms (50 × 9 cm; length × width) surrounded by 10-cm high transparent Plexiglass walls. The three arms radiated from a central triangle (8 cm on each side) and were spaced 120° from each other. Access to each arm from the central area could be blocked by a removable opaque barrier wall. The floor of the maze was covered with sawdust bedding, which was changed between both the testing phases and the trials. The maze was elevated 90 cm above the floor and was positioned in a well-lit room enriched with distal spatial cues. For each retention interval tested (see below), the experiment was performed in a different room with a distinct set of extra-maze cues surrounding the Y-maze, to avoid confounds by familiar visual cues. A digital camera was mounted above the Y-maze apparatus. Images were captured at a rate of 5 Hz and transmitted to a PC running the EthoVision tracking system (Noldus Information Technology), which calculated the time spent and distance moved in the three arms and center zone of the Y-maze. The test of spatial recognition memory in the Y-maze consisted of two phases, called the sample and choice phases. The allocation of arms (start, familiar and novel arm) to a specific spatial location was counterbalanced across the subjects.

  • Sample phase: The animals were allowed to explore two arms (referred to as ‘start arm’ and ‘familiar arm’). Access to the remaining arm (‘novel arm’) was blocked by a barrier wall door. To begin a trial, the subject was introduced at the end of the start arm and was allowed to freely explore both the start and the familiar arms for 5 min. Test timing was initiated once the subject had made an entry into the central triangular area, as detected by the EthoVision tracking system. The subject was then removed and kept in a holding cage during the specific retention intervals (see below) prior to the choice phase. The barrier door was removed and the sawdust flooring changed to avoid olfactory cues.
  • Choice phase: Following a specific retention interval (see below), the test subject was introduced to the maze again. During the choice phase, the barrier wall was removed so that the animals could freely explore all arms of the maze for 2 min. The subject was then removed from the maze and returned to the home cage. The sawdust flooring was changed in preparation for the next trial. On each trial, the time spent in each of the three arms was recorded. The relative time spent in the novel arm during the choice phase was calculated by the formula [time spent in the novel arm/[time spent in all arms]) × 100 and used as the index for spatial novelty preference. In addition, total distance moved on the entire maze was recorded and analyzed in order to assess general locomotor activity.

To manipulate the retention demand in the temporal domain, the interval between the two phases (i.e. sample and choice phases) of the Y-maze test was varied. First, a minimal interval of 1 min was used. This minimal delay referred to the time needed to change the sawdust flooring between the two phases in order to avoid confounds by olfactory cues. The interval between the two phases was then increased to 15 min, 30 min and 24 h.

PPI of the acoustic startle reflex

Sensorimotor gating was assessed using the paradigm of PPI of the acoustic startle reflex. PPI of the acoustic startle reflex refers to the reduction in startle reaction in response to a startle-eliciting pulse stimulus when it is shortly preceded by a weak prepulse stimulus (Hoffman & Searle 1965). PPI deficits have been consistently reported in acute and chronic schizophrenia patients, as well as in psychosis-prone individuals (Braff et al. 2001). Therefore, the study of PPI in Nurr1-deficient mice relative to WT mice would allow us to gain insight into the potential importance of Nurr1 in modulating the expression of sensorimotor gating with relevance to human psychotic disease.

The apparatus consisted of four startle chambers for mice (San Diego Instruments, San Diego, CA, USA). The test apparatus has been fully described elsewhere (Meyer et al. 2005). In the demonstration of PPI of the acoustic startle reflex, subjects were presented with a series of discrete trials comprising a mixture of four trial types. These included pulse-alone trials, prepulse-plus-pulse trials, prepulse-alone trials and no-stimulus trials in which no discrete stimulus other than the constant background noise was presented. The pulse and prepulse stimuli used were in the form of a sudden elevation in the broadband white noise level (sustaining for 40 and 20 ms, respectively) from the background (65 dBA), with a rise in time of 0.2–1.0 milliseconds. In all, three different intensities of pulse (100, 110 and 120 dBA) and three intensities of prepulse (71, 77 and 83 dBA, which corresponded to 6, 12 and 18 dBA above background, respectively) were used. The stimulus-onset asynchrony of the prepulse and pulse stimuli on all prepulse-plus-pulse trials was 100 milliseconds (onset-to-onset).

A session began with the animals being placed into the Plexiglass enclosure. They were acclimatized to the apparatus for 2 min before the first trial began. The first six trials consisted of six startle-alone trials, comprising two trials of each of the three possible pulse intensities. These trials served to habituate and stabilize the animals' startle response and were not included in the analysis. Subsequently, the animals were presented with 10 blocks of discrete test trials. Each block consisted of the following: three pulse-alone trials (100, 110 or 120 dBA), three prepulse-alone trials (+6, +12 or +18 dBA above background), nine possible combinations of prepulse-plus-pulse trials (three levels of prepulse × three levels of prepulse) and one no stimulus. The 16 discrete trials within each block were presented in a pseudorandom order, with a variable interval of a mean of 15 seconds (ranging from 10 to 20 seconds). For each of the three pulse intensities (100, 110 or 120 dBA), PPI was indexed by per cent inhibition of the startle response obtained in the pulse-alone trials by the following expression: 100% × (1 − [mean reactivity on prepulse-plus-pulse trials/mean reactivity on pulse-alone trials]), for each subject and at each of the three possible prepulse intensities (+6, +12 or +18 dBA above background).

Social interaction and recognition

Social behavior is commonly referred to behavior that takes place in a social context and results from the interaction between individuals (of the same species). As mice (like most other rodents) are highly social animals, social interaction can be efficiently studied under experimental conditions (Crawley 2007). Importantly, evaluation of social interaction has clear relevance to schizophrenia because deficient social interaction is one of the hallmark negative symptoms in schizophrenia (Foussias & Remington 2010), so that the social interaction and recognition test used here served to capture possible abnormalities in this behavioral domain.

The apparatus was made of opaque Plexiglass and consisted of three identical arms (50 × 9 cm; length × width) surrounded by 10-cm high Plexiglass walls. The three arms radiated from a central triangle (8 cm on each side) and spaced 120° from each other. The floor of the maze was covered with sawdust bedding, which was changed between each individual habituation and test trial. Two of the three arms contained rectangular wire grid cages (13 × 8 × 10 cm, length × width × height; bars horizontally and vertically spaced 9 mm apart). The third arm did not contain a metal wire cage and served as the start zone (see below).

The subjects were first habituated to the apparatus by being allowed to explore the Y-maze for 8 min on two consecutive days. This served to familiarize the test animals with the apparatus and to reduce novelty-related locomotor hyperactivity that may potentially confound social interaction during the critical test phase. The rectangular wire cages (located at the end of two arms) were left empty during the habituation phase.

The test of social interaction consisted of two phases, namely the ‘dummy phase’ and the ‘test phase’. The allocation of arms (start, familiar and novel arm) to a specific spatial location was counterbalanced across subjects.

  • Dummy phase: The animals were allowed to explore the three arms (referred to as ‘start arm’, ‘dummy arm’ and ‘live arm’). During this phase, one metal wire cage contained an unfamiliar C57BL/6NIH mouse (‘live mouse’); and the other wire cage contained an inanimate object (‘dummy mouse’), which was made of black LEGO™ (Billund, Denmark) bricks and took the shape of a mouse. To begin a trial, the subject was introduced at the end of the start arm and was allowed to freely explore all three arms for 5 min. Test timing was initiated once the subject had made an entry into the central triangular area, as detected by the EthoVision XT (Noldus Information Technology) tracking system. Social interaction was defined as nose detection within a 4-cm interaction zone adjacent to the metal wire cages. In addition, total distance traveled in the three arms and central zone of the test apparatus were recorded in order to obtain measures of general locomotor activity. On completion of the ‘dummy phase’, the subject was removed and kept in a holding cage, during which the sawdust flooring was changed to avoid olfactory cues.
  • Test phase: The inanimate dummy mouse was now replaced by another unfamiliar C57BL/6NIH mouse, which is referred to as the ‘novel mouse’ during the test phase. The other cage contained the ‘familiar mouse’ previously used in the ‘dummy phase’. The allocation of the ‘novel mouse’ and ‘familiar mouse’ to the two wire cages was counterbalanced across experimental groups.

To start the ‘test phase’, the subject was introduced into the maze again. During the ‘test phase’, the subject was allowed to freely explore all three arms for 5 min. Exploration was recorded using the EthoVision XT (Noldus Information Technology) tracking system and was defined as nose detection within a 4-cm interaction zone adjacent to the metal wire cages as described before.

Latent inhibition in the CAA paradigm

LI is a selective learning paradigm, in which prior repeated pre-exposures (PEs) to the to-be-conditioned stimulus (CS) retards subsequent development of the conditioned response (CR) following explicit pairings between the same CS and an unconditioned stimulus (US). Deficiency in LI has been related to the attentional deficits commonly observed in schizophrenic patients (Feldon & Weiner 1992; Weiner 2003), which may be central to the patients' difficulties in forming, maintaining and/or retrieving selective associations between events. Therefore, the assessment of LI in Nurr1-deficient mice relative to WT mice was used to capture potential abnormalities in the domains of selective attention relevant to schizophrenia.

The first LI experiment used the CAA procedure according to methods established before (Meyer et al. 2008b). The apparatus consisted of four identical two-way shuttle boxes (model H10-11M-SC; Coulbourn Instruments, Allentown, PA, USA). The internal dimensions of each shuttle box, as measured from the raised grid floor, were 35.5 × 18 × 2.5 cm. The box was separated by an aluminum wall into two identical compartments, which were interconnected by an opening (6.5 × 8 cm) in the partition wall, thus allowing the animal to freely traverse from one compartment to the other. The grid floor was made of stainless steel rods, 4 mm in diameter and spaced 7 mm apart (center to center), and was connected to a constant current shock generator (model H10-1M-XX-SF; Coulbourn Instruments). Through the grid floor an electric shock (0.25 mA) could be delivered. The CS was an 80-dBA white noise delivered by a speaker placed on the floor of the cubicle and behind the shuttle box. Shuttle response was detected by a series of photocells (H20-95X; Coulbourn Instruments) mounted on the side of both shuttle compartments. The boxes were illuminated throughout the experimental session by two diffuse light sources (1.1 W), each mounted 19 cm above the grid floor on the sidewall on each of the two compartments.

The test procedures consisted of two phases: PE and conditioning, conducted 24 h apart. Animals from each of the two treatment groups (WTs and KOs) were allocated to one of two conditions: CS-PE and non-pre-exposure (NPE). In the PE phase, the PE subjects were placed in the test chambers and presented with 100 PEs to a 5-second noise CS according to a random interstimulus interval schedule of 40 ± 15 seconds. The NPE subjects were confined to the chamber for an equivalent period of time without any stimulus presentation. The total number of shuttle responses performed during the PE session was recorded.

On the conditioning day, the subjects were returned to the same shuttle boxes and received a total of 50 avoidance trials presented with an intertrial interval (ITI) of 40 ± 15 seconds. Each avoidance trial began with the onset of the white noise CS. If an animal made a shuttle into the adjacent compartment within 5 seconds of the CS onset, the CS would be terminated without any delivery of the shock-US and an avoidance response was scored. Avoidance failure led immediately to an electric foot-shock presented coinciding with the CS. This could last for a maximum of 2 seconds, but would be terminated by a shuttle response during this period, which would then be counted as an escape response. To index conditioned avoidance learning, the mean avoidance response per 10-trial block was submitted to statistical analysis.

Discrimination reversal learning in the water T-maze

In reversal learning, subjects first learn to respond differentially to two stimuli of opposing valence and are then confronted with the same two stimuli but with the reversed valence. The ability to recognize an unexpected consequence from previously established associative learning rule and then to switch the response contingency accordingly is crucial to reversal learning (Brown & Bowman 2002; Dalley et al. 2004). Because such behavioral/cognitive switching is impaired in patients with schizophrenia (Crider 1997), the reversal learning test used here was used to capture the potential impact of Nurr1 on schizophrenia-relevant abnormalities in the domains of behavioral/cognitive switching.

The apparatus consisted of a modified white circular tank measuring 1 m in diameter. It was filled daily with fresh water to a depth of 19 cm, with the temperature of the water maintained at ∼ 24°C. Opaque Plexiglass partitions were positioned vertically into the water to form three interconnected corridors (10 cm wide) in the shape of a ‘T’. The partition walls extended to a height of 18 cm above the water surface. The vertical start arm was 30 cm long and the two parallel choice arms were 45 cm long. They joined at the center of the water tank. A piece of cylindrical clear Plexiglass, 7 cm in diameter and 18.5 cm high, was positioned at the end of either the left or the right goal arm and served as the escape platform.

In acquisition training, the animals were required to learn to discriminate the left and right goal arms, with only one of them leading to an escape platform hidden at the far end. For one half of the animals in each group, the left arm was correct, and for the other half, the right arm was correct. The platform location remained unchanged until the end of acquisition. There were six trials per daily session, conducted at an ITI of 10 min. The first session was conducted on 2 testing days of three trials each, to allow the animals to acclimatize to the water. To begin a trial, the animals were placed at the beginning of the start arm and allowed up to 1 min to choose between the two goal arms. Once its entire body had entered into an arm, a guillotine door was lowered to prevent the animal from retracing. If the animal chose the correct arm and swam onto the platform, it was immediately removed from the maze and returned to a holding cage. When the incorrect arm was chosen, the animal was confined to the arm for 50 seconds and then removed from the maze to a holding cage. Acquisition training continued until an animal had reached criterion performance of 11 correct responses across two consecutive days (i.e. of 12 trials). Afterward, the platform location was moved to the other, previously incorrect, arm to assess reversal learning. Reversal training continued until an animal reached criterion performance once again. The percentage of correct arm choices as well as the errors to criterion performance was recorded for each animal during acquisition and reversal training.

Behavioral Testing: Cohort II

Latent inhibition in the CTA paradigm

We also assessed LI in a second paradigm, namely the CTA paradigm. Here, we chose the CTA procedure as this paradigm is arguably only minimally confounded by changes in locomotor behavior. This is relevant in as much as Nurr1-deficient mice display spontaneous hyperactivity (Rojas et al. 2007), which may undermine the interpretation of the data obtained in the first LI experiment using the active avoidance procedure (see above).

The apparatus consisted of Macrolon type II cages with modified grid tops, which allowed the efficient placement or removal of the drinking tubes (two per cage). The drinking tubes were made from 15 ml polypropylene test tubes (Cellstars; Greiner Bio-One, Frickenhausen, Germany) and equipped with an air-tight screwed top. An opening of 2.5 mm in diameter was made at the end of the tubes, thus allowing the animal access to the liquid contained therein without leakage. Two acrylic rings (20 mm in inner diameter) were mounted in-between the metal grids of the cage top to allow the efficient placement and removal of the drinking tubes, which could be fitted smoothly into the lumen of the rings and remained in place with the tube cap resting on top of the ring. When the two drinking tubes were in place, a distance of 40 mm separated the openings of the two drinking tubes, at a level of approximately 50 mm above the cage floor that was covered with sawdust. The placement was such that the animals could easily switch drinking from one tube to the other. Liquid consumption from a given drinking tube was calculated by taking the difference in its weight before and after a drinking session. The apparatus has been fully described elsewhere (Meyer et al. 2004).

The animals were kept singly in the modified cages throughout the experiment. One day after the animals were switched to single caging, they were gradually introduced to a water deprivation regime over a 6-day period to achieve 22.5 h water deprivation on the fifth day. On all subsequent days, the animals were allowed daily access to fluid in two 45-min periods, separated by a 6-h interval. During the first drinking period, the animals had free access to two tubes and the second period always consisted of water only, which was presented in regular water bottles.

The experimental procedures consisted of four stages: baseline, PE, conditioning and test.

  • Baseline: On day 1, the animals were allowed access to water during the two drinking periods. The allocation of the subjects to the two PE conditions was counterbalanced in each experimental group according to the animals' baseline performance.
  • PE: On day 2, animals allocated to the NPE group had access to water as described above. Animals in the PE group were given access to 10% sucrose solution in both drinking tubes during the first drinking period. During the second drinking period, all animals were allowed to drink from the water bottles.
  • Conditioning: On day 3, all animals were given access to 10% sucrose solution for 45 min in the first drinking period, followed by a lithium chloride (LiCl) injection 1 min later. LiCl was dissolved in 0.9% NaCl to achieve a concentration of 0.25 M. It was injected in a volume of 2% v/w body weight. The solution was freshly prepared on the conditioning day.
  • Test: On days 4–7, conditioned aversion to the sucrose taste was measured in a two-choice test with one drinking tube filled with 10% sucrose solution and the other with water. Conditioned aversion to the sucrose taste was indexed by sucrose consumption as a proportion (in per cent) of total liquid consumed during the first drinking period when the animals were confronted with a choice between sucrose and water.

The data collected on the four phases were analyzed separately.

Locomotor response to dizocilpine (MK-801) using a within-subject design

We evaluated the behavioral sensitivity to treatment with low doses of the non-competitive N-methyl-d-aspartate (NMDA) receptor antagonist dizocilpine (MK-801) by assessing the animals' locomotor responses to the drug challenge. MK-801 was chosen because patients with schizophrenia are known to display enhanced sensitivity to glutamatergic drug challenge (Krystal et al. 1994; Lahti et al. 2001).

The locomotor responses to acute systemic treatment with MK-801 were evaluated in identical open-field arenas described above. The animals were first habituated to the arenas for a period of 20 min while drug free. Next, the animals were briefly removed from the arena, injected with sterile 0.9% NaCl (saline) solution via the intraperitoneal (i.p.) route and allowed to explore the arena for another 20 min. Finally, the subjects were again briefly removed from the arena and injected with MK-801 solution. The subjects were then immediately returned to the same arenas again and the locomotor responses to the acute drug challenge were monitored for a period of 80 min. MK-801 (Merck, Sharp & Dohme, Hertfordshire, UK) was dissolved in isotonic 0.9% NaCl solution to achieve the desired concentration for injection. MK-801 was administered i.p. at a dose of 0.2 mg/kg. The selected dose MK-801 does not produce any ceiling effects on locomotor activity enhancement (Meyer et al. 2008c). This readily facilitates the assessment of modulatory effects of the genetic manipulation on MK-801-induced changes in locomotor activity. The volume of injection was 5 ml/kg for both solutions. All solutions were freshly prepared on the day of testing.

Behavioral testing: cohort III

Locomotor response to dizocilpine (MK-801) using a between-subject design

An additional experiment was performed to evaluate the animals' locomotor response to acute systemic treatment with MK-801 using a between-subject treatment design. This was performed in identical open-field arenas described above. All animals were first habituated to the arenas for a period of 1 h while drug free. Next, the animals were briefly removed from the arena and half of them were injected i.p. with sterile 0.9% NaCl (saline) solution and the other half with MK-801 solution. The subjects were then immediately returned to the same arenas again and the locomotor responses to the acute vehicle (saline) or drug (MK-801) challenge were monitored for a period of 2 h. MK-801 (Merck, Sharp & Dohme) was dissolved in isotonic 0.9% NaCl solution to achieve the desired concentration for injection as described above, and it was administered i.p. at a dose of 0.2 mg/kg. All solutions were freshly prepared on the day of testing and were administered using an injection volume of 5 ml/kg.

Statistical analysis

All data were analyzed using parametric analysis of variance (anova), followed by Fisher's least significant difference post hoc comparisons or restricted anovas whenever appropriate. Statistical significance was set at P < 0.05. All statistical analyses were performed using the statistical software StatView (version 5.0) implemented on a PC running the Windows XP operating system.

Results

Cohort I

Spontaneous locomotor activity in the open field

The total distance moved in the entire arena was expressed as a function of 5-min bins and analyzed using a 2 × 2 × 12 (sex × genotype × 5-min bins) repeated-measures anova. Animals from all groups displayed a clear habituation effect across the 1-h testing period (main effect of 5-min bins: F11,528 = 100.37; P < 0.001). The overall total distance moved was significantly increased in Nurr1 heterozygous KO relative to WT animals (main effect of genotype: F1,48 = 7.716; P < 0.01; Fig. 1). This effect was similarly present in male and female subjects.

Figure 1.

Spontaneous locomotor behavior assessed in the open field. The line plot shows distance moved (cm) as a function of 5-min bins, and the bar plot depicts the mean distance traveled (cm) per bin. Mice with a heterozygous constitutive Nurr1 deletion (KO) display a significant overall increase in distance moved compared to WT mice. **P < 0.01, reflecting the significant main effect of genotype. N(WT) = 26 (10 m, 16 f), N(KO) = 26 (12 m, 14 f). All values are means ± SEM. f, females; m, males.

Spatial recognition memory in the Y-maze

The critical measure of spatial recognition memory in the Y-maze test is the relative time spent in the novel (previously unexplored) arm during the choice phase of this test. The relative exploration time was analyzed using a 2 × 2 × 4 (sex × genotype × delays) anova of the per cent time spent in the novel arm. This analysis showed no significant main effect or interactions involving the genotype variable, suggesting that heterozygous Nurr1 deficiency did not affect spatial recognition memory in the Y-maze. Indeed, Fig. 2 shows that both Nurr1 heterozygous KO and WT animals displayed a clear and highly comparable preference for the novel arm at all retention intervals tested. anova only showed a main effect of delays (F3,141 = 3.67, P < 0.05), reflecting differential overall performance as a function of delays (Fig. 2).

Figure 2.

Spatial recognition memory in the Y-maze. The graph shows the per cent time spent in the novel (previously unexplored) arm during the choice phases of a 1-min, 15-min, 30-min and 24-h retention interval test. Mice with a heterozygous constitutive Nurr1 deletion (KO) and WT mice similarly performed above chance level (indicated by the dashed line) at all retention intervals tested. N(WT) = 26 (10 m, 16 f), N(KO) = 25 (11 m, 14 f). All values are means ± SEM. f, females; m, males.

PPI of the acoustic startle reflex

Per cent PPI (%PPI) was analyzed using a 2 × 2 × 3 × 3 (sex × genotype × pulse level × prepulse level) anova. Heterozygous Nurr1 deletion induced a sex-dependent effect on %PPI (interaction between sex and genotype: F1,48 = 4.59, P < 0.05), with male but not female subjects being affected. More specifically, male Nurr1 heterozygous KO animals displayed a significant overall reduction in %PPI relative to male WT mice (Fig. 3a), as supported by the main effect of genotype (F1,20 = 8.648, P < 0.01) in the anova restricted to males. In contrast, %PPI levels did not significantly differ between female Nurr1 heterozygous KO and WT mice (F1,28 = 0.895, P = 0.35; Fig. 3b).

Figure 3.

PPI, startle reactivity and prepulse-elicited reactivity in male mice with a heterozygous constitutive Nurr1 deletion (KO) and WT mice. Line plots on the left show %PPI as a function of three different pulse levels (P-100, P-110 and P-120, which correspond to pulse stimulus intensities of 100, 110 and 120 dBA). The graphs also show mean %PPI across all prepulse and pulse levels tested, as well as startle reactivities to the three different pulses (P-100, P-110 and P-120). The line plots on the right show reactivity to prepulse-alone trials (0, +6, +12 and +18 dBA above background of 65 dBA). (a) Male Nurr1 heterozygous KO mice display a significant overall reduction in %PPI compared to male WT mice. **P < 0.01, reflecting the significant main effect of genotype in males. In addition, male KO mice show a significant reduction in the reactivity to highest prepulse stimulus (+18 dBA above background of 65 dBA). Male KO and WT mice do not differ in startle reactivity to pulse-alone trials. N(WT) = 10, N(KO) = 12. (b) Female Nurr1 heterozygous KO mice do not significantly differ from female WT mice in the measures of %PPI, startle reactivity to pulse-alone trials and reactivity to prepulse-alone trials. N(WT) = 16, N(KO) = 14. All values are means ± SEM.

The Nurr1 genotype did not significantly alter startle reactivity to pulse-alone trials, neither in males nor in females (Fig. 3a,b). The 2 × 2 × 3(sex × genotype × pulse level) anova of startle reactivity did not yield significant effects involving the between-subjects factor of genotype. However, heterozygous Nurr1 deletion significantly reduced the reaction to prepulse-alone trials, and this effect seemed most pronounced at the highest prepulse level. A 2 × 2 × 4 (sex × genotype × prepulse level) anova of reactivity to prepulse-alone trials (including background reactivity) showed a significant main effect of genotype (F1,53 = 3.97, P = 0.05) and its interaction with prepulse levels (F3,159 = 4.71, P < 0.01). In keeping with the sex-specific effects of Nurr1 deficiency on %PPI (Fig. 3a), we further analyzed reactivity to prepulse-alone trials separately for males and females. The 2 × 4 (genotype × prepulse level) anova of prepulse reactivity restricted to the males showed a main effect of genotype (F1,25 = 6.54; P < 0.05), and a significant prepulse level × genotype interaction (F3,75 = 4.75; P < 0.01). Subsequent post hoc analyses confirmed that male Nurr1 heterozygous KO animals displayed a significant (P < 0.01) reduction in prepulse reactivity at the highest prepulse level relative to male WT mice (Fig. 3a). The analyses restricted to females did not show any significant main effects or interactions involving the between-subject factor of genotype, suggesting that reactivity to prepulse-alone trials was not significantly altered in female Nurr1-deficient mice relative to female WT mice (Fig. 3b).

Social interaction and memory

In the social interaction and memory test, exploration time was expressed as a function of 1-min bins and analyzed using 2 × 2 × 2 × 5(sex× genotype × object × bins) anova. Exploration time was separately analyzed for the initial ‘dummy phase’, in which the relative exploration time between a congenic mouse and an inanimate dummy object was assessed, and for the subsequent ‘test phase’, in which the relative exploration time between a novel congenic mouse and a familiar congenic mouse was evaluated.

During the ‘dummy phase’ of the test, both Nurr1 heterozygous KO and WT mice showed a clear preference for the congenic mouse relative to the inanimate dummy (Fig. 4a), as supported by the highly significant main effect of object (F1,48 = 30.207; P < 0.001). No main effects or interactions involving genotype attained statistical significance, indicating intact social interaction in Nurr1-deficient animals. Likewise, both Nurr1 heterozygous KO and WT mice animals spent significantly more time exploring the novel mouse than the familiar mouse (Fig. 4b), leading to a significant main effect of object (F1,48 = 4.169; P < 0.05) during the ‘test phase’. Again, no main effects or interactions involving genotype reached statistical significance, suggesting that heterozygous Nurr1 deletion did not affect social memory.

Figure 4.

Social interaction and recognition in mice with a heterozygous constitutive Nurr1 deletion (KO) and WT mice. (a) The bar plot depicts the relative exploration time between an unfamiliar congenic mouse (‘mouse’) and an inanimate dummy object (‘dummy’). Both KO and WT mice show a significant preference for the mouse relative to the dummy object. ***P < 0.001, signifying the significant main effect of object. (b) The bar plot shows the relative exploration time between a novel and a familiar congenic mouse. Both KO and WT mice show a significant preference for the novel mouse relative to the familiar mouse. *P < 0.05, signifying the significant main effect of object. N(WT) = 26 (10 m, 16 f), N(KO) = 26 (12 m, 14 f). All values are means ± SEM. f, females; m, males.

LI in the two-way active avoidance paradigm

CAA responses were expressed as a function of 10-trial blocks and analyzed using a 2 × 2 × 2 × 5 (sex × genotype × PE × 10-trial blocks) repeated-measures anova. Acquisition of conditioned avoidance responding was apparent in all experimental groups by the increase in avoidance responses performed on successive CS–US trials. This led to a significant main effect of 10-trial blocks (F4,176 = 95.13, P < 0.001). A significant LI effect (reflected by increased avoidance responding to the NPE subjects relative to the PE subjects) was evidenced by the presence of a significant main effect of PE (F1,44 = 4.72, P < 0.05). Notably, anova did not show any significant main effects or interactions involving the genotype variable, suggesting that the Nurr1 genotype did not significantly modulate LI in the active avoidance paradigm (Fig. 5). The heterozygous Nurr1 deletion also did not induce a general deterioration in instrumental conditioning as such because a 2 × 2 × 5 (genotype × sex × 10-trial blocks) repeated-measures anova of conditioned avoidance responses restricted to NPE animals showed no significant main effect or interactions involving genotype. Furthermore, general activity levels (as indexed by the analysis of spontaneous ITI shuttles) during conditioning phase were not significantly different between Nurr1 heterozygous KO and WT mice (data not shown).

Figure 5.

LI measured in the conditioned active avoidance paradigm. (a) The line plots show the numbers of avoidance shuttles performed by NPE or PE WT and heterozygous Nurr1-deficient (KO) animals. (b) The bar plot depicts the mean avoidance shuttles performed by NPE or PE subjects. A significant LI effect (i.e. a reduction in conditioned avoidance responding to PE relative to NPE subjects) was present in both WT and KO animals. *P < 0.05, representing the significant main effect of PE. N(WT − NPE) = 12 (4 m, 8 f), N(WT-PE) = 14 (6 m, 8 f), N(KO-NPE) = 12 (5 m, 7 f), N(KO-PE) = 14 (7 m, 7 f). All values are means ± SEM. f, females; m, males.

Discrimination reversal learning in the water T-maze

  • Acquisition phase: Acquisition of the left–right discrimination task was analyzed by a 2×2×5 (sex × genotype × days) repeated-measure anova of per cent correct trials to criterion. As shown in Fig. 6, animals from both genotype conditions similarly improved the performance as a function of test days, as supported by the presence of a highly significant main effect of days (F4,188 = 43.73, P < 0.001). Notably, Nurr1 deficiency did not significantly influence the acquisition of the discrimination task; anova did not show any significant main effect or interaction involving the between-subject factor of genotype. This impression was also confirmed by the 2 × 2 (sex × genotype) anova of the total number of errors to criterion, which did not show any significant outcomes (Fig. 6).
  • Reversal phase: Reversal of the discrimination task was analyzed by a 2×2×9 (sex × genotype × days) repeated-measures anova of per cent correct trials to criterion. Again, animals from both genotype conditions similarly improved the performance as a function of test days. anova of per cent correct trials showed a highly significant main effect of days (F8,376 = 190.82, P < 0.001), but did not show a significant main effect or interaction involving the genotype variable. Likewise, the 2×2 (sex × genotype) anova of the total number of errors to criterion failed to show any statistical significances. Hence, reversal learning of left–right discrimination was not significantly affected by heterozygous Nurr1 deletion.
Figure 6.

Left–right discrimination reversal learning in the water T-maze. (a) The line plot shows the performance expressed as the percentage correct responses per day. Performance during both the initial acquisition and the subsequent reversal phase was highly similar between mice with a heterozygous constitutive Nurr1 deletion (KO) and WT mice. A clear reversal effect was evident in both groups as suggested by the below-chance performance (represented by the dashed line) on the first 2 days of reversal training. (b) The bar plot displays the total errors made during the initial acquisition and subsequent reversal learning phase. N(WT) = 25 (10 m, 15 f), N(KO) = 26 (12 m, 14 f). All values are means ± SEM. f, females; m, males.

Cohort II

LI in the CTA paradigm

  • Baseline phase: The amount of liquid consumption (water for all subjects) was subjected to a 2×2 (sex × genotype) anova, which showed a main effect of genotype (F1,60 = 11.22, P < 0.01). This reflected the significant increase in water consumption displayed by Nurr1 heterozygous KO mice relative to WT mice. The mean ± SEM of water consumption in Nurr1 heterozygous KO and WT animals was 1.43 ± 0.04 and 1.19 ± 0.05 ml, respectively.
  • PE phase: The amount of liquid consumption (sucrose solution for PE subjects and water for NPE subjects) was subjected to a 2 × 2 × 2 (sex × genotype × PE) anova of total liquid consumption. Again, Nurr1-deficient mice drank significantly more than WT animals, as shown by the main effect of genotype (F1,56 = 8.58, P < 0.01). The mean ± SEM of fluid consumption in Nurr1 heterozygous KO and WT animals was 1.52±0.07 and 1.23±0.07 ml, respectively. Liquid consumption was significantly higher in NPE subjects compared to PE subjects (main effect of PE: (F1,56 = 4.84, P < 0.05). The mean ± SEM of fluid consumption in NPE and PE subjects was 1.48 ± 0.08 and 1.26 ± 0.07 ml, respectively.
  • Conditioning phase: The amount of sucrose intake on the conditioning session was subjected to a 2 × 2 × 2 (sex × genotype × PE) anova. No main effects or interactions attained statistical significance.
  • Test phase: CTA was indexed by sucrose consumption as a proportion (in per cent) of total liquid consumed when the animals were confronted with a choice between sucrose solution and water and was analyzed using a 2 × 2 × 2 × 4 (sex × genotype × PE × test days) repeated-measures anova. Per cent sucrose consumption increased as a function of successive test days, leading to a significant main effect of test days (F3,162 = 11.05, P < 0.001). A significant LI effect (reflected by reduced per cent sucrose consumption in NPE subjects relative to PE subjects) was evidenced by the presence of a significant main effect of PE (F1,54 = 11.53, P < 0.001). This LI effect was noticeable in both Nurr1 heterozygous KO and WT animals (Fig. 7), and anova did not show significant interactions between PE, genotype, sex and/or days. The heterozygous Nurr1 downregulation also did not significantly affect the expression of the CR as such because a 2 × 2 × 2 × 4 (sex × genotype × PE × test days) repeated-measure anova of percentage sucrose consumption restricted to the NPE animals showed no significant main effects or interactions involving the factor of genotype.
Figure 7.

LI measured in the conditioned taste aversion paradigm. (a) The line plots show the per cent sucrose solution intake (%) in NPE or PE WT and heterozygous Nurr1-deficient (KO) animals as a function of four consecutive test days. (b) The bar plot depicts the mean per cent sucrose intake in NPE or PE subjects. A significant LI effect (i.e. reduced per cent sucrose intake in NPE relative to PE subjects) was present in both WT and KO animals. ***P < 0.001, representing the significant main effect of PE. N(WT-NPE) = 15 (7 m, 8 f), N(WT-PE) = 16 (8 m, 8 f), N(KO-NPE) = 15 (7 m, 8 f), N(KO-PE) = 16 (9 m, 7 f). All values are means ± SEM. f, females; m, males.

Sensitivity to MK-801-induced hyperlocomotor activity using a within-subjects design

  • Habituation phase: Total distance moved during the initial habituation phase was expressed as a function of 5-min bins and analyzed using a 2×2×4 (sex × genotype × 5-min bins) repeated-measure anova. Consistent with our previous findings (see Fig. 1), Nurr1 heterozygous KO displayed a significant increase in the total distance moved compared to WT animals (main effect of genotype: F1,27 = 11.91, P < 0.01). Animals from both genotype groups displayed a clear habituation effect across the 20-min habituation period (Fig. 8), as supported by the significant main effect of bins (F3,81 = 51.75; P < 0.001).
  • Saline phase: Subsequent to the habituation phase, animals were injected with saline solution, and total distance moved was again recorded for a 20-min period and analyzed using a 2 × 2 × 4 (sex × genotype × 5-min bins) repeated-measure anova. This analysis only yielded a significant main effect of bins (F3,81 = 3.91, P < 0.05).
  • MK-801 phase: Total distance moved during the 80-min MK-801 phase was also expressed as a function of 5-min bins and analyzed using a 2 × 2 × 16 (sex × genotype × 5-min bins) repeated-measure anova. Following injection of MK-801, locomotor activity generally increased as a function of bins and reached peak levels between 50 and 80 min postinjection. Most interestingly, the locomotor-enhancing effect of systemic MK-801 treatment was significantly enhanced in Nurr1 heterozygous KO mice relative to WT mice, as supported by the presence of a significant main effect of genotype (F1,27 = 4.24, P < 0.05) and its interaction with bins (F15,405 = 1.90, P < 0.05). The main effect of bins also attained statistical significance (F15,405 = 26.86, P < 0.001).
Figure 8.

Locomotor reaction to systemic treatment with dizocilpine (MK-801) using a within-subject treatment design. The line plot shows locomotor activity in the open field expressed as distance traveled (cm) per 5-min bin during the initial habituation phase, following saline (SAL) administration, and following acute MK-801 (0.2 mg/kg, i.p.) treatment. The bar plot illustrates the overall means across the MK-801 period. Mice with a heterozygous constitutive Nurr1 deletion (KO) displayed a significant overall increase in the distance moved following MK-801 treatment compared to drug-challenged WT mice. *P < 0.05, representing the significant main effect of genotype during the MK-801 phase. N(WT) = 15 (7 m, 8 f), N(KO) = 16 (9 m, 7 f). All values are means ± SEM. f, females; m, males.

Cohort III

Sensitivity to MK-801-induced hyperlocomotor activity using a between-subject design

  • Habituation phase: Total distance moved during the initial habituation phase was expressed as a function of 5-min bins and analyzed using a 2 × 12 (genotype × 5-min bins) repeated-measure anova. In line with our previous findings (see Fig. 1), Nurr1 heterozygous KO showed a significant increase in the total distance moved compared to WT animals (main effect of genotype: F1,19 = 4.79, P < 0.05) during the drug-free acclimatization phase. Animals from both genotype groups displayed a clear habituation effect across the 1-h habituation period (Fig. 9), as supported by the significant main effect of bins (F11,209 = 43.45; P < 0.001).
  • Drug treatment phase: Subsequent to the habituation phase, animals were either injected with saline or MK-801 solution, and total distance moved was recorded for a 2-h period and analyzed using a 2 × 2 × 24 (genotype × drug treatment × 5-min bins) repeated-measure anova. This analysis showed significant main effects of genotype (F1,17 = 7.48; P < 0.05) and drug treatment (F1,17 = 15.18; P < 0.01), as well as a significant genotype × drug treatment interaction (F1,17 = 7.12; P < 0.05). The three-way interaction between genotype, drug treatment and bins also attained statistical significance (F23,391 = 2.75; P < 0.001). Subsequent post hoc group comparisons of the mean distance moved confirmed that Nurr1-deficient animals treated with MK-801 solution displayed a significant increase in locomotor activity compared to WT animals injected with MK-801 (P < 0.001) and compared to saline-treated Nurr1-deficient or WT mice (both P < 0.001) (Fig. 9).
Figure 9.

Locomotor reaction to systemic treatment with dizocilpine (MK-801) using a between-subject treatment design. The line plot shows locomotor activity in the open field expressed as distance traveled (cm) per 5-min bin during the initial habituation phase (bins 1–12) and following saline (SAL) or MK-801 (0.2 mg/kg, i.p.) administration (bins 13–36). The bar plot illustrates the overall means across the 2-h period following saline or MK-801 treatment. Mice with a heterozygous constitutive Nurr1 deletion (KO) displayed a significant overall increase in the distance moved during the initial habituation phase compared to WT animals. *P < 0.05, representing the significant main effect of genotype during the initial habituation phase. In the subsequent drug treatment phase, Nurr1-deficient animals treated with MK-801 solution displayed a significant increase in locomotor activity compared to WT animals injected with MK-801, and compared to saline-treated Nurr1-deficient or WT mice. ***P < 0.001, based on post hoc group comparisons. N(WT-SAL) = five males, N(WT-MK-801) = five males, N(KO-SAL) = five males, N(KO-MK-801) = six males. All values are means ± SEM.

Discussion

This study used a comprehensive test battery to evaluate whether heterozygous Nurr1-deficient mice may display multiple behavioral, cognitive and pharmacological phenotypes relevant to schizophrenia and related psychotic disturbances. Our study confirmed the presence of increased spontaneous locomotor activity in Nurr1-deficient mice relative to WT mice (Bäckman et al. 2003; Eells et al. 2002; Rojas et al. 2007), suggesting that spontaneous hyperactivity is a robust phenotype of constitutive heterozygous Nurr1 deficiency. In addition, we found that Nurr1 heterozygous KO animals displayed a significant potentiation of the locomotor-enhancing effects of acute treatment with the non-competitive NMDA receptor antagonist MK-801. This novel phenotype of Nurr1-deficient mice was confirmed using both within-subject and between-subject treatment designs and corroborates the findings of enhanced behavioral sensitivity of Nurr1 heterozygous KO mice to the NMDA receptor antagonist phencyclidine (PCP) (Rojas et al. 2007). It thus appears that the presence of marked locomotor hyperactivity following pharmacologically induced NMDA receptor antagonism is another robust phenotype associated with constitutive heterozygous Nurr1 deficiency.

Here, we also report a novel phenotype of Nurr1-deficient mice, namely impairment in PPI of the acoustic startle reflex. This phenotype was found to be clearly sex-specific as it only emerged in male but not female Nurr1 heterozygous KO mice. Interestingly, although Nurr1-deficient animals did not significantly differ in startle reactivity to pulse-alone trials compared to WT subjects, they showed a significant sex-dependent reduction in prepulse-elicited reactivity (PPER), with male but not female animals being affected. The effects of Nurr1 deficiency on PPI and PPER resemble the concomitant effects of acute PCP or MK-801 treatment on these measures in mice in as much as such pharmacological blockade of NMDA receptors is known to disrupt PPI and, at the same time, weaken the reactivity to the prepulse stimuli relative to the baseline no-stimulus conditions (Yee et al. 2004). It is therefore intriguing to raise the possibility that the sensorimotor gating deficits seen in male Nurr1-deficient mice may have, at least in part, a glutamatergic component and that such PPI impairment may arise via a failure to detect the prepulse stimulus (Yee & Feldon 2009). However, our data here do not directly explore the neuronal mechanisms underlying the PPI deficits in Nurr1-deficient mice nor do they fully support the possibility that this deficit may be linked to reduced detection of the prepulse stimulus. Indeed, if the PPI deficits in Nurr1-deficient mice were primarily attributable to a failure to detect the prepulse, one would expect PPI deficiency to correlate with reductions in PPER. This was notably not the case: while the PPI deficit in male Nurr1-deficient mice emerged largely independently of the precise prepulse levels used, a significant deficit in PPER was only detected at the highest prepulse level (+18 dB above a background of 65 dB).

Although the effects of constitutive heterozygous Nurr1 deletion on PPI impairment appear relatively strong in magnitude (∼30% reduction in mean PPI in KO relative to WT mice), our findings are in contrast with previous attempts that failed to show significant PPI deficiency in (group-raised) Nurr1-deficient mice (Eells et al. 2006). Several parameters could explain the discrepancies between our findings here and the previous report by Eells et al. (2006), considering that various biological factors, such as age, sex and circadian cycle have been reported as modulating variables in the expression of PPI (Geyer & Swerdlow 1998; Swerdlow et al. 2001). First, in the study by Eells et al. (2006), the animals were tested ∼3 months after weaning (i.e. at ∼PND 110), which is in contrast to our study, in which KO and WT underwent PPI testing at ∼PND 75. Second, our animals were subjected to behavioral testing exclusively in the dark phase of the light/dark cycle, whereas Eells et al. (2006) investigated the behavioral effects of heterozygous Nurr1 deletion during the light phase. Finally, it is not clear from the data reported by Eells et al. (2006) whether their cohort of Nurr1-deficient mice displayed reduced reactivity to prepulse-alone trials, thereby facilitating PPI attenuation via failure to detect the prepulse stimulus.

The constellation of the behavioral and pharmacological abnormalities identified here suggests that Nurr1-deficient mice display some phenotypic characteristics reminiscent of schizophrenia-like dysfunctions (Rojas et al. 2007). Indeed, deficits in PPI have been consistently reported in acute and chronic schizophrenia patients, as well as in psychosis-prone individuals (Braff et al. 1992, 2001; Swerdlow et al. 1998). Interestingly, there is also recent evidence for impaired PPER in unmedicated schizophrenic patients (Csomor et al. 2009), similar to what we report here in Nurr1 heterozygous KO animals. Furthermore, in analogy to the presence of hyperactive locomotor behavior measured in the open-field test, a subset of schizophrenic patients exhibits psychomotor agitation that includes hyperactivity and/or increased stereotypic movements (Powell & Miyakawa 2006; but see also Perry et al. 2009). Patients with schizophrenia also display enhanced behavioral and cognitive sensitivity to acute treatment with NMDA receptor antagonist such as ketamine (Lahti et al. 1995, 2001), which fits with our findings of increased locomotor responsiveness to acute MK-801 treatment in Nurr1-deficient mice. The behavioral hypersensitivity of Nurr1-deficient mice to systemic MK-801 administration may further suggest that these mice may display low-tonic NMDA-mediated signaling (Mohn et al. 1999). Together with previous reports (Rojas et al. 2007), our results thus support the face validity of the heterozygous Nurr1-deficient mouse model for specific symptomatological features of schizophrenia and related psychotic disease.

Despite this, we would like to emphasize that constitutive heterozygous Nurr1 deletion is clearly associated with a restricted schizophrenia-related phenotype. This is because neither male nor female Nurr1-deficient mice displayed overt abnormalities in other cardinal behavioral and cognitive functions known to be impaired in schizophrenic disease (Tamminga & Holcomb 2005; Tandon et al. 2009), including social interaction and recognition, spatial recognition memory (as assessed in a spatial Y-maze test), selective attention (as assessed by the LI paradigm) or discrimination reversal learning (as assessed in a left–right discrimination test in the water T-maze). The use of a comprehensive test battery has thus allowed us to show an appreciable specificity of constitutive Nurr1 deficiency in the development of abnormal brain functions relevant to schizophrenia. In total, our findings suggest that heterozygous deletion of Nurr1 selectively modulates spontaneous and glutamatergic drug-induced locomotor behavior as well as sensorimotor gating functions, while sparing numerous other behavioral and cognitive functions relevant to schizophrenic disease.

We believe that the presence of a restricted schizophrenia-related phenotype in Nurr1-deficient mice does not readily undermine the validity of this genetic model in the context of basic schizophrenic research; rather, it may be useful for experimental studies on specific aspects of schizophrenia symptomatology which are related to locomotor behavior, psychotomimetic drug sensitivity and sensorimotor gating. However, it is important to emphasize that in addition to schizophrenia, Nurr1 may also have relevance to other neurological and/or neuropsychiatric disorders, especially Parkinson's disease (PD) (Jankovic et al. 2005) and depression-related dysfunctions (Rojas et al. 2010). A variety of experimental studies in mice with constitutive or conditional Nurr1 knockdown show that Nurr1 deficiency increases the vulnerability of the midbrain dopaminergic neurons to degenerate, which is a hallmark pathological feature in PD (Jankovic et al. 2005; Jiang et al. 2005; Kadkhodaei et al. 2009). Mutations in human NR4A2 (the human analogue of the murine Nurr1 gene) and a decrease in Nurr1-positive neurons of the substantia nigra have also been shown in some (rare) PD cases (Chu et al. 2006; Grimes et al. 2006; Huang et al. 2004). It remains controversial, however, whether PD patients display impaired sensorimotor gating in the form of PPI disruption (Leng et al. 2004; Perriol et al. 2005; Valls-Soléet al. 2004), and the administration of selective NMDA receptor antagonists such as MK-0657 do not appear to exert a significant modulatory impact on motor functions in PD patients (Addy et al. 2009). Given this, we suggest that the PPI impairment and potentiation of MK-801 sensitivity identified in Nurr1-deficient mice may be more closely related to schizophrenia-like than PD-like dysfunctions. This impression is further supported by the findings that the potentiation of psychotomimetic drug sensitivity in Nurr1-deficient mice can be normalized by the typical antipsychotic drug haloperidol (Rojas et al. 2007), supporting the predictive validity of the Nurr1 deficiency model for human psychotic disease. A role of (transient) developmental Nurr1 deficiency has also been described in two neurodevelopmental animal models of schizophrenia, namely the prenatal immune activation and developmental vitamin D deficiency models (Cui et al. 2010; Meyer et al. 2008a; Vuillermot et al. 2010). These models are known to mimic a broad spectrum of behavioral, pharmacological and neuroanatomical features of schizophrenia, and they encompass intrinsic etiological significance and predictive validity for the disorder (Eyles et al. 2009; Meyer & Feldon 2009, 2010). Hence, there is converging evidence from various lines of experimental research and parallel investigations in humans supporting a role of Nurr1 in the development of schizophrenia-relevant brain dysfunctions (Bäckman et al. 2003; Buervenich et al. 2000; Cui et al. 2010; Eells et al. 2006; Meyer et al. 2008a; Vuillermot et al. 2010; Xing et al. 2006).

In conclusion, our study highlights that constitutive heterozygous deletion of Nurr1 selectively modulates spontaneous and glutamatergic drug-induced locomotor behavior as well as sensorimotor gating functions, while sparing numerous other behavioral and cognitive functions relevant to schizophrenic disease. In view of the circuital role of Nurr1 in early development and maintenance of the midbrain dopamine system (Kadkhodaei et al. 2009; Saucedo-Cardenas et al. 1998; Zetterström et al. 1996), the Nurr1 deficiency model may be particularly useful to elucidate the relative contribution of early dopaminergic dysfunctions to the development of distinct symptomatological features implicated in schizophrenia and related psychotic disorders.

Acknowledgments

This study was supported by the Swiss National Science Foundation (grant 3100AO-100309 and grant 3100AO-116719) and ETH Zurich (grant 11 07/03). J.F. received additional support from a 2009 NARSAD Distinguished Investigator Award. All authors declare that they have no conflicts of interest to disclose.

Ancillary