The Effect of Neuregulin 1 on PCP Induced Alterations of Locomotion, Neurotransmission and Cell Growth As Relevant for Schizophr

University of Wollongong Research Online

University of Wollongong Thesis Collection University of Wollongong Thesis Collections

2014 The effect of Neuregulin 1 on PCP induced alterations of locomotion, neurotransmission and cell growth as relevant for schizophrenia Martin Engel University of Wollongong

Recommended Citation Engel, Martin, The effect of Neuregulin 1 on PCP induced alterations of locomotion, neurotransmission and cell growth as relevant for schizophrenia, Doctor of Philosophy thesis, School of Medicine, University of Wollongong, 2014. http://ro.uow.edu.au/theses/4395

Research Online is the open access institutional repository for the University of Wollongong. For further information contact the UOW Library: [email protected]

The effect of Neuregulin 1 on PCP induced alterations of locomotion, neurotransmission and cell growth as relevant for schizophrenia

Martin Engel

This thesis is presented as part of the requirement for the Award of the Degree

Doctor of Philosophy

from the University of Wollongong School of Medicine

July, 2014

CERTIFICATION

I, Martin Engel, declare that this thesis, submitted in partial fulfilment of the re- quirements for the award of Doctor of Philosophy, in the School of Medicine, Uni- versity of Wollongong, is wholly my own work unless otherwise referenced or acknowledged. The document has not been submitted for qualifications at any other academic institution.

Martin Engel 23rd of July 2014

ABSTRACT

Schizophrenia is a serious psychiatric disorder causing a wide range of devastating impairments. A combination of several genetic and environmental factors is considered to result in a range of neurobiological changes underlying the symptom spectrum of schizophrenia. Differences in the γ-aminobutyric acid (GABA)ergic, glutamatergic and dopaminergic neurotransmitter systems have been most consistently observed in patients with schizophrenia. Current pharmacological treatment options for patients with schizophrenia are, however, still limited in their ability to address the range of symptoms and identified neurobiological differences. Substantial evidence from human studies has linked the neurotrophic factor neuregulin 1 (NRG1) to the pathophysiology of schizophrenia, with further support from animal studies. Acute NRG1 signalling has been shown to affect GABAergic and glutamatergic neurotransmission in vitro in the prefrontal cortex (PFC) and hippocampus, two regions strongly linked to the symptomatology of schizophrenia. Due to the acute signalling properties of NRG1 in attenuating schizophrenia-relevant neurotransmission, a therapeutic potential of the NRG1 signalling system for the treatment of schizophrenia has been hypothesized, but never explored. In the present PhD thesis I assessed the possible antipsychotic-like characteristics of NRG1 treatment on schizophrenia-relevant behavioural and molecular impairments focusing on GABAergic and glutamatergic signalling. I modelled schizophrenia-like impairments in adult male mice by acute phencyclidine (PCP). Acute PCP treatment in mice has been shown to induce a wide spectrum of schizophrenia-relevant symptoms, including hyperlocomotion, impaired sensorimotor gating and neurotransmission impairments. This treatment paradigm is therefore widely used as a first step in antipsychotic drug development studies. In my first study, I examined the effects of concurrent NRG1 and PCP treatment on locomotor activity and sensorimotor gating. Central and peripheral NRG1 administration dose-dependently prevented PCP-induced hyperlocomotion and prepulse inhibition impairment. Advancing on these findings, I then studied the effects of NRG1 application on acute neurotransmission in the presence of PCP by microdialysis in the PFC and hippocampus of freely moving mice. In the PFC, NRG1 reduced glutamate and glycine levels, while combined NRG1 and PCP treatment triggered an increase in extracellular glycine. In the hippocampus, PCP treatment triggered a sig-

III

nificant increase in GABA levels without affecting glutamate or glycine. The GABA increase was not observed following combined NRG1 and PCP treatment, suggesting that NRG1 prevented the PCP effect. An increase in glycine was again observed following combined treatment, here accompanied by a reduction in glutamate levels. In my final study, I treated cells from primary frontal cortex cultures with NRG1 in the presence of PCP, measuring components of the GABAergic system relevant to the pathophysiology of schizophrenia. PCP changed the mRNA expression of GABA producing enzymes and the calcium binding protein parvalbumin. Combined treatment with NRG1 dose-dependently prevented these PCP-induced expression changes in the frontal cortex cells. Prevention of PCP-induced hyperlocomotion and PPI impairments demonstrates a clear involvement of the NRG1 signalling pathway in neuronal circuits relevant for spontaneous behaviour. The region specific influence of NRG1 on glutamate and glycine levels confirms the involvement of NRG1-ERBB signalling in adult excitatory neurotransmission. Furthermore, preventing the PCP-induced GABA increase in the hippocampus provides support for a functional role in inhibitory neurotransmitter signalling. An involvement of NRG1 in the regulation of inhibitory neurotransmission was further supported by in vitro frontal cortex experiments where NRG1 treatment prevented changes to GABA producing enzymes. The combined outcomes of my experiments provide initial support for potential antipsychotic-like characteristics of NRG1 treatment in a mouse model with induced schizophrenia-relevant behavioural and molecular impairments. The observed efficacy of peripherally administered NRG1 to prevent PCP-induced behavioural deficits indicates its potential as a new therapy for schizophrenia patients. Exploring the consequences of long-term NRG1 treatment on a broader spectrum of behavioural performance and molecular changes, particularly of the dopaminergic system, will be required to advance on the present findings and take the next step towards a potential suitability of NRG1 treatment in a clinical setting.

ACKNOWLEDGEMENTS

I would like to acknowledge and thank all the people who have taken part, assisted and supported me throughout the journey of my PhD. First and foremost, I would like to thank my supervisors Elisabeth Frank and Xu-Feng Huang, who have made this PhD project possible. For their support me despite the unexpected start, through many turns and changes, from early Monday mornings till late Sunday nights, from Amsterdam to Wollongong, from a skype call in Cardiff to presentations on three continents, who supported the idea before the first data point was anywhere in sight. My collaborators Tim Karl, Andrew Jenner and Lezanne Ooi, who have joined along the way from different groups and institutes: thank you for your constant support and giving me the opportunity to learn new methods and strategies for scientific investigations. Vincenzo Crunelli, Angelique Hoolahan and Simon Kaja, for the valuable scientific discussions and helpful advice, despite having no stakes in the project, but only wanting to see me progress. Thank you to the owners of the many hands that have helped me day in and day out, particularly Tracy, Maria, Zhixiang, the CTN and IHMRI team, for providing and maintaining an atmosphere I was happy to return to, even when I left it less than 12 hours ago. Thank you to my fellow PhD students who made surviving the endless hours in the lab bearable, with great company and hot chocolate, especially Peta, Na- talie, Jess, Colin and Amy. And finally, to all my family and friends, especially Louise, for tolerating my spend- ing so many late nights and weekends in the lab. My sincere gratitude for your good humour, support and acceptance while I completed this project!

TABLE OF CONTENTS

CERTIFICATION ...... II ABSTRACT ...... III ACKNOWLEDGEMENTS ...... V TABLE OF CONTENTS ...... VI LIST OF FIGURES ...... X LIST OF TABLES ...... XII List of abbreviations ...... XIII 1 Introduction ...... 1 1.1 Schizophrenia ...... 1 1.1.1 The symptoms of schizophrenia ...... 1 1.1.2 The neurobiology of schizophrenia ...... 2 1.1.3 Current Pharmacological Treatment Situation of Schizophrenia ...... 9 1.1.4 Current attempts in targeting GABA for treating schizophrenia ...... 12 1.2 Neuregulin 1 and ERBB4 signalling ...... 13 1.2.1 ERBB4 ...... 13 1.2.2 Neuregulin 1 ...... 14 1.2.3 Acute Neuregulin 1-ERBB4 signalling ...... 16 1.2.4 Schizophrenia and the NRG1-ERBB4 pathway ...... 21 1.3 General aim ...... 23 1.4 Pharmacological animal model strategy for antipsychotic drug development ...... 23 1.4.1 Chronic vs acute rodent models ...... 23 1.4.2 The acute PCP schizophrenia rodent model...... 24 1.5 Summary and aims ...... 29 2 The effect of NRG1 on PCP-induced behavioural Impairments ...... 30 2.1 Introduction ...... 30 2.2 Materials and Methods ...... 32 2.2.1 Subjects ...... 32 2.2.2 Drugs ...... 32 2.2.3 I.C.V. Surgery ...... 33 2.2.4 Locomotor activity ...... 33 2.2.5 Sensorimotor gating (prepulse inhibition of startle assessment) ...... 34 VI

2.2.6 Drug treatment scheme for open field and PPI ...... 35 2.2.7 Statistical Analysis ...... 35 2.3 Results ...... 36 2.3.1 NRG1 prevents PCP-induced hyperlocomotion following peripheral administration ...... 36 2.3.2 NRG1 prevents PCP-induced hyperlocomotion following central administration ...... 38 2.3.3 NRG1 prevents PCP-induced deficit of prepulse inhibition of startle ... 40 2.4 Discussion ...... 43 2.4.1 NRG1 treatment prevents PCP-induced hyperlocomotion ...... 43 2.4.2 NRG1 treatment prevents PCP-induced PPI impairments ...... 44 2.4.3 Behavioural consequences of increased NRG1-ERBB signalling – treatment implications ...... 46 3 Neuregulin 1 treatment affects GABA and glutamate release after PCP co- treatment in freely moving mice ...... 49 3.1 Introduction ...... 49 3.2 Materials and Method ...... 52 3.2.1 Subjects ...... 52 3.2.2 Drugs ...... 52 3.2.3 Microdialysis and amino acid quantification ...... 52 3.2.4 Statistical Analysis ...... 56 3.3 Results ...... 60 3.3.1 Extracellular levels of GABA, glutamate and glycine in the prefrontal cortex 60 3.3.2 Extracellular levels of GABA, Glutamate and Glycine in the dorsal hippocampus ...... 66 3.4 Discussion ...... 72 3.4.1 NRG1 treatment affects extracellular neurotransmitter levels in the PFC in the presence of PCP ...... 72 3.4.2 NRG1 treatment effects extracellular neurotransmitter levels in the dCA1 in the presence of PCP ...... 80 4 NRG1 prevents PCP-induced changes in GABAergic interneurons in vitro ...... 88 4.1 Introduction ...... 88

VII

4.2 Methods ...... 91 4.2.1 Subjects ...... 91 4.2.2 Drugs ...... 91 4.2.3 Frontal cortex culture ...... 91 4.2.4 Cell culture treatment and mRNA expression analysis ...... 92 4.2.5 Selection of reference genes...... 93 4.2.6 Real Time Quantitative PCR method and data analysis ...... 93 4.2.7 Statistical Analysis ...... 94 4.3 Results ...... 95 4.3.1 NRG1 treatment prevents PCP-induced increase in GAD67 mRNA expression levels ...... 95 4.3.2 NRG1 treatment prevents PCP-induced decrease in GAD65 mRNA expression levels ...... 97 4.3.3 NRG1 treatment prevents PCP-induced increase in PV mRNA expression levels ...... 99 4.3.4 DISC1 mRNA expression levels were not affected by NRG1 or PCP treatments ...... 101 4.3.5 Synaptophysin mRNA expression levels were not affected by NRG1 or PCP treatments ...... 102 4.3.6 NMDA-NR1 mRNA expression levels were not affected by NRG1 or PCP treatments ...... 103 4.4 Discussion ...... 104 4.4.1 NRG1 treatment prevents PCP-induced mRNA expression changes relevant for GABAergic signalling ...... 104 4.4.2 Pre- and post-synaptic markers under the influence of PCP and NRG1 108 5 Overall Discussion and Conclusion ...... 110 5.1 The Antipsychotic Potential of NRG1 Treatment ...... 110 5.1.1 Possible mechanisms underlying the effect of NRG1 on PCP impairments ...... 113 5.1.2 NRG1 treatment potential beyond schizophrenia ...... 114 5.2 Recommendation for Future Research ...... 114 5.2.1 The role of dopamine in the treatment effect of NRG1 ...... 114

VIII

5.2.2 NRG1 treatment effect on cognitive performance ...... 115 5.2.3 Long-term effects of NRG1 treatment ...... 115 5.2.4 Potential treatment effect of NRG1 on existing impairments ...... 116 5.2.5 NRG1 treatment effect on human nervous system ...... 117 5.3 Conclusion ...... 117 6 References ...... 119

LIST OF FIGURES

Figure 1: Neurotransmitter system changes in the prefrontal cortex and hippocampus of schizophrenia ...... 4 Figure 2: Schematic diagram of neuregulin 1 protein family members expressed in the frontal cortex and hippocampus ...... 15 Figure 3: Schematic diagram of ERBB4 intracellular signalling following neuregulin 1 binding on interneuron post-synaptic membrane ...... 18 Figure 4: Schematic of the open field experiment schedule ...... 33 Figure 5: Schematic of the sensorimotor gating experiment schedule ...... 34 Figure 6: Peripheral administered neuregulin 1 (NRG1) prevents PCP-induced hyperlocomotion ...... 37 Figure 7: Central administered neuregulin 1 (NRG1) prevents PCP-induced hyperlocomotion ...... 39 Figure 8: Startle response to pulse intensity and pulse timing ...... 41 Figure 9: Peripheral neuregulin 1 (NRG1) injection prevents PCP-induced prepulse inhibition (PPI) impairment ...... 42 Figure 10: Dose-response effect of neuregulin 1 (NRG1) vs PCP on locomotor activity ...... 47 Figure 11: Microdialysis probe placements ...... 53 Figure 12: Microdialysis treatment and sampling scheme for prefrontal cortex (PFC) and dorsal hippocampus CA1 (dCA1) ...... 54 Figure 13: Mass spectrometry analysis of GABA ...... 57 Figure 14: Mass spectrometry analysis of glutamate ...... 58 Figure 15: Mass spectrometry analysis of glycine ...... 59 Figure 16: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular GABA levels in the prefrontal cortex of freely moving mice...... 61 Figure 17: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glutamate levels in the prefrontal cortex in vivo...... 63 Figure 18: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glycine levels in the prefrontal cortex in vivo...... 65

Figure 19: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular GABA levels in the dorsal hippocampus CA1 in vivo...... 67 Figure 20: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glutamate levels in the dorsal hippocampus CA1 in vivo...... 69 Figure 21: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glycine levels in the dorsal hippocampus CA1 in vivo...... 71 Figure 22: Summary diagram of treatment-induced neurotransmitter level differences ...... 72 Figure 23: GAD67 mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture...... 96 Figure 24: GAD65 mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture...... 98 Figure 25: Parvalbumin (PV) mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture...... 100

LIST OF TABLES

Table 1: Effect of current antipsychotics on acute PCP-induced hyperlocomotion and sensorimotor gating impairments in rodents ...... 26 Table 2: Effect of current antipsychotics on neurotransmitter levels following acute PCP treatment in rodents ...... 28 Table 3: Effect of common antipsychotics and NRG1 on neurotransmitter levels in the PFC of rodents ...... 75 Table 4: Effect of common antipsychotics and NRG1 on neurotransmitter levels in the hippocampus of rodents ...... 83 Table 5: Hydrolysis probes for real time mRNA quantification ...... 93 Table 6: Unchanged DISC1 mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures ...... 101 Table 7: Unchanged synaptophysin mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures .... 102 Table 8: Unchanged NMDA receptor subunit 1 (NR1) mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures ...... 103

XII

LIST OF ABBREVIATIONS

ACSF (artificial cerebrospinal fluid) ADAM (A Disintegrin and metalloproteinase) AMPAR (α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor) ANOVA (analysis of variance) ARIA (acetylcholine-receptor-inducing activator) ASR (acoustic startle response) AUC (area under the curve) BACE1 (β-secretase 1) CA (cornuammonis) CNS (central nervous system) CSF (cerebrospinal fluid) Cq (quantification cycle) D2 (dopamine receptor D2) DISC1 (disrupted in schizophrenia 1) EGF (epidermal growth factor) ERBB (epidermal growth factor receptor tyrosine kinase) GABA (γ-aminobutyric acid) GAD (glutamate decarboxylase) GAT (GABA transporter) GGF (glial growth factor) HIP (hippocampus) i.c.v. (intracerebroventricular) i.p. (intraperitoneal) ISI (inter-trail stimulus interval) LMM (Linear Mixed Model) MRM (multiple reaction monitoring) mRNA (messenger ribonucleic acid) nAChR (nicotinic acetylcholine receptor) NDF (Neu differentiation factor) NR1 (essential NMDA receptor subunit 1) NRG1 (neuregulin 1) NMDAR (N-Methyl-D-aspartic acid receptor) PCP (phencyclidine) PET (positron emission tomography) PI3K (phosphoinositide 3-kinase) PFC (Prefrontal cortex) PPI (prepulse inhibition) PSD95 (postsynaptic density protein 95) PV (parvalbumin) SAL (saline solution, 0.9% NaCl) SYP (synaptophysin) TACE (tumour necrosis factor α-converting enzyme)

XIII

1 INTRODUCTION

1.1 Schizophrenia

Schizophrenia is a severe psychiatric disorder leading to life-long burden for the affected patients and their social environment. The risk of developing schizophrenia is 0.4 to 0.7% for the general population independent of world region or economic development (Bhugra 2005; Saha et al. 2005). This risk increases by having a family member with schizophrenia, depending on the degree of consanguinity, with a life- time risk of close to 50% for a monozygotic twin of an affected patient (Cardno & Gottesman 2000; Kendler & Gardner 1997). Despite the strong heritability (about 80%), no consistent genetic differences have been identified across schizophrenia patient populations (Fromer et al. 2014; Purcell et al. 2014). In fact, a recent report identified over 1500 single nucleotide differences which contribute to the risk of developing schizophrenia (Ripke et al. 2013). In addition to the genetic susceptibility, a variety of environmental factors have been identified which contribute to the actual incidence rate of schizophrenia; these include (but are not limited to) prenatal infec- tion (Meyer et al. 2009; Patterson 2009), cannabis consumption (Barkus & Murray 2010; Casadio et al. 2011),vitamin D deficiency during early development (McGrath et al. 2010), urbanicity and migration (Tandon et al. 2008; Réthelyi et al. 2013; Lin- scott & van Os 2013).

1.1.1 The symptoms of schizophrenia

The combination of genetic and environmental factors that contribute to the development of schizophrenia result in a broad range of clinical symptoms. The types of symptoms and their severity vary between individual patients as well as throughout the progress of the disorder. Prior to the onset of schizophrenia, retrospective studies report delayed motor coordination and impaired neuromotor function in individuals who later developed schizophrenia (Schenkel & Silverstein 2004). These early signs are followed by minor cognitive impairments, such as language based deficits, deficits in attention and problem solving abilities, all of which are present prior the first psychotic episode (Schenkel & Silverstein 2004; Tandon et al. 2009).

The onset of schizophrenia is marked by an initial psychotic episode often in late adolescence or early adulthood, during which patients experience a range of symptoms including hallucinations, anxiety, impaired social abilities, disordered thoughts, depression and cognitive impairments (Tandon et al. 2009). Lacking any cure, patients suffer mostly for the rest of their life, a period which is marked by recurring psychotic episodes with varying frequencies and intensities (Tandon et al. 2009; Tandon et al. 2008). The period between psychotic episodes is dominated by an initial worsening of cognitive symptoms in combination with anxiety, depression and social impairments, which later on plateau with a only small chance of recovery (Tandon et al. 2009). Furthermore, patients with schizophrenia suffer from an increased risk of suicide (12 times the general average) (Saha et al. 2007) and 50% of patients suffer from substance abuse (compared to 15% in the general population) (Tyrer et al. 2011). However, the cognitive symptoms, such as impaired working and procedural memory and reduced language and executive function abilities, have a long-lasting dire impact on the social interactions and vocational opportunities of affected individuals (Millan et al. 2012; Tandon et al. 2010).

1.1.2 The neurobiology of schizophrenia

Several neurobiological factors have been implicated in the pathophysiology underlying the broad spectrum of clinical schizophrenia symptoms. However, identifying consistent biological differences in patients with schizophrenia has proven very chal- lenging. Firstly, the range of genetic and environmental factors which contribute to the risk of developing schizophrenia varies between patients. Secondly, the first psychotic episode occurs predominantly during late adolescence and early adulthood, a period of increased cortical plasticity (Uhlhaas & Singer 2011), making it difficult to identify pre-existing neurobiological differences. Finally, differences in antipsychotic treatment regimens between patients limit the ability of post-mortem studies to shed light on the symptom-causing pathophysiology (Smieskova et al. 2009). Nonetheless, variations in specific neurotransmitter signalling systems have been reported from a growing number of patient cohorts, allowing the identification of chemical differences in relation to the observed symptomatology. Among the investigated neurotransmitter systems, differences in the dopaminergic, glutamatergic and γ-aminobutyric acid (GABA) signalling pathways have been the most consistently 2

observed in patients with schizophrenia. Studying the neurotransmitter systems has been made possible by post-mortem protein and messenger ribonucleic acid (mRNA) analysis of regional expression profiles for synthesizing enzymes, membrane-bound receptors, membrane transporters and associated intracellular signalling elements. Furthermore, recent in vivo imaging methods such as positron emission tomography (PET), magnetic resonance spectroscopy and single photon emission computed tomography enabled longitudinal evaluation of neurochemical changes in patients (Mcguire et al. 2008).

1.1.2.1 Dopaminergic system and schizophrenia The dopaminergic system was the first neurotransmitter system implicated in the pathophysiology of schizophrenia (Carlsson & Carlsson 2006). Dopamine is a monoamine neurotransmitter that modulates the fast neurotransmission of glutamate and GABA. Dopamine exerts its effect through G-protein coupled receptors which are found on neurons in areas critical for voluntary movement, reward, attention, working memory and learning (Beaulieu & Gainetdinov 2011). In vivo and post-mortem studies in patients with schizophrenia have reported slightly increased levels of dopamine receptors in the striatum and prefrontal cortex (PFC) (Chana et al. 2013; Gurevich et al. 1997; Howes & Kapur 2009; Toda & Abi- Dargham 2007) (Figure 1). Furthermore, increased dopamine synthesis, release and extracellular levels have been detected in the striatum of patients, while reduced dopamine levels have been identified in the PFC of patients (Howes & Kapur 2009; Howes et al. 2013) (Figure 1). These findings indicate that pre-synaptic dopaminergic mechanisms are consistently altered in patients with schizophrenia (Howes & Kapur 2009). Differences in the modulatory action of long-range dopaminergic midbrain projections on to local glutamatergic and GABAergic neurons have therefore been proposed as a fundamental element underlying a range of schizophrenia symptoms (Howes & Kapur 2009).

Figure 1: Neurotransmitter system changes in the prefrontal cortex and hippocampus of schizophrenia Schematic summary diagram of the predominant cell types in the frontal cortex and hippocampus and variations in neurotransmitter signalling systems identified in patients with schizophrenia in these cells. Cell types: pyramidal neurons (P, blue), GA- BAergic interneurons (i, orange), astrocytes (A, green) and oligodendrocytes (O, navy). Protein expression and neurotransmitter level differences have been identified in the dopaminergic (green insert; D2: dopamine D2 receptor), glutamatergic (blue insert; NR1: NMDA receptor subunit NR1), GABAergic (orange insert; GAD67: glutamate decarboxylase, PV: parvalbumin, GAT: GABA transporter, GABAA-R: GABA receptor subtype A), serotonergic (5HT-R: serotonin receptor) and cholinergic (α7 nACh-R: acetylcholine receptor subtype α7) system. Arrows indicate predominantly ↓reduced or ↑increased levels, or ↑↓inconsistent findings in schizophrenia patients compared to healthy controls. For expression details and references see 1.1.2. Diagram modified from (Meyer-Lindenberg 2010).

1.1.2.2 Glutamatergic system and schizophrenia After observing psychotic-like symptoms following the recreational use of substances with strong N-Methyl-D-aspartic acid receptor (NMDAR) antagonism, the involvement of the glutamatergic system in the pathophysiology of schizophrenia has received increasing attention (Moghaddam & Javitt 2012). Glutamate is an excitatory neurotransmitter throughout the nervous system with binding affinity for pre-, post- and extrasynaptic receptors, which include fast- and slow acting ion channels and G- protein coupled receptors (Traynelis et al. 2010). Several post-mortem studies provide support for a NMDAR hypofunction hypothesis in patients with schizophrenia. Protein and mRNA expression of the essential NMDAR NR1 subunit has been reported to be reduced in the hippocampus of patients, with growing evidence for reduced NMDAR levels also in the dorsolateral PFC (Geddes et al. 2011; Weickert et al. 2013) (Figure 1). Investigating the levels of available NMDA receptors in vivo supports the notion of reduced glutamate signalling in schizophrenia, after discovering reduced NMDAR levels in the hippocampus of unmedicated patients and reduced levels in the frontal cortex of clozapine-treated patients (Pilowsky et al. 2005) (Figure 1). Glutamate levels in different brain regions of patients with schizophrenia have been studied in vivo in recent years. In the PFC of patients with schizophrenia, glutamate levels reportedly differ compared to controls depending on subregions and antipsychotic medication. Glutamate in the dorsolateral PFC appeared unchanged in unmedicated patients but reduced in patients taking antipsychotics (Poels, Kegeles, Kantrowitz, Javitt, et al. 2014). Medial PFC and anterior cingulate cortex glutamate levels however, appear to be elevated in unmedicated patients, while unchanged in medicated patients (Kegeles et al. 2012; Poels, Kegeles, Kantrowitz, Javitt, et al. 2014). Glutamate levels in other cortical areas appeared not to be altered in patients with schizophrenia (Poels, Kegeles, Kantrowitz, Slifstein, et al. 2014). In sub-cortical areas, glutamate levels have been measured in the hippocampus, thalamus and basal ganglia (Poels, Kegeles, Kantrowitz, Slifstein, et al. 2014). The results of these studies have been inconsistent, but appear to trend towards reduced glutamate levels (Poels, Kegeles, Kantrowitz, Slifstein, et al. 2014). While the NMDAR hypofunction hypothesis receives strong support from observed reductions in NMDAR levels, the lack of significant glutamate level differences suggests an involvement of additional

neurotransmitter systems. Glutamatergic pyramidal neurons are under the tonic influence of dopamine as described above. Phasic control over the activity of these neurons is exerted by rapid GABA release from GABAergic interneurons (Roth & Draguhn 2012). GABAergic signalling directly influences the response of pyramidal neurons to both glutamate and dopamine, particularly in the hippocampus and PFC (Moghaddam & Javitt 2012; Paz et al. 2008). The observed differences of the glutamatergic and dopaminergic system are therefore possibly related to differences in the GABAergic system.

1.1.2.3 GABAergic system and schizophrenia Investigations into the role of the GABAergic system in the pathophysiology of schizophrenia have recently gained momentum. A small number of in vivo neuroimaging studies have assessed differences in GABA levels in cortical and subcortical regions. Increased as well as unchanged GABA levels have been found in the PFC of patients (Goto et al. 2009; Kegeles et al. 2012; Kegeles et al. 2012). Decreased as well as unchanged GABA levels have also been observed in the basal ganglia of patients (Goto et al. 2009; Tayoshi et al. 2010). These observations are in line with an older study measuring GABA levels in the cerebrospinal fluid (CSF) of unmedicated, recently diseased patients, which identified reduced levels during the early stages of the disorder (van Kammen et al. 1982). The in vivo knowledge on GABA-A receptor levels is limited, with a recent PET study identifying reduced GABA-A receptor levels in the striatum, but unchanged levels in the hippocampus of subjects with high risk of developing schizophrenia (Kang et al. 2013). These small numbers of studies indicate a role of GABA signalling in the critical stage of the disease progression. Post-mortem investigations have identified expression differences of proteins critical to the normal functioning of the GABAergic signalling in the PFC and subregions of the hippocampus. In particular, reduced expression of the GABA synthesizing enzyme glutamate decarboxylase (GAD) and the calcium binding protein parvalbumin (PV) have repeatedly been identified in frontal cortex regions of patients with schizophrenia (Akbarian et al. 1995; Beasley & Reynolds 1997; Dracheva et al. 2004; Duncan et al. 2010; Guidotti et al. 2000; Hashimoto, Arion, et al. 2008; Hashimoto, Bazmi, et al. 2008; Hashimoto et al. 2003; Mellios et al. 2009; Sakai et al. 2008; 6

Straub et al. 2007; Volk et al. 2012; Volk et al. 2000; Woo et al. 2008) (Figure 1). In the hippocampus, GAD and PV have also been found to be decreased in some patients (Heckers et al. 2002; Zhang & Reynolds 2002; Wang et al. 2011; Thompson Ray et al. 2011; Knable et al. 2004; Torrey et al. 2005) (Figure 1). While expression differences appear to differ between patient cohorts, antipsychotic treatment differences are likely the cause for this discrepancy (Heckers et al. 2002; Schreiber et al. 2011). Furthermore, GABA-A receptors have found to be increased while GABA transporters (GAT) appear decreased in the PFC and hippocampus of patients (Benes, Khan, et al. 1996; Benes, Vincent, et al. 1996; Duncan et al. 2010; Volk et al. 2001; Hashimoto, Arion, et al. 2008; Hashimoto, Bazmi, et al. 2008; Schleimer et al. 2004) (Figure 1). The differences in GABA-A receptors and GAT are likely a compensatory consequence to the reduced GABA production indicated by reduced GAD expression. Noteworthy, GABA-A receptors are expressed on pyramidal neurons, interneurons and astrocytes (Lee, Schwab, et al. 2011; Lewis et al. 2005; Pettit & Augustine 2000), GAD and GAT are expressed in GABAergic interneurons and GABA-producing astrocytes (Fish et al. 2011; Lee, McGeer, et al. 2011), with PV exclusively expressed in a subset of GABAergic interneurons (Lewis et al. 2005; Markram et al. 2004). This expression profile indicates a cell type specific contribution to the schizophrenia pathophysiology. While the cause for reduced GAD and PV expression remains unknown, it likely indicates a reduction in GABAergic interneuron activity, leading to reduced GABA production and a lower demand for intracellular calcium management. Local populations of GABAergic interneurons are critical for coordinated signalling inside- as well as across brain regions, in particular for areas strongly associated with symptoms of schizophrenia such as the PFC and the hippocampus (Godsil et al. 2013; Gonzalez-Burgos et al. 2011; Inan et al. 2013). Dif- ferences in the activity level of these interneurons in patients with schizophrenia would consequentially contribute to a range of the observed symptoms.

1.1.2.4 Other neurotransmitters While focusing on the three most widely studied neurotransmitter systems in schizophrenia, changes to many other neurotransmitters have been reported, including serotonin, acetylcholine and glycine, which in turn can influence dopamine, glutamate and GABA signalling. 7

Serotonin receptor activity can influence dopamine and glutamate release in several brain areas (de Bartolomeis et al. 2013). Serotonin receptors have consequentially been implicated in the pathophysiology of schizophrenia and proposed as possible treatment targets (de Bartolomeis et al. 2013; Meltzer et al. 2003). Patients with schizophrenia appear to have reduced expression levels of serotonin receptors in the frontal cortex and hippocampus (López-Figueroa et al. 2004; Ngan et al. 2000). However, the knowledge about differences in the serotonergic system in schizophrenia is limited by the lack of specificity of serotonin receptor ligands, as well as the effect of antipsychotic treatment on the serotonergic system (Chana et al. 2013).

The involvement of acetylcholine neurotransmission in the symptomatology of schizophrenia has been suggested as a possible explanation for the high incidence of smoking seen in patients with schizophrenia (de Leon & Diaz 2005). Post-mortem studies of acetylcholine receptor expression, especially of the α7 nicotinic acetylcholine receptor (nAChR), have identified reductions in several frontal cortex and hippocampus regions in patients with schizophrenia (Thomsen et al. 2010). Since α7 nAChR is highly expressed on GABAergic interneurons in healthy brains (Thomsen et al. 2010), altered acetylcholine signalling via this receptor might contribute to the reported activity difference in these interneurons described earlier.

Glycine acts as an allosteric modulator on NMDAR, providing tonic modulation to the receptor channel opening capacity (Labrie & Roder 2010). Patients with schizophrenia reportedly have lower blood serum glycine levels, which appear to be negatively correlated with symptom severity (Hons et al. 2010; Kantrowitz & Javitt 2010). The observed reduction in glycine levels has been suggested to contribute to the NMDAR hypofunction hypothesis described above (Labrie & Roder 2010).

1.1.2.5 Brain connectivity and schizophrenia Activation of NMDAR on GABAergic interneurons in the PFC enables precise feedback control over local pyramidal neuron excitability (recently reviewed by (Bartos et al. 2007; Chamberland & Topolnik 2012; Isaacson & Scanziani 2011)). PV expressing interneurons synapse with over 1000 pyramidal neurons, synchronizing the action potential generation across populations of pyramidal neurons (Bartos et al. 8

2007). The synchronized excitatory-inhibitory balance of the PFC is considered to be the corner stone of several functions impaired in schizophrenia such as working memory, attention and sensory gating (Gonzalez-Burgos et al. 2011; Uhlhaas & Singer 2010). Several of these abilities appear to additionally depend on afferent and efferent connections between the PFC and the hippocampus (Godsil et al. 2013). Deficits in cortical synchrony in the PFC of patients with schizophrenia have been identified by electrophysiological and neuroimaging studies (Farzan et al. 2010; Kirihara et al. 2012; Lee et al. 2003; Uhlhaas & Singer 2010; Woo et al. 2010). Fur- thermore, neuroimaging studies utilizing diffusion tensor imaging and magne- toencephalography have provided evidence suggesting that the interaction between the PFC and the hippocampus of patients with schizophrenia differs compared to control subjects (Ellison-Wright & Bullmore 2009; Kumra et al. 2004; Olbrich et al. 2008; Scheel et al. 2013; White et al. 2007) (recently reviewed by (Del Arco & Mora 2009; Javitt et al. 2008; Meyer-Lindenberg 2010; Schmitt et al. 2011)). Together with local GABAergic and glutamatergic neurotransmission imbalances, these connectivity and synchronicity deficits are thought to significantly contribute to the symptom spectrum of schizophrenia (Gonzalez-Burgos et al. 2011; Schmitt et al. 2011).

1.1.3 Current Pharmacological Treatment Situation of Schizophrenia

Sixty years of pharmacological research has resulted in a range of treatment options, which, however, still have limited efficacy in treating the symptom spectrum of schizophrenia. All currently available antipsychotics share a common antagonistic action for the dopamine D2 receptor, with their D2 affinity predicting their clinical potential (Tandon et al. 2010). The initial generation of antipsychotics, including substances like chlorpromazine and haloperidol, provided the very first pharmacological treatment option for patients with schizophrenia. Patients were able to reduce the severity of their psychotic episodes, particularly hallucinations and delusions (Miyamoto et al. 2012). However, serious motor control side effects, such as tremor, tardive dyskinesia and Parkinson- ism, were common to the first generation of antipsychotics (Farde et al. 1992). Research aiming at reducing the adverse reactions of the early antipsychotics while maintaining the beneficial D2 antagonism resulted in the development of the second 9

generation of antipsychotics, such as olanzapine, clozapine and risperidone (Tandon et al. 2010). These newer substances have broader neurotransmitter receptor profiles targeting serotonergic, histaminergic and muscarinic receptors with reduced affinity for D2 receptors, resulting in fewer motor control side effects (Miyamoto et al. 2012). Additionally, these second generation antipsychotics were of benefit for patients with schizophrenia who remained unresponsive to the first generation of D2 antagonistic antipsychotics (Meltzer 1992). However, the newer group of drugs have been shown to cause a range of metabolic side effects, including weight gain and sedation, to a larger extend than the first generation of antipsychotics (Leucht et al. 2013). The retained beneficial D2 antagonisms led to the formation of the first neurotransmitter hypothesis of schizophrenia, proposing that dysregulation of the dopaminergic system underlies the schizophrenia symptomatology, as described in section 1.1.2.1 (Seeman 1987). The first and second generation antipsychotics show a strong therapeutic potential against the so called positive symptom spectrum, which describes reality distorting perceptions such as auditory and visual hallucinations, thought insertion and delusion of control (Leucht et al. 2013; Tandon et al. 2010; Tandon et al. 2009). However, antipsychotics are less efficient in reducing anxiety, depression and social impairments and show very little effect against cognitive symptoms (Millan et al. 2012; Miyamoto et al. 2012; Tandon et al. 2010). This treatment deficit suggests that not all neurotransmission differences are addressed by the available antipsychotics.

The effect of antipsychotics on neurotransmitter levels and network activity in the frontal cortex and hippocampus formation in patients with schizophrenia has been assessed by several neuroimaging studies. Haloperidol, clozapine and aripiprazole have been reported to reduce the relative metabolic activity within the PFC of patients with schizophrenia (Cohen et al. 1999; Kim et al. 2013; Lahti et al. 2005). However, olanzapine and sertindole appear to increase the relative metabolic activity within the PFC in patients (Buchsbaum et al. 2009; Buchsbaum et al. 2007). The observed difference in PFC activation by the different antipsychotics likely contributes to the varying effects of these treatments on the symptom spectrum in patients with schizophrenia (Bilder et al. 2002; Carpenter & Buchanan 2008; Keefe et al. 2004; Leucht et al. 2009; Ramaekers et al. 1999). Interestingly, while antipsychotic treat-

ment did not appear to effect glutamate concentration in the anterior cingulate cortex (which is connected to the PFC) (Tayoshi et al. 2009), GABA concentration in the same region appeared to negatively correlate with the dose of antipsychotics in patients with schizophrenia (Tayoshi et al. 2010). In light of the proposed reduction of GABA levels in the PFC of patients described earlier, this finding suggests that symptoms caused by an altered excitatory-inhibitory balance in the PFC might not be sufficiently addressed by classical antipsychotics. Since this aspect might be of critical importance for antipsychotic treatment development, further research will need to investigate this observed interaction. The effect of antipsychotic treatment on hippocampal neurotransmitter levels has not been well studied, with some indication for increased glutamate levels following chronic antipsychotic treatment (Poels, Kegeles, Kantrowitz, Slifstein, et al. 2014). However, as summarized in a recent review by Poels and colleagues, large differences in experimental and subject parameters between studies limit the reliable identification of the neurotransmitter level and receptor changes caused by antipsychotics (Poels, Kegeles, Kantrowitz, Javitt, et al. 2014). Yet, animal studies assessing the effects of chronic antipsychotic treatment indicate lasting differences to neurotransmitter systems. Despite the common antagonism of D2 receptors, increased and decreased GABA receptors levels, g-protein coupled receptor levels and NMDA receptor phosphorylation, among others have been reported following chronic antipsychotic treatment (reviewed by Lieberman et al. (2008)).

Despite over 60 available antipsychotics and over 100,000 research publications about the cause and management of schizophrenia (Scopus, 2014), the current treatments are insufficient in fully addressing the severely debilitating symptom spectrum. The strong focus of the current antipsychotics on the dopaminergic and serotonergic system leaves differences in the GABAergic and glutamatergic system un- dertreated. It has been discussed that this likely contributes to the limited ability of current antipsychotics to sufficiently improve the cognitive impairments, which are a strong predictor of the poor social outcomes of schizophrenia (Tandon et al. 2009).

1.1.4 Current attempts in targeting GABA for treating schizophrenia

The proposed role of GABAergic neurotransmission in the cognitive symptoms of schizophrenia has promoted the ionotropic GABA receptor subtype A to a prime target for antipsychotic treatments. The post-synaptical expression of GABA-A receptors facilitates the synchronization between interneurons and pyramidal neurons in frontal cortex and hippocampal regions (Benardo 1997; Cobb et al. 1995). Targeting the synaptic GABA-A receptors with benzodiazepines, however, causes several adverse reactions such as sedation, ataxia and cognitive impairments (Möhler et al. 2002). As reported earlier, reduced expression of α-subtype containing GABA-A receptors has been found in some cohorts of patients with schizophrenia (Duncan et al. 2010; Hashimoto, Arion, et al. 2008; Volk et al. 2002). A small number of GABA-A α- specific clinical trials, however, failed to show consistent improvements to the symptoms of patients with schizophrenia (Rudolph & Knoflach 2011; Vinkers et al. 2010). While targeting the GABA-A receptor to address GABAergic interneuron differences in schizophrenia has so far been unsuccessful, different strategies for the GABA aspect of the schizophrenia pathophysiology will have to be considered (Vinkers et al. 2010). In particular, modifying GABAergic neurotransmission by targeting the activity of GABAergic interneurons instead of directly affecting GABA-A receptors might form a fruitful strategy. The epidermal growth factor receptor tyrosine kinase ERBB4 has been detected on GABAergic interneurons in the PFC and hippocampus, while not being expressed on pyramidal neurons in the same regions (Neddens & Buonanno 2010; Neddens et al. 2011; Vullhorst et al. 2009; Abe, Namba, et al. 2009; Longart et al. 2007; Nagano et al. 2007; Yau et al. 2003). The restricted neuronal expression profile of ERBB4 to GABAergic interneurons with reported deficits in schizophrenia raises the possibility that its ligand neuregulin 1 (NRG1) might selectively alter GABAergic interneuron activity as could be relevant for schizophrenia signalling aberrations.

1.2 Neuregulin 1 and ERBB4 signalling

1.2.1 ERBB4

ERBB4 is part of a family of membrane-spanning tyrosine kinase receptors (ERBB1- 4), homologous to epidermal growth factor (EGF) receptors. ERBB4 homodimers and ERBB4-ERBB2,3 heterodimers can be activated by NRG1 and other EGF-like factors of the neuregulin family, (reviewed by (Iwakura & Nawa 2013)). ERBB2 and ERBB3 build functional receptors by forming heterodimers with other ERBB proteins, which then determine their binding ligands (Guy et al. 1994; Iwakura & Nawa 2013; Tzahar et al. 1996). ERBB receptors show a subtype specific expression pattern across the body. In the periphery ERBB receptors are expressed on neuromuscular junctions, including car- diomyocytes, influencing cholinergic neurotransmission between peripheral nerves and their target muscle (Bersell et al. 2009; Wadugu & Kühn 2012; Zhu et al. 1995). ERBB receptors have further been detected in endocrine organs (Zhao 2013) and tumour cells (Hegde et al. 2013; Seshacharyulu et al. 2012). Additionally, macrophag- es also appear to express ERBB receptors, suggesting an involvement of NRG1 signalling in immune system actions (Calvo et al. 2010). In the central nervous system, ERBB receptors are found on a range of glial cell types in different cortical and subcortical regions, with ERBB4 expressed on astrocytes, microglia and oligodendrocytes (Calvo et al. 2010; Dimayuga et al. 2003; Ger- ecke et al. 2001; Raabe et al. 2004; Sharif et al. 2009; Vartanian et al. 1997). Neuronal expression of ERBB4 has been identified on cortical and hippocampal GABAergic interneurons (Abe, Nawa, et al. 2009; Longart et al. 2007; Nagano et al. 2007; Yau et al. 2003). The neuronal ERBB4 expression in both regions has been confirmed to be limited to post-synaptic densities of GABAergic interneurons (Faz- zari et al. 2010; Garcia et al. 2000; Neddens et al. 2011; Vullhorst et al. 2009). Inter- neurons expressing the calcium binding protein PV contain the majority of ERBB4 found in the PFC and hippocampus, with interneurons expressing the calcium binding proteins calretinin or somatostatin containing only a fraction of available ERBB4 receptors (Fazzari et al. 2010). Initial observations of ERBB4 being expressed on pyramidal neurons or pre-synaptic sites of interneurons have been revised due to unspecific antibody binding (Neddens et al. 2011). Outside of the cortex, ERBB4 has 13

also been found on dopaminergic neurons in the striatum (Abe, Namba, et al. 2009; Depboylu et al. 2012; Gerecke et al. 2001; Steiner et al. 1999), PV-negative neurons in the amygdala (Shamir et al. 2012), cerebellar granule cells (Fenster et al. 2012), and on neuroprogenitor cells of the dentate gyrus (Mahar et al. 2011). The ERBB4 expression data of some of these studies, however, relies mainly on the anti-ERBB4 antibody sc-283 from Santa Cruz, which has been shown to exhibit non-specific la- belling of other subunits (Neddens et al. 2011) and will therefore need additional confirmation.

1.2.2 Neuregulin 1

NRG1 is part of a family of growth and differentiation factors which are transcribed by the NRG1-4 genes. The NRG1 genes encode six types of proteins (I-VI) with pre- dicted 31 isoforms (Mei & Xiong 2008). The commonly known isoforms of NRG1 include Neu differentiation factor (NDF), heregulin, glial growth factor (GGF) and acetylcholine-receptor-inducing activator (ARIA), with ARIA being the first one identified (Falls et al. 1990). Each NRG1 isoform consists of an extracellular type- specific sequence and a transmembrane domain followed by an intracellular cytoplasmic tail (Figure 2). Each member of the NRG1 family contains an EGF-like domain between the transmembrane domain and the extracellular tail in either α or β configuration (Figure 2). The type-specific segment of NRG1 type III contains a second transmembrane domain and intracellular segment, leading to an extracellular loop exposing the EGF-like domain to extracellular space (Figure 2) (reviewed in detail by (Buonanno 2010; Mei & Xiong 2008; Talmage 2008).

Figure 2: Schematic diagram of neuregulin 1 protein family members expressed in the frontal cortex and hippocampus Schematic diagram of the predominant cell types in the frontal cortex and hippocampus and neuregulin 1 protein family members expressed by these cells. Cell types: pyramidal neurons (P, blue), GABAergic interneurons (I, orange), astrocytes (A, green) and oligodendrocytes (O, navy). Insert: schematic structure of Neuregulin 1 type I to IV proteins expressed on the cell membranes of pyramidal neurons, astrocytes and interneurons. References for signalling, cleavage and expression details see 1.2.2.

All NRG1 subtypes signal via their extracellular EGF-like domain, which can be cleaved off the transmembrane domain by different metalloproteases including AD- AM10, 17, 19/TACE (tumour necrosis factor α-converting enzyme), β-secretase 1 15

(BACE1), γ-secretase (Fleck et al. 2013; Fleck et al. 2012; Luo et al. 2011) and neuropsin (Tamura et al. 2012). NRG1 type III is additionally equipped for intracellular signalling via the second, cysteine-rich intracellular domain, following cleavage by Aph1B/C γ secretase (Dejaegere et al. 2008) (Figure 2).

NRG1 expression has been detected in cells of the peripheral and central nervous system (reviewed by (Talmage 2008)), the cardiac system (Wadugu & Kühn 2012) as well as the endocrine and gastrointestinal system (Zhao 2013). The majority of NRG1 isoforms in the nervous system contain the β-EGF-like domain configuration (Meyer & Birchmeier 1994; Shinoda et al. 1997). Localisation of the EGF-like domain of all NRG1 isoforms shows a broad expression pattern in the human brain across the PFC, cingulate cortex, hippocampus, thalamus, amygdala, cerebellum and brain stem (Law et al. 2004). The specific expression patterns of the different NRG1 isoforms vary throughout the nervous system development with peak expression for all forms shortly after birth (X. Liu et al. 2011). The NRG1 type I-IV isoforms have been detected in varying ratios in the PFC, hippocampus, thalamus, hypothalamus, amygdala and cerebellum (Bare et al. 2011; Chen et al. 2008; Eilam et al. 1998; Hashimoto et al. 2004; X. Liu et al. 2011; Loeb et al. 1999; Mei & Xiong 2008; Parlapani et al. 2010; Woo et al. 2007). The expression patterns of NRG1 type V and VI have so far not been published. In the cortical grey matter, pyramidal cells are the primary producer of NRG1 (Pankonin et al. 2009). NRG1 isoforms have primarily been found in the cell soma and dendrites, with NRG1 type III also expressed on axons (Bare et al. 2011; Brinkmann et al. 2008; X. Liu et al. 2011; Shamir & Buonanno 2010). The different patterns of NRG1 isoform and ERBB isoform expression across the nervous system and inside cells suggest distinct functions for NRG1-ERBB signalling in different cell types and anatomical locations. For the present study, ERBB4 receptors are of particular interest due to their confirmed expression on GABAergic interneurons in the PFC and hippocampus.

1.2.3 Acute Neuregulin 1-ERBB4 signalling

The NRG1 type I / ARIA isoform was found to strongly affect nervous system functions in the developed brain when signalling through ERBB4. Indeed, the NRG1 type 16

I expression is strongly influenced by neuronal activity during development and adulthood (Eilam et al. 1998; X. Liu et al. 2011), which suggests that this isoform is a likely candidate for neurotransmission related functions. Upon binding of NRG1 to the post-synaptically expressed ERBB4 receptor, several intracellular reactions can be triggered, which are summarized in Figure 3. The range of actions include phosphorylation of ERBB4 tyrosine kinase (Bennett 2009), trans- location of the intracellular domain to the nucleus for gene expression regulation (Al- lison et al. 2011), interaction with the postsynaptic density protein 95 (PSD95), altering NMDA-R receptor subunit phosphorylation (Bjarnadottir et al. 2007; Pitcher et al. 2011; Yu et al. 1997) and less studied mechanisms such as protein kinase C interactions (Mitchell et al. 2013) (Figure 3). Acute consequences of these reactions on neurotransmitter signalling have been identified in different brain regions of rodents, primarily by stimulation with the recombinant NRG1 type 1 β-EGF extracellular domain.

Figure 3: Schematic diagram of ERBB4 intracellular signalling following neuregulin 1 binding on interneuron post-synaptic membrane Schematic diagram of the predominant cell types in the frontal cortex and hippocampus and the intracellular signalling cascade triggered by neuregulin 1 type 1 (NRG1). NRG1 binding (1) triggers formation of ERBB3-ERBB4 heterodimers (2) bound to postsynaptic density protein 95 (PSD95), triggering a range of different intracellular signalling cascades, including: (3) gene expression regulation by cytoplasmic domain 1 (CYT1) via Jak/Stat pathway; (4) regulation of mTor pathway and intracellular Ca2+ levels by cytoplasmic domain 2 (CYT2) via PI3K and AKT phosphorylation and (5) altering NMDA receptor (NMDAR) activity through phosphorylation of family kinases SRC and FYN. References for signalling, binding and expression details see 1.2.1 and 1.2.3

1.2.3.1 Effect of NRG1-ERBB4 signalling on GABAergic neurotransmission In cortical slice preparations, increased depolarization-evoked GABA release was observed following application of NRG1, while basal GABA release remained unaffected (Woo et al. 2007). Further experiments confirmed that this effect was caused by activating ERBB4 on PV-expressing interneurons (Wen et al. 2010). The increase in GABA release by NRG1 treatment also triggered an increase in GABA-A receptor currents (Chen et al. 2010; Woo et al. 2007), consequentially reducing the action potential rate of the post-synaptic pyramidal neuron (Wen et al. 2010). Additionally, the stimulation of ERBB4 by NRG1 has been shown to directly result in the internalization of GABA-A receptors on the post-synaptic membrane of hippocampal interneurons (Mitchell et al. 2013). This reduction of GABA-A receptors was visible even after prolonged in vitro treatment with NRG1 for six days in hippocampal slices (Okada & Corfas 2004). Also, treatment of hippocampal neurons with NRG1 increased ubiquitin mRNA, an enzyme which can regulate GABA-A receptor synapse expression (Allison et al. 2011). In contrast, treatment of cerebellar granule cells with NRG1 has been found to increase GABA-A receptor expression (Rieff et al. 1999; Xie et al. 2004). These studies show that NRG1-ERBB4 signalling can acutely regulate GABA-A mediated neurotransmission in a region specific manner.

1.2.3.2 Effect of NRG1-ERBB4 signalling on glutamatergic neurotransmission GABA neurotransmission by GABAergic interneurons in the PFC and hippocampus directly affects glutamatergic pyramidal neurons (Markram et al. 2004). The effect of NRG1-ERBB4 signalling on GABAergic neurotransmission therefore suggests a possible influence on glutamatergic neurotransmission in both the PFC and hippocampus. While in vitro experiments have shown that NRG1 treatment has no effect on glutamate release (Gu et al. 2005; Pitcher et al. 2011), a recent experiment with transgenic animals indicated that NRG1 can reduce the fusion of glutamate vesicles with the pre-synaptic membrane resulting in glutamatergic dysfunction (Yin et al. 2013). Baseline excitatory post-synaptic potentials in hippocampal pyramidal neurons in vitro, however, do not appear to be affected by the presence of NRG1 (Shamir et al. 2012).

NMDA receptor currents in the PFC on the other hand were reduced following prolonged NRG1 application, which has been suggested to result in a reduction of functional NMDAR receptors on the post-synaptic membrane (Gu et al. 2005). Altered glutamatergic signalling has also been proposed to result directly from ERBB4 activation. ERBB4 and NMDAR are connected via PSD95 in GABAergic interneurons (Garcia et al. 2000; Huang et al. 2000). Activation of ERBB4 by NRG1 has been shown to inhibit phosphorylation of the non-receptor protein kinase SRC and FYN, resulting in dose-dependent changes of NMDAR function (Bennett 2009; Bjarnadottir et al. 2007; Pitcher et al. 2011). Regarding the second dominant post- synaptic glutamate receptor α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPAR), acute NRG1 treatment does not appear to effect AMPAR currents, while preventing NRG1-ERBB4 signalling destabilizes AMPAR surface expression (Fen- ster et al. 2012; Gu et al. 2005; Li et al. 2007). The potential effect on glutamate release as well as the observed changes in NMDAR function by NRG1-ERBB4 signalling suggests that local NRG1-ERBB4 signalling can influence efferent pyramidal neuron connections to other cortical and subcortical structures. NRG1-ERBB4 signalling could consequentially influence different behaviour and cognitive abilities relying on these networks.

1.2.3.3 Effect of NRG1-ERBB4 signalling on other neurotransmitter systems

1.2.3.3.1 Dopamine The identification of ERBB4 on dopaminergic neurons in the striatum has triggered investigations in to the effect of NRG1-ERBB4 signalling on dopaminergic neurotransmission. Microdialysis experiments have shown that local infusion of NRG1 into the striatum and hippocampus and peripheral NRG1 injections result in increased extracellular dopamine levels in both regions (Carlsson et al. 2011; Kwon et al. 2008; Yurek et al. 2004). The co-expression of ERBB4 and D4 on interneurons in the hippocampus might contribute to this effect, which also appears to influence network signalling underlying cognitive functions (Andersson et al. 2012).

1.2.3.3.2 Acetylcholine Hippocampal interneurons express both ERBB4 and α7 nAChR and an interaction between both signalling systems has therefore been studied in vitro. Application of 20

NRG1 type I appears to alter the surface expression levels of α7 nAChR and acetylcholine currents in the hippocampus, although the direction of this change is debated (Chang & Fischbach 2006; Liu et al. 2001). Furthermore, recent work suggests that endogenous NRG1 type III signalling is required for sustained membrane expression levels of α7 nAChR and acetylcholine currents (Hancock et al. 2008; Zhong et al. 2008).

In summary, the in vitro and in vivo studies in rodents suggest that acute NRG1- ERBB4 signalling influences several neurotransmitter systems that are strongly implicated in the symptomatology of schizophrenia. The consequences of acute NRG1-ERBB4 signalling have also been studied in transgenic animal models with over and under-expression of NRG1 isoforms and ERBB4 (Deng et al. 2013). However, the critical role of NRG1 signalling in central nervous system development could have altered the response to acute NRG1-ERBB4 interventions in adulthood. The identified differences in neurotransmitter receptor expression and function in these transgenic animals are therefore of limited value for our understanding of the acute effect of NRG1-ERBB4 signalling on neurotransmission in the developed brain.

1.2.4 Schizophrenia and the NRG1-ERBB4 pathway

Additionally to the reported effect of NRG1-ERBB4 signalling on schizophrenia- relevant neurotransmitter systems, a genetic association between NRG1-ERBB4 and schizophrenia has been identified.

1.2.4.1 Genetic association between NRG1, ERBB4 and schizophrenia The NRG1 gene has been considered as a potential risk factor for developing schizophrenia following a discovery by Stefansson and colleagues (2002). This group identified an association and linkage between a haplotype in the NRG1 gene on chromo- some 8 and schizophrenia in an Icelandic population of patients with schizophrenia (Stefansson et al. 2002). Following this initial finding, the same and other DNA sequences on the NRG1 gene have been associated with schizophrenia (Agim et al. 2013; Alaerts et al. 2009; Bousman et al. 2013; Georgieva et al. 2008; Huertas- 21

Vazquez et al. 2013; Law et al. 2006; Munafò et al. 2008; Munafò et al. 2006; Nico- demus et al. 2009; Petryshen et al. 2005; Stefansson et al. 2003; Walker et al. 2010; Walss-Bass et al. 2006; Williams et al. 2003). Additionally, particular polymorphisms in the ERBB4 gene have also been associated with schizophrenia (Law et al. 2007; Nicodemus et al. 2006; Silberberg et al. 2006). However, not all patient populations show a genetic associations between NRG1 and schizophrenia (Gardner et al. 2006; Gong et al. 2009; Kampman et al. 2004; Loh et al. 2013; Moon et al. 2011; Munafò et al. 2006; Walker et al. 2010) nor ERBB4 and schizophrenia (Chen et al. 2012; Stefanis et al. 2013).

1.2.4.2 Molecular association between NRG1, ERBB4 and schizophrenia The majority of genetic polymorphisms associating NRG1 with a risk for developing schizophrenia have been found in the non-coding region of the NRG1 gene, which suggests a change in expression of NRG1 isoforms, rather than a change in the amino acid sequence of the proteins themselves (Buonanno 2010). Several studies have therefore been performed to assess the mRNA and protein expression levels of the different NRG1 isoforms in the brain and periphery of patient with schizophrenia. NRG1 type I appears to be increased in both the PFC and hippocampus, while type II, III and IV remain unaltered in patients with schizophrenia (Hashimoto et al. 2004; Law et al. 2006). Interestingly, reduced NRG1 type I levels have been observed in the blood plasma of schizophrenia patients (Shibuya et al. 2010). Increased as well as unaltered ERBB4 expression levels have been observed in the PFC of patients with schizophrenia (Chong et al. 2008; Hahn et al. 2006; Joshi et al. 2014), with no data available on hippocampus levels. The NRG1 and ERBB4 protein quantification studies are hampered by the unavaila- bility of reliable isoform-specific antibodies. This is highlighted by studies with isoform-unspecific NRG1 antibodies which have identified increased, decreased and unchanged NRG1 levels in the PFC of patients (Barakat et al. 2010; Bertram et al. 2007; Chong et al. 2008; Hahn et al. 2006). While the results of these expression studies are challenged by specificity issues, and are at times inconsistent in their findings, they still support the genetic association studies in suggesting an involvement of altered NRG1-ERBB4 signalling system in the schizophrenia pathophysiology. 22

1.3 General aim

Impaired dopaminergic, GABAergic and glutamatergic neurotransmission has been observed in patients with schizophrenia. While currently available antipsychotics primarily address the dopaminergic (and serotonergic) system, the GABAergic and glutamatergic systems are not being targeted. Acute signalling of the neurotrophic factor NRG1 via its ERBB4 receptor has been shown to affect, among others, GA- BAergic and also glutamatergic neurotransmission in schizophrenia-relevant brain areas, particularly in the PFC and hippocampus. The aim of my thesis is therefore to assess potential antipsychotic-like characteristics of NRG1 treatment with a focus on GABAergic and glutamatergic neurotransmission.

1.4 Pharmacological animal model strategy for antipsychotic drug development

The discovery process of a novel antipsychotic pharmacological treatments includes several stages from single cell response measurements in vitro to treating modelled schizophrenia-relevant behavioural impairments in vivo (Sarter & Tricklebank 2012). For acute NRG1 application, the effect on different neurotransmitter systems has been assessed in vitro and partially in vivo (see 1.2.3). Furthermore, several studies have administered the active NRG1β1-EGF domain to healthy mice, showing that peripheral treatments have the potential to cross the blood-brain barrier and interact with central ERBB receptors (Carlsson et al. 2011; Cui et al. 2013; Kastin et al. 2004; Kato et al. 2011; Rösler et al. 2011). Based on this, the first step for me to identify potential antipsychotic properties was to assess the effect of NRG1 treatment on schizophrenia-relevant behavioural impairments in mice.

1.4.1 Chronic vs acute rodent models

Schizophrenia is characterized by a combination of different symptoms that vary in their severity throughout the progress of the disorder, as described in 1.1. Animal modelling attempts have therefore focused either on triggering a psychotic episode- like state (acute models), which causes a range of functional impairments, or on cre-

ating long-lasting behavioural and molecular differences (chronic or sub-chronic models), which would represent psychotic deficits between episodes (Pratt et al. 2012). Chronic or non-acute modelling strategies include the introduction of schizophrenia- relevant genetic mutations, applying long-term chemical treatments, altering developmental processes or environmental interventions (Arguello & Gogos 2011; Ar- guello & Gogos 2006; Pratt et al. 2012). However, as schizophrenia is a multifaceted and unique human disorder, there is ongoing debate of how to achieve construct validity for the complex combination of environmental and genetic risk factors in non- acute animal models (Harrison et al. 2012; Jones et al. 2011). Modelling the acute psychotic state in animals is primarily achieved by administering illicit/psychodelic drugs with pharmacological actions relevant to alterations of neurotransmission observed in patients with schizophrenia (Arguello & Gogos 2006). Phencyclidine (PCP), also known as angel dust, is a commonly used substance for this purpose, providing an environment for early treatment validation (Adell et al. 2012). Use of PCP has been considered for animal models after observing that acute and chronic use can trigger schizophrenia-like symptoms, such as disorganization of thoughts and body-image changes, and prolonged psychosis in healthy adults (Javitt & Zukin 1991; Luby et al. 1959; Smith & Wesson 1980; Yago et al. 1981). Further- more, PCP has also been shown to increase symptom severity in patients with schizophrenia (Javitt et al. 2012; Luby et al. 1959; Murray 2002).

1.4.2 The acute PCP schizophrenia rodent model

PCP is a potent non-competitive NMDAR antagonist and an agonist for the dopamine D2 receptors (Kapur & Seeman 2002; Seeman et al. 2009; Seeman et al. 2005), with additional lower binding affinity for several other targets such as acetylcholine, kappa opioid and muscarinic receptors (Amir et al. 1985; Hustveit et al. 1995; Maa- yani & Weinstein 1979; Zukin 1982). PCP treatment in rodents has therefore been used to model the glutamate hypofunction and dopamine hyperfunction hypotheses of schizophrenia (Moghaddam & Krystal 2012; Seeman 2009). The combination of pharmacological actions and observed behavioural consequences in humans provides face validity to the acute PCP treatment for modelling acute psychotic episodes of schizophrenia (Adell et al. 2012; Javitt et al. 2012; Pratt et al. 2012). 24

1.4.2.1 Behavioural effects of acute PCP treatment in rodents Acute PCP treatment in rodents has been shown to induce a wide spectrum of schizophrenia-relevant symptoms, including hyperlocomotion and impaired sensorimotor gating (Geyer et al. 2001; Gleason & Shannon 1997). Increased locomotor activity after acute PCP treatment has been considered a model for aspects of positive symptoms experienced by patients with schizophrenia during the acute psychotic state of a psychotic episode (Adell et al. 2012). The relevance of this behavioural change has recently been highlighted in a human-equivalent for the rodent locomotor activity test, observing increased activity of patients with psychosis (Perry et al. 2010; Perry et al. 2009). A range of studies have therefore been performed to assess the ability of current antipsychotics to reduce or prevent the PCP-induced hyperlocomotion impairments in rodents, summarized in Table 1. Patients with schizophrenia also show impaired sensorimotor gating (Braff et al. 2001). Sensorimotor gating describes the ability to “gate” trivial or currently irrele- vant sensory information from awareness to focus attention on situation-relevant information (Braff & Geyer 1990). Sensorimotor gating experiments utilise the inhibition of startle response which occurs when a startling stimulus is preceded by a weak pre-stimulus, so called prepulse inhibition (PPI). Impairments in PPI have been observed in patients with schizophrenia (Braff et al. 2001; Swerdlow et al. 2008). PPI deficits can be induced by acute PCP treatment in rodents (Yee et al. 2004) and prevented or reduced by several current antipsychotics (Table 1). In humans, antipsychotic drugs have shown to improve PPI impairments in some (Wynn et al. 2007), but not all patients with schizophrenia (Mackeprang et al. 2002; Geyer et al. 2001). This has made the PCP-induced hyperlocomotion and sensorimotor gating impairment widely used drug screening paradigms.

Table 1: Effect of current antipsychotics on acute PCP-induced hyperlocomotion and sensorimotor gating impairments in rodents Drug Species Locomotor activity Prepulse inhibition PCP (mg/kg) Mouse ↑ (3, 5, 7.5)f, i, k, w, h ↓ (2.5, 3, 4, 5, 8, 10)a, f, n, r, s Rat ↑ (1.25, 1.5, 2, 2.5 3, ↓ (1, 1.25, 1.5, 2, 2.5, 3, 7.5)c, e, o, t, u, x, y, z 5)b, d, g, l, m, o, p, q, t, v, x Aripiprazole Mouse ˅a Rat ˅x, ~u ~x Clozapine Mouse ˅f, h, i, k, w, ˅f, ~a, Rat ˅y, z , ~c ˅b, d, p, v, ~l, q Haloperidol Mouse ˅i, k, w ˅n, ~a, Rat ˅c, t, y, z, ~e ˅p, ~l, m, t Ketanserin Mouse ˅k Rat ~b Olanzapine Mouse ˅k ~a Rat ~u ˅g, o, p Risperidone Mouse ˅h ˅p Rat ~u Ritanserin Mouse ˅k Ziprasidone Mouse ˅e PCP treatment (mg/kg either intraperitoneal or subcutaneous) effect direction: ↑ increase, ↔ no change, ↓ reduction; Antipsychotic treatment effect direction: ˅ reduces PCP effect, ~ no change to PCP effect, ˄ increases PCP effect References: a(Fejgin et al. 2007), b(Bakshi et al. 1994), c(Abekawa et al. 2006), d(Ballmaier et al. 2007), e(Abdul-Monim et al. 2003), f(Adage et al. 2008), g(Bakshi & Geyer 1995), h(Fell et al. 2008), i(Vanover et al. 2004), k(Gleason & Shannon 1997), l(Johansson et al. 1994), m(Keith et al. 1991), n(Klamer et al. 2004), o(Wiley & Kennedy 2002), p(Yamada et al. 1999), q(Wiley 1994), r(Burban et al. 2010), s(Curzon & Decker 1998), t(Furuya et al. 1998), u(Gyertyán et al. 2011), v(Leng et al. 2003), w(Marquis et al. 2007), x(Nordquist et al. 2008), y(Cartmell et al. 1999), z(Chiusaroli et al. 2010)

1.4.2.2 Neurobiological effects of acute PCP treatment in rodents The observed behavioural differences following acute PCP treatment in rodents have been attributed to neurotransmission changes similar to those observed in patients with schizophrenia (Adell et al. 2012; Javitt et al. 2012). Microdialysis experiments have shown that acute PCP treatment affects the extracellular levels of several neurotransmitters in a range of brain regions, including the PFC and hippocampus (Table 2). These findings confirm that the diverse pharmacological profile of PCP broadly disrupts neurotransmission across the central nervous system. The caused differences in neurotransmitter systems are similarly found altered in patients with schizophrenia (see 1.1.2). These neurotransmission alterations are of particularl relevance for the model validity, as they have been identified in symptom-relevant brain areas (Table 2). While the majority of microdialysis studies have assessed neurotransmitter levels in the PFC and striatum, post-mortem studies add support for the involvement of the hippocampus in the effect of acute PCP treatment (Zavitsanou et al. 2008). The acute effect of PCP treatment on neurotransmission therefore supports the face validity of the acute PCP rodent model for treatment development studies. The ability of current antipsychotic treatments to alter PCP-induced neurotransmitter level changes has been assessed in a small number of rodent studies. PCP-induced changes to glutamate and serotonin levels in the PFC were amendable by current antipsychotics (Table 2). However, the effect of these antipsychotics on hippocampal neurotransmitter levels following PCP-treatment is unknown. The lack of available data is possibly due to the limited effect of current antipsychotics on hippocampus- associated schizophrenia symptoms such as long-term memory impairments (Boyer et al. 2007; Désaméricq et al. 2014). Furthermore, the ability of these pharmacological substances to influence GABA levels in the presence of PCP has also not been investigated.

Table 2: Effect of current antipsychotics on neurotransmitter levels following acute PCP treatment in rodents Drug GABA Glutamate Dopamine Serotonin Acetylcholine PCP ↓ PFC (rd) p, u; ↔ (s.c.) s ↑ mPFC (s.c. & i.p.) ↑ NAc (rd) c, (i.p.) f, g, l, v, (s.c.)h; ↔ (i.p.) m ↔ mPFC (rd) a, ↑ (rd) p ↑ NAc (i.p.) m ↑ Str (rd) n; ↔ (rd) q, (i.p.) n b, d, e, j, l, t ↑ mPFC (i.p.) f, i, j, k, l, o, v, (rd) p, u, (s.c.) r ↑ mPFC (i.p.)a, e, f, j, v, ↓ Hip (s.c.) s ↔ PFC(rd) p ↑ Str (rd) n, (i.p.) f, n, v, (s.c.) r; ↔ (i.p.) i, v (s.c.) s ↔ NAc l ↔ Str (i.p.) f, u ↔ Str (rd) n, q, (i.p.) n ↑ NAc (i.p.) v ↑ Hip (s.c.) s Clozapine ˅ mPFC (i.p.) d, e ˅ mPFC (s.c.) a Haloperidol ˅ mPFC (i.p.) d, ~ e ~ mPFC (s.c.) a Olanzapine ˅ mPFC (s.c.) a Prazosin ˅ mPFC (s.c.) a Ritanserin ~ mPFC (i.p.) i ˅ mPFC (i.p.) a Effect direction: ↑ increase, ↔ no change, ↓ reduction; ˅ reduce PCP effect, ~ no change to PCP effect, ˄ increase PCP effect Brain region: mPFC: medial prefrontal cortex, NAc: nucleus accumbense, Str: striatum Administration type: rd: retrodialysis, infusion via microdialysis probe; i.p. intraperitoneal, s.c. subcutaneous References: a(Amargos-Bosch et al. 2006), b(Amitai et al. 2012), c(McCullough & Salamone 1992), d(Abekawa et al. 2006), e(Adams & Moghaddam 2001), f(Etou et al. 1998), g(Jacobsen et al. 2005), h(Javitt et al. 1999), i(Kuroki et al. 1999), j(Li et al. 2010), k(Jentsch et al. 2008), l(Adams & Moghaddam 1998), m(Del Arco et al. 2007), n(Lillrank et al. 1994), o(Maeda et al. 2003), p(Zhu et al. 2004), q(Yamamoto et al. 1999), r(Hagino et al. 2010), s(Martin et al. 1998), t(Baker et al. 2008), u(Yonezawa et al. 1998), v(Millan et al. 1999)

The consequence of prolonged PCP exposure on the expression of receptors and other neurotransmission-related proteins has been studied following several different treatment regimes. Expression levels of GABA-A receptors (Abe et al. 2000; Yu et al. 2002), GAD67 (Amitai et al. 2012), PV (Reynolds et al. 2004; Thomsen et al. 2009; Cochran et al. 2003), NMDA-R (Lindahl & Keifer 2004; Wang et al. 1999) and other GABAergic and glutamatergic system elements have been reported to be altered by different chronic and sub-chronic PCP treatments (Javitt et al. 2012; Mouri et al. 2007). The variation in treatment timing and concentration as well as the use of different species and the inconsistency of the observed differences make the chronic and sub-chronic PCP treatments a less robust schizophrenia model than the acute PCP treatment. However, little data is available on the consequence of acute PCP on neurotransmission-related protein expression levels.

1.5 Summary and aims

Several aspects of the symptomatology and pathophysiology of schizophrenia are not addressed by current antipsychotic drugs. The strong focus of current pharmacological treatments on the dopaminergic system leaves the prominent differences in the GABAergic and glutamatergic system untreated. GABAergic and glutamatergic neurotransmission have been reported to be influenced by NRG1, particularly through its receptor ERBB4. In regards to these considerations, the specific aims of my thesis are to: (1) assess the potential of acute NRG1 application to prevent PCP-induced schizophrenia-like behavioural impairments, (2) study the effect of NRG1 treatment on GABAergic and glutamatergic neurotransmission in the presence of PCP and finally (3) treat frontal cortex networks with NRG1 in the presence of PCP to identify its effect on components of the GABAergic and glutamatergic system relevant to the pathophysiology of schizophrenia.

2 THE EFFECT OF NRG1 ON PCP-INDUCED BEHAVIOURAL IMPAIR- MENTS

2.1 Introduction

Addressing behavioural impairments in schizophrenia-relevant rodent models is fundamental and therefore a common strategy to explore potential antipsychotic abilities of pharmacological treatments (Jones et al. 2011). While it is important for novel antipsychotics to address identified biological alterations in patients with schizophrenia, pharmacological treatments ultimately have to improve the behaviour of patients. Initial drug screening experiments therefore rely on models with quantifiable behavioural differences. A range of interventions, each with demonstrable construct validity from clinical studies, have been developed to model behavioural differences. The suitability of a pharmacological agent can initially be tested in a broad model, like acute administration of PCP to rodents, which is widely used for early treatment identification studies (Adell et al. 2012; Javitt et al. 2012). Successful compounds can then be assessed in a more specialised environment, modelling single disease components while the potential of unsuccessful substances can be doubted early in the drug development process. Acute PCP treatment in rodents has been shown to induce a wide spectrum of schizophrenia-relevant phenotypes, including hyperlocomotor activity and impairment of sensorimotor gating (Gleason & Shannon 1997; Yee et al. 2004). Several studies have therefore assessed the ability of current and novel antipsychotics to reduce or prevent PCP-induced hyperlocomotion and sensorimotor gating impairment (Table 1). Recent reviews of the acute PCP model have supported the justification of these preclinical studies by proposing that improving these PCP-induced impairments has translational validity for the clinical setting (Adell et al. 2012; Javitt et al. 2012; Pratt et al. 2012). Transgenic and knockout animal models with mutations in the NRG1 or ERBB4 genes express sensorimotor gating impairments as well as increased locomotor activity (Chen et al. 2008; Deakin et al. 2009; Duffy et al. 2008; Karl et al. 2007; Kato et al. 2010; Stefansson et al. 2002). While developmental compensatory mechanisms likely contribute to these impairments, these studies indicate that unbalanced NRG1- 30

ERBB4 signalling can have behavioural consequence. Furthermore, overexpressing NRG1 in mice appears to have a different effect on adult neurotransmission compared to NRG1 administration during adulthood (Kwon et al. 2008; Kato et al. 2010), which supports the involvement of acute NRG1-ERBB4 signalling in behaviour- relevant CNS structures. The aim of this first study was therefore to assess potential antipsychotic characteristics of acute NRG1 treatment on schizophrenia-relevant locomotor activity and PPI impairments induced by acute PCP treatment. Using these behavioural paradigms in my study allowed me to set effects of NRG1 treatment in relation to treatment outcomes observed in previous studies for a range of existing antipsychotics (see Table 1). While receptors for NRG1 are expressed in several brain regions (see 1.2.1), they have also been found on muscle tissue (Bersell et al. 2009; Wadugu & Kühn 2012; Zhu et al. 1995). Peripheral NRG1 treatment might therefore directly affect muscle cells, masking possible central nervous system (CNS) responses. To separately ex- amine the effect of central and peripheral NRG1 administration on PCP-induced impairments, I performed independent experiments with intracerebroventricular (i.c.v.) and intraperitoneal (i.p.) NRG1 administration. Several studies have reported that i.p. PCP treatment triggers hyperlocomotion and impairs sensory motor gating in mice (Table 1). I therefore hypothesized that PCP treatment would also induce hyperlocomotion and PPI impairments in the present study. The small number of in vitro and in vivo experiments studying the effect of NRG1 on neurotransmission indicate that NRG1 treatment could alter acute neurotransmission relevant to these behaviours (see 1.2.3). I therefore hypothesized that combined treatment of NRG1 and PCP would reduce the PCP effect on locomotor activity and PPI. Finally, as NRG1 has been reported to readily cross the blood- brain-barrier, I hypothesized that peripheral and central treatment will have similar therapeutic effects.

2.2 Materials and Methods

2.2.1 Subjects

Ten-week old male C57BL/6J mice were purchased from Animal Resource Centre (Perth, WA, Australia) and Australian BioResources (Moss Vale, NSW, Australia). Animals were housed two to four per open-lid cage with minimal environmental enrichment (M1 polypropylene cages with sawdust bedding; Able Scientific). Animals were maintained on a 12 h light/dark cycle (light on at 0600 hours) with food and water available ad libitum in the animal facilities of the University of Wollongong (Wollongong, Australia) and Neuroscience Research Australia (Sydney, Australia). All research and animal care procedures were approved by either the Animal Ethics Committee of the University of Wollongong (AE 11/18) or University of New South Wales (11/80B) and were in agreement with the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’. All steps were taken to ameliorate any suffering in all work involving animals in my studies. All animals were acclimatised to the animal facilities for at least one week prior to behavioural experiments and testing order was pseudo-randomised to account for potential diurnal variations.

2.2.2 Drugs

All compounds were obtained from Sigma (Castle Hill, NSW, Australia) except where specified otherwise. The principal isoforms of NRG1 expressed within the nervous system are those containing the β1-EGF domain, which I therefore selected for my treatment experiments (Meyer & Birchmeier 1994; Shinoda et al. 1997). Drugs were freshly prepared on each treatment day. Recombinant human NRG1 (#396 NRG1-β1/HRG1-β1 EGF domain carrier free; R&D Systems, USA) and PCP were dissolved in either artificial cerebrospinal fluid (ACSF, 119mM NaCl; 2.5mM

CaCl2; 2.5mM KCl; 1.3mM MgSO4; 1mM NaH2PO4; 26.2mM NaHCO3; 11mM glucose, 7.4pH, 290mOsm/kg) for i.c.v. injections or in 0.9% NaCl (vehicle) for i.p. injections. Each mouse received no more than three injections with an inter-test interval of at least six days.

2.2.3 I.C.V. Surgery

Mice destined for i.c.v. treatment underwent surgical implantation of an i.c.v. cannula one week prior to experiments. Animals were anaesthetised with an inhalation 1.5% isoflurane (Delvet, USA)/98.5% oxygen mix. Using a stereotactic frame (Kopf Instruments, USA), a 21G guide cannula (1cm) was implanted into the brain targeting the right lateral ventricle (Bregma -0.25mm rostral, -1mm medial, -1.5mm from top of skull (Paxinos & Franklin 2001)). The guide cannula was fixed to the skull with three anchor screws (2mm, Microbiotech, Sweden) and dental acrylic (Ketac Cem, Halas Dental Limited, Australia). All animals were injected with 0.1ml of a depot-antibiotic substance (Medicam®, Boehriner Ingelheim, Germany) and which was also added to their drinking water (0.1%) for three days following surgery. Dummy cannulae (28G), were inserted into the guide to keep it clean and prevent occlusion. Following surgery, mice were individually housed and their recovery was monitored daily until the behavioural experiments commenced.

2.2.4 Locomotor activity

For locomotor activity assessment during the light phase, each mouse was placed in a round open field chamber (black PVC, grey floor, 40cm diameter, 9lx for a least anxiogenic environment, Figure 4A). After 30min habituation, the mice received an i.c.v and/or i.p. injection according to their treatment scheme and their activity was monitored over a subsequent 30min period (Figure 4B).

A B Treatment (i.p. ± i.c.v.)

Baseline

‐30 min ‐10 min 0 min +30 min Habituation Experiment

Figure 4: Schematic of the open field experiment schedule (A) For the locomotor activity experiments, all animals were placed in a round open field arena and distance travelled was video recorded. (B) Animals were habituated to the arena for 30min, followed by i.p. or i.p. + i.c.v. treatment and further 30min in

the arena. The distance travelled during the 30min post-treatment response time was normalised to the 10min baseline before injections.

Distance was analysed from video recordings using the EthoVision tracking system (Noldus, Netherlands). Activity data were collected in 5min intervals. The average distance 10min before treatment was set as baseline for each animal, with post- treatment intervals normalised to this baseline (Figure 4).

2.2.5 Sensorimotor gating (prepulse inhibition of startle assessment)

Acoustic startle response (ASR) and PPI were tested during the light phase in special mouse startle chambers (SR-Lab: San Diego Instruments, San Diego, USA). Animal enclosures were cleaned with detergent between animals. Startle response was measured as the average amplitude of the startle. The duration of the prepulse was 20ms and 40ms for the startle. The inter-trial interval was averaged over 10–25s. ASR was calculated as the startle amplitude in arbitrary units averaged over the test trial. For ASR habituation, blocks of startle responses were averaged at the beginning, middle and end of the PPI protocol (3 startle response blocks). Percentage of PPI (%PPI) was calculated as [(ASR 120dB – prepulse response) × 100/ASR 120dB].

Treatment (i.p.)

30min 1 week 5min 30min

Habituation Experiment

Figure 5: Schematic of the sensorimotor gating experiment schedule One week prior to the beginning of each experiment, all animals were habituated to the startle chambers for 30min at 70dB white noise. On the experiment day, animals were placed in the startle chamber following i.p. treatment, with experiments com- mencing after an additional 5min habituation period at 70dB.

Following habituation to the startle chamber (30min one week before testing, Figure 5), and a 5min acclimatisation period (70dB background noise) upon test, the protocol was as following: The session started with five 120dB startle pulses after which 34

three startle pulses (70/100/120dB) were presented five times each in a pseudo- randomised order. After this, 75 PPI response trials (prepulse intensities of 74/82/86dB followed by a 120dB startle pulse) were presented five times in a quasi- randomised order employing five different inter-trail stimulus intervals (ISIs) (32/64/128/256/512ms) followed by final five 120dB startle pulses.

2.2.6 Drug treatment scheme for open field and PPI

For the i.p. experiments, all animals (n=8/group) received an i.p. injection of either PCP (3mg/kg), PCP+NRG1 (5, 25 or 50µg/kg), NRG1 (5, 25 or 50µg/kg) or vehicle. For the i.c.v. experiments, all animals (n=6-8/group) received an i.c.v. injection of either NRG1 (3, 30 or 300ng/kg) or ACSF and simultaneously an i.p. injection of PCP (3mg/kg) or vehicle. The PCP concentration of 3mg/kg has been selected based on previous reports on its suitability for the locomotor activity and PPI paradigms without causing unwanted adverse reactions (Gleason & Shannon 1997; Kitaichi et al. 1994; Yee et al. 2004).

2.2.7 Statistical Analysis

Pre-injection habituation locomotor activity and startle response following pulse- only trails were assessed by one-way analysis of variance (ANOVA), and followed by Bonferroni’s multiple comparison post-hoc analysis if applicable. Differences in post-treatment locomotion were analysed using a General Linear Model for repeated measurements, followed by Tukey’s post-hoc analysis when applicable. Differences in startle habituation and PPI, averaged for all ISIs, were assessed using two-way repeated measure ANOVA (with the factors treatment and prepulse intensity or pulse timing), and followed by Bonferroni’s multiple comparison post-hoc analysis where applicable. Outliers were screened as mean ±2 standard deviations and removed.

2.3 Results

2.3.1 NRG1 prevents PCP-induced hyperlocomotion following peripheral administration

Spontaneous locomotion of i.p. treated animals during habituation (mean=101.4 cm/min) did not differ between treatment groups (F(3, 28)=0.43, P=0.73) (Figure 6). Post-injection time (F(5, 140)=2.321, P<0.05) and treatment (F(3, 28)=7.86, P<0.01) affected locomotion and an interaction of time and treatment was observed (F(15, 140)=2.817, P<0.001). PCP administration led to an increase in distance travelled compared to vehicle treatment and remained above vehicle levels for the entire recording period (P<0.001 vs vehicle) (Figure 6). Combining PCP with NRG1 i.p. altered the PCP effect in a concentration-dependant manner (Figure 6). Locomotor activity following combined treatment of PCP and 25µg/kg NRG1 was significantly less than PCP only treatment (P<0.01) and similar to vehicle treated animals (P=0.91) (Figure 6A). PCP combined with 5 or 50 μg/kg NRG1 led to locomotor activity similar to PCP only treated animals (5 μg/kg: P=0.77, 50 μg/kg: P=0.71) (Figure 6B). Furthermore, animals treated with 50µg/kg NRG1 and PCP showed significantly increased locomotor activity compared to vehicle treated animals (P<0.001). Mice treated with different concentrations of NRG1 i.p. showed no difference in locomotor activity compared to vehicle treated animals (P>0.9 for 5, 25 and 50 μg/kg NRG1 vs vehicle) (Figure 6B).

Figure 6: Peripheral administered neuregulin 1 (NRG1) prevents PCP-induced hyperlocomotion After habituation, mice (adult male, n=8/group) received intraperitoneal (i.p.) treatment and were returned to the open field arena. Locomotion after i.p. treatment was normalised to individual baseline performance. Mice were administered veh (0.9% NaCl), PCP (3 mg/kg). NRG1+PCP (5, 25 or 50μg/kg NRG1 + 3mg/kg PCP) or NRG1 (5, 25 or 50μg/kg). (A) Post-treatment distance moved during 5 min intervals following NRG1 (25μg/kg), NRG1 (25μg/kg) + PCP, PCP or Veh is presented as mean percentage difference to baseline locomotion ± SEM. (B) Average distance moved across the post-treatment period is presented as mean percentage difference to baseline locomotion + SEM; NRG1 concentration displayed as μg/kg i.p. **P < 0.01 vs PCP; ###P<0.001 vs Veh.

2.3.2 NRG1 prevents PCP-induced hyperlocomotion following central administration

In the second experiment, spontaneous locomotor activity during habituation (mean=98.3 cm/min) did not differ between treatment groups (F(6, 44)=1.17, P=0.34) when tested one week after i.c.v. surgery. Post-injection time (F(5, 140)=3.84, P<0.01) and treatment (F(3, 28)=14.7, P<0.001) affected locomotion, but no interaction of time and treatment was observed (F(15, 140)=0.771, P=0.71). Post-hoc analysis revealed a significant increase in locomotor activity following PCP treatment compared to vehicle treatment (P<0.001) (Figure 7A), with distance re- maining above vehicle levels for entire post-treatment period. Combined treatment of PCP with NRG1 i.c.v. affected the PCP-induced locomotor activity changes in a dose-dependent manner (Figure 7B). Locomotor activity following combined treatment of PCP i.p. and 30ng/kg NRG1 i.c.v was significantly less than PCP only treatment (P<0.01) and similar to vehicle treated animals (P=0.1) (Figure 7A). However, locomotor activity following PCP with either 3ng/kg or 300ng/kg NRG1 i.c.v. was not different to the PCP only treatment group (P=0.74 and P=0.24 respectively), but significantly increased compared to vehicle treated animals (P<0.01 and P<0.05 respectively) (Figure 7B). Spontaneous locomotor activity following NRG1 i.c.v. treatment was indistinguisha- ble to vehicle treated animals (30 ng/kg: P=0.71; 300 ng/kg: P=0.86) (Figure 7B).

Figure 7: Central administered neuregulin 1 (NRG1) prevents PCP-induced hyperlocomotion After habituation, mice (adult male, n=8/group) received intracerebroventricular (i.c.v.) and intraperitoneal (i.p.) treatment and were returned to the open field arena. Locomotion after i.c.v. and i.p. treatment was normalised to individual baseline performance. Mice were administered vehicle (artificial cerebrospinal fluid (aCSF) i.c.v. + 0.9% NaCl, i.p.), PCP (aCSF, i.c.v. + 3mg/kg, i.p.), NRG1+PCP (3, 30 or 300ng/kg NRG1, i.c.v. + 3mg/kg PCP, i.p.) or NRG1 (30 or 300ng/kg NRG1, i.c.v. + 0.9% NaCl, i.p.). (A) Post-treatment distance moved during 5 min intervals following NRG1 (30ng/kg i.c.v.), NRG1 (30ng/kg i.c.v.) + PCP, PCP or Veh is presented as mean percentage difference to baseline locomotion ± SEM. (B) Average distance moved across the post-treatment period is presented as mean percentage difference to baseline locomotion + SEM; NRG1 concentration displayed as ng/kg i.c.v. **P < 0.01 vs PCP; #P<0.05, ##P<0.01, ###P<0.001 vs Veh.

2.3.3 NRG1 prevents PCP-induced deficit of prepulse inhibition of startle

Increase in startle pulse intensity resulted in an increased startle reflex response across all treatments as revealed by one-way ANOVA for repeated measures (startle pulse intensity: F(2, 116)=188.0, P<0.001; 70dB=9.1±0.6, 100dB=22.5±1.2, 120dB=68.7±5.49) (Figure 8A). Two-way ANOVA for repeated measures revealed a significant effect of time (F(2, 116)=23.29, P<0.001) but not treatment (F(7, 58)=1.845, P=0.1), nor an interaction for time x treatment (F(14, 116)=0.62, P=0.84) on startle reflex habituation throughout the experiment session (Figure 8B). Critically, there was no treatment effect on startle reflex response in pulse-only trials when averaged across the experiment session (F(7, 58)= 1.502, P=0.2) (Figure 9).

Using a two-way ANOVA for repeated measures, I identified a significant effect of treatment (F(6, 45)=5.554, P<0.001) and prepulse intensity on PPI (F(2, 90)=371.4, P<0.001), but no interaction between treatment and prepulse intensity (F(12, 90)=1.128, P=0.35). Post-hoc analysis revealed a significant reduction in PPI following PCP treatment for each tested prepulse intensity compared to vehicle (74dB: 14.2±3.1%, P<0.01; 82dB: 36.1±4%, P<0.001; 86dB: 42±4%, P<0.01) (Figure 9A). Combined treatment of NRG1 and PCP had a concentration dependent effect on PCP-induced PPI disruption. PPI following combined PCP and 25µg/kg NRG1 treatment was significantly higher than following PCP only treatment for 12dB and 16dB prepulse intensity (82dB: 51.5±3.4%, P<0.01; 86dB: 54.1±3%, P<0.05 vs PCP) (Figure 9B). PPI following combined PCP and 50µg/kg NRG1 differed to PCP-only treatment at 82dB prepulse intensity (51.3±3.8%, P<0.05 vs PCP) (Figure 9B). PPI following combined treatment of PCP with 5µg/kg NRG1 did not differ for any prepulse intensities to PCP-only treatment (P>0.8 for each prepulse intensity vs PCP) (Figure 9B). NRG1 treatment at 25µg/kg and 50µg/kg had no effect on prepulse inhibition compared to vehicle (Figure 9B). 5µg/kg NRG1 treatment impaired prepulse inhibition at 74dB and 82dB prepulse intensity (74dB: 19.2±3.2%, 82dB: 44.8±3.1%, P<0.05 vs vehicle each) (Figure 9B).

Veh PCP (3 mg/kg) NRG1 (5 µg/kg) + PCP NRG1 (25 µg/kg) + PCP NRG1 (50 µg/kg) + PCP NRG1 (5 µg/kg) NRG1 (25 µg/kg) NRG1 (50 µg/kg)

Figure 8: Startle response to pulse intensity and pulse timing Mice (adult male, n=7-8/group) received intraperitoneal (i.p.) treatment and were placed in the startle chambers for 30min experiments. Mice were administered Veh (0.9% NaCl), PCP (3mg/kg), NRG1+PCP (5, 25 or 50µg/kg NRG1 + 3mg/kg PCP) or NRG1 (5, 25 or 50µg/kg NRG1). (A) Startle development to pulse-only trials at three different pulse intensities, presented as startle + SEM. (B) Startle habituation to 120dB pulse-only trials at the beginning, middle and end of the experiment session, presented as startle + SEM.

Startle (arbituary value) Prepulse inhibition (%)

Figure 9: Peripheral neuregulin 1 (NRG1) injection prevents PCP-induced prepulse inhibition (PPI) impairment Mice (adult male, n=7-8/group) received intraperitoneal (i.p.) treatment with veh (0.9% NaCl), PCP (3mg/kg), NRG1+PCP (5, 25 or 50μg/kg NRG1 + 3mg/kg PCP) or NRG1 (5, 25 or 50μg/kg) before being placed into the startle chamber. Acoustic startle reactivity was tested in 100 trials following a 5 min acclimatisation period to white noise (70 dB). (A) Treatments had no effect on average startle response to a 120 dB pulse (One-way ANOVA, F(7, 58)=1.502, P=0.18). Values are presented as mean of startle intensity + SEM. (B) Treatment effect on PPI following different prepulse intensities. Values are presented as mean percentage of PPI + SEM. *P<0.05, **P < 0.01 vs PCP; #P<0.05, ##P<0.01, ###P<0.001 vs veh.

2.4 Discussion

My behaviour experiments showed that PCP-induced hyperlocomotor activity and PPI impairments were prevented by combined treatment with NRG1 in a dose- dependent manner. Animals treated with NRG1-only showed no behavioural differences compared to vehicle treated animals.

2.4.1 NRG1 treatment prevents PCP-induced hyperlocomotion

I observed an increase in locomotor activity following acute PCP treatment (i.p. experiment: 1.54 times, i.c.v. experiment 2.8 times), which is within the reported range of existing publications for the same 3mg/kg i.p. PCP concentration (1.3 times (Mouri et al. 2011), 2 times (Allen et al. 2011; Nagai et al. 2003), 3 times (Gleason & Shannon 1997)). The increase in distance travelled following PCP treatment was larger for animals which had undergone i.c.v. surgeries prior to testing than those only receiving i.p. injections (Figure 6 and Figure 7). PCP-induced hyperlocomotion has been shown to be dose, strain and environment dependent, thus the extent can differ between studies (Koseki et al. 2012; Mouri et al. 2011). It is therefore possible that surgery-related handling made the animals more susceptible to the psychotic-like actions of PCP. Currently available antipsychotics have shown to be effective in reducing PCP- induced hyperlocomotor activity following 2-7.5 mg/kg PCP i.p. (Table 1). I therefore selected 3mg/kg PCP i.p. as the concentration of choice for reliably inducing hyperlocomotor activity effect without causing any of the reported adverse effects on general motor function as seen with higher PCP concentrations (Gleason & Shannon 1997; Kitaichi et al. 1994; Marquis et al. 2007). Several experiments have assessed the effect of antipsychotics on PCP-induced hyperlocomotion when administered prior to the PCP injection (Abdul-Monim et al. 2003; Abekawa et al. 2002). While this procedure is likely due to the pharmacokinet- ics of the assessed substance, the translational value of such a treatment regime is limited. The behavioural consequences of acute PCP treatment are considered to simulate an acute psychotic episode (Adell et al. 2012; Javitt et al. 2012). I therefore decided to administer NRG1 and PCP simultaneously to assess the effect of NRG1 from the onset of the PCP interference. 43

When testing NRG1 for its efficacy against PCP, I observed that i.p. and i.c.v. NRG1 treatments were successful in preventing PCP-induced impairments in the open field arena (Figure 6 and Figure 7). NRG1 at 25µg/kg i.p. and 30ng/kg i.c.v. completely prevented the induction of hyperlocomotion by PCP. The effectiveness of NRG1 at these concentrations against PCP-induced impairments is similar to findings in publications where existing antipsychotics were used (see Table 1). The NRG1 treatment effect was concentration dependent as the high (50µg/kg i.p. or 300ng/kg i.c.v.) or low (0.5 and 5µg/kg i.p. or 3ng/kg i.c.v.) doses of NRG1 did not prevent PCP- induced hyperlocomotion (Figure 6 and Figure 7).

The comparable outcomes of i.p. and i.c.v. NRG1 administration against PCP- induced hyperlocomotion suggest that central rather than peripheral NRG1 receptors mediated the beneficial treatment effects. Both administration methods completely prevent i.p. PCP from inducing hyperlocomotion, which suggests a rapid uptake of i.p. NRG1 across the blood-brain-barrier in treatment-relevant quantities. This finding is in accordance with existing studies, which have shown that peripherally administered human recombinant NRG1 can readily cross the blood-brain-barrier in mice (Kastin et al. 2004; Rösler et al. 2011; Kato et al. 2011; Cui et al. 2013). My results provide evidence for the suitability of peripheral NRG1 administration with behavioural consequence, which is further supported by the observed comparable timing of i.p. and i.c.v. injected NRG1 in preventing the PCP-induced hyperlocomotion (Figure 6 and Figure 7).

2.4.2 NRG1 treatment prevents PCP-induced PPI impairments

I observed a reduction in PPI following acute PCP treatment for each prepulse intensity, with values similar to existing publications (Adage et al. 2008; Fejgin et al. 2007; Klamer et al. 2004). Several currently available antipsychotics have shown to be effective in reducing the PCP-induced PPI impairments in rodents following 1-5 mg/kg PCP i.p. (Table 1). However, the effects have not always been reproducible nor are all current antipsychotics effective against PCP-induced PPI impairments (Fejgin et al. 2007; Johansson et al. 1994; Keith et al. 1991; Wiley & Kennedy 2002). I selected 3mg/kg PCP i.p. as optimal concentration for reliably inducing a PPI deficit (Curzon & Decker 1998; Johansson et al. 1994) and to remain compara- 44

ble to my open field experiments. The startle response during the single pulse trials was not affected by PCP, confirming the suitability of 3mg/kg PCP for my experiment as general reflex mechanisms were not impaired (Figure 8). In line with my locomotor activity experiments, I opted for simultaneous treatment of NRG1 and PCP to identify potential antipsychotic-like actions of NRG1, even though several sensorimotor gating studies administered the drug of interest prior to PCP (Bakshi & Geyer 1997; Bakshi & Geyer 1995; Ballmaier et al. 2007; Fejgin et al. 2007; Johansson et al. 1994; Klamer et al. 2004; Wiley & Kennedy 2002; Yamada et al. 1999). Startle response was not affected by NRG1 injections, confirming the suitability of the selected NRG1 concentrations (Figure 8). This confirmation is of critical importance as several current antipsychotics have shown to dose-dependently reduce the startle reflex in pulse-only trials (Fejgin et al. 2008). In my study, simultaneous administration of NRG1 and PCP prevented the PCP- induced PPI impairments (Figure 9). The treatment effect was concentration dependent as PPI following high (50µg/kg i.p.) and low (5µg/kg i.p.) doses of NRG1 in the presence of PCP was impaired (Figure 9). NRG1 at 25µg/kg i.p. prevented the reduction of PPI by PCP for the 12 and 16dB prepulse intensities. The effectiveness of NRG1 at this concentration against PCP-induced impairments is similar to published effects of existing antipsychotics (see Table 1). In line with my locomotor activity experiments, startle response and PPI were not affected by 25μg/kg NRG1, the most effective concentration against PCP, throughout the experiment (Figure 8). Furthermore, startle response at the start and end of the experiment were similar for 25μg/kg NRG1, indicating that NRG1 treatment had no delayed negative effect on relevant CNS circuits (Figure 8). Preventing sensorimotor gating impairment in this widely used acute psychosis model suggests a critical potential for NRG1 treatment to beneficially address a common schizophrenia symptom present in patients (Swerdlow & Geyer 2011). The observed effect of NRG1 on PCP-induced PPI reduction provides additional support for the suitability of peripheral NRG1 treatment in targeting CNS structures relevant to the schizophrenia pathophysiology (Swerdlow et al. 2001).

2.4.3 Behavioural consequences of increased NRG1-ERBB signalling – treatment implications

The open field and prepulse inhibition experiments have shown that NRG1 administration can prevent behavioural impairments otherwise induced by PCP. This supports my initial hypothesis suggesting that NRG1 treatment can alter behaviour- relevant neurotransmission. Since I am the first one to assess the behavioural implication of acute NRG1 administration on locomotor activity and sensorimotor gating relevant to schizophrenia, it is also critical to study the effects of NRG1-only treatment on the examined behaviours. This is of particular interest since several current antipsychotics can have adverse effects on behavioural performance in healthy animals (Adage et al. 2008; Amitai et al. 2007; Bakshi & Geyer 1995; Fejgin et al. 2007; Geyer et al. 2001; Gleason & Shannon 1997; Johansson et al. 1995; Johansson et al. 1994; Linn et al. 2003; Mutlu et al. 2011; Nordquist et al. 2008; Okuyama et al. 1999; Swerdlow et al. 1991; Wiley 1994). The results from the i.c.v. and i.p. experiments show that acute NRG1 administration at different concentrations has no effect on spontaneous locomotor activity (Figure 6 and Figure 7), which had also been indicated by a previous study (Mahar et al. 2011). Furthermore, neither startle response nor PPI differed in animals treated with the most effective NRG1 concentrations (Figure 8 and Figure 9). The outcome of my experiments suggests that NRG1 has a system-dependent effect within a specific concentration range. Administration of NRG1 appears capable of improving specifically impaired conditions at concentrations which show no effect on the healthy phenotype. In vitro experiments have so far provided conflicting information regarding the effect of acute NRG1 on neuronal activity. Some studies suggest that NRG1 has little effect on baseline activity of neuronal networks (Kwon et al. 2005; Kwon et al. 2008; Fisahn et al. 2009), while others show alterations to neurotransmitter signalling (such as GABA and glutamate) even at baseline following NRG1 application (Mitchell et al. 2013; Yin et al. 2013). The concentration-dependent effect of NRG1 signalling observed in the present study might contribute to the varying effect of NRG1 on neuronal activity.

In my experiments I explored the dose-dependent effect of acute NRG1 treatment on locomotor activity and sensorimotor gating. Existing antipsychotics show clear dose- response effects in preventing PCP-induced impairments (Fejgin et al. 2007; Gleason & Shannon 1997; Johansson et al. 1994), and more importantly in reducing symptom severity in patients with schizophrenia (Kinon et al. 2004; Miyamoto et al. 2012). I observed that the efficiency of NRG1 treatment in preventing PCP-induced impairments is also dose-dependent, following a u-shaped dose-response curve in the locomotor activity and PPI paradigms (Figure 10), with 25μg/kg i.p. and 30ng/kg i.c.v being the most effective dose (Figure 6, Figure 7 and Figure 9). To identify the precise NRG1 treatment window, additional treatment concentrations are required to support the indicated u-shaped dose-response curve. NRG1 has been proposed to cross the blood-brain barrier via a saturable active transport mechanism (Kastin et al. 2004). In the peripheral treatment paradigms, the effect of 25µg/kg NRG1+PCP i.p differs to 50μg/kg NRG1+PCP on locomotor activity and PPI (Figure 6, Figure 9). This observed dose-response to NRG1 indicates that 25µg/kg NRG1 i.p. might not fully exhaust the active transport mechanism which enables blood-brain-barrier crossing.

AB 200

150 150

100 100 PCP effect

50 50 Locomotion deviation to deviation Locomotion vehicle (% of PCP effect) of(% vehicle 0 0 Vehicle 52550 330300 (µg/kg i.p.) (ng/kg i.c.v.) NRG1 co-treatment

Figure 10: Dose-response effect of neuregulin 1 (NRG1) vs PCP on locomotor activity U-shaped dose-response curves: Locomotion depicted as deviation from vehicle treatment with varying doses of i.p. NRG1 (A, µg/kg) or i.c.v. NRG1 (B, ng/kg) administered for NRG1+PCP co-treatment. Shades indicate vehicle and PCP-group SEM respectively. For detailed dose-effects see Figure 6 and Figure 7.

Additionally to the NRG1 treatment concentration relevance from my study, treatment duration also appears to be relevant to behavioural regulation. Knockout mouse studies have shown that constant increase as well as decrease in NRG1 production can cause hyperlocomotion and PPI impairments (Luo et al. 2013; Wen et al. 2010; Yin et al. 2013). Altered NRG1 signalling during development is likely to trigger compensatory mechanisms, which potentially contribute to the observed behavioural impairments in knockout animals. However, Rescuing NRG1 overexpression during adulthood also rescued PPI and locomotion impairments in some transgenic animals (Luo et al. 2013; Yin et al. 2013), suggesting that acute NRG1 signalling is relevant for behaviour regulation.

Remarkably, a dose-response effect similar to the present study was recently reported in a phase II clinical trial of NRG1. Patients with chronic heart failure have been treated successfully with recombinant NRG1, where the medium dose showed the highest efficacy and the least side effects, with a lasting therapeutic effect beyond treatment termination (Gao et al. 2010; Jabbour et al. 2011). The early nature of these clinical trials with intravenous NRG1 administration requires further exploration of NRG1-concentration dependent effects.

In summary, my behavioural results provide the first evidence for antipsychotic-like characteristics of NRG1 treatment by demonstrating that NRG1 has neuromodulatory properties with schizophrenia-relevant behavioural consequences. By improving PCP-induced hyperlocomotion and PPI impairments NRG1 shows characteristics of existing antipsychotics without causing observable adverse reactions. Further behaviour experiments with longer NRG1 treatment durations are required to assess its potential suitability for the clinical setting. However, in my next study I investigated the effect of NRG1 on GABAergic and glutamatergic neurotransmission, since both have been implicated in locomotor activity regulation and sensorimotor gating (Asi- nof & Paine 2013; Geyer et al. 2001; Swerdlow et al. 2001; Tanner 1979; Viggiano 2008; Yeomans et al. 2010).

3 NEUREGULIN 1 TREATMENT AFFECTS GABA AND GLUTAMATE RELEASE AFTER PCP CO-TREATMENT IN FREELY MOVING MICE

3.1 Introduction In my behaviour experiments, I have shown that NRG1 treatment improves schizophrenia-relevant behavioural impairments in mice. These findings indicate for the first time an antipsychotic-like ability of NRG1 treatment in the PCP rodent model of schizophrenia. PCP has a range of binding affinities for neurotransmission-relevant receptors and channels as described in detail in chapter 1.4.2.2. The effect of PCP treatment on the GABAergic, glutamatertic and dopaminergic system in schizophrenia-relevant brain areas has been proposed to model aspects of neurotransmission changes during an acute psychotic episode experienced by patients with schizophrenia (Adell et al. 2012; Javitt et al. 2012). Differences in the GABAergic, glutamatergic and dopaminergic system have been identified in the PFC and hippocampus of patients with schizophrenia by in vivo neuroimaging and post-mortem studies (see 1.1.2 for details). Additionally, the activity of both regions as well as the interaction between the PFC and the hippocampus in patients with schizophrenia also appear to differ compared to healthy controls (El- lison-Wright & Bullmore 2009; Kumra et al. 2004; Olbrich et al. 2008; Scheel et al. 2013; White et al. 2007) (recently reviewed by (Del Arco & Mora 2009; Javitt et al. 2008; Meyer-Lindenberg 2010; Schmitt et al. 2011)). Effective pharmacological treatment options would therefore need to address impairments in neurotransmission in both regions. Current antipsychotics, however, primarily address differences in the dopaminergic system with only indirect effects on GABAergic and glutamatergic system alterations (see 1.1.3 for details). This restricted pharmacological profile has been discussed as a likely reason for their limited treatment benefit on a range of schizophrenia symptoms (Vinkers et al. 2010). NRG1 treatment has shown to affect GABA release and GABA receptor expression levels in acute PFC and hippocampus in vitro slice preparations (Allison et al. 2011; Wen et al. 2010; Woo et al. 2007). Furthermore, NRG1 signalling has also been shown to alter glutamatergic neurotransmission in vitro (see 1.2.3.2 for details). 49

Assessing the effect of NRG1-ERBB4 signalling on acute neurotransmission in vivo has shown that an increase in NRG1 can lead to increased extracellular dopamine levels in the hippocampus, with no data available on GABA or glutamate levels (see 1.2.3 for details). Advancing on my behavioural findings, the aim of this study was to assess the effect of NRG1 treatment on acute neurotransmitter levels in vivo. Specifically, the effect of NRG1 treatment on extracellular levels of GABA, glutamate and glycine in the PFC and hippocampus and critically, the ability of NRG1 to alter extracellular neurotransmitter levels in both brain regions in the presence of PCP.

The existing literature strongly supports the broad effect of acute PCP treatment on neurotransmission in schizophrenia-related brain regions in rodents (Table 2). I have therefore selected the acute PCP treatment to model schizophrenia-like neurotransmission alterations in vivo. Furthermore, as outlined above, the excitatory-inhibitory balance of the GABAergic-glutamatergic systems in the PFC and hippocampus are important for normal behaviour processes. I therefore administered NRG1 and PCP directly into these brain areas via the microdialysis probe (retrodialysis) to study the treatment effect on local extracellular transmitter levels in freely moving mice. The hippocampus is subdivided into functionally discrete subregions, with distinctly different afferent and efferent connections to other cortical and subcortical structures (Fanselow & Dong 2010). The CA1 region of the dorsal hippocampus forms mon- osynaptic connections with the PFC (Fanselow & Dong 2010; Small et al. 2011). Critically, activity levels of the CA1 region correlate with symptom severity of patients with schizophrenia, which supports the reported post-mortem protein expression differences in this region (Harrison 2004; Schobel et al. 2009). Consequentially, I selected the dorsal CA1 region for the neurotransmitter analysis in my microdialysis experiment. The observed effect of NRG1 on local neurotransmitter levels in vitro is likely mediated by ERBB receptors expressed on GABAergic interneurons and astrocytes in the PFC and hippocampus (Longart et al. 2007; Neddens et al. 2011; Sharif et al. 2009). Based on these findings I hypothesised that NRG1 administration into the PFC and hippocampus would alter the extracellular levels of GABA, glutamate and glycine.

Peripherally and centrally administered PCP has been shown to affect several neurotransmitter levels in the PFC and hippocampus of rodents (Table 2). While no data exists on changes to glycine levels and effects on glutamate levels are inconclusive, I hypothesised that PCP administration would increase GABA, with a potential to effect both glutamate and glycine levels. Neurotransmission changes in both brain regions have been suggested to contribute to hyperlocomotion and PPI impairment (Bast & Feldon 2003; Howland et al. 2004; Jodo et al. 2005; Viggiano 2008). My behaviour data has shown that NRG1 treatment at the same time as PCP treatment can prevent the behavioural changes otherwise induced by PCP. Consequentially, I hypothesised that NRG1 treatment would reduce the effect of PCP on neurotransmitter level changes.

3.2 Materials and Method

3.2.1 Subjects

Ten-week old male C57BL/6J mice were purchased from Australian BioResources (Moss Vale, NSW, Australia). Animals were housed two per cage with minimum enrichment (M1 polypropylene cages with sawdust bedding; Able Scientific) and maintained on a 12h light/dark cycle (light on at 0600 hours) with food and water available ad libitum in the animal facility of the University of Wollongong (Wollon- gong, Australia). All research and animal care procedures were approved by the An- imal Ethics Committee of the University of Wollongong (AE10/07) and were in agreement with the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’. All steps were taken to ameliorate any suffering in all work involving animals in my studies.

3.2.2 Drugs

All compounds were obtained from Sigma (Castle Hill, NSW, Australia) except where specified otherwise. Drugs were freshly prepared on each treatment day. Re- combinant human NRG1 (NRG1-β1/HRG1-β1 EGF domain CF; R&D Systems, USA) and PCP were dissolved in ACSF (119mM NaCl; 2.5mM CaCl2; 2.5mM KCl; 1.3mM MgSO4; 1mM NaH2PO4; 26.2mM NaHCO3; 11mM glucose, 7.4pH, 290mOsm/kg) for retrodialysis.

3.2.3 Microdialysis and amino acid quantification

3.2.3.1 Surgery and housing All animals were acclimatised for one day prior surgery to their purpose built microdialysis cages (25x25cm, grey PVC with clear front, 40cm high). Animals were anaesthetised with an inhalation 1.5% isoflurane (Delvet, USA) /98.5% oxygen mix. Using a stereotactic frame (Kopf Instruments, USA) a single guide cannula (MAB 6.14.IC Microbiotech/SE, Sweden) was implanted into the brain to accommodate a microdialysis probe targeted at either the medial PFC or dorsal CA1

(mPFC/prelimbic & infralimbic cortex: 1.7mm rostral to bregma, 0.5mm medial to bregma, 1.5mm beneath the surface of the skull; dCA1: 2mm caudal to bregma, 1.5mm medial to bregma, 2mm beneath the surface of the skull (Paxinos & Franklin 2001)) (Figure 11). The guide cannula was fixed to the skull with three anchor screws (2mm, Microbiotech, Sweden) and dental acrylic cement (Ketac Cem, Halas Dental Limited, Australia). A dummy (Microbiotech, Sweden) was inserted into the guide cannula to keep it clean and prevent occlusion. Following surgery, all animals were injected with 0.1ml of a depot-antibiotic substance (Medicam®, Boehriner Ingelheim, Germany) and additionally received the antibiotic in their drinking water (0.1%) for three days following surgery. Mice were individually housed in purpose- built microdialysis cages for all experimental procedures which allowed free access to food and water.

Figure 11: Microdialysis probe placements Microdialysis guide cannulas were placed either in the prefrontal cortex (PFC, prelimbic (PrL) and infralimbic (IL) segments) or into the dorsal hippocampus (cornuammonis 1 (CA1) segment) in anaesthetized mice. One week post-surgery, microdialysis probes were inserted into the respective guide cannulae 2.5h before the first

treatment (Figure 12) and attached to inflow (blue) and outflow (purple) tubing. Structural diagrams adapted from (Paxinos & Franklin 2001).

3.2.3.2 Dialysis sampling After seven days of recovery, the microdialysis probe (1mm membrane length, 15kDa cut off, MAB 6.14.1, Microbiotech, Sweden) was inserted into the guide cannula and lowered into the respective target area of the mPFC/dCA1 2.5h before the first treatment. It was connected via low volume in- and outflow tubing (FEP-tubing, 4001004, Microbiotech, Sweden, fluid swivel 375/D/22QM, Instechlabs, USA) via a counter-balanced lever arm to a syringe pump (Harvard Instruments, USA). This allowed the animals to freely move during sample collection. The probes were perfused with ACSF at a ﬂow rate of 1μl/min. Microdialysis samples (10µl, 1µl/min) were collected 2.5h after probe insertion, with treatment and sampling performed according to Figure 12. Animals received 10min retrodialysis infusion with either NRG1 (10nM), PCP (10µM) or NRG1+PCP, separated by 2.5h of ACSF each, with samples being collected for 60min from the start of each treatment (Figure 12). Treatment order was constant for all animals to allow the comparison of molecular responses.

Probe Insertion & ACSF Flow

NRG1 (10 nM) infusion PCP (10 µM) infusion NRG1 + PCP infusion

2.5h 2.5h 2.5h B -10 0 10 20 30 40 50 B -10 0 10 20 30 40 50 B -10 0 10 20 30 40 50

Figure 12: Microdialysis treatment and sampling scheme for prefrontal cortex (PFC) and dorsal hippocampus CA1 (dCA1) Mice (adult male, n=12/treatment) received microdialysis probes one week post- surgery into the respective target area of the PFC/dCA1 2.5h before the first treatment. The probes were perfused with artificial cerebrospinal fluid (ACSF) at a ﬂow rate of 1 μl/min. After a 2.5h perfusion period without sampling, samples were collected every 10min, with two baseline samples prior the infusion of each treatment and six samples after each treatment, with a 2.5h inter-treatment interval. Neuregulin 1 (NRG1, 10nM), Phencyclidine (PCP, 10µM) and NRG1+PCP were infused through the microdialysis probe for 10min into the respective brain area.

3.2.3.3 Potassium depolarization and location confirmation After the final experiment ACSF containing high potassium (50mM KCl (Stanford et al. 2000)) was infused and sampled to confirm depolarisation evoked neurotransmitter release from the surrounding tissue. Following this, the brains were removed, fro- zen in liquid nitrogen and stored at -80C until cut in 50µm sections using a cryostat (Leica, USA). The placement of the microdialysis probe was confirmed by Nissl staining of the relevant coronal sections. Animals with incorrect probe placements and unresponsive to potassium treatment were excluded from the analysis.

3.2.3.4 GC-MS/MS sample processing and analysis Derivatisation followed a previously published method (Eckstein et al. 2008). Sam- ples were defrosted on ice, supplemented with heavy isotopes (20pmole GABA-d6, 200pmole Glycine-C13-2, N15, 2pmole Glutamate-d5), and diluted in acetoni- trile/methanol (1:1) before centrifugation. The supernatant was dried off at 37°C under nitrogen before derivatisation with pentafluoropropionic acetic anhydride and 2,2,3,3,3-pentafluoro-1-propanol (PFAA and PFPOH, PM Separations PTY, Capala- ba, Australia). Following derivatisation, samples were resuspended in toluene for analysis. Derivatised samples were analysed on an Agilent 7001B mass selective detector in- terfaced with an Agilent 7890A gas chromatograph. The injection port, MS source and GC-MS/MS interface were kept at 250°C, 150°C and 250°C respectively. Sepa- rations were carried out on a fused silica capillary column (20m x 180µm i.d.) coated with cross-linked 5% phenylmethylsiloxane (film thickness 0.18µm), (Agilent, HP5MS), with Helium as the carrier gas. Column temperature was initially held at 60°C for 0.5min, increased to 210°C at 15°C/min, further increased to 300°C at 30°C/min, and held for 4min. Specific multiple reaction monitoring (MRM) transitions were monitored using the negative chemical ionization mode at 70eV with ar- gon as the reagent gas (1.25ml/min) and the ion source and quadrapole at 150°C. Se- lected MRM transitions were performed to monitor the GABA ion transition at m/z 190.0 → 109.0, qualifier transitions at m/z 341.0 → m/z 208.0 and m/z 321.0 → m/z 301.0, and m/z 327 → m/z 306 for the corresponding isotopic d6 labelled internal standard (Figure 13). Glutamate ion transition was detected at m/z 537.0 → m/z

217.0, qualifier transitions at m/z 537.0 → m/z 313.0 and m/z 379.0 → m/z 188.0, and m/z 542 → m/z 221 for the corresponding isotopic d5 labelled internal standard (Figure 14). Glycine ion transition was detected at m/z 333.0 → m/z 313.0, qualifier transitions at m/z 333.0 → 200.0 and m/z 365.0 → 232.0, and m/z 336.0 → m/z 316.0 for the corresponding isotopic labelled internal standard (Figure 15). The area under the curve for peaks of GABA, glutamate and glycine were normalised by their respective internal standard peaks, followed by normalisation to the relevant baseline samples.

3.2.4 Statistical Analysis

Treatment effect in neurotransmitter levels was assessed using Linear Mixed Model (LMM) on repeated measurements for each brain area and transmitter. Single time point differences were assessed by one way ANOVA for repeated measurements, followed by Tukey’s post-hoc analysis when appropriate. Analyses were performed using SPSS 19.0. Statistical significance was accepted at P<0.05, data are presented as means±SEM. Outliers were screened as mean ±2 standard deviations and removed.

Figure 13: Mass spectrometry analysis of GABA Representative spectra of one cerebrospinal fluid sample following derivatization and gas chromatography/mass spectrometry detection in negative chemical ionization. Multiple-reaction monitoring (MRM) segments were defined for the amino acids of interest. Segment four spans the acquisition time for GABA at 5min (A), monitoring the reactions for GABA (B) and GABA-d6 (C). The area under the curve for the product ion of (D) GABA and (E) GABA-d6 was used for the data analysis. Data presented as counts over acquisition time (A, D, E) and counts over mass (B, C).

Figure 14: Mass spectrometry analysis of glutamate Representative spectra of one cerebrospinal fluid sample following derivatization and gas chromatography/mass spectrometry detection in negative chemical ionization. Multiple-reaction monitoring (MRM) segments were defined for the amino acids of interest. Segment six spans the acquisition time for glutamate at 6min (A), monitoring the reactions for glutamate (B) and glutamate-d5 (C). The area under the curve for the product ion of (D) glutamate and (E) glutamate-d5 was used for the data analysis. Data presented as counts over acquisition time (A, D, E) and counts over mass (B, C).

Figure 15: Mass spectrometry analysis of glycine Representative spectra of one cerebrospinal fluid sample following derivatization and gas chromatography/mass spectrometry detection in negative chemical ionization. Multiple-reaction monitoring (MRM) segments were defined for the amino acids of interest. Segment one spans the acquisition time for glycine at 3.9min (A), monitoring the reactions for glycine (B) and glycine-d3 (C). The area under the curve for the product ion of (D) glycine and (E) glycine-d3 was used for the data analysis. Data presented as counts over acquisition time (A, D, E) and counts over mass (B, C).

3.3 Results

After confirming the microdialysis probe location and K+ response for each animal, I excluded the samples of two animals from the PFC group and one animal of the dCA1 group.

3.3.1 Extracellular levels of GABA, glutamate and glycine in the prefrontal cortex

3.3.1.1 Treatment effect on extracellular GABA levels in the PFC A one-way repeated measure ANOVA confirmed that the baseline GABA values were not different before each treatment (F(1.4, 9.8)=1.16; P=0.33). The LMM analysis revealed that time (F(5, 121.8)=2.74; P<0.05), but not treatment (F(3, 123.9)=0.56; P=0.64) significantly effected extracellular GABA levels in the PFC, with no interaction between both (F(15, 122.3)=1.21; P=0.27) (Figure 16). Further analysis of individual time point differences by one-way ANOVA revealed that PCP administration resulted in a significant reduction of GABA levels 20min after treatment (37.6±8.4% of baseline, P<0.05) (Figure 16A). Interestingly, this reduction was not observed following combined NRG1 and PCP treatment (Figure 16A). NRG1 treatment caused an increase in extracellular GABA levels 10min after treatment (194.2±47.8% of baseline, P<0.05) (Figure 16A). Furthermore, I assessed GABA level differences across the treatment via LMM analysis of the area under the curve (AUC). None of the three treatment options altered extracellular GABA levels compared to baseline-equivalent levels (F(3, 18.7)=0.89; P<0.46) (Figure 16B). Of note, in a biased approach, I observed an increase in extracellular GABA levels during PCP infusion in three animals (250±34.6% of baseline, P<0.05), which was prevented by combined NRG1 and PCP treatment (84.3±23.7% of baseline, P=0.66).

Figure 16: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular GABA levels in the prefrontal cortex of freely moving mice. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the prefrontal cortex by retrodialysis. GABA levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. #P<0.05 PCP vs baseline; &P<0.05 NRG1 vs baseline.

3.3.1.2 NRG1 treatment reduced extracellular glutamate levels in the PFC A one-way repeated measure ANOVA confirmed that the baseline glutamate values were not different before each treatment (F(1.4, 9.8)=2.69; P=0.13). The LMM analysis revealed that treatment (F(3, 145.8)=11.4; P<0.001) and time (F5, 144.4)=3.7; P<0.001) significantly effected extracellular glutamate levels, with no interaction between both factors (F(15, 144.4)=1; P=0.46) (Figure 17A). Detailed analysis showed that NRG1 treatment caused a persistent reduction in extracellular glutamate levels in the PFC throughout the recording period (50.8±12% of baseline, P<0.001) (Figure 17A). Neither PCP (118.3±9.8%) nor combined NRG1 with PCP (101.1±10.5%) treatment had a significant effect on glutamate levels throughout the recording period (Figure 17A). This observation was confirmed by a LMM analysis of the AUC, showing that NRG1 significantly reduced glutamate levels (52.2±9.8% of baseline equivalent, P<0.01), with no difference observed after PCP (114.7±12.2%) or NRG1 and PCP (109.35±14%) treatment (Figure 17B).

Figure 17: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glutamate levels in the prefrontal cortex in vivo. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the prefrontal cortex by retrodialysis. Glutamate levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. &&P<0.01 NRG1 vs baseline equivalent.

3.3.1.3 NRG1 treatment alters extracellular glycine levels in the PFC Baseline extracellular glycine values were not different before each treatment (F(1.2, 8.6)=1.1; P=0.35). Glycine levels in the PFC were significantly affected by treatment (F(3, 145.1)=6.85; P<0.001), but not sampling time (F(5, 144.5)=1.26; P=0.29), as revealed by a LMM analysis (Figure 18A). Post-hoc analysis revealed a significant increase in glycine levels after the onset of the combined NRG1 and PCP treatment (190±29% of baseline, P<0.05) (Figure 18A). Neither PCP (90.6±24.4%, P=1) nor NRG1 (57.1±24.9%, P=0.84) showed a significant effect on glycine levels throughout the recording (Figure 18A). An AUC analysis (F(1.577, 11.04)=3.97; P=0.05), however, identified a significant reduction in overall glycine levels triggered by NRG1 treatment (57.5±9.7% of baseline equivalent, P<0.01) (Figure 18B). The AUC for glycine following PCP (94.24±23.5%) or combined PCP and NRG1 (125.1±10.4%) treatment was similar to the baseline equivalent levels (Figure 18B).

Figure 18: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glycine levels in the prefrontal cortex in vivo. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the prefrontal cortex by retrodialysis. Glycine levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. *P<0.05 NRG1+PCP vs baseline; &&P<0.01 NRG1 vs baseline equivalent.

3.3.2 Extracellular levels of GABA, Glutamate and Glycine in the dorsal hippocampus

3.3.2.1 NRG1 treatment prevents PCP-induced increase in extracellular GABA levels in the dCA1 A one-way repeated measure ANOVA confirmed that the baseline GABA values were not different before each treatment (F(1.8, 19.9)=2.22; P=0.14). GABA levels in the dorsal CA1 region of the hippocampus were not affected by the treatments (F(3, 147.4)=1.39; P=0.25), but varied significantly over the sampling time (F(5, 144.8)=10.6; P<0.001) according to a LMM analysis. Furthermore, I identified a significant interaction between treatment and sample time on GABA levels (F(15, 144.4)=3.34; P<0.001) (Figure 19A). However, detailed analysis by one way ANO- VA showed that GABA levels during the retrodialysis period were significantly affected by the treatment regimens (F(3, 25)=3.35; P<0.05) (Figure 19A). PCP treatment resulted in a significant increase in extracellular GABA levels during the treatment period (313.3±82.8% of baseline, P<0.01) (Figure 19A). Combined treatment with NRG1 prevented the PCP-induced GABA increase (103.0±15.7% of baseline, P=0.9) (Figure 19A). A trend towards increased GABA levels was observed during NRG1 treatment (212.3±67.2%, P=0.09) (Figure 19A). The LMM analysis of the AUC showed a trend towards a treatment effect on the dCA1 GABA levels across the recording period (F(3, 19.9)=2.8; P=0.06) (Figure 19B). Neither PCP (83.24±8.6%) nor NRG1 (76.1±8.1%) by themselves showed any effect on GABA levels. A trend towards reduced GABA levels was identified following combined NRG1 and PCP treatment (67±8% of baseline equivalent, P=0.07).

Figure 19: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular GABA levels in the dorsal hippocampus CA1 in vivo. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the dorsal hippocampus CA1 by retrodialysis. GABA levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. ##P<0.01, PCP vs baseline; TP=0.07, NRG1+PCP vs baseline equivalent.

3.3.2.2 NRG1 treatment reduces extracellular glutamate levels in the dCA1 A one-way repeated measure ANOVA indicated that baseline glutamate values did not differ before each treatment (F(1.8, 19.6)=2.6; P=0.112). Extracellular glutamate levels were significantly affected by the treatments (F(3, 158.9)=16.8; P<0.001) and varied over the sampling time (F(5, 157.1)=2.4; P<0.05), with no interactions between both factors as identified by LMM analysis (F(15, 156.7)=1.3; P=0.21) (Figure 20A). Post-hoc analysis showed that NRG1 (61.5±7% of baseline, P<0.001) and combined NRG1 and PCP (52.1±8% of baseline, P<0.001) significantly reduced glutamate levels across the sampling period (Figure 20A). PCP, however, had no overall effect on glutamate levels (92.4±8% of baseline, P=1). Detailed single-time point analysis showed that combined treatment with NRG1 and PCP significantly reduced glutamate levels during the treatment period (28.3±18% of baseline, P<0.001), with extracellular levels only returning to baseline levels at the end of the sampling period (Figure 20A). PCP treatment triggered an increase in glutamate levels 20min after the treatment (135±12.8% of baseline, P<0.05). This peak was not present following the combined treatment of NRG1 and PCP (56.8±12.6% of baseline, P<0.01) (Figure 20A). The AUC analysis found a significant treatment effect (F(3, 23.1)=5.9; P<0.01). Post-hoc comparison confirmed the observation that NRG1 and combined NRG1 and PCP treatment resulted in glutamate levels below baseline values (36.3±5% of baseline equivalent, P=0.07 and 28.5±6.7% of baseline equivalent P<0.05 respectively) (Figure 20B). PCP treatment had no effect on overall glutamate levels (96.7±9.4%).

Figure 20: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glutamate levels in the dorsal hippocampus CA1 in vivo. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the dorsal hippocampus CA1 by retrodialysis. Glutamate levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. ***P<0.001, *P<0.05 NRG1+PCP vs baseline; T#P<0.05, PCP vs baseline; &P<0.05, T&P=0.07 NRG1 vs baseline.

3.3.2.3 NRG1 treatment reduces extracellular glycine levels in the dCA1 A one-way repeated measure ANOVA confirmed that the baseline glycine values were not different before each treatment (F(1.4, 11.2)=2.3; P=0.15). Extracellular glycine levels in the dCA1 were significantly affected by the treatment regimens (F(3, 159)=6.7; P<0.001) and varied across the sampling time (F(5, 157.1)=3.1; P<0.01). Furthermore, the LMM analysis revealed a trend for an interaction between both factors (F(15, 158.8)=1.6; P=0.08). Post-hoc analysis showed that none of the three treatment options altered glycine levels compared to baseline values (PCP: 127.5±12%; NRG1+PCP: 114±12%; NRG1: 74.1±12% of baseline) (Figure 21A). Detailed single-time point one-way ANOVA revealed that PCP treatment caused an increase in glycine levels 50min after the treatment (186.5±47.1% of baseline, P<0.05), an effect not observed following either NRG1 and PCP or NRG1 treatment (Figure 21A). Furthermore, glycine levels were increased in the presence of the combined treatment of NRG1 and PCP (195.6±54% of baseline, P<0.05) (Figure 21A). The AUC LMM analysis revealed a significant treatment effect on overall glycine levels (F(3, 15.3)=7.3; P<0.01) (Figure 21B). Post-hoc comparison identified a significant reduction in glycine levels following NRG1 treatment (68.4±3.8%, P<0.001), but not following the other treatments into the dCA1 (PCP: 121.7±112.5%, P=0.36; NRG1+PCP: 94.6±9.8%, P=1; of baseline equivalent) (Figure 21B).

Figure 21: Treatment effect of neuregulin 1 (NRG1) and phencyclidine (PCP) infusion on extracellular glycine levels in the dorsal hippocampus CA1 in vivo. Mice received a local infusion of NRG1 (10nM), PCP (10µM) or NRG1+PCP for 10min into the dorsal hippocampus CA1 by retrodialysis. Glycine levels are expressed as a percentage of the respective baseline level (see Figure 12) and represented as the means ± SEM: (A) for each sampling time point or (B) for total area under the curve analysis. Dotted lines indicate baseline levels. *P<0.05 NRG1+PCP vs baseline; #P<0.05, PCP vs baseline; &&&P<0.001 NRG1 vs baseline.

3.4 Discussion

My microdialysis experiments showed that PCP-induced alterations to extracellular GABA levels were prevented by combined treatment with NRG1 in the dCA1 and in a subgroup of animals in the PFC (Figure 22). Furthermore, NRG1 treatment in the presence of PCP altered extracellular glutamate and glycine levels in both brain regions (Figure 22).

Figure 22: Summary diagram of treatment-induced neurotransmitter level differences Summary diagram of the neurotransmitter level differences following neuregulin 1 (10nM, blue), phencyclidine (10µM, black) or combined (red) infusion into either the prefrontal cortex (PFC) or dorsal CA1 of the hippocampus. Treatment effects on extracellular GABA, glutamate or glycine levels during infusion period: ↑ above baseline, ↓ below baseline. Neurotransmitter area under the curve differences across sampling period: ≈ equal to baseline, ˅ below baseline, ˄ above baseline levels.

3.4.1 NRG1 treatment affects extracellular neurotransmitter levels in the PFC in the presence of PCP

3.4.1.1 Extracellular GABA levels remained unchanged following NRG1 and PCP treatment Local infusion of PCP into the PFC had little effect on extracellular GABA levels in the majority of animals (Figure 16A). I observed a PCP-induced increase in extracellular GABA levels in the PFC of a subgroup of animals (Figure 16C), which will need to be explored in subsequent experiments.

Published data on the effect of local PCP infusion on extracellular GABA levels by Yonezawa et al. as well as Zhu and colleagues have shown a reduction in GABA levels following local PCP administration in the rat PFC (Yonezawa et al. 1998; Zhu et al. 2004). Both studies have significant methodological differences to my experiment. Firstly, significant reduction in GABA levels was observed after 40min of consistent PCP retrodialysis (Yonezawa et al. 1998; Zhu et al. 2004), with no significant difference after 10min, which is similar to my experiment. Secondly, PCP was infused at 50µM (Zhu et al. 2004) and 100-400µM (Yonezawa et al. 1998), compared to 10µM in my study. Finally, microdialysis treatment and sampling was performed 18-20h post-surgery (Yonezawa et al. 1998; Zhu et al. 2004), allowing very little recovery time compared to the seven day period in my experiments. These methodological differences are the likely cause for the observed treatment effect difference. Additionally, peripheral PCP treatment has shown to have no effect on PFC GABA levels in rats (Martin et al. 1998), further supporting that GABA response to PCP varies with method differences. The lack of an overall effect of PCP on extracellular GABA levels hampered the efforts of assessing the NRG1 treatment effect on PCP-induced GABA differences. However, since PCP interferes with several neurotransmitter receptors to cause schizophrenia-relevant neurotransmission changes, I still assessed the effect of NRG1 on neurotransmitter levels in the presence of PCP. Extracellular GABA in the PFC remained at baseline levels following combined NRG1 and PCP treatment (Figure 16A-B). In the subgroup of animals where PCP treatment induced a GABA level increase, combining NRG1 with PCP prevented this increase, keeping GABA at baseline levels. While these effects were not observed in the majority of animals, the finding indicates a potential interaction which will need to be explored in future work. NRG1 receptors are primarily located on GABA and glutamate releasing cells in the PFC (Neddens et al. 2011; Sharif et al. 2009), which also express the dominant PCP target NMDAR. The reported functional link between ERBB4 and NMDAR might be the cause for the observed finding in this subgroup (Geddes et al. 2011), but further investigations are required to assess this interaction. Understanding the effect of NRG1-ERBB signalling on neurotransmission in the PFC during physiological normal conditions is of critical importance for recent investigations into the use of NRG1 for neurological and cardiological treatments (Gao

et al. 2010; Jabbour et al. 2011; Gauthier et al. 2013). Furthermore, such data would allow the comparison with existing antipsychotics, which have shown to alter PFC neurotransmitter levels in healthy rodents (Table 3). Since no published data exists on the effect of NRG1 signalling on acute neurotransmission in vivo for the PFC, I treated freely moving mice via retrodialysis with NRG1-only. In vitro experiments have shown that NRG1 administration can lead to an increase in depolarization-evoked GABA release via ERBB4 activation on PFC interneurons (Wen et al. 2010; Woo et al. 2007). In my study, this effect was only observable in the first post-treatment samples, showing elevated GABA levels in the PFC 10min after the NRG1 only treatment (Figure 16A). This result indicates that a sudden increase in extracellular NRG1 does not directly influence GABA levels in vivo. The delayed treatment effect, however, suggests that NRG1 indirectly affects GABA levels, possibly through interaction of ERBB4 with intracellular phosphatases and neurotransmitter receptors. Further experiments are required to explore the detailed signalling cascade underlying this finding in relation to the published in vitro results. The antipsychotic zotepine has been shown to increase GABA levels in healthy animals (Table 3), an effect which has been attributed to an increased risk of inducing seizures (Yamamura, Ohoyama, Hamaguchi, Nakagawa, et al. 2009). While the zotepine-induced increase in GABA levels occurred in the presence of the drug, NRG1 treatment only led to an increase in GABA levels after the treatment cessation for one sampling time point (Figure 16A). It is therefore less likely that NRG1 treatment would cause a similar adverse reaction as has been associated with the GABA- level increase following Zotepine treatment. .

Table 3: Effect of common antipsychotics and NRG1 on neurotransmitter levels in the PFC of rodents Name GABA Glutamate Glycine Dopamine Serotonin Noradrenaline Acetylcholine

Aripiprazole ↔ oralf ↔ oral o ↑ oralf ↔ s.c. p, r ↑ s.c. p, r Clozapine ↓ rd i ↔ oral k ↔ s.c.e ↑ s.c.e, q ↔g; ↓ s.c. g, l ↑ s.c. q, s ↑ j; ↔ g; ↓e s.c. ↑ i.v. m ↑ i.p. h ↑ rd m ↑ rd g Haloperidol ↑ s.c.e ↔ oralf ↔ s.c.e ↔ s.c.e, q ↓ oralf ↔ i.p. a ↔ s.c. s ↔ i.p. a ↔ s.c. e, g, j ↔ i.p. a ↔ g; ↓ l ↓ s.c. q ↓ rd i ↔ i.p. a, h ↔ i.v. m ↔ i.p. a ↔ rd g ↔ rd g Olanzapine ↓ s.c.e ↔ oralf ↔ s.c.e ↑ oral o ↔ oralf ↑ s.c. t ↔ s.c. r, t ↔ s.c.e ↑ s.c. e, r, t ↔ s.c. l, t ↑ s.c. s ↑ i.v. m ↑ rd m Quetiapine ↔ i.p.d ↑ i.p.d ↑ i.p.d ↔ i.p.d ↑ i.p.d Risperidone ↑ s.c. r ↑ i.p.c ↑ i.p. c ↑ s.c. r, s ↑ i.p.c ↑ rd m Ritanserin ↔ i.p.b ↑ s.c. l Zotepine ↑ i.p. a ↑ i.p.a ↑ i.p. a ↔ i.p. a ↑ i.p. a Neuregulin 1 ↔ rd x ↓ rd x ↓ rd x

Effect on extracellular neurotransmitter levels: ↔ no effect, ↑ increase, ↓ reduction; drug administration: s.c. subcutaneous; i.p. intraperitoneal; i.v. intravenous; oral; rd: retrodialysis (local infusion via microdialysis probe). References: a(Yamamura, Ohoyama, Hamaguchi, Nakagawa, et al. 2009), b(Kuroki et al. 1999), c(Ohoyama et al. 2011), d(Yamamura, Ohoyama, Hamaguchi, Kashimoto, et al. 2009), e(Heidbreder et al. 2001), f(Carli, Calcagno, Mainolfi, et al. 2011), g(López-Gil et al. 2007), h(Abekawa et al. 2006), i(Bourdelais & Deutch 1994), j(Daly & Moghaddam 1993), k(Carli, Calcagno, Mainini, et al. 2011), l(Amargos-Bosch et al. 2006), m(Bortolozzi et al. 2010), n(Gessa et al. 2000), o(Jordan et al. 2004), p(Li et al. 2004), q(Chung et al. 2004), r(Huang et al. 2006), s(Ichikawa et al. 2002), t(Koch et al. 2004) x(my study)

3.4.1.2 NRG1 treatment reduces extracellular glutamate levels in the PFC PCP retrodialysis did not affect local glutamate levels in the PFC (Figure 17). This observation is in line with existing studies reporting unchanged extracellular glutamate levels following PCP and local NMDA receptor agonist administration in the PFC of rats (López-Gil et al. 2007; Zhu et al. 2004). However, an increase in PFC glutamate levels has been observed following peripheral PCP treatment (Abekawa et al. 2007; Adams & Moghaddam 2001; Amitai et al. 2012; Baker et al. 2008; Li et al. 2010), which was preventable by clozapine and haloperidol treatment (Abekawa et al. 2007; Amitai et al. 2012) (Table 2). The present and published data indicate that peripheral and central PCP treatments differently affect glutamate and GABA levels in the PFC. This strongly suggests that excitatory afferent connections contribute largely to the systemic PCP effect on neurotransmission in the PFC (Suzuki et al. 2002). Changes to other neurotransmitters following local PCP infusions are therefore also likely to differ compared to systemic injections. PCP retrodialysis allows precise control over treatment onset and intensity when interfering with local neurotransmission, avoiding the treatment duration and intensity ambiguity of systemic injections. However, the clear difference between published neurotransmitter differences following systemic PCP and the present findings following local PCP infusion show that the broad nervous system-wide effect of PCP is of importance when modelling schizophrenia-relevant neurotransmission impairments. Future experiments will therefore need to include this route of PCP administration. While PCP alone had no effect on extracellular glutamate levels, the combined infusion of NRG1 and PCP also did not affect extracellular glutamate levels in the PFC (Figure 17). NRG1 by itself however, reduced glutamate levels in the PFC (Figure 17). The reduction in extracellular glutamate levels lasted longer than the infusion period. This suggests that activation of NRG1 receptors triggers intracellular mechanisms which possibly outlast the direct receptor stimulation duration. However, since glutamate levels in the samples collected 2.5h after the NRG1 infusion (i.e. baseline samples before PCP treatment) did not differ compared to pre-NRG1 treatment samples, the intracellular reactions appear to return to their pre-treatment state within 2h. The possible mechanisms behind the observed reduction in glutamate have been studied by a small number of in vitro experiments. Bath application of NRG1 has shown to reduce NMDA receptor currents on pyramidal cells (Gu et al. 2005), while 77

not affecting glutamate release in electrophysiological slice preparations (Gu et al. 2005; Pitcher et al. 2011). It has therefore been proposed that the current reduction is caused by internalization of the essential NMDA receptor NR1 subunit (Gu et al. 2005). Since the reduction in glutamate was not visible in the presence of PCP, it appears likely that NMDAR signalling and dopamine signalling are involved in the NRG1 effect on glutamate. However, recent work has shown that activation of the pre-synaptic protein LIMK1 following NRG1 application reduces the release of glutamate, which might be the contributing element in the present results (Yin et al. 2013). Additional neurotransmitter actions present in vivo and interactions between ERBB4 and several intracellular kinases (Bjarnadottir et al. 2007; Liu & Kern 2002; Pitcher et al. 2011; Ting et al. 2011) are also likely involvement in the observed reduction. Identifying the precise mechanism will require further in vivo experiments, combining NRG1 retrodialysis with inhibitors for glutamate-signalling relevant enzymes such as FYN, SRC and others.

3.4.1.3 Combined treatment of NRG1 and PCP increases extracellular glycine levels in the PFC Glutamatergic neurotransmission can be influenced by the extracellular levels of glycine, an allosteric modulator of NMDAR (Labrie & Roder 2010; Lee et al. 2009). Retrodialysis of PCP had little effect on extracellular glycine levels in the PFC (Figure 18). While no published data exists about the effect of local PCP on glycine levels, PFC glycine levels are also unaffected by peripheral PCP administration (Li et al. 2010). Glycine levels were significantly increased during combined infusion of NRG1 and PCP (Figure 18A). Several clinical trials have been performed aiming to increase activation of the glycine binding site on NMDAR by using direct agonists or glycine uptake transport inhibitors (Labrie & Roder 2010). The ability of NRG1 to increase glycine in the presence of PCP therefore indicates a potential for this line of clinical investigations. To address the lack of knowledge about the effect of NRG1 on glycine levels, I also quantified glycine level differences following NRG1-only infusion into the PFC. NRG1 treatment lead to a reduction in extracellular glycine levels (Figure 18A). In- terestingly, while the combined NRG1 and PCP treatment led to a significant in- 78

crease during the infusion period, NRG1 treatment significantly reduced glycine throughout the sampling period (Figure 18B). When comparing the glycine with the glutamate results, it becomes apparent that both neurotransmitters are simultaneously reduced following NRG1 treatment (Figure 17 and Figure 18). Interestingly, this pattern is not maintained following the combined NRG1 and PCP treatment, which triggers an increase in glycine levels, without affecting glutamate levels (Figure 17 and Figure 18). The cause for this treatment difference will need to be assessed in separate experiments incorporating receptor antagonists for the different PCP targets. In relation to the behavioural experiments in chapter 2, increasing glycine levels has been shown to prevent the behavioural impairments as well as dopamine level increases following PCP treatment in rodents (Boulay et al. 2008; Javitt et al. 1999). The observed increase in glycine following the combined NRG1 and PCP treatment might have therefore potentially contributed to the prevention of PCP-induced behavioural impairments by NRG1.

The observed treatment effects on neurotransmitter levels in the PFC occurred during different sampling time points as described above. The differences observed during the perfusion period suggest that local PFC cell populations are involved in the treatment responses, while the post-infusion differences indicate that afferent and efferent connections to other cortical and subcortical structures are additionally involved in the present results (Hoover & Vertes 2007). The effect of PCP on PFC neurotransmission has been reported to depend on PFC-hippocampal connections (Jodo et al. 2005). My results suggest that the NRG1 infusion also triggers differing feedback responses from other structures, which might have contributed to the observed differences in neurotransmitter levels. Unlike hypothesised, PCP retrodialysis had no significant effect on the extracellular levels of the three studied neurotransmitters. While I chose to use local PCP infusion to have precise control over the treatment timing, the treatment effect in the PFC differed to published systemic PCP treatment studies. This prevented the analysis of a NRG1 treatment effect on PCP-induced neurotransmitter level changes in the PFC. The antipsychotic treatment potential of NRG1 against impaired PFC neurotransmission will therefore need to be assessed in a future experiment with peripheral PCP treatment.

3.4.2 NRG1 treatment effects extracellular neurotransmitter levels in the dCA1 in the presence of PCP

Direct interactions between the PFC and the hippocampus have been implicated in several schizophrenia-relevant cognitive functions, including spatial working memory and decision making (Churchwell & Kesner 2011; Driesen et al. 2008; God- sil et al. 2013). Additionally, neurotransmission differences in the hippocampus have been observed in patients with schizophrenia (Tamminga et al. 2010). I therefore assessed the ability of NRG1 to alter GABAergic, glutamatergic and glycinergic neurotransmission in vivo in the hippocampus in the presence of PCP.

3.4.2.1 NRG1 treatment prevents PCP-induced GABA level increase in the dCA1 Retrodialysis of PCP triggered a three-fold increase in extracellular GABA levels in the dCA1 during the treatment period (Figure 19), without affecting GABA levels post-treatment. This increase suggests that the GABAergic interneurons of the CA1 respond differently to local PCP stimulation than PFC interneuron populations, where I only observed a similar effect in a subgroup of samples. Indeed, GABAergic interneurons of the CA1 have been shown to be more sensitive to NMDAR antagonists than pyramidal neurons in the same region (Grunze et al. 1996), while GA- BAergic interneurons in the PFC do not share this characteristic (Rotaru et al. 2011). This difference could explain why the increase in GABA levels following PCP infusion in the hippocampus was not observed in the majority of PFC samples. No published study has so far described the effect of local PCP treatment on GABA levels in the hippocampus in vivo. The effect of peripheral PCP treatment has been reported by one study, observing a 25% reduction in hippocampal GABA levels for 3h (Martin et al. 1998). In my study, PCP increased GABA levels exclusively during the treatment period, which suggests a direct treatment effect on local neurotransmitter release or uptake mechanisms in the dCA1. Peripheral injections of PCP, as in the study by Martin et al. (1998), leads to prolonged alterations of neurotransmitter levels (Hagino et al. 2010; Martin et al. 1998), likely affecting intracellular signalling cascades which have not been triggered in my study. Furthermore, it is possible that peripheral PCP could have caused a brief increase in GABA as seen in my study and 80

then returned to baseline levels. Such an effect would likely be hidden by the 30min sampling interval in the study by Martin and colleagues (Martin et al. 1998). Extracellular levels of GABA have been shown to be influenced by dopamine D2 receptor activation, with D2 agonists increasing and D2 antagonist decreasing extracellular levels (Grobin & Deutch 1998). The agonistic action of PCP on D2 receptors might therefore be a possible mechanism of increased GABA levels observed in my experiment. No data has been published on the effect of local or peripheral administered antipsychotics on local or peripheral PCP-induced changes to hippocampal neurotransmitter levels (Table 2). However, in patients with schizophrenia, activity levels of the CA1 region correlate with their symptom severity (Schobel et al. 2009). It was therefore of importance to locally infuse NRG1 into the hippocampus under the modelled condi- tion caused by PCP in an attempt to identify possible treatment-relevant characteristics of NRG1. Combining NRG1 with PCP completely prevented PCP from inducing an increase in GABA levels in the dCA1 (Figure 19). While the agonistic action of PCP on dopamine receptors together with the antagonistic action on NMDAR receptors might be the cause for the GABA increase, it remains unclear how NRG1 signalling via ERBB receptors can counteract the PCP effect. Interestingly, a recent study with GABA transporter knockout animals has shown that the reversal of increased GABA signalling prevents several behavioural impairments otherwise present in these mice (Yu et al. 2013). Preventing the PCP-induced GABA increase might therefore be one of the mechanisms contributing to the observed behavioural effects of combined NRG1 and PCP treatment in Chapter 2.

To further explore the effect of NRG1 signalling on extracellular neurotransmitter levels, I analysed GABA levels following NRG1 infusion during physiological normal conditions. NRG1 treatment had no overall effect on extracellular GABA levels in the dCA1 (Figure 19), which is similar to what I observed in the PFC samples (Figure 16). Several in vitro experiments have indicated that increasing NRG1 signalling would alter, possibly even increase GABA signalling (Woo et al. 2007; Wen et al. 2010; Okada & Corfas 2004). The difference between published in vitro results and the

present in vivo findings highlights the importance of a complete neuronal network when studying the effects of NRG1 on neurotransmission. Limited knowledge is available about the effect of current antipsychotics on neurotransmitter levels in the hippocampus of healthy animals, following either local infusion or peripheral administration (Table 4). However, the reported effect of these antipsychotics on acetylcholine levels in the hippocampus suggests that the levels of other neurotransmitters might also be altered by these drugs (Lawrence 2008). For NRG1, one study has shown that retrodialysis with the same NRG1 isoform as in my study resulted in an increase in extracellular dopamine levels in the dCA1 of rats (Kwon et al. 2008). Since no data on other neurotransmitters are available, it remains unclear if this increase in dopamine caused to the treatment effect against PCP- induced GABA increase or if effects on other neurotransmitters contributed.

Table 4: Effect of common antipsychotics and NRG1 on neurotransmitter levels in the hippocampus of rodents Name GABA Glutamate Glycine Dopamine Serotonin Noradrenaline Acetylcholine

Aripiprazole ↑ s.c. b, e ↔ s.c. b, e Chlorpromazine ↑ s.c. a Clozapine ↑ s.c. c ↓ i.p. f ↑ s.c. a, c, d ↑ rd d Haloperidol ↔ s.c. c ↔ s.c. a, ↑c Olanzapine ↑ s.c. e ↑ s.c.a, d, e ↑ rd d Raclopride ↔ i.p. f ↔ s.c. a Risperidone ↑ s.c. a ↔ rd d Thioridazine ↑ s.c. a, d ↑ rd d Ziprasidone ↔ s.c. a, d ↑ i.p. a Neuregulin 1 ↔ rd x ↓ rd x ↔ rd x ↑ rd g Effect on extracellular neurotransmitter levels: ↔ no effect, ↑ increase, ↓ reduction; drug administration: s.c. subcutaneous; i.p. intraperitoneal; rd: retrodialysis (local infusion via microdialysis probe). References: a(Shirazi-Southall et al. 2002), b(Li et al. 2004), c(Chung et al. 2004), d(Johnson et al. 2005), e(Huang et al. 2006), f(Assié et al. 1997), g(Kwon et al. 2008), x(this study)

3.4.2.2 NRG1 treatment reduces extracellular glutamate levels in the dCA1 A small number of clinical studies indicate that glutamate levels differ in the hippocampus regions of patients with schizophrenia following antipsychotic treatment (Poels, Kegeles, Kantrowitz, Javitt, et al. 2014). It is therefore of relevance to assess the effect of NRG1 on glutamate levels both in the PCP-modelled and physiological healthy conditions. PCP retrodialysis had no effect on extracellular glutamate levels in the dCA1 (Figure 20), a result similar to what I observed in the PFC (Figure 17). Interestingly, while PCP did not effect glutamate levels in either of the two brain regions, the effect on GABA is markedly different, increasing levels in the dCA1 without affecting levels in the PFC. This difference supports the previously made observation that GABAer- gic interneuron populations in both regions respond differently to PCP while glutamate-releasing pyramidal neurons do not appear to share this characteristics, as discussed above. No published study has so far described the effect of local or peripheral PCP treatment on glutamate levels in the hippocampus in vivo. It is therefore not possible to say whether a treatment difference between local and peripheral PCP on glutamate levels, as observed in the PFC, exists in the dCA1. While it might be inter- esting to explore the lack of an effect of PCP on local glutamate levels in future experiments, it is of limited relevance for the purpose of the present study on NRG1. The combined treatment of NRG1 and PCP prevented the PCP-induced increase in GABA as discussed above, while at the same time reducing extracellular glutamate levels (Figure 20). Such a reduction in glutamate levels was not observed in the PFC samples following the combined treatment, which supports the discussed response difference of neuronal populations in both regions. Glutamatergic synapses from CA3 pyramidal neurons are the primary source of glutamate in the CA1 region (Takács et al. 2012). While NRG1 overexpression has shown to reduce glutamate release in the CA1 via pre-synaptic intracellular mechanism in one study (Yin et al. 2013), how exogenous NRG1 might cause this effect remains unclear since ERBB3- 4 are not expressed on glutamatergic pre-synaptic terminals (Fazzari et al. 2010). The range of pharmacological actions of PCP combined with the activation of ERBB3-4 receptors on different cell types in the CA1 does not allow me to conclude on the mechanism behind the glutamate reduction following the combined treatment.

To explore the involvement of NRG1 in the observed glutamate level reduction, I separately treated animals with NRG1 under physiologically healthy conditions. NRG1 retrodialysis resulted in a trend towards reduced extracellular glutamate levels (Figure 20), but to a lesser extent than observed in the PFC (Figure 17). The reduction in glutamate was also less pronounced than following the combined NRG1 and PCP treatment in the dCA1. This result indicates that exogenous NRG1 was partially responsible for the reduction following the combined NRG1 and PCP retrodialysis, but that its effect was amplified by the pharmacological actions of PCP. It remains to be assessed whether exogenous NRG1 reduced glutamate levels by altering the control of local GABAergic interneurons over connected glutamatergic synapses from the CA3 region or via feedback mechanisms to the CA3, entorhinal cortex or thalamus (Chamberland & Topolnik 2012; Klausberger & Somogyi 2008).

3.4.2.3 Combined NRG1 and PCP treatment increased extracellular glycine levels in the dCA1 Variations in hippocampal glycine levels have been reported to alter glutamate release via pre-synaptic mechanisms (Lee et al. 2009), without being affected by GABA (Tanaka et al. 2003). The glycine signalling pathway offers an additional possibility for NRG1 to influence glutamate levels and I therefore quantified extracellular glycine levels following the three different treatment options. PCP retrodialysis had no overall effect on the extracellular levels of glycine aside from a brief increase at the end of the sampling period (Figure 21). The combined infusion of NRG1 and PCP however, caused an increase in extracellular glycine levels during infusion period (Figure 21). This increase in glycine following the combined infusion was of similar intensity and duration as observed in the PFC samples (Figure 18). While glycine has been reported to trigger glutamate release in the hippocampus (Lee et al. 2009), the actions of PCP and NRG1 appear to alter this response as glutamate levels dropped following the combined treatment (Figure 20). Interestingly, glycine levels remained unchanged following NRG1 retrodialysis in the dCA1 (Figure 20). This treatment response again differs between both regions, with significantly reduced glycine levels in the PFC (Figure 18). Neither the strong reduction in glutamate nor the increase in glycine levels triggered by the combined 85

NRG1 and PCP treatment was observed following the NRG1-only treatment. This suggests that the effect of NRG1 on neurotransmitter levels depends significantly on the current state of the neuronal network. Existing in vitro experiments have indicated that NRG1 administration is more likely to affect neurotransmitter receptor availability than neurotransmitter release in hippocampus slice preparations (Chang & Fischbach 2006; Liu et al. 2001; Okada & Corfas 2004; Pitcher et al. 2011; Shamir et al. 2012). The observed differences in rapid neurotransmitter level changes between PCP and NRG1-PCP treatment is therefore remarkably. However, the temporal resolution of our samples limits our ability to draw conclusions on the exact order of neurochemical events, such as GABA or glycine control over glutamate levels (Lee et al. 2009; Tanaka et al. 2003). The timing between treatments as well as the order of treatments could also have potentially influenced the neurotransmitter responses to the separate treatments. While there was no difference between the pre-treatment baseline samples for each analysed neurotransmitter, I cannot exclude the possibility that other transmitters and intracellular mechanisms were still influenced by the previous treatment after the 2.5h recovery period. Furthermore, continuous microdialysis has been reported to drain the extracellular space around the probe (Chefer et al. 2009). While this limita- tion primarily occurs with larger probe pore sizes and faster fluid transport than used in my study (Chefer et al. 2009), the possibility of this effect to alter neurotransmitter production cannot be excluded. Critically, a separate class of CNS cells has to be considered as contributing factor to the observed treatment effects. Cortical astrocytes have been reported to express receptors for NRG1 (Sharif et al. 2009), GABA receptors and synthesizing enzymes (Lee, McGeer, et al. 2011), NMDA receptors (Lalo 2006) and dopamine D2 receptors (Mladinov et al. 2010; Shao et al. 2012) among others. Furthermore, cortical astrocytes are able to release GABA (Le Meur et al. 2012; Lee, McGeer, et al. 2011), glutamate and glycine (reviewed in (Hamilton & Attwell 2010)). Although I did maintain natural osmotic levels to avoid stimulating GABA release from astrocytes by lower osmotic pressure (Wang et al. 2013), it is likely that neurochemical responses of astrocytes have either directly or indirectly been triggered by NRG1 and PCP treatment. However, as astrocyte function has also been implicated in the pathophysiology of schizophrenia (Halassa et al. 2007; Katsel et al. 2011), their contribu-

tion to the observed effects does not deter from the clinical potential of NRG1 treatment. With the present microdialysis experiment I was able to study the treatment effects on local neurotransmission in vivo. Retrodialysis administration of NRG1 and PCP enabled me to control the localised treatment onset and duration to assess the acute treatment effect in two distinct regions. The effect of PCP on local neurotransmitter levels, was, however, not as hypothesised, likely due to the restricted treatment location. Future studies are therefore needed to assess the ability of NRG1 signalling to reduce impairments in glutamatergic neurotransmission. However, the present study shows that glutamatergic and glycinergic neurotransmissions are affected by NRG1 in vivo and that this effect on neurotransmitter levels differs when PCP is present. The precise underlying mechanism needs to be explored in future experiments using selective inhibitors for NRG1-ERBB signalling pathway elements. The microdialysis findings support the observations from the behaviour studies in that NRG1 treatment effects depend on the current state of the neuronal network.

4 NRG1 PREVENTS PCP-INDUCED MRNA EXPRESSION CHANGES IN GABAERGIC INTERNEURONS IN VITRO

4.1 Introduction

My behaviour experiments show that central and peripheral administration of human recombinant NRG1 can prevent hyperlocomotion and PPI impairments induced by acute PCP treatment in mice. Several CNS structures which have been implicated in the regulation of locomotor activity and sensorimotor gating have also been associated with the pathophysiology of schizophrenia (Bast & Feldon 2003; Goulding 2009; Swerdlow et al. 2008; Swerdlow et al. 2001). Exemplary are the frontal cortex and hippocampus structures for their involvement in several symptom domains of schizophrenia and the assessed behavioural paradigms (Gonzalez-Burgos et al. 2011; Hall et al. 2010; Keshavan et al. 2008). I therefore performed in vivo microdialysis of the PFC and dCA1 described in the previous chapter, assessing the effect of NRG1 and PCP on extracellular neurotransmitter levels in freely moving mice. NRG1 retrodialysis affected glutamate and glycine levels and prevented PCP-induced changes to extracellular GABA levels. The behaviour and microdialysis experiments indicate that NRG1 signalling can influence schizophrenia-relevant behaviour and neurotransmission and prevent modelled impairments triggered by acute PCP. This ad- dresses elements of acute psychotic episodes as experienced by patients with schizophrenia. Differences in protein and mRNA expression levels relevant to GABAergic and glutamatergic neurotransmission have been identified in frontal cortex regions of patients with schizophrenia (see 1.1.2 for details). GABAergic interneurons have received particular attention after differences in GAD, PV and GAT1 expression levels where identified in frontal cortex regions of patients (Akbarian et al. 1995; Beasley & Reynolds 1997; Dracheva et al. 2004; Duncan et al. 2010; Guidotti et al. 2000; Hashimoto, Arion, et al. 2008; Hashimoto, Bazmi, et al. 2008; Hashimoto et al. 2003; Mellios et al. 2009; Sakai et al. 2008; Straub et al. 2007; Volk et al. 2012; Volk et al. 2000; Woo et al. 2008). In combination with reported expression differences of GABAA receptors (Benes, Vincent, et al. 1996; Hashimoto, Bazmi, et al. 2008; Lewis et al. 2004), NMDAR (Geddes et al. 2011; Weickert et al. 2013) and the 88

postsynaptic scaffold protein disrupted in schizophrenia 1 (DISC1) (Lipska et al. 2006; Nakata et al. 2009; Sawamura et al. 2005), these post-mortem findings suggest altered neurotransmission between GABAergic interneurons and pyramidal neurons in schizophrenia. Furthermore, structural differences in the frontal cortex interneuron-pyramidal neuron network in transgenic and lesion animal models suggest a vital role for GABAergic interneurons in the symptom spectrum of schizophrenia (reviewed by (Bissonette & Powell 2012; Marín 2012)). The GABAergic system is insufficiently addressed by the majority of available antipsychotics (Vinkers et al. 2010), which primarily target dopamine and serotonin receptors (see 1.1.3 for details). Closing this gap in the current treatment spectrum is of critical importance since the limited effect on GABA-relevant targets might contribute to low treatment efficacy against the cognitive symptoms domain of schizophrenia (Lewis & Moghaddam 2006). The post-synaptic density of GABAergic interneurons is the predominant site of neuronal ERBB4 expression in the frontal cortex (Fazzari et al. 2010; Neddens & Buonanno 2010; Neddens et al. 2011; Vullhorst et al. 2009). With NRG1 being released from pyramidal neurons in the PFC (X. Liu et al. 2011), this suggests a possible role for NRG1-ERBB4 signalling in the interneuron-pyramidal neuron network. The aim of this chapter was therefore to assess the ability of NRG1 to alter the expression levels of GABAergic system components and related elements in frontal cortex neurons. To study the effect of NRG1 treatment on local cell populations, I used primary frontal cortex cultures. This excludes confounding effects of afferent synapses from other cortical and subcortical regions. Furthermore, to assess the treatment effect of NRG1 during schizophrenia-relevant neurotransmission interruption, I administered NRG1 together with PCP. Prolonged PCP treatment has shown to reduce NMDAR activity and alter PV and GAD expression levels in rodent frontal cortex regions (Cochran et al. 2003; Ka- rasawa et al. 2002; Thomsen et al. 2009), modelling differences similar to those found in patients with schizophrenia (see 1.1.2.3 for details on the GABAergic system in schizophrenia). I investigated possible treatment effects of NRG1 on key elements of GABAergic interneurons, namely GAD67, GAD65 and PV, which indicate differences in intracellular mechanisms of these neurons. PCP has also been found to

impair synapse development in frontal cortex regions in several animal experiments (Wang et al. 2004), while NRG1 has been reported to positively affect synapse structures (Ozaki et al. 1997; Yang et al. 1998; Ting et al. 2011). I therefore included the presynaptic vesicle protein synaptophysin (SYP), DISC1 and the essential NMDA receptor subunit 1 (NR1) in my analysis to identify potential interactions in the treatment effects of NRG1 and PCP on synapses. Furthermore, as expression levels of SYP (Chambers et al. 2005; Glantz & Lewis 1997; Honer et al. 1999; Landén et al. 2002; Rao et al. 2013; Vawter et al. 1999; Webster et al. 2001), DISC1 (Lipska et al. 2006; Nakata et al. 2009; Sawamura et al. 2005) and NR1 (recently reviewed by Geddes et al. (2011)) have found to be altered in some patients with schizophrenia, a possible effect of NRG1 treatment on these proteins would have potential relevance for a clinical translation (Figure 1). I hypothesised that PCP treatment will cause schizophrenia-relevant differences in mRNA expression levels in frontal cortex cells due to its effect on several neurotransmitter receptors (see 1.4.2.2 for details). Since NRG1 treatment prevented behavioural and neurochemical changes induced by PCP in the previous experiments, I further hypothesised that NRG1 treatment will alter the PCP effect in frontal cortex cultures. I hypothesized that this effect would be particularly observable in GA- BAergic interneuron-relevant targets due to the location of the NRG1 receptor ERBB4 on these neurons.

4.2 Methods

4.2.1 Subjects

Pregnant female C57BL/6J mice were purchased from Australian BioResources (Moss Vale, NSW, Australia). Animals were single housed in open-top cages with minimum enrichment (M1 polypropylene cages with sawdust bedding; Able Scien- tific) and maintained on a 12h light/dark cycle (light on at 0600 hours) with food and water available ad libitum in the animal facility of the University of Wollongong (Wollongong, Australia). All research and animal care procedures were approved by the Animal Ethics Committee of the University of Wollongong (AE 10/19) and were in agreement with the ‘Australian Code of Practice for the Care and Use of Animals for Scientific Purposes’. All steps were taken to ameliorate any suffering in all work involving animals in my studies.

4.2.2 Drugs

All compounds were obtained from Sigma (Castle Hill, NSW, Australia) except where specified otherwise. Drugs were freshly prepared on each treatment day. Re- combinant human NRG1 (#396 NRG1-β1/HRG1-β1 EGF domain CF; R&D Sys- tems, USA) and PCP were dissolved in ACSF (119mM NaCl; 2.5mM CaCl2; 2.5mM

KCl; 1.3mM MgSO4; 1mM NaH2PO4; 26.2mM NaHCO3; 11mM glucose, 7.4pH, 290mOsm/kg) for cell culture treatments.

4.2.3 Frontal cortex culture

On post natal day 1 or 2 mouse pups were decapitated and their brain removed asep- tically. The brain was placed in ice cold dissection solution (Hilgenberg & Smith 2007) and the frontal cortex area of both hemispheres dissected in to 1-2mm blocks following the careful removal of meninges and other superficial blood vessels. The tissue pieces were incubated for 30min at 37°C in papain-containing enzyme solution for tissue digestion, followed by enzyme inhibitor washing steps. Tissue pieces were then dissociated in growth medium by trituration using glass pipettes. Dissociated cells were collected in warm growth medium consisting of Neurobasal™ (21103, 91

gibco, Life Technologies Australia Pty Ltd, Mulgrave, Australia) supplemented with 1% B-27 supplement (Life Technologies Australia Pty Ltd, Mulgrave, Australia). Cells were plated in complete medium on poly-d-lysine coated 96-well culture plates (F-bottom, Cellstar, greiner bio-one, Germany) at 5 × 104 cells/well for mRNA analysis and maintained at 37°C in a humidiﬁed atmosphere of 5% CO2 and 95% air. One day after the plating, the medium was replaced with new conditioned Neuro- basal medium with B-27 supplement (cNBM/B27, (Hilgenberg & Smith 2007)) and again in 48h intervals. For mRNA expression experiments, neurons were cultured and maintained for 6 days before the drug treatment started. I used one week old high density cultures to allow sufficient time for the development of excitatory and inhibitory responses to GABA signaling (Ben-Ari et al. 2007; Ting et al. 2011).

4.2.4 Cell culture treatment and mRNA expression analysis

All drugs were freshly prepared on each treatment day. NRG1 and PCP were dissolved and diluted in ACSF to prevent interference with growth factors in NBM/B27. The suitability of ACSF for mixed neuronal culture treatment over 24h has recently been confirmed by Pamenter and colleagues (Pamenter et al. 2013). Drug treatments were added to the cultures by replacing the existing medium with fresh ACSF containing NRG1 (0.01, 0.1 or 1nM), PCP (10µM), NRG1 (0.01, 0.1, 1 or 10nM) +PCP or no drug. For the analysis of the neurotransmission signalling marker mRNA expression levels, the mRNA isolations were carried out with a Cell-to-Ct kit (Ambion, Life Technolo- gy Pty Ltd, Mulgrave, Australia) according to the manufacturer’s instructions. Van Peer and colleagues reported that cDNA synthesis from cell lysis matches the accu- racy and sensitivity of classic RNA purification and cDNA synthesis from < 1x105 cells when using this kit (Van Peer et al. 2012). Reverse transcription reactions con- tained 15µl of lysis solution, with 3µl of the resulting cDNA-containing solution being used for each qPCR reaction. Hydrolysis probes (Cat# 4331182, Ambion, Life Technology Pty Ltd, Mulgrave, Australia) were used in duplex reactions for target- specific amplification. TaqMan Probe details can be found in Table 5.

Table 5: Hydrolysis probes for real time mRNA quantification Target UniGene ID Assay ID Reference Sequence gene PAK1IP1 Mm.24789 Mm0050941 NM_026550.3 UBC Mm.331 Mm01198158_m1 NM_019639.4 DISC1 Mm.233504 Mm00533313_m1 NM_174853.2 GAD1/ Mm.272120 Mm00725661_s1 NM_008077.4 GAD67 GAD2/ Mm.4784 Mm00484623_m1 NM_008078.2 GAD65 GRIN1 Mm.278672 Mm00433790_m1 NM_001177656.1 PVALB Mm.2766 Mm00443100_m1 NM_013645.3 SYP Mm.223674 Mm00436850_m1 NM_009305.2

4.2.5 Selection of reference genes

The reference genes UBC and PAK1IP1 were selected after testing UBC, Pak1IP1, AP3D1 and CsnK2a2 (GeNorm mouse kit ge-SY-6, Integrated Sciences Pty. Ltd., Australia) across a range of representative samples. The stability was assessed using the GeNorm software. Validated reference genes exhibited average expression stability values (M) under 1.5 (Vandesompele et al. 2002).

4.2.6 Real Time Quantitative PCR method and data analysis

The amount of double stranded PCR product synthesised in each cycle was measured using TaqMan Universal Mastermix (4440038, Ambion, Life Technology Pty Ltd, Mulgrave, Australia) on a Lightcycler 480 real time PCR machine (Roche, Australia) in 20µl reaction volume per technical replicate (n=3) according to the Cell-to-Ct kit instructions. Samples with quantification cycles (Cq) of 35 and above were excluded from the analysis (average Cq for DISC1=30.3, GAD1=30.4, GAD2=27.8, NR1=28.2, PAK1IP1=28.4, PVALB=26.6, SYP=25.6 and UBC=22.6). Cq values for each gene of interest were normalised to the Cq values of both reference genes from the same cDNA preparations, considering the individual reaction efficiencies and

inter-run calibrator values according to the 2-ΔCq method (Schmittgen & Livak 2008).

4.2.7 Statistical Analysis

Normalised quantification cycle values for each target gene were further normalised to the relevant vehicle treatment values of either the 3h or 24h treatment group to assess fold change differences following each treatment intervention. Differences in mRNA expression were separately analysed for the 3h and 24h treatment groups by using one-ay ANOVA for each gene of interest. Bonferroni’s multiple comparison was employed to identify individual differences when applicable. Analyses were performed using SPSS 19.0. Statistical significance was accepted at P<0.05, data are presented as means±SEM. Outliers were screened as mean ±2 standard deviations and removed.

4.3 Results

4.3.1 NRG1 treatment prevents PCP-induced increase in GAD67 mRNA expression levels

Drug treatment for 3h had no significant effect on GAD67 mRNA expression levels (F(10, 51)=1.28; P=0.26) (Figure 23). Drug treatment for 24h had a significant effect on GAD67 mRNA expression levels (F(8, 36)=20.28; P<0.001) (Figure 23). PCP treatment over 24h significantly increased GAD67 mRNA levels compared to vehicle treated cultures (203.6±14.9% of vehicle, P<0.001) (Figure 23). NRG1 prevented the PCP-induced increase in mRNA expression in a dose- dependent manner: GAD67 mRNA expression levels following treatment with 0.01nM and 0.1nM NRG1 combined with PCP were similar to levels in vehicle treated cells (73.79±15.63% of vehicle, P>0.99; 108.4±14.9% of vehicle, P>0.99 respectively) (Figure 23). Combined treatment of PCP with 1 or 10nM NRG1 resulted in a significant increase in GAD67 mRNA expression levels compared to vehicle, similar to PCP-only treated cells (191.5±15.6% of vehicle, P<0.001; 219.1±21.5% of vehicle, P<0.001 respectively) (Figure 23). NRG1 treatment for 24h did not change GAD67 mRNA levels at 0.1, and 1nM, but significantly reduced expression levels at 0.01 nM (122.5±15.2% of vehicle, P>0.9; 110.5±15.2% of vehicle, P>0.9; 51.4±7.6% of vehicle, P<0.001 respectively) (Figure 23).

Figure 23: GAD67 mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture. Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), phencyclidine (PCP), NRG1+PCP or NRG1 (treatment concentration) for (A) 3h or (B) 24h after six days in vitro. mRNA expression levels of glutamic acid decarboxylase 67 (GAD67) were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment + SEM. ##P<0.01, ###P<0.001 vs Vehicle; ***P<0.001 vs PCP.

4.3.2 NRG1 treatment prevents PCP-induced decrease in GAD65 mRNA expression levels

Drug treatment for 3h had a significant effected on GAD65 mRNA expression (F(10, 50)=5.57; P<0.001) (Figure 24). A post-hoc analysis revealed a trend for reduced GAD65 mRNA expression following 0.01nM NRG1 (82.5±1%, P=0.06 vs vehicle) and 0.01nM NRG1 combined with PCP (78.3±3.2%, P=0.06 vs vehicle). Treatment for 24h had a significant effect on GAD65 mRNA expression levels (F(8, 38)=4.68; P<0.001) (Figure 24). PCP treatment over 24h significantly reduced GAD65 mRNA levels compared to vehicle treatment (69.7±8.6% of vehicle, P<0.05) (Figure 24). NRG1 prevented the PCP-induced reduction in mRNA expression in a dose- dependent manner: combining PCP with 0.1nM NRG1 resulted in GAD65 mRNA levels similar to vehicle treatment (92.1±7.9% of vehicle, P>0.9) (Figure 24). How- ever, GAD65 mRNA levels following combined treatment of PCP with 0.01, 1 or 10nM NRG1 were similar to PCP-only treated cells (83.9±8.6% of vehicle, P>0.9; 86.4±9% of vehicle, P>0.9; 79.3±10.5% of vehicle, P>0.9 respectively) (Figure 24). NRG1 treatment for 24h dose-dependently affected GAD65 mRNA levels. NRG1 at 0.1nM and 1nM had no effect on GAD65 mRNA expression (93.9±6.2% of vehicle, P=0.7; 114.5±6.2% of vehicle, P=0.07 respectively) while 0.01nM NRG1 significantly reduced expression levels of GAD65 mRNA (72.6±7.6% of vehicle, P<0.01) (Figure 24).

Figure 24: GAD65 mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture. Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), phencyclidine (PCP), NRG1+PCP or NRG1 (treatment concentration) for (A) 3h or (B) 24h after six days in vitro. mRNA expression levels of glutamic acid decarboxylase 65 (GAD65) were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment + SEM. #P<0.05, vs Vehicle; TP=0.057 vs PCP.

4.3.3 NRG1 treatment prevents PCP-induced increase in PV mRNA expression levels

Drug treatment for 3h significantly affected PV mRNA expression levels (F(10, 50)=5.4; P<0.001) (Figure 25). Post-hoc analysis showed that 0.1nM NRG1 treatment significantly increased PV mRNA levels (131.4±7.7% of vehicle, P<0.001) (Figure 25), while none of the other treatments affected PV mRNA levels. Drug treatment for 24h had a significant effect on PV mRNA expression levels (F(8, 37)=4.8; P<0.001) (Figure 25). PCP treatment over 24h significantly reduced PV mRNA levels compared to vehicle treated cells (133.4±11.8% of vehicle, P<0.01) (Figure 25). NRG1 prevented the PCP-induced reduction in PV mRNA expression in a dose- dependent manner: combining PCP with 0.1nM NRG1 led to PV mRNA levels significantly lower than PCP-only treated cells (105.6±10.1% of vehicle, P<0.05) (Figure 25). However, PV mRNA levels following combined treatment of PCP with 0.01, 1 and 10nM NRG1 were similar to PCP-only treatment (107.9±10.6% of vehicle, P=0.07; 134.8±10.6% of vehicle, P>0.9; 154±12.4% of vehicle, P=0.2 respectively) (Figure 25). NRG1 treatment for 24h had no effect on PV mRNA levels at any of the tested concentrations (Figure 25).

Figure 25: Parvalbumin (PV) mRNA expression after 3h and 24h of neuregulin 1 (NRG1) and PCP treatment in primary frontal cortex culture. Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), PCP, NRG1+PCP or NRG1 (treatment concentration) for (A) 3h or (B) 24h after six days in vitro. mRNA expression levels of PV were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment + SEM. ###P<0.001, ##P<0.01, #P<0.05 vs Vehicle; *P<0.05, vs PCP.

100

4.3.4 DISC1 mRNA expression levels were not affected by NRG1 or PCP treatments

Drug treatment for 3h had a significant effect on DISC1 mRNA expression levels in mice frontal cortex cells (F(10, 47)=2.03; P=0.05) (Table 6). Post-hoc analysis, however, failed to identify a significant treatment effect following any of the applied treatments (Table 6). Furthermore, DISC1 mRNA expression levels were not affected after 24h exposure to any of the tested treatment combinations (F (8, 37)=0.97; P=0.5) (Table 6).

Table 6: Unchanged DISC1 mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures Treatment 3h 24h ACSF 100% 100% PCP (10µM) 99.8±4.4% 90.6±5.6% NRG1 (0.01nM) + PCP 109.3±1.6% 101.1±19.8% NRG1 (0.1nM) + PCP 106.5±5.7% 84.1±2.8% NRG1 (1nM) + PCP 121.1±7.5% 105.2±3.8% NRG1 (10nM) + PCP 98.9±8.6% 109.4±24.8% NRG1 (0.01nM) 89.8±3.7% 108.5±7.8% NRG1 (0.1nM) 86.4±13.1% 103.7±2.9% NRG1 (1nM) 124±6.5% 103.6±3.6% Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), PCP, NRG1+PCP or NRG1 (treatment concentration) for 3h or 24h after six days in vitro. mRNA expression levels of DISC1 were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment ± SEM.

101

4.3.5 Synaptophysin mRNA expression levels were not affected by NRG1 or PCP treatments

Drug treatment for 3h had no significant effect on SYP mRNA expression levels in mice frontal cortex cells (F(10, 50)=1.4; P=0.2) (Table 7). Furthermore, SYP mRNA expression levels were not affected by any of the tested treatment combinations after 24h (F (8, 40)=1.48; P=0.2) (Table 7).

Table 7: Unchanged synaptophysin mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures Treatment 3h 24h ACSF 100% 100% PCP (10µM) 105.7±7.5% 100.8±15.6% NRG1 (0.01nM) + PCP 93.6±2.1% 83.9±6.8% NRG1 (0.1nM) + PCP 106.3±4.5% 96.6±11.4% NRG1 (1nM) + PCP 99.5±3.5% 88.6±5.3% NRG1 (10nM) + PCP 101.2±11% 133.6±29.3 NRG1 (0.01nM) 80.1±2.2% 92.1±11.9% NRG1 (0.1nM) 132.7±5.9% 98.1±3.2% NRG1 (1nM) 120.3±3.8% 112.6±3.7% Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), PCP, NRG1+PCP or NRG1 (treatment concentration) for 3h or 24h after six days in vitro. mRNA expression levels of SYP were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment ± SEM.

102

4.3.6 NMDA-NR1 mRNA expression levels were not affected by NRG1 or PCP treatments

Drug treatment for 3h had no significant effect on NR1 mRNA expression levels in mice frontal cortex cells (F(10, 50)=0.63; P=0.78) (Table 8). Treatment for 24h had a significant effect on NR1 mRNA expression levels (F (8, 38)=3.43; P<0.05). How- ever, no significant NR1 mRNA expression level differences were observed between vehicle and drug treated cells (Table 8).

Table 8: Unchanged NMDA receptor subunit 1 (NR1) mRNA expression levels after 3h and 24h NRG1 and PCP treatment compared to vehicle in primary frontal cortex cultures Treatment 3h 24h ACSF 100% 100% PCP (10µM) 101.8±3.8% 82.2±4.4% NRG1 (0.01nM) + PCP 104.7±2.1% 80.2±7.5% NRG1 (0.1nM) + PCP 101.9±3.9% 97.5±6.4% NRG1 (1nM) + PCP 94.1±2.5% 90.6±5.9% NRG1 (10nM) + PCP 94.9±6.6% 104.7±3.2% NRG1 (0.01nM) 79.9±3.4% 108.7±5.6% NRG1 (0.1nM) 117.3±4.8% 109.8±7.3% NRG1 (1nM) 105.5±6.7% 113.4±9.6% Primary cell cultures of frontal cortex cells from mice (post natal day 1-2, n=5- 7/group) were treated with vehicle (artificial cerebrospinal fluid), PCP, NRG1+PCP or NRG1 (treatment concentration) for 3h or 24h after six days in vitro. mRNA expression levels of NR1 were analysed via quantitative real time polymerase chain reaction (qPCR). Values are presented as the mean percentage expression difference compared to vehicle treatment ± SEM.

103

4.4 Discussion

My frontal cortex cell culture experiments showed that PCP changed gene expression that is relevant for GABAergic signalling. Combined treatment with NRG1 prevented the PCP effect in a dose-dependent manner. Elements of interneuron and pyramidal neuron synapses were, however, unaffected by the applied treatments.

4.4.1 NRG1 treatment prevents PCP-induced mRNA expression changes relevant for GABAergic signalling

PCP treatment caused an increase in GAD67 and PV mRNA expression while reducing GAD65 mRNA expression in the frontal cortex cultures after 24h (Figure 23, Figure 24 and Figure 25). The altered expression of these genes indicates that PCP treatment influenced the activity level of the GABAergic interneurons in the primary cell culture. To ensure that the observed differences in gene expression are the direct consequence of the applied treatments, I took several precautions to minimise the impact of confounding factors: all cultures were treated in medium-free ACSF to avoid that factors of the enriched medium could mask the actual treatment effect. Additionally, the molarity of the treatment solution was adjusted to be equivalent to naturally pro- duced CSF (290 mOsm/kg, (Rosen & Andrew 1990)), ensuring normal electrochemical cell responses. Finally, as PCP has been reported to have neurotoxic properties, frontal cortex cultures were treated with a PCP concentration below its cytotoxic levels (Väisänen et al. 1999).

The effect of local PCP administration on GABAergic interneuron mRNA or protein expression levels in vitro has not previously been reported, limiting the ability to evaluate the observed PCP effect against existing data. Systemic sub-chronic administration of PCP in vivo has been reported to reduce GAD67 and PV mRNA and protein expression in the frontal cortex of adult rodents (Abdul-Monim et al. 2007; Amitai et al. 2012; Cochran et al. 2003; McKibben et al. 2010; Reynolds et al. 2004; Thomsen et al. 2010; Thomsen et al. 2009). While these treatment effects were not observed in all PCP rodent studies (Benneyworth et al. 2011; Peat & Gibb 1983), they show a clear difference to the present cell culture experiment. The differing PCP 104

treatment effects between my in vitro results and published systemic in vivo studies on GAD67 and PV are likely due to interactions between cortical and subcortical regions following systemic PCP treatment which are excluded in our experiment (Amitai et al. 2012; Benneyworth et al. 2011; Cochran et al. 2003; Del Arco et al. 2007; Martin et al. 1998; McKibben et al. 2010; Thomsen et al. 2010; Thomsen et al. 2009). Neurochemical changes following peripheral PCP treatment are the consequence of local receptor actions and several excitatory afferent connections to the PFC, as discussed in relation to my microdialysis findings (see 3.4.1.2 for details), differing in vitro findings are therefore to be expected. Furthermore, the significantly longer treatment duration in the animal experiments (24h vs 7 to 21 days) is also likely to contribute to the treatment effect difference.

4.4.1.1 NRG1 signalling has the potential to address GABAergic signalling impairments Altered mRNA and protein expression levels relevant to GABAergic signalling have been identified in frontal cortex regions of patients with schizophrenia (see 1.1.2.3 for details) (Fung et al. 2010; Hashimoto, Arion, et al. 2008; Konradi et al. 2011; Woo et al. 2004). Existing antipsychotics have consequentially been assessed on their ability to reduce or prevent some of the reported GABAergic system alterations in pre-clinical studies. So far, only clozapine has been reported to be able to reverse PV and GAD67 expression alterations in the sub-chronic PCP rat model, while neither haloperidol nor risperdine reduced the PCP treatment effect (Cochran et al. 2003; Amitai et al. 2012; McKibben et al. 2010). Remarkably, combined treatment with NRG1 prevented the PCP-induced mRNA expression alteration of GAD67, GAD65 and PV (Figure 23, Figure 24 and Figure 25). This might be an indication that the treatment effect of NRG1 is comparable to the described clozapine treatment effect (Cochran et al. 2003) My results further suggest that the NRG1 signalling system might have the potential of addressing pathophysiological alterations of the GABAergic interneuron system.

4.4.1.2 The NRG1 treatment effect against PCP is dose-dependent The ability of NRG1 treatment to prevent the PCP-induced mRNA expression differences was dose-dependent. I identified a u-shaped-like concentration effect of NRG1 in preventing the GAD67 mRNA increase (Figure 23), GAD65 mRNA decrease 105

(Figure 24) and PV mRNA increase (Figure 25) in the presence of PCP. NRG1 at 0.1nM was most effective in preventing the induced changes for each target. The gene expression level differences between vehicle treatment and low (0.01nM) or high (1nM) NRG1 with PCP was larger than compared to 0.1nM NRG1 with PCP treatment. The resulting concentration effect is similar to the observed u-shaped dose-response curve from the behaviour experiments (Figure 10). A possible molecular explanation for this dose-response effect has been suggested in earlier publications, which proposed a u-shaped activation pattern for ERBB4 receptor phosphorylation (Role & Talmage 2007). ERBB4 stimulation by NRG1 leads to phosphorylation of intracellular signalling kinases such as phosphoinositide 3-kinase (PI3K) (Krivosheya et al. 2008), protein kinase B (AKT) regulating intracellular Ca2+ levels (Schapansky et al. 2009) and gene expression (Allison et al. 2011) (Figure 3). Fur- thermore, ERBB4 activation leads to (likely indirect) phosphorylation of neurotransmitter receptor kinases including proto-oncogene tyrosine-protein kinases SRC and FYN which influence functional properties of NMDA receptors (Bjarnadottir et al. 2007; Pitcher et al. 2011; Yu et al. 1997) (Figure 3). The reported dose-dependent ERBB4 activation by NRG1 might consequentially lead to differing kinase phosphorylation patterns resulting in the observed dose-response effect seen for GAD67, GAD65 and PV mRNA levels.

4.4.1.3 Treatment effect on GABAergic system is duration dependent NRG1 retrodialysis had an immediate effect on extracellular neurotransmitter levels in the PFC in my microdialsyis study. I therefore assess the effects different NRG1 and PCP treatment durations might have on GABAergic system relevant mRNA expression after 3h and 24h. Interestingly, the observed treatment effect on GAD67, GAD65 and PV mRNA expression appeared to be strongly influenced by the treatment duration. None of the differences observed after 24h treatment were present after 3h treatment, with GAD65 and GAD67 mRNA expression levels being indis- tinguishable between treated and untreated cultures. NRG1 at 0.1nM significantly increased PV mRNA expression levels after the 3h, but not 24h treatment period (Figure 25). The increase in PV might be a compensatory consequence of short-term PI3K stimulation and resulting release of intracellular Ca2+ at this concentration (Gu et al. 2005).

106

PCP treatment has been reported to have a treatment duration-dependent effect on mRNA expression after treatment onset. Adachi and colleagues observed increases in brain derived neurotrophic factor expression 6h, but not 3h, after PCP injection (Adachi et al. 2013). While Gao and colleagues showed that increased expression of the early response gene cfos peaks between 2 and 3h after PCP treatment (Gao et al. 1998). This work, together with my results indicate that structural responses to PCP treatment can be expected within 24h and are more likely to occur several hours into the presence of PCP. This observation also suggests that a longer treatment duration in the microdialysis study could potentially have triggered a change in GABA levels in the PFC. The relevance of treatment duration has also been reported for local NRG1 administration relevant to neurotransmitter responses. In a hippocampal neuronal cell culture study, Schapansky and colleagues assessed the ability of NRG1 to modulate intracellular calcium regulation in response to acute glutamatergic neurotransmission (Schapansky et al. 2009). Cells treated with AMPA receptor agonists showed a marked reduction in intracellular Ca2+ response after NRG1 treatment for 24h, but not 1h (Schapansky et al. 2009). Furthermore, responses to NMDA receptor agonists remained unaltered (Schapansky et al. 2009). These observations suggest an interaction of NRG1 signalling with neurotransmitter receptors beyond acute neurotransmitter release modulation. It is possible that such a modulatory action of NRG1 caused the observed mRNA expression difference in my frontal cortex cultures, especially as the three targets are critical to the electrochemical responsiveness of GABAergic interneuron. Interestingly, the experiment by Schapansky and colleagues shows that intracellular Ca2+ regulation responds differently to extracellular glutamate after exposure to NRG1 for 24h, but not 1h (Schapansky et al. 2009). As PCP treatment has been reported to influence intracellular Ca2+ levels by blocking Ca2+ ATPase in cerebellar interneurons (Pande et al. 1999), the ability of NRG1 to influence intracellular Ca2+ regulation might contribute to the observed treatment effect in my experiment. Fu- ture studies into the effect of NRG1 on neurotransmission could therefore include in vitro Ca2+ imaging to investigate this possible interaction.

107

Existing antipsychotics such as clozapine, haloperidol and aripiprazole have shown to increase the expression of GAD67 mRNA (Peselmann et al. 2013; Zink et al. 2004), but not PV (Cahir et al. 2005), when administered to healthy animals. Future experiments will need to assess how this effect contributes to the antipsychotic potential of these drugs. NRG1 treatment did not show such an effect in my study at the treatment-relevant concentration. However, the long-term effect of NRG1 treatment on these targets in unimpaired interneurons remains to be determined, as well as possible treatment response differences following in vivo administration.

4.4.2 Pre- and post-synaptic markers under the influence of PCP and NRG1

The mRNA expression of pre-synaptic (SYP) and post-synaptic (DISC1, NR1) markers assessed in my experiment was unaffected by PCP treatment (Table 6, Table 7 and Table 8). Several studies have reported differences to SYP (Fattorini et al. 2009; Thomsen et al. 2010; Thomsen et al. 2009; Wang et al. 2004) and NR1 (Hana- nia et al. 1999; Lindahl & Keifer 2004; Wang et al. 2005; Wang et al. 1999; Yu et al. 2002) mRNA and protein expression following peripheral PCP treatment in rodents, with no data on DISC1. Treatment duration (3h or 24h compared to 5 and more days), PCP concentration and in vivo vs in vitro administration differences are the most-likely reasons for the varying outcomes. The three pre- and post-synaptic targets are expressed in different cell types, including GABAergic interneurons (Steinecke et al. 2012; Geddes et al. 2011; Bragina et al. 2010). Since the expression of the three targets was not affected by PCP, I can assume that the observed differences in GAD67, GAD65 and PV expression are likely due to a change in demand for the respective proteins and not due to a change in cell or synapse numbers in the primary cultures. NRG1 treatment had no effect on the mRNA expression of SYP, NR1 or DISC1. A previous study utilizing a cancer cell line had shown that NRG1 treatment for 2h induced a lasting DISC1 increase (Seshadri et al. 2010), but no data from neuronal cultures is available. Using frontal cortex cultures, a recent study has shown that treatment with NRG1 for 48h induces synaptogenesis (Ting et al. 2011). This finding suggests that a longer treatment duration might be required to trigger such an effects.

108

Furthermore, it remains to be seen if NRG1 can induce synaptogenesis in neurons in the presence of PCP or other modelling mechanisms. Current antipsychotics have shown to alter the expression of SYP (Eastwood et al. 2000; Ozdemir et al. 2012; Park et al. 2013; Scarr & Dean 2012), NR1 (Fumagalli et al. 2008; O’Connor et al. 2006; Schmitt et al. 2003) and DISC1 (Chiba et al. 2006) both in vitro and in vivo. These studies have been performed in physiological healthy environments and it therefore remains unclear if these effects contribute to the therapeutic or adverse treatment response. In summary, the frontal cortex cell culture study shows that stimulating the NRG1 signalling pathway can prevent PCP-induced gene expression changes relevant to GABAergic signalling. The molecular mechanisms behind the observed dose- response curve will need to be studied in future experiments to understand if similar mechanisms might have contributed to the dose-dependent treatment effect of NRG1 in the behaviour studies. No differences have been observed in the pre- and postsynaptic targets for any of the treatment combinations. A longer treatment duration might be necessary to assess the potential of NRG1 treatment to address impairments schizophrenia-relevant impairments in synapse structures.

109

5 OVERALL DISCUSSION AND CONCLUSION

5.1 The Antipsychotic Potential of NRG1 Treatment

Current antipsychotics are of limited benefit against a wide range of schizophrenia symptoms (Tandon et al. 2009). This restriction has been attributed to their predominantly anti-dopaminergic action, which does not address the GABAergic and glutamatergic neurotransmission differences observed in patients with schizophrenia (reviewed in 1.1.3). Drug development efforts have therefore shifted towards altering GABAergic and glutamatergic neurotransmission (reviewed in 1.1.4 and by Vinkers et al. (2010). Past research has shown that NRG1 administration can alter dopaminergic transmission in vivo as well as GABAergic and glutamatergic transmission in vitro (reviewed in 1.2.3 and by (Buonanno 2010; Rico & Marín 2011)). I hypothesised that NRG1 treatment has antipsychotic-like potential in a rodent schizophrenia model. Limited knowledge is available on the effects of NRG1 on neurotransmission in the PFC, hippocampus and striatum. However, neither the behavioural consequences of NRG1 administration nor the effect of NRG1 treatment on schizophrenia-relevant paradigms have been explored. I therefore studied the effects of NRG1 treatment against PCP-induced schizophrenia-like behaviour impairments in mice. Animals receiving central or peripheral injection of NRG1 together with a peripheral PCP injection, displayed normal locomotor activity, unlike animals only receiving PCP. Furthermore, peripheral NRG1 prevented sensorimotor gating impairment otherwise observed following PCP alone. Preventing schizophrenia-relevant behaviour impairments otherwise caused by PCP indicates antipsychotic-like characteristics of NRG1 treatment, since several available antipsychotics show similar treatment effects against PCP in rodents (Table 1). Critically, while several of these antipsychotics negatively affect healthy behaviours (Adage et al. 2008; Nordquist et al. 2008; Okuyama et al. 1999), NRG1 treatment caused no change in locomotor activity nor sensorimotor gating. Finally, the rapid treatment response following peripheral injections of NRG1 suggests a potential suitability as a peripherally administrable substance against CNS impairments.

110

The outcome of the behaviour experiments formed the foundation for biochemical investigations into the antipsychotic potential of NRG1 treatment with a focus on the GABAergic and glutamatergic systems. Altered neurotransmitter levels and neurotransmission-related protein expression have been observed in the PFC and hippocampus of patients with schizophrenia. The GABAergic and glutamatergic neurotransmitter systems appear particularly affected in both of these regions (reviewed in 1.1.2.2 and 1.1.2.3). I therefore studied the effect of NRG1 treatment on acute GABAergic and glutamatergic neurotransmission in the presence of PCP in the PFC and dCA1 of mice. In the PFC, NRG1 reduced glutamate and glycine levels, while combined NRG1 and PCP treatment triggered an increase in extracellular glycine. Despite a range of pharmacological actions of PCP treatment, extracellular GABA, glutamate and glycine levels were unaffected. As there was no PCP-induced neurotransmission deficit in the PFC, it is not possible to say whether NRG1 treatment would have attenuated altered levels of these neurotransmitters. The NRG1 treatment effect on glutamate and glycine, however, provides direct in vivo evidence for the proposed involvement of NRG1 signalling in adult neurotransmission and might have contributed to the prevention of PCP-induced behavioural impairments (see 5.1.1 below for details). In the hippocampus, PCP treatment triggered a significant increase in GABA levels without affecting glutamate or glycine. The GABA increase was not observed following combined NRG1 and PCP treatment, suggesting that NRG1 prevented the original PCP effect. An increase in glycine was again observed following combined treatment, this time accompanied by a reduction in glutamate levels. The differences in neuronal network structure and cellular make up between the PFC and hippocampus likely accounts for the variation in neurotransmitter level differences between these regions. Several pharmacological treatment options are currently under development aiming to increase glycine and adjust GABA-glutamate signalling in patients with schizophrenia (reviewed in (Javitt 2009; Miyamoto et al. 2012; Rudolph & Knoflach 2011)). The ability of NRG1 treatment to alter GABA, glutamate and glycine levels in the presence of PCP also suggests that NRG1 treatment has the potential to alter disrupted neurotransmission in patients with schizophrenia. In the present study, the behavioural and microdialysis experiments focused on the treatment effects of NRG1

111

in the acute PCP rodent model with relevance to acute psychotic episodes experienced by patients with schizophrenia (Adell et al. 2012; Javitt et al. 2012; Pratt et al. 2012). However, neurotransmission impairments are present throughout the course of the disease, as suggested by differences in neurotransmission-relevant protein expression discovered in post-mortem tissue of patients (see 1.1.2 for details). The purpose of the final experiment of my PhD was thus to form an initial investigation into the effects of prolonged NRG1 treatment on altered neurotransmission. I treated frontal cortex networks with NRG1 in the presence of PCP, measuring components of the GABAergic system relevant to the pathophysiology of schizophrenia. PCP changed the mRNA expression of GABA producing enzymes and the calcium binding protein PV. Combined treatment with NRG1 dose-dependently prevented these PCP-induced expression differences in the dissociated frontal cortex cultures. While results from primary cell culture experiments can only be considered as an initial step in the treatment development process, the outcome of this study shows that extended stimulation of the NRG1 signalling pathway has the potential to change GABAergic interneuron activity during modelled neurotransmission impairments. Altered neurotransmission and aberrant inhibitory synaptic function in the PFC and the hippocampus are highly implicated in the symptoms of schizophrenia (Hall et al. 2010; Keshavan et al. 2008; Gonzalez-Burgos et al. 2011). The ability of NRG1 to improve aspects of the GABAergic inhibitory system in frontal cortex neurons in vitro is therefore of strong relevance for its potential as an antipsychotic treatment option against altered GABA signalling observed in patients (Kegeles et al. 2012). The identified effect of NRG1 on GABAergic signalling also opens the possibility of a combined NRG1-antipsychotic therapy, since current antipsychotics have little effect on impaired GABAergic signalling in individuals with schizophrenia.

The present work is an initial exploration into the treatment potential of NRG1 against PCP-induced impairments in mice with relevance to schizophrenia. Using three separate experimental approaches enabled me to study the treatment potential of NRG1 on different CNS complexity levels. This strategy, however, limited the possibility of drawing conclusions about the mechanisms underlying the behaviour and molecular findings. In future experiments, including inhibitors for neurotransmitter transporters (e.g. GAT) and receptors (e.g. NMDAR and D2) as well as intracel-

112

lular kinases (e.g. FYN/SRC and AKT), will aid the identification of the mechanisms underlying the observed treatment effect following combined NRG1 and PCP administration.

5.1.1 Possible mechanisms underlying the effect of NRG1 on PCP impairments

While the experiments of my PhD focused on the antipsychotic potential of NRG1, the observed results allow me to hypothesise regarding the underlying molecular mechanism of the NRG1-PCP interaction. PCP is a commonly used substance to induce psychotic-like deficits against which antipsychotics can be tested for their therapeutic potential. PCP has been reported to inhibit membrane bound Ca2+ ATPase on GABAergic interneurons (Pande et al. 1999), additionally to its pharmacological action on neurotransmitter receptors (see 1.4.2.2 for details). Consequently, PCP-induced increase in intracellular Ca2+ could have triggered the observed demand for Ca2+ binding proteins (Figure 25) as well as the elevation in GABA levels (Figure 19) (Bardo et al. 2002). Increased GABA signalling has been previously related to hyperlocomotion and disrupted PPI (Yeomans et al. 2010; Yu et al. 2013). The effect of NRG1 against the PCP-induced changes could have been mediated by reducing intracellular Ca2+ through activation of PI3K via ERBB4 (Z. Liu et al. 2011), subsequently preventing an increase in GABA release and behavioural impairments. Additionally, the NRG1-induced increase in glycine in the presence of PCP (Figure 18 and Figure 21) is likely to have contributed to the reduction of PCP-induced behavioural impairments. Increase in glycine has been shown to attenuate PPI impairments following treatment with the NMDAR antagonist MK-801, prevent PCP- induced hyperlocomotion and prevent PCP-induced increases in extracellular dopamine (Javitt et al. 1999; Kanahara et al. 2008; Toth & Lajtha 1986). However, since PCP exerts its effect through a range of targets on both neurons and glia, additional molecular mechanisms are likely contributing to the observed NRG1- PCP treatment effects. While conclusions on the exact mechanisms underlying the observed results can surely not be drawn from my studies, they present a basis and allow the formation of a hypothesis for promising further investigation.

113

5.1.2 NRG1 treatment potential beyond schizophrenia

With my PhD I have focused on antipsychotic-relevant consequences of NRG1 signalling by using the PCP-induced schizophrenia rodent model. In recent years, however, several other research groups have studied the treatment potential of the NRG1- ERBB4 signalling pathway for a number of brain disorders. For example, Carlsson and colleagues have shown that NRG1 treatment can be neuroprotective in a Parkin- son’s disease mouse model by using the same NRG1 isoform as in my experiments (Carlsson et al. 2011). In regards to Alzheimer’s disease pathology, NRG1 treatment has also shown to be neuroprotective in pre-clinical in vitro and in vivo rodent models against amyloid protein neurotoxicity (Woo et al. 2012; Chen et al. 2011; Min et al. 2011; Ryu et al. 2012; Cui et al. 2013). The neuroprotective properties of NRG1 have also been explored in relation to neurological disorders such as stroke and nerve injury, showing promising early pre-clinical treatment effects (Li et al. 2012; Ronchi et al. 2013; Gauthier et al. 2013; Shyu et al. 2004; Xu et al. 2004; Guo et al. 2006; Syed et al. 2010). The most promising finding for the treatment relevance of NRG1, however, has been made in relation to cardiovascular disease (Wadugu & Kühn 2012). Patients with chronic heart failure have been treated with NRG1 for ten days leading to lasting improvements in cardiac function (Gao et al. 2010; Jabbour et al. 2011). No indication of nervous system impairments have been reported during and following the treatment period. While its efficacy for central nervous system pathologies remains to be evaluated, this suggests that treatment with NRG1 is tolerated by humans, which overcomes a critical hurdle in the drug development process.

5.2 Recommendation for Future Research

5.2.1 The role of dopamine in the treatment effect of NRG1

Differences in the dopaminergic neurotransmitter system in patients with schizophrenia are well documented, as descripted in section 1.1.2.1. The present experiments, however, assess the effect of NRG1 on the GABAergic and glutamatergic neurotransmitter system in the schizophrenia-relevant PCP model. The results from the microdialysis study suggest that additional transmitters are involved in the NRG1 114

treatment effect. Dopamine is a likely candidate, particularly as NRG1 retrodialysis has been shown to trigger an increase in extracellular dopamine levels in the hippocampus of mice (Kwon et al. 2008). A logical next step would therefore be to repeat the microdialysis experiment with NRG1 and PCP retrodialysis and quantify extracellular dopamine levels. Further, recent experiments indicate an involvement of the dopamine receptor D4 in the effect of NRG1 on hippocampal extracellular DA levels (Andersson et al. 2012). Including antagonists selective for different dopamine receptors in the retrodialysis solution would assist with the identification of the treatment-relevant NRG1 signalling pathway.

5.2.2 NRG1 treatment effect on cognitive performance

The outcome of the behaviour study indicates that NRG1 treatment has the potential to be beneficial against symptoms experienced during an acute psychotic episode, as modelled by PCP treatment in mice. The period between psychotic episodes is dominated by impairments to cognitive abilities (Tandon et al. 2009). It is therefore of strong clinical relevance to study the effect of NRG1 treatment on cognitive functions. Memory deficits and attention deficits have been modelled in rodents through different strategies, such as chronic amphetamine and PCP treatment, neonatal hippocampal lesion and perinatal immune activation (Mouri et al. 2013). Assessing the effect of NRG1 treatment on the performance in latent inhibition, 5-choice serial reaction time task and radial arm maze experiments with these animal models will shed light on the treatment potential against schizophrenia-relevant cognitive impairments.

5.2.3 Long-term effects of NRG1 treatment

The existing pre-clinical and clinical studies, as well as the experiments from my PhD, focus predominantly on acute or short-term NRG1 treatment effects. The neurotrophic role of NRG1 suggests the possibility that long-term changes in NRG1 signalling could result in lasting structural differences in several peripheral and central nervous system structures. Current transgenic and knockout animal models with mutations in the NRG1 or ERBB4 genes alter this signalling pathway from early development onwards and are therefore not suited to shed light on the consequences of NRG1 treatment on the ful- 115

ly developed organisms. Furthermore, the effect of NRG1 appears dose-dependent, which cannot easily be studied in transgenic animals. Recent animal work indicates that even a short duration of NRG1 treatment can induce lasting structural differences, with four days NRG1 treatment resulting in increased dentate gyrus neurogenesis 28 days later (Mahar et al. 2011). While this structural change has been suggested to be of potential benefit in relation to depression, structural changes in other regions could be of potential disadvantage. Furthermore, the observed reduction in glutamate by NRG1 observed in my microdialysis study and in vitro experiments might lead to compensatory mechanisms on both pre- and postsynaptic neurons (with unknown consequences ), if maintained for extended periods. The next step in evaluating the suitability of NRG1 as a treatment option for the mentioned disorders and especially schizophrenia, will therefore require detailed behavioural and molecular analysis following long-term peripheral NRG1 administration.

5.2.4 Potential treatment effect of NRG1 on existing impairments

My experiments show that NRG1 treatment can prevent PCP-induced behavioural and molecular changes in mice. To advance on the clinical relevance of the observed NRG1 treatment effect, future research will require the administration of NRG1 after the formation of clinical relevant impairments. By treating non-acute schizophrenia animal models, it will be possible to assess the ability of NRG1 treatment to restore pre-existing impairments. For example, the pre-natal immune challenge and pre-natal vitamin D deficiency models trigger lasting neurochemical differences such as altered dopamine metabolism and GABAergic interneuron function (McGrath et al. 2010; Meyer & Feldon 2010). The involvement of the NRG1 signalling in cell mechanisms beyond acute neurotransmission (e.g. synapse formation and nerve migration (Mei & Xiong 2008; Cahill et al. 2013)), might attenuate impairments caused by schizophrenia modelling strategies and would therefore have stronger relevance to a clinical setting.

116

5.2.5 NRG1 treatment effect on human nervous system

Different acute and chronic modelling strategies can provide environments where potential treatment characteristics of the NRG1 signalling pathway can be explored. While the ERBB4 expression pattern appears consistent across species (Neddens et al. 2011), differences have been observed in the NRG1 isoforms expressed, with NRG1 type IV only found in humans, not rodents (Tan et al. 2007; Shamir & Buo- nanno 2010). This highlights the need for comparing the effect of NRG1 treatment between the rodent and primate brain. Additionally, the broad distribution of ERBB3-4 receptors throughout peripheral tissue types will require careful analysis of potential species-specific adverse effects, such as those indicated by breast cancer studies in humans (Montero et al. 2008).

5.3 Conclusion

With the present series of experiments I have demonstrated that NRG1 treatment can prevent PCP-induced behavioural impairments, PCP-induced neurotransmission changes and PCP-induced disruptions to the frontal cortex neuronal network. Prevention of PCP-induced hyperlocomotion and PPI impairments demonstrates a clear involvement of the NRG1 signalling pathway in neuronal circuits relevant for spontaneous behaviour. The region-specific influence of NRG1 on glutamate and glycine levels confirms the involvement of NRG1-ERBB signalling in adult excitatory neurotransmission. Furthermore, preventing the PCP-induced GABA level increase in the hippocampus provides support for a functional role in inhibitory neurotransmitter hippocampal signalling. An involvement in the regulation of inhibitory neurotransmission of NRG1 was further supported by in vitro frontal cortex experiments where NRG1 treatment prevented changes to the expression of GABA producing enzymes. The treatment effects of NRG1 against PCP were dose-dependent in most experiments, suggesting that NRG1-ERBB signalling is tightly regulated in the adult nervous system. The combined outcomes of my experiments provide initial support for potential antipsychotic-like characteristics of NRG1 treatment in a mouse model with induced schizophrenia-relevant behavioural and molecular impairments. The observed effica- 117

cy of peripherally administered NRG1 to prevent PCP-induced behavioural deficits indicates its potential as a new therapy for schizophrenia patients. Exploring the consequences of NRG1 treatment in the long-term, on a broader spectrum of behavioural tasks and further schizophrenia-relevant molecular changes will be required to advance on the present findings and finally evaluate the suitability of NRG1 treatment in the clinic.

118

6 REFERENCES

Abdul-Monim, Z., Neill, J.C. & Reynolds, G.P., 2007. Sub-chronic psychotomimetic phencyclidine induces deficits in reversal learning and alterations in parvalbumin-immunoreactive expression in the rat. Journal of Psychopharmacolo- gy, 21(2), pp.198–205.

Abdul-Monim, Z., Reynolds, G.P. & Neill, J.C., 2003. The atypical antipsychotic ziprasidone, but not haloperidol, improves phencyclidine-induced cognitive deficits in a reversal learning task in the rat. Journal of Psychopharmacology, 17(1), pp.57–65.

Abe, S. et al., 2000. Differential expression of GABA(A) receptor subunit mRNAs and ligand binding sites in rat brain following phencyclidine administration. Synapse, 38(1), pp.51–60.

Abe, Y., Namba, H., et al., 2009. In situ hybridization reveals developmental regulation of ErbB1-4 mRNA expression in mouse midbrain: Implication of ErbB receptors for dopaminergic neurons. Neuroscience, 161(1), pp.95–110.

Abe, Y., Nawa, H. & Namba, H., 2009. Activation of epidermal growth factor receptor ErbB1 attenuates inhibitory synaptic development in mouse dentate gyrus. Neuroscience research, 63(2), pp.138–148.

Abekawa, T. et al., 2002. Effect of MS-153 on the development of behavioral sensi- tization to locomotion- and ataxia-inducing effects of phencyclidine. Psycho- pharmacology, 160(2), pp.122–131.

Abekawa, T., Ito, K. & Koyama, T., 2007. Different effects of a single and repeated administration of clozapine on phencyclidine-induced hyperlocomotion and glutamate releases in the rat medial prefrontal cortex at short- and long-term withdrawal from this antipsychotic. Naunyn Schmiedebergs Arch Pharmacol, 375(4), pp.261–71.

Abekawa, T., Ito, K. & Koyama, T., 2006. Role of the simultaneous enhancement of NMDA and dopamine D1 receptor-mediated neurotransmission in the effects of clozapine on phencyclidine-induced acute increases in glutamate levels in the rat medial prefrontal cortex. Naunyn-Schmiedeberg’s Archives of Phar- macology, 374(3), pp.177–193.

Adachi, N. et al., 2013. Phencyclidine-induced decrease of synaptic connectivity via inhibition of BDNF secretion in cultured cortical neurons. Cerebral cortex (New York, N.Y.: 1991), 23(4), pp.847–858.

Adage, T. et al., 2008. In vitro and in vivo pharmacological profile of AS057278, a selective D-amino acid oxidase inhibitor with potential anti-psychotic properties. European Neuropsychopharmacology, 18(3), pp.200–214.

Adams, B.W. & Moghaddam, B., 1998. Corticolimbic dopamine neurotransmission is temporally dissociated from the cognitive and locomotor effects of phencyclidine. Journal of Neuroscience, 18(14), pp.5545–5554.

119

Adams, B.W. & Moghaddam, B., 2001. Effect of clozapine, haloperidol, or M100907 on phencyclidine-activated glutamate efflux in the prefrontal cortex. Biol Psychiatry, 50(10), pp.750–7.

Adell, A. et al., 2012. Is the acute NMDA receptor hypofunction a valid model of schizophrenia? Schizophrenia bulletin, 38(1), pp.9–14.

Agim, Z.S. et al., 2013. Discovery, validation and characterization of Erbb4 and Nrg1 haplotypes using data from three genome-wide association studies of schizophrenia. PloS one, 8(1), p.e53042.

Akbarian, S. et al., 1995. Gene expression for glutamic acid decarboxylase is reduced without loss of neurons in prefrontal cortex of schizophrenics. Archives of general psychiatry, 52(4), pp.258–266.

Alaerts, M. et al., 2009. Support for NRG1 as a susceptibility factor for schizophrenia in a northern Swedish isolated population. Archives of General Psychia- try, 66(8), pp.828–837.

Allen, J.A. et al., 2011. Schizophrenia risk gene CAV1 is both pro-psychotic and required for atypical antipsychotic drug actions in vivo. Translational psychiatry, 1, p.e33.

Allison, J.G. et al., 2011. The ERBB4 intracellular domain (4ICD) regulates NRG1- induced gene expression in hippocampal neurons. Neuroscience research, 70(2), pp.155–163.

Amargos-Bosch, M. et al., 2006. Clozapine and olanzapine, but not haloperidol, suppress serotonin efflux in the medial prefrontal cortex elicited by phencyclidine and ketamine. Int J Neuropsychopharmacol, 9(5), pp.565–73.

Amir, A. et al., 1985. Effects of phencyclidine and analog drugs on acetylcholine receptor of cultured muscle cells. Biochemical pharmacology, 34(7), pp.949– 954.

Amitai, N. et al., 2012. Repeated phencyclidine administration alters glutamate release and decreases GABA markers in the prefrontal cortex of rats. Neuro- pharmacology, 62(3), pp.1422–31.

Amitai, N., Semenova, S. & Markou, A., 2007. Cognitive-disruptive effects of the psychotomimetic phencyclidine and attenuation by atypical antipsychotic medications in rats. Psychopharmacology, 193(4), pp.521–537.

Andersson, R.H. et al., 2012. Neuregulin and dopamine modulation of hippocampal gamma oscillations is dependent on dopamine D4 receptors. Proceedings of the National Academy of Sciences of the United States of America, 109(32), pp.13118–13123.

Del Arco, A. et al., 2007. Stimulation of D2 receptors in the prefrontal cortex reduces PCP-induced hyperactivity, acetylcholine release and dopamine metabolism

120

in the nucleus accumbens. Journal of Neural Transmission (Vienna, Austria: 1996), 114(2), pp.185–193.

Del Arco, A. & Mora, F., 2009. Neurotransmitters and prefrontal cortex-limbic system interactions: implications for plasticity and psychiatric disorders. J Neu- ral Transm, 116(8), pp.941–52.

Arguello, P.A. & Gogos, J.A., 2006. Modeling madness in mice: one piece at a time. Neuron, 52(1), pp.179–196.

Arguello, P.A. & Gogos, J.A., 2011. Psychiatric genetics and the generation of mutant animal models. In Animal Models of Schizophrenia and Related Disor- ders. Neuromethods. Totowa, NJ: Humana Press, pp. 189–209.

Ben-Ari, Y. et al., 2007. GABA: a pioneer transmitter that excites immature neurons and generates primitive oscillations. Physiological reviews, 87(4), pp.1215– 1284.

Asinof, S.K. & Paine, T.A., 2013. Inhibition of GABA synthesis in the prefrontal cortex increases locomotor activity but does not affect attention in the 5- choice serial reaction time task. Neuropharmacology, 65, pp.39–47.

Assié, M.B., Cosi, C. & Koek, W., 1997. 5-HT1A receptor agonist properties of the antipsychotic, nemonapride: comparison with bromerguride and clozapine. European journal of pharmacology, 334(2-3), pp.141–147.

Baker, D.A. et al., 2008. Contribution of cystine-glutamate antiporters to the psychotomimetic effects of phencyclidine. Neuropsychopharmacology, 33(7), pp.1760–1772.

Bakshi, V.P. & Geyer, M.A., 1995. Antagonism of phencyclidine-induced deficits in prepulse inhibition by the putative atypical antipsychotic olanzapine. Psycho- pharmacology, 122(2), pp.198–201.

Bakshi, V.P. & Geyer, M.A., 1997. Phencyclidine-induced deficits in prepulse inhibition of startle are blocked by prazosin, an alpha-1 noradrenergic antagonist. The Journal of pharmacology and experimental therapeutics, 283(2), pp.666– 674.

Bakshi, V.P., Swerdlow, N.R. & Geyer, M.A., 1994. Clozapine antagonizes phencyclidine-induced deficits in sensorimotor gating of the startle response. Jour- nal of Pharmacology and Experimental Therapeutics, 271(2), pp.787–794.

Ballmaier, M. et al., 2007. Cannabinoid receptor antagonists counteract sensorimotor gating deficits in the phencyclidine model of psychosis. Neuropsychophar- macology, 32(10), pp.2098–2107.

Barakat, A. et al., 2010. Decreased Neuregulin 1 C-terminal fragment in Brodmann’s area 6 of patients with schizophrenia. Schizophrenia Research, 124(1-3), pp.200–207.

121

Bardo, S., Robertson, B. & Stephens, G.J., 2002. Presynaptic internal Ca2+ stores contribute to inhibitory neurotransmitter release onto mouse cerebellar Purkinje cells. British journal of pharmacology, 137(4), pp.529–537.

Bare, D.J., Becker-Catania, S.G. & DeVries, G.H., 2011. Differential localization of neuregulin-1 type III in the central and peripheral nervous system. Brain Re- search, 1369, pp.10–20.

Barkus, E. & Murray, R.M., 2010. Substance use in adolescence and psychosis: clari- fying the relationship. Annual Review of Clinical Psychology, 6, pp.365–389.

De Bartolomeis, A., Buonaguro, E.F. & Iasevoli, F., 2013. Serotonin-glutamate and serotonin-dopamine reciprocal interactions as putative molecular targets for novel antipsychotic treatments: from receptor heterodimers to postsynaptic scaffolding and effector proteins. Psychopharmacology, 225(1), pp.1–19.

Bartos, M., Vida, I. & Jonas, P., 2007. Synaptic mechanisms of synchronized gamma oscillations in inhibitory interneuron networks. Nature reviews. Neurosci- ence, 8(1), pp.45–56.

Bast, T. & Feldon, J., 2003. Hippocampal modulation of sensorimotor processes. Progress in neurobiology, 70(4), pp.319–345.

Beasley, C.L. & Reynolds, G.P., 1997. Parvalbumin-immunoreactive neurons are reduced in the prefrontal cortex of schizophrenics. Schizophrenia Research, 24(3), pp.349–355.

Beaulieu, J.-M. & Gainetdinov, R.R., 2011. The physiology, signaling, and pharmacology of dopamine receptors. Pharmacological reviews, 63(1), pp.182–217.

Benardo, L.S., 1997. Recruitment of GABAergic inhibition and synchronization of inhibitory interneurons in rat neocortex. Journal of neurophysiology, 77(6), pp.3134–3144.

Benes, F.M., Khan, Y., et al., 1996. Differences in the subregional and cellular distribution of GABAA receptor binding in the hippocampal formation of schizophrenic brain. Synapse (New York, N.Y.), 22(4), pp.338–349.

Benes, F.M., Vincent, S.L., et al., 1996. Up-regulation of GABAA receptor binding on neurons of the prefrontal cortex in schizophrenic subjects. Neuroscience, 75(4), pp.1021–1031.

Bennett, M., 2009. Positive and negative symptoms in schizophrenia: the NMDA receptor hypofunction hypothesis, neuregulin/ErbB4 and synapse regression. The Australian and New Zealand Journal of Psychiatry, 43(8), pp.711–721.

Benneyworth, M.A. et al., 2011. Failure of NMDA receptor hypofunction to induce a pathological reduction in PV-positive GABAergic cell markers. Neuroscience Letters, 488(3), pp.267–271.

Bersell, K. et al., 2009. Neuregulin1/ErbB4 signaling induces cardiomyocyte proliferation and repair of heart injury. Cell, 138(2), pp.257–270. 122

Bertram, I. et al., 2007. Immunohistochemical evidence for impaired neuregulin-1 signaling in the prefrontal cortex in schizophrenia and in unipolar depression. Annals of the New York Academy of Sciences, 1096, pp.147–156.

Bhugra, D., 2005. The global prevalence of schizophrenia. PLoS medicine, 2(5), p.e151; quiz e175.

Bilder, R.M. et al., 2002. Neurocognitive effects of clozapine, olanzapine, risperidone, and haloperidol in patients with chronic schizophrenia or schizoaffec- tive disorder. The American journal of psychiatry, 159(6), pp.1018–1028.

Bissonette, G.B. & Powell, E.M., 2012. Reversal learning and attentional set-shifting in mice. Neuropharmacology, 62(3), pp.1168–1174.

Bjarnadottir, M. et al., 2007. Neuregulin1 (NRG1) signaling through Fyn modulates NMDA receptor phosphorylation: differential synaptic function in NRG1+/- knock-outs compared with wild-type mice. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 27(17), pp.4519–4529.

Bortolozzi, A. et al., 2010. Dopamine release induced by atypical antipsychotics in prefrontal cortex requires 5-HT1A receptors but not 5-HT2A receptors. In- ternational Journal of Neuropsychopharmacology, 13(10), pp.1299–1314.

Boulay, D. et al., 2008. Characterization of SSR103800, a selective inhibitor of the glycine transporter-1 in models predictive of therapeutic activity in schizophrenia. Pharmacology Biochemistry and Behavior, 91(1), pp.47–58.

Bourdelais, A.J. & Deutch, A.Y., 1994. The effects of haloperidol and clozapine on extracellular GABA levels in the prefrontal cortex of the rat: an in vivo microdialysis study. Cerebral cortex (New York, N.Y.: 1991), 4(1), pp.69–77.

Bousman, C.A. et al., 2013. Effects of NRG1 and DAOA genetic variation on transition to psychosis in individuals at ultra-high risk for psychosis. Translational psychiatry, 3, p.e251.

Boyer, P. et al., 2007. Hippocampal abnormalities and memory deficits: new evidence of a strong pathophysiological link in schizophrenia. Brain research reviews, 54(1), pp.92–112.

Braff, D.L. & Geyer, M.A., 1990. Sensorimotor gating and schizophrenia. Human and animal model studies. Archives of general psychiatry, 47(2), pp.181–188.

Braff, D.L., Geyer, M.A. & Swerdlow, N.R., 2001. Human studies of prepulse inhibition of startle: normal subjects, patient groups, and pharmacological studies. Psychopharmacology, 156(2-3), pp.234–258.

Bragina, L. et al., 2010. Heterogeneity of glutamatergic and GABAergic release ma- chinery in cerebral cortex: analysis of synaptogyrin, vesicle-associated membrane protein, and syntaxin. Neuroscience, 165(3), pp.934–943.

123

Brinkmann, B.G. et al., 2008. Neuregulin-1/ErbB Signaling Serves Distinct Func- tions in Myelination of the Peripheral and Central Nervous System. Neuron, 59(4), pp.581–595.

Buchsbaum, M.S. et al., 2009. FDG-PET and MRI imaging of the effects of sertindole and haloperidol in the prefrontal lobe in schizophrenia. Schizophrenia research, 114(1-3), pp.161–171.

Buchsbaum, M.S. et al., 2007. FDG-PET in never-previously medicated psychotic adolescents treated with olanzapine or haloperidol. Schizophrenia research, 94(1-3), pp.293–305.

Buonanno, A., 2010. The neuregulin signaling pathway and schizophrenia: from genes to synapses and neural circuits. Brain Research Bulletin, 83(3-4), pp.122–131.

Burban, A. et al., 2010. Modulation of prepulse inhibition and stereotypies in rodents: no evidence for antipsychotic-like properties of histamine H-3-receptor inverse agonists. Psychopharmacology, 210(4), pp.591–604.

Cahill, M.E. et al., 2013. Neuregulin1 signaling promotes dendritic spine growth through kalirin. Journal of neurochemistry, 126(5), pp.625–635.

Cahir, M. et al., 2005. Chronic haloperidol or clozapine treatment does not alter parvalbumin immunoreactivity in the rat frontal cortex or hippocampus. Neuro- science letters, 373(1), pp.57–60.

Calvo, M. et al., 2010. Neuregulin-ErbB signaling promotes microglial proliferation and chemotaxis contributing to microgliosis and pain after peripheral nerve injury. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 30(15), pp.5437–5450.

Cardno, A.G. & Gottesman, I.I., 2000. Twin studies of schizophrenia: from bow-and- arrow concordances to star wars Mx and functional genomics. American journal of medical genetics, 97(1), pp.12–17.

Carli, M., Calcagno, E., Mainolfi, P., et al., 2011. Effects of aripiprazole, olanzapine, and haloperidol in a model of cognitive deficit of schizophrenia in rats: Rela- tionship with glutamate release in the medial prefrontal cortex. Psychophar- macology, 214(3), pp.639–652.

Carli, M., Calcagno, E., Mainini, E., et al., 2011. Sertindole restores attentional performance and suppresses glutamate release induced by the NMDA receptor antagonist CPP. Psychopharmacology, 214(3), pp.625–637.

Carlsson, A. & Carlsson, M.L., 2006. A dopaminergic deficit hypothesis of schizophrenia: the path to discovery. Dialogues in Clinical Neuroscience, 8(1), pp.137–142.

124

Carlsson, T. et al., 2011. Systemic administration of neuregulin-1β1 protects dopaminergic neurons in a mouse model of Parkinson’s disease. Journal of Neu- rochemistry, 117(6), pp.1066–1074.

Carpenter, W.T. & Buchanan, R.W., 2008. Lessons to take home from CATIE. Psy- chiatric services (Washington, D.C.), 59(5), pp.523–525.

Cartmell, J., Monn, J.A. & Schoepp, D.D., 1999. The metabotropic glutamate 2/3 receptor agonists LY354740 and LY379268 selectively attenuate phencyclidine versus d-amphetamine motor behaviors in rats. Journal of Pharmacolo- gy and Experimental Therapeutics, 291(1), pp.161–170.

Casadio, P. et al., 2011. Cannabis use in young people: the risk for schizophrenia. Neuroscience and Biobehavioral Reviews, 35(8), pp.1779–1787.

Chamberland, S. & Topolnik, L., 2012. Inhibitory control of hippocampal inhibitory neurons. Frontiers in Neuroscience, 6, p.165.

Chambers, J.S. et al., 2005. Growth-associated protein 43 (GAP-43) and synaptophysin alterations in the dentate gyrus of patients with schizophrenia. Progress in neuro-psychopharmacology & biological psychiatry, 29(2), pp.283–290.

Chana, G. et al., 2013. Biomarker investigations related to pathophysiological pathways in schizophrenia and psychosis. Frontiers in cellular neuroscience, 7, p.95.

Chang, Q. & Fischbach, G.D., 2006. An acute effect of neuregulin 1 beta to suppress alpha 7-containing nicotinic acetylcholine receptors in hippocampal interneurons. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 26(44), pp.11295–11303.

Chefer, V.I. et al., 2009. Overview of brain microdialysis. Current protocols in neuroscience / editorial board, Jacqueline N. Crawley ... [et al.], Chapter 7, p.Unit7.1.

Chen, P. et al., 2012. Analysis of association between common SNPs in ErbB4 and bipolar affective disorder, major depressive disorder and schizophrenia in the Han Chinese population. Progress in Neuro-Psychopharmacology & Biologi- cal Psychiatry, 36(1), pp.17–21.

Chen, Y. et al., 2011. Interfering effect and mechanism of neuregulin on experimental dementia model in rats. Behavioural Brain Research, 222(2), pp.321– 325.

Chen, Y.-J. et al., 2010. ErbB4 in parvalbumin-positive interneurons is critical for neuregulin 1 regulation of long-term potentiation. Proceedings of the Nation- al Academy of Sciences of the United States of America, 107(50), pp.21818– 21823.

Chen, Y.-J.J. et al., 2008. Type III neuregulin-1 is required for normal sensorimotor gating, memory-related behaviors, and corticostriatal circuit components. The

125

Journal of Neuroscience: The Official Journal of the Society for Neurosci- ence, 28(27), pp.6872–6883.

Chiba, S. et al., 2006. Effect of antipsychotic drugs on DISC1 and dysbindin expression in mouse frontal cortex and hippocampus. Journal of neural transmission (Vienna, Austria: 1996), 113(9), pp.1337–1346.

Chiusaroli, R. et al., 2010. Antipsychotic-like effects of the N-methyl-d-aspartate receptor modulator neboglamine: An immunohistochemical and behavioural study in the rat. Pharmacological Research, 61(5), pp.430–436.

Chong, V.Z. et al., 2008. Elevated neuregulin-1 and ErbB4 protein in the prefrontal cortex of schizophrenic patients. Schizophrenia Research, 100(1-3), pp.270– 280.

Chung, Y.-C. et al., 2004. Clozapine increases both acetylcholine and dopamine release in rat ventral hippocampus: role of 5-HT1A receptor agonism. Brain research, 1023(1), pp.54–63.

Churchwell, J.C. & Kesner, R.P., 2011. Hippocampal-prefrontal dynamics in spatial working memory: interactions and independent parallel processing. Behav- ioural brain research, 225(2), pp.389–395.

Cobb, S.R. et al., 1995. Synchronization of neuronal activity in hippocampus by individual GABAergic interneurons. Nature, 378(6552), pp.75–78.

Cochran, S.M. et al., 2003. Induction of metabolic hypofunction and neurochemical deficits after chronic intermittent exposure to phencyclidine: differential modulation by antipsychotic drugs. Neuropsychopharmacology: Official Pub- lication of the American College of Neuropsychopharmacology, 28(2), pp.265–275.

Cohen, R.M. et al., 1999. The brain metabolic patterns of clozapine- and fluphena- zine-treated female patients with schizophrenia: evidence of a sex effect. Neuropsychopharmacology: official publication of the American College of Neuropsychopharmacology, 21(5), pp.632–640.

Cui, W. et al., 2013. Neuregulin1beta1 antagonizes apoptosis via ErbB4-dependent activation of PI3-kinase/Akt in APP/PS1 transgenic mice. Neurochemical research, 38(11), pp.2237–2246.

Curzon, P. & Decker, M.W., 1998. Effects of phencyclidine (PCP) and (+)MK-801 on sensorimotor gating in CD-1 mice. Progress in Neuro- Psychopharmacology and Biological Psychiatry, 22(1), pp.129–146.

Daly, D.A. & Moghaddam, B., 1993. Actions of clozapine and haloperidol on teh extracellular levels of excitatory amino acids in the prefrontal cortex and striatum of conscious rats. Neuroscience Letters, 152(1-2), pp.61–64.

Deakin, I.H. et al., 2009. Behavioural characterization of neuregulin 1 type I overexpressing transgenic mice. Neuroreport, 20(17), pp.1523–1528.

126

Dejaegere, T. et al., 2008. Deficiency of Aph1B/C-gamma-secretase disturbs Nrg1 cleavage and sensorimotor gating that can be reversed with antipsychotic treatment. Proceedings of the National Academy of Sciences of the United States of America, 105(28), pp.9775–9780.

Deng, C. et al., 2013. Neuregulin-1 signalling and antipsychotic treatment: potential therapeutic targets in a schizophrenia candidate signalling pathway. Psycho- pharmacology, 226(2), pp.201–215.

Depboylu, C. et al., 2012. Neuregulin-1 receptor tyrosine kinase ErbB4 is upregulated in midbrain dopaminergic neurons in Parkinson disease. Neuroscience letters, 531(2), pp.209–214.

Désaméricq, G. et al., 2014. Long-term neurocognitive effects of antipsychotics in schizophrenia: a network meta-analysis. European journal of clinical pharmacology, 70(2), pp.127–134.

Dimayuga, F.O. et al., 2003. The neuregulin GGF2 attenuates free radical release from activated microglial cells. Journal of neuroimmunology, 136(1-2), pp.67–74.

Dracheva, S. et al., 2004. GAD67 and GAD65 mRNA and protein expression in cerebrocortical regions of elderly patients with schizophrenia. Journal of neuroscience research, 76(4), pp.581–592.

Driesen, N.R. et al., 2008. Impairment of working memory maintenance and response in schizophrenia: functional magnetic resonance imaging evidence. Biological psychiatry, 64(12), pp.1026–1034.

Duffy, L. et al., 2008. Behavioral profile of a heterozygous mutant mouse model for EGF-like domain neuregulin 1. Behavioral Neuroscience, 122(4), pp.748– 759.

Duncan, C.E. et al., 2010. Prefrontal GABA(A) receptor alpha-subunit expression in normal postnatal human development and schizophrenia. Journal of Psychi- atric Research, 44(10), pp.673–681.

Eastwood, S.L., Burnet, P.W. & Harrison, P.J., 2000. Expression of complexin I and II mRNAs and their regulation by antipsychotic drugs in the rat forebrain. Synapse (New York, N.Y.), 36(3), pp.167–177.

Eckstein, J.A. et al., 2008. Simultaneous profiling of multiple neurochemical pathways from a single cerebrospinal fluid sample using GC/MS/MS with elec- tron capture detection. Journal of Mass Spectrometry: JMS, 43(6), pp.782– 790.

Eilam, R. et al., 1998. Activity-dependent regulation of Neu differentiation factor/neuregulin expression in rat brain. Proceedings of the National Academy of Sciences of the United States of America, 95(4), pp.1888–1893.

127

Ellison-Wright, I. & Bullmore, E., 2009. Meta-analysis of diffusion tensor imaging studies in schizophrenia. Schizophrenia research, 108(1-3), pp.3–10.

Etou, K. et al., 1998. Ceruletide inhibits phencyclidine-induced dopamine and serotonin release in rat prefrontal cortex. Pharmacology Biochemistry and Behav- ior, 61(4), pp.427–434.

Falls, D.L. et al., 1990. Mr 42,000 ARIA: a protein that may regulate the accumula- tion of acetylcholine receptors at developing chick neuromuscular junctions. Cold Spring Harbor symposia on quantitative biology, 55, pp.397–406.

Fanselow, M.S. & Dong, H.-W., 2010. Are the dorsal and ventral hippocampus functionally distinct structures? Neuron, 65(1), pp.7–19.

Farde, L. et al., 1992. Positron emission tomographic analysis of central D1 and D2 dopamine receptor occupancy in patients treated with classical neuroleptics and clozapine. Relation to extrapyramidal side effects. Archives of general psychiatry, 49(7), pp.538–544.

Farzan, F. et al., 2010. Evidence for gamma inhibition deficits in the dorsolateral prefrontal cortex of patients with schizophrenia. Brain: a journal of neurology, 133(Pt 5), pp.1505–1514.

Fattorini, G. et al., 2009. Acute phencyclidine administration reduces extracellular glutamate levels and the expression of synaptophysin and SNAP-25 in rat frontal cortex. Schizophrenia research, 108(1-3), pp.288–289.

Fazzari, P. et al., 2010. Control of cortical GABA circuitry development by Nrg1 and ErbB4 signalling. Nature, 464(7293), pp.1376–1380.

Fejgin, K. et al., 2008. Nitric oxide signaling in the medial prefrontal cortex is involved in the biochemical and behavioral effects of phencyclidine. Neuropsy- chopharmacology, 33(8), pp.1874–83.

Fejgin, K. et al., 2007. The atypical antipsychotic, aripiprazole, blocks phencyclidine-induced disruption of prepulse inhibition in mice. Psychopharmacology, 191(2), pp.377–385.

Fell, M.J. et al., 2008. Evidence for the role of metabotropic glutamate (mGlu)2 not mGlu3 receptors in the preclinical antipsychotic pharmacology of the mGlu2/3 receptor agonist (-)-(1R,4S,5S,6S)-4-amino-2- sulfonylbicyclo[3.1.0]hexane-4,6-dicarboxylic acid (LY404039). Journal of Pharmacology and Experimental Therapeutics, 326(1), pp.209–217.

Fenster, C., Vullhorst, D. & Buonanno, A., 2012. Acute neuregulin-1 signaling influences AMPA receptor mediated responses in cultured cerebellar granule neurons. Brain Research Bulletin, 87(1), pp.21–29.

Fisahn, A. et al., 2009. Neuregulin-1 modulates hippocampal gamma oscillations: implications for schizophrenia. Cerebral Cortex (New York, N.Y.: 1991), 19(3), pp.612–618.

128

Fish, K.N., Sweet, R.A. & Lewis, D.A., 2011. Differential distribution of proteins regulating GABA synthesis and reuptake in axon boutons of subpopulations of cortical interneurons. Cerebral cortex (New York, N.Y.: 1991), 21(11), pp.2450–2460.

Fleck, D. et al., 2012. BACE1 dependent neuregulin processing: review. Current Alzheimer research, 9(2), pp.178–183.

Fleck, D. et al., 2013. Dual Cleavage of Neuregulin 1 Type III by BACE1 and AD- AM17 Liberates Its EGF-Like Domain and Allows Paracrine Signaling. The Journal of Neuroscience, 33(18), pp.7856–7869.

Fromer, M. et al., 2014. De novo mutations in schizophrenia implicate synaptic networks. Nature, 506(7487), pp.179–184.

Fumagalli, F. et al., 2008. Dynamic regulation of glutamatergic postsynaptic activity in rat prefrontal cortex by repeated administration of antipsychotic drugs. Mo- lecular pharmacology, 73(5), pp.1484–1490.

Fung, S.J. et al., 2010. Expression of Interneuron Markers in the Dorsolateral Pre- frontal Cortex of the Developing Human and in Schizophrenia. American Journal of Psychiatry, 167(12), pp.1479–1488.

Furuya, Y. et al., 1998. Differential effects of the strychnine-insensitive glycine site antagonist (+)-HA-966 on the hyperactivity and the disruption of prepulse inhibition induced by phencyclidine in rats. Brain Research, 781(1-2), pp.227– 235.

Gao, R. et al., 2010. A Phase II, randomized, double-blind, multicenter, based on standard therapy, placebo-controlled study of the efficacy and safety of recombinant human neuregulin-1 in patients with chronic heart failure. Journal of the American College of Cardiology, 55(18), pp.1907–1914.

Gao, X.M., Hashimoto, T. & Tamminga, C.A., 1998. Phencyclidine (PCP) and dizocilpine (MK801) exert time-dependent effects on the expression of immediate early genes in rat brain. Synapse (New York, N.Y.), 29(1), pp.14–28.

Garcia, R.A., Vasudevan, K. & Buonanno, A., 2000. The neuregulin receptor ErbB-4 interacts with PDZ-containing proteins at neuronal synapses. Proceedings of the National Academy of Sciences of the United States of America, 97(7), pp.3596–3601.

Gardner, M. et al., 2006. Extreme population differences across Neuregulin 1 gene, with implications for association studies. Molecular Psychiatry, 11(1), pp.66– 75.

Gauthier, M.-K. et al., 2013. Dysregulation of the neuregulin-1-ErbB network modulates endogenous oligodendrocyte differentiation and preservation after spinal cord injury. The European journal of neuroscience, 38(5), pp.2693–2715.

129

Geddes, A.E., Huang, X.-F. & Newell, K.A., 2011. Reciprocal signalling between NR2 subunits of the NMDA receptor and neuregulin1 and their role in schizophrenia. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 35(4), pp.896–904.

Georgieva, L. et al., 2008. Support for neuregulin 1 as a susceptibility gene for bipolar disorder and schizophrenia. Biological Psychiatry, 64(5), pp.419–427.

Gerecke, K.M. et al., 2001. ErbB transmembrane tyrosine kinase receptors are differentially expressed throughout the adult rat central nervous system. J Comp Neurol, 433(1), pp.86–100.

Gessa, G.L. et al., 2000. Dissociation of haloperidol, clozapine, and olanzapine effects on electrical activity of mesocortical dopamine neurons and dopamine release in the prefrontal cortex. Neuropsychopharmacology, 22(6), pp.642– 649.

Geyer, M.A. et al., 2001. Pharmacological studies of prepulse inhibition models of sensorimotor gating deficits in schizophrenia: a decade in review. Psycho- pharmacology (Berl), 156(2-3), pp.117–54.

Glantz, L.A. & Lewis, D.A., 1997. Reduction of synaptophysin immunoreactivity in the prefrontal cortex of subjects with schizophrenia. Regional and diagnostic specificity. Archives of general psychiatry, 54(7), pp.660–669.

Gleason, S.D. & Shannon, H.E., 1997. Blockade of phencyclidine-induced hyperlocomotion by olanzapine, clozapine and serotonin receptor subtype selective antagonists in mice. Psychopharmacology, 129(1), pp.79–84.

Godsil, B.P. et al., 2013. The hippocampal-prefrontal pathway: the weak link in psychiatric disorders? European neuropsychopharmacology: the journal of the European College of Neuropsychopharmacology, 23(10), pp.1165–1181.

Gong, Y.G. et al., 2009. A two-method meta-analysis of Neuregulin 1(NRG1) association and heterogeneity in schizophrenia. Schizophrenia Research, 111(1- 3), pp.109–114.

Gonzalez-Burgos, G., Fish, K.N. & Lewis, D.A., 2011. GABA neuron alterations, cortical circuit dysfunction and cognitive deficits in schizophrenia. Neural Plasticity, 2011, p.723184.

Goto, N. et al., 2009. Reduction of brain gamma-aminobutyric acid (GABA) concentrations in early-stage schizophrenia patients: 3T Proton MRS study. Schizo- phrenia research, 112(1-3), pp.192–193.

Goulding, M., 2009. Circuits controlling vertebrate locomotion: moving in a new direction. Nature reviews. Neuroscience, 10(7), pp.507–518.

Grobin, A.C. & Deutch, A.Y., 1998. Dopaminergic regulation of extracellular γ- aminobutyric acid levels in the prefrontal cortex of the rat. Journal of Phar- macology and Experimental Therapeutics, 285(1), pp.350–357.

130

Grunze, H.C. et al., 1996. NMDA-dependent modulation of CA1 local circuit inhibition. The Journal of neuroscience: the official journal of the Society for Neu- roscience, 16(6), pp.2034–2043.

Gu, Z. et al., 2005. Regulation of NMDA receptors by neuregulin signaling in prefrontal cortex. The Journal of Neuroscience: The Official Journal of the Soci- ety for Neuroscience, 25(20), pp.4974–4984.

Guidotti, A. et al., 2000. Decrease in reelin and glutamic acid decarboxylase67 (GAD67) expression in schizophrenia and bipolar disorder: a postmortem brain study. Archives of general psychiatry, 57(11), pp.1061–1069.

Guo, W.-P. et al., 2006. Neuroprotective effects of neuregulin-1 in rat models of fo- cal cerebral ischemia. Brain Research, 1087(1), pp.180–185.

Gurevich, E.V. et al., 1997. Mesolimbic dopamine D3 receptors and use of antipsychotics in patients with schizophrenia. A postmortem study. Archives of general psychiatry, 54(3), pp.225–232.

Guy, P.M. et al., 1994. Insect cell-expressed p180erbB3 possesses an impaired tyrosine kinase activity. Proceedings of the National Academy of Sciences of the United States of America, 91(17), pp.8132–8136.

Gyertyán, I. et al., 2011. Cariprazine (RGH-188), a potent D 3/D 2 dopamine receptor partial agonist, binds to dopamine D 3 receptors in vivo and shows antipsychotic-like and procognitive effects in rodents. Neurochemistry Interna- tional, 59(6), pp.925–935.

Hagino, Y. et al., 2010. Essential role of NMDA receptor channel ε4 subunit (GluN2D) in the effects of phencyclidine, but not methamphetamine. PloS One, 5(10), p.e13722.

Hahn, C.-G. et al., 2006. Altered neuregulin 1-erbB4 signaling contributes to NMDA receptor hypofunction in schizophrenia. Nature Medicine, 12(7), pp.824–828.

Halassa, M.M., Fellin, T. & Haydon, P.G., 2007. The tripartite synapse: roles for gli- otransmission in health and disease. Trends in molecular medicine, 13(2), pp.54–63.

Hall, J. et al., 2010. Hippocampal function in schizophrenia and bipolar disorder. Psychol Med, 40(5), pp.761–70.

Hamilton, N.B. & Attwell, D., 2010. Do astrocytes really exocytose neurotransmitters? Nature Reviews. Neuroscience, 11(4), pp.227–238.

Hanania, T., Hillman, G.R. & Johnson, K.M., 1999. Augmentation of locomotor activity by chronic phencyclidine is associated with an increase in striatal NMDA receptor function and an upregulation of the NR1 receptor subunit. Synapse (New York, N.Y.), 31(3), pp.229–239.

131

Hancock, M.L. et al., 2008. Presynaptic type III neuregulin1-ErbB signaling targets {alpha}7 nicotinic acetylcholine receptors to axons. The Journal of General Physiology, 181(3), pp.511–21.

Harrison, P.J. et al., 2012. Genetic mouse models relevant to schizophrenia: taking stock and looking forward. Neuropharmacology, 62(3), pp.1164–1167.

Harrison, P.J., 2004. The hippocampus in schizophrenia: a review of the neuropatho- logical evidence and its pathophysiological implications. Psychopharmacolo- gy, 174(1), pp.151–162.

Hashimoto, R. et al., 2004. Expression analysis of neuregulin-1 in the dorsolateral prefrontal cortex in schizophrenia. Molecular Psychiatry, 9(3), pp.299–307.

Hashimoto, T., Arion, D., et al., 2008. Alterations in GABA-related transcriptome in the dorsolateral prefrontal cortex of subjects with schizophrenia. Molecular psychiatry, 13(2), pp.147–161.

Hashimoto, T., Bazmi, H.H., et al., 2008. Conserved regional patterns of GABA- related transcript expression in the neocortex of subjects with schizophrenia. The American Journal of Psychiatry, 165(4), pp.479–489.

Hashimoto, T. et al., 2003. Gene expression deficits in a subclass of GABA neurons in the prefrontal cortex of subjects with schizophrenia. The Journal of neuroscience: the official journal of the Society for Neuroscience, 23(15), pp.6315– 6326.

Heckers, S. et al., 2002. Differential hippocampal expression of glutamic acid decarboxylase 65 and 67 messenger RNA in bipolar disorder and schizophrenia. Archives of general psychiatry, 59(6), pp.521–529.

Hegde, G.V. et al., 2013. Blocking NRG1 and other ligand-mediated Her4 signaling enhances the magnitude and duration of the chemotherapeutic response of non-small cell lung cancer. Science translational medicine, 5(171), p.171ra18.

Heidbreder, C.A. et al., 2001. Increased responsiveness of dopamine to atypical, but not typical antipsychotics in the medial prefrontal cortex of rats reared in iso- lation. Psychopharmacology, 156(2-3), pp.338–351.

Hilgenberg, L.G.W. & Smith, M.A., 2007. Preparation of dissociated mouse cortical neuron cultures. Journal of visualized experiments: JoVE, (10), p.562.

Honer, W.G. et al., 1999. Synaptic and plasticity-associated proteins in anterior frontal cortex in severe mental illness. Neuroscience, 91(4), pp.1247–1255.

Hons, J. et al., 2010. Glycine serum level in schizophrenia: relation to negative symptoms. Psychiatry research, 176(2-3), pp.103–108.

Hoover, W.B. & Vertes, R.P., 2007. Anatomical analysis of afferent projections to the medial prefrontal cortex in the rat. Brain structure & function, 212(2), pp.149–179. 132

Howes, O.D. et al., 2013. Midbrain dopamine function in schizophrenia and depression: a post-mortem and positron emission tomographic imaging study. Brain: a journal of neurology, 136(Pt 11), pp.3242–3251.

Howes, O.D. & Kapur, S., 2009. The dopamine hypothesis of schizophrenia: version III--the final common pathway. Schizophrenia Bulletin, 35(3), pp.549–562.

Howland, J.G. et al., 2004. Electrical stimulation of the hippocampus disrupts prepulse inhibition in rats: Frequency- and site-dependent effects. Behavioural Brain Research, 152(2), pp.187–197.

Huang, M. et al., 2006. Effects of divalproex and atypical antipsychotic drugs on dopamine and acetylcholine efflux in rat hippocampus and prefrontal cortex. Brain research, 1099(1), pp.44–55.

Huang, Y.Z. et al., 2000. Regulation of neuregulin signaling by PSD-95 interacting with ErbB4 at CNS synapses. Neuron, 26(2), pp.443–455.

Huertas-Vazquez, A. et al., 2013. A Common Missense Variant in the Neuregulin1 Gene is associated with Both Schizophrenia and Sudden Cardiac Death. Heart rhythm: the official journal of the Heart Rhythm Society.

Hustveit, O., Maurset, A. & Oye, I., 1995. Interaction of the chiral forms of ketamine with opioid, phencyclidine, sigma and muscarinic receptors. Pharmacology & toxicology, 77(6), pp.355–359.

Ichikawa, J. et al., 2002. Atypical, but not typical, antipsychotic drugs increase cortical acetylcholine release without an effect in the nucleus accumbens or striatum. Neuropsychopharmacology, 26(3), pp.325–339.

Inan, M., Petros, T.J. & Anderson, S.A., 2013. Losing your inhibition: linking cortical GABAergic interneurons to schizophrenia. Neurobiology of disease, 53, pp.36–48.

Isaacson, J.S. & Scanziani, M., 2011. How inhibition shapes cortical activity. Neu- ron, 72(2), pp.231–243.

Iwakura, Y. & Nawa, H., 2013. ErbB1-4-dependent EGF/neuregulin signals and their cross talk in the central nervous system: pathological implications in schizophrenia and Parkinson’s disease. Frontiers in cellular neuroscience, 7, p.4.

Jabbour, A. et al., 2011. Parenteral administration of recombinant human neuregulin- 1 to patients with stable chronic heart failure produces favourable acute and chronic haemodynamic responses. European Journal of Heart Failure, 13(1), pp.83–92.

Jacobsen, J.P.R. et al., 2005. Monoaminergic dysregulation in glutathione-deficient mice: Possible relevance to schizophrenia? Neuroscience, 132(4), pp.1055– 1072.

133

Javitt, D.C., 2009. Glycine transport inhibitors for the treatment of schizophrenia: symptom and disease modification. Current opinion in drug discovery & development, 12(4), pp.468–478.

Javitt, D.C. et al., 2012. Has an angel shown the way? Etiological and therapeutic implications of the PCP/NMDA model of schizophrenia. Schizophrenia bulletin, 38(5), pp.958–966.

Javitt, D.C. et al., 2008. Neurophysiological biomarkers for drug development in schizophrenia. Nature reviews. Drug discovery, 7(1), pp.68–83.

Javitt, D.C. et al., 1999. Reversal of phencyclidine-induced effects by glycine and glycine transport inhibitors. Biological Psychiatry, 45(6), pp.668–679.

Javitt, D.C. & Zukin, S.R., 1991. Recent advances in the phencyclidine model of schizophrenia. Am J Psychiatry, 148(10), pp.1301–8.

Jentsch, J.D. et al., 2008. Clonidine and guanfacine attenuate phencyclidine-induced dopamine overflow in rat prefrontal cortex: Mediating influence of the alpha- 2A adrenoceptor subtype. Brain Research, 1246(C), pp.41–46.

Jodo, E. et al., 2005. Activation of medial prefrontal cortex by phencyclidine is mediated via a hippocampo-prefrontal pathway. Cerebral cortex (New York, N.Y.: 1991), 15(5), pp.663–669.

Johansson, C. et al., 1995. Prepulse inhibition of acoustic startle, a measure of sensorimotor gating - effects of antipsychotics and other agents in rats. Pharma- cology Biochemistry and Behavior, 52(4), pp.649–654.

Johansson, C., Jackson, D.M. & Svensson, L., 1994. The atypical antipsychotic, remoxipride, blocks phencyclidine-induced disruption of prepulse inhibition in the rat. Psychopharmacology, 116(4), pp.437–442.

Johnson, D.E. et al., 2005. The role of muscarinic receptor antagonism in antipsychotic-induced hippocampal acetylcholine release. European journal of pharmacology, 506(3), pp.209–219.

Jones, C.A., Watson, D.J.G. & Fone, K.C.F., 2011. Animal models of schizophrenia. British journal of pharmacology, 164(4), pp.1162–1194.

Jordan, S. et al., 2004. In vivo effects of aripiprazole on cortical and striatal dopaminergic and serotonergic function. European Journal of Pharmacology, 483(1), pp.45–53.

Joshi, D., Fullerton, J.M. & Weickert, C.S., 2014. Elevated ErbB4 mRNA is related to interneuron deficit in prefrontal cortex in schizophrenia. Journal of Psy- chiatric Research, 53, pp.125–132.

Van Kammen, D.P. et al., 1982. CSF levels of gamma-aminobutyric acid in schizophrenia. Low values in recently ill patients. Archives of general psychiatry, 39(1), pp.91–97.

134

Kampman, O. et al., 2004. Neuregulin genotype and medication response in Finnish patients with schizophrenia. Neuroreport, 15(16), pp.2517–2520.

Kanahara, N. et al., 2008. Glycine and D-serine, but not d-cycloserine, attenuate prepulse inhibition deficits induced by NMDA receptor antagonist MK-801. Psychopharmacology, 198(3), pp.363–374.

Kang, J.I. et al., 2013. Reduced Binding Potential of GABA-A/Benzodiazepine Re- ceptors in Individuals at Ultra-high Risk for Psychosis: An [18F]- Fluoroflumazenil Positron Emission Tomography Study. Schizophrenia bulletin.

Kantrowitz, J.T. & Javitt, D.C., 2010. N-methyl-d-aspartate (NMDA) receptor dysfunction or dysregulation: the final common pathway on the road to schizophrenia? Brain research bulletin, 83(3), pp.108–121.

Kapur, S. & Seeman, P., 2002. NMDA receptor antagonists ketamine and PCP have direct effects on the dopamine D(2) and serotonin 5-HT(2)receptors- implications for models of schizophrenia. Molecular psychiatry, 7(8), pp.837–844.

Karasawa, J. et al., 2002. MS-377, a selective sigma receptor ligand, indirectly blocks the action of PCP in the N-methyl-D-aspartate receptor ion-channel complex in primary cultured rat neuronal cells. Life Sciences, 70(14), pp.1631–1642.

Karl, T. et al., 2007. Altered motor activity, exploration and anxiety in heterozygous neuregulin 1 mutant mice: implications for understanding schizophrenia. Genes, Brain, and Behavior, 6(7), pp.677–687.

Kastin, A.J., Akerstrom, V. & Pan, W., 2004. Neuregulin-1-beta1 enters brain and spinal cord by receptor-mediated transport. Journal of Neurochemistry, 88(4), pp.965–970.

Kato, T. et al., 2010. Phenotypic characterization of transgenic mice overexpressing neuregulin-1. PloS One, 5(12), p.e14185.

Kato, T. et al., 2011. Transient exposure of neonatal mice to neuregulin-1 results in hyperdopaminergic states in adulthood: implication in neurodevelopmental hypothesis for schizophrenia. Molecular Psychiatry, 16(3), pp.307–320.

Katsel, P. et al., 2011. Astrocyte and glutamate markers in the superficial, deep, and white matter layers of the anterior cingulate gyrus in schizophrenia. Neuro- psychopharmacology: official publication of the American College of Neuro- psychopharmacology, 36(6), pp.1171–1177.

Keefe, R.S.E. et al., 2004. Comparative effect of atypical and conventional antipsychotic drugs on neurocognition in first-episode psychosis: a randomized, double-blind trial of olanzapine versus low doses of haloperidol. The Ameri- can journal of psychiatry, 161(6), pp.985–995.

135

Kegeles, L.S. et al., 2012. Elevated prefrontal cortex γ-aminobutyric acid and glutamate-glutamine levels in schizophrenia measured in vivo with proton magnetic resonance spectroscopy. Archives of general psychiatry, 69(5), pp.449– 459.

Keith, V.A., Mansbach, R.S. & Geyer, M.A., 1991. Failure of haloperidol to block the effects of phencyclidine and dizocilpine on prepulse inhibition of startle. Biological Psychiatry, 30(6), pp.557–566.

Kendler, K.S. & Gardner, C.O., 1997. The risk for psychiatric disorders in relatives of schizophrenic and control probands: a comparison of three independent studies. Psychological medicine, 27(2), pp.411–419.

Keshavan, M.S. et al., 2008. Schizophrenia, “just the facts”: what we know in 2008 Part 3: neurobiology. Schizophrenia Research, 106(2-3), pp.89–107.

Kim, E. et al., 2013. The relationship between antipsychotic D2 occupancy and change in frontal metabolism and working memory : A dual [(11)C]raclopride and [(18) F]FDG imaging study with aripiprazole. Psycho- pharmacology, 227(2), pp.221–229.

Kinon, B.J. et al., 2004. Dose response and atypical antipsychotics in schizophrenia. CNS drugs, 18(9), pp.597–616.

Kirihara, K. et al., 2012. Hierarchical organization of gamma and theta oscillatory dynamics in schizophrenia. Biological psychiatry, 71(10), pp.873–880.

Kitaichi, K. et al., 1994. Effects of risperidone on phencyclidine-induced behaviors - comparison with haloperidol and ritanserin. Japanese Journal of Pharmacol- ogy, 66(2), pp.181–189.

Klamer, D., Engel, J.A. & Svensson, L., 2004. Phencyclidine-induced behaviour in mice prevented by methylene blue. Basic and Clinical Pharmacology and Toxicology, 94(2), pp.65–72.

Klausberger, T. & Somogyi, P., 2008. Neuronal diversity and temporal dynamics: the unity of hippocampal circuit operations. Science (New York, N.Y.), 321(5885), pp.53–57.

Knable, M.B. et al., 2004. Molecular abnormalities of the hippocampus in severe psychiatric illness: postmortem findings from the Stanley Neuropathology Consortium. Molecular Psychiatry, 9(6), pp.609–620.

Koch, S., Perry, K.W. & Bymaster, F.P., 2004. Brain region and dose effects of an olanzapine/fluoxetine combination on extracellular monoamine concentrations in the rat. Neuropharmacology, 46(2), pp.232–242.

Konradi, C. et al., 2011. Hippocampal interneurons are abnormal in schizophrenia. Schizophrenia research, 131(1-3), pp.165–173.

Koseki, T. et al., 2012. Exposure to enriched environments during adolescence prevents abnormal behaviours associated with histone deacetylation in phency- 136

clidine-treated mice. The international journal of neuropsychopharmacology / official scientific journal of the Collegium Internationale Neuropsycho- pharmacologicum (CINP), 15(10), pp.1489–1501.

Krivosheya, D. et al., 2008. ErbB4-neuregulin signaling modulates synapse development and dendritic arborization through distinct mechanisms. The Journal of Biological Chemistry, 283(47), pp.32944–32956.

Kumra, S. et al., 2004. Reduced frontal white matter integrity in early-onset schizophrenia: a preliminary study. Biological psychiatry, 55(12), pp.1138–1145.

Kuroki, T. et al., 1999. Effects of the serotonin(2A/2C) receptor agonist and antagonist on phencyclidine-induced dopamine release in rat medial prefrontal cortex. Progress in Neuro-Psychopharmacology and Biological Psychiatry, 23(7), pp.1259–1275.

Kwon, O.B. et al., 2008. Neuregulin-1 regulates LTP at CA1 hippocampal synapses through activation of dopamine D4 receptors. Proceedings of the National Academy of Sciences of the United States of America, 105(40), pp.15587– 15592.

Kwon, O.B. et al., 2005. Neuregulin-1 reverses long-term potentiation at CA1 hippocampal synapses. J Neurosci, 25(41), pp.9378–83.

Labrie, V. & Roder, J.C., 2010. The involvement of the NMDA receptor D- serine/glycine site in the pathophysiology and treatment of schizophrenia. Neuroscience and biobehavioral reviews, 34(3), pp.351–372.

Lahti, A.C. et al., 2005. Functional effects of single dose first- and second-generation antipsychotic administration in subjects with schizophrenia. Psychiatry research, 139(1), pp.19–30.

Lalo, U., 2006. NMDA Receptors Mediate Neuron-to-Glia Signaling in Mouse Cor- tical Astrocytes. Journal of Neuroscience, 26(10), pp.2673–2683.

Landén, M. et al., 2002. Reduction of the synaptophysin level but normal levels of glycerophospholipids in the gyrus cinguli in schizophrenia. Schizophrenia research, 55(1-2), pp.83–88.

Law, A.J. et al., 2007. Disease-associated intronic variants in the ErbB4 gene are related to altered ErbB4 splice-variant expression in the brain in schizophrenia. Hum Mol Genet, 16(2), pp.129–41.

Law, A.J. et al., 2006. Neuregulin 1 transcripts are differentially expressed in schizophrenia and regulated by 5’ SNPs associated with the disease. Proceedings of the National Academy of Sciences of the United States of America, 103(17), pp.6747–6752.

Law, A.J. et al., 2004. Neuregulin-1 (NRG-1) mRNA and protein in the adult human brain. Neuroscience, 127(1), pp.125–36.

137

Lawrence, J.J., 2008. Cholinergic control of GABA release: emerging parallels between neocortex and hippocampus. Trends in neurosciences, 31(7), pp.317– 327.

Lee, E.-A. et al., 2009. Presynaptic glycine receptors facilitate spontaneous glutamate release onto hilar neurons in the rat hippocampus. Journal of Neuro- chemistry, 109(1), pp.275–286.

Lee, K.-H. et al., 2003. Synchronous gamma activity: a review and contribution to an integrative neuroscience model of schizophrenia. Brain research. Brain research reviews, 41(1), pp.57–78.

Lee, M., McGeer, E.G. & McGeer, P.L., 2011. Mechanisms of GABA release from human astrocytes. Glia, 59(11), pp.1600–1611.

Lee, M., Schwab, C. & McGeer, P.L., 2011. Astrocytes are GABAergic cells that modulate microglial activity. Glia, 59(1), pp.152–165.

Leng, A. et al., 2003. Effect of the 5-HT6 receptor antagonists Ro 04-6790 and Ro 65-7199 on latent inhibition and prepulse inhibition in the rat: Comparison to clozapine. Pharmacology Biochemistry and Behavior, 75(2), pp.281–288.

De Leon, J. & Diaz, F.J., 2005. A meta-analysis of worldwide studies demonstrates an association between schizophrenia and tobacco smoking behaviors. Schiz- ophrenia research, 76(2-3), pp.135–157.

Leucht, S. et al., 2013. Comparative efficacy and tolerability of 15 antipsychotic drugs in schizophrenia: a multiple-treatments meta-analysis. Lancet, 382(9896), pp.951–962.

Leucht, S. et al., 2009. Second-generation versus first-generation antipsychotic drugs for schizophrenia: a meta-analysis. Lancet, 373(9657), pp.31–41.

Lewis, D.A., Hashimoto, T. & Volk, D.W., 2005. Cortical inhibitory neurons and schizophrenia. Nature Reviews. Neuroscience, 6(4), pp.312–324.

Lewis, D.A. & Moghaddam, B., 2006. Cognitive dysfunction in schizophrenia: con- vergence of gamma-aminobutyric acid and glutamate alterations. Archives of neurology, 63(10), pp.1372–1376.

Lewis, D.A., Volk, D.W. & Hashimoto, T., 2004. Selective alterations in prefrontal cortical GABA neurotransmission in schizophrenia: a novel target for the treatment of working memory dysfunction. Psychopharmacology (Berl), 174(1), pp.143–50.

Li, B. et al., 2007. The neuregulin-1 receptor erbB4 controls glutamatergic synapse maturation and plasticity. Neuron, 54(4), pp.583–597.

Li, Y. et al., 2012. Neuregulin-1 is neuroprotective in a rat model of organophos- phate-induced delayed neuronal injury. Toxicology and applied pharmacology, 262(2), pp.194–204.

138

Li, Z. et al., 2004. Aripiprazole, a novel antipsychotic drug, preferentially increases dopamine release in the prefrontal cortex and hippocampus in rat brain. Eu- ropean journal of pharmacology, 493(1-3), pp.75–83.

Li, Z. et al., 2010. The novel neurotensin analog NT69L blocks phencyclidine (PCP)- induced increases in locomotor activity and PCP-induced increases in monoamine and amino acids levels in the medial prefrontal cortex. Brain Research, 1311, pp.28–36.

Lieberman, J.A. et al., 2008. Antipsychotic drugs: comparison in animal models of efficacy, neurotransmitter regulation, and neuroprotection. Pharmacological reviews, 60(3), pp.358–403.

Lillrank, S.M. et al., 1994. In vivo effects of local and systemic phencyclidine on the extracellular levels of catecholamines and transmitter amino acids in the dorsolateral striatum of anaesthetized rats. Acta physiologica Scandinavica, 150(2), pp.109–115.

Lindahl, J.S. & Keifer, J., 2004. Glutamate receptor subunits are altered in forebrain and cerebellum in rats chronically exposed to the NMDA receptor antagonist phencyclidine. Neuropsychopharmacology: official publication of the Ameri- can College of Neuropsychopharmacology, 29(11), pp.2065–2073.

Linn, G.S. et al., 2003. Reversal of phencyclidine-induced prepulse inhibition deficits by clozapine in monkeys. Psychopharmacology, 169(3-4), pp.234–239.

Linscott, R.J. & van Os, J., 2013. An updated and conservative systematic review and meta-analysis of epidemiological evidence on psychotic experiences in children and adults: on the pathway from proneness to persistence to dimen- sional expression across mental disorders. Psychological Medicine, 43(6), pp.1133–1149.

Lipska, B.K. et al., 2006. Expression of DISC1 binding partners is reduced in schizophrenia and associated with DISC1 SNPs. Human molecular genetics, 15(8), pp.1245–1258.

Liu, J. & Kern, J.A., 2002. Neuregulin-1 activates the JAK-STAT pathway and regulates lung epithelial cell proliferation. American Journal of Respiratory Cell and Molecular Biology, 27(3), pp.306–313.

Liu, X. et al., 2011. Specific Regulation of NRG1 Isoform Expression by Neuronal Activity. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 31(23), pp.8491–8501.

Liu, Y. et al., 2001. Neuregulins increase alpha7 nicotinic acetylcholine receptors and enhance excitatory synaptic transmission in GABAergic interneurons of the hippocampus. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 21(15), pp.5660–5669.

Liu, Z. et al., 2011. Neuregulin-1β prevents Ca(2+) overloading and apoptosis through PI3K/Akt activation in cultured dorsal root ganglion neurons with

139

excitotoxicity induced by glutamate. Cellular and molecular neurobiology, 31(8), pp.1195–1201.

Loeb, J.A. et al., 1999. Expression patterns of transmembrane and released forms of neuregulin during spinal cord and neuromuscular synapse development. De- velopment (Cambridge, England), 126(4), pp.781–791.

Loh, H.C. et al., 2013. Neuregulin-1 (NRG-1) and its susceptibility to schizophrenia: a case-control study and meta-analysis. Psychiatry research, 208(2), pp.186– 188.

Longart, M. et al., 2007. Regulation of ErbB-4 endocytosis by neuregulin in GA- BAergic hippocampal interneurons. Brain Research Bulletin, 73(4-6), pp.210–219.

López-Figueroa, A.L. et al., 2004. Serotonin 5-HT1A, 5-HT1B, and 5-HT2A receptor mRNA expression in subjects with major depression, bipolar disorder, and schizophrenia. Biological psychiatry, 55(3), pp.225–233.

López-Gil, X. et al., 2007. Clozapine and Haloperidol Differently Suppress the MK- 801-Increased Glutamatergic and Serotonergic Transmission in the Medial Prefrontal Cortex of the Rat. Neuropsychopharmacology, 32(10), pp.2087– 2097.

Luby, E.D. et al., 1959. Study of a new schizophrenomimetic drug; sernyl. A.M.A. archives of neurology and psychiatry, 81(3), pp.363–369.

Luo, X. et al., 2011. Cleavage of neuregulin-1 by BACE1 or ADAM10 protein produces differential effects on myelination. The Journal of biological chemistry, 286(27), pp.23967–23974.

Luo, X. et al., 2013. Reversible Overexpression of Bace1-Cleaved Neuregulin-1 N- Terminal Fragment Induces Schizophrenia-Like Phenotypes in Mice. Biolog- ical psychiatry.

Maayani, S. & Weinstein, H., 1979. Some structure activity relationships of phencyclidine derivatives as anticholinergic agents in vitro and in vivo. Progress in clinical and biological research, 27, pp.91–106.

Mackeprang, T., Kristiansen, K.T. & Glenthoj, B.Y., 2002. Effects of antipsychotics on prepulse inhibition of the startle response in drug-naive schizophrenic patients. Biological Psychiatry, 52(9), pp.863–873.

Maeda, J. et al., 2003. Different roles of group I and group II metabotropic glutamate receptors on phencyclidine-induced dopamine release in the rat prefrontal cortex. Neuroscience Letters, 336(3), pp.171–174.

Mahar, I. et al., 2011. Subchronic peripheral neuregulin-1 increases ventral hippocampal neurogenesis and induces antidepressant-like effects. PloS One, 6(10), p.e26610.

140

Marín, O., 2012. Interneuron dysfunction in psychiatric disorders. Nature Reviews. Neuroscience, 13(2), pp.107–120.

Markram, H. et al., 2004. Interneurons of the neocortical inhibitory system. Nature Reviews Neuroscience, 5(10), pp.793–807.

Marquis, K.L. et al., 2007. WAY-163909 [(7bR,10aR)-1,2,3,4,8,9,10,10a-octahydro- 7bH-cyclopenta-[ b][1,4]diazepino[6,7,1hi]indole]: A novel 5- hydroxytryptamine 2C receptor-selective agonist with preclinical antipsychotic-like activity. Journal of Pharmacology and Experimental Therapeu- tics, 320(1), pp.486–496.

Martin, P., Carlsson, M.L. & Hjorth, S., 1998. Systemic PCP treatment elevates brain extracellular 5-HT: A microdialysis study in awake rats. NeuroReport, 9(13), pp.2985–2988.

McCullough, L.D. & Salamone, J.D., 1992. Increases in extracellular dopamine levels and locomotor activity after direct infusion of phencyclidine into the nucleus accumbens. Brain research, 577(1), pp.1–9.

McGrath, J.J. et al., 2010. Developmental vitamin D deficiency and risk of schizophrenia: a 10-year update. Schizophrenia Bulletin, 36(6), pp.1073–1078.

Mcguire, P. et al., 2008. Functional neuroimaging in schizophrenia: diagnosis and drug discovery. Trends in Pharmacological Sciences, 29(2), pp.91–98.

McKibben, C.E. et al., 2010. Effect of pretreatment with risperidone on phencyclidine-induced disruptions in object recognition memory and prefrontal cortex parvalbumin immunoreactivity in the rat. Behavioural brain research, 208(1), pp.132–136.

Mei, L. & Xiong, W.C., 2008. Neuregulin 1 in neural development, synaptic plasticity and schizophrenia. Nat Rev Neurosci, 9(6), pp.437–52.

Mellios, N. et al., 2009. Molecular determinants of dysregulated GABAergic gene expression in the prefrontal cortex of subjects with schizophrenia. Biological Psychiatry, 65(12), pp.1006–1014.

Meltzer, H.Y. et al., 2003. Serotonin receptors: their key role in drugs to treat schizophrenia. Progress in neuro-psychopharmacology & biological psychiatry, 27(7), pp.1159–1172.

Meltzer, H.Y., 1992. Treatment of the neuroleptic-nonresponsive schizophrenic patient. Schizophrenia bulletin, 18(3), pp.515–542.

Le Meur, K. et al., 2012. GABA release by hippocampal astrocytes. Frontiers in computational neuroscience, 6, p.59.

Meyer, D. & Birchmeier, C., 1994. Distinct isoforms of neuregulin are expressed in mesenchymal and neuronal cells during mouse development. Proceedings of the National Academy of Sciences of the United States of America, 91(3), pp.1064–1068. 141

Meyer, U. & Feldon, J., 2010. Epidemiology-driven neurodevelopmental animal models of schizophrenia. Prog Neurobiol, 90(3), pp.285–326.

Meyer, U., Feldon, J. & Yee, B.K., 2009. A review of the fetal brain cytokine imbal- ance hypothesis of schizophrenia. Schizophrenia Bulletin, 35(5), pp.959–972.

Meyer-Lindenberg, A., 2010. From maps to mechanisms through neuroimaging of schizophrenia. Nature, 468(7321), pp.194–202.

Millan, M.J. et al., 2012. Cognitive dysfunction in psychiatric disorders: characteristics, causes and the quest for improved therapy. Nature reviews. Drug discovery, 11(2), pp.141–168.

Millan, M.J. et al., 1999. Contrasting mechanisms of action and sensitivity to antipsychotics of phencyclidine versus amphetamine: Importance of nucleus accumbens 5-HT2A sites for PCP-induced locomotion in the rat. European Journal of Neuroscience, 11(12), pp.4419–4432.

Min, S.S. et al., 2011. Neuregulin-1 prevents amyloid β-induced impairment of long- term potentiation in hippocampal slices via ErbB4. Neuroscience Letters, 505(1), pp.6–9.

Mitchell, R.M. et al., 2013. ErbB4 reduces synaptic GABAA currents independent of its receptor tyrosine kinase activity. Proceedings of the National Academy of Sciences of the United States of America, 110(48), pp.19603–19608.

Miyamoto, S. et al., 2012. Pharmacological treatment of schizophrenia: a critical review of the pharmacology and clinical effects of current and future therapeutic agents. Molecular psychiatry, 17(12), pp.1206–1227.

Mladinov, M. et al., 2010. Astrocyte Expression of D2-Like Dopamine Receptors in the Prefrontal Cortex. Translational Neuroscience, 1(3), pp.238–243.

Moghaddam, B. & Javitt, D., 2012. From revolution to evolution: the glutamate hypothesis of schizophrenia and its implication for treatment. Neuropsycho- pharmacology: official publication of the American College of Neuropsycho- pharmacology, 37(1), pp.4–15.

Moghaddam, B. & Krystal, J.H., 2012. Capturing the angel in angel dust: Twenty years of translational neuroscience studies of NMDA receptor antagonists in animals and humans. Schizophrenia Bulletin, 38(5), pp.942–949.

Möhler, H., Fritschy, J.M. & Rudolph, U., 2002. A new benzodiazepine pharmacology. The Journal of pharmacology and experimental therapeutics, 300(1), pp.2–8.

Montero, J.C. et al., 2008. Neuregulins and cancer. Clinical cancer research: an official journal of the American Association for Cancer Research, 14(11), pp.3237–3241.

142

Moon, E. et al., 2011. Lack of association to a NRG1 missense polymorphism in schizophrenia or bipolar disorder in a Costa Rican population. Schizophrenia Research, 131(1-3), pp.52–57.

Mouri, A. et al., 2013. Animal models of schizophrenia for molecular and pharmacological intervention and potential candidate molecules. Neurobiology of Dis- ease, 53, pp.61–74.

Mouri, A. et al., 2011. Mouse strain differences in phencyclidine-induced behavioural changes. The International Journal of Neuropsychopharmacology / Official Scientific Journal of the Collegium Internationale Neuropsychopharmacolog- icum (CINP), 15(6), pp.767–79.

Mouri, A. et al., 2007. Phencyclidine animal models of schizophrenia: approaches from abnormality of glutamatergic neurotransmission and neurodevelopment. Neurochemistry international, 51(2-4), pp.173–184.

Munafò, M.R. et al., 2006. Association of the NRG1 gene and schizophrenia: a meta- analysis. Molecular Psychiatry, 11(6), pp.539–546.

Munafò, M.R., Attwood, A.S. & Flint, J., 2008. Neuregulin 1 genotype and schizophrenia. Schizophr Bull, 34(1), pp.9–12.

Murray, J.B., 2002. Phencyclidine (PCP): a dangerous drug, but useful in schizophrenia research. J Psychol, 136(3), pp.319–27.

Mutlu, O. et al., 2011. Effects of olanzapine, sertindole and clozapine on MK-801 induced visual memory deficits in mice. Pharmacology, biochemistry, and behavior, 99(4), pp.557–565.

Nagai, T. et al., 2003. Effect of AD-5423 on animal models of schizophrenia: phencyclidine-induced behavioral changes in mice. Neuroreport, 14(2), pp.269– 272.

Nagano, T. et al., 2007. In vivo administration of epidermal growth factor and its homologue attenuates developmental maturation of functional excitatory synapses in cortical GABAergic neurons. The European journal of neuroscience, 25(2), pp.380–390.

Nakata, K. et al., 2009. DISC1 splice variants are upregulated in schizophrenia and associated with risk polymorphisms. Proceedings of the National Academy of Sciences of the United States of America, 106(37), pp.15873–15878.

Neddens, J. et al., 2011. Conserved interneuron-specific ErbB4 expression in frontal cortex of rodents, monkeys, and humans: implications for schizophrenia. Bio- logical psychiatry, 70(7), pp.636–645.

Neddens, J. & Buonanno, A., 2010. Selective populations of hippocampal interneurons express ErbB4 and their number and distribution is altered in ErbB4 knockout mice. Hippocampus, 20(6), pp.724–744.

143

Ngan, E.T. et al., 2000. Decreased serotonin 2A receptor densities in neuroleptic- naive patients with schizophrenia: A PET study using [(18)F]setoperone. The American journal of psychiatry, 157(6), pp.1016–1018.

Nicodemus, K.K. et al., 2009. A 5’ promoter region SNP in NRG1 is associated with schizophrenia risk and type III isoform expression. Molecular Psychiatry, 14(8), pp.741–743.

Nicodemus, K.K. et al., 2006. Further evidence for association between ErbB4 and schizophrenia and influence on cognitive intermediate phenotypes in healthy controls. Molecular Psychiatry, 11(12), pp.1062–1065.

Nordquist, R.E. et al., 2008. Effects of aripiprazole/OPC-14597 on motor activity, pharmacological models of psychosis, and brain activity in rats. Neurophar- macology, 54(2), pp.405–416.

O’Connor, J.A. et al., 2006. Region specific regulation of NR1 in rhesus monkeys following chronic antipsychotic drug administration. Biological psychiatry, 60(6), pp.659–662.

Ohoyama, K. et al., 2011. Effect of novel atypical antipsychotic, blonanserin, on extracellular neurotransmitter level in rat prefrontal cortex. European Journal of Pharmacology, 653(1-3), pp.47–57.

Okada, M. & Corfas, G., 2004. Neuregulin1 downregulates postsynaptic GABAA receptors at the hippocampal inhibitory synapse. Hippocampus, 14(3), pp.337–344.

Okuyama, S. et al., 1999. A selective dopamine D-4 receptor antagonist, NRA0160: A preclinical neuropharmacological profile. Life Sciences, 65(20), pp.2109– 2125.

Olbrich, H.M. et al., 2008. Frontolimbic glutamate alterations in first episode schizophrenia: evidence from a magnetic resonance spectroscopy study. World J Biol Psychiatry, 9(1), pp.59–63.

Ozaki, M. et al., 1997. Neuregulin-beta induces expression of an NMDA-receptor subunit. Nature, 390(6661), pp.691–694.

Ozdemir, H. et al., 2012. Differential effects of antipsychotics on hippocampal pre- synaptic protein expressions and recognition memory in a schizophrenia model in mice. Progress in neuro-psychopharmacology & biological psychiatry, 39(1), pp.62–68.

Pamenter, M.E. et al., 2013. DIDS (4,4-Diisothiocyanatostilbenedisulphonic Acid) Induces Apoptotic Cell Death in a Hippocampal Neuronal Cell Line and Is Not Neuroprotective against Ischemic Stress. PloS one, 8(4), p.e60804.

Pande, M. et al., 1999. Inhibition of calcium ATPase by phencyclidine in rat brain. Molecular and cellular biochemistry, 194(1-2), pp.173–177.

144

Pankonin, M.S. et al., 2009. Differential distribution of neuregulin in human brain and spinal fluid. Brain Research, 1258, pp.1–11.

Park, S.W. et al., 2013. Effects of antipsychotic drugs on the expression of synaptic proteins and dendritic outgrowth in hippocampal neuronal cultures. Synapse (New York, N.Y.), 67(5), pp.224–234.

Parlapani, E. et al., 2010. Gene expression of neuregulin-1 isoforms in different brain regions of elderly schizophrenia patients. The World Journal of Biological Psychiatry: The Official Journal of the World Federation of Societies of Bio- logical Psychiatry, 11(2 Pt 2), pp.243–250.

Patterson, P.H., 2009. Immune involvement in schizophrenia and autism: etiology, pathology and animal models. Behavioural Brain Research, 204(2), pp.313– 321.

Paxinos, G. & Franklin, K., 2001. The Mouse Brain in Stereotaxic Coordinates, San Diego: Academic Press.

Paz, R.D. et al., 2008. Glutamatergic dysfunction in schizophrenia: from basic neuroscience to clinical psychopharmacology. European neuropsychopharmacology: the journal of the European College of Neuropsychopharmacology, 18(11), pp.773–786.

Peat, M.A. & Gibb, J.W., 1983. The effects of phencyclidine on glutamic acid decarboxylase activity in several regions of the rat brain. Neuroscience Letters, 35(3), pp.301–306.

Van Peer, G., Mestdagh, P. & Vandesompele, J., 2012. Accurate RT-qPCR gene expression analysis on cell culture lysates. Scientific reports, 2, p.222.

Perry, W. et al., 2009. A reverse-translational study of dysfunctional exploration in psychiatric disorders: from mice to men. Archives of General Psychiatry, 66(10), pp.1072–1080.

Perry, W. et al., 2010. Quantifying over-activity in bipolar and schizophrenia patients in a human open field paradigm. Psychiatry Research, 178(1), pp.84–91.

Peselmann, N. et al., 2013. Aripiprazole differentially regulates the expression of Gad67 and γ-aminobutyric acid transporters in rat brain. European archives of psychiatry and clinical neuroscience, 263(4), pp.285–297.

Petryshen, T.L. et al., 2005. Support for involvement of neuregulin 1 in schizophrenia pathophysiology. Molecular Psychiatry, 10(4), pp.366–374, 328.

Pettit, D.L. & Augustine, G.J., 2000. Distribution of functional glutamate and GABA receptors on hippocampal pyramidal cells and interneurons. Journal of neurophysiology, 84(1), pp.28–38.

Pilowsky, L.S. et al., 2005. First in vivo evidence of an NMDA receptor deficit in medication-free schizophrenic patients. Molecular Psychiatry, 11(2), pp.118– 119. 145

Pitcher, G.M. et al., 2011. Schizophrenia susceptibility pathway neuregulin 1-ErbB4 suppresses Src upregulation of NMDA receptors. Nature Medicine, 17(4), pp.470–478.

Poels, E.M.P., Kegeles, L.S., Kantrowitz, J.T., Javitt, D.C., et al., 2014. Glutama- tergic abnormalities in schizophrenia: A review of proton MRS findings. Schizophrenia research, 152(2-3), pp.325–332.

Poels, E.M.P., Kegeles, L.S., Kantrowitz, J.T., Slifstein, M., et al., 2014. Imaging glutamate in schizophrenia: review of findings and implications for drug discovery. Molecular psychiatry, 19(1), pp.20–29.

Pratt, J. et al., 2012. Advancing schizophrenia drug discovery: optimizing rodent models to bridge the translational gap. Nature reviews. Drug discovery, 11(7), pp.560–579.

Purcell, S.M. et al., 2014. A polygenic burden of rare disruptive mutations in schizophrenia. Nature, 506(7487), pp.185–190.

Raabe, T.D. et al., 2004. Localization of neuregulin isoforms and erbB receptors in myelinating glial cells. Glia, 45(2), pp.197–207.

Ramaekers, J.G. et al., 1999. Psychomotor, Cognitive, extrapyramidal, and affective functions of healthy volunteers during treatment with an atypical (amisul- pride) and a classic (haloperidol) antipsychotic. Journal of clinical psychopharmacology, 19(3), pp.209–221.

Rao, J.S. et al., 2013. Increased neuroinflammatory and arachidonic acid cascade markers, and reduced synaptic proteins, in the postmortem frontal cortex from schizophrenia patients. Schizophrenia research, 147(1), pp.24–31.

Réthelyi, J.M., Benkovits, J. & Bitter, I., 2013. Genes and environments in schizophrenia: The different pieces of a manifold puzzle. Neuroscience and Biobe- havioral Reviews, 37(10 Pt 1), pp.2424–2437.

Reynolds, G.P. et al., 2004. Calcium binding protein markers of GABA deficits in schizophrenia--postmortem studies and animal models. Neurotoxicity research, 6(1), pp.57–61.

Rico, B. & Marín, O., 2011. Neuregulin signaling, cortical circuitry development and schizophrenia. Current Opinion in Genetics & Development, 21(3), pp.262– 270.

Rieff, H.I. et al., 1999. Neuregulin induces GABA(A) receptor subunit expression and neurite outgrowth in cerebellar granule cells. The Journal of neuroscience: the official journal of the Society for Neuroscience, 19(24), pp.10757– 10766.

Ripke, S. et al., 2013. Genome-wide association analysis identifies 13 new risk loci for schizophrenia. Nature genetics, 45(10), pp.1150–1159.

146

Role, L.W. & Talmage, D.A., 2007. Neurobiology: new order for thought disorders. Nature, 448(7151), pp.263–265.

Ronchi, G. et al., 2013. ErbB2 receptor over-expression improves post-traumatic peripheral nerve regeneration in adult mice. PloS one, 8(2), p.e56282.

Rosen, A.S. & Andrew, R.D., 1990. Osmotic effects upon excitability in rat neocortical slices. Neuroscience, 38(3), pp.579–590.

Rösler, T.W. et al., 2011. Biodistribution and brain permeability of the extracellular domain of neuregulin-1-β1. Neuropharmacology, 61, pp.1413–1418.

Rotaru, D.C. et al., 2011. Glutamate Receptor Subtypes Mediating Synaptic Activa- tion of Prefrontal Cortex Neurons: Relevance for Schizophrenia. Journal of Neuroscience, 31(1), pp.142–156.

Roth, F.C. & Draguhn, A., 2012. GABA metabolism and transport: effects on synaptic efficacy. Neural plasticity, 2012, p.805830.

Rudolph, U. & Knoflach, F., 2011. Beyond classical benzodiazepines: novel therapeutic potential of GABAA receptor subtypes. Nature reviews. Drug discovery, 10(9), pp.685–697.

Ryu, J. et al., 2012. Neuregulin-1 exerts protective effects against neurotoxicities induced by C-terminal fragments of APP via ErbB4 receptor. Journal of pharmacological sciences, 119(1), pp.73–81.

Saha, S. et al., 2005. A systematic review of the prevalence of schizophrenia. PLoS medicine, 2(5), p.e141.

Saha, S., Chant, D. & McGrath, J., 2007. A systematic review of mortality in schizophrenia: is the differential mortality gap worsening over time? Archives of general psychiatry, 64(10), pp.1123–1131.

Sakai, T. et al., 2008. Changes in density of calcium-binding-protein- immunoreactive GABAergic neurons in prefrontal cortex in schizophrenia and bipolar disorder. Neuropathology: official journal of the Japanese Socie- ty of Neuropathology, 28(2), pp.143–150.

Sarter, M. & Tricklebank, M., 2012. Revitalizing psychiatric drug discovery. Nature reviews. Drug discovery, 11(6), pp.423–424.

Sawamura, N. et al., 2005. A form of DISC1 enriched in nucleus: altered subcellular distribution in orbitofrontal cortex in psychosis and substance/alcohol abuse. Proceedings of the National Academy of Sciences of the United States of America, 102(4), pp.1187–1192.

Scarr, E. & Dean, B., 2012. Altered neuronal markers following treatment with mood stabilizer and antipsychotic drugs indicate an increased likelihood of neurotransmitter release. Clinical psychopharmacology and neuroscience: the official scientific journal of the Korean College of Neuropsychopharmacology, 10(1), pp.25–33. 147

Schapansky, J. et al., 2009. Neuregulin beta1 enhances peak glutamate-induced intracellular calcium levels through endoplasmic reticulum calcium release in cultured hippocampal neurons. Canadian Journal of Physiology and Phar- macology, 87(10), pp.883–891.

Scheel, M. et al., 2013. Myelination deficits in schizophrenia: evidence from diffusion tensor imaging. Brain structure & function, 218(1), pp.151–156.

Schenkel, L.S. & Silverstein, S.M., 2004. Dimensions of premorbid functioning in schizophrenia: a review of neuromotor, cognitive, social, and behavioral domains. Genetic, social, and general psychology monographs, 130(3), pp.241– 270.

Schleimer, S.B. et al., 2004. GABA transporters GAT-1 and GAT-3 in the human dorsolateral prefrontal cortex in schizophrenia. Neuropsychobiology, 50(3), pp.226–230.

Schmitt, A. et al., 2003. Effects of long-term antipsychotic treatment on NMDA receptor binding and gene expression of subunits. Neurochemical research, 28(2), pp.235–241.

Schmitt, A. et al., 2011. Schizophrenia as a disorder of disconnectivity. European archives of psychiatry and clinical neuroscience, 261 Suppl 2, pp.S150–154.

Schmittgen, T.D. & Livak, K.J., 2008. Analyzing real-time PCR data by the comparative C(T) method. Nature protocols, 3(6), pp.1101–1108.

Schobel, S.A. et al., 2009. Differential targeting of the CA1 subfield of the hippocampal formation by schizophrenia and related psychotic disorders. Archives of General Psychiatry, 66(9), pp.938–946.

Schreiber, S. et al., 2011. Increased density of GAD65/67 immunoreactive neurons in the posterior subiculum and parahippocampal gyrus in treated patients with chronic schizophrenia. The world journal of biological psychiatry: the official journal of the World Federation of Societies of Biological Psychiatry, 12(1), pp.57–65.

Seeman, P., 1987. Dopamine receptors and the dopamine hypothesis of schizophrenia. Synapse, 1(2), pp.133–152.

Seeman, P., 2009. Glutamate and dopamine components in schizophrenia. Journal of psychiatry & neuroscience: JPN, 34(2), pp.143–149.

Seeman, P., Guan, H.C. & Hirbec, H., 2009. Dopamine D2High receptors stimulated by phencyclidines, lysergic acid diethylamide, salvinorin A, and modafinil. Synapse, 63(8), pp.698–704.

Seeman, P., Ko, F. & Tallerico, T., 2005. Dopamine receptor contribution to the action of PCP, LSD and ketamine psychotomimetics. Molecular psychiatry, 10(9), pp.877–883.

148

Seshacharyulu, P. et al., 2012. Targeting the EGFR signaling pathway in cancer therapy. Expert Opinion on Therapeutic Targets, 16(1), pp.15–31.

Seshadri, S. et al., 2010. Disrupted-in-Schizophrenia-1 expression is regulated by beta-site amyloid precursor protein cleaving enzyme-1-neuregulin cascade. Proceedings of the National Academy of Sciences of the United States of America, 107(12), pp.5622–5627.

Shamir, A. et al., 2012. The Importance of the NRG-1/ErbB4 Pathway for Synaptic Plasticity and Behaviors Associated with Psychiatric Disorders. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 32(9), pp.2988–2997.

Shamir, A. & Buonanno, A., 2010. Molecular and cellular characterization of Neu- regulin-1 type IV isoforms. Journal of Neurochemistry, 113(5), pp.1163– 1176.

Shao, W. et al., 2012. Suppression of neuroinflammation by astrocytic dopamine D2 receptors via αB-crystallin. Nature, 494(7435), pp.90–94.

Sharif, A. et al., 2009. Differential erbB signaling in astrocytes from the cerebral cortex and the hypothalamus of the human brain. Glia, 57(4), pp.362–379.

Shibuya, M. et al., 2010. Measurement and comparison of serum neuregulin 1 immunoreactivity in control subjects and patients with schizophrenia: an influence of its genetic polymorphism. Journal of Neural Transmission (Vienna, Austria: 1996), 117(7), pp.887–895.

Shinoda, J. et al., 1997. Multiple isoforms of neuregulin are expressed in developing rat dorsal root ganglia. Journal of neuroscience research, 50(5), pp.673–683.

Shirazi-Southall, S., Rodriguez, D.E. & Nomikos, G.G., 2002. Effects of typical and atypical antipsychotics and receptor selective compounds on acetylcholine efflux in the hippocampus of the rat. Neuropsychopharmacology, 26(5), pp.583–594.

Shyu, W.-C. et al., 2004. Neuregulin-1 reduces ischemia-induced brain damage in rats. Neurobiology of Aging, 25(7), pp.935–944.

Silberberg, G. et al., 2006. The involvement of ErbB4 with schizophrenia: association and expression studies. American Journal of Medical Genetics. Part B, Neuropsychiatric Genetics: The Official Publication of the International So- ciety of Psychiatric Genetics, 141B(2), pp.142–148.

Small, S.A. et al., 2011. A pathophysiological framework of hippocampal dysfunction in ageing and disease. Nat Rev Neurosci, 12(10), pp.585–601.

Smieskova, R. et al., 2009. The effects of antipsychotics on the brain: what have we learnt from structural imaging of schizophrenia?--a systematic review. Cur- rent pharmaceutical design, 15(22), pp.2535–2549.

149

Smith, D.E. & Wesson, D.R., 1980. PCP abuse: Diagnostic and psychopharmacolog- ical treatment approaches. Journal of Psychedelic Drugs, 12(3-4), pp.293– 299.

Stanford, J.A., Giardina, K. & Gerhardt, G.A., 2000. In vivo microdialysis studies of age-related alterations in potassium-evoked overflow of dopamine in the dorsal striatum of Fischer 344 rats. International Journal of Developmental Neu- roscience: The Official Journal of the International Society for Developmen- tal Neuroscience, 18(4-5), pp.411–416.

Stefanis, N.C. et al., 2013. Schizophrenia candidate gene ERBB4: covert routes of vulnerability to psychosis detected at the population level. Schizophrenia bulletin, 39(2), pp.349–357.

Stefansson, H. et al., 2003. Association of neuregulin 1 with schizophrenia confirmed in a Scottish population. American Journal of Human Genetics, 72(1), pp.83–87.

Stefansson, H. et al., 2002. Neuregulin 1 and susceptibility to schizophrenia. Ameri- can Journal of Human Genetics, 71(4), pp.877–892.

Steinecke, A. et al., 2012. Disrupted-in-Schizophrenia 1 (DISC1) is necessary for the correct migration of cortical interneurons. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 32(2), pp.738–745.

Steiner, H. et al., 1999. Differential expression of ErbB3 and ErbB4 neuregulin receptors in dopamine neurons and forebrain areas of the adult rat. Experi- mental Neurology, 159(2), pp.494–503.

Straub, R.E. et al., 2007. Allelic variation in GAD1 (GAD67) is associated with schizophrenia and influences cortical function and gene expression. Molecu- lar psychiatry, 12(9), pp.854–869.

Suzuki, Y. et al., 2002. Acute administration of phencyclidine induces tonic activation of medial prefrontal cortex neurons in freely moving rats. Neuroscience, 114(3), pp.769–779.

Swerdlow, N.R. et al., 1991. Effects of spiperone, raclopride, SCH 23390 and clozapine on apomorphine inhibition of sensorimotor gating of the startle response in the rat. The Journal of pharmacology and experimental therapeutics, 256(2), pp.530–536.

Swerdlow, N.R. et al., 2008. Realistic expectations of prepulse inhibition in translational models for schizophrenia research. Psychopharmacology, 199(3), pp.331–388.

Swerdlow, N.R. & Geyer, M.A., 2011. Using an Animal Model of Deficient Sen- sorimotor Gating to Study the Pathophysiology and New Treatments of Schizophrenia. Journal of Clinical Pathology, 64(6), pp.285–301.

150

Swerdlow, N.R., Geyer, M.A. & Braff, D.L., 2001. Neural circuit regulation of prepulse inhibition of startle in the rat: current knowledge and future challenges. Psychopharmacology, 156(2-3), pp.194–215.

Syed, N. et al., 2010. Soluble neuregulin-1 has bifunctional, concentration-dependent effects on Schwann cell myelination. The Journal of Neuroscience: The Offi- cial Journal of the Society for Neuroscience, 30(17), pp.6122–6131.

Takács, V.T. et al., 2012. Extrinsic and local glutamatergic inputs of the rat hippocampal CA1 area differentially innervate pyramidal cells and interneurons. Hippocampus, 22(6), pp.1379–1391.

Talmage, D.A., 2008. Mechanisms of neuregulin action. Novartis Foundation Sym- posium, 289, pp.74–84; discussion 84–93.

Tamminga, C.A., Stan, A.D. & Wagner, A.D., 2010. The hippocampal formation in schizophrenia. The American journal of psychiatry, 167(10), pp.1178–1193.

Tamura, H. et al., 2012. Processing of neuregulin-1 by neuropsin regulates GABAer- gic neuron to control neural plasticity of the mouse hippocampus. The Jour- nal of neuroscience: the official journal of the Society for Neuroscience, 32(37), pp.12657–12672.

Tan, W. et al., 2007. Molecular cloning of a brain-specific, developmentally regulated neuregulin 1 (NRG1) isoform and identification of a functional promoter variant associated with schizophrenia. The Journal of biological chemistry, 282(33), pp.24343–24351.

Tanaka, S. et al., 2003. GABAergic modulation of hippocampal glutamatergic neurons: an in vivo microdialysis study. European Journal of Pharmacology, 465(1-2), pp.61–67.

Tandon, R., Keshavan, M.S. & Nasrallah, H.A., 2008. Schizophrenia, “just the facts” what we know in 2008. 2. Epidemiology and etiology. Schizophrenia Re- search, 102(1-3), pp.1–18.

Tandon, R., Nasrallah, H.A. & Keshavan, M.S., 2009. Schizophrenia, “just the facts” 4. Clinical features and conceptualization. Schizophrenia Research, 110(1-3), pp.1–23.

Tandon, R., Nasrallah, H.A. & Keshavan, M.S., 2010. Schizophrenia, “just the facts” 5. Treatment and prevention. Past, present, and future. Schizophrenia Re- search, 122(1-3), pp.1–23.

Tanner, T., 1979. GABA-induced locomotor activity in the rat, after bilateral injection into the ventral tegmental area. Neuropharmacology, 18(5), pp.441–446.

Tayoshi, S. et al., 2010. GABA concentration in schizophrenia patients and the effects of antipsychotic medication: a proton magnetic resonance spectroscopy study. Schizophrenia research, 117(1), pp.83–91.

151

Tayoshi, S. et al., 2009. Metabolite changes and gender differences in schizophrenia using 3-Tesla proton magnetic resonance spectroscopy (1H-MRS). Schizo- phrenia research, 108(1-3), pp.69–77.

Thompson Ray, M. et al., 2011. Decreased BDNF, trkB-TK+ and GAD67 mRNA expression in the hippocampus of individuals with schizophrenia and mood disorders. Journal of psychiatry & neuroscience: JPN, 36(3), pp.195–203.

Thomsen, M.S. et al., 2009. alpha(7) Nicotinic acetylcholine receptor activation prevents behavioral and molecular changes induced by repeated phencyclidine treatment. Neuropharmacology, 56(6-7), pp.1001–1009.

Thomsen, M.S. et al., 2010. Cognitive improvement by activation of alpha7 nicotinic acetylcholine receptors: from animal models to human pathophysiology. Cur- rent Pharmaceutical Design, 16(3), pp.323–343.

Ting, A.K. et al., 2011. Neuregulin 1 promotes excitatory synapse development and function in GABAergic interneurons. The Journal of Neuroscience: The Offi- cial Journal of the Society for Neuroscience, 31(1), pp.15–25.

Toda, M. & Abi-Dargham, A., 2007. Dopamine hypothesis of schizophrenia: making sense of it all. Current psychiatry reports, 9(4), pp.329–336.

Torrey, E.F. et al., 2005. Neurochemical markers for schizophrenia, bipolar disorder, and major depression in postmortem brains. Biological psychiatry, 57(3), pp.252–260.

Toth, E. & Lajtha, A., 1986. Antagonism of phencyclidine-induced hyperactivity by glycine in mice. Neurochemical research, 11(3), pp.393–400.

Traynelis, S.F. et al., 2010. Glutamate receptor ion channels: structure, regulation, and function. Pharmacological reviews, 62(3), pp.405–496.

Tyrer, P. et al., 2011. Psychosis with coexisting substance misuse; NICE clinical guideline 120, National Institute for Health and Care Excellence. Available at: guidance.nice.org.uk/cg120.

Tzahar, E. et al., 1996. A hierarchical network of interreceptor interactions deter- mines signal transduction by Neu differentiation factor/neuregulin and epidermal growth factor. Molecular and cellular biology, 16(10), pp.5276–5287.

Uhlhaas, P.J. & Singer, W., 2010. Abnormal neural oscillations and synchrony in schizophrenia. Nature reviews. Neuroscience, 11(2), pp.100–113.

Uhlhaas, P.J. & Singer, W., 2011. The development of neural synchrony and large- scale cortical networks during adolescence: relevance for the pathophysiology of schizophrenia and neurodevelopmental hypothesis. Schizophrenia bulletin, 37(3), pp.514–523.

Väisänen, J. et al., 1999. Excitatory actions of NMDA receptor antagonists in rat entorhinal cortex and cultured entorhinal cortical neurons. Neuropsychophar-

152

macology: official publication of the American College of Neuropsycho- pharmacology, 21(1), pp.137–146.

Vandesompele, J. et al., 2002. Accurate normalization of real-time quantitative RT- PCR data by geometric averaging of multiple internal control genes. Genome biology, 3(7).

Vanover, K.E. et al., 2004. Pharmacological characterization of AC-90179 2-(4- methoxyphenyl)-N-(4-methyl-benzyl)-N-(1-methyl-piperidin-4 -yl)acetamide hydrochloride : A selective serotonin 2A receptor inverse agonist. Journal of Pharmacology and Experimental Therapeutics, 310(3), pp.943–951.

Vartanian, T. et al., 1997. Axonal neuregulin signals cells of the oligodendrocyte lin- eage through activation of HER4 and Schwann cells through HER2 and HER3. The Journal of cell biology, 137(1), pp.211–220.

Vawter, M.P. et al., 1999. Alterations of hippocampal secreted N-CAM in bipolar disorder and synaptophysin in schizophrenia. Molecular psychiatry, 4(5), pp.467–475.

Viggiano, D., 2008. The hyperactive syndrome: metanalysis of genetic alterations, pharmacological treatments and brain lesions which increase locomotor activity. Behavioural brain research, 194(1), pp.1–14.

Vinkers, C.H. et al., 2010. The inhibitory GABA system as a therapeutic target for cognitive symptoms in schizophrenia: investigational agents in the pipeline. Expert opinion on investigational drugs, 19(10), pp.1217–1233.

Volk, D. et al., 2001. GABA transporter-1 mRNA in the prefrontal cortex in schizophrenia: decreased expression in a subset of neurons. The American journal of psychiatry, 158(2), pp.256–265.

Volk, D.W. et al., 2000. Decreased glutamic acid decarboxylase67 messenger RNA expression in a subset of prefrontal cortical gamma-aminobutyric acid neurons in subjects with schizophrenia. Archives of general psychiatry, 57(3), pp.237–245.

Volk, D.W. et al., 2012. Deficits in transcriptional regulators of cortical parvalbumin neurons in schizophrenia. The American journal of psychiatry, 169(10), pp.1082–1091.

Volk, D.W. et al., 2002. Reciprocal alterations in pre- and postsynaptic inhibitory markers at chandelier cell inputs to pyramidal neurons in schizophrenia. Cer- ebral cortex (New York, N.Y.: 1991), 12(10), pp.1063–1070.

Vullhorst, D. et al., 2009. Selective expression of ErbB4 in interneurons, but not pyramidal cells, of the rodent hippocampus. J Neurosci, 29(39), pp.12255–64.

Wadugu, B. & Kühn, B., 2012. The role of neuregulin/ErbB2/ErbB4 signaling in the heart with special focus on effects on cardiomyocyte proliferation. American

153

journal of physiology. Heart and circulatory physiology, 302(11), pp.H2139– 2147.

Walker, R.M. et al., 2010. Association analysis of Neuregulin 1 candidate regions in schizophrenia and bipolar disorder. Neuroscience Letters, 478(1), pp.9–13.

Walss-Bass, C. et al., 2006. A novel missense mutation in the transmembrane domain of neuregulin 1 is associated with schizophrenia. Biological Psychiatry, 60(6), pp.548–553.

Wang, A.Y. et al., 2011. Bipolar disorder type 1 and schizophrenia are accompanied by decreased density of parvalbumin- and somatostatin-positive interneurons in the parahippocampal region. Acta neuropathologica, 122(5), pp.615–626.

Wang, C. et al., 2004. Blockade of N-methyl-D-aspartate receptors by phencyclidine causes the loss of corticostriatal neurons. Neuroscience, 125(2), pp.473–483.

Wang, C. et al., 1999. Chronic phencyclidine increases NMDA receptor NR1 subunit mRNA in rat forebrain. Journal of neuroscience research, 55(6), pp.762–769.

Wang, C., Fridley, J. & Johnson, K.M., 2005. The role of NMDA receptor upregulation in phencyclidine-induced cortical apoptosis in organotypic culture. Bio- chemical pharmacology, 69(9), pp.1373–1383.

Wang, Y.-F. et al., 2013. GABAergic inhibition through synergistic astrocytic neuronal interaction transiently decreases vasopressin neuronal activity during hypoosmotic challenge. The European journal of neuroscience, 37(8), pp.1260–1269.

Webster, M.J. et al., 2001. Synaptophysin and GAP-43 mRNA levels in the hippocampus of subjects with schizophrenia. Schizophrenia research, 49(1-2), pp.89–98.

Weickert, C.S. et al., 2013. Molecular evidence of N-methyl-D-aspartate receptor hypofunction in schizophrenia. Molecular Psychiatry, 18(11), pp.1185–1192.

Wen, L. et al., 2010. Neuregulin 1 regulates pyramidal neuron activity via ErbB4 in parvalbumin-positive interneurons. Proceedings of the National Academy of Sciences of the United States of America, 107(3), pp.1211–1216.

White, T. et al., 2007. Disruption of hippocampal connectivity in children and adolescents with schizophrenia--a voxel-based diffusion tensor imaging study. Schizophrenia research, 90(1-3), pp.302–307.

Wiley, J.L., 1994. Clozapine’s effects on phencyclidine-induced disruption of prepulse inhibition of the acoustic startle response. Pharmacology Biochemistry and Behavior, 49(4), pp.1025–1028.

Wiley, J.L. & Kennedy, K.L., 2002. Evaluation of the efficacy of antipsychotic attenuation of phencyclidine-disrupted prepulse inhibition in rats. Journal of Neural Transmission, 109(4), pp.523–535.

154

Williams, N.M. et al., 2003. Support for genetic variation in neuregulin 1 and susceptibility to schizophrenia. Molecular Psychiatry, 8(5), pp.485–487.

Woo, R.-S. et al., 2007. Neuregulin-1 enhances depolarization-induced GABA release. Neuron, 54(4), pp.599–610.

Woo, R.-S. et al., 2012. Neuregulin-1 protects against neurotoxicities induced by Swedish amyloid precursor protein via the ErbB4 receptor. Neuroscience, 202, pp.413–423.

Woo, T.-U.W., Kim, A.M. & Viscidi, E., 2008. Disease-specific alterations in glutamatergic neurotransmission on inhibitory interneurons in the prefrontal cortex in schizophrenia. Brain research, 1218, pp.267–277.

Woo, T.-U.W., Spencer, K. & McCarley, R.W., 2010. Gamma oscillation deficits and the onset and early progression of schizophrenia. Harvard review of psychiatry, 18(3), pp.173–189.

Woo, T.-U.W., Walsh, J.P. & Benes, F.M., 2004. Density of glutamic acid decarboxylase 67 messenger RNA-containing neurons that express the N-methyl-D- aspartate receptor subunit NR2A in the anterior cingulate cortex in schizophrenia and bipolar disorder. Archives of General Psychiatry, 61(7), pp.649– 657.

Wynn, J.K. et al., 2007. Effects of olanzapine, risperidone and haloperidol on prepulse inhibition in schizophrenia patients: a double-blind, randomized controlled trial. Schizophrenia research, 95(1-3), pp.134–142.

Xie, F., Raetzman, L.T. & Siegel, R.E., 2004. Neuregulin induces GABAA receptor beta2 subunit expression in cultured rat cerebellar granule neurons by activating multiple signaling pathways. Journal of Neurochemistry, 90(6), pp.1521– 1529.

Xu, Z. et al., 2004. Neuregulin-1 is neuroprotective and attenuates inflammatory responses induced by ischemic stroke. Biochemical and Biophysical Research Communications, 322(2), pp.440–446.

Yago, K.B. et al., 1981. The urban epidemic of phencyclidine (PCP) use: Clinical and laboratory evidence from a public psychiatric hospital emergency ser- vice. Journal of Clinical Psychiatry, 42(5), pp.193–196.

Yamada, S. et al., 1999. Involvement of serotonin 2A receptors in phencyclidine- induced disruption of prepulse inhibition of the acoustic startle in rats. Bio- logical Psychiatry, 46(6), pp.832–838.

Yamamoto, Y., Kakigi, T. & Maeda, K., 1999. Intra-striatal phencyclidine inhibits N-methyl-D-aspartic acid- stimulated increase in glutamate levels of freely moving rats. Progress in Neuro-Psychopharmacology and Biological Psychi- atry, 23(1), pp.161–174.

155

Yamamura, S., Ohoyama, K., Hamaguchi, T., Kashimoto, K., et al., 2009. Effects of quetiapine on monoamine, GABA, and glutamate release in rat prefrontal cortex. Psychopharmacology, 206(2), pp.243–258.

Yamamura, S., Ohoyama, K., Hamaguchi, T., Nakagawa, M., et al., 2009. Effects of zotepine on extracellular levels of monoamine, GABA and glutamate in rat prefrontal cortex. British Journal of Pharmacology, 157(4), pp.656–665.

Yang, X. et al., 1998. A cysteine-rich isoform of neuregulin controls the level of expression of neuronal nicotinic receptor channels during synaptogenesis. Neu- ron, 20(2), pp.255–270.

Yau, H.-J. et al., 2003. Neural development of the neuregulin receptor ErbB4 in the cerebral cortex and the hippocampus: preferential expression by interneurons tangentially migrating from the ganglionic eminences. Cerebral cortex (New York, N.Y.: 1991), 13(3), pp.252–264.

Yee, B.K., Chang, D.L.T. & Feldon, J., 2004. The effects of dizocilpine and phencyclidine on prepulse inhibition of the acoustic startle reflex and on prepulse- elicited reactivity in C57BL6 mice. Neuropsychopharmacology, 29(10), pp.1865–1877.

Yeomans, J.S. et al., 2010. GABA receptors and prepulse inhibition of acoustic startle in mice and rats. The European journal of neuroscience, 31(11), pp.2053– 2061.

Yin, D.-M. et al., 2013. Reversal of behavioral deficits and synaptic dysfunction in mice overexpressing neuregulin 1. Neuron, 78(4), pp.644–657.

Yonezawa, Y. et al., 1998. Involvement of γ-aminobutyric acid neurotransmission in phencyclidine-induced dopamine release in the medial prefrontal cortex. Eu- ropean Journal of Pharmacology, 341(1), pp.45–56.

Yu, B. et al., 2002. Adaptation to chronic PCP results in hyperfunctional NMDA and hypofunctional GABA(A) synaptic receptors. Neuroscience, 113(1), pp.1–10.

Yu, X.M. et al., 1997. NMDA channel regulation by channel-associated protein tyrosine kinase Src. Science (New York, N.Y.), 275(5300), pp.674–678.

Yu, Z. et al., 2013. GABA transporter-1 deficiency confers schizophrenia-like behavioral phenotypes. PloS one, 8(7), p.e69883.

Yurek, D.M. et al., 2004. Supranigral injection of neuregulin1-beta induces striatal dopamine overflow. Brain Research, 1028(1), pp.116–119.

Zavitsanou, K. et al., 2008. Rapid cortico-limbic alterations in AMPA receptor densities after administration of PCP: implications for schizophrenia. Journal of Chemical Neuroanatomy, 36(2), pp.71–76.

Zhang, Z.J. & Reynolds, G.P., 2002. A selective decrease in the relative density of parvalbumin-immunoreactive neurons in the hippocampus in schizophrenia. Schizophrenia Research, 55(1-2), pp.1–10. 156

Zhao, W.-J., 2013. The expression and localization of neuregulin-1 (Nrg1) in the gastrointestinal system of the rhesus monkey. Folia histochemica et cytobiologi- ca / Polish Academy of Sciences, Polish Histochemical and Cytochemical So- ciety, 51(1), pp.38–44.

Zhong, C. et al., 2008. Presynaptic type III neuregulin 1 is required for sustained enhancement of hippocampal transmission by nicotine and for axonal targeting of alpha7 nicotinic acetylcholine receptors. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 28(37), pp.9111–9116.

Zhu, G. et al., 2004. Hyperactivity of endoplasmic reticulum associated exocytosis mechanism contributes to acute phencyclidine intoxication. Journal of Phar- macological Sciences, 95(2), pp.214–227.

Zhu, X. et al., 1995. Neuregulin receptors, erbB3 and erbB4, are localized at neuromuscular synapses. The EMBO journal, 14(23), pp.5842–5848.

Zink, M. et al., 2004. Differential effects of long-term treatment with clozapine or haloperidol on GABAA receptor binding and GAD67 expression. Schizo- phrenia research, 66(2-3), pp.151–157.

Zukin, S.R., 1982. Differing stereospecificities distinguish opiate receptor subtypes. Life sciences, 31(12-13), pp.1307–1310.

157