Comparing the Effects of Subchronic Phencyclidine and Medial Prefrontal Cortex Lesions on Cognitive Tests Relevant to Schizophrenia

Journal: Psychopharmacology

Manuscript ID: Psych-2015-00226.R2

Manuscript Type: Original Investigation

Date Submitted by the Author: 01-Jul-2015

Complete List of Authors: McAllister, Kathryn; The University of Cambridge, Department of Psychology; The University of Cambridge, MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute Mar, Adam; The University of Cambridge, Department of Psychology; The University of Cambridge, MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute Theobald, David; The University of Cambridge, Department of Psychology; The University of Cambridge, MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute Saksida, Lisa; The University of Cambridge, Department of Psychology; The University of Cambridge, MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute Bussey, Timothy; The University of Cambridge, Department of Psychology; The University of Cambridge, MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute

SCHIZOPHRENIA, PHENCYCLIDINE, PREFRONTAL CORTEX, OBJECT Keywords: RECOGNITION, RAT, ANIMAL MODEL, ASSOCIATIVE LEARNING, GLUTAMATE, LEARNING AND MEMORY, DISCRIMINATION

Page 1 of 37 Psychopharmacology

1 2 3 Comparing the Effects of Subchronic Phencyclidine and Medial 4 5 Prefrontal Cortex Dysfunction on Cognitive Tests Relevant to 6 Schizophrenia 7 8 ab ab ab a,b a,b 9 McAllister KAL , Mar AC , Theobald DE , Saksida LM , Bussey TJ 10

11 a 12 University of Cambridge Department of Psychology, Downing Street, 13 Cambridge, CB2 3EB, UK 14 15 b MRC and Wellcome Trust Behavioural and Clinical Neuroscience Institute, 16 University of Cambridge, Cambridge CB2 3EB, UK 17 18 19 Acknowledgements: KAL McAllister received funding from the Cambridge 20 Commonwealth Trusts and University of Cambridge Overseas Studentship 21 22 Programme. LM Saksida and TJ Bussey also received funding from the 23 Innovative Medicine Initiative Joint Undertaking under grant agreement no 24 115008 of which resources are composed of EFPIA in-kind contribution and 25 financial contribution from the European Union’s Seventh Framework 26 Programme (FP7/2007-2013). 27 28 29 Disclosures: LMS and TJB consult for Campden Instruments Ltd. 30 31 32 Communicating author: 33 34 KAL McAllister 35 36 20 Manchester Sq. 37 London, UK, W1U 3PZ 38 39 [email protected] 40 +44 (0)7898 989 195 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 1 Psychopharmacology Page 2 of 37

1 2 3 Abstract 4 5 Rationale 6 It is becoming increasingly clear that the development of treatments for 7 cognitive symptoms of schizophrenia requires urgent attention, and that valid 8 animal models of relevant impairments are required. With subchronic PCP 9 10 (scPCP) a putative model of such impairment, the extent to which changes 11 following scPCP do or do not resemble those following dysfunction of the 12 prefrontal cortex is of importance. 13 Objectives 14 15 The present study carried out a comparison of the most common scPCP 16 dosing regimen with excitotoxin-induced medial prefrontal cortex (mPFC) 17 dysfunction in the rat, across several cognitive tests relevant to schizophrenia. 18 19 Methods 20 ScPCP subjects were dosed i.p. with 5mg/kg PCP or vehicle twice daily for 21 one week followed by 1 week washout prior to behavioural testing. mPFC 22 23 dysfunction was induced via fibre-sparing excitotoxin infused into the 24 prelimbic and infralimbic cortex. Subjects were tested on spontaneous novel 25 object recognition, touchscreen object-location paired-associates learning, 26 and touchscreen reversal learning. 27 28 Results 29 A double-dissociation was observed between object-location paired- 30 associates learning and object recognition: mPFC dysfunction impaired 31 acquisition of the object-location task, but not spontaneous novel object 32 recognition, while scPCP impaired spontaneous novel object recognition, but 33 34 not object-location associative learning. Both scPCP and mPFC dysfunction 35 resulted in a similar facilitation of reversal learning. 36 Conclusions 37 38 The pattern of impairment following scPCP raises questions around its 39 efficacy as a model of cognitive impairment in schizophrenia, particularly if 40 importance is placed on faithfully replicating the effects of mPFC dysfunction. 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 2 Page 3 of 37 Psychopharmacology

1 2 3 Introduction 4 5 6 It is becoming increasingly clear that the development of treatments for 7 the cognitive symptoms of schizophrenia requires urgent attention 8 (Minzenberg and Carter 2012; Schaefer et al. 2013; Vingerhoets et al. 2013). 9 10 To achieve this aim we need valid animal models of the cognitive impairments 11 in schizophrenia (Nestler and Hyman 2010). The psychotomimetic agent 12 phencyclidine (PCP) has been widely used to model schizophrenia in rodents 13 (Castner et al. 2004; Neill et al. 2010). In humans, PCP and other N-methyl- 14 D-aspartate receptor (NMDAR) antagonists can induce the symptoms of 15 schizophrenia and cause relapse for schizophrenics stabilized on anti- 16 psychotic medication (Itil et al. 1967; Javitt and Zukin 1991; Krystal et al. 17 1994; Lahti et al. 1995 & 2001; Malhotra et al. 1996 & 1997; Snyder 1980; 18 Tamminga 1998). These observations are consistent with the ‘glutamate 19 20 hypothesis of schizophrenia’, which posits that NMDAR function is disrupted 21 in schizophrenia (Javitt et al. 2012; Poels et al. 2014; Singh and Singh 2011). 22 Subchronic PCP (scPCP) in particular has many prima facie 23 advantages. For example, cognitive impairments relevant to schizophrenia 24 have been observed following scPCP (Arnt et al. 2010; Barnes et al. 2012; 25 26 Beninger et al. 2010; Howes et al. 2015; Laurent and Podhorna 2004; 27 McKibben et al. 2010; Meltzer et al. 2013) and the model, like other 28 subchronic treatments, allows assessment of the effects of the treatment 29 when the drug is no longer ‘on board’, thus minimizing motoric and/or 30 motivational changes which can seriously confound cognitive testing and 31 subsequent interpretation of the results. 32 33 Another attraction of the scPCP model is the assumption that it induces 34 dysfunction of the prefrontal cortex (Amitai et al. 2012; Young et al. 2015), a 35 major feature of the schizophrenic brain often cited as a major contributor to 36 cognitive deficits (Glahn et al. 2005; Hill et al. 2004; Ingvar and Franzen 37 1974). Therefore the extent to which changes following scPCP do or do not 38 resemble those found following dysfunction of the prefrontal cortex is of some 39 importance to researchers in this area. Therefore in the present study we 40 41 carried out a head-to-head comparison of the effects of the most common 42 scPCP dosing regimen (5mg/kg twice daily for seven days followed by 7 days 43 washout), with the effects of fibre-sparing excitotoxin-induced medial 44 prefrontal cortex (mPFC) dysfunction in the rat, across several cognitive tests 45 relevant to schizophrenia: object-location paired-associates learning, 46 spontaneous novel object recognition, and reversal learning. 47 48 Object-location learning (Experiment 1):Object-location learning, for 49 example as studied using the CANTAB PAL test (Sahakian et al. 1988), can 50 be used to investigate the MATRICS domain of visual learning and memory 51 and the CNTRICS domain of long term memory. Impairments in object- 52 location paired-associates learning (PAL) have been found in chronic 53 schizophrenia (Aubin et al. 2009; Donohoe et al. 2008; Haring et al. 2014; 54 Wood et al. 2002) and can also be present at first onset of the disorder and 55 may predict following clinical severity (Barnett et al. 2005). Additionally, 56 57 impairments on CANTAB PAL have been observed during the prodromal 58 phase and may be present in relatives at risk of developing schizophrenia 59 60 3 Psychopharmacology Page 4 of 37

1 2 3 (Bartok et al. 2005). Normal performance of CANTAB PAL depends on the 4 medial temporal lobe (Wood et al. 2002) and prefrontal cortex (Owen et al. 5 1995). 6 7 The rodent object-location paired-associates learning task (PAL) was 8 inspired by the CANTAB PAL task (Talpos et al. 2009) and has been 9 recommended by CNTRICS for studying long-term memory in animal models 10 of schizophrenia (Bussey et al. 2013). 11 12 Object Recognition (Experiment 2) . Spontaneous Object Recognition 13 (SOR) has been used to examine non-spatial memory impairments linked to 14 schizophrenia (Dere et al. 2007; Lyon et al. 2012). In studies of SOR and 15 scPCP immediately post-washout, significant impairments were observed 16 following various retention periods (Arnt et al. 2010; Horiguchi et al. 2011; 17 McKibben et al. 2010; Snigdha et al. 2010). Additionally, object recognition 18 impairments following scPCP are attenuated by clozapine and risperidone, 19 20 but not haloperidol (Grayson et al. 2007). 21 Reversal Learning (Experiment 3) : Reversal learning assesses the 22 subject’s ability to acquire a simple ‘rule’ and then adopt a new strategy when 23 the rule no longer holds. Reversal learning is impaired in schizophrenia 24 (Leeson et al. 2009; McKirdy et al. 2009; Murray et al. 2008; Waltz and Gold 25 26 2007) and has been recommended by CNTRICS for studying executive 27 function in animal models of schizophrenia (Gilmour et al. 2013). First, rats 28 learn to discriminate between an S+ and S- by perceptually discriminating 29 between S+ and S- and learning the reward contingency. Once the subject 30 has acquired the discrimination, the reward contingency is reversed (S+ 31 becomes the S-). The reversal stage requires the subject to inhibit 32 responding to the previously rewarded stimulus, and switch to the new reward 33 contingencies. 34 35 Reversal learning ability is dependent on the orbitofronal cortex 36 (Bissonette et al. 2008; Burke et al. 2009; Chudasama and Robbins 2003) 37 and striatum (Ragozzino 2007; Ragozzino et al. 2002). Previous studies have 38 found impairments in reversal learning in schizophrenia (McKirdy et al. 2009; 39 Murray et al. 2008; Schlagenhauf et al. 2014; Waltz and Gold 2007), which 40 41 are present following correction for IQ (Leeson et al. 2009), consistent with 42 hypofrontality in schizophrenia. 43 44 45 Methods 46 47 48 Apparatus: Rat touchscreen chambers were the same as those used 49 in previous touchscreen studies (McAllister et al 2013; Bussey et al 2008; 50 Talpos et al 2010). The object recognition Y-maze apparatus was the same 51 as used by Forwood et al. (2005) and Winters et al. (2004). Three identical 52 copies were used of each object with no natural significance or previous 53 54 presentation. 55 Surgery: mPFC lesions generally followed the adaptation of Birrell and 56 Brown’s protocol (2000), used in previous touchscreen studies centred on 57 prelimbic and infralimbic cortex (McAllister et al 2013). Bilateral injections of 58 59 60 4 Page 5 of 37 Psychopharmacology

1 2 3 0.2µL of 0.06M ibotenic acid (Sigma, UK) or vehicle were made at AP +3.5, L 4 ±0.6, V -5.2; AP+ 2.5, L ±0.6, V -5.0mm, relative to skull surface Bregma via 5 custom infusing line and 10µL Hamilton syringe and Harvard Instruments 6 Pump 11 (Holliston, Massachusetts) pump. Subjects recovered for at least 7 one week with ad libitum food and water prior to behavioural testing. 8 9 Histology: Protocol followed previous touchscreen studies of lesions 10 centred on mPFC (McAllister et al 2013). Coronal sections (60 µm) were 11 stained with NeuN and lesion locations were mapped onto standardized 12 sections of the (Paxinos and Watson 2007) atlas of the rat brain. 13 14 Subchronic Phencyclidine Protocol: Rats were dosed with 5mg/kg 15 PCP (Sigma, UK) via intraperitoneal injection (5mg/ml) or 1ml/kg vehicle twice 16 daily (7:30am/pm) for seven days, followed by a seven-day washout period 17 prior to behavioural testing. 18 19 Touchscreen Testing Protocol: Rats were group-housed on a reverse 20 light-dark schedule (light 7pm-7am) and tested during dark phase. A 21 restricted diet maintained rats at no less than 85% of free-feeding weight, with 22 water available ad libitum. Pretraining touchscreen use prior to object- 23 location or reversal learning tasks followed established protocols (Bussey et 24 al. 2008; Horner et al. 2013; McAllister et al 2013). This experiment was 25 26 conducted in accordance with the United Kingdom Animals (Scientific 27 Procedures) Act, 1986. 28 29 30 Object-location learning (Experiment 1) 31 scPCP treatment : 16 male 250-275g Lister Hooded rats were obtained 32 33 from Harlan, UK. After completion of pretraining rats were pseudorandomly 34 allocated to PCP or vehicle treatment groups. Following scPCP protocol rats 35 completed two days of ‘punish for incorrect’ pretaining at criterion levels 36 before PAL acquisition. PAL acquisition generally followed protocols 37 established by Talpos et al (2009) with the exception that rats were tested 38 twice daily over 11 blocks of 5 sessions each. 39 40 mPFC lesions: 16 male 250-275g Lister Hooded rats were obtained 41 from Harlan, UK. Behavioural methods were initially identical to those in 42 Experiment 1 scPCP, but as a difference between groups was observed, 43 testing was extended to 17 blocks to explore the persistence of the deficit. 44 Following acquisition of dPAL, an object-location challenge took place: the foil 45 stimulus was replaced with the correct stimulus for all trials (sPAL protocol; 46 Talpos et al., 2009; figure 1). 47 48 Behavioural testing : In this task three images are used as objects, each 49 correct when presented in one of three response locations, and incorrect 50 when presented in either of the other two. There are six unique trial types of 51 one object in its correct location and a different foil object in an incorrect 52 location (the ‘dPAL’ version of the task; Talpos et al 2009, figure 1). Talpos et 53 54 al (2009) demonstrated that pharmacological manipulations of the 55 hippocampus could impair performance of PAL, consistent with human 56 imaging studies showing the hippocampus activated during encoding and 57 parahippocampal gyrus during retrieval of the CANTAB PAL task (de Rover et 58 59 60 5 Psychopharmacology Page 6 of 37

1 2 3 al. 2011). Dependency on mPFC has not previously been investigated. In this 4 first experiment, we tested the effects of scPCP on the touchscreen PAL test. 5 6 7 8 9 10 11 12 13 14 15 Fig. 1 Trial-types in touchscreen object-location paired-associates learning 16 (dPAL left and sPAL right). S+ rewarded; S- results in ‘time out’. Adapted 17 from Talpos et al (2009). 18 19 20 21 Object recognition (Experiment 2) 22 The same scPCP and mPFC rats were used as in Experiment 1 and 23 maintained under the same housing conditions. Experiment 2 took place 24 immediately following PAL acquisition, with initial habituation beginning the 25 26 day following the last session of PAL. 27 Three daily 5-minute habituation sessions followed by two sets of 28 discriminations with novel object, order and novel side counterbalanced. 29 Discrimination trials consisted of 3-minute exposure to 2 copies of the same 30 31 object (sample phase), delay, then 3-minute choice phase with a third copy of 32 the pre-exposed object and a novel object. The apparatus and objects were 33 wiped down with a dilute ethanol solution between trials to prevent olfactory 34 trails. A minimum of 20 seconds exploration during sample phase was 35 required for analysis, defined as nose oriented towards the object within 2cm. 36 37 38 Discrimination scores were calculated for the total time as 39 well as the first minute. 40 41 In addition, this experiment allowed testing of the persistence of the 42 functional effects of scPCP, as it is conceivable that the lack of effect of 43 scPCP on PAL in Experiment 1 was due to the effects of scPCP decreasing 44 over time. Indeed, longitudinal testing of the effects of scPCP post-washout is 45 not common, with most testing taking place fewer than seven days post- 46 washout. Thus we tested the same rats used in Experiment 1 on 47 spontaneous object recognition at 2 time-points: 40 and 105 days post- 48 washout. 49 50 51 52 Reversal learning (Experiment 3) 53 32 male 250-275g Lister Hooded rats were obtained from Harlan, UK 54 and housed as in PAL protocol, above. 55 56 Behavioural testing took place in the touchscreen chambers used in 57 Experiment 1, with screen masking adapted for 2 choices. Acquisition 58 59 60 6 Page 7 of 37 Psychopharmacology

1 2 3 consisted of daily 1 hour sessions of up to 100 trials until all subjects reached 4 stable criterion of greater than 80% across 3 consecutive days. Subjects 5 were pseudorandomly assigned to treatment groups, counterbalancing for 6 performance. All rats immediately re-baselined at criterion and reversal 7 began on day 4 post-washout/11 post-lesion. Maximum session length was 8 9 100 trials or 1 hour, with the first session split into shorter sessions of 25, 25, 10 and 50 trials (one session/day) to ensure completion of trials (summed to 11 session 1). Rats were tested for 18 sessions. 12 13 14 Results 15 16 17 Experiment 1: Effects of scPCP on acquisition of the PAL task 18 19 Repeated-measures ANOVA did not demonstrate a significant effect of 20 scPCP in acquisition of object-location paired-associates learning (F (1-14) = 21 0.675, p=0.43) and both groups reached 80% correct by block 11. At this 22 point, rats moved on to Experiment 2. Log transformed reaction times 23 significantly decreased across acquisition (F( 2.37,33.12) = 16.31, p<0.001) with 24 no significant effect of scPCP (F( 1, 14) = 0.89, p= 0.361). Magazine latency 25 also decreased across acquisition (F = 2.67, p= 0.005), with no significant 26 (10,14) 27 effect of scPCP (F( 1,14) = 0.11, p=0.92). 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 Fig. 2 Effects of scPCP on acquisition of PAL. Blocks of 5 sessions are 47 plotted. Error bars ±1 SEM. 48 49 50 51 Experiment 1: mPFC Dysfunction and PAL Acquisition 52 Histological results: Histology confirmed lesions of the medial prefrontal 53 54 cortex in all rats, centred on the prelimbic and infralimbic cortices with 55 damage to overlying anterior cingulate cortex. Shams did not show significant 56 damage to any of those regions. One lesion was excluded from data analysis 57 as largely unilateral, resulting in 8 sham and 7 lesions. 58 59 60 7 Psychopharmacology Page 8 of 37

1 2 3 4 5 Fig. 3 mPFC lesions in PAL. 6 Damage common to all subjects in 7 solid black, with maximum extent 8 of any damage shown by the black 9 line. Coronal sections at 3.72mm, 10 2.76mm, and 1.08mm anterior to 11 Bregma. Images adapted from Paxinos and Watson 12 (2007). Composite photomicrograph presents contrast 13 between sham (left) and lesion (right) for demonstration 14 15 purposes. 16 17 18 19 20 21 22 23 Behavioural results: Repeated measures ANOVA demonstrated 24 significant main effect of lesion (F(1,12)= 5.01, p=0.045) and session block 25 (F(7.87,54)= 63.8, p<0.001), with no interaction (F(7.87,54)= 1.74, p=0.099). 26 Performance appeared to asymptote slightly below 90% correct for shams, 27 and 80% correct for lesions (no significant learning over blocks 17-19: shams 28 (F(2,12)= 0.25, p=0.78; lesion F(2,12)= 0.23, p=0.80). Response latencies 29 reduced over the course of acquisition (F(7.42,81.6)= 1.94, <0.001), with no 30 significant difference between sham and lesion ( F(1,11)= 1.46, p=0.25 or 31 latency to magazine (F(1,11)= 0.32, p=0.86). 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 8 Page 9 of 37 Psychopharmacology

1 2 3 Fig. 4 Effects of mPFC 4 5 dysfunction on acquisition of PAL 6 and sPAL object-location 7 challenge. Blocks of 5 sessions 8 are plotted. Error bars ±1 SEM. 9 *= p<0.05 (S-N-K). A significant 10 difference in standard deviation of 11 performance between trial-types 12 during PAL acquisition. 13 14 15 16 17 18 19 20 During the mPFC object-location challenge, there was no significant 21 difference in performance between dPAL and sPAL for sham (t (12) =0.46, 22 p=0.65) or lesion (t (12) =0.59, p=0.56), suggesting both groups used an object- 23 location strategy (rather than, e.g., a conditional strategy). The lesion group 24 was impaired on both sPAL and dPAL (dPAL t =3.05, p=0.01; sPAL 25 (12) 26 t(12) =2.14, p=0.044). To further probe the nature of task acquisition, analysis 27 of individual trial types was then undertaken. 28 Analysis showed different rates of learning of each trial-type-- not 29 unexpected, as it is likely an animal would, for various reasons, work out 30 some of the object-location bindings before others. Interestingly, however, 31 32 initial examination by trial-type across acquisition found greater variance in 33 the lesion group versus shams, suggesting that the lesion group had not fully 34 solved all of the object-location paired-associates. 35 To quantify this variability we calculated, for each rat, mean standard 36 37 deviation on each of the 6 trial types throughout acquisition (figure 4). There 38 was a significant effect of block (ANOVA F (8.4,108) =2.94, p<.01) and a 39 significant lesion x block interaction (F (8.4,108) =2.53, p=0.01) as the sham group 40 learned to perform all trial-types with greater proficiency, whereas the rats in 41 the lesion group learned some trial types but not others. 42 43 44 Experiment 2: scPCP and Object Recognition 45 46 Both 1 minute and 3 minute discrimination scores showed significant 47 decrease in preference for the novel object in the scPCP group at six but not 48 fifteen weeks post-washout (figure 5). There was no significant difference 49 between sham and lesion in overall exploration time during sample or choice 50 at six weeks (sample t (12) =0.086, p= 0.51; choice t (12) = 0.82, p= 0.43). scPCP 51 significantly impaired novel object recognition at a 1 hour delay compared to 52 vehicle t = 2.55, p< 0.01, with the scPCP discrimination score not 53 (12) 54 significantly greater than chance (t (6) =1.69, p=0.14). 55 56 57 58 59 60 9 Psychopharmacology Page 10 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Fig. 5 ScPCP and object recognition 40 and 105 days post-washout. Means 19 20 of first minute discrimination scores. ±1SEM. *=p <0.01. 21 22 23 Re-testing three months (105 days) post washout found no impairment 24 in scPCP at a 1 hour or 24 hour delay (figure 5; 1 hour t (12) = 0.05, p= 0.95; 24 25 hour t (12) = 0.07 p= 0.92), and no difference in exploration time (1 hour sample: 26 t(12) =1.21, p= 0.25, choice: t (12) = 1.43, p= 0.18; 24 hour sample t (12) =0.74, p= 27 0.47, choice t (12) = 0.32, p= 0.75). 28 29 30 Experiment 2: mPFC Dysfunction and Object Recognition 31 32 1 minute and 3 minute discrimination scores showed no significant 33 difference between lesion and sham rats at 1 hour or 24 hour delays (1 hour 34 t(12) =1.51, p= 0.62, t (.02) = 0.98 or 24 hour t (12) =0.93, p= 0.37, t (12) =0.55, p= 35 0.60). 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 10 Page 11 of 37 Psychopharmacology

1 2 3 Fig. 6 mPFC dysfunction and object recognition. Means of first minute 4 discrimination scores. ±1SEM. 5 6 7 There were no significant differences in exploration time in sample or 8 choice phase at a 1 hour (sample: t =1.53, p= 0.15; choice: t = 0.78, p= 9 (12) (12) 10 0.45) or 24 hour delay (sample: t (12) = 0.52, p= 0.61; choice: t (12) =1.03, 11 p=0.32). Discrimination scores at the 24 hour delay were significantly greater 12 than chance for both sham and lesion (sham t (6) =4.51, p=0.004; lesion t (6) = 13 3.14, p=0.016). 14 15 16 Experiment 3: scPCP and Reversal Learning 17 18 scPCP rats were significantly faster to reach the reversal learning 19 criterion (80% accuracy over 2 consecutive sessions) than control rats both in 20 errors to criterion (t (14) = 1.78, p=0.048) and sessions to criterion (t (14) = 1.95, 21 p=0.035). 22 23 Performance across days of acquisition (figure 7) revealed a significant 24 main effect of drug treatment on overall accuracy (F (1,10) = 5.2, p= 0.046) and 25 number of correction trials (F (1,10) = 6.4, p= 0.03) but no drug x session 26 interactions (accuracy F (8.09, 80.9) = 0.58, p= 0.77, correction trials F (3.91, 39.2) = 27 0.95, p= 0.44). 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Fig. 7 Effects of scPCP on reversal learning acquisition, sessions and errors 52 to criterion. Means ±1 SEM. *= p<0.05. 53 54 55 Analysis was also conducted in terms of stage of reversal, through 56 early sessions (sessions 1-2), and mid-late sessions (considering sessions 3- 57 6 and 7-18 independently gave the same results as pooling). There was no 58 59 60 11 Psychopharmacology Page 12 of 37

1 2 3 significant difference in accuracy or number of correction trials in early trials 4 (accuracy F (1,13) = 0.068, p=0.799; correction trials F (1,14) = 2.11, p= 0.17), but 5 the effect of scPCP on accuracy was significant through mid-late sessions 6 (F (1,10) = 4.93, p=0.041). Analysis of the perseveration index, the ratio of 7 correction trials to initially incorrect responses, showed no significant 8 9 difference overall (F (1,10) = 1.17, p= 0.30) or in early sessions (F (1,10) = 4.62, p= 10 0.067) between sham and lesion. 11 A significant effect of drug treatment on log transformed response 12 latency was seen in early sessions (F = 5.71, p= 0.038) with vehicle rats 13 (1,10) initially responding significantly more rapidly. This initial difference was not 14 15 observed through later sessions F (1,10) = 4.23, p=0.67) and there was no 16 significant drug x session interaction F (5.89,70.74) =1.74, p=0.60). No significant 17 effect of drug was seen on log transformed magazine latency in early 18 sessions (F (1,10) = 0.20, p=0.664), late sessions (F (1,10) = 2.03, p=0.184), or 19 overall (F (1,10) = 1.79, p= 0.210). 20 21 22 Experiment 3: mPFC Dysfunction and Reversal Learning 23 24 Histological results: Analysis of histology confirmed lesions of the 25 medial prefrontal cortex in all rats, centred on the prelimbic cortex, with 26 damage to the infralimbic cortex and overlying anterior cingulate (figure 8. 27 Sham animals did not show damage to any of those regions, and one 28 lesioned animal was excluded due to more unilateral lesion than common. 29 30 31 Fig. 8 mPFC lesions in reversal 32 33 learning. Damage common to all 34 subjects in solid black, with 35 maximum extent of any damage 36 shown by the black line. Coronal 37 sections taken at 3.72mm, 38 2.76mm, and 1.08mm anterior to Bregma. Images 39 adapted from Paxinos and Watson (2007). Composite 40 photomicrograph presents a contrast between sham 41 (left) and lesion (right) for demonstration purposes. 42 43 44 45 46 47 48 49 50 51 Behavioural results: Rats were rebaselined for three sessions of the 52 original discrimination post-surgery, which all subjects completed at criterion 53 level (>80%). mPFC rats were significantly faster in number of sessions 54 required to reach the reversal learning criterion (t =2.52, p=0.013), and also 55 (12) 56 showed a near-significant reduction in the number of errors to criterion 57 (t (12) =1.71, p=0.056). 58 59 60 12 Page 13 of 37 Psychopharmacology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22

23 24 25 Fig. 9 Effects of mPFC lesions on reversal learning; errors and sessions to 26 criterion. Means ±1 SEM. *= p<0.05. 27 28 29 30 Analysis of accuracy over the course of reversal found no significant 31 effect of lesion during early trials (F (1,11) = 4.18, p= 0.086), mid trials (3-6) 32 F(1,11) = 4.12, p= 0.067), late trials (7-18) F (1,11) = 2.57, p= 0.14) or across mid- 33 late sessions (F (1,11) = 3.2, p= 0.10). There was no significant difference in 34 number of correction trials in early trials only (F(1,11) = 0.53, p= 0.48) or mid- 35 late trials only (F (1,11) = 2.22, p=0.16). 36 37 The perseveration index showed no significant effect of lesion 38 (F (1,10) =0.44, p= 0.52). There was no difference in reaction time (F (1,11) = 39 0.074, p= 0.79) with reaction time decreasing as the reversal progressed 40 (F (7.7,84.9) = 34.1, p< 0.001). There was no effect of lesion on magazine latency 41 (F (1,11) = 0.001, p= 0.98) with latency roughly remaining constant as the 42 reversal progressed in both groups (F (1.23,13.53) = 4.01, p=0.059). 43 44 45 Discussion 46 47 48 49 There was a double-dissociation between object-location paired- 50 associates learning and object recognition: mPFC dysfunction impaired 51 acquisition of the object-location task, but not spontaneous novel object 52 recognition, while scPCP impaired spontaneous novel object recognition, but 53 not object-location associative learning. Both scPCP and mPFC dysfunction 54 resulted in a similar facilitation during the late stages of reversal learning. 55 56 57 58 59 60 13 Psychopharmacology Page 14 of 37

1 2 3 4 scPCP mPFC Lesion 5 6 Object-Location Paired-Associates No effect Impairment 7 Learning: Acquisition 8 9 10 Object-Location Paired-Associates No effect No effect 11 Learning: Motivation/Latencies 12 13 14 Object Recognition: Discrimination Impairment No Effect 15 16 Object Recognition: Exploration No effect No effect 17 18 19 Reversal Learning Facilitation Facilitation 20 21 Table 1 Cognitive profiles of scPCP and mPFC lesions generated in the 22 present study. 23 24 25 26 Object-location learning (Experiment 1) 27 Whereas most tests of cognitive function following scPCP are 28 conducted immediately post-washout, observed impairments on object 29 recognition following PAL acquisition indicate the lack of effect of scPCP on 30 PAL was not due to a decrease in general efficacy of the manipulation across 31 32 time. 33 Previous studies have suggested that scPCP can impair object-location 34 learning. For example, 5mg/kg scPCP induced a deficit in rats’ ability to 35 simultaneously associate object-place-context (O-P-C) to identify novel 36 objects (Le Cozannet et al. 2010). That task, however, relies on one-trial 37 38 spontaneous object recognition (known to be impaired following scPCP, see 39 below). This lower-order failure could underlie the observed impairment in 40 other studies of object-location learning. 41 While a significant main effect of lesion was observed, the effect of the 42 43 lesion manifests as a selective impairment in the later stages of acquisition. 44 To investigate further how this task is solved, and thus perhaps illuminate the 45 impairment seen following mPFC dysfunction, we carried out an object- 46 location challenge (‘sPAL’; Talpos et al 2009). With only six unique trial types 47 it is conceivable that animals were responding to trial types as a whole picture 48 or scene (a ‘gestalt’) rather than individual objects situated in individual 49 locations, perhaps applying conditional rules by trial type. 50 51 Correct object-location associations should result in completion of 52 sPAL with little difficulty, whereas a strategy based on the combination of 53 objects presented during the trial would not work. Analysis showed no 54 impairment in either group switching to sPAL, suggesting that both sham and 55 lesion groups were not using a ‘gestalt’ trial-type strategy. The mPFC group 56 was still impaired relative to sham, and their performance on sPAL dropped 57 roughly the same amount (but not to chance). This pattern suggests that the 58 59 60 14 Page 15 of 37 Psychopharmacology

1 2 3 residual performance of mPFC lesioned group was due to some preserved 4 object-location learning, but that this learning was not a rapid as in controls. 5 6 In addition, analysis of variability in performance across trial types was 7 quite revealing: in sham animals, performance on the different trial-types was 8 highly variable during first sessions, then appeared to briefly converge, before 9 performance across all trial types increased more uniformly. This pattern may 10 reflect a transition from an ‘elemental’ strategy, or perhaps stimulus bias 11 (responding preferentially to particular stimuli), to an object-location 12 configural/associative solution. 13 14 Interestingly, the PFC-lesioned rats showed a very different pattern of 15 performance (figure 4). Like shams, rats in the PFC group appear to have 16 approached the task with an elemental strategy or stimulus bias (indicated by 17 the high variability across trial types during initial acquisition). The variance 18 then decreases as rats relinquish this strategy or bias. Thereafter, however, 19 20 performance differs markedly: whereas the sham animals’ performance 21 increases and the trial-type variance decreases, the mPFC rats’ performance 22 increases little and the variance remains the same. This pattern of results, 23 and the pattern obtained in the sPAL probe analysis, suggest that in the later 24 stages of PAL acquisition, rats master object-location associations, and that 25 mPFC dysfunction impairs the ability to progress through this stage of the 26 task. The idea that mPFC dysfunction impairs object-location binding is 27 consistent with previous reports (Barker et al. 2007; Barker and Warburton 28 2011). 29 30 31 32 Object Recognition (Experiment 2) 33 scPCP treatment produced a robust impairment on object recognition 34 in keeping with previous studies (Arnt et al. 2010; Grayson et al. 2014; 35 Grayson et al. 2007; Horiguchi et al. 2011; Snigdha et al. 2010; Snigdha et al. 36 37 2011). This impairment is not the result of general differences in the 38 exploration of objects, as evidenced by the roughly equal exploration times 39 during the sample phase. The effects of scPCP treatment were shown to be 40 surprisingly long-lasting, affecting SOR 40 days post washout. However, the 41 effects of scPCP was also time-limited, with no effect at even a 24 hour delay 42 after 105 days. A caveat is that this comparison is made within-subject, 43 without a between-subject control for order of testing. However it seems 44 unlikely that the substantial difference between effects of scPCP at 40 and 45 150 days was due to this factor alone. 46 47 The neurobiological changes underlying this impairment remain 48 unclear, and myriad effects are observed following scPCP treatment (Egerton 49 et al. 2005; Neill et al. 2010; Neill et al. 2014). In particular, studies have 50 shown that scPCP treatment downregulates striatal D1 and upregulates 51 cortical 5-HT1A receptor binding in rats (Choi et al. 2009). Investigation of the 52 53 amelioration of scPCP-induced impairments in novel object recognition 54 implicate these mechanisms, as the D1 receptor agonist asenapine 55 ameliorates a scPCP-induced impairment in object recognition, unless co- 56 treated with the D1 antagonist SCH-23390 (Snigdha et al. 2011). Co- 57 administration of atypical antipsychotics with PCP prevent development of 58 59 60 15 Psychopharmacology Page 16 of 37

1 2 3 novel object recognition deficits (Oyamada et al. 2015). Further studies with 4 D1 agonists (McLean et al. 2009) or atypical antipsychotics (Grayson et al. 5 2007) have also demonstrated amelioration of impairments in object 6 recognition. Additionally, the 5-HT6 receptor antagonist Lu AE58054 has 7 been shown to rescue scPCP induced impairment in object recognition (Arnt 8 9 et al. 2010). 10 A lack of effect of mPFC lesions on object recognition is in keeping the 11 literature published on various forms of spontaneous object recognition 12 (Barker et al. 2007; Ennaceur et al. 1997; Hannesson et al. 2004). 13 14 15 16 A double-dissocation between object-location learning and object recognition 17 The effects of scPCP dissociated across tasks: while scPCP did not 18 impair object-location paired-associates learning, it did impair spontaneous 19 object recognition. This may appear paradoxical, as object recognition could 20 21 seem to be required for paired-associates learning (for example, lesions to the 22 perirhinal cortex impaired both object recognition and an object-in-place task 23 (Barker and Warburton 2011), although perirhinal cortex lesions have not 24 been tested with the touchscreen PAL task used here). However the repeated 25 exposure to trial types and gradual learning in touchscreen paired-associates 26 learning may be a fundamental difference between these tasks. Unlike object 27 recognition, which is regarded (rightly or wrongly; Bussey et al. 2013) as one- 28 trial learning, PAL was learned across sessions over many days. Perirhinal 29 cortex dysfunction can impair discrimination learning across sessions in the 30 31 touchscreen (Winters et al. 2010), although the photographic stimuli used in 32 that study were more complex than those used here, and it has been shown 33 that perirhinal lesions can impair complex, but not simple discriminations 34 (Bussey et al. 2003). Another possibility is that scPCP affected structures 35 other than perirhinal cortex that are important for object recognition, for 36 example mediodorsal thalamus (Aggleton and Mishkin 1983; Mumby et al. 37 1993). 38 39 The effects of mPFC cortex dysfunction were also observed to 40 dissociate across tasks, but in the opposite direction: lesions impaired PAL, 41 but not object recognition. This likely reflects the higher order nature of the 42 PAL task, which requires integrating information about object and location. 43 This interpretation is consistent with the findings of Barker et al (2007), who 44 report impairment following PFC dysfunction on an object-location task (but 45 not on simpler object or location recognition tasks). 46 47 Thus, Experiments 1 and 2 demonstrate a double dissociation between 48 scPCP and PFC dysfunction on PAL and object recognition tasks. The object 49 recognition data show that scPCP affects structures outside the PFC that are 50 important for object recognition. More surprisingly, perhaps, the PAL 51 experiments show that if scPCP induces mPFC dysfunction, it is not of 52 53 sufficient magnitude to mimic the impairments found after excitotoxic mPFC 54 dysfunction on this task. 55 56 57 Reversal learning (Experiment 3) 58 59 60 16 Page 17 of 37 Psychopharmacology

1 2 3 Rodent models of schizophrenia have often demonstrated impaired 4 acquisition, impaired reversal, and/or increased perseveration in various 5 testing paradigms (Brigman et al. 2006 & 2008; Zhuo et al. 2007). In contrast, 6 in the present experiment, scPCP facilitated performance during the later 7 sessions of reversal. scPCP-treated animals did not display any differences 8 9 from vehicle-treated animals when performing below 50% correct, indicating 10 that increased perseveration is not resulting from the scPCP treatment. 11 While certain other models of schizophrenia have not displayed 12 increased perseveration (e.g., Brigman et al. 2008), they have shown 13 impaired learning of the new S+ v S- contingency. However, scPCP 14 15 treatment did not impair acquisition or reversal of a visual discrimination in 16 mice (Brigman et al. 2009) or rats (Fellini et al. 2014). Unlike these previous 17 studies, in the present study performance during late reversal was actually 18 facilitated, with PCP-treated rats reaching asymptote at a higher level of 19 performance than their vehicle treated controls. However, selective late-stage 20 facilitation has been previously observed in touchscreen reversal learning 21 conducted under similar conditions. In mice, both mild stress and 22 ventromedial prefrontal cortex lesions selectively facilitated late reversal 23 (Graybeal et al. 2011). The authors note that in other studies chronic stress 24 25 can promote habit-like responding in rats (coupled with atrophy of mPFC and 26 the ‘associative’ striatum and hypertrophy of the ‘sensorimotor’ striatum, as in 27 (Dias-Ferreira et al. 2009). Indeed, lesions of the dorsolateral striatum 28 impaired touchscreen reversal in mice; see Graybeal et al (2011) 29 supplementary materials. The effects of scPCP might be understood using a 30 similar explanation: dysfunction of the vmPFC removes inhibition on the 31 dorsolateral striatum and thus facilitates striatal-mediated habit formation late 32 in the reversal. However combined with the results from Experiment 1, such 33 34 an explanation suggests that mPFC dysfunction induced by scPCP was 35 sufficient to facilitate transition to habit, but insufficient to produce an 36 impairment in PAL. Alternatively, other mechanisms such as direct 37 disinhibition of the striatum might underlie the scPCP-induced improvement. 38 Like scPCP treatment, and consistent with previous studies, mPFC 39 40 dysfunction did not impair reversal learning (Boulougouris et al. 2007; 41 Chudasama et al. 2001) and indeed, according to sessions and errors to 42 criterion, significantly facilitated it. Visual inspection indicates that like scPCP, 43 facilitation appears in the late, but not the early, perseverative stage of 44 reversal. A difference between the pattern of data in the two experiments is 45 that the (sham-lesioned) control rats eventually reached the same 46 performance level as the mPFC-lesioned rats, whereas the vehicle-treated 47 rats never quite reached the same level as the scPCP-treated rats. There 48 was also more variability in the mPFC lesion group than there was in the 49 50 scPCP-treated group. Taking these considerations into account, and the 51 uncommon pattern of late reversal facilitation in both experiments, we 52 conclude that the effects of scPCP and mPFC dysfunction were similar. 53 However neither manipulation looks like what one would expect from a 54 ‘complete’ model of schizophrenia, in which reversal learning is, if anything, 55 impaired (Leeson et al. 2009; Murray et al. 2008; Waltz and Gold 2007). 56 57 That mPFC dysfunction did not produce impairment in touchscreen 58 reversal learning is consistent with previous literature (Bussey et al., 1997; 59 60 17 Psychopharmacology Page 18 of 37

1 2 3 although impairments were obtained with perceptually difficult stimuli). 4 Previously, dissociations have been observed between regions of the 5 prefrontal cortex: orbitofrontal cortex lesions impair reversal learning, while 6 mPFC lesions impair attentional set shifting (Bissonette et al. 2013; Dias et al. 7 1996). Indeed, ventromedial prefrontal cortex (vmPFC) lesions in mice led to 8 9 a late facilitation similar to that observed here (Graybeal et al. 2011). Coupled 10 with the finding that BDNF infusions into the vmPFC prevent reversal 11 facilitation by stress, and the effects of striatal lesions, Graybeal et al 12 proposed a mechanism whereby vmPFC dysfunction disinhibits habit 13 formation by striatal systems (see discussion of Experiment 3 above, 14 (Graybeal et al. 2011). A similar mechanism may underlie the late facilitation 15 seen in the scPCP-treated rats, and/or the trend toward facilitation shown by 16 the mPFC-lesioned rats, in the present study. 17 18 19 20 Evaluating scPCP as a Model for Schizophrenia 21 The use of phencyclidine to model schizophrenia has received some 22 criticism (Brigman et al. 2010; Domino and Luby 2012). As schizophrenia is 23 characterized by hypofunction of the prefrontal cortex (Andreasen et al. 1992; 24 Carter et al. 1998; Glahn et al. 2005; Hill et al. 2004), in the present 25 26 experiments we asked the specific question of whether the cognitive 27 impairments following scPCP resemble those following mPFC dysfunction. 28 These findings may be useful to consider when choosing a rodent model for 29 studies into schizophrenia, particularly if importance is placed on the model 30 faithfully reproducing the effects of mPFC dysfunction. Considering each of 31 the three tasks in turn: 32 33 Object-location learning: Although scPCP had no effect on PAL, 34 schizophrenic patients are impaired on the CANTAB version of PAL (Bartok et 35 al. 2005). Great caution must be taken, however, in using this contrast alone 36 to conclude that scPCP is a poor model, as we do not yet have a complete 37 idea to what extent rodent PAL and CANTAB-PAL depend on the same 38 psychological and neural mechanisms. Although some evidence suggests 39 they do, for example performance of rodent PAL and CANTAB-PAL both 40 41 involve the hippocampus (de Rover et al. 2011; Owen et al. 1995; Talpos et 42 al. 2008; Kim et al., this issue) and mutations in the schizophrenia-related 43 gene dlg2 can lead to impairments in both rodent PAL and CANTAB-PAL 44 (Nithianantharajah and Grant 2013). However, scPCP did not replicate the 45 effects of mPFC dysfunction on this task. 46 47 Object Recognition: scPCP clearly and reproducibly impaired OR. Thus 48 scPCP combined with SOR may be a good model for testing schizophrenia- 49 relevant cognitive enhancers. However the lack of effect of mPFC dysfunction 50 on OR suggests that the effect of scPCP was not due to dysfunction of the 51 mPFC. Furthermore, a caveat regarding the use of SOR is that it has been 52 argued that schizophrenics are not impaired on familiarity-based item 53 recognition as measured by SOR, but are impaired only when recollective 54 processes are required (Huron et al. 1995; van Erp et al. 2008). 55 56 Reversal learning: Schizophrenic patients are impaired on touchscreen 57 reversal learning, yet scPCP in the present experiment not only did not impair 58 59 60 18 Page 19 of 37 Psychopharmacology

1 2 3 reversal learning, it facilitated switching to the new contingency. Other 4 researchers have similarly observed a lack of impairment following scPCP on 5 touchscreen reversal learning (Brigman et al. 2009; Fellini et al. 2014) and in 6 attentional set-shifting paradigms (Egerton et al. 2005; McLean et al. 2012; 7 Rodefer et al. 2005 & 2008). Thus to the extent that researchers wish a 8 9 model to recapitulate the impairments in reversal learning seen in 10 schizophrenia, scPCP may not be the best choice. It should be noted, 11 however, that reversal impairments following scPCP have been observed 12 using other methodology, for example in operant chambers (Abdul-Monim et 13 al. 2007; Idris et al. 2010; McLean et al. 2009 & 2010 and in some studies 14 using the bowl-digging version of the task (Dawson et al. 2010; however cf. 15 Egerton et al. 2005 & 2008; McLean et al. 2008 & 2012; Pedersen et al. 2009; 16 Rodefer et al. 2005 & 2008). 17 18 To conclude, there was a double-dissociation between scPCP and 19 mPFC dysfunction on object-location paired associates learning and object 20 recognition, yet similar effects on reversal learning. The pattern of impairment 21 following scPCP raises questions around its efficacy as a model of cognitive 22 impairment in schizophrenia, particularly if importance is placed on the model 23 faithfully reproducing the effects of mPFC dysfunction. 24 25 26 27 28 Abdul-Monim, Z., Neill, J. C., & Reynolds, G. P. (2007). Sub-chronic 29 psychotomimetic phencyclidine induces deficits in reversal learning and 30 31 alterations in parvalbumin-immunoreactive expression in the rat. 32 Psychopharmacology, 21 (2), 198-205. 33 Aggleton, J. P., & Mishkin, M. (1983). Memory impairments following 34 restricted medial thalamic lesions in monkeys. Exp Brain Res, 52 (2), 35 199-209. 36 37 Amitai, N., Kuczenski, R., Behrens, M. M., & Markou, A. (2012). Repeated 38 phencyclidine administration alters glutamate release and decreases 39 GABA markers in the prefrontal cortex of rats. Neuropharmacology, 40 62 (3), 1422-1431. 41 42 Andreasen, N. C., Rezai, K., Alliger, R., Swayze, V. W., 2nd, Flaum, M., 43 Kirchner, P., et al. (1992). Hypofrontality in neuroleptic-naive patients 44 and in patients with chronic schizophrenia. Assessment with xenon 133 45 single-photon emission computed tomography and the Tower of 46 London. Arch Gen Psychiatry, 49 (12), 943-958. 47 48 Arnt, J., Bang-Andersen, B., Grayson, B., Bymaster, F. P., Cohen, M. P., 49 DeLapp, N. W., et al. (2010). Lu AE58054, a 5-HT6 antagonist, 50 reverses cognitive impairment induced by subchronic phencyclidine in 51 a novel object recognition test in rats. Int J Neuropsychopharmacology, 52 13 (8), 1021-1033. 53 54 Aubin, G., Stip, E., Gelinas, I., Rainville, C., & Chaparro, C. (2009). Daily 55 functioning and information-processing skills among persons with 56 schizophrenia. Psychiatric Services, 60 (6), 817-822. 57 58 59 60 19 Psychopharmacology Page 20 of 37

1 2 3 Barker, G. R., Bird, F., Alexander, V., & Warburton, E. C. (2007). Recognition 4 memory for objects, place, and temporal order: a disconnection 5 analysis of the role of the medial prefrontal cortex and perirhinal cortex. 6 J Neuroscience, 27 (11), 2948-2957. 7 8 Barker, G. R., & Warburton, E. C. (2011). When Is the Hippocampus Involved 9 in Recognition Memory? J Neuroscience, 31 (29), 10721-10731. 10 11 Barnes, S. A., Young, J. W., & Neill, J. C. (2012). Rats tested after a washout 12 period from sub-chronic PCP administration exhibited impaired 13 performance in the 5-Choice Continuous Performance Test (5C-CPT) 14 when the attentional load was increased. Neuropharmacology, 62 (3), 15 1432-1441. 16 17 Barnett, J. H., Sahakian, B. J., Werners, U., Hill, K. E., Brazil, R., Gallagher, 18 O., et al. (2005). Visuospatial learning and executive function are 19 independently impaired in first-episode psychosis. Psychol Med, 35 (7), 20 1031-1041. 21 22 Bartok, E., Berecz, R., Glaub, T., & Degrell, I. (2005). Cognitive functions in 23 prepsychotic patients. Prog Neuropsychopharmacol Biol Psychiatry, 24 29 (4), 621-625. 25 Beninger, R. J., Beuk, J., Banasikowski, T. J., van Adel, M., Boivin, G. A., & 26 27 Reynolds, J. N. (2010). Subchronic phencyclidine in rats: alterations in 28 locomotor activity, maze performance, and GABA(A) receptor binding. 29 Behav Pharmacology, 21 (1), 1-10. 30 Bissonette, G. B., Martins, G. J., Franz, T. M., Harper, E. S., Schoenbaum, 31 G., & Powell, E. M. (2008). Double dissociation of the effects of medial 32 33 and orbital prefrontal cortical lesions on attentional and affective shifts 34 in mice. J Neuroscience, 28 (44), 11124-11130. 35 Bissonette, G. B., Powell, E. M., & Roesch, M. R. (2013). Neural structures 36 underlying set-shifting: roles of medial prefrontal cortex and anterior 37 38 cingulate cortex. Behavioural Brain Research, 250 , 91-101. 39 Boulougouris, V., Dalley, J. W., & Robbins, T. W. (2007). Effects of 40 orbitofrontal, infralimbic and prelimbic cortical lesions on serial spatial 41 reversal learning in the rat. Behavioural Brain Research, 179 (2), 219- 42 228. 43 44 Brigman, J. L., Feyder, M., Saksida, L. M., Bussey, T. J., Mishina, M., & 45 Holmes, A. (2008). Impaired discrimination learning in mice lacking the 46 NMDA receptor NR2A subunit. Learn Mem, 15 (2), 50-54. 47 48 Brigman, J. L., Graybeal, C., & Holmes, A. (2010). Predictably irrational: 49 assaying cognitive inflexibility in mouse models of schizophrenia. Front 50 Neuroscience, 4 . 51 52 Brigman, J. L., Ihne, J., Saksida, L. M., Bussey, T. J., & Holmes, A. (2009). 53 Effects of Subchronic Phencyclidine (PCP) Treatment on Social 54 Behaviors, and Operant Discrimination and Reversal Learning in 55 C57BL/6J Mice. Front Behav Neuroscience, 3 , 2. 56 57 58 59 60 20 Page 21 of 37 Psychopharmacology

1 2 3 Brigman, J. L., Padukiewicz, K. E., Sutherland, M. L., & Rothblat, L. A. (2006). 4 Executive functions in the heterozygous reeler mouse model of 5 schizophrenia. Behav Neuroscience, 120 (4), 984-988. 6 7 Brown, V. J., & Bowman, E. M. (2002). Rodent models of prefrontal cortical 8 function. Trends Neuroscience, 25 (7), 340-343. 9 10 Burke, K. A., Takahashi, Y. K., Correll, J., Leon Brown, P., & Schoenbaum, G. 11 (2009). Orbitofrontal inactivation impairs reversal of Pavlovian learning 12 by interfering with 'disinhibition' of responding for previously 13 unrewarded cues. Eur J Neuroscience, 30 (10), 1941-1946. 14 Bussey, T. J., Barch, D. M., & Baxter, M. G. (2013). Testing long-term 15 16 memory in animal models of schizophrenia: suggestions from 17 CNTRICS. Neurosci Biobehav Rev, 37 (9 Pt B), 2141-2148. 18 Bussey, T. J., Muir, J. L., Everitt, B. J., & Robbins, T. W. (1997). Triple 19 dissociation of anterior cingulate, posterior cingulate, and medial frontal 20 21 cortices on visual discrimination tasks using a touchscreen testing 22 procedure for the rat. Behav Neuroscience, 111 (5), 920-936. 23 Bussey, T. J., Padain, T. L., Skillings, E., Winters, B. D., Morton, A. J., & 24 Saksida, L. M. (2008). The touchscreen cognitive testing method for 25 rodents: How to get the best out of your rat. Learn Mem, 15 (7), 516- 26 27 523. 28 Bussey, T. J., Saksida, L. M., & Murray, E. A. (2003). Impairments in visual 29 discrimination after perirhinal cortex lesions: testing 'declarative' vs. 30 'perceptual-mnemonic' views of perirhinal cortex function. Eur J 31 Neuroscience, 17 (3), 649-660. 32 33 Carter, C. S., Perlstein, W., Ganguli, R., Brar, J., Mintun, M., & Cohen, J. D. 34 (1998). Functional hypofrontality and working memory dysfunction in 35 schizophrenia. Am J Psychiatry, 155 (9), 1285-1287. 36 37 Castner, S. A., Goldman-Rakic, P. S., & Williams, G. V. (2004). Animal 38 models of working memory: insights for targeting cognitive dysfunction 39 in schizophrenia. Psychopharmacology), 174 (1), 111-125. 40 41 Choi, Y. K., Snigdha, S., Shahid, M., Neill, J. C., & Tarazi, F. I. (2009). 42 Subchronic effects of phencyclidine on dopamine and serotonin 43 receptors: implications for schizophrenia. J Mol Neuroscience, 38 (3), 44 227-235. 45 46 Chudasama, Y., Bussey, T. J., & Muir, J. L. (2001). Effects of selective 47 thalamic and prelimbic cortex lesions on two types of visual 48 discrimination and reversal learning. European Journal of 49 Neuroscience, 14 , 1009-1020. 50 51 Chudasama, Y., & Robbins, T. W. (2003). Dissociable contributions of the 52 orbitofrontal and infralimbic cortex to pavlovian autoshaping and 53 discrimination reversal learning: further evidence for the functional 54 heterogeneity of the rodent frontal cortex. J Neuroscience, 23 (25), 55 8771-8780. 56 57 Dawson, N., Thompson, R. J., McVie, A., Thomson, D. M., Morris, B. J., & 58 Pratt, J. A. (2010). Modafinil Reverses Phencyclidine-Induced Deficits 59 60 21 Psychopharmacology Page 22 of 37

1 2 3 in Cognitive Flexibility, Cerebral Metabolism, and Functional Brain 4 Connectivity. Schizophr Bulletin . 5 6 de Rover, M., Pironti, V. A., McCabe, J. A., Acosta-Cabronero, J., Arana, F. 7 S., Morein-Zamir, S., et al. (2011). Hippocampal dysfunction in patients 8 with mild cognitive impairment: a functional neuroimaging study of a 9 visuospatial paired associates learning task. Neuropsychologia, 49 (7), 10 2060-2070. 11 12 Dere, E., Huston, J. P., & De Souza Silva, M. A. (2007). The pharmacology, 13 neuroanatomy and neurogenetics of one-trial object recognition in 14 rodents. Neurosci Biobehav Rev, 31 (5), 673-704. 15 16 Dias, R., Robbins, T. W., & Roberts, A. C. (1996). Primate analogue of the 17 Wisconsin Card Sorting Test: effects of excitotoxic lesions of the 18 prefrontal cortex in the marmoset. Behav Neuroscience, 110 (5), 872- 19 886. 20 21 Dias-Ferreira, E., Sousa, J. C., Melo, I., Morgado, P., Mesquita, A. R., 22 Cerqueira, J. J., et al. (2009). Chronic stress causes frontostriatal 23 reorganization and affects decision-making. Science, 325 (5940), 621- 24 625. 25 Domino, E. F., & Luby, E. D. (2012). Phencyclidine/Schizophrenia: One View 26 27 Toward the Past, The Other to the Future. Schizophr Bulletin . 28 Donohoe, G., Spoletini, I., McGlade, N., Behan, C., Hayden, J., O'Donoghue, 29 T., et al. (2008). Are relational style and neuropsychological 30 performance predictors of social attributions in chronic schizophrenia? 31 Psychiatry Res, 161 (1), 19-27. 32 33 Egerton, A., Reid, L., McKerchar, C. E., Morris, B. J., & Pratt, J. A. (2005). 34 Impairment in perceptual attentional set-shifting following PCP 35 administration: a rodent model of set-shifting deficits in schizophrenia. 36 Psychopharmacology, 179 (1), 77-84. 37 38 Egerton, A., Reid, L., McGregor, S., Cochran, S., Morris, B. J., & Pratt, J. A. 39 (2008). Subchronic and chronic PCP treatment produces temporally 40 distinct deficits in attentional set shifting and prepulse inhibition in rats. 41 Psychopharmacology, 198(1), 37-49. 42 43 44 45 Ennaceur, A., Neave, N., & Aggleton, J. P. (1997). Spontaneous object 46 recognition and object location memory in rats: the effects of lesions in 47 the cingulate cortices, the medial prefrontal cortex, the cingulum bundle 48 and the fornix. Exp Brain Research, 113 (3), 509-519. 49 Fellini, L., Kumar, G., Gibbs, S., Steckler, T., & Talpos, J. (2014). Re- 50 51 evaluating the PCP challenge as a pre-clinical model of impaired 52 cognitive flexibility in schizophrenia. Eur Neuropsychopharmacology, 53 24 (11), 1836-1849. 54 Forwood, S. E., Winters, B. D., & Bussey, T. J. (2005). Hippocampal lesions 55 56 that abolish spatial maze performance spare object recognition 57 memory at delays of up to 48 hours. Hippocampus, 15 (3), 347-355. 58 59 60 22 Page 23 of 37 Psychopharmacology

1 2 3 Gilmour, G., Arguello, A., Bari, A., Brown, V. J., Carter, C., Floresco, S. B., et 4 al. (2013). Measuring the construct of executive control in 5 schizophrenia: defining and validating translational animal paradigms 6 for discovery research. Neurosci Biobehav Rev, 37 (9 Pt B), 2125-2140. 7 8 Glahn, D. C., Ragland, J. D., Abramoff, A., Barrett, J., Laird, A. R., Bearden, 9 C. E., et al. (2005). Beyond hypofrontality: a quantitative meta-analysis 10 of functional neuroimaging studies of working memory in 11 schizophrenia. Hum Brain Mapping, 25 (1), 60-69. 12 13 Graybeal, C., Feyder, M., Schulman, E., Saksida, L. M., Bussey, T. J., 14 Brigman, J. L., et al. (2011). Paradoxical reversal learning 15 enhancement by stress or prefrontal cortical damage: rescue with 16 BDNF. Nature Neuroscience, 14 , 1507-1509. 17 18 Grayson, B., Adamson, L., Harte, M., Leger, M., Marsh, S., Piercy, C., et al. 19 (2014). The involvement of distraction in memory deficits induced by 20 NMDAR antagonism: relevance to cognitive deficits in schizophrenia. 21 Behavioural Brain Research, 266 , 188-192. 22 23 Grayson, B., Idris, N. F., & Neill, J. C. (2007). Atypical antipsychotics 24 attenuate a sub-chronic PCP-induced cognitive deficit in the novel 25 object recognition task in the rat. Behavioural Brain Research, 184 (1), 26 31-38. 27 28 Hannesson, D. K., Howland, J. G., & Phillips, A. G. (2004). Interaction 29 between Perirhinal and Medial Prefrontal Cortex Is Required for 30 Temporal Order But Not Recognition Memory for Objects in Rats. J 31 Neuroscience, 24 (19), 4569-4606. 32 33 Haring, L., Mottus, R., Koch, K., Trei, M., & Maron, E. (2014). Factorial 34 validity, measurement equivalence and cognitive performance of the 35 Cambridge Neuropsychological Test Automated Battery (CANTAB) 36 between patients with first-episode psychosis and healthy volunteers. 37 Psychol Med icine 1-11. 38 39 Hill, K., Mann, L., Laws, K. R., Stephenson, C. M., Nimmo-Smith, I., & 40 McKenna, P. J. (2004). Hypofrontality in schizophrenia: a meta- 41 analysis of functional imaging studies. Acta Psychiatr Scand, 110 (4), 42 243-256. 43 44 Horiguchi, M., Huang, M., & Meltzer, H. Y. (2011). Interaction of mGlu2/3 45 agonism with clozapine and lurasidone to restore novel object 46 recognition in subchronic phencyclidine-treated rats. 47 Psychopharmacology), 217 (1), 13-24. 48 Horner, A. E., Heath, C. J., Hvoslef-Eide, M., Kent, B. A., Kim, C. H., Nilsson, 49 50 S. R., et al. (2013). The touchscreen operant platform for testing 51 learning and memory in rats and mice. Nature Protocols, 8 (10), 1961- 52 1984. 53 Howes, O., McCutcheon, R., & Stone, J. (2015). Glutamate and dopamine in 54 55 schizophrenia: an update for the 21st century. Psychopharmacology, 56 29 (2), 97-115. 57 58 59 60 23 Psychopharmacology Page 24 of 37

1 2 3 Huron, C., Danion, J., Giacomoni, F., & Grange, D. (1995). Impairment of 4 recognition memory with, but not without, conscious recollection in 5 schizophrenia. American Journal of Psychiatry, 152 (2), 1737-1742. 6 7 Idris, N., Neill, J., Grayson, B., Bang-Andersen, B., Witten, L. M., Brennum, L. 8 T., et al. (2010). Sertindole improves sub-chronic PCP-induced 9 reversal learning and episodic memory deficits in rodents: involvement 10 of 5-HT(6) and 5-HT (2A) receptor mechanisms. Psychopharmacology, 11 208 (1), 23-36. 12 13 Ingvar, D. H., & Franzen, G. (1974). Abnormalities of cerebral blood flow 14 distribution in patients with chronic schizophrenia. Acta Psychiatr 15 Scand, 50 (4), 425-462. 16 17 Itil, T., Keskiner, A., Kiremitci, N., & Holden, J. M. (1967). Effect of 18 phencyclidine in chronic schizophrenics. Can Psychiatr Assoc J, 12 (2), 19 209-212. 20 21 Javitt, D. C., & Zukin, S. R. (1991). Recent advances in the phencyclidine 22 model of schizophrenia. Am J Psychiatry, 148 (10), 1301-1308. 23 Javitt, D. C., Zukin, S. R., Heresco-Levy, U., & Umbricht, D. (2012). Has an 24 Angel Shown the Way? Etiological and Therapeutic Implications of the 25 PCP/NMDA Model of Schizophrenia. Schizophrenia Bulletin, 38 (5), 26 27 958-966. 28 Kim, C.H., Heath, C.J., Kent, B.A., Bussey, T.J., Saksida, L.M. (2015). The 29 role of the dorsal hippocampus in two versions of the touchscreen 30 automated paired associates learning (PAL) task for mice. 31 Psychopharmacology , in press. 32 33 Krystal, J. H., Karper, L. P., Seibyl, J. P., Freeman, G. K., Delaney, R., 34 Bremner, J. D., et al. (1994). Subanesthetic effects of the 35 noncompetitive NMDA receptor antagonist, ketamine, in humans: 36 Psychotomimetic, perceptual, cognitive, and neuroendro- crine 37 38 responses. Arch Gen Psychiatry, 51 , 199-214. 39 Lahti, A. C., Koffel, B., LaPorte, D., & Tamminga, C. A. (1995). Subanesthetic 40 doses of ketamine stimulate psychosis in schizophrenia. 41 Neuropsychopharmacology, 13 (1), 9-19. 42 43 Lahti, A. C., Weiler, M. A., Tamara Michaelidis, B. A., Parwani, A., & 44 Tamminga, C. A. (2001). Effects of ketamine in normal and 45 schizophrenic volunteers. Neuropsychopharmacology, 25 (4), 455-467. 46 47 Laurent, V., & Podhorna, J. (2004). Subchronic phencyclidine treatment 48 impairs performance of C57BL/6 mice in the attentional set-shifting 49 task. Behav Pharmacology, 15 (2), 141-148. 50 51 Le Cozannet, R., Fone, K. C., & Moran, P. M. (2010). Phencyclidine 52 withdrawal disrupts episodic-like memory in rats: reversal by donepezil 53 but not clozapine. Int J Neuropsychopharmacology, 13 (8), 1011-1020. 54 Leeson, V. C., Robbins, T. W., Matheson, E., Hutton, S. B., Ron, M. A., 55 56 Barnes, T. R., et al. (2009). Discrimination learning, reversal, and set- 57 shifting in first-episode schizophrenia: stability over six years and 58 59 60 24 Page 25 of 37 Psychopharmacology

1 2 3 specific associations with medication type and disorganization 4 syndrome. Biol Psychiatry, 66 (6), 586-593. 5 6 Lyon, L., Saksida, L. M., & Bussey, T. J. (2012). Spontaneous object 7 recognition and its relevance to schizophrenia: a review of findings 8 from pharmacological, genetic, lesion and developmental rodent 9 models. Psychopharmacology, 220 (4), 647-672. 10 11 Malhotra, A. K., Pinals, D. A., Adler, C. M., Elman, I., Clifton, A., Pickar, D., et 12 al. (1997). Ketamine-induced exacerbation of psychotic symptoms and 13 cognitive impairment in neuroleptic-free schizophrenics. 14 Neuropsychopharmacology, 17 (3), 141-150. 15 16 Malhotra, A. K., Pinals, D. A., Weingartner, H., Sirocco, K., Missar, C. D., 17 Pickar, D., et al. (1996). NMDA receptor function and human cognition: 18 the effects of ketamine in healthy volunteers. 19 Neuropsychopharmacology, 14 (5), 301-307. 20 21 McAllister, K.A.L., Saksida, L.M., Busset, T.J. (2013). Dissociation between 22 memory retention across a delay and pattern separation following 23 medial prefrontal cortex lesions in the touchscreen TUNL task. 24 Neurobiology Learn Mem, 101 , 120-126. 25 McAlonan, K., & Brown, V. J. (2003). Orbital prefrontal cortex mediates 26 27 reversal learning and not attentional set shifting in the rat. Behavioural 28 Brain Research, 146 (1-2), 97-103. 29 McKibben, C. E., Jenkins, T. A., Adams, H. N., Harte, M. K., & Reynolds, G. 30 P. (2010). Effect of pretreatment with risperidone on phencyclidine- 31 induced disruptions in object recognition memory and prefrontal cortex 32 33 parvalbumin immunoreactivity in the rat. Behavioural Brain Research, 34 208 (1), 132-136. 35 McKirdy, J., Sussmann, J. E., Hall, J., Lawrie, S. M., Johnstone, E. C., & 36 McIntosh, A. M. (2009). Set shifting and reversal learning in patients 37 38 with bipolar disorder or schizophrenia. Psychol Medicine, 39 (8), 1289- 39 1293. 40 McLean, S. L., Beck, J. P., Woolley, M. L., Neill, J. C., (2008). A preliminary 41 investigation into the effects of antipsychotics on sub-chronic 42 phencyclidine-induced deficits in attentional set-shifting in female rats. 43 44 Behav Brain Research, 189(1 ), 152-158. 45 McLean, S. L., Idris, N. F., Grayson, B., Gendle, D. F., Mackie, C., Lesage, A. 46 S., et al. (2012). PNU-120596, a positive allosteric modulator of alpha7 47 nicotinic acetylcholine receptors, reverses a sub-chronic phencyclidine- 48 induced cognitive deficit in the attentional set-shifting task in female 49 50 rats. Psychopharmacology, 26 (9), 1265-1270. 51 McLean, S. L., Idris, N. F., Woolley, M. L., & Neill, J. C. (2009). D1-like 52 receptor activation improves PCP-induced cognitive deficits in animal 53 models: Implications for mechanisms of improved cognitive function in 54 55 schizophrenia. European Neuropsychopharmacology, 19 (6), 440-450. 56 McLean, S. L., Neill, J. C., Idris, N. F., Marston, H. M., Wong, E. H., & Shahid, 57 M. (2010). Effects of asenapine, olanzapine, and risperidone on 58 59 60 25 Psychopharmacology Page 26 of 37

1 2 3 psychotomimetic-induced reversal-learning deficits in the rat. 4 Behavioural Brain Research, 214 (2), 240-247. 5 6 Meltzer, H. Y., Rajagopal, L., Huang, M., Oyamada, Y., Kwon, S., & 7 Horiguchi, M. (2013). Translating the N-methyl-D-aspartate receptor 8 antagonist model of schizophrenia to treatments for cognitive 9 impairment in schizophrenia. Int J Neuropsychopharmacology, 16 (10), 10 2181-2194. 11 12 Minzenberg, M. J., & Carter, C. S. (2012). Developing treatments for impaired 13 cognition in schizophrenia. Trends Cogn Sci, 16 (1), 35-42. 14 Mumby, D. G., Pinel, J. P., & Dastur, F. N. (1993). Mediodorsal thalamic 15 16 lesions and object recognition in rats. Psychobiology, 21 (1), 27-36. 17 Murray, G. K., Cheng, F., Clark, L., Barnett, J. H., Blackwell, A. D., Fletcher, 18 P. C., et al. (2008). Reinforcement and reversal learning in first-episode 19 psychosis. Schizophr Bulletin, 34 (5), 848-855. 20 21 Neill, J. C., Barnes, S., Cook, S., Grayson, B., Idris, N. F., McLean, S. L., et 22 al. (2010). Animal models of cognitive dysfunction and negative 23 symptoms of schizophrenia: focus on NMDA receptor antagonism. 24 Pharmacol Ther, 128 (3), 419-432. 25 26 Neill, J. C., Harte, M. K., Haddad, P. M., Lydall, E. S., & Dwyer, D. M. (2014). 27 Acute and chronic effects of NMDA receptor antagonists in rodents, 28 relevance to negative symptoms of schizophrenia: a translational link to 29 humans. Eur Neuropsychopharmacology, 24 (5), 822-835. 30 31 Nestler, E. J., & Hyman, S. E. (2010). Animal models of neuropsychiatric 32 disorders. Nat Neurosci, 13 (10), 1161-1169. 33 34 Nithianantharajah, J., & Grant, S. G. (2013). Cognitive components in mice 35 and humans: combining genetics and touchscreens for medical 36 translation. Neurobiol Learn Mem, 105 , 13-19. 37 Owen, A. M., Sahakian, B. J., Semple, J., Polkey, C. E., & Robbins, T. W. 38 (1995). Visuo-spatial short-term recognition memory and learning after 39 40 temporal lobe excisions, frontal lobe excisions or amygdalo- 41 hippocampectomy in man. Neuropsychologia, 33 (1), 1-24. 42 Oyamada, Y., Horiguchi, M., Rajagopal, L., Miyauchi, M., & Meltzer, H. Y. 43 (2015). Combined serotonin (5-HT)1A agonism, 5-HT2A and dopamine 44 45 D2 receptor antagonism reproduces atypical antipsychotic drug effects 46 on phencyclidine-impaired novel object recognition in rats. Behavioural 47 Brain Research, 285 , 165-175. 48 Paxinos, G., & Watson, C. (2007). The Rat Brain in Stereotaxic Coordinates . 49 London, UK: Academic Press. 50 51 Pedersen, C.S., Goetghebeur, P., Dias, R. (2009). Chronic infusion of PCP 52 via osmotic mini-pumps: A new rodent model of cognitive deficit in 53 schizophrenia characterized by impaired attentional set-shifting (ID/ED) 54 performance. Journal of Neuroscience Methods, 185(1) , 66-69. 55 56 Poels, E. M., Kegeles, L. S., Kantrowitz, J. T., Slifstein, M., Javitt, D. C., 57 Lieberman, J. A., et al. (2014). Imaging glutamate in schizophrenia: 58 59 60 26 Page 27 of 37 Psychopharmacology

1 2 3 review of findings and implications for drug discovery. Mol Psychiatry, 4 19 (1), 20-29. 5 6 Ragozzino, M. E. (2007). The contribution of the medial prefrontal cortex, 7 orbitofrontal cortex, and dorsomedial striatum to behavioral flexibility. 8 Ann N Y Acad Sci, 1121 , 355-375. 9 10 Ragozzino, M. E., Jih, J., & Tzavos, A. (2002). Involvement of the 11 dorsomedial striatum in behavioral flexibility: role of muscarinic 12 cholinergic receptors. Brain Research, 953(1-2), 205-214. 13 Rodefer, J. S., Murphy, E. R., & Baxter, M. G. (2005). PDE10A inhibition 14 reverses subchronic PCP-induced deficits in attentional set-shifting in 15 16 rats. Eur J Neuroscience, 21 (4), 1070-1076. 17 Rodefer, J. S., Nguyen, T. N., Karlsson, J. J., & Arnt, J. (2008). Reversal of 18 subchronic PCP-induced deficits in attentional set shifting in rats by 19 sertindole and a 5-HT6 receptor antagonist: comparison among 20 21 antipsychotics. Neuropsychopharmacology, 33 (11), 2657-2666. 22 Sahakian, B. J., Morris, R. G., Evenden, J. L., Heald, A., Levy, R., Philpot, M., 23 et al. (1988). A Comparative-Study of Visuospatial Memory and 24 Learning in Alzheimer-Type Dementia and Parkinsons-Disease. Brain, 25 111 , 695-718. 26 27 Schaefer, J., Giangrande, E., Weinberger, D. R., & Dickinson, D. (2013). The 28 global cognitive impairment in schizophrenia: consistent over decades 29 and around the world. Schizophr Res, 150 (1), 42-50. 30 31 Schlagenhauf, F., Huys, Q. J., Deserno, L., Rapp, M. A., Beck, A., Heinze, H. 32 J., et al. (2014). Striatal dysfunction during reversal learning in 33 unmedicated schizophrenia patients. Neuroimage, 89 , 171-180. 34 35 Singh, S. P., & Singh, V. (2011). Meta-analysis of the efficacy of adjunctive 36 NMDA receptor modulators in chronic schizophrenia. CNS Drugs, 37 25 (10), 859-885. 38 Snigdha, S., Horiguchi, M., Huang, M., Li, Z., Shahid, M., Neill, J. C., et al. 39 40 (2010). Attenuation of phencyclidine-induced object recognition deficits 41 by the combination of atypical antipsychotic drugs and pimavanserin 42 (ACP 103), a 5-hydroxytryptamine(2A) receptor inverse agonist. J 43 Pharmacol Exp Ther, 332 (2), 622-631. 44 45 Snigdha, S., Idris, N., Grayson, B., Shahid, M., & Neill, J. C. (2011). 46 Asenapine improves phencyclidine-induced object recognition deficits 47 in the rat: evidence for engagement of a dopamine D1 receptor 48 mechanism. Psychopharmacology, 214 (4), 843-853. 49 Snyder, S. H. (1980). Phencyclidine. Nature, 285 (5764), 355-356. 50 51 Talpos, J. C., Dias, R., Bussey, T. J., & Saksida, L. M. (2008). Hippocampal 52 lesions in rats impair learning and memory for locations on a touch- 53 sensitive computer screen: the "ASAT" task. Behavioural Brain 54 Research, 192 (2), 216-225. 55 56 Talpos, J. C., Winters, B. D., Dias, R., Saksida, L. M., & Bussey, T. J. (2009). 57 A novel touchscreen-automated paired-associate learning (PAL) task 58 59 60 27 Psychopharmacology Page 28 of 37

1 2 3 sensitive to pharmacological manipulation of the hippocampus: a 4 translational rodent model of cognitive impairments in 5 neurodegenerative disease. Psychopharmacology, 205 (1), 157-168. 6 7 Tamminga, C. A. (1998). Schizophrenia and glutamatergic transmission. Crit 8 Rev Neurobiol, 12(1-2), 21-36. 9 10 van Erp, T. G., Lesh, T. A., Knowlton, B. J., Bearden, C. E., Hardt, M., 11 Karlsgodt, K. H., et al. (2008). Remember and know judgments during 12 recognition in chronic schizophrenia. Schizophr Res, 100 (1-3), 181- 13 190. 14 Vingerhoets, W. A., Bloemen, O. J., Bakker, G., & van Amelsvoort, T. A. 15 16 (2013). Pharmacological Interventions for the MATRICS Cognitive 17 Domains in Schizophrenia: What's the Evidence? Front Psychiatry, 4 , 18 157. 19 Waltz, J. A., & Gold, J. M. (2007). Probabilistic reversal learning impairments 20 21 in schizophrenia: further evidence of orbitofrontal dysfunction. 22 Schizophr Res, 93 (1-3), 296-303. 23 Winters, B. D., Bartko, S. J., Saksida, L. M., & Bussey, T. J. (2010). Muscimol, 24 AP5, or scopolamine infused into perirhinal cortex impairs two-choice 25 visual discrimination learning in rats. Neurobiology Learn Memory, 26 27 93 (2), 221-228. 28 Wood, S. J., Proffitt, T., Mahony, K., Smith, D. J., Buchanan, J. A., Brewer, 29 W., et al. (2002). Visuospatial memory and learning in first-episode 30 schizophreniform psychosis and established schizophrenia: a 31 functional correlate of hippocampal pathology? Psychol Med, 32 (3), 32 33 429-438. 34 Young, A. M., Stubbendorff, C., Valencia, M., & Gerdjikov, T. V. (2015). 35 Disruption of medial prefrontal synchrony in the subchronic 36 phencyclidine model of schizophrenia in rats. Neuroscience, 287 , 157- 37 38 163. 39 Zhuo, J. M., Prescott, S. L., Murray, M. E., Zhang, H. Y., Baxter, M. G., & 40 Nicolle, M. M. (2007). Early discrimination reversal learning impairment 41 and preserved spatial learning in a longitudinal study of Tg2576 APPsw 42 mice. Neurobiol Aging, 28 (8), 1248-1257. 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 28 Page 29 of 37 Psychopharmacology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 Fig. 1 Trial-types in touchscreen object-location paired-associates learning (dPAL left and sPAL right). S+ 23 rewarded; S- results in ‘time out’. Adapted from Talpos et al (2009). 24 83x36mm (300 x 300 DPI) 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Psychopharmacology Page 30 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Fig. 3 mPFC lesions in PAL. Damage common to all subjects in solid black, with maximum extent of any damage shown by the black line. Coronal sections at 3.72mm, 2.76mm, and 1.08mm anterior to 48 Bregma. Images adapted from Paxinos and Watson (2007). Composite photomicrograph presents contrast 49 between sham (left) and lesion (right) for demonstration purposes. 50 41x73mm (300 x 300 DPI) 51 52 53 54 55 56 57 58 59 60 Psychopharmacology Page 32 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 Fig. 3 mPFC lesions in PAL. Damage common to all subjects in solid black, with maximum extent of any 30 damage shown by the black line. Coronal sections at 3.72mm, 2.76mm, and 1.08mm anterior to 31 Bregma. Images adapted from Paxinos and Watson (2007). Composite photomicrograph presents contrast 32 between sham (left) and lesion (right) for demonstration purposes. 33 41x27mm (300 x 300 DPI) 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 33 of 37 Psychopharmacology

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 Fig. 8 mPFC lesions in reversal learning. Damage common to all subjects in solid black, with maximum extent of any damage shown by the black line. Coronal sections are taken at 3.72mm, 2.76mm, and 48 1.08mm anterior to Bregma. Images adapted from Paxinos and Watson (2007). Composite 49 photomicrograph presents a contrast between sham (left) and lesion (right) for demonstration purposes. 50 41x79mm (300 x 300 DPI) 51 52 53 54 55 56 57 58 59 60 Psychopharmacology Page 34 of 37

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 Fig. 8 mPFC lesions in reversal learning. Damage common to all subjects in solid black, with maximum 33 extent of any damage shown by the black line. Coronal sections are taken at 3.72mm, 2.76mm, and 1.08mm anterior to Bregma. Images adapted from Paxinos and Watson (2007). Composite 34 photomicrograph presents a contrast between sham (left) and lesion (right) for demonstration purposes. 35 41x31mm (300 x 300 DPI) 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 35 of 37 Psychopharmacology

1 2 3 4 5 6 7 scPCP and Reversal Sessions to Criterion 8 9 20 10 100 15 11 90 12 10 13 80 Criterion

Sessions Sessions to 5 14 70 15 16 60 0 PCP 17 50 18 Errors to Criterion 19 40 Percent Correct 20 30 21 500 22 20 400 23 PCP 300 10 24 Vehicle 200 Criterion 25 0 Errors to 26 1 3 5 7 9 11131517 100 27 Session 0 28 PCP 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Psychopharmacology Page 36 of 37

1 2 3 4 5 6 Sessions7 to Criterion 8 9 10 11 12 13 14 15 16 Vehicle 17 Errors18 to Criterion 19 20 21 22 23 24 25 26 27 28 Vehicle 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 Page 37 of 37 Psychopharmacology

1 2 3 4 5 6 7 8 9 mPFC Lesions and Reversal mPFC and Reversal 10 Sessions to Criterion 11 100 12 90 20 13 15 14 80 15 10 70

16 Criterion 5 Sessions Sessions to 17 60 0 18 mPFC Sham 19 50 20 40 mPFC and Reversal 21 Percent Correct Errors to Criterion 22 30 23 20 750 24 Lesion 25 10 500 26 Sham 0 250

27 Criterion Errors to 28 1 3 5 7 9 11131517 29 Session 0 mPFC Sham 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60